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 IV.

PUBLISHED AT THE REQUEST OF THE

COLONIAL GOVERNMENT

BY

THE ROYAL SOCIETY.

LONDON :
1905.

/ o 7 <j

c,

[ iii ]

CONTElNTS of PART IV.

PEARL OYSTER REPORT.

Page

Preface

The Great Pearl Fishery of 1905

SUPPLEMENTARY REPORTS.

XXIII —On the Isopoda. By Revel. T. R. R. Stebbing, F.R.S. (Twelve Plates) . 1

XXIV.— On the Macrura. By Joseph Pearson, B.Sc. (Two Plates) 65

XXV. — On the Antipatharia. By J. Arthur Th >ms< in, M.A., and J. J. Simpson,

M.A. (One Plate) 93

XXYI. — On the Polyzoa. By Laura R. Thornely (One Plate) .... . 107

XXVII.— On the Medusa. By E. T. Browne, B. A. (Four Plates) 131

XXVIII. — On the Alcyonaria — Supplementary. By J. Arthur Thomson, M.A.

(One Plate) 167

XXIX.— On the Solitary Corals. By G. C. Bourne, D.Sc. (Four Plates) ... 187

XXX.— On the Poi.ych.eta. By Arthur Willey, D.Sc., F.R.S. (Eight Plates) . 243

With a Note on Polydora armata. By Arnold T. Watson, F.L S. . 325

a 2

i: v ]

PREFACE.

When the last Part of this Keport was issued, early in the present year, I still
hoped that it would be possible to compress the remaining Supplementary Keports,
along with the conclusion of the Pearl Oyster Keport proper, within the limits of
Part IV., and an intimation to that effect was made in the Preface. During the
summer, however, when the Authors of the Supplementary Reports began to send in
their manuscripts, it became plain that I had been too sanguine, and that a fifth and
concluding Part would be necessary to finish the series. Under these circumstances
I have decided to issue all the Supplementary Reports that are ready in the present
Part — making it up to a volume of about the usual size of 300 pages and 30 plates —
and to publish the few remaining Reports along with the several concluding sections
of the Pearl-oyster work in Part V., early in 1906.

The Reports (all well advanced) which I still expect are : — On the Brachyura, by
Mr. Douglas Laurie; on the Schizopoda and Stomatopoda, by Mr. Walter
Tattersall ; on the Anomura, and on the Actiniaria ; on the Marine Insects, by
Professor G. Carpenter ; on the Tunicata, by myself ; a list of the Molluscan Shells,
by Mr. R. Standen and Mr. A. Leicester ; and a list of the Foraminifera, by
Mr. W. J. Dakin. There will be a further, and final, instalment on the Parasites of
the Pearl Oyster, by Mr. Shipley and Mr. Hornell, and I propose to add a general
discussion of the faunistic results. These sections, with the remainder of the Pearl-
oyster work, and the Summary of Results and Recommendations as to the Conservation
of the Banks, should then complete the Report.

The valuable memoirs in the present volume will speak for themselves, but I desire
to express here my very cordial thanks to my friends the various contributors for
their kindness in undertaking the work, and for their skill in carrying it out. The
only case that calls for any further remark is the last in the volume. In order to
prevent delay, I arranged with Dr. Willey that the proofs of the Report on the
Polychseta should not be sent out to Ceylon. As the "copy" was supplied to me
in the form of a corrected print marked " for press," and as Mr. Arnold Watson
and Mr. Cyril Crossland — both familiar with the group — have kindly helped me to
read the proofs, I hope that the chance of errors having escaped notice has been
minimised. I have added at the last moment an interesting note by Mr. Arnold
Watson on the Polychaste worm which is commensal with the sponge Aulospongus

[ vi ]

tiibulatus — a familiar object on the pearl banks. This worm, when first found, was
supposed to be an Oligochaete, but now proves, as Mr. Watson has shown, to be a
Polychaete allied to the Leucodore that burrows in the pearl oyster shells.

During the present year, since the issue of Part III., important changes have taken
place in Ceylon, both on the pearl banks and in the arrangements proposed for future
fisheries. The recommendations made, on biological grounds, in the article on ' The
Present Condition of the Pearl Banks' (Part III., p. 44) were not fully adopted by
the Government of Ceylon.

The South Cheval Paar, which it was hoped would be left for another year, to
supply a fishery in 1906, was stripped along with the Modragam Paars and some
other sections of the Cheval — the result being that the great fishery in the Spring of
1905 now holds the record both for the total number of oysters fished and also for
the amount of revenue brought in to the Government. The prices obtained were very
high, and it is, of course, under such circumstances, a great temptation to a Government
to fish and sell as many oysters as it is possible to obtain in the limited time permitted
by the weather. It must always be a difficult matter to decide whether oysters
present in abundance but admittedly immature should be secured at once or left to
have an additional year of growth and probable pearl-formation. We can only hope
that on the present occasion the decision has been a wise one, and that the clearing of
the South Cheval Paar has not imperilled the success of next year's fishery.

Another factor which may have an important bearing upon the future history of
the Ceylon Pearl Fisheries is the proposal that the pearl banks should be leased for a
period of years to a Syndicate, which will be bound by the terms of the lease to
expend a considerable sum annually in the cultivation and exploitation of the fishing
grounds. It seems probable at the time of issuing this volume that the next Ceylon
Pearl Fishery will be held under the auspices of the Syndicate.

W. A. HEKDMAN.
The University, Liverpool.
November, 1905.

[ vii ]

REPORT ON THE PEARL OYSTER FISHERIES
OF THE GULF OF MANAAR -PART 1Y.

THE GREAT PEARL FISHERY OF 1905.

(Drawn up from the Government Official Keports and Mr. Hornell's Letters,

with Additions and Remarks.)

The Pearl Fishery which was held at Ceylon in the spring of 1905 may well be called
"the great fishery." The temporary "fishery-town" erected at Marichchukaddi was
larger than it is known to have ever been before, the inhabitants congregated there
were at least 10,000 more than in the previous "record" fishery of 1904, the numbers
of divers (4991) and their attendant manducks (4894) and the fleet of boats (318)
were much greater, and the totals of oysters fished and of rupees obtained for the
Government were far beyond all previous records. There were forty-seven fishing
days as against thirty-three in 1904. The number of oysters collected (upwards of
eighty-one and a half millions) is nearly double that obtained at any previous fishery,
and the revenue derived from the Government share alone was nearly two and a half
million rupees, which beats all other known fisheries with nearly 15 lacs of rupees
to spare.

This is the last of a series of three highly successful fisheries in consecutive years,
and it does not seem likely to be rivalled by any prospective fishery of the oysters
now in sight upon the grounds. The fishery of 1903 yielded 41,180,137 oysters and
the Government share of the revenue was about £55,303; in 1904 the oysters totalled
41,039,085 and the Government revenue ,£71,050; in 1905 the number of oysters was
81,580,716 and the revenue over £167,381. From these fisheries, taken together, the
Government of Ceylon has derived fully £293,735, without taking into account the
revenue derived from postal returns (which at these three fisheries were considerable),
and other receipts indirectly connected with, but consequent upon, the industry.

These figures look magnificent, and they show what a valuable possession the

(2.)

1904

(3.)

1814

(4.)

1891

(5.)

1808

(6.)

1903

(7.)

1888

41,039,085

#

44,311,441

#

41,180,137

22,052,769

-[viii] CEYLON PEARL OYSTER REPORT.

Ceylon Government has in the pearl banks, but it must be remembered that this
series of three very profitable fisheries followed on an interval of eleven years during
which there were no returns, and that the average yield for the last three years is far
above that of the last hundred or so — the period for which we have any accurate
record.

It may be of interest to quote here the results of the chief fisheries of that
century : —

(1.) 1905 yielded 81,580,716 oysters. Kevenue = 2,510,727 Rs.

= 1,065,751
= 1,051,876
= 963,748
= 842,577
= 829,548
= 804,247

* Particulars as to number of oysters fished not available previous to 1835.

After that the revenue rapidly drops to seven, five, four, three, and two hundred-
thousand rupees, and there are a number of fisheries on the list with between one and
two hundred-thousand. As an example of a very poor fishery, we may take the one
held in 1884 off Chilaw, yielding only 636,000 oysters and a revenue of 17,153 rupees.
Many single days in the recent fishery far exceed the entire proceeds of certain
previous fisheries.

The Report of Mr. G. P. Lewis, Government Agent, writing as Superintendent of
the fishery, puts in this way " the extraordinary achievements of this fishery by
comparison with the chief fisheries of former years, the fisheries of 1808 and 1903
were beaten in ten days, that of 1891 in eleven, and those of 1904 and 1814, which
had hitherto occupied the first and second places respectively, in twelve." The fishery
lasted for 35 days after that. Mr. Hornell,, in his Report, puts it another way, and
says that the total number fished " all but equals the combined totals of the two
great fisheries of 1903 and 1904, and is not far from double the highest number
taken in any other fishery concerning which we have authentic information, namely,
44,311,441 in 1891 " ('Ceylon Sessional Papers,' 1905, p. 36).

Enough has been said to show the exceptional nature of the recent fishery, but it
is quite a pertinent question whether this phenomenal success has not been attained,
to some extent at least, at the expense of the next few years. It will be remembered
that in the section of this Report entitled ' Present Condition of the Pearl Banks '
(Part ILL, p. 37), written a year ago as the result of Mr. Hornell's inspection in
November, 1904, certain reasons were given for fishing in 1905 the two Modragams
and the sections of the Cheval defined as south-west, mid-east, and south-central
(see fig. 1), leaving for 1906 the sixty-odd million oysters upon the South and South-

THE GREAT PEARL FISHERY OF 1905.

L«l

Fig. 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 months old, the dotted areas are those with oysters about 3 years old, the
solid black indicates those beds of oysters over 5 years old.

east Cheval.
follows : —

The oysters on the banks in question were then estimated to be as

Paar.

South Modragam

North Modragam

South-west Cheval

Mid-east and South-central Cheval

South Cheval

South-east Cheval

Number of oysters.

Age in March, 1 905.

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

Over 3 years.

,, 3 „

„ 5 „

3 to 6 „

Over 3 „
„ 3 „

Of these, the oysters on the two Modragam paars were thought to be unhealthy
and liable to die off; those on the South-west Cheval were very old and clearly had
to be secured at once or not at all ; and those on the Mid-east and South-central
Cheval were apparently being smothered by young oysters. Consequently the
recommendation was that these forty-three million oysters should be fished first,

b

[x] CEYLON PEARL OYSTER REPORT.

leaving the rest which seemed healthy, vigorous, and safe for a further year of growth
and pearl -production.

On the other hand, the November valuation of the pearls from the samples collected
gave 24.65 Rs. per thousand to the South Cheval, as against 17.86 lis. for the South
Modragam which stood next. That fact was sure to cause pressure to be brought by
the pearl merchants upon the Government in favour of fishing the South Cheval ;
and it is, of course, impossible to avoid sympathising with the view that 40 million
oysters in the hand may be worth a good deal more than the chance of getting them
next year at an enhanced value. Our recommendation of last year accordingly ended
with these words : — " 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 be 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" (Part III., p. 46).

The ' Gazette' of 16th December, 1904, gave notice, however, that the fishery, to
begin on 20th February, 1905, would include the South Cheval, the South-west
Cheval, the Mid-east Cheval, and the North and South Modragams. We can now
only hope that the decision of Government to fish the South Cheval in 1905, in
addition to the beds we had recommended, was a wise one, and that it has not too
seriously affected the fishery prospects of the next few years. It is to be feared that
the oysters fished from the South Cheval will scarcely maintain in pearl-yield the
reputation of their bank, and it can scarcely be doubted that, had they been left, and
had they survived, their value next year would have been much greater.

The inspection in February, immediately before the fishery, showed fortunately
that no catastrophic change had taken place. The numbers for the South Cheval
and North Modragam were increased, those for the Mid-east Cheval and South
Modragam were reduced somewhat, while the very old oysters on the South-west
Cheval, as was to be expected, had continued to dwindle and were now so few and so
scattered as to be scarcely fishable.

The most noteworthy observation in this inspection was, as Mr. Hornell states,
that on the south-east side of the South Modragam " the bed extended beyond the
limit of the ground inspected, and there was therefore a probability that the actual
numbers would prove considerably in excess of the estimate." I discuss this particular
bed of oysters further on (see below, p. xi.).

The following statement of the numbers* estimated at the November and February
inspections, and the actual numbers fished from the beds, may be worth recording : —

* The numbers recorded by Mr. Lewis and by Mr. Hornell respectively in their reports differ some-
what in detail on account of the methods adopted in assigning the oysters to their banks.

THE GEEAT PEARL FISHERY OF 1905.

[xi]

Paar.

Novemher.

February.

Fished.

South Cheval

40,220,000

13,750,000

4,700,000

21,000,000

Not estimated.

| 47,500,000 |

9,000,000

10,000,000

19,700,000

Not estimated.

29,383,444

1,503,590

2,975,849

7,280,817

Say, 10,000,000

„ 30,732,820

Adjacent part of South-east

Mid-east and South-central Cheval . .

North Modragam

South Modragam

81,876,520

Mr. Hornell estimates that about forty-one and a half millions of adult oysters
are left on the banks. It is to be feared that a considerable number of these will die
off during the year, but a solid bed of 10,000,000 remains on the South-east Cheval,
and will probably be available for fishing next spring. There may be other patches
on other parts of the Cheval still remaining in sufficient numbers to be worth fishing,
and there ought to be a good many left on the South Modragam and the Kutiramalai
paars, but these are said to be largely overgrown with a younger generation, and it is
quite doubtful whether a sufficient number of the old will survive to yield a fishery
in 1906 on these banks.

I now come to a question connected with the large number of oysters recorded in
Mr. Lewis's report as fished from the South Modragam Paar, upon which I wish
to make some observations. I shall quote first a couple of passages I find in
Mr. Hornell's report (p. 38) in regard to the finding and fishing of these oysters : —

" 34. On 22nd February a considerable number of boats fished on the ground to the
south of that portion of the South Modragam inspected earlier in the month. The
whole of this area is composed of purely sandy bottom, whereon lay oysters free from
commingling with any of a younger generation ; all were a little over three years
old. In some places they lay in loose bunches ranging from five to twelve individuals
in each ; elsewhere the oysters clung to the projecting edges of large wedge-shaped
shells of Pinna bicolor, rooted upright in the sand by the narrow end. The former
condition delighted the hearts of the divers ; the latter gave those ui^rovided with
protecting finger-stalls considerable trouble, the sharp edges of the Pinna shells
inflicting frequent cuts and scratches on the hands tearing them from their sandy
foothold. In both cases the divers filled their bags with remarkable celerity,
40 seconds in many cases sufficing to fill the diver's bag with 60, 70, or even 100
oysters. As a result, the day's catch, aggregating the enormous total of 4,574,460
oysters, broke every known record."

"35. Work, on this sandy area, continued with feverish activity for the remainder
of the week. The daily catches never fell below 4,000,000 per day, while on 24th
February high-water mark was reached with the enormous take of 5,005,685 oysters.

b 2

[xii] CEYLON PEAKL OYSTEE REPOET.

Such a large total for one day's fishing establishes a record that is likely to remain
unsurpassed for many years to come."

It is exceedingly interesting to find thus, from the official reports, that a considerable
number of the oysters (from twenty-seven to thirty millions) credited (in the report
of the Superintendent, p. 2) to the South Modragam Paar were really fished from an
area extending to the south-east far beyond the usual limit of that paar. We must
either conclude that the oysters have extended over the sandy ground lying between
the South Modragam Paar and Kutiramalai point or that a new bed of oysters in
this position has been fished. The difference may be considered to be unimportant,
as being little more than a point in nomenclature. I notice that Mr. Lewis in his
report prefers to regard the whole of the large irregular-shaped area as being the
South Modragam Paar ; while Mr. Hornell gives to this new southern part a new
and quite appropriate name, " Kutiramalai Paar," on the analogy of Aripu Paar,
Vankali Paar, Karativo Paar, Chilaw Paar, and others — all named after the land off
which they lie. I think the latter course is the better simply because it does less
violence to our existing ideas, charts and definitions. I reproduce here (fig. 2) a
tracing from Captain Donnan's chart with which I worked when on the pearl banks
in 1902. It shows the North and South Modragams as two little areas of approxi-
mately equal size, and the South is certainly nothing like, either in shape or position,
the area from which, judging from Mr. Hornell's sketch-plan given as " Annexure I "
on p. 49 of his report (Colombo, 1905), the oysters in question were obtained. In
the definitions of these paars which I gave in the first volume of this Report I find
that the North Modragam is described (Part I., 1903, p. 105) as lying " south-east of
the central part of the Cheval Paar, at from \ mile to 1 mile distant, and is nearly
1 mile in diameter. It is about 8|- miles west of Kallar tower. The depth is from
5f to 6| fathoms," &c. The description of the South Modragam runs (loc. cit.,
p. 106): "This lies 1 mile south-south-east of the North Modragam, and is about
^ to f mile in diameter. It is about 7 miles north-north-west of Kodramallai Point,
and has a depth of 5|- to 6 fathoms. The bottom is rocky," &c. All this agrees with
Captain Donnan's chart and, I think, with previous records of these paars, but not
with the area from which the thirty million oysters were fished this year, as shown
in Mr. Hornell's sketch-plan which 1 here reproduce (fig. 3) for comparison with
fig. 2. Consequently I would favour the application of the new name " Kutiramalai
Paar " to this very considerable southerly extension of the area hitherto known as
the South Modragam Paar. It has this year proved itself to be of very much greater
importance than many paars which have for long enjoyed distinctive names.

Now to turn to a more interesting point than mere nomenclature. Have these
oysters on the Kutiramalai Paar, which had apparently not been inspected and
estimated last year, but which have been fished this year along with those of the
South Modragam, ever been seen before ? / believe they have, and that I found
them myself in March, 1902, when dredging along with Sir William Twynam,

THE GREAT PEARL FISHERY OF 1905.

[xiii]

Captain Donnan, and Mr. Hornell, in the " Lady Havelock." When I read of
these oysters in the Reports of Mr. Lewis and Mr. Hornell, I at once recollected

CK

* *

;n m!

»^ m\

Fig. 2. Captain DoNNAN's outlines and positions
of the Modragams (KM. and S.M.) and Cheval
Paar (Ch.) in relation to Kutiramalai Point
(K.P.), the Shoal Buoy (S.B.). The positions of
the dredgings made from the " Lady Havelock "
on March 17th, 1902, are shown by the four
crosses lying to the south-east of the South
Modragam. Scale one-fourth of an inch to the
nautical mile.

Fig. 3. Mr. Hornell's sketch-plan of the fishing
ground of 1905, reduced to the same scale as
fig. 2, one-fourth of an inch to the nautical mile.
The thick lines show the areas fished, the thin
line in the South-east Cheval the area with
oysters over three years old left unfisherl, and
the dotted line in the South Modragaru the
ground occupied by oysters six months old.
Below that is the large area which Mr. Hornell
has called the " Kutiramalai Paar."

the occasion, and on turning to my notes I find (see this Report, Part I., ' Narrative,'
p. 84, published in 1903) :—

" After rejoining Captain Donnan and the inspection boats in the South Cheval
district, we took four hauls of the dredge between the South Modragam Paar and
Kodramallai Point. These may be united as Station LXIV. From between South
Modragam and Jaggerboom Paars, along a line south-east towards Kodramallai Point ;
depth 4 \ to 5^ fathoms ; bottom coarse sand, with much fine green weed and small
pearl oysters."

" The fine green weed from the bottom had very young spat of pearl oysters on it.
The small oysters dredged were 8 or 9 months old, and were in quantity at about
3^ miles off Kodramallai Point." I had brought Sir W. Twynam from Jaffna that

[xiv] CEYLON PEARL OYSTER REPORT.

same morning (March 17th, 1902) and wished to show him how the dredge could
bring up oysters or whatever else lay on the bottom, and I believe that these first
dredged young oysters tbat we showed him " between South Modragam Paar and
Kodramallai Point " were a sample of the oysters which have been fished this year
from the " Kutiramalai Paar," and have contributed much to the success of the
present fishery. The age of the oysters agrees, and the position on the chart is about
right. The four little crosses on fig. 2 show the four dredgings united in the
' Narrative ' as Station LXIV.

Finally, I would make a general remark upon paar-ground in the Gulf of Manaar.
It is, that we need never be surprised to find that boundaries of banks alter, and that
oysters appear on occasions in new spots. I have already pointed out in a previous
part of this Report that the whole of the shallow shelf within, say, the 12-fathom line
is potential paar-ground. The bottoms shift to some extent and may change their
character, sand may be washed over a paar, or again may be swept away, leaving a
hard bottom. One of my first observations when dredging on the " Lady Havelock"
was that we found oysters in the Cheval district where they had no right to be,
according to Captain Donna n's oiitline of the paar. This statement may seem to
cut at the root of my own charts and definitions ; but it is not really so. The
outlines of the paars are justifiable and, in fact, necessary as representing the normal,
departures from which may be expected, but are exceptional.

It is evident that there is some difference of opinion amongst the authorities in
Ceylon as to whether or not " dredging is economically a more sound method of
fishing than is diving." I am inclined to think that the operation has not yet had a
fair trial ; but even though it may not, under the conditions in vogue in the East, be
able to compete economically with native diving, I must emphasise what I have said
elsewhere in this report, that the utility of dredging is by no means confined to
obtaining a supply of adult oysters for the market, but is really fourfold, consisting as
it does : — ■

(a) In fishing oysters ;

(6) In cleaning the ground and removing enemies ;

(c) In thinning out overcrowded beds ; and

(d) In spat transplantation.

Its value is not properly assessed if account be taken of the first of these alone, or
even of the first and the last. Finally, it must be remembered that several of these
important operations can usually be carried on in the same series of dredgings.

Mr. Hornell is acting in accordance with these views, and at the recent fishery it
is evident that a good deal of transplanting, cleaning, and thinning out went on
simultaneously with the fishing for oysters. I shall quote a few sentences in regard
to this work in Mr. Hornell's own words : —

"90. On 15th March enough of mature oysters had been removed by the divers

THE GREAT PEARL FISHERY OF 1905. [xv]

from the central portion of the South Cheval to permit of the commencement of
operations for the establishment there by transplantation of a bed of young ones.
Accordingly, I instructed Captain Jelstrup to dredge for young oysters aged about
six to seven months, which I knew to be in profusion in the South Modragam Paar,
to convey them to the South Cheval and there to throw them overboard within the
limits which I defined by a series of mark buoys.

" 91. The method adopted was to bag all the young oysters taken, to keep them
in a cool place covered with wet sacks, and to steam twice a day, noon and evening,
to the South Cheval, throwing them out as the vessel manoeuvred between the flags.
The young oysters stood the treatment well ; there was practically no mortality, as I
ascertained by sending divers down from time to time to ascertain the condition of
these young, and to bring me samples.

" 92. The total quantity transplanted was upward of ten millions. I propose to
concentrate attention upon this bed during the next season, in order to give the
experiment fair treatment. I hope to transplant thereto in November next an
additional 10 to 15 millions ; after that we may hope that enough work shall have
been accomplished to ensure a small fishery two years later — a result that would be
due entirely to cultural methods, and not to the fortuitous interaction of currents and
other natural influences.

" 93. During the course of the dredging, a great amount of good was done by
the capture and destruction of large numbers of starfishes and carnivorous gastropod
molluscs, noted enemies of the pearl oyster. Frequently between 200 and 300
starfishes were taken and destroyed in one day. By my instructions these were
retained on board for twenty-four hours in order to insure that life should of a
certainty be extinct, when they were returned to the sea."

In regard to the bed of young transplanted oysters, Mr. Hornell adds : —

" 109. Prior to leaving the banks, I directed the Government divers to examine the
area in question. They reported young oysters apparently numerous, the individuals
they brought up were in good health, exhibiting no ill-effects consequent on the
transplanting operation. Some had attached to fragments of cultch rubble laid out
during the course of the fishery.

" 110. (b) Thinning out. — This operation went on concurrently with transplantation,
the abundance of young oysters on the South Modragam being so inordinate as
to constitute, through overcrowding, a grave danger to their own continued prosperity.
Transplantation, by entailing a reduction in the number's upon the South Modragam,
should re-act favourably, and if the density of population be found still too great
when next the bed is inspected, further transplantation should be resorted to. So
great is the present profusion, that the numbers taken in the dredges upon the last
day appeared as great as on the day transplantation was begun, and indicated no
appreciable diminution in the fertility of the bed.

"111. (c) Cultching. — During the fishery a quantity of rubble obtained from the

[xvi] CEYLON PEARL OYSTER REPORT.

indurated limestone strata of Kayts, near Jaffna, was laid down, principally upon the
sandy portion of the South-central Cheval, a region which has never yet yielded a
pearl fishery. The quantity contracted for was 180 cubes; unfortunately the
contractor experienced such difficulties in obtaining vessels to convey the stone to its
destination that he was able to deliver less than half the specified quantity, although
I helped him materially by extending the time limit by a week.

" 112. The stone, on the whole, was satisfactory in quality, and likely to prove a
durable cultch material. In size the blocks approximated 3|- in. by 3^ in. by 2^ in.,
but several consignments contained a proportion of blocks of excessive dimensions,
and these I rejected.

" 113. The total amount laid down was just under 300 tons. Some 40 tons of the
friable semi-calcareous sandstone of Kalpitiya was also used. The greater part was
deposited on the South-central Cheval, some 20 tons on a sandy patcli near the
centre of the South Cheval, and about the same quantity on the north-east quarter
of the South Modragam Paar, a locality where there is no outcropping rock on the
bottom. Several times towards the end of the fishery I received fragments of this
cultch from the divers with young oysters attached to the surface."

This fishery was phenomenal not only in the number of oysters obtained, but also
in the prices that they fetched. While the valuation of the samples taken at the
November and February inspections showed a range of from 8 to 24'65 rupees per
1000, the oysters at the fishery sold at very much higher prices — the lowest being 24
and the highest 124 rupees per 1000. The average price over the whole fishery was
48-89 rupees. Of these oysters sold by the Government, Mr. Lewis states that
" India took a hundred times the quantity taken by Ceylon." It is to be hoped that
India will not be sorry that it took them. Judging from the probable pearl-yield of
these oysters, it is difficult to see how the business can be made to pay at such
inflated prices. However, the Bombay pearl merchants, Kilakarai Moormen and
Paumben Chetties probably know best what they are about, and the mysteries of the
Indian pearl market may justify even more remarkable proceedings than the paying
of 124 rupees for oysters valued by the experts at 1270, 27'41, and 31"10 rupees
per 1000.

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

REPORT

ON THE

ISOPODA

COLLECTED BY

Professor HEEDMAN, at CEYLON, in 1902.

BY

The Rev. THOMAS R. R. STEBBING, M.A., F.R.S., Sec. L.S., F.Z.S.,

FELLOW OF KING'S COLLEGE, LONDON.

[With TWELVE PLATES.]

The interest of the present collection is not to be measured by the number of species
or the number of specimens, still less by the size of the animals. The species are not
numerous, the examples in many cases are few, and of some the dimensions have
proved to be afflictingly small. On the other hand, there is no want of variety, since
the thirty-four species more or less definitely discriminated are spread over five very
distinct tribes and divided among sixteen families. There is no want of novelty, since
two of the families, three of the genera, and fourteen of the species are now added to
science on what appear to be satisfactory grounds. Certainly the family Amesopodid^e,
founded on some small creatures allied to Idotea, must be thought well worthy of
notice.

The only family illustrated by a rather large supply of specimens is that of the
Sphferomidse. But the supply has not thrown much convincing illumination on the
difficulties of this group, which have of late years arrested the attention of several
writers. Any person of impatient temper who has ever attempted, when pressed for
time, to disentangle with unskilful fingers a knotted skein of string, may understand
the plight of a busy naturalist who has Sphaeromidre to classify. There is always the
chance in regard to this family that, after struggling with the complexities of the
situation, one may have done more harm than good, by adding to the confusion instead
of lessening it. It is, indeed, a general disadvantage attending the description of a
local fauna, or of a special collection, that it may involve the fragmentary treatment
of problems which can only properly be solved by a monographic survey.

The tribal division of the Isopoda is here accepted from the invaluable ' Crustacea
of Norway,' by Professor G. O. Sars.

B

2 CEYLON PEARL OYSTER REPORT.

DESCRIPTION OF THE SPECIES.

ISOPODA ANOMALA.

" Tribe : CHELIFERA.
Family: TANAID.E.

Tanais, Audouin and Milne-Edwards.

1828, Tanais, Audouin and Milne-Edwards, 'Resume' d'Entomologie,' p. 182, pi. 29, fig. 1, and
'Precis d'Entomologie,' vol. 1, p. 46, pi. 29, fig. 1.

The genus was named Anisocheirus by Westwood in 1832 ('Ann. Sci. Nat.,'
vol. 27), Zcu.ro by Templeton in 1836 ('Trans. Entom. Soc.,' vol. 2, p. 201), and
Crossurus by Rathke in 1843 ('Fauna Norwegens,' p 35). Sars, when defining it
in 1896 ('Crustacea of Norway,' vol. 2, part i., p. 11), remarked that it was well
distinguished from all the other genera of the family, especially by the circumstance
that the pleon was composed of only five segments and carried only three pairs of
pleopods. He assigned to it only four sjDecies : T. tomentosus, Kroyer ; T. cavolinii,
Milne-Edwards ; T. dulongii (Audouin) ; and T. novce-zealandice, Thomson.
According to Dollfus ('Bull. Soc. Zool. France,' vol. 21, p. 207, 1897) the first of
them is a synonym of the second. But several other species are on record, T. macro-
cheles, Nicolet, 1849; T. brasiliensis, Dana, 1849; T. loricatus, Bate, 1864;
T. gracilis, Heller, 1866; T. willemoesii, Studer, 1884; T. hirsutus, Beddard,
1886 ; and since Sars wrote, several additions have been made to the list, namely,
T. robustus, Moore, 1894; T. grimaldii, Dollfus, 1897; T. chevreuxi, Dollfus,
1897; T. testudinicola, Dollfus, 1897; T. alascensis, H. Richardson, 1899;
T. stamfordi, H. Richardson, 1901; T. philetcerus, Stebbing, 1904; T. normani,
H. Richardson, ] 905. In this group it must be noticed that brasiliensis, gracilis,
willemoesii, hirsutus, robustus, testudinicola, normani, are all excluded from the genus,
not as defined by Dana, but as restricted by Sars, since they all have six segments
in the pleon instead of five. It is not improbable that other more or less correlated
differences will be found to exist, but the material here at disposal would not justify
interference in the matter. It may, however, be observed that T. testudinicola,
Dollfus, is evidently the same species as T. robustus, Moore. T. loricatus, Bate,
described from an imperfect specimen, still awaits fuller description. T. vovcb-
zealandice, G. M. Thomson ('Trans. N. Zealand Institute,' vol. 13, p. 207, plate 7,
fig. 3), according to the figure, has the pleon distinctly six-jointed, but no notice of
this is taken in the text, which includes in the generic definition the character " pleon
five-jointed."

ISOPODA. 3

Tanais gracilis, Heller — Plate I. (D).

1866, Tanais gracilis, Heller, ' Novara Exp., Zool.,' vol. 2, pt. 3 (Crustacea), p. 133, pi. 12, fig. 3.

Heller's description is as follows : — " The head very short, the roundish black
eyes placed forward on the somewhat projecting lateral angles, the front bluntly
triangular. The lower antenna? five-jointed, thinner than, but almost as long as, the
upper, which are three-jointed. The first [coalesced] pera?on segment is the largest of
all, narrowed forward, the second [first free] segment the shortest, the fifth, sixth and
seventh sub-equal one to the other. The pleon narrows gradually backwards. The
first three pleon segments shorter than those of the perreon, but longer than the
fourth and fifth pleon segment, the last (sixth) again larger and apically rounded.
The uropods (Schwanzanhiinge) five-jointed, their basal joint thick, triangular, the
first joint of the appendage (des Anhanges) tolerably long, towards the end somewhat
thickened, the four following somewhat shorter than the first, among themselves
subequal, cylindrical, the last blunt-ended, all the joints beset with long seta?. The
first chelipeds large and strong. The legs little setose. The colour of the body on
the surface yellowish or brownish, with dark points and markings. Locality : —
St. Paul." Length, 3 millims.

In Heller's account there are one or two ambiguities, for, while he speaks of and
figures six segments in the pleon, he uses the expression " fourth and fifth pleon
segment " as if intending to mention a composite segment, and, while he distinctly
says that the uropods are five -jointed, he appears to distinguish a peduncular joint
from a five-jointed ramus. His figure shows only five joints in all, as in our
specimen, which was not otherwise separable from Heller's. The mouth organs
agree with those in the genus Tanais. The upper antenna? have a fourth joint in the
shape of a tubercle, representing the flagellum. The finger in the hinder group of
pera?opods is strongly uncinate. The pleopods are delicate in structure. Whether
there were more than three pairs I did not succeed in ascertaining.

Length of specimen, 2 millims.

Locality : — Gulf of Manaar.

In T. brasiliensis and T. normani the uropods are six-jointed, in T, willemoesii
nine-jointed, in T. hirsutus about twelve-jointed, in T. robustus four-jointed.

Heterotanais, G. O. Sars.

1880, Heterotanais, Sars, 'Arch. Naturv. Kristian.,' vol. 7 (1881), separat, p. 28.

1886, Heterotanais, Saks, 'Arch. Naturv. Kristian.,' vol. 11, p. 333.

1880, Heterotanais, Norman and Stebeixg, 'Trans. Zool. Soc. London,' vol. 12, pt. 4, p. 108.

1896, Heterotanais, Sars, ' Crustacea of Norway,' vol. 2, pt. 1, p. 13.

1897, Heterotanais, Dollfds, ' M^m. Soc. Zool. France,' vol. 11, p. 38.
1901, Heterotanais, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, p. 501.

The leading feature in this genus is the sexual difference in the first gnathopods,
which are small and normally chelate in the female, but large and complexly

B 2

4 CEYLON PEARL OYSTER REPORT.

subchelate or not normally chelate in the male. It is, however, prohahle that up to
a certain stage the males may not exhibit this distinction. The species referred to
the genus by Sars are H. orstedi (Kroyer), //. limicola (Harger), H. tenuis
(G. M. Thomson), //. anomalus, Sars. To these Dollfus, in 1897, adds //. algiricus
and LI. provincialis, but, upon a comparison of the descriptions and figures of
H. algiricus and H. anomalus, there is no character given by which they can be
distinguished. H. V. Hodgson ('Nat. Hist, of the "Southern Cross,'" p. 240, 1902)
remarks that but for the structure of the uropods he would have placed his own
Paratanais antarctica and Beddard's P. dimorphus in Sars' genus Heterotanais.

The sjjecies now offered as an additional member of the genus labours under the
considerable disadvantage of not being represented by any adult male specimen,
apart from which it is not easy to say whether junior or female specimens should be
allotted to Heterotanais or Leptochelia. As will be seen by the description, if the
minute character of the maxillary palp can be trusted, the species belongs to
Heterotanais. As the first gnathopods are remarkably stout, one may be glad to
rescue it from a genus like Leptochelia that derives its name from the slenderness of
those appendages.

Heterotanais crassicornis, n. sp. — Plate I. (A).

In the cephalothorax the line of junction between the head and the first perseon
segment is slightly indicated. The antepenultimate and penultimate segments of
the perseon are the longest. The first five segments of the pleon are subequal, the
telsonic segment not elongate.

The eyes are socketed, at apex in dorsal view apparently, but not really, acute,
lenses about 12 in number.

First antennae : first joint unusually stout, not twice as long as broad, more than
twice as long as the second ; third not longer than the second, very narrow, tapering,
with three setae, one of which is, perhaps, attached to a scarcely perceptible flagellar
joint.

Second antennas shorter and much narrower than the first, second joint the widest,
fourth the longest, but not very long ; a minute flagellar joint is tipped with a very
long seta.

Mandibles nearly as figured by Sars (' Crustacea of Norway,' pi. 6) for H orstedi,
but on the left mandible the cutting edge and accessory plate are far less distinctly
denticulate, and on the right the cutting edge shows a little tooth, the serration of
the upper border being barely perceptible. There is no accessory plate on the right,
and on both mandibles the molar is very prominent, as in the species described by
Sars.

The first maxilla has the palp terminated by one seta, in accordance with the
generic character given by Sars, whereas in the female of Leptochelia this organ is
tipped with two setae.

IROPODA. 5

The first gnathopods are remarkably stout, the fifth joint being a little longer than
broad, but almost quadrate, the sixth subequal to it in length but less broad, with
the trunk a little longer than its rather broad thumb. The thumb has a slightly
serrate inner margin, ends in a small tooth, and is furnished with a row of three setse
on the surface and one seta on the outer margin. The finger is more slender, with
undulating inner margin.

The second gnathopods are very slender, with second joint scarcely longer than the
fifth or sixth, which are subequal ; the fourth joint is much shorter than these, but a
little longer than the third ; the slender curved finger is about as long as the sixth
joint.

The pera?opods are small, with the second joint decidedly longer and, especially in
the last three pairs, broader than any of the other joints ; the fourth and fifth joints
differ little in length, each being shorter than the sixth, which is also considerably
longer than the finger.

The pleopods resemble those of Leptocheha, having the inner margin of each oval
ramus fringed with setse, and the outer margin of the outer ramus carrying one seta
placed high up.

The uropods have a six-jointed inner ramus, of which the first joint is the stoutest
and the last the longest, carrying at its apex one seta as long as the ramus and one
or two shorter setae. The outer ramus, which is also tipped with a long and a short
seta, is composed of two joints, the first very short, the second more than twice as
long, the two together rather longer than the first joint of the inner ramus.

Length : — Of four specimens the largest measured a little over 2 millims., and the
smallest a little over 1 millim. in length.

Locality : — Gulf of Manaar.

The specific name refers to the remarkable stoutness of the first antennae, which

appears to distinguish the species from any hitherto described in this genus or

Leptochelia.

Leptochelia, Dana.

1849, Leptochelia, Dana, ' Amer. J. Rci.,' ser. 2, vol. 8, p. 425.

1900, Leptochelia, Stebbing, in Willey's 'Zoological Results,' pt. 5, p. 614.

1900, Leptochelia, Borradaile, 'Proc. Zool. Roc.,' London, p. 797.

1902, Leptochelia, H. F. Moore, 'Bull. U.R. Fish Conim.,' vol. 20 (for 1900), p. 165.

1902, Leptochelia, H. Richardson, ' Trans. Connecticut Ac. Rci.,' vol. 11, p. 279.

For a fuller synonymy of this genus down to 1900, Willey's ' Zoological Results'
may be consulted. Mr. H. F. Moore's recently established Porto Rican species,
L. incerta, is hesitatingly referred by Miss H. Richardson to L. dubia (Kroyer).
The undesigned coincidence of dubiety and uncertainty in the specific names with
the note of interrogation in the reference very appropriately marks the position of the
systematist in dealing with this genus. When the generic name is justified by the
presence of the extraordinarily elongated first gnathopods of the mature male, the
standing ground is tolerably firm. This applies to four species, but in those leaves

6 CEYLON PEARL OYSTER REPORT.

the determination of female and immature specimens in obscurity. In regard to the
uropods the outer branch seems to vary not only in different species but even in the
same species, being sometimes two-jointed and sometimes one-jointed. The inner
branch has always more than three joints.

Leptochelia mirabilis, n. sp. — Plate I. (B).

While in many of its features resembling L. minuta, Dana, and L. forresti
(Stebbing), the present species is easily distinguishable from them both and from
L. rapax, Harger, by characters of the first and second antennae, of the first
gnathopods and the uropods, as well as by its much greater size.

The cephalothorax has the front rather broadly angled, not reaching beyond the
eyes, the sides at first slightly concave, then considerably bulging. In the dorsal line
the free segments of the peraeon are successively longer to the antepenultimate, with
a slight successive decrease in the remaining two. The first five segments of the pleon
are subequally short, the telsonic segment equal to two of them combined, apically
angular. The pleon at the centre is slightly wider than the peraeon.

The eyes are movably socketed, dark, composed of a few large lenses.

The first antennae are once and two-thirds as long as the body, the first joint a
little swollen and bent at the base, about nine times as long as the third joint, the
more slender second being about eight times the third ; the flagellum of thirteen joints,
carrying sensory filaments, is between two and three times as long as the third joint.

The second antennae are about one-fourth as long as the first, the fourth joint of
the peduncle longer than the three preceding joints combined and more than twice
as long as the fifth joint, which is a little shorter than the two-jointed needle-like
flagellum, not including its two or three long apical seta? attached to the minute
second joint.

The first gnathopods are of very surprising length, being much more than twice as
long as the whole body of the animal, and while, considered in themselves, they are
very slender, on the other hand, when compared with the frame that carries them,
their stoutness becomes a matter for wonder. The basal joints are short, but the
three terminal joints are of enormous length. The pair are not symmetrical and both
members are damaged, so that exact measurements cannot be given. In the larger
one the slender, apically curved, movable finger is equal in length to the first joint of
the first antennas ; it is shorter than the trunk of the hand, which widens to the hinge
of the finger, and is produced to a long slender thumb or immovable finger, the apex
of which is broken. The existing jjortion of the antepenultimate joint is longer than
the trunk of the hand, and is narrower near the base than in the greater part of its
length. In the shorter member the hand widens more abruptly and shows a little
gap at the base of the fingers, the movable finger being sinuous, and having three
little tubercles on the inner margin near the base.

Second gnathopods of quite insignificant size, agreeing in character with the first

ISOPODA. 7

and .second peraeopods, but rather longer, all the joints slender, the third very short,
the sixth shorter than the second, but longer than the fourth or fifth, these being
subequal, each a little longer than the very slender slightly curved finger.

The third, fourth, and fifth peraeopods are shorter and stouter than the preceding
pairs, with curved spines round the apex of the fifth, and small seta? round that of the
sixth joint ; the finger tolerably stout.

The pleopods, as in the other species and in Heterotanais, have the two narrowly
oval subequal rami fringed on the inner margin with long plumose setae, the outer
ramus having high up on its outer margin a single adpressed seta.

The uropods have the inner ramus seven-jointed, the joints carrying numerous setae
(the apex of the seventh broken); the outer ramus has two joints, longer than the first
two of the inner ramus.

Length, (5 millims. to 7 millims. For L. minuta and L. forresti the recorded length
does not exceed 2'5 millims.

Locality : — Gulf of Manaar.

Leptochelia lifuensis, Stebbing — Plate I. (C).

1900, Leptochelia lifuensis, Stebbing, in Willey's ' Zool. Results,' pt. 5, p. 616, pi. 64c, d, pi. 65b.
1900, Leptochelia, sp., Borradaile, 'Proc. Zool. Soc. London,' p. 797, pi. 51, figs. 2-2c.

This species has recently been described and figured, and the illustrations here
given will, I think, show that the Ceylon specimens are in substantial agreement with
those from Lifu. They also show the considerable contrast between the antennas and
gnathopods of the male in this species and those in L. mirabdis. The question,
however, remains open as to a possible dimorphism in the males, which would diminish
or destroy the contrast.

Specimens in the collection from eight stations are referred to this species.

Locality, &c. : — On baskets of oysters hung to buoy at Galle : Two specimens,
female ; one 3-35 millims. long with seven large eggs. From pearl oysters, East Cheval
Paar, Gulf of Manaar : A male, 2 millims. long ; the two teeth on the thumb of
the cheliped rather close together, and on one of the chelipeds not fully formed ;
outer ramus of uropod one-jointed, inner five-jointed. Female from same station,
2"5 millims. long and agreeing with description in Willey's 'Zoological Results.'
From pearl oysters, Gulf of Manaar : Male, 2-5 millims. long ; two teeth on thumb
of cheliped close together ; outer ramus of uropod one-jointed, very short ; inner
ramus imperfect. Female, 3 millims. long. Oft' Mutwal Island, March 19, 1902 :
A female, 4 millims. long ; outer ramus of uropod two-jointed, inner six-jointed.
Cheval Paar, Gulf of Manaar : A female, 3 millims. long ; outer ramus of uropod
one-jointed, inner five-jointed, the last two joints rather long ; the one-jointed ramus
very short. Also a female, 1'75 millims. long; outer ramus one-jointed, inner four-
jointed. At each of two other points in the Gulf of Manaar a female specimen was
taken. By " female " should be understood specimens without any distinctively
male characters.

8 CEYLON PEARL OYSTER REPORT.

ISOPODA GENUINA.

Tribe : FLABELLIFERA.

Family : ANTHURID^E.

Calathura, Norman and Stebbing.

1886, Calathura, Norman and Stebbing, 'Trans. Zool. Soc.,' London, vol. 12, pt. 4, p. 122.
1897, Calathura, Sars, 'Crustacea of Norway,' vol. 2, p. 44.

1901, Calathura, Whitelegge, ' Mem. Australian Mus.,' vol. 4, pt. 2, p. 225.
1901, Calathura, H. Richardson, ' Proo. U.S. Mus.,' vol. 23, p. 509.

To the species Calathura brachiata (Stimpson) Sars has added his C. nor neglect,
Whitelegge his C. gigas, and Miss Richardson her C. crenulata. Bonnier's
C. affinis (1896) seems rather to belong to Paranthura.

Calathura, sp.

Two specimens occur in the collection, one from weed-bearing oysters off south-east
of Modragam Paar, the other from north end of Chilaw Paar, February 2, 1902,

9 to 11 fathoms. Both have dark eyes. The former is about 14 millims. in length,
the latter about 9 millims. They may be the species which Haswell has described
in 'Proc. Linn. Soc. N. S. Wales,' vol. 5, p. 478, plate 10, fig. 5, 1881, as
Paranthura (?) crassicornis, sp. nov. The Australian Anthuridse will no doubt before
long be more fully described, and I have therefore left over these specimens till a
more favourable opportunity offers for their specific identification.

Family : GNATHIID,F.

Gnathia, Leach.

1814, Gnathia, Leach, ' Edinb. Encycl.,' vol. 7, p. 402.

1835, Gnathia, Westwood, Loudon's 'Mag. Nat. Hist.,' vol. 8, p. 273.

1885, Anceus, Haswell, 'Proc. Linn. Soc. N. S. Wales,' vol. 9, pt. 4, p. 1005.

1887, Anceus, H. J. Hansen, ' Dijmphna-Togtets Krebsdyr,' p. 205.
1896, Gnathia, Bonnier, 'Ann. Univ. Lyon,' vol. 26, p. 571.

1900, Gnathia, Stebbing, in Willey's ' Zoological Results,' pt. 5, p. 625.

1901, Gnathia, Ohlin, 'Bihang K. Svenska Vet.-Akad. Handlingar,' vol. 26, No. 12, p. 20.

1902, Gnathia, Hodgson, 'Nat. Hist, of the "Southern Cross,'" p. 241.

The references here given are supplementary to the much longer list supplied in
Willey's ' Zoological Results.' As the material in the present collection is limited to
a single small specimen, it would not be a suitable opportunity for an extended
review of the genus. Especial attention, however, should be directed to (inathia
ferox (Haswell), since it is evidently a near ally of the species about to be described.
The details of the mandibles are similar in the two, but by the relative size and
position of these features the appendages in question are strongly distinguished.
The maxillipeds and first gnathopods of G. ferox are as yet undescribed. It is not

ISOPODA. 9

improbable that a separate genus will be thought desirable for these species, when
they are more fully known.

Gnathia insolita, n. sp. — Plate XII. (B).

The head is very broad, markedly emarginate between the mandibles, the minute
rostral point in the emargination only becoming visible after mounting, which
involved some flattening of the frontal border. The telson is acutely triangular.
Eyes small, dark, oval, slightly oblique in position, facets numerous.

Upper antennas : Second joint shorter than first, first shorter than third, flagellum
with very short first joint, long second, which is subequal to remaining three joints
combined.

Lower antennas : These are longer than the upper, with second joint of the four-
jointed peduncle much the shortest ; of the six-jointed flagellum the first joint is the
longest.

The mandibles implanted wide apart are of uncommon pattern. There is no tooth
on the outer margin ; about midway on the inner is a quadrate process that might
represent the molar ; the apical part presents two strongly divergent teeth, with a
small convex lamina at the bottom of the cavity between them.

The maxillipeds are of very delicate structure, except for the strong muscles in the
second joint. This joint is broader than long, not showing any distinct apical lobe.
In the four-jointed palp each joint is successively narrower than its predecessor.
They have the outer margin fringed with setae, 3 on the first, 5 on the second, 8 or 9
on the third, and 7 on the rounded incurved fourth joint.

The first gnathopods are very distinctive of the species, not having the ordinary
tapering character, but the principal joint being as broad as it is long. Within are
seen the chitinized areas, probably indicative of original joints now in coalescence.
The middle area is the most extensive, the lowest has a feebly indicated companion.
The upper half of the rounded margin is furnished with 21 graduated setae. This
margin is the upper and inner when these valve-like limbs are closed together, but it
is not to be inferred that that is the proper description according to homology. There
is a very small apical joint, with no trace of a division into two joints.

The second gnathopods are rather longer than the four pairs of peraeopods, of
which the second pair are slightly the shortest. In general character all five pairs of
trunk-limbs agree.

The first pleopods have the rami rather longer than those of the other four pairs.
All five agree with the uropods in carrying rather long setae, which is contrary to
custom in the adult male of this genus.

Length barely 2 millims., breadth a little less than half the length.

Locality : — Gulf of Manaar, off Karativo, from a sponge.

The specific name refers to the unusual characters of the mandibles and first
gnathopods.

C

10

CEYLON PEARL OYSTER REPORT.

Family : ETJRYDICID.E.
(Cirolanidse of Harger, Hansen, Bonnier, Stebbing, Norman, and others.)

Among the genera that have been assigned to this family, Eurydice, Leach, 1815,
is beyond dispute the earliest, so that the name Cirolanidse, notwithstanding the
distinction conferred upon it as the title of Hansen's notable work in 1890, is bound
to give way to Eurydicidae. Another generic name has " page precedence " over
Cirolana, since in 1818 Leach established Nelocira as his tenth genus of Cymothoadae,
and Cirolana as the eleventh, assigning a single species to each respectively, Nelocira
swainso?ri and Cirolana cranchii. These are now considered to be one and the same
species. The only distinction between them which Leach supplied was that the
pleon of Nelocira. had five segments and that of Cirolana six. As, therefore, the
generic character of Nelocira is erroneous, it is just that the preference should be
given to the correctly described Cirolana, and this has been done by general consent.
It is, moreover, convenient, because Nelocira is apt to be confused with the very
similarly named but quite distinct genus which Leach at the same date called
Nerocila.

The family at present includes seven closely connected genera, distinguished in the
following synoptic table : —

r Pleon forming only two distinct segments. 1. Colopisthus, H. Richardson, 1
1 < Pleon forming more than two distinct
I segments — 2.

fEyes absent; peduncle of uropods with 2. Cirolanides, Benedict, 1896.
inner apex not produced.
Eyes present ; peduncle of uropods with
inner apex produced — 3.
' Peduncle of second antennae four-jointed ;
maxillipeds without hooks on second
joint,
Peduncle of second antennae five-jointed ;*
maxillipeds with hooks on second
joint — 4.
r First pleopods with inner branch broad. 4. Cirolana, Leach, 1818.
4< First pleopods with inner branch nar-
L row — 5.

f Head and trunk broad.
5 <

LHead and trunk narrow — 6.

902.

3. Eurydice, Leach, 1815.

5. Hansenolana, Stebbing, 1900.

* With regard to a sixth joint in this peduncle in Cirolcma, Conilera and Bathynomus, see Hansen,
'Journ. Linn. Soc.,' vol. 29, p. 339.

ISOPUDA. 1 1

First pleopods indurated ; second with

male appendix attached at base of inner

ramus. li. Conilera, Leach, 1818.

First pleopods not indurated ; second with

male appendix attached far from base

of inner ramus. 7. Conilorpheus, n. gen., 1905.

In addition to these some authors include in this family the huge-eyed Bathynomus,
A. Milxe-Edwards, 187'J, which has the two abnormal characters of an accessory
branch on the first antennae and supplementary branchiae on the pleopods, the blind
Anuropus, Beddard, 1886, in which the uropods resemble the pleopods in character,
and the maxillipeds have a one-jointed palp, and Branchuropus, H. F. Moore, 1902,
which agrees with Anuropus in the characters just mentioned, but differs from it by
possessing eyes and in the general habit of body. I am disposed to allot Bathynomus
to a separate family Bathynomidse, and the other two genera to a family Anuropidse,
as already suggested for the former of them in 1893.

A. Dollfus ('Ann. Sci. Nat.,' Zool., Ser. 8, vol. 20, p. 271) now transfers his genus
Sphceromides, 1898, to the "family Cirolanidae, and institutes in that family a new
genus Faucheria for C<Bcospk(eroma faucheri, Dollfus and Vire, 1900, to which be
believes that SpelceosphcBroma julium, Feruulio, 1904, is nearly allied.

Cirolana, Leach.

1818, Cirolana, Lkach, ' Diet. Sci. Nat.,' vol. 12, p. 347.

1881, Cirolana, Miers, 'Ann. Nat. Hist.,' ser. 5, vol. s, p. .369.

1896, Cirolana, Bonxier, " Fdriophthalmes dn ' Caudan,'" 'Ann. Univ. Lyon,' vol. 26, p. 574.

1900, Cirolana, Stebbing, in Willky's ' Zoological Results,' part 5, p. 629.

1902, Cirolana, F. Moore, ' Bull. U.S. Fish Commission,' vol. 20 (for 1900), pt. 2, p. 166.

1902, Cirolana, Stebbing, ' South African Crustacea,' part 2, p. 49.

1903, Cirolana, Dollfus, 'Bull. Soc. Zool. France,' vol. 28, p. 5.

1904, Cirolana, Stebbing, in Willey's ' Spolia Zeylanica,' vol. 2, pt. 5, p. 11.

1904, Cirolana, Stebbing, in Gardiner's ' Fauna, Maldive and Laccadive Archip.,' vol. 2, pt. 3, p. 701.
1904, Cirolana, H. Richardson, 'Proc. U.S. Mus.,' vol. 27, p. 35.

1904, Cirolana, NORMAN, 'Ann. Nat. Hist,,' ser. 7, vol. 14, p. 437.

1905, Cirolana, Hansen, ' Journ. Linn. Soc. London,' Zool., vol. 29, p. 339.

The species assigned to this genus are now very numerous, and it must be
considered a fortunate circumstance that there is no necessity for transferring them
to Xeloeira, which lapses as practically a nomeu nudum. The references above given
will enable the student to trace a far longer list. The blind C cubensis, W. P. Hay,
1903, needs to be compared with C. area. Dollfus, 1903.

Cirolana sulcaticauda, Stebbing.

1900, Cirolana sulcaticauda, STEBBING, in Gardixek's 'Fauna, Maldive and Laccadive
Archip.,' vol. 2, pt. 3, p. 701, pi. 49b.
Of this species a specimen measuring about 6 millims. in length was taken at

c 2

12 CKVLON PEAEL OYSTER EEPOET.

Cheval Paar, and five specimens occurred in a tow-net gathering oti' Marichchukaddi.
With the five there were four young specimens, two of which did not certainly
helong to the same species. In this latter gathering there were also two specimens
of a young Idotea.

Cirolana parva, H. J. Hansen.

1890, Cirolana parva, Hansen, 'Vid. Selsk. Skr.,' ser. 6, vol. 3, pp. 321, 340, pi. 2, fig. 6-66;
pi. 3, fig. 1-lrf.

1901, Cirolana parva, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, p. 511 (Localities),

1902, Cirolana parva, H. F. Moom:, 'Bull. U.S. Fish. Com., vol. 20 (for 1900), pt. 2, pp. 166,
167, pi. 8, figs. 6-8.

This species is known from both the East and West Indies, and is probably to some
extent variable, since Hansen and Moore agree in saying that the broad, sub-
triangular telsonic segment has the rounded apical margin furnished with about eight
spines, while in our specimens the number appears to be uniformly six.

Moore describes the uropods as "short, reaching hardly to end of telson : rami
subequal, narrow at ends, bifid, their margins furnished with spines and a few short
seta?." But our specimens agree with Hansen's account, according to which the
inner ramus reaches beyond the outer and is broader, and with his figure which shows
it to lie very much broader, so as to make the term subequal quite inapplicable.

The first antenna? have the third joint of the peduncle about as long as the
composite first and second joint, the nagellum little shorter than the peduncle,
nine-jointed in both adult male and female. Hansen gives it as eleven-jointed,
much shorter than the peduncle, Moore as eleven- to twelve-jointed. The second
antenna? have the first three joints of the peduncle very short, fourth a little shorter
than the fifth, nagellum in male twenty-six-jointed, in female twenty- to twenty-two-
jointed. A specimen not fully adult, having the fifth pera?opods unarmed and much
shorter than the fourth, had the nagellum of the first antenna? seven-jointed, that of
the second fifteen-jointed.

In the maxillipeds the antepenultimate joint is furnished with only four or five seta?
on the outer margin.

The first gnathopods have five blunt spines on the hind margin, the bluntness
probably due to wear. At the apex of the inner margin the sixth joint in all the
limbs has a rather strong spine.

The male appendix of the second pleopods is slender, the acute apex slightly
incurved, reaching beyond the rami.

Length, male about 6 '5 millims., breadth 2 millims. ; female about the same size;
specimen not adult, 3 millims. long, 1*25 millims. broad.

Localities: — Male, Chilaw Paar, Station LXIX., 8 to 11 fathoms; female with
young, Muttuvaratu Paar, 1U fathoms. Specimen not adult, Talaivillu Paar.

Cirolana, sp.

The telsonic segment is smooth, triangular, with the apex almost acute, but not

ISOPODA. 13

quite, since a high magnification shows it narrowly truncate, carrying two spines.
Each side also carries two spines just above the apex.

The eyes are rather small, lateral. The first antennae have the third joint about
as long as the composite first and second, and the fiagellum of seven joints, most of
which carry sensory filaments. In the second antennae the ultimate joint of the
peduncle is slightly longer than the penultimate ; the fiagellum is rather long,
fifteen-jointed. The mouth organs are of the pattern characteristic of Cirolana. The
limbs are rather slender, not very strongly setose or spinose, the first gnathopods
being, as usual, discriminated from the succeeding legs by the position of the fifth
joint. The uropods, like the telson, have a fringing of fine spines and feathered setae ;
the inner apex of the peduncle is rather sharply produced about to the middle of the
broad inner ramus, which narrows apically and extends beyond the telsonic segment
and the narrower inner ramus. The integument is covered with scale-like markings.

The length is 5 millims., with a breadth of about 2 millims. The specimen carried
several young ones, still enclosed in membranous capsules, but with the eyes already
visible. I hesitate about applying a specific name, whether new or old, to this small
< »vigerous specimen.

Locality : — Deep water, south of Galle, depth up to 100 fathoms.

Conilorpheus, n. gen.

Onlv the male known. Both first and second antennas short. First maxillae
carrying four strong plumose setae on the inner plate. The maxillipeds having the
plate of the second joint furnished with hooked spines. The male appendix of the
second pleopods is attached far down on the inner ramus. The uropods have the
process of the peduncle very elongate and the outer ramus much smaller than the
inner.

The body is almost cylindrical, with the basal segment of the short pleon covered
by the laciniated seventh segment of the peraeon. In the first pleopods the peduncle
is not longer than broad, and neither ramus is hard.

The generic name refers to the combination of characters, here presented, partly
pointing to the genus Conilera and partly to the genus Eurydice, which takes
its name from the wife of Orpheus. The first maxillae, like those of Bathynomus,
show an unusual feature in carrying four setae instead of three on the inner plate,
and the laciniate border of the last peraeon segment is also uncommon in this family.

Conilorpheus herdmani, n. sp. — Plate II. (A).

The head is produced into a narrow distaby widened process between the first
antennae ; its breadth is not much greater than its length. The second and third
segments of the parallel-sided peraeon are the shortest ; the first segment even at the
forward produced sides is not longer than the seventh, of which the postero-dorsal

14 CEYLON PEARL OYSTER REPORT.

margin is cut into four acute lappets ; between these the pleon segments, second to
fifth, are partially seen. The telsonic segment narrows near the base, forming a broad
oval, of which the serrate apical margin is beset with setae and small spines.

The eyes are not large, wide apart, round.

The short stout first antennae have the second joint of the peduncle as it were
embedded in the broader first joint; the tapering fiagellum is composed of seven short
joints, most of them carrying sensory filaments. The second antennas are about twice
the length of the first. The third joint of the peduncle is the longest, and not much
longer than broad, the fourth joint is of equal length and breadth, shorter but much
broader than the fifth joint and on its obliquely truncate distal margin carrying
some elongate setae ; one such seta is similarly situate on the fifth joint. The
fiagellum is shorter than the peduncle, nine-jointed.

The mouth-parts are as in Cirolana, except that the first maxillae have four stout
plumose setae on the inner plate. The spines on the outer plate are slender. The
maxillipeds are well furnished with setae and spines, and the antepenultimate joint is
large.

The first gnathopods have the second joint distally widened and the third still more
so, the latter having the front apex and hind margin fringed with setae. The fourth
joint has four blunt spines on the hind margin and a long spine on the front apex ;
it entirely overlaps the small fifth joint which carries two blunt spines, one of them
minute. The sixth joint has a stout spine between two thin ones, and some long
setae. The finger is slightly curved, nearly as long as the sixth joint, and, like the
other limbs, has on the inner margin a small tooth or spine at the base of the short
curved nail.

The fourth and fifth peraeopods are nearly alike and about equal in length, but the
second and fourth joints are longer and the fifth and sixth joints are shorter in the
fifth pair than in the fourth. The side-plates are distally subquadrate in the sixth
peraeon segment, but are more produced and distally triangular in the seventh.

The first pleopods have the peduncle of equal length and breadth, with four delicate
coupling spines. The inner ramus is rather shorter than the outer and less than half
as broad. Both have the rounded distal margin fringed with jelumose setae. In the
second pair the peduncle is broader than long ; the inner ramus is almost as long as
the outer and much more than half as broad. The male appendix is slender,
apically acute, attached just above the middle of the inner margin of the inner ramus
and extending considerably beyond both rami, which are armed as in the preceding
pair. In all five pairs the outer ramus is broad, more or less vaulted and distally
fringed with setae. The inner ramus has some distal setae in all but the last pair.

The uropods have the short base of the peduncle produced on the inner side into a
very long process. The inner ramus is large, oval, reaching nearly to the end of the
telsonic segment, and having all its free margin similarly serrate and armed with
short spines and long plumose setae. It has two or three dark markings. The outer

ISOPODA. 15

ramus is only half as broad and half as long as the inner, with similar armature on
the distal half.

Length 6 inillims., breadth 1*5 millims.

Locality :— Station V., north end of Chilaw Paar, February 2, 1902, 9 to 11
fathoms.

The specific name is chosen out of respect to the president of the Linnean Society,
to whose ever watchful activity in the interests of science the collection of species
here described is due.

Hansenolana, Stebbing.

1900, Hansenolana, Stubbing, in Wii.t.ey's 'Zoological Results,' pt. 5, p. 634.

The species H. anisopous, for which this genus was originally established, shows, as
I observed at the time, various points of resemblance to Oirolana sphceromiformis,
Hansen. While retaining the latter in the genus Oirolana, Hansen himself forcibly
calls attention to the singularity of its appearance. He had at command only one
specimen, a female, not ovigerous, 4 "25 millims. long, from St. Thomas, in the West
Indies. The examination of a male specimen in Professor Herdman's collection
induces me now to transfer the species to Hansenolana, the definition of which will
in consecpience require to be modified as follows : —

Head transverse, produced into a process between the two pairs of antennas ; first
five segments of pleon very short. Eyes small and wide apart. Mouth organs nearly
as in Oirolana, but on the inner plate of the first maxillre the three seta? are spine-
like and not coarsely plumose. In the second maxilke the outer pan of plates carry
very few spines. First gnathopods with the sixth joint broad. First pleopods with
narrow rami. Male appendix of the second pleopods attached at some distance from
the base of the ramus.

The two species which at present fall under this definition are easily distin-
guished : —

Apex of telsonic segment broad. 1. II. anisopous, Stebbing.
Apex of telsonic segment narrow. 2. H. sphceromiformis (Hansen).

Between these two may, perhaps, be placed Oirolana hanseni, Bonnier (' Ann.
Univ. Lyon,' vol. 26, p. 574, 1896), originally described from an obviously juvenile
specimen, and now (1905) re-described by Hansen from specimens probably still
immature. In this species the telsonic segment has an apex much less broad than
that of H. anisopous, but considerably broader than that of H. spluevomiformis.

Hansenolana sphseroniiformis (Hansen) — Plate II. (B).

1890, Cirolana sphaeromiformis, H. J. Hansen, ' Vid. Selsk. Skr.,' ser. 6, vol. 5, pp. 319, 351, pi. 4, fig. 3-3</.

1900, Cirolana sphaeromiforinis, Stebbing, in Wu, ley's 'Zoological Results,' pt. 5, p. G34.

1901, Cirolana sphaeromiformis, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, p. 512.

Hansen's full description, together with Ids figures of the female, and those here

1G CEYLON PEARL OYSTER REPORT.

given of the male, leaves only a few points needing to be submitted or explained.
The flagellum of the second antennae has in our specimen ten joints instead of twelve.
The epistome and upper lip I did not clearly make out. Hansen does not call
attention to the almost complete smoothness of the three seta? on the inner plate of
the first maxillae. The second maxillae have three spines on the outermost plate and
four on its companion. In the second pleopods the inner ramus is narrow and has the
male appendix attached about one-third of its length from the base, reaching beyond
both rami, and having apparently a bifid apex. The first four pairs of pleojDods have
setae on both rami ; the fifth pair escaped observation. In the telsonic segment,
which is covered with squamose markings rather more conspicuously than the rest
of the body, the three dorsal carinae are clearly developed. But here the lateral pair
appear to run without divergence and none of the three quite reach the margin of the
segment, in these respects differing from the female represented by Hansen.
According to Hansen the apex is not spinose, and he says the same of the uropods.
In our specimen both the segment and its appendages carry several minute spines,
which are only visible under high magnification. This is also the case in a specimen
sent me from Antigua by W. R. Forrest, Esq.

In the first gnathopods the spines on the inner margin of the broad hand show a
variety of minute deuticulations. In this and the other limbs the point of junction
between nail and finger is marked on the convex margin by a group of microscopic
setules. The first gnathopods cannot claim to be subchelate, though otherwise by
breadth and compactness they make some approach to the corresponding limbs in
//. cmisopous. But, whereas in that species the two following pairs are very
differently and more slenderly constructed, here they show a very near agreement, in
all three pairs the fifth joint underriding the sixth.

Length scarcely 3 millims., breadth 1*5 millims., the proportion being 27:14,
Hansen's specimen was 4-25 millims. long.

Locality : — Among compound Ascidians and with other Isopods at Reef, Galle.

The small size and deceptive appearance of this species, rather than any actual
rarity or any marvel of distribution, may account for the fact that the East Indies
have now revealed a solitary specimen of the male fourteen years after the West
Indies had disclosed a solitary specimen of the female.

Family : AEGATHONID^, nov.

Mandibles, with cutting edge bidentate or simple ; molar represented by a feeble
blade, not serrate. First maxillae with inner plate broadly truncate, outer strongly
produced, ending in an unguis with a small curved spine at its base. Second maxillae
very short, ending in a broadly rounded single lobe. Maxillipeds six-jointed, second
joint not elongate, fourth and fifth joints fused together, seventh well-developed,
blunt. Male appendix of second pleopods affixed at the base of the ramus.

By the character of the mandibles this family offers a connecting link between the

ISOPODA. 17

Eurydicidse and other neighbouring families, such as the Corallanidae, but in every one
of the mouth parts it presents some distinctive feature. It is at present represented
only by a single specimen of a single species, so that it is not possible to say what
amount of sexual dimorphism may occur.

Argathona, n. gen.

The characters of the family will at present suffice for those of the single genus.
Argathona is a nymph or half- goddess, so recently sprung from the brain of
Mr. Justin Huntly McCarthy, that her name is not likely to have been hitherto
borrowed for scientific purposes.

Argathona normani, n. sp. — Plate III. (A).

The first peraeon segment is rather the longest, and the last rather the shortest.
The side-plates are diagonally furrowed, those of the second and third segments less
deep than the rest, and not produced beyond their segments ; the last four pairs are
rhomboidal. The fourth pleon segment overlaps the fifth at the sides. The telsonic
segment, with sinuous sides, becomes rather narrowly triangular as it approaches the
rounded apex. The dorsal surface of the animal from one end to the other is beset
with spines large or small, the only segment free from them being the first of the
pleon, but also a basal area is left free where one segment slides under another, and a
sinuous free area marks what is probably the boundary between the sixth pleon
segment and the true telson. The sixth and seventh segments of the peraeon have
each six pale tubercles among the fringing spines of the hind margin ; the fourth
pleon segment has the same number, but less regularly spaced ; the fifth has two
that are submedian and much larger than those already mentioned, and the sixth
segment has a pair which are quite near to the sides. The telson carries numerous
spines in the serrate border besides those that belong to the dorsal cover, and is
likewise fringed with long plumose setae.

The eyes are dark, rather large, set wide apart.

The short first antennae have the first and second joints apparently fused into one
thick joint, not much longer than the following more slender joint ; the flagellum,
rather longer than the peduncle, is twelve-jointed.

The second antennae have the fourth joint rather longer than the three preceding
joints combined, and the fifth rather longer than the fourth ; the twenty-nine-joiuted
flagellum is once and two-thirds the length of the peduncle.

The frontal lamina is pentagonal, not very large. The epistome forms two arms
widely divergent, reaching beyond the membranaceous upper lip, in which a trans-
verse area is perceptible of normal form, probably more highly chitinized than the
remainder of the appendage.

The lower lip is longer than broad, the lobes elongate piriform, flattened on the
confronting margins, the rounded apices not as usual fringed with setules, but

D

18 CEYLON PEARL OYSTER REPORT.

carrying a shorf booth or spine at the inner angle, and a longer one inserted a little
behind and outside the other.

The mandibles are elongate, that on the left bidentate, that on the right with the
cutting edge undivided ; each has a narrow, transparent, apparently very feeble blade
representing the molar, but devoid of the saw-teeth which are conspicuous in
Eurydice and Oirolana. The palp is implanted near the base of the trunk, its
second joint much longer either than the first or falciform third.

The maxillipeds are narrow, the three terminal joints setose, none very widely
expanded.

The first gnathopods are moderately robust, with four short stout spines conspicuous
on the stout fourth joint ; the fifth joint short, not produced along the inner side of
the sixth joint, which is very slightly armed : the finger curved, simple. The second
gnathopods and first peneopods agree with the preceding limb. The four following-
pairs have a different character, with less tendency to geniculation, except between
the second and third joints. There is little difference in length between the joints
from the second to the sixth, the first three of these having the expanded distal
margin beset with spines of varying length, and some of the spines, especially on the
hinder apex of the fifth joint, are serrate. The finger shows a little projection at the
base of the nail.

The pleopods have large rami. In the second pair the male appendix, affixed close
to the base of the ramus, reaches a little beyond it, and is abruptly narrowed to a
short linear apex ; near the base its margin is fringed with minute setules.

The uropods have the peduncle considerably produced, spinose on its outer part,
but dorsally almost clear, the long process having lateral and apical armature. The
rami are strongly fringed like the telson, and the broad inner ramus, which reaches
a little beyond the telson, is dorsally sprinkled with short spines, but the much
narrower outer ramus has much of its dorsal surface smooth, evidently to suit its
habit of folding underneath its companion.

The smooth ventral surface is orange-coloured, the spiny coating of the back dark
brown, the limbs quite pale.

Length of the specimen in slightly bent position, It) millims. , which is about two and
a-third times the breadth. A second specimen measured 12 millims. in length by
6 millims. in breadth.

Locality : — The smaller specimen was from coral reefs, Gulf of Manaar, the larger
from Station XXXIX., south of Galle, up to 30 fathoms.

The clothing of this species gives it, when under the microscope, a very striking-
appearance ; especially its caudal fan, by the grouping and variety of the spines and
the addition of the long feathered setre in more or less symmetrical arrangement,
produces a particularly agreeable effect on the eye. I have named it in honour of
my friend the Rev. A. M. Nokman, h.t'.L., F.R.S., whose services to the zoology of
invertebrates are justly celebrated.

ISOPODA. 1 lJ

Family: OOKALLANID/E.

It is only necessary here to recall that the family embraces the genera Corallana,
Dana, Taclum, Schiodte and Metnert, Alcirona, Hansen, and Lanocira, Hansen,
with Hansen's definition of the family Alcironidre transferred to the modified family
( lorallanidse.

Lanocira, Hansen.

1890, Lanocira, Hansen, ' Vid. Selsk. Skr.,' ser. 6, vol. 5, pt. 3, pp. 287, 313, 391, 395.

1904, Lanocira, Stebring, in Gardiner's 'Fauna, Maldive and Laccadive Archip.,' vol. 2, pt. 3, p. TOG.

The species may he distinguished as follows : —

TThe hinder part of the body not setigerous. 1. L. kroyeri, Hansen.

I The hinder part of the body setigerous '2.

„ f Head (of male) with frontal horn. 2. L. gardineri, Stebbixc.

I Head without frontal horn — 3.

„ / Telsonic segment with broadly rounded apex-. 3. L. rotundicauda, StebbiNg.

L Telsonic segment with narrowly rounded apex. 4. L. zeylanica, n. sp.

Lanocii-a gardineri, Stebbing.

1904, Lanocira gardineri, Stebbing, in Gardiner's ' Fauna, Maldive and Laccadive Archip.,'
vol. 2, pt, 3, p. TOG, pi. 51 A.

For the features which may apparently be relied on for distinguishing this species
from L. zeylanica, see the next following account of the latter form.

Locality : — One specimen of L. gardineri was taken at the Galle reef, with
compound Ascidians and some other Isopods. A second was taken along with some
small sphreromids at Cheval Paar, and a third was labelled " Gulf of Manaar,"

Lanocira zeylanica, n. sp. — Plate V. (B).

The general resemblance of this species to L. gardineri, which I have recently
described from the Maldive-Laccadive Archipelagoes, is extremely close. The dis-
tinguishing features are that the present form has the body from the fifth peraeon
segment to the extremity of the pleon far more strongly setigerous ; that it has the
front of the head with a well-marked margin and a faintly indicated longitudinal
depression behind it, but no upturned frontal horn and no pair of dorsal tubercles
between the eyes; that the first maxillse are stronger; and that the fifth perseopods
are armed on the third and fourth joints with far longer spines. As in the other
species the second maxillae are tipped with two setse, but the difference in length
between the two is greater here. The apical spine of the fourth joint in the first
gnathopods is here stronger.

The eyes are dark. The first antennas have a flagellum of five joints carrying
sensory filaments. The mandibles with bi'oad base and slender trunk exhibit an apical

D 2

20 CEYLON PEARL OYSTER REPORT.

tooth, and alongside of the plate which carries this a thin membrane, which appears
to be prolonged backward into a slightly curved lingual representative of the molar,
the whole apparatus being somewhat obscured by entanglement with the lower lip.
The upper lip is slightly emarginate.

The curved finger of the first gnathopod shows two minute prominences on the
inner margin, and one such prominence is visible on the other limbs, faintly marking
the base of the nail, but the feature is not peculiar to this species. The telsonic
segment has six spines at the apex.

Length G millims., breadth 275 millims.

Locality : — The dissected specimen was a male from Jokkenpiddi Paar. Two other
specimens, apparently of the same species, were obtained at the south end of Cheval
Paar.

Family: JEGIDJE.

The accepted genera may be distinguished as follows : —

. fPleon abruptly narrower than perreon. 1. Syscenns, Harger, 1880.

L Pleon not abruptly narrower than perseon — 2.
(' First antenna3 with flagellum of not more than six

joints ; maxillipeds of not more than four joints. 2. Rocinela, Leach, 1818.
1 First antennae with flagellum of more than six

joints; maxillipeds of not less than six joints. 3. sEga, Leach, 1815.

Syscenus agrees with Rocinela in respect to the maxillipeds. It is now recognised
that the Rocinela lilljeborgii described by Bovallius in 1885 is identical with
Syscenus infelix, Harger, 1880, but it seems to have been overlooked that Bovallius
himself was the first to acknowledge the generic identity, and to point out the
probability that Havponyx pranizoides, Sars, would prove to be a young Syscenus
(' Bihang till k. Svenska Vet. Akad. Handl.,' vol. 11, No. 17, p. 17, 1887). jEgioclms,
Bovallius, is not accepted as distinct from sEga. Acherusia, Lucas, 1849, and
Alitropus, Milne-Edwards, 1840, are regarded as synonyms of Rocinela; Pterelas,
Guerin, 1836, and JEgacylla, Dana, 1856, as synonyms of JEga.

Mga,, Leach.

1815, JEga, Leach, ' Trans. Linn. Soc. London,' vol. 11, p. 369.

1879, iEga, Schiodte and Meinert, ' Naturhist. Tidsskr.,' ser. 3, vol. 12, p. 334.

1882, JEga, Haswell, 'Proc. Linn. Soc. N.S. Wales,' vol. 6, p. 11.

1890, .ffiga, H. J. Hansen, ' Vid.-Selsk. Skr.,' ser. 6, vol. 5, p. 316.

1897, iEga, Sars, ' Crustacea of Norway,' vol. 2, p. 58.

In recent years species have been added to this genus by Hansen, Whitelegge,
H. Richardson and Norman. Many other references to it may be traced under
those given above and in company with those given below for the genus Rocinela.

In defining the genus Leach laid stress on the ample development of the first two

ISOPODA. 21

joints of the peduncle of the first antennae. The dilatation of these joints is used as
a character also by Schiodte and Meinert. Sars, however, employs the qualified
statement, " the first two peduncular joints more or less expanded," and, in fact,
describes the antennae as very slender both in Mga arctica, Lutken, and JEga
ventrosa, M. Sars. The absence of any expansion from the two peduncular joints in
question is conspicuous in the figures given by Schiodte and Meinert of their species
AUga nodosa. They further speak of the frontal lamina, that is, the plate above the
upper lip between the bases of the second antennae, as large or very large in Mga,
but minute or evanescent in Rocinela. But, taking all the species of the two genera
together, this distinction does not seem to be stable.

In the species about to be described the peduncular joints of the first antennae are
not specially dilated and the frontal lamina is not very large. But while in these
respects it makes an approach to Rocinela, it is clearly separated from that genus by
the flagellum of the first antenna? and by the maxillipeds. Its peculiarities tempted
me to make it the type of a new genus, but I am content to leave it for the present
as a very distinct unit among the many species of the genus /Ega.

JEga, ommatophylax, n. sp. — Plates IV., V. (A),

The very marked and at present seemingly unique feature of this species pertains
to the first peraeon segment of the male. The anterior border of this segment
projects a sub-median pair of cylindrical processes over the large contiguous eyes.
The specific name has been chosen to suggest that their function is protective to the
organs of vision. In Rocinela cornuta, Richardson, the antero-lateral angles of the
first peraeon segment are extended straight forwards, probably with the same object.
The defence obtained is presumably worth the interference with sight that must
result from it.

Male. The head projects a distally widened round-ended frontal process slightly
upturned, with the exception of this process having its dorsal surface almost
completely and its ventral surface partially covered by the dark eyes. The first
peraeon segment, without iucluding its slightly convergent antero-dorsal processes, is
longer than any of the other segments, these varying little among themselves in
length or breadth. The first five segments of the pleon are but little narrower than
the peraeon and are subequal one to the other, somewhat wider than the telsonic
segment, which is broader than long, with its broadly rounded apical margin serrate,
carrying spines and setae and having the central point a little produced.

The eyes meet in the middle line of the head, leaving a little triangular interval
above, but occupying all the hind margin.

The frontal lamina is not large. The bases of the first antennae are concealed from
above by the front of the head, and have a slender peduncle with flagellum thirteen-
to fourteen-jointed. In the second pair the joints of the slender peduncle increase in
length from the second to the fifth, and the flagellum is thirty-two-jointed.

22 CEYLON PEARL OYSTER REPORT.

The upper lip appears to be rounded, membranaceous. The mandibles have the palp
planted near the base of the trunk, with the first joint nearly as long as the second,
the third not very short. The first maxillae are long and slender, the small apex
carrying three hooked spines and lour spinules. The second maxilla? are much broader
than the first, having the inner margin of the apex armed with three little hooks,
in addition to which a small, narrowly oval, movable inner plate is tipped with two
hooks. The maxillipeds have an irregularly rounded epipod, as wide as the second
joint, which is itself wide, elongate, and produced into a narrowly tapering process
tipped with two setules. The third joint is short, distinct, the fourth and fifth have
distinct outlines, but are, perhaps, only apically separate, the fifth carrying outward
curved spines at its apex, its outer margin forming a continuous curve with the
faintly separable sixth and seventh joints, which together have a free inner margin
tipped with two outward curving spines. As will be seen from the figures, this
accounl of the mouth organs in the male has been a little supplemented from the
mouth organs of another specimen possibly of a different sex.

The first gnathopods have a few short finely plumose seta3 along the front margin
of the rather narrow second joint, and the same garniture seems to occur on the
corresponding margin in the other limbs. The third joint is rather longer than the
fourth and carries a single spine at its front apex. The fourth joint has two stout
spines on the hind margin, the fifth is small, underriding the sixth, but not overlapped
in front by the fourth ; its spines like those of the hand are slight. The finger is
nearly as long as the hand, with setules marking the base of the nail.

The second gnathopods are very similar to the first, but stronger and more spinose.
The third joint has a stout spine on the hind margin, and the fourth has four or five
such spines; the fifth joint does not underrkle the sixth. The first perseopod is of
nearly the same appearance. On the left side of the specimen each of these limbs
has, on the inner apex of the short fifth joint, an articulated obtuse process about
twice as long as it is broad, giving the limb a subchelate character. On the right
side of the specimen these processes are not present. Whether they are abnormal
growths on the left side or are accidentallv missing from the right I cannot deter-
mine.* The side-plates of the second gnathopods are round-ended. In the following
limbs they tend to become less and less obtuse and those of the fifth perreopods are
subacute, produced over the first segment of the pleon, of which, however, the angles
are free.

In the last four pairs of pereeopods the third joint attains a considerable length,
this anil the three following joints being armed with numerous well developed spines,
of which a group on the hind apex of the fifth joint, though not very long, are

* It is worth noting for comparison that Sars ('Crustacea of Norway,' vol. 2, p. CI) describes the three
anterior pairs of legs in JEga crenulata, Lutken, as "distinguished by a very conspicuous cultriform spine,
issuing from the end of the propodos, inside the base of the daetylus," and 'WliiTELEGriE (' Mem. Australian
.Museum,' iv., pt. 2, p. 233) describes a similar process on first peiwopods of his Mga ungnstata,

TSOPODA. 23

distinguished by their pectinate character. The genital papilla* on the ventral side
of the last perseon segment are short and broad.

In the tirst pleopods the sinuous inner margin of the peduncle carries seven coupling
.spines and as many plumose seta-. The outer branch, as in the following pairs, is
fringed round most of its margin with plumose setae. The inner branch in tins and
the next pair has a fringe of seta' on the lower half of the inner margin and on tin'
apical holder. The second pair are distinguished by the extraordinary length of the
slender male appendix, which is twice as long as the trunk of the supporting branch.

The uropods have the inner apex of the peduncle greatly produced into an acute
process, of which the inner margin is setose. The broad inner ramus is mostly fringed
with plumose setae and carries eleven spines in the serrate part of its margin. The
shorter and much narrower outer ramus is fringed also with plumose seta? and carries
nine spines.

Length 14 millims.. breadth about G millims.

Locality : — The single specimen, a male, was taken in " deep water off Galle."

A specimen Avhich I deem to be the female or a younger form of the foregoing was
" dredged off Mutwal Island,'' and measured 12 millims. in length, with a width of
about 4-5 millims. It is devoid of the frontal process of the head, and the first
perseon segment is without the two submedian dorsal processes. The last three
segments of the peraeon more decidedly surpass in length the preceding three
segments than in the form already described; the first antenna? have twelve and the
second antenna? have twenty-six joints to the fiagellum. The peculiar process of the
w rist in the second gnathopod and first perseopod is wanting. No male appendix
could be discerned in the (undissected) pleon. Otherwise the agreement of the two
specimens is extremely close. The remarkable eyes are alike in both, and though
they agree with those of Rociuela vigilans, Haswell, that much larger species, if the
figure of the maxillipeds can lie trusted, must lie generically distinct. The tenacitv
with which some of the mouth parts in this and kindred species cling together makes
satisfactory dissection difficult. But there is little reason to doubt that the frontal
lamina, epistome, and rounded membranaceous upper lip figured from the second
specimen would equally well represent those parts in the first, had they there been in
a condition for figuring. Mga cyclops. Haswkll, is described as Inning the eyes
confluent, the telsonic segment sub-triangular.

Rociuela, Leach.

1616, Rocinela, Leach, 'Diet. Sci. Nat.,' vol. 12, p. 349 ("BocinMe," p. 348).

1625, Rocinela, Desmarest, 'Consid. gen. Crust.,' p. 301.

1S40, Acherusia, Li van, • Explor. Algerie, Crust.,' p. 78.

1867, Rocinela, Bate and WESTWOOD, ' Brit. Sessile-eyed Crust.,' part 18, vol. 2, p. 289.

1879, Rocinela, ScmODTE ami Meinert, ' Naturhist. Tidsskr.,' ser. 3, vol. 12, p. 380.

1860, Rocinela, Haswell, 'Proc. Linn. Soc. N. S. Wales,' vol. 5, p. 472.

1663. Rocinela, Hargek, 'Bull. Mus. Conip. Zool.,' vol. 9, art. 23, p. 97.

24 CEYLON PEARL OYSTER REPORT.

1890, Rocinela, H. J. Hansen, 'Vid. Selsk. Skr.,' ser. 6, vol. 5, pp. 298, 316, 406.
1893, Rocinela, Stbbbing, 'History of Crustacea,' p. 34*.

1896, Rocinela, Bonnier, " Edriophthalmes du 'Caudan,'" 'Ann. Univ. Lyon,' vol. 26, p. 578.

1897, Rocinela, H. J. Hansen, ' Bull. Mus. Comp. Zool.,' vol. 31, p. 108.

1898, Rocinela, H. Richardson, ' Proc. Amer. Philos. Soc.,' vol. 37, No. 157, p. 8.

1899, Rocinela, Sars, 'Crustacea of Norway,' vol. 2, p. 65.

1899, Rocinela, H. Richardson, 'Proc. U.S. Mus.,' vol. 21, p. 827.

1900, Rocinela, H. Richardson, 'American Naturalist,' vol. 34, No. 399, p. 218.

1901, Rocinela, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, pp. 520, 523.

1902, Rocinela, H. F. Moore, 'Bull. U.S. Fish Commission,' vol. 20 (for 1900), pt. 2, p. 171.

1903, Rocinela, H. Richardson, 'Bull. U.S. Fish Comm. for 1903,' p. 49.

1904, Rocinela, H. Richardson, 'Proc. U.S. Mus.,' vol. 27, p. 33.

In 1898 Miss Harriet Richardson reckoned the then known species of this genus
at nineteen, of which she provided a useful analytic key. Bonnier's blind species,
R. typhlops, 1896, had probably at that time not come under her notice, as it is not
included in the list. On the other hand, one of the accepted nineteen is the
Tasmanian species which G. M. Thomson named R. spongicola in 1892 (1893).
Since this has about fourteen joints to the nagellum of the first antennae and a
distinct third joint to the maxillipeds, it should be transferred to the genus AZga.
In 1899 Miss Richardson made R. alashensis (Lockington) a synonym of
R. belliceps, which Stimpson had assigned to the genus ASga. In 1903 she described
R. hawaiiensis as a new species, near to R. orientalis, and in 1904 established a new
species, R. affinis, near to Harger's R. oculata, and gave the name R. angustata to a
form which she had previously supposed identical with Hansen's R. laticauda. In
1902 Lanchester instituted R. mundana ('Proc. Zool. Soc. London,' p. 378, pi. 35,
figs. 9-9«).

Rocinela orientalis, Schiodte and Meinert. — Plate VI. (C).

1879, Rocinela orientalis, Schiodte and Meinert, 'Naturhist. Tidsskr.,' ser. 3, vol. 12, pp. 383,

395, pi. 13, figs. 1, 2.
1898, Rocinela orientalis, H. Richardson, ' Proc. Amer. Philos. Soc.,' vol. 37, No. 157, pp. 9, 1 1.

This species belongs to a group distinguished by the Danish authors as having the
eyes clearly separated, the nagellum of the second antennte composed of fourteen to
sixteen joints, and the sixth joint in the first three pairs of trunk-limbs armed
with three or four spines. To distinguish it from its neighbours in this group,
Miss Richardson notes that the frontal margin is not produced as it is in R. dumerilii
(Lucas), and that, whereas Schiodte and Meinert's R. maculata and R. americana
have the telsonic segment " linguate," and both branches of the uropods crenulate on
their exterior margins, in the present species the telsonic segment is subtriangular
and the branches of the uropods are not crenulate on their exterior margins. The
Danish authors mention the crenulation in the two former species, but neither deny
nor affirm it in regard to R. orientalis, the telsonic segment of which they describe as
subtriangular, with rounded sides, a phraseology quite as applicable to their figures of

ISOPODA. 25

that segment in E. a/mericana and R. maculata. A more tangible distinction appears
to lie in the proportions of the uropods, which in both the species last mentioned are
said to have the inner branch a little longer and broader than the outer, while in the
present species it is described as much longer and a little broader. Probably it was
not the absolute length of each branch that was taken into comparison, but the inner
branch was reckoned the longer by all that part of it which extended beyond
the outer branch. It will be seen by the figures here given, that, if the specific
determination is correct, the telsonic segment is subject to some variation, the apical
margin passing from the subtriangular to a rather broadly rounded contour. In
either case the margin is minutely serrate and fringed with minute spines and short
plumose setae. The uropods have the margins a little more strongly serrate and the
armature rather stronger. Their peduncle is greatly produced at the inner apex, and
the long process is rather strongly fringed with setae.

In the first antennae the flagellum is six-jointed, but in the smaller and perhaps
not fully adult specimen five-jointed. The second antennae have a long spine on the
apex of the fourth joint of the peduncle, the flagellum in the male specimen with
fourteen joints on one antenna and fifteen on the other, many of the joints fringed
with setules. In the smaller specimen the flagella were twelve- and thirteen-jointed.

In the mandibles the first joint of the palp is the longest. The terminal joint of
the maxillipeds has two outward curving apical spines, and a similar spine below the
apex (this, however, not being clearly discerned in the male specimen).

The fingers of the prehensile legs are strongly hooked. The ambulatory legs have
the third joint very elongate, esjaecially in the last pair.

The first pleopods have both branches narrow, fringed with setae, except on much
of the outer border of the inner branch. The second pleopods have the outer
branch longer and much broader than the inner. In both the margin is somewhat
irregular and much of it fringed. The male appendix does not reach the end of the
inner branch.

Length of largest specimen, male, 13 millims., with a breadth of 7 millims. The
smaller specimen figured was 11 "3 millims. long by 5 millims. broad. A specimen
laden with eggs was 12 millims. long, 5 '7 5 millims. broad.

Localities: — Station I., off Negombo, 20 fathoms; Station II., off Uluwitti,
8 fathoms ; Station V., Chilaw Paar, 10 fathoms ; on weed bearing oyster spat, S. E.
of Modragam Paar.

Family: CYMOTHOIDiE.

Anilocra, Leach.

1818, Anilocra, Leach, 'Diet. Sci. Nat.,' vol. 12, pp. 348, 350.

1900, Anilocra, Stebbing, in Willey's 'Zoological Results,' pt. 5, p. 639.

1901, Anilocra, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, p. 528.

1902, Anilocra, H. F. Moore, 'Bull. U.S. Fish Comm.,' vol. 20 (for 1900), pt. 2, p. 172.

E

26 CEYLON PEARL OYSTER REPORT.

Anilocra dimidiata, Bleeker.

1857, Anilocra dimidiata, Bleeker, 'Acta Soc. >Sci. Iwlo-Neerl., ' vol. 2, art. 5, pp. 30, 31.
1881, Anilocra dimidiata, Schiodte and Meinert, ' Nat. Tidsskr.,' ser. 3, vol. 13, p. 103, 111,

pi. 8 (15), figs. 5, 6.
1900, Anilocra dimidiata, Stebbing, in Willey's 'Zoological Results,' pt. 5, p. 639.

The two specimens obtained agree closely with the description given by Schiodte
and Meinert, having, in accordance with their conspectus of the species in this
genus, the first antennae geniculate, the coxas not carinate but simple, and the fingers
in the first four pairs of trunk -legs inflated in the middle, the inflation here forming
a single nodule, not two or three nodules as in A. leptosoma, Bleeker. The
colouring also agrees, not merely in being yellow, bespattered with minute dark
specks, which is common to so many preserved species, but in the much more peculiar
character of being very much darker on the right side of the animal than on the
other, in accordance with Bleeker's description. Schiodte and Meinert write as
though either side might be the darker.

The Danish authors speak of their specimens being more or less twisted to the
right, whereas ours are quite straight as in Bleeker's figure. They attribute only
nine joints to the second antennas, while ours have these appendages ten-jointed.
Both first and second antennas, and the latter especially, are much compressed. In
one specimen both members of the second pair, and in the other one member, have
the antepenultimate joint shorter than either of its neighbours.

The fifth pleon segment has the postero-lateral angle produced a little over the
telsonic segment, which is broader than represented by Schiodte and Meinert, but
is, as they describe, obscurely carinate, with raised lateral margins. The peduncle
of the uropods is only shortly produced on the inner apex ; the long narrow rami
are perfectly smooth, approximately equal in length, the outer a little the narrower.

One specimen, carrying elongate eggs, measured 24 millims. in length, the other
being 22 millims. long. In each case the greatest breadth was 8 millims.

Locality : — Palk Bay, 6 fathoms.

Rhiothra, Schiodte and Meinert.
1884, Rhiothra, Schiodte and Meinert, 'Nat. Tidsskr.,' ser. 3, vol. 14, p. 223, 318.
This genus is placed by the Danish authors in the Cymothoinas, the second tribe of
their family Cymothoidas. In the somewhat conjectural reference to it of a single
specimen, a male not fully grown, it would be out of place to indulge in any long
discussion of the characters, for which the original work should be consulted. It may
be remarked that the generic definition refers to the female, not to the adolescent
male, in which the second antennas have a larger number of joints.

Rhiothra callipia, Schiodte and Meinert. — Plate VI. (A).
1884, Rhiothra callipia, Schiodte and Meinert, ' Nat. Tidsskr.,' ser. 3, vol. 14, p. 319, pi. 12, figs. 8-13.
The single specimen here available agrees with the account of the " mas adolescens"

ISOPODA. 27

given by the Danish authors as well as could be expected considering its smaller size.
The slender second antennae, however, consist of thirteen joints instead of twelve.
The eight-jointed first antennas have the five joints of the flagellum each apically
furnished with a spray of sensory filaments. The male appendix of the second
pleopods, which the above-named authorities describe as very thin, hooked, scarcely
reaching the end of the rami, is, in our example, only a third as long as the rami, not
especially thin, and not showing any perceptible hook. The coupling spines of the
peduncle are numerous. The delicately laminar rami of the uropods have fringes of
finely plumose setae and are scarcely, or not at all, shorter than the telsonic segment,
the broadly rounded hind margin of which is fringed with very short but finely
plumose setae. Its base carries dorsally numerous setules, of which there are a few
on the preceding segments. The pleopods are without setae, according to the custom
of the family Cymothoidae.

A feature of our specimen, to which Schiodte and Meinert make no allusion, is,
that in all the limbs the fifth joint has the inner apex protruding, acutely in the first
gnathopod, broadly in the fifth peraeopod, where it is armed with three spines. The
third joint is remarkably short in the former, but tolerably long in the latter pair of
limbs. In all the limbs the finger is strongly uncinate.

Colour orange yellow, lightly sprinkled with small dark flecks, especially on the
sides, the limbs pale.

Length about 6 '7 5 millims., breadth about 2 '75 millims.

Locality : — Station LVIIL, off Karativo Paar, 9 to 26 fathoms.

Irona, Schiodte and Meinert.

1884, Irona, Schiodte and Meinert, 'Nat. Tidsskr.,' ser. 3, vol. 14, pp. 327, 381.
1897, Irona, H. J. Hansen, 'Bull. Mus Comp. Zool. Harvard,' vol. 31, p. 110.
1901, Irona, H. Richardson, 'Proc. U.S. Mus.,' vol. 23, pp. 525, 531.

This genus was placed by its authors in the Livonecinae, the third tribe of their
family Cymothoidae. From the seven other genera which they assign to the same
tribe it is distinguished by one or more of the following characters : — Segments of
the pleon clearly separate, fifth pereeopods subecmal in length to the preceding legs,
or a little longer, with uncinate fingers, the body all moderately convex, the front
broadly or shortly rounded, the pleon deeply immersed in the peraeon.

In the definition of the genus nothing is said as to the mouth organs or the
character of the pleopods. Four species were placed in the genus, to which Hansen
has since added a fifth, 7. foveolata. This last agrees with the species about to be
described in a rather striking feature, of which Hansen gives the following account : —
The side-plates of the sixth, and especially of the seventh, segment are much broader
and posteriorly much more produced than the others, besides on each side rising
considerably above the more lateral part of the dorsal surface of the thorax [peraeon],
which is brought about by the curious fact that these epimera are turned outwards

e 2

28 CEYLON PEARL OYSTER REPORT.

and somewhat upwards." In Cterissa pterygoid (Koelbel) there is a similar
expansion of the side plates, hut it applies to all six of them on one side of the
animal and to none on the other side.

Irona nanoides, n. sp. — Plate VI. (B).

The specimen, a female, having its pouch enormously distended with young ones,
was slightly distorted so as to make the outline of the left side very convex. The
middle of the back is raised considerably above the lateral parts of the segments.
The head has a short, very broad front. The peraeon is very broad, the side-plates of
the fifth segment approaching the character of the two following pairs. The first two
segments of the pleon are overlapped by the last segment of the peraeon, the next
three are very short but wider than the almost semicircular telsonic segment.

The eyes are wide apart, not very large, black. The first antennae are rather stout,
especially as to the first three of the eight joints. The ten-jointed second pair are
slighter, subequal in length. The upper lip has a four-lobed margin as in Ar&locra
cuvieri, Leach, and in Renocila periophthalmi, Stebbing. The mandibles have a
stout first joint to the palp, the second much thinner and a little shorter, the third
shorter and thinner than the second, and armed with a few spines. The trunk thins
out in advance of the palp, apparently carrying a quasi-molar not very remote from
the pointed cutting-plate. The slender first maxilla is tipped with five spinules.
The second maxilla appears to have a membranous apical margin accompanied by a
process carrying small hooked spines. The maxillipeds have the composite second
and third joints long and broad, followed by a joint which is about equal in length
and breadth, narrowed at the rounded apex, to which is attached the narrow terminal
bearing two outward bent hooks at its summit and one such hook on its side.

The gnathopods and perteopods are all very similar in appearance and structure,
the hinder pairs having some superiority of size. The second joint is substantial, but
not conspicuously expanded ; the third in the gnathopods is as long as the hand, but
in the hinder peraeopod is longer than any one of the joints that follow it ; the fourth
joint is short but wide, being especially bulging in the fifth peraeopod ; the fifth joint
is of insignificant size, tending slightly to underride the short curved hand, which
does not exceed in length the simple but strongly hooked finger.

The pleopods are all of remarkable breadth, both branches similar in structure and
devoid of setae. The coupling spines of the short peduncles are small. The uropods
are short, with two subequal oval branches, little longer than the stout peduncle, of
which the inner apex is not produced. There are some tiny spinules on the branches,
of which the inner is decidedly not longer than the outer. Colour in spirit yellow,
with the upturned side-plates whitish.

Length 10 millims., greatest breadth 5 "5 millims.

Locality: — Station XXXIX., Gallehogalle Bank, 16 to 20 fathoms.

The young ones taken from the mother's pouch have the broad front to the head as

ISOPODA. 29

in the adults and the eight-jointed first antennae. In the limbs the nail is rather
more distinct from the trunk of the finger than it is in the full-grown animal. The
seventh segment of the person is, as usual, without limbs, and resembling the
segments of the pleon which at this stage are much narrower than the perseon. The
telsonic segment shows a broadly rounded or very obtusely-angled apical margin, which
like those of the uropods, and possibly also those of the pleopods, is feebly and
microscopically fringed with setules. There is no subapical constriction of the telsonic
segment as in the " pullus stadii primi" of Irona foveolata, and the inner branch of
the uropod is broader than in the young of that species.

In Irona nana, Schiodte and Meinert, the adult female has the outer branch of
the uropod much longer than the inner ; in Hansen's species the inner is considerably
longer than the outer, so that both species may be easily distinguished from the one
here described.

Family: SPHyEROMIDyE.

The genera that with more or less acceptance have maintained places in this family
are Sphoerorna, Bosc, 1802; Campecopea, Leach, 1813; Cymodoce, Leach, 1814;
Dyaamene, Leach, 1814 ; Ncesa, Leach, 1815 (for Nescea, Leach, 1813, preoccupied) ;
Ciliccea, Leach, 1818; Zazara, Leach, 1818; Cerceis. Amphoroidea, Cassidina
Ancinus, all four instituted by Milne-Edwards in 1840; Monolistra, Gerstaecker
1856; Isocladus, Miers, 1876; Ceratocephalus, Woodward, 1877 (not preoccupied
by Ceratocephala, Warder, 1838, and therefore taking precedence of Bregmocerella
Haswell, 1885); Cycloidura, Stebbing, 1878 (for Cyclura, Stebbing, 1874,
preoccupied); Scutidoidea, Chilton, 1882; Plakarthrium, Chilton, 1883 (of which
Chelonidium, Pfeffer, 1887, is a synonym, so that the family Chelonidiidas if
maintained must be named Plakarthriidee) ; Haswellia, Miers, 1884 (for Calyptura,
Haswell, 1881, preoccupied); CymodoceUa, Pfeffer, 1887; No?sicopea, Stebbing
1893; Ccecosphceroma, Dollfus, 1896 (part); Tecticeps, H. Richardson, 1897 ■
Exosphceroma, Stebbing, 1900; CassidineUa, Whitelegge, 1901; Chitonopsis,
Whitelegge, 1902; Parasphceroma, Stebbing, 1902; Vireia, Dollfus, 1905.

This rather unwieldy group suffers at present under various difficulties, towards the
solution of which only a few suggestions can here be volunteered. Eugene Hesse in
1872 undertook to prove, with a reserve which he evidently scarcely entertained that
Sphceroma represented the female of Cymodoce and Dyaamene the female of Ncesa.
The discovery of undoubted males in several species of Sphceroma has shown that the
first part of his hypothesis is untenable, but for the second part there is much to be
said. It is exceedingly probable that Dynamene montagui, Leach, is the young male
and that Dynamene rubra and viridis, Leach, are young forms, female or male of
Ncesa bidentata (Adams), which is the adult male. The colouring, the general
structure, and very frequent occurrence under similar conditions of these four forms
give warrant to this belief (see ' Journ. Linn. Soc.,' London, vol. 12, p. 148, 1874).
From acceptance of this view will follow the necessity of cancelling one of the generic

30 CEYLON PEAEL OYSTER REPORT.

names. Nescea, Leach, has priority, but is preoccupied. Ncesa was substituted for it.
But Dynamene has priority over Ncesa, although its title is a little peculiar. It
was indeed defined in advance of the substituted naming of Ncesa, but the species for
which it was defined were not specified by name until 1818. The simplest issue out
of the complication seems to be by reducing the two genera and their four repre-
sentative species to a single genus and species under the name Dynamene bidentata
(Adams). Thus Ncesa disappears, and Dynamene in its place acquires an intelligible
status.

The relief, however, is not very great, because there are several other species that
have been assigned to this dimorphic genus before its dimorphism was understood,
and of these the true generic position remains uncertain.

Although sexual dimorphism is not conspicuous in SphcBroma, Exosplueroma, or
Parasphceroma, in many other genera of the family it has been more or less clearly
established, and there exists at least a possibility that the females of different genera
may be nmch less divergent in appearance than the males. The latter sex is dis-
tinguished in Dynamene and Campecopea by having the sixth segment of the perseon
dorsally produced, in Zuzara, Cycloidura, Isocladus, and Hasivellia by having the
perseon's seventh segment so produced. In Dynamene, Campecopea, Ciliccea, and
Noesicopea the males have the inner ramus of the uropods degraded (with an exception
subsequently mentioned). In Ancinus that ramus is wanting, but the sexes have
not as yet been discriminated either in that genus or in Tecticeps, which shows a
near relationship to it. The uropods in Tecticeps are biramose, with the inner branch
much the shorter.

In regard to some isolated members of the family, it may be suggested that
Sphceroma algoense, Stebbing, 1875, Cymodocella tubicauda, Pfeffer, 1887, and
Splueroma (?) egregia, Chilton, 1891, must all belong to the same genus, and may
possibly deserve to be united under the name Cymodocella algoensis. Cymodocea
antarctica, Hodgson, 1902, also appears to approach Cymodocella more nearly than
Cymodoce. Exosphceroma amplifrons, Stebbing, would, according to my present
view, stand better in the genus Cymodoce, and, in any case, I agree with my friend
Dr. H. J. Hansen that it cannot properlv be retained under Exosplueroma.

Since the above was written, Mr. Holmes has kindly sent me his interesting essay
on the sexes of Sphseromids (' Proc. California Ac. Sci.,' ser. 3, Zool., vol. 3, p. 295).
He takes the view that the name Dynamene should be accepted in place of Ncesa,
hut further extends it to supersede Ciliccea, a procedure which can scarcely be
accepted without more consideration and argument. As regards the male sex,
Cymodoce seems fairly distinguishable by superficial characters from Dynamene and
both sections of Ciliccea, but whether there are stable and sufficient marks for
separating either the female or the juvenile forms of Cymodoce and Ciliccea in all
cases is less clear. The addition of new species without tolerably full description and
figures is rather to be deprecated than welcomed.

ISOPODA. 31

Sphaeroma, Bosc.

1802, Sphaeroina, Bosc, 'Hist. Nat. des Crustacds,' vol. 2, p. 182.
1873, Sphaeroma, Harger, 'Amer. Jouin. Sci.,' ser. 3, vol. 5, p. 314.
1880, Sphaeroma, Harger, 'Eep. U.S. Fish. Comm. for 1878,' p. 368.

1900, Sphaeroma (sensu restrido), Stebbing, 'Proc. Zool. Soc. London,' p. 552.

1901, Sphaeroma, Stebbing, in 'Spolia Zeylanica,' vol. 2, pt. 5, p. 15.

1901, Sphaeroma, Stebbing, in Gardiner's 'Fauna, Maldive and Laccadive Arehip.,' vol. 2, pt. 3, p. 710.

The definition of the restricted genus may he formulated as follows : — Sexual
dimorphism not conspicuous. Telsonic segment without apical sinus. Maxillipeds
having the fourth and fifth joints fringed on the inner margin with long setae, hut
neither these two joints nor the sixth produced into lobes. First and second
gnathopods having the third and fourth joints fringed on the front margin with long
setse. Uropods with subequal rami.

Conforming to these characters are the type species, S. serratum (Fabricius),
S. terebrans, Bate, S. quadridei datum, Say, and S. walkeri, n. sp. In the multi-
tude of species which from early dates down to the last two or three years have been
assigned to Sphceroma, there are many which are obviously excluded from that genus
as above defined. But there are several of which not enough is known to enable
us to say whether they belong to it or not. There is a presumption in favour of
S. quoianum, Milne-Edwards, since Heller in re-describing that species says that
the first three pairs of feet have the middle joints fringed on the outer side with long
hairs, and there is a similar presumption in regard to S. sieboldii, Dolleus, 1889,
and S. pentodon, H. Richardson, 1904. On the other hand, Whitelegge, describing
S. australe, S. latifrons, and S. phimosum in 1902, speaks of the lobes in the palp
of the maxillipeds as well developed. At least the first of these three may be referred
with some certainty to Exosplmroma.

Sphaeroma walkeri, n. sp. — Plate VII.

Head not very broad or long. Peraeon convex, its segments not greatly differing
in length, although a contrary impression may be produced by the telescoping or
extension of a particular segment. The unsutured side-plates of the first segment
are as usual much the longest, being subacutely outdrawn backwards and forwards.
The next two pairs are apically narrowed, the following three somewhat squared, and
the seventh pair with a rather deep trituberculate hind margin. The transverse
tuberculatum of the segments, which is scarcely perceptible on the first segment,
gradually increases in prominence, and fringes the seventh segment in a very
pronounced manner. In the front division of the pleon, representing five faintly
distinguishable segments consolidated into one, there is a curved transverse row of
tubercles, belonging chiefly to the fourth segment. The telsonic segment is only
moderately convex, and becomes slightly concave near the broadly rounded, slightly

32 CEYLON PEARL OYSTER RETORT.

crenulate apical margin. This shield is ornamented hy four longitudinal rows of
tubercles, two submedian, of about eight tubercles apiece, and two sublateral of four,
which lead to an encircling apical ridge, the sloping sides of the shield carrying at
the upper part short divergent lines of tubercles, and usually a little tubercle outside
each sublateral line.

The eyes are dark, deeply inserted in the front margin of the peraeon.
The first antennae have a broad basal joint, probably representing two joints con-
solidated. The following joint is small, scarcely as long as broad. The flagellum in
the specimen examined consisted of fourteen joints, the first much the longest, as
long as the basal and twice as long as the terminal joint of the peduncle.

The slender second antenna? are rather longer than the first ; the fourth and fifth
joints of the peduncle are equal in length ; the flagellum is fourteen-jointed, rather
longer than the peduncle.

The epistome is somewhat longer than broad ; the upper lip has a feebly trilobed
margin.

The gastric spines on the folds of the stomach at its entrance display a variety of
shapes, which, however, may be customary.

The mandibles have the molar strong and prominent, the palp less slight than
usual. In both pairs of maxillae the plates appear to be somewhat broader than is
usually the case.

The maxillipeds have the plate of the second joint broad, strongly setose round the
convex distal margin and down the surface at some distance from the inner margin.
The latter is armed with an exceptionally long and slender upward-bent coupling
spine. The third joint is exceedingly small, the fourth as broad as long, with inner
margin narrower than the outer, the fifth about square, like the preceding joint
having its inner margin densely fringed with long setae ; the sixth and seventh joints
are subequal, apically setose, considerably longer and narrower than the fifth joint.

The first gnathopods have the hind margin of the second joint setose in its upper
part, the third joint nearly as long, with the plumose setse of the front margin very
long ; the much shorter fourth joint has long plumose setae on the convex front
margin ; the triangular fifth joint is quite small ; the sixth joint is about as long as
the fourth, with short plumose setae distally on the front margin and a serrate spine
at apex of hind margin ; the finger has minute setules along the concave hind
margin.

The second gnathopod differs from the first chiefly by the cylindrical fifth joint,
which is nearly as long as the sixth, the former carrying plumose setae at its front
apex, and both being setose along the hind margin.

The peraeopods are fringed with setae on the front margin of the second, third, and
fourth joints, and on the hind margin of all joints from the second or third to the
sixth in the first two pairs, and on the corresponding but inverted margins of the
other three pairs. The second and third joints are more robust than in the gnathopods.

ISOPODA. 33

The fifth joint is uniformly shorter than the sixth, which in the fifth pair is con-
siderably elongated, being as long, though not as bi'oad, as the third joint.

The fixed inner branch of the uropods has two or three tubercles on its upper
surface ; the outer branch is fringed with setse along the inner margin and has the
outer divided into six or seven teeth.

The colouring (as preserved) is a symmetrical dark-grey mottling on a pale ground.

The specimen figured was 9 millims. long and 5 millims. broad. The second
pleopods showed no trace of a male appendix.

Numerous examples were obtained at Jokkenpiddi Paar, one in tow-net gathering
at Marichcbukaddi, two at Cheval Paar, one in Galle Harbour, and one elsewhere.

The tuberculation of the dorsal surface is not incapable of being regarded as
forming a series of transverse lines, but the general effect produced is that of a
striking contrast between longitudinal lines on the telsonic segment and transverse
lines on the rest of the body. I name this prettily sculptured species in honour of
A. O. Walker, Esq., F.L.S., a carcinological colleague whose cheering friendship I
have for many years enjoyed.

Cilicaea, Leach.

1818, Cilicaea, Leach, ' Dictionnaire des Sciences Naturelles,' vol. 12, p. 312 (" Ciliate," p. 311).

1818, Nassa (part), Say, ' Journ. Acad. Sci. Philad.,' vol. 1, p. 182.

1825, Cilicaea, Desmarest, 'Consid. gen. Crust.,' p. 295 (" Cicilcea," p. 442).

1829, Cilicaea, Latreii.le, 'Le Regne Animal,' vol. 4, p. 138.

1836, Cilicaea, Guerix, 'Iconogr. Regne Animal, Crust.,' pi. 30.

1840, Ncesea (part), Milne-Edwards, 'Hist. Nat. Crust.,' vol. 3, p. 216 ("Nesea," p. 628).

1853, Nesaea (part), Dana, 'U.S. Expl. Exp.,' vol. 13, p. 749.

1879, Nesea, G-. M. Thomson, ' Trans. New Zealand Inst.,' vol. 11, p. 234.

1881, Cilicaea, Haswell, 'Proc. Linn. Soc. N.S. Wales,' vol. 5, p. 475.

1882, Cilicaea, Haswell, 'Proc. Linn. Soc. N.S. Wales,' vol. 6, p. 1.
1882, Ciliccea, Haswell, 'Catal. Australian Crust.,' p. 295.

1884, Cilicaea, Miers, ' Zool. of the " Alert," ' p. 308.

1886, Cymodocea, Beddard, '"Challenger" (Isopoda) Reports,' vol. 17, p. 145.

1891, Cymodocea, Ives, 'Proc. Acad. Sci. Philad.,' pp. 188, 194.

1899, Cilicaea, H. Richardson, 'Proc. U.S. Nat, Mus.,' vol. 21, pp. 831, 838.

1900, Cilicaea, II. Richardson, 'The American Naturalist,' vol. 34, No. 399, p. 222.

1900, Cilicaea, Steering, Willey's 'Zoological Results,' pt. 5, p. 643.

1901, Cilicaea, H. Richardson, 'Proc. U.S. Nat. Mus.,' vol. 23, pp. 532, 535.

1902, Cilicea, H. F. Moore, ' Bull. U.S. Fish Comm.,' vol. 20 (for 1900), p. 172.
1902, Cilicaea, H. Richardshn, 'Trans. Connect. Acad.,' vol. 11, p. 291.

1902, Ciliccea, Whitelegge, 'Mem. Australian Mus.,' Mem. 4, p. 265.

To include the numerous species now assigned to this genus, the following definition
is offered : —

Sexual dimorphism conspicuous. Segments of perason devoid of dorsal processes.
Telsonic segment with an apical sinus. Maxillipeds with fourth, fifth, and sixth joints

F

34 CEYLON PEARL OYSTER REPORT.

produced into apically setose lobes. Gnathopods without fringes of long plumose
seta?. Uropods of male (except in C spinulosa) having only the outer ramus strongly
developed.

Leach having only C latreillii on which to found his genus, availed himself of
characters which would exclude many of the forms now grouped under this generic
name. The four marks which he used for distinguishing Cilic(ea among the
Sphseromidae were : first, the approximate ecmality of the sixth and seventh segments
of the perason ; second, the prolonged medio-dorsal process on the anterior division of
the pleon ; third, the apical sinus with central lobe in the hinder division of the pleon ;
and fourth, the rudimentary character of the inner ramus of the uropods. The second
of these characters is conspicuous in fewer than half the species at present assigned to
the genus.

The lobe within the apical sinus of the pleon is found in oidy four of the species,
and the fourth character is subject to one curious exception, since in C spinulosa the
inner ramus of the uropods is rather longer than the outer. Leach himself evidently
suspected that the long process of the pleon might be peculiar to the male, and this
has proved to be the case. But the females, so far as is known, besides being without
the dorsal process, have the apical sinus, at least usually, simple, and the rami of the
uropods subequal. Whitelegge, however, says that " the sexual differences in
C. hystrix are very slight," and that in C. stylifera "the female does not differ materially
from the male." He thinks it highly probable that the form figured by Haswell as
the female of C. hystrix may really be the female of C. spinulosa, Haswell in his text
leaving the point ambiguous. Miers regards SpJueroma pubescens, Milne-Edwards,
the Cymodocea pubescens of Haswell, as with scarcely any doubt the female of
C. latreillii. H. F. Moore explains Cymodocea bermudcnsis, Ives, as female of
C caudata (Say), and suggests that Dynamene nodidosa, Richardson, is the female
of C. caudata- gil liana, Richardson, nodulosa being apparently named by a slip of
the pen instead of tuberculosa. But S. J. Holmes makes it fairly certain that
Dynamene tuberculosa is the female of Ciliccea cordata, Richardson. As tuberculosa
has page precedence, it will supersede cordata, and not without advantage to the
genus, since to many ears cordata and caudata are indistinguishable. The confusion
caused by the sexual dimorphism in this genus will not, perhaps, be very easily
disentangled.

The following synoptic view of the species rather suggests that, for practical
convenience, those which are devoid of the great dorsal process might be grouped
under a sejmrate generic name : —

Anterior division of pleon in male with

long medio-dorsal process — 2.
Anterior division of pleon in male

without such process — 11.

2s

ISOPODA.
Apical sinus in male with prominent

central lobe — 3.
Apical sinus in male with central lobe

very small or absent — 5.

I The medio-dorsal process with apex
o j simple.

The medio-dorsal process with apex
bifid— 4.

r Outer ramus of uropods sub-apically
4 <^ notched.

L Outer ramus of uropods not notched.

I Apical sinus of pleon in male with
r ! very small central lobe.

I Apical sinus of pleon in male without

i. central lobe — 6.

f Surface of perseon conspicuously spinu-
lose — 7.
Surface of perseon smooth or mode-
(. rately granular — 8.
I Inner ramus of the uropods in male

35

1. C. latreillii, Leach, 1818.

C. crassct, Haswell, 1882.
C. longispina, Miers, 1884.

4. C. antennalis, Mieks, 1884.

not shorter than the outer.
Inner ramus of the uropods in male
shorter than the outer.

J Apex of medio-dorsal process not dis-
o I tinctly bifid — 9.

Apex of medio-dorsal process distinctly
I bifid— 10.

[ Apex of medio-dorsal process truncate,
g J slightly indented.

Apex of medio-dorsal process truncate,
t slightly trifid.

I Outer ramus of uropods in male with
. . I bifurcate apex.

Outer ramus of uropods in male with
[_ simple apex. 10.

'Apical sinus of pleon in male with

minute central lobe. 11.

} Apical sinus of pleon in male with
large central lobe — 12.
Apical sinus of pleon in male without
central lobe — 13.

p 2

5. C. spinulosa, Haswell, 1882.

6. C. hystrix, Haswell, 1882.

7. C. caniculata (Thomson), 1879.

8. C. granulata, Whitelegge, 1902.

9. C. tenuicaudata, Haswell, 1881.
C. whiteleggei, n. sp., 1905.
C. sculpta (Holmes), 1904.

36

CEYLON PEARL OYSTER REPORT.

15<|

' Central lobe with the flanking apices

bifid.
12<^

Central lobe with the flanking apices

simple.

f Apical sinus in male simple — 14.
1_ Apical sinus in male sculptured — 17.

r Apical sinus visible from above — 15.
\ Apical sinus dorsally concealed — 16.

Movable ramus of uropods in male

stiliform.
Movable ramus of uropods in male
uncinate.
f Apical sinus concealed by a trilobed
» J process.

Apical sinus concealed by a simple
^ acute process.

'Apical sinus with two teeth in the
border.
Apical sinus with four teeth in the

border.
Apical sinus with six teeth in the
border — 18.
f Teeth in the sinus boldly cut.
|_ Teeth in the sinus minute.

12. C. granulosa, H. Richardson, 1899.

13. C. li7iguicauda,IL. Rtchardson, 1901.

17^

14. C. stylifera, Whitelegge, 1902.

15. C. carinata, H. Richardson, 1900.

16. C. curtispina, Haswell, 1882.

17. C. ornata, Whitelegge, 1902.

18. C. beddardi, n. sp., 1905.

19. C. caudata (Say), 1818.

20. C. tuberculosa (H. Richardson), 1899.

21. C. gilliana, H. Richardson, 1899.

In regard to this list it should be mentioned that Miers treats Haswell's
C. crassicaudata and his own C. longispina as two varieties of C. latreittii. The
specific name of C. antennalis he adopts from White, who published it in 1847
without a description. The species described by G. M. Thomson, in 1879, as Nesea
caniculata was recorded by Thomson and Chilton, in 1886, as Ncesa canalicidata,
with a notice that Miers supposed it to belong to Cilicaea, and that the type specimen
had apparently been lost. Miss Richardson's C. gilliana was described by her
under the name C. caudata gilliana, as a new sub-species of C. caudata (Say).
Species that may belong to the genus, but which have as yet only been described in
the female form, are not included in the list.

Cilicsea latreillii, Leach — Plates III. (B), VIII.

1818, Cilicaea latreillii, Leach, 'Diet. Sci. Nat.,' vol. 12, p. 342.

1825, Cilicsea latreillii, Desmarest, 'Consid. g&i. Crust.,' p. 296, pi. 48, fig. 3.

1836, Cilicaea latreillii, Guerin, ' Iconogr. Regne Animal,' pi. 30, fig, 4 (by error marked

2 on plate ; see explanation of plates, p. 32, second edition).
1840, Nffisea latreillii, Milne-Edwards, 'Hist. Nat. Crust.,' vol. 3, p. 218.

ISOPODA. 37

1881, Cilicaea crassicaudata, Haswell, ' Proe. Linn. Soc, N.S. Wales,' vol. 5, p. 475, pi. 17, fig. 3.
18S4, Cilicaea latreillei, MlEKS, ' Zool. of the "Alert,"' p. 308, var. crassicaudata, p. 309.
1902, Ciliccea crassicaudata, Whitelegge, ' Mem. Austral. Mus.,' Mem. 4, p. 273, fig. 35.

Male. — Miers says, " the segments of the body are covered with a very short stiff
pubescence." In the Ceylon specimen the integument is clothed with plumose setules
accompanied by thin pellucid undulating fringes, the precise character of which is not
easy to determine. Their function may be to retain a concealing coverlet of mud on
the animal's coat, as suggested by Doflein (' Brachyura of the " Valdivia," ' p. 203),
for the " leaf-hairs " of the Dromiacea and Oxyrrhyncha. The hind region of the last
three perseon segments is ornamented with transverse rows of small granules. The
long unsutured side-plates of the first segment have the narrowly-squared front apex
commonly found in this and some other genera. In contrast to these, the side-plates
of the next three segments are extremely short, those of the third segment ending
acutely, with the points less produced than those of their neighbours on either side.
The anterior division of the pleon projects a thick, apically obtuse, medio-dorsal
process over and beyond the telsonic segment, leaving exposed to view the pair of
sub-lateral bosses on that segment, but covering both the central lobe and the sides of
its apical notch.

The first antennae in the specimen figured have a flagellum of twenty-one joints.
The second antennas in the same specimen show a want of symmetry, in one member of
the pair the fourth and fifth joints of the peduncle being considerably longer than any
of the preceding joints, the flagellum ten-jointed, shorter than the peduncle, while in
the other member the fourth and fifth joints of the peduncle are each shorter than
the second, and the flagellum is twelve-jointed, longer than the peduncle. Normally,
as shown by another specimen, and in Desmarest's figure, the first antennae are not,
as in this instance, longer, but shorter than the second.

Mandibles with dark horn-coloured cutting edge with dentation obscure, accessory
plate bidentate on one mandible, simple on the other, a tuft of spines between this
and the prominent molar, palp slender.

Maxillipeds : The second joint has a sinuous outer margin to the stem, its plate is
of moderate breadth, and the ultimate joint of the palp reaches well beyond the lobe
of the penultimate joint.

First gnathopods : Third joint elongate, hind margin carrying many little groups of
spinules, front distally channeled and above the groove produced into an obtuse spine-
tipped process, fourth joint broader than long, its spine-tipped front apex overlapping
the fifth joint, with seven spines along the hind margin ; small fifth joint with five,
and sixth joint with six spines along the hind margin, of the last set the sixth being
shorter than the one before it, but with this exception, the three successive sets of
slightly serrate spines beginning with smaller and ending with larger forms. The
finger is strong, as long as the stout sixth joint, ending in two horny nails, of which
the short inner one is serrate on its inner margin.

38 CEYLON PEARL OYSTER REPORT.

The second guathopods are less robust, the spines of the front margin more
developed, those of the hind margin more numerous, but for the most part not so
strong, the fifth joint not differing greatly in size and character from the fourth, the
finger not quite so long as the sixth joint, and its inner nail showing no serration.

The fifth perseopod is the longest of the limbs, the process of the third joint by
successive reduction in prominence here almost disappearing, the fourth, fifth, and
sixth joints attaining their greatest elongation, and the finger being decidedly shorter
than the preceding joint.

The first pleopods exhibit an interesting feature of the inner branch in that its
inner margin, instead of joining the outer either by a continuous straight line or a
convex curve, here makes a slightly concave sweep from the point where the fringe
of long, feathered seta? begins. The second pleopods, as already described and figured
by Mr. Whitelegge, are remarkable for the great length and corrugated appearance
of the male stilet. This in our specimen, as in that described by Mr. Whitelegge,
was stiffened into a strongly bent position. The three following pairs have the
covering branch sutured and lying very close to the branchial plate. In the fifth
pair the short, but broad, portion beyond the suture is bilobed, and from the junction
of the lobes rises a forward-pointing lappet, this, with the border of the inner lobe,
and two tubercles just in advance of the suture, being also closely beset with little
teeth or prickles in regular arrangement.

The uropods have the short stout peduncle produced on the inner side to a short
thick process representing the inner ramus, of which the appearance varies consider-
ably with the point of view. The stout, blunt outer ramus has a tooth on the outer
margin at some distance from the apex, often obscure, and, according to Miers,
occasionally obsolete.

Length, 11 millims. to 13 millims. From the tendency of the specimens to fold up,
the exact length is not very easily measured.

Localities : — South of Manaar ; Coral reef, Gulf of Manaar ; Pearl banks, Palk Bay.
One specimen was obtained at each station.

Cilicsea latreillii, Leach, juv. — Plate III. (B).

This small specimen is remarkable for its coat. This may be described as thickly
beset with short, stiff, irregularly blunt setse covering all the dorsal surface,
except such parts as are adapted for sliding under neighbouring parts. In the
' Zoology of H.M.S. "Alert,"' pp. 308-310, 1884, Mr. E. J. Miers makes Splwroma
pubescens, Milne-Edwards, 1840, doubtfully, a synonym of Ciliccea latreillii, and
Haswell's Cymodocea pubescens, 1881, with conviction, another synonym. His
remarks on the variations of this species according to age, sex, and individuality,
should be carefully studied by anyone interested in the subject. Haswell gives the
length of his longest specimen of Cymodocea pubescens as 1 inch. That the setose
covering should be worn down in large specimens might be easily explained, but it is

ISOrODA. 39

less easy to understand in an evidently young specimen, such as that in the present
collection. Probably, therefore, the peculiar setae are not stumps due to attrition,
but an inchoate stage of the undulating fringes noticed above. It will be observed
that the spines of the first gnathopod are not blunted, and the finger clearly shows a
delicate serration of the inner margin and small setules on the outer. In the adult
male the serration is less easy to observe, the finger having become thicker and coarser.

Length 6 millims., breadth 3 millims.

Locality: — Coast of Ceylon, under 100 fathoms.

Cilicaea whiteleggei, n. sp. — Plate IX. (A), (B).

Male. — This species is easily distinguished from C. latreUlii by its much slighter
general structure, as well as by the character of the very elongate dorsal process in
the front division of the pleon, the terminal part of which is not conical but almost
parallel-sided and sharply bifid at the apex. Also, the apical notch of the telsonic
segment is simply semicircular without median lobe. The three species which make
a nearer approach are C. tenuicaudata, Haswell, distinguished by the bifid apices of
the uropods ; C. caniculata (G. M. Thomson), with practically truncate apex to the
dorsal process and the free ramus of the uropods short and thick ; and lastly
C. granulata, Whitelegge, which also has the dorsal process truncate, " with three
small terminal subspiniform granules." This last species attains a» length of
13 millims., therefore greatly exceeding in size that which I am here naming after
Mr. Whitelegge out of respect for his useful researches in this group.

The surface is very inconspicuously granular and hairy, only on the pleon and
uropods showing these characters at all distinctly. The side-jjlates are similar to
those of the preceding species, but having in addition a subcarinate appearance along
the upper portion. The pleon, which is as long as head and peraeon combined, has a
rounded tooth or projection on either side of the dorsal process at its base, which
must be regarded as an additional distinction between this species and C. granulata.
It has " a submedian pair of tubercles transversely disposed behind the middle of the
terminal segment," such as Whitelegge considers distinctive of the female in his
species.

The first antennae have the flagellum consisting of ten or eleven unecpial joints,
with sensory filaments on the last five. The longer second antennae have the last
joint of the peduncle decidedly longer than the penultimate, the flagellum longer than
the peduncle, fifteen-jointed.

In the mouth organs it was probably a casual abnormality that the first maxilla
had only three setae instead of the usual four on the inner plate.

The maxillipeds differ from those of C. latreillii by the stronger stem of the second
joint, which has a straight outer margin and carries a much wider plate ; also the
lobes of the fifth and sixth joints are longer than in the older species.

First gnathopods : The third joint projects and carries a spine at the middle of

40 CEYLON PEAEL OYSTER REPORT.

the front margin, but has no process. The fourth joint has a spine at the front apex
and three on the hind margin, the small triangular fifth has two spines, and the sixth
three on the hind margin ; the finger has a horny principal nail, but the accessory
nail is only represented by a small pellucid spine.

The second gnathopods have similar fingers, but the four preceding joints more
elongate, the fifth joint being of the same character as in the perseopods. Most of
these limbs show a fur-like clothing of the inner margin of the fourth and two
following joints ; the spines are quite small and not very numerous ; the proportions
of the joints are variable, in the long fifth perseopods the fourth joint being as long as
the third or the sixth and a little longer than the fifth.

The first pleopods have the inner branch triangular with no distal emargination of
the inner border. Both branches are covered with scale-like markings. The male
appendix of the second pair is produced only a little way beyond the branches and is
subapically widened but apically narrowed to a short obtuse point, thus differing
from the longer, not subapically widened, stilet in C. granulata. The third pleopods
differ from those of C. latreillii by having the inner branch considerably shorter than
the outer, and the part of the outer branch below the suture with a breadth much
less in excess of its length.

The uropods have a dorsal tubercle on the peduncle ; the peduncle is widened
above on the inner side, and is still more widened below into the short unjointed
inner ramus. The outer ramus is very long and nearly straight, the tolerably acute
tip tending to bend outwards.

Length at full stretch from front of head to end of dorsal process of pleon,
8 '5 millims. The uropods reach little beyond the dorsal process.

Localities : — Cheval Paar, Gulf of Manaar ; deep water off Galle ; off Foul Point,
Trincomalee, Station XXIV.

Female. — The form which I regard as the female of this species is devoid of the
long dorsal process and has the rami of the uropods subequal. It is also considerably
smaller than the male. But in other respects the agreement between the two forms
is so close as to leave little room for doubt that they are the same species. The
maxillipeds are especially characteristic. The antennae, first gnathopods, and fifth
peraeopods, as will be seen by the illustrative figures, are in close agreement. The
pleopods, in addition to agreement in shape, show the same scale-like markings.

The uropods have the fixed branch squarely truncate, the outer narrowly ovate,
with its apex acute, turned slightly outward. The front division of the pleon has a
faintly marked medio-dorsal projection of its hind margin. The telsonic segment has
two prominent submedian bosses and a semicircular apical notch.

Localities : — Trincomalee, Station XXIV., off Foul Point; deep water off Galle.

Cilicsea beddardi, n. sp. — Plate X. (A), (B).
Male. — This species is nearly related to C. caudata (Say), but, according to the

ISOPODA. 41

more recent descriptions and figures of the latter by Ives (' Proc. Ac. Sci., Pliilad.,'
1891, pp. 188, 194, pi. 6, figs. 11-16), and by H. F. Moore ('Bull. U.S. Fish.
Comra.,' vol. 20 for 1900, p. 172, pi. 10, figs. 5-8, 1902), the present form must be
considered distinct. Say, in his original account, speaks of the telsonic segment as
" marked by a deep sinus, within which are two or four small teeth." Ives, who had
six male specimens from Bermudas, takes no notice of the alternative two teeth, but
says, " there appears to be a tendency in the four spines within the sinus of the
posterior abdominal segment to become double." Moore says, " the apical notch is
furnished with four teeth, two small ones at the base, and two larger ones outside of
them and at a slightly lower level. The two limbs forming the borders of the notch
are notched at their tips and furnished with a tuft of setae." These notches at the
tips are also very clearly shown in the figure given by Ives. In the Ceylon species
there are decidedly only two teeth within the sinus, and the tips of the sinus are
not notched, but carry dorsally an upright tuft of setules seemingly seated on a
tubercle.

A few setules rise from the dorsal surface throughout the length of the animal.
There is little difference in length between the segments of the peraeon, except in
regard to the seventh, which is the shortest, and has its side-plates rounded, less
produced than the rest. The anterior division of the pleon has a transverse row of
five small setulose tubercles. To these succeed on the telsonic segment three longi-
tudinal ridges divided into tubercles, some of which carry setules. Behind the ridges
the segment is depressed and has a group of setules in advance of the sinus which
has been above described.

The first antennae have an eight-jointed flagellum. In the second antennae one
member had the flagellum ten-jointed, while in the other it was seven-jointed.

The epistome is conspicuously tri-lobed above. The distal margin of the upper lip
is evenly convex.

One of the mandibles has both cutting edge and accessory plate tri-dentate. On
the other the dentation of these parts is obscure. They are succeeded by a bunch of
spines. The molar is well developed ; the palp slender, with its first joint not much
longer than the second or third.

The first maxillae have the usual four feathered setae on the inner plate, the apical
spines of the outer plate slender.

The maxillipeds are slightly constructed, with the last four joints apically
setiferous, the three preceding the terminal joint being produced into narrow lobes.

The limbs are very similar to those of Ciliccea ivhiteleggei, but rather less robust.
The second peraeopod on the right side of the specimen figured has the sixth joint
reduced to an oval stump, carrying no finger.

The first pleopods are remarkable for having the inner ramus twice as broad as it
is long, agreeing with the same appendage as described by Whitelegge for Zuzara
emarginata, Haswell. The second pleopods also have the inner ramus much

G

42 CEYLON PEARL OYSTER REPORT.

broader than long, with the male appendix a little longer than the breadth of the
ramus, transparent, smooth, of uniform width, with the apex turned a little away
from the ramus. The outer ramus has the serration of its distal outer margin
produced into prominent teeth. The third pleopods have the portion of the outer
ramus beyond the suture nearly as long as it is broad. The fourth and fifth pairs
have branchial folds on both rami.

The uropods are very hairy. The short peduncle is produced into a quadrate
process representing the inner ramus, which has its lower angle a little acutely
produced and beset with short feathered setae. The movable ramus is long,
cylindrical, slightly curved, the outer margin being convex.

Length 5 millims., breadth 2 '5 millims.

Locality : — Cheval Paar ; Muttuvaratu Paar.

Cilicsa (?), sp. juv.— Plate IX. (C).

The small specimen figured is no doubt immature, as may be judged from the
unfurnished condition of the maxillipeds. The pleon differs little from that of the
female or young C. whiteleggei, except that the outer ramus of the uropods shows no
outward curving. The first antennae have a massive peduncle and a flagellum of
eight joints or, perhaps, more. The flagellum of the second antennas is twelve-
jointed. In the finger of the gnathopods the little spine or tooth at the base of the
nail is comparatively strong.

Length, in a somewhat bent position, about 5-5 millims.

Locality : — Palk Bay.

Cymodoce, Leach.

1814, Cymodoce, Leach, ' Edinburgh Encycl.,' vol. 7, p. 433.

1902, Cymodoce, Stebbing, ' South African Crustacea,' part 2, p. 73.

1904, Cymodoce, Stebbing, in Gardiner's 'Fauna, Maldive and Laccadive Archip.,' vol. 2, pt. 3, p. 712.

Cymodoce bicarinata, Stebbing — Plate X. (C).

1904, Cymodoce bicarinata, Stebbing, in Gardiner's 'Fauna, Maldive and Laccadive Archip.,' vol. 2,
pt. 3, p. 712, pi. 52b.

The specimen here figured differed from the type by its greater proportionate
breadth, being 4 millims. broad by 6 millims. long, while the flagellum of the first
antennae was only eleven -jointed, and that of the second fourteen -jointed. But the
long genital papillae* of the seventh peraeon segment and the greatly produced
inward curving male appendix of the second pleopods and various other features
showed close agreement. To judge by other specimens, the breadth is variable.

Milne-Edwards, 'Hist. Nat. Crust.,' vol. 3, p. 213, 1840, describes C. pilosa as

* This term is introduced by Dr. H. C. Williamson in the ' Twenty-second Annual Report of the
Fishery Board for Scotland,' part iii., p. 101, 1904. For an earlier vague use of it by v. Willemoes Suhm,
see ' "Challenger" Amphipoda,' p. 438.

ISOPODA. 43

follows : — " Body very flexible and almost smooth in front, but granular and setose
in the hinder half. Front obtuse as in Sphceroma ; hind margin of the first pleon
segment furnished with two rounded tubercles ; two tubercles similar but much
larger, more salient and above all more elongated, situated on the last pleon segment
and separated by a longitudinal furrow, at the extremity of which is found a boss
furnished with a pencil of long setse. Terminal incision of the pleon very wide ; the
elongate median process almost cylindrical, rounded at the end, and terminating at
the level of the extremity of tbe two teeth formed by the sides of the incision.
Terminal plates of the uropods reaching much beyond the extremity of the pleon ;
the inner large and obtuse ; the outer much broader, thin on the inner side, but very
thick towards the outer margin and armed with a conical tooth at its extremity.
Length about 6 lines. Habitat, the Mediterranean." A length of 12'5 millims.,
compared with 6 millims., makes the difference in size very considerable, but in other
respects there is a close resemblance between the Pacific species and that described
from the Mediterranean.

Localities : — East of Shoal buoy ; off Chilavaturai ; on trap sunk to bottom, Shoal buoy.

Cymodoce inornata, Whitelegge.

1902, Cymodoce inornata, Whitelegge, ' Mem. Australian Mus.,' iv., pt. 4, p. 263, fig. in text.

The specimen referred to this species agrees well with Mr. Whitelegge's account
and figure. It has the body minutely hairy, the pigmented area of the eyes with
the narrow end directed backwards, the process in the apical sinus of the telsonic
segment small and rounded, the outer ramus of the uropods bidentate, with the outer
tooth smaller and higher up than the inner.

On the other hand, it should be stated that the apical sinus is much more marked
than that which Mr. Whitelegge represents, and the tip of the median process
scarcely reaches the level of the apices of the sinus. The outer ramus of the uropods
is not so forcibly bidentate as in Mr. Whitelegge's figure, and above the semi-
circular depression, which that author mentions as extending from one uropod to the
other, the telsonic segment is here bilobed. In the Australian memoir it is spoken of
simply as convex.

Length about 12 millims., equal to twice the breadth. The specimen was crowded
with young ones, in which the eyes were already visible.

Locality : — Coral reefs, Gulf of Manaar. A second specimen was taken south of
Adam's Bridge.

Tribe: VALVIFERA.

To the tribe, in 1897, Sars assigned the three families Idotheidse (synonymous
with Idoteidse), Arcturidse (with suggestion of AstacillidaB as the proper substitute),
and Chaetiliidse. The late Axel Ohlin in 1901 added a new family, Pseudidotheidse,
to receive the three species, Idothea miersii, Studer, Pseudidothea bonniem, Ohlin,
and Arcturides cornutus, Studer. He was, however, almost convinced that the first

G 2

44 CEYLON PEARL OYSTER REPORT.

two were identical, and suspected that in any case both would have to be transferred
to Arcturides, the genus of the third. He recognised that in the latter event the
family name would have to be changed. It would rather inconveniently become
Arcturididse. Another family, likewise intermediate between the Idoteidse and
Astacillidae, is now required. This is remarkable for the negative character in which
it agrees with the Amphipoda Caprellidea, a common result being no doubt referable
not to any close tie of consanguinity, but to the simple fact that in each instance
nature has enabled a species to get rid of limbs for which it had no further use. For
the genera Antarcturus and Pseudoprion, zur Strassen, see 'Zool. Anzeiger,' vol. 25,
p. G86 ; vol. 26, p. 31 ; 1902-03.

Family : AMESOPODID^E, nov.

Body not geniculate, but antennae, mouth organs, first gnathopods, and appendages
of pleon nearly as in the Astacillidae. Second gnathopods ambulatory, not setose, not
fully jointed. First and second pairs of perseopods unrepresented, except by the
marsupial plates in the female.

Amesopous, n. gen.

To the characters of the family none can be added for generic distinction, so long as
the family contains but a single genus. It may be noticed that the first segment of
the peraeon is coalesced with the head, that all the segments of the pleon are fused
into one, and that the wrist and hand of the first gnathopods are fringed with
conspicuously trifid setae.

The name is framed to express the default of the median pairs of legs. In Cleantis,
Dana, a genus of the Idoteidae, which in general facies makes some approach to the
present, the second perseopods are the smallest of the limbs. In Arcturides the head
is only incompletely separated from the first perseon segment, but Ohlin finds this
fusion complete in his Astacilla falclandica, and nearly so in A. magellanica.

Amesopous richardsonae, n. sp. — Plate XI. (A).

Head united to first peraeon segment without apparent suture, rostral point minute,
lateral lobes produced about to the end of the first joint of the upper antennae.
Body in male narrowly cylindrical, the limbless segments of the perseon the smallest,
but in the ovigerous female the cephalothorax widens distally, and the second, third,
and fourth peraeon segments which carry large marsupial plates are much wider, the
third and fourth being also considerably longer than any of the three following
segments. The pleon is narrowly ovoid, narrowly rounded at the apex. In lateral
view the dorsal outline is corrugated, and the female has a pair of dorso-lateral
tubercles on each of the second, third, and fourth peraeon segments.

The eyes are dark, laterally protuberant just below the frontal lobes of the head.
The facets are small and numerous.

The first antennae reach some way along the third joint of the second pair which

ISOPODA. 45

they overlie. The first joint is stout, a little longer than broad, the second and third
shorter and much narrower, together as long as the one-jointed nagellum, which
carries three apical sensory filaments.

The second antennas have a very short first joint, the second not long, the fifth
rather longer than the third and rather shorter than the fourth, the three-jointed
nagellum being as long as the peduncle's third joint. The very short apical joint is
tipped with a curved spine.

The upper lip appears to be rounded. The lower lip forms two broadly rounded
lobes.

The mandibles are without palp, with small tridentate cutting plate and narrow
accessory plate, close to which is a strong but not elongate molar with finely
denticulate crown.

The first maxilla? show only two plumose setae on the apex of the inner plate, and
nine not very elongate spines on that of the outer plate.

The second maxillae are remarkably short, with short comparatively broad plates,
the outermost tipped with two long setae, the middle one with four that are not so
long, and the innermost with five that are shorter and more spine-like.

The maxillipeds have the lobe of the second joint produced about to the end of the
fifth joint and armed on the inner margin with setae and with three or sometimes
only two hooks. The third joint is small, the fifth broadly oval, the sixth much
shorter, nearly as broad as long, the seventh almost tubercular ; the fifth and sixth
are' well fringed with setae on the inner margin. The epipod is cpiadrately oval in
the female, but in the male balloon-like, being very narrow at the base.

The first gnathopods are closely applied to the mouth. They have the fourth joint
somewhat cup-like, much broader than the third, the fifth joint longer but not so
wide, the sixth narrower than the fifth but subequal to it in length, and both of
these notable for the trifid setae along the inner margin. The middle branch of the
setae is the longest. The narrowed apical part of the sixth joint has curved setae on
the outer margin. The finger is short and conical, and tipped with a spine.

The second gnathopods display only five joints. The finger has its inner margin
denticulate and ends in a very small curved unguis, agreeing with the finger in the
third, fourth, and fifth peraeopods. The two preceding joints have the inner margin
denticulate or serrate. Of the four joints preceding the finger the second is not
longer than broad, the third is shorter than the first, and the first than the fourth.
Whether the first represents a coalescence of the first and second or of the second
and third joints, or whether the fourth may be a fusion of the true fifth and sixth
joints it is, perhaps, vain to speculate.

The three hinder peraeopods are almost exactly alike, the second joint in the female
decidedly longer than the sixth, but scarcely so in the male. The third joint is
longer than the fourth and the fourth than the fifth.

The first and second pleopods have two slender branches with long apical setae.

46 CEYLON PEARL OYSTER REPORT.

The others appear to be simply branchial, narrowly oval, without setae. A male
pleopod figured shows one of the branches apically divided for a short distance. The
smooth inner division is probably the male stilet in preparation.

The uropods are elongate, the narrowly triangular terminal division being about a
quarter as long as the peduncle. Within it is a plate about two-thirds as long and
half as broad, with a long seta on its rounded apex.

The female is brown, with numerous conspicuous white spots, and three dorsal
longitudinal dark bands. The males did not exhibit any white spots, and had two
dorso-lateral dark bands.

The length of the female was 6 millims., not including the second antennae which
were 275 millims. long. The longest male was 4 millims. From this the detail
figures of the male are taken. The male figured measured 2 '5 millims. in length.
The sex of the female specimen was beyond question, as it was provided with
fourteen large eggs. As to the other much smaller specimens one must speak with
more reserve, as they might be young ones of either sex.

Locality : — From pearl oysters, East Cheval Paar.

The specific name is given out of respect to the assiduous work which
Miss Harriet Richardson has devoted to this tribe of the Isopoda.

Family: IDOTEID^E.

Idotea, Fabricius.
1798, Idotea, Fabricius, ' Supplenientum Ent. Syst.,' p. 302.

Idotea sp.

In a tow-net gathering off Marichchukaddi there were two specimens of a young
Idotea, 2 millims. long, in which the last pair of peraeopods were not yet visible.

Family: ASTACILLIILE.

Astacilla, Cordiner.

1795, Astacilla, Cordiner, 'Remarkable Ruins . . . and Singular Subjects of Natural History.'

Section, " AstacillaB," etc.
1893, Astacilla, Stebbing, 'History of Crustacea,' p. 370.
1897, Astacilla, Sars, ' Crustacea of Norway,' vol. 2, p. 87.
1904, Astacilla, Norman, 'Ann. Nat. Hist.,' ser. 7, vol. 14, p. 447.

Astacilla amblyura, n. sp. — Plate XI. (B).

The head has a minute rostral point and the usual broad lateral lobes in advance of
the dark protuberant eyes ; it is rather gibbous between the eyes and apparently has
a pair of tubercles wide apart in advance of the hump. The specimen was rather foul
with adhesive extraneous matter, by which the excrescences were in part obscured, in
part exaggerated, and there was some risk, in clearing away what was adventitious,

ISOPODA. 47

of removing what really belonged to the animal. A strong groove (but not necessarily
an articulation) separates the first segment of the perseon from the head. Tbe long
fourth segment has a dorsal hump in advance of the middle and a small medio-dorsal
tubercle closely flanked by two others a little to the rear of it on the hind margin.
The sixth perseon segment appears to have two lateral tubercles, the fifth and seventh
segments each one such tubercle on either side. No transverse dorsal divisions of the
pleon could be discerned, but on each lateral margin there are three projections ; to
the last and most prominent of these a short obtusely ending apical triangle succeeds,
very different from the acute ending of the pleon in the four Norwegian species figured
by Professor G. 0. Sars, but agreeing with the apex in A. granulata, Sars, and
A. marionensis, Beddard.

The first antennae are normal, with stout first joint, the second and third
successively narrower, the one-jointed flagellum fringed with thirteen or fourteen
sensory filaments.

The second antennas are very little shorter than the body, the third joint much
stouter distally than at its base, the fourth joint the longest of all, but much narrower
than the third and curving to a little conspicuous tooth not far from the base ; the
fifth joint is narrower, intermediate in length between the third and fourth, about
twice as long as the three-jointed flagellum.

In the mouth parts and limbs there is scarcely any tangible difference from the
corresponding structures figured by Sars for A. lovgicomis (Sowerby).

On one of the mandibles, between the accessory plate and the strong molar, in this
species two or three short spines are to be seen crowded into a very narrow space.
The cutting edere has three or four close-set teeth. The first maxillae have three setae
on the inner plate, nine very short spines on the outer. The second maxillae have the
inner plate much broader and more setose than the other two plates. In the maxillipeds
the broad plate of the second joint is armed with a remarkably large coupling hook ;
the seventh joint is well developed, not unguiform.

The first gnathopods, of which the last four joints are by no means unlike the
corresponding four of the maxillipeds, have serrate spines on the broad fifth joint.
The seventh joint has one conspicuous spine among many that are smaller. The
second gnathopods and the first and second peraeopods are, as usual, alike, very
slender, with the seventh joint minute, this and the three preceding joints being
furnished with very long setae. The fourth, fifth and sixth joints are subequal in
length. The third peraeopods are a little longer than the fourth or fifth, all three
pairs being stoutly built, with the second and sixth joints longest, but none very
elongate or conspicuously armed, except the finger which has a stout apical tooth on
the inner margin in addition to the short curved nail. The finger is also tuberculate
on the inner margin, and this is perhaps the case with some of the preceding joints.

The pleopods agree very nearly with those which have been figured by Sars for the
male of A. longicomis and A. granulata, the first two pairs having slender rami, with

48 CEYLON PEARL OYSTER REPORT.

the setas of the apical border elongate, and the masculine appendix of the second pair
forming a narrow stilet which reaches to the end of the ramus and carries two long
apical setas.

The uropods are rounded above, widest below the middle, then rather rapidly
narrowed to the rami, of which the external is very small, acutely triangular, and
reaching to the apex of the pleon. The internal ramus is still smaller. Colour in
spirit a dull yellowish.

Length 9 millims., apart from the antennae, the lower of which are nearly as long
as the body, about 8 "5 millims.

Locality : — Periya Paar, Gulf of Manaar.

A specimen, 4 millims. long, delicately pink in colour, from East Cheval Paar, is no
doubt the young of this species. It has the first perason segment not marked off
from the head, the fourth segment very elongate and smooth. The flagellum of the
first antennae is armed about the apex with only three or four sensory filaments.
The fifth perasopods are still imperfectly articulated, very small, ending obtusely
without a nail.

Another specimen, from East Cheval Paar, is only 3 millims. long, with the fifth
perasopods in a still more inchoate condition. Here, however, the fourth segment of
the perason is not especially elongate and shows traces of median and terminal
tuberculation. The colour is a delicate pink. Neither in this nor the other juvenile
specimen is there a tooth on the fourth joint of the lower antennas.

The specific name is from the Greek d/i./3\i;s, blunt, and obpd, tail. Apart from
size and arrangement of tubercles, the distinguishing characters of this species depend
on the antennal tooth just mentioned, the decided groove between head and perason,
the solidarity of the pleon, and the extension of the uropods further back than
appears to be the case in any of the sj)ecies hitherto described.

Tribe: ASELLOTA.
Family: JANIRID/E.

Sars, when separating this family in 1897 from the Asellkke, incidentally men-
tioned Stenetrium as belonging to the latter. Miss Richardson, in 1902, without
comment transferred the genus to the newer division within the Asellota. It certainly
seems to conform in many important respects to the following definition which Sars
himself gives of the Janiridas : —

" General habitus that of the Asellidas, but the lateral parts of the cephalon always
laniellarly expanded. Eyes, when present, subdorsal. Superior antennas sometimes
well developed, with the flagellum multiarticulate, sometimes very small, with rudi-
mentary flagellum. Inferior antennas always longer than the superior, with the
peduncle six-articulate, and generally carrying a small accessory appendage (scale)
outside the third joint. Oral parts normal. Legs subequal in length, with the
dactylus generally bi- or tri-unguiculate ; first pair sometimes differing from the others

ISOPODA. 49

in being prehensile. First pair of uropoda [pleopoda] in female transformed into a
single, large, opercular plate ; in male, constituting the median piece of the compound
operculum, the lateral pieces of which are formed by the copulative appendages.
The three succeeding pairs very delicate, the last pair forming simple, smooth
Lamellae, the two preceding ones with the outer ramus narrow and confluent with the
basal part. Uropoda biramous, more or less developed."

From this characterization Miss Richardson's account varies in one or two respects,
assigning to the Asellidce and Janiridse in common a feature which Sars only
attributes to the former, namely, that the first antennae issue close together, which
cannot be predicated of all the Janiridse, and omitting a feature on which Sars lays
stress, namely, that the peduncle of the second antenna is six-jointed. In two species
of Stenetrium, the small fourth joint of this peduncle has been clearly observed, but
in three of the species, including the one first assigned to the genus, it is either not
present or has been overlooked.

Sars speaks of nine or ten genera as being included in the family, and since he
wrote, the genus Carpias, Richardson, 1902, has been added. In 1901, Dr. Ortmann
(' Proc. Ac. Philad.,' p. 157) introduced the new generic name Tole to take the place
of "Janthc, Bovallius," on the ground of preoccupation. In this he is followed by
Miss Richardson in 1905. But the genus which Bovallius instituted in 1881 is
Ianthe, not Janthc, so that no change is required. Moreover, in his key to the
species of Tole, the first species which Dr. Ortmann mentions is "J. bovctllii
(Studer)," which was named Ianthopsis bovallii by Beddard in 1886 (' "Challenger"
Isopoda,' vol. 17, pt. 48, p. 15, pi. 5, fig. 5). Seeing that Dr. Ortmann expressly
adopts the type species of Ianthe as the type of Tole, that name must lapse as a
synonym, and could not properly be revived in case either Tole libbeyi, Ortmann,
or Tole holmesi, Richardson, should in future be transferred from Ianthe to a distinct

genus.

Janira, Leach.

1814, Janira, Leach, ' Edinburgh Encyel.,' vol. 7, p. 434.

1886, Janira, Beddard, ' " Challenger," Isopoda, Reports,' vol. 17, part 48, p. 5.

1897, Janira, Sars, ' Crustacea of Norway,' vol. 2, p. 98.

1898, Janira, A. 0. Walker, 'Trans. Biol. Soc. Liverpool,' vol. 12, p. 280.

The species about to be described agrees in many respects with this genus as
defined by Sars, but the fifth joint or wrist of the first gnathopods is not sub-
fusiform ; it is distally expanded, so as to form a kind of bidentate palm. The single
specimen in the collection appears to be an adult male. It was, however, so exceed-
ingly small and defective, having lost most parts of the second antennae, all the last
three pairs of perseopods, and the uropods, that it was not well suited for initiating a
new genus. The parallel-sided perseon agrees with what is found in Jceropsis,
Stenetrium, and Iais. Hasw ell's Stenetrium inerme may be congeneric, but the
great difference in size makes specific agreemeivt very unlikely.

h

50 CEYLON PEARL OYSTER REPORT.

Janira (?) nana, n. sp. — Plate III. (C).

The general appearance agrees with that of Jaeropsis curvicomis, in company with
which the specimen was taken, but the segments of the peraeon are not so markedly
separated. The pleon is nearly circular, not serrate, with a small apical convexity, on
either side of which the uropods probably protrude.

The eyes are pale orange coloured, differing from those of Iais pubescens by having
not two but thirty-eight components.

The first antennae have a stout basal joint, the second much smaller, and the third
almost like a flagellar joint ; the flagellum is longer than the peduncle, its seven
joints unequal in length, all slender, the last tipped with a couple of long setae or
filaments.

The upper lip is rounded, and seemed to be projected forward with the lower lip.
The mandibles have a prominent cutting-plate divided into five teeth, the accessory
plate on the left similarly divided, but bidentate on the right. The spine-row has
five or six spines. The molar is prominent, denticulate. The three joints of the palp
are subequal in length. The first maxillae have a slender inner plate, and nine,
mostly denticulate, spines on the outer. The outer and middle plates of the second
maxillae carry each three setae. In the maxillipeds the plate of the second joint is
rather large, with several plumose setae on the distal margin, and one hook on the
inner margin ; the third, fourth, and fifth joints are broad, the fifth having its outer
margin longer than the inner ; the sixth and seventh joints are narrow.

The first gnathopods have the fifth joint much broader but not longer than the
sixth, which in closing down would reach much beyond the palmar margin. The
finger is much the same in all the known limbs of the peraeon, having a short trunk
with two distinct nails. The second gnathopods agree with the first and second
peraeopods in structure, but are longer, especially in the second, fifth, and sixth
joints ; the sixth joint is narrower than the preceding joint.

The male operculum is composed of the first two pairs of pleopods. The first pair
are narrow and more or less tapering, but with a constriction below the middle.
They end in two pairs of overlapping shortly lanceolate lobes. The second pair are
semicircular, with a long sinuous almost filiform masculine appendix.

Length 1*5 millims. Haswell's Stenetrium inerme is described as -^g-ths of an inch
in length.

Locality : — Gulf of Manaar.

Jaeropsis, Koehler.

1885, Jaeropsis, Koehler, 'Ann. Sci. Nat.,' se>. 6, Zool., vol. 19, Art. 1, p. 2.

1886, Jaeropsis, Beddard, ' " Challenger," Isopoda, Reports,' vol. 17, p. 20.
1891, Jaeropsis, Chilton, 'Trans. New Zealand Inst.,' vol. 24, p. 267.
1893, Jaeropsis, Steebing, 'History of Crustacea,' p. 379.

1899, Jaeropsis, H. Richardson, 'Proc. U.S. Mus.,' vol. 21, p. 857.

ISOPODA. 51

1899, Jaeropsis, Norman, 'Ann. Nat. Hist.,' ser. 7, vol. 4, p. 291.

1900, Jaeropsis, H. RICHARDSON, ' Amer. Naturalist,' vol. 34, No. 400, p. 298.
1902, Jaeropsis, H. Richardson, 'Trans. Connect. Acad. Sci.,' vol. 11, p. 298.

By the addition, which is well justified, of Jcera curvicornis, Nicolet, Miss
Richardson is able to say iu 1902 that " six species of this genus have been hereto-
fore described." They are J. curvicornis (Nicolet) ; ./. brevicornis, Koehler ;
./. marionis, Beddard ; J. neo-zelanica, Chilton ; J. lobata, Richardson ; J. dolljusi,
Norman ; to which on the same occasion Miss Richardson adds J. rathburue.
Jcera antarctica, Pfeffer, may perhaps belong to the group, but the description and
figures leave its generic location quite uncertain. All the species have many features
in common. They range in size from 2 millims. to a little over 4 millims. The sides
of the middle body or perseon are nearly or quite parallel, with the segments very
distinctly separated. Both pairs of antennae are short. The appendages of the
peraeon are truly isopodous, without any real distinction between gnathopods and
peraeopods. The uropods are small, carrying two minute dissimilar rami, and
occupying emarginations in the distal border of the caudal shield. In the second
antennae the joint numbered second by Koehler, third by Chilton, Beddard, and
Richardson, fourth by Canon Nohman, is broadly expanded, unless J. marionis be
an exception, for in that species the joint is figured as cylindrical rather than
laminar.

It may perhaps be objected that there is a want of authority for the statement that
the uropods occupy emarginations in the telsonic segment. Miss Richardson indeed
savs that her species, J. lobata, differs from Koehler's " in the shape of the
terminal segment, which is perfectly rounded in J. brevicornis," while in ./. lobata
" there are two posterior incisions for the reception of the uropods." But one may
easily press too far the differences shown in the habitus figures of very minute
animals. When the highly magnified figure of the uropod of J. brevicornis is
considered, it will be noticed that the outer margin is serrate, and this makes it
probable that here as in other species it has its share in completing the curve of the
tail-piece.

Jaeropsis curvicornis (Nicolet) — -Plate XI. (C).

1819, Jaera curvicornis, Nicolet, in Gay's ' Hist. fis. y pol. de Chile,' Zool., vol. 3, p. 263, pi. 3, fig. 10.

1891, Jaeropsis neo-zelanica, Chilton, ' Trans. New Zealand Instit.,' p. 267.

1902, Jaeropsis curvicornis, H. Richardson, 'Trans. Connect. Acad. Sci.,' vol. 11, p. 298.

The body, as described by Nicolet, forms a rounded longitudinal medio-dorsal
elevation. This is not particularly easy to see, but, when the specimen is placed
back downwards, its rolling from side to side is evidence of the shape in question.
The head corresponds with Chilton's description as being " produced slightly into a
rostrum between the bases of the antennae ; end of rostrum emarginate, and with a

H 2

52 CEYLON PEARL OYSTER REPORT.

rounded lobe fitting into the emargination." Norman says ot the head in J. dollfusi
" the anterior margin is emarginate, and in front of this the buccal organs are
conspicuously projected," but his figure also shows the rounded lobe in the emargina-
tion. The figure of J. dollfusi shows the pleon more sharply contracted towards the
apex and the sides more deeply serrate than is the case with the present species.

The eyes are not large, not dark, placed near the front angles of the head. The
first antennse have a broad basal joint, seemingly denticulate at the front corners.
The second joint is much shorter and much narrower, the third smaller than the
second, and the two remaining joints very insignificant, but tipped with two long
filaments.

The second antennae are of the typical form, seemingly with three short basal
joints, followed by the characteristic large dilated joint with thin outer margin
slightly crenulate, not strongly as in J. dollfusi; to this succeeds a much smaller,
apically expanded joint, helping to form a double geniculation. The five remaining
joints, perhaps, constitute the flagellum, but the first, which is very far the largest of
them, has usually been accounted the terminal joint of the peduncle. It is, however,
not very usual for the penultimate joint of the peduncle to be shorter than the joint
preceding as well as the joint following it.

The upper lip has a rounded distal margin. The mandibles have the cutting edge
cut into five teeth, eight spines in the spine-row, the three-jointed palp very small.
The first maxillae have three short spine-like seta? on the inner plate, and nine to
eleven spines, mostly denticulate, on the outer plate. The second maxilla? are
notable for the shortness of the inner plate ; each of the three plates carries four
apical seta?. The maxillipeds have a very large second joint with extremely broad
plate, the distal margin slightly and irregularly crenulate with a gentle curvature or
sinuosity, distinct from the quadrate character displayed in J. dollfusi. The coupling
hooks are two. The fourth joint is distally produced on the inner margin, the sixth
joint is very narrow and the seventh minute.

The limbs of the peraeon differ but little, the fourth joint being shorter than the
third, fifth, or sixth. The finger has two conspicuous nails and one that is
inconspicuous.

The operculum of the pleopods in the female is broadly rounded for nearly two-
thirds of its length, and then contracts to a narrowly truncate apex carrying four
setules. It does not show marks of a longitudinal or a transverse suture, such as are
said by Dr. Chilton to be indistinctly visible in his specimen.

The uropods fill the emarginations of the pleon. The peduncle is more strongly
serrate on the inner than on the outer margin. The inner ramus is hook-like ; the
outer, which is even smaller than the inner, carries a bunch of setae.

Length 2 millims. Nicolet gives 2 lines for the length of his specimen ; Chilton
about 2 '5 millims. for his.
Locality : — Gulf of Manaar.

ISOPODA. 53

Family: STENETRIID.E *
Stenetrium, Haswell.

1881, Stenetrium, Haswell, 'Pr. Linn. Soc. N.S. Wales,' vol. 5, p. 478.

1882, Stenetrium, Haswell, 'Cat. of Australian (Malacostracan) Crustacea,' p. 308.

1884, Stenetrium, Chilton, 'Trans. N. Zealand Instit.,' vol. 1G, p. 251.

1885, Stenetrium, Haswell, 'Pr. Linn. Soc. N.S. Wales,' vol. 9, p. 1009.

1886, Stenetrium, Beddakd, ' " Challenger " Isopoda, Reports,' vol. 17, pt. 48, p. 8.
1895, Stenetrium, Hansen, ' Isopoden der Plankton-Exp.,' p. 6.

1902, Stenetrium, H. Richardson, 'Trans. Connect. Ac. Sci.,' vol. 11, p. 295.
1905, Stenetrium, Hansen, 'Proc. Zool. Soc. London,' pp. 303, 316.

Five species have been assigned to this genus, S. armatum, Haswell, S. inerme,
Haswell, S. fractum, Chilton, S. haswelli, Beddard, and S. stebbingi, Richardson.
But Haswell's S. inerme differs from his other species in having rounded lateral
eyes, the antepenultimate joint of the maxillipeds distally narrowed, and perhaps, also
by having the rostrum subacute. It appears to belong to the genus Notasellus,
Pfeffer, 1887. The union of the other four species in a single genus is probably
justifiable, though in each case some important evidence is wanting. For S. armatum
Haswell has twice figured the mandible, and on each occasion gives no indication
of its possessing a molar. In S. haswelli, and in the species about to be described
from Ceylon, this part of the mandible is strongly developed and too conspicuous to
be overlooked. In the descriptions of S. fractum and S. stebbingi the presence or
absence of this structure, is not discussed. For the last-mentioned species no account
is given of the pleopods, and for the other species the accounts of these organs are
variable or uncertain. Including the new species, which is nearly allied to what is
known of S. fractum, the genus may be defined as follows : —

Body depressed, parallel-sided. Pleon consolidated. Head bluntly rostrate. Eyes
obliquely dorsal. First antennae short, inserted close to the rostrum on either side
of it. Second antennae elongate, with exopod on the third joint. Mandible with
palp. Maxillipeds with third to fifth joints broad, sixth and seventh narrow. First
gnathopods simply or complexly subchelate. Second gnathopods and all the
peraeopods slender, ambulatory, biunguiculate. Pleopods not in every case biramose.
Uropods biramose, not adjacent, inserted apically on the telsonic segment.

The uncertainty attending the characters in some of the species makes it difficult to

* After the manuscript of this paper had passed out of my hands, I received the luminous essay, " On
the Morphology and Classification of the Asellota-group of Crustaceans, with Descriptions of the Genus
SU a' triii in, IIasw., and its Species," by Dr. H. J. Hansen ('Proc. Zool. Soc. London,' p. 302, April 18,
1905). In this the new family Stenetriidse is defined (Joe. cit., p. 315), and nine species of Stenetrium are
described, with illustrative figures of several and a conspectus of them all. Five are new, S. rneditermneum,
S. serniliiin, S. occidmtale, S. antillense, S. siamense. For a more accurate account of the pleopods than I
had myself arrived at I am now indebted to Hansen's instructive treatise.

54 CEYLON PEAEL OYSTER REPORT.

produce a useful synoptic table. The following is offered, therefore, with all necessary-
reserves : — -

Telsonic segment without lateral notch. 1. S. haswelli, Beddard.

Telsonic segment with lateral notch — 2.

2-

First gnathopod with hind margin of wrist

produced. 2. S. stebbingi, Richardson.

First gnathopod with hind margin of wrist not
produced — 3.

„J First antenna, second joint as long as first. 3. S. fractum, Chilton.

L First antenna, second joint shorter than first — 4.

I Margin of head convex between rostrum and antero-
. j lateral angles. 4. S. armatum, Haswell.

I Margin of head angular between rostrum and
[_ antero-lateral angles. 5. S. chiltoni, n. sp.

In the adult male the first gnathopods strikingly distinguish S. haswelli,
S. stebbingi, S. armatum. But in the female of the last-named species these
gnathopods do not appear to differ from those of S. fractum, and from those of
S. chiltoni only by the greater robustness of the hands.

Stenetrium chiltoni, n. sp. — Plate XII. (A).

The whole body, dorsally and at the sides, is beset with rather long stiff setae.
The head has a rather broad, blunt rostrum not reaching so far forward as the broad
epistome. The antero-lateral angles of the head are acutely incurved, and between
each of these and the rostrum the margin is produced to a point, thus forming sockets
for the first antennae. The segments of the peraeon differ but little in length or
breadth ; the sides are nearly straight, with the anterior angles of the first four
pointing acutely forwards. The telsonic segment has the lateral margins regularly
but quite microscopically serrate, and, as in all the species except S. haswelli, each of
these margins far down is produced into a tooth. The indentation or pocket thus
formed is followed lower down by a small setiferous indent, to which succeed the
rounded corners of the broad apical margin, with its shallow convex projection between
the uropods.

The eyes are narrowly bean-shaped, placed obliquely near the middle of the convex
lateral ridges that run below the anterior lobes of the head.

The first antennae agree closely with those of S. armatum, the second joint being
much shorter than the first and not so long as the third ; the flagellum is obscurely
six-jointed. The second antennae have the first joint acutely produced on the outer
side, but the short second joint and the longer third are not produced on either side.
The exopod is conical, having its truncate point tipped with a pencil of setae.
Haswell has apparently overlooked the second joint and described the third as

ISOPODA. 55

produced like the first externally and distally into a slender acute process, the process
ending in a hair-like appendage. He does not mention the articulated scale or
exopod. For S. fractum Chilton describes and figures the third joint as "produced
acutely at its antero-distal angle, bearing on the outer edge an articulated appendage,
which has the end rounded and supplied with a few long setse." The small fourth
joint, which is seen in the present species and in S. hasioelli, is not noticed or figured
in connexion with the other three species, as noted in the discussion of the family.
The fifth and sixth joints of the peduncle are elongate, the sixth slightly longer than
the fifth, the flagellum three times as long as the sixth joint, rather longer than the
whole peduncle, composed of very many little scarcely separated joints, setose.

The upper lip is apically rounded. The two broad lobes of the lower lip have the
usual armature of minute spines.

Left mandible with dentate accessory plate like the cutting edge but smaller, spine-
row of five serrate spines ; right mandible without accessory plate, unless it be
represented by the first of the five spines in the spine-row ; cutting edge with four or
five crowded teeth, but within the mandible the new teeth in preparation for the
moult are spread out in one plane ; molar long and prominent ; palp of three long
joints, the second carrying five short spines between two long ones, the third falciform,
with long spines at apex, short ones fringing the margin.

First maxilla with three spines, a little tooth and some setules on apex of inner
plate, and nine more or less denticulate spines on apex of outer plate.

Second maxilla with about four slender spines on apex of outer plate, and also on
that of the middle one, the rather broader and more oval inner plate carrying several
spines along the inner margin.

Maxillipeds with large distally narrowed epipods reaching nearly to the apical
border of the broad lobes which surmount the second joint and considerably overtop
the fourth joint ; the third joint is short but broad, the fourth larger than the fifth,
both of them broad and widened distally, the sixth and seventh being abruptly much
narrower.

The first gnathopods have the second joint moderately long, the three following
joints short, the fourth subacutely produced on the front margin ; the fifth joint is
setose on the hind margin ; the sixth joint is less than twice as long as its greatest
breadth ; the front margin is curved and carries a few setules, the hind margin
straight, furnished with many setse ; the palm, defined by a long spine, carries several
smaller pectinate spines sloping towards this palmar spine ; the finger, which curves
over the palm and ends in a small simple nail (broken in the specimen), has a few
setules on the convex margin and several microscopical spines on the concave border.
Chilton speaks of the palm of his species as " armed with strong serrated setse," and
the finger as having the "inner edge thickly fringed with strong denticulated setoe," but
these expressions may refer to the armature as it appears when very highly magnified.

The second gnathopods appear to have a round-lobed first joint ; the second joint is

56 CEYLON PEARL OYSTER REPORT.

about as long as the third and fourth combined, the third being much longer than
the fourth, nearly as long as the sixth, which is slightly shorter than the fifth. The
finger is less than half as long as the sixth joint ; it curves to a sharp apical point,
which is overhung by an unguisdike sjjine, while on the concave margin of the finger
there is a small spine. The peraeopods differ little in character from the second
gnathopods, except that the second joint is less elongated and the third joint is more
nearly subequal in length to the sixth. To the first and second gnathopods and first
and second perreopods in one of the specimens four pairs of marsupial plates were
attached, the third pair being the largest, but the fourth also of considerable size.
In dissection of the pleon there came away a linear ring, which, perhajjs, represents a
degraded first pleon segment. Dorsally two such segments are indicated.* A small
unpaired plate, square above and triangular below, without any trace of longitudinal
or other suture, must be regarded as representing the first pair of pleopods.f The
second pleopods are wanting, as in other females of this tribe. The third pleopods
form a very large pair of biramose appendages, the peduncle small, the inner ramus
branchial, with three or four seta? on the narrow apex, the outer ramus of great size,
with slightly oblique transverse suture below the middle, but starting just above the
apex of the inner ramus. The fourth pair are biramose, and have the oval inner
branchial ramus much broader and not shorter than the outer ramus, which shows a
transverse suture above the middle and has the tapering lower division fringed with
several long setae. In the fifth pair each pleopod consists of a single branchial ramus,
possibly representing a coalescence of two rami, the outer margin raised and distally
fringed with setae.j

The uropods are inserted a little within the distal margin of the telsonic segment,
separated by the convexity which may he considered an equivalent of the actual
telson. The peduncle is rather stout, shorter than the rami, of which the inner is the
larger, both being well furnished with tufts of long setae on sides and apex.

Description of the uropods and the complete second antennae is based on a specimen
of the same dimensions as the one figured, but which did not come to light till after
the less complete example had been figured. This second specimen was straight, but
a third, rather smaller specimen with it had a distortion similar to that shown in the
plate.

Length 4'5 millims., breadth 1'5 millims.

Locality : — Reef, Galle, with Ascidians ; and Coral banks, Gulf of Manaar.

* For the genus at large Hansen says, " two rudimentary segments are observed in front of the large
abdominal shield" (loc. cit., p. 304).

t Haswell (loc. cit., p. 1010) says: "The bases of the first pair of abdominal appendages are
covered in both cases by a broad plate, with a bifid apex attached to the posterior border of the last
thoracic segment." By "both cases" no doubt the two sexes are intended, and "the first pair of
abdominal appendages " are really the third pair of pleopods.

X Hansen in his character of the family says in regard to the pleopods, "fifth pair with only one
ramus, in all probability the exopod " (loc. cit., p. 315).

ISOPODA. 57

The specific name is given out of respect to my friend Dr. Charles Chilton,
whose Stenetrium fractum has a name only too suggestive of the mishaps to which
these delicate isopods are liable. Haswell's species is described as half-an-inch long,
Chilton's as about a sixth of an inch. Though it remains a little doubtful whether
the species here described belongs to Haswell's genus, the possibility is also open
that S. armatum, S. fractum, and S. chiltoni may all be the same species.

Family : MUNNID/E.
Pleurocope, A. 0. Walker.

1901, Pleurocope, Walker, ' Journ. Linn. Soc.,' London, vol. 28, p. 297.
Mr. Walker remarks that " this genus differs from Pleurogonium, its nearest ally,
in the large size and peculiar appendages of the head, the different relative
proportion and structure of the antenna?, in the form of the caudal segment, and in
the position and size of the uropods, which are unusually large for the family." It
may, however, be observed that in the genus Dendrotion, Sars, the uropods are
larger and more conspicuous than in the present genus.

Pleurocope dasyura, Walker.

1901, Pleurocope dasyura, Walker, 'Journ. Linn. Soc.,' vol. 28, p. 297, pi. 27, figs. 12 to 18.

The description by Mr. A. 0. Walker, and the excellent figures by Mr. Andrew
Scott which accompany it, place the identification of this species beyond doubt.
Beyond verification I have nothing to add, except that the perreon displayed four
stiff upstanding dorsal setse. A point of interest would have been to ascertain the
character of the mandibles. But at the very moment when I was arranging the
specimen for dissection, it disappeared like a dream, and defied all the efforts made for
its re-discovery.

The length was a little over 1 millim., therefore approximately the same as
Mr. Walker's type specimen from the Mediterranean. It came into my hands
already named by Mr. A. Scott.

Locality : — Gulf of Manaar.

Tribe: ONISCLDEA.

Family: LIGIID^E.

Ligia, Fabricius.

1798, Ligia, Fabricius, ' Supplementum Ent. Syst.,' p. 301.
1885, Ligia, Budde Lund, 'Isopoda Terrestria,' p. 258.

Ligia exotica, Roux.

1828, Ligia exotica, Eoux, 'Crust. MeVlit.,' livr. 3, pi. 18, f. 9.
1885, Ligia exotica, Budde Lund, 'Isopoda Terrestria,' p. 267.

A mutilated specimen occurs in the collection, which appears with little doubt to
belong to this widely distributed species.

Locality : — Station XXXIX., Gallehogalle Bank, 16 to 30 fathoms

58

CEYLON PEARL OYSTER REPORT.
INDEX.

^Ega

JEgidsa ....
amblyura (Astacilla)

Amesopodidse . .
Amesopous . . .
Anilocra ....
Anthuriche . . .
Argathona . . .
Argathonidse
Asellota ....
Astacilla ....
Astacillidoe . . .

beddardi (Cilicsea) . . .
bicarinata (Cymodoce) .

Calathura

callipia (Rhiothra) . . .

Chelifera

ekiltoni (Stenetrium) . .

Cilicrea

Cirolana

Conilorpheus ....

Corallanidse

crassicornis (Heterotanais)
curvicornis (Jseropsis).

Cymodoce

Cymothoidce

age
20
20

46, PL XL (B).
44
44
25
8
17
16
48
46
46

40, PL X. (A), (B).
42, PL X. (C).

8
26, PL VI. (A).
2

54, PL XII. (A).

33

11

13

19

4, PL I. (A).
51, PL XL (C).
42
25

dasyura (Pleurocope) .
dimidiata (Anilocra) .

Eurydicidoe . .
exotica (Ligia) .

Flabellifera

gardineri (Lanocira)
Gnathia ....
GnathiicLe . . .
gracilis (Tanais)

Han.senolana

herdmani (Conilorphcus)
Heterotanais

Idotea

Idoteida;

inornata (Cymodoce) .

57
26

10
57

8

19

3, PL I. (D).

15

13, PL II. (A).
3

46
46
43

insolita (Gnathia)
Iron a ....

Page
9, PL XII. (B).

27

Jseropsis
Janira .
Janiridse

Lanocira

latreillii (Ciliesea) . . .

Leptochelia

lifuensis (Leptochelia) .

Ligia

Ligiidas

mirabilis (Leptochelia) .
Munnidse

nana (Janira V) . . . .
nanoides (Irona) . . .
normani (Argathona) .

ommatophylax (iEga) . .

Oniscidea

orientalis (Rocinela) . .

parva (Cirolana) . . .
Pleurocope

Rhiothra

richardsome (Amesopous)
Rocinela

Sphseroma

Sphreroniidse

sphasromiformis (Hanseno-

lana)

Stenetriidas

Stenetrium

sulcaticauda (Cirolana) .

Tanaidse

Tanais .

50
49
48

19

36, Pis. III. (B), VIII.

5

7, PL I. (C).
57
57

6, PL I. (B).
57

50, PL III. (C).
28, PL VI. (B).
17, PL III. (A).

21, Pis. IV., V. (A).

57

24, PL VI. (C).

12

57

26

44, PL XL (A).

23

31

29

15, PL II. (B).

53

53

11

2
2

Valvifera 43

walkeri (Sphseroma) .

whiteleggei (Cilica-a) .

zeylanica (Lanocira) .

31, PL VII.

39, PL IX. (A), (B).

19, PL V. (B).

ISOPODA. 59

EXPLANATION OF PLATES.
PLATE I.

A. Heierolanais crassicomis, n. sp. — n.s., natural size of specimen figured in lateral and in dorsal view;

a.s., a.i., upper and lower antennae more highly magnified ; gn. 1, gn. 2, first and second gnathopods ;
prp. 1, prp. 3, first and third peraeopods; urp., uropod.

All the details are to the same scale, except the separate thumb and finger of gn. 1, and the
separate outer branch of the uropod, which are more magnified than the other parts.

B. Leptochclia mirabilis, n. sp. — A.n.s., natural size of specimen figured in dorso-lateral view, with first

gnathopod of the right side supplied from fragments; gn. 1, part of first gnathopod of the left side,
probably belonging to the above specimen and drawn to the same scale; a.s., a.i., third joint of
peduncle and the flagellum of upper antenna, and the lower antenna — these and the following details
on a higher scale of magnification ; gn. 2, prp. 5, second gnathopod and part of fifth peraeopod ; pip.,
urp., one of the pleopods and a uropod ; B.n.s., natural size of specimen figured in dorsal view,
showing lower antennae and base of right upper antenna ; uropods broken.

C. Leptochclia lifuensis. — a.s. ? , a.i. $ , upper and lower antennae of female; gn. 1, ? , gn. 2, $ , first and

second gnathopods of female; urp. ?, uropod of female; a.s. J1, a.i. <$ , upper and lower antennae
of male; gn. 1, g , first gnathopod of male; urp., uropod of male.

All the above are magnified to the same scale as the general details in Plate I., A, except the
separate ramus of the male uropod, which is magnified on the same scale as the corresponding
ramus in Plate I., A.

D. Tanais gracilis, Heller. — n.s., natural size of specimen figured in lateral and dorsal views ; a.s., a.i.,

upper and lower antennae very highly magnified ; gn. 1, gn. 2, pip. 4, first and second gnathopods
and fourth peraeopod ; I'L, urp., dorsal view of pleon and uropods, to the same scale as the pre-
ceding details ; in., mxp., mandible and maxillipeds, exopod of the latter detached and incomplete.
These and the separate portions of the second gnathopod and fourth peraeopod are more magnified
than the other details.

PLATE II.

A. Conilorpheus herdmani, n. gen. et sp. — n.s., lines indicating natural size of specimen figured below in

dorsal and lateral view; C, dorsal view of the head; Pa: s. 7, seventh segment of peraeon in dorsal
view, and in lateral view with the fifth peraeopod and pleopods showing below; PL, pleon in dorsal
view; a.s., a.i., first and second antennae ; m., m., mx. 1, mx. 2, mxp., the two mandibles, first and
second maxillae, and maxillipeds, the mandible on the right figured from the inner side; gn. 1,
prp. 4, pip. 5, pip. 1, pip. 2, urp., first gnathopod, fourth and fifth peraeopods, first and second
pleopods, and uropods, the last in ventral view.

The mouth organs are magnified on a higher scale than the other appendages.

B. Hansenolana sphceromiformis (Hansen). — n.s., lines indicating natural size of specimen figured in dorsal

view; PL, pleon in dorsal view; a.s., a.i., first and second antennae ; m., mx. 1, mx. 2, mxp., mandible,
first and second maxillae, maxillipeds; gn. 1, gn. 2, prp. 5, pip. 1, pip. 2, pip. i, urp., first and second
gnathopods, part of fifth peraeopod, first, second, and fourth pleopods, and uropods. Below the full
figure of gn. 1 from the outside is given a more enlarged figure of the other member of the pair
from the inner side. The portion of this gn. 1 and the portion of prp. 5 are enlarged on the same
scale as the mouth organs.

I 2

60 CEYLON PEARL OYSTER REPORT.

PLATE III.

A. Argalhona narmani, n. gen. et sp. — n.s., lines indicating natural size of the specimen represented by tho

adjoining figures in dorsal and lateral views ; PL, pleon in dorsal view, more highly magnified ;
a.s., a.L, the first and second antennae ; l.s., the upper lip, with the epistome surmounted by the
frontal lamina; /.?'., the lower lip; m., in., the left mandible entire, and part of the right mandible ;
nt.r. 1, mx. 2, mxp., the first and second maxillae and the maxillipeds ; (in. 1, gn. 2, prp. 5, the first
gnathopod without side-plate, the second gnathopod and fifth perreopod each with its side-plate ;
pip. 2, the second pleopod.

The mouth parts are magnified on a higher scale than the other appendages, but the uropods
figured in attachment to the pleon are enlarged on a lower scale than the rest.

B. Cilicrca latreillii, Leach, juv. — n.s., lines indicating natural size of specimen figured below in dorsal and

dorso-lateral aspects; PL, pleon in dorsal view; a.s., a.L, first and second antenn;e ; l.s., upper lip
with epistome; map., maxillipeds; gn. 1, first gnathopod.

The mouth parts are more highly magnified than the other appendages.

C. Janira (1) nana, n. sp. — n.s., line indicating natural size of the specimen ; PL, pleon, without appendages ;

a.s., a.L, first antenna and four basal joints of second; l.s., Li., upper and lower lips; m., in., mx. 1,
mx. 2, mxp., the mandibles, first and second maxilla?, and a maxilliped ; gn. 1, gn. 2, pip. 1, first and
second gnathopods and first persaopod, the finger of the first gnathopod and that of the first
perseopod more enlarged; pip. 1, pip. 2, the first and (one of the) second pleopods, the two pairs
together forming the male operculum.

The mouth parts are more highly magnified than the other appendages, being on the same scale
as the more enlarged finger of the first gnathopod.

PLATE IV.

JEga omtmatophylax, n. sp., £ . — n.s., lines indicating natural size of the specimen figured in dorsal aspect ;
ft, the head in dorsal aspect; C.L., lateral view of the head in conjunction with the first two
segments of the peraaon ; Per.s. 1., anterior part of first segment of perason in dorsal view;
g.p., ventral plate of the seventh perreon segment, with the genital papillae also more highly
magnified ; PL, Pleon in ventral view, after removal of the pleopods ; a.s., a.L, first and second
antennas ; m., mx. 1, mx. 2, mx. 2, mxp., mandible, first maxilla, second maxilla in two positions,
maxillipeds in ventral aspect. These organs are magnified on a higher scale than the other details
in general, and the distal parts of mandible, first maxilla, and maxillipeds are again more highly
magnified. In the mandible the third joint of the pjalp is missing. The further enlargement of
the maxillipeds is from the dorsal aspect, gn. 1, gn. 2, gn. 2, prp. 3, prp. 5, pip. 1, 2, 5, urp., the
first and second gnathopods, third and fifth perseopods, first, second and fifth pleopods, and the
uropod. The portions of these appendages which required further enlargement are on the same
scale as the principal figures of the mouth organs. Both members of the second pair of gnathopods
are figured, to show the difference mentioned in the text.

PLATE V.

A. dEga ommalophylax, n. sp., $ (*!). — n.s., lines indicating natural size of specimen figured above in dorsal
view ; C, ventral view of the head ; l.s., upper lip surmounted by the epistome and frontal lamina ;
m., m., the two mandibles, with higher magnification of a seta from second joint of palp, and of

ISOPODA. 61

apical portion of the trunk. In both mandibles a rounded lobe is shown below the apical margin,
but this lobe was only indefinitely made out. mx. 1, mx. 2, map., first and second maxillae, with the
apices more highly magnified, and the maxillipeds.
B. Lanocira zeyh/nica, n. sp. — n.s., lines indicating natural size of specimen figured above in dorsal view;
PL, dorsal view of pleon more highly magnified; a.S., a.i., first and second antennae; l.s., upper
lip; m., ?«., mx. 1, mx. 1, mx. 2, mxp. the mandibles, first maxilla?, one of the second maxillae, and
the maxillipeds; <jh. 1, prp. 5, first gnathopod and fifth perajopod ; pip. I, pip. 2, first and second
pleopods.

The mouth organs are more highly magnified than the other appendages.

PLATE VI.

A. Rhiothra cdllipia, Sciiiodte and Meinert. — n.s., lines indicating natural size of male specimen figured

above in dorsal view; ft, the head, stripped of its appendages, in dorsal view; a.s., a.i., the first
and second antenna1, with the terminal portion of each more highly magnified ; m., a mandible in
connexion with the epistome ; mx. 1, mx. 2, mxp., the first and second maxilhe and the maxillipeds;
gn. 1, pp. 5, the first gnathopod and the fifth peraeopod ; pip. 2, the second pleopod, a more
highly magnified portion showing the numerous coupling-spines of the peduncle and the male
appendix of the inner branch, urp., one of the uropods.

The mouth organs are more highly magnified than the full figures of the other appendages.

B. Irona nanoides, n. sp. — n.s., lines indicating natural size of the female specimen figured at the centre

in dorsal view; ft, the head, stripped of its appendages, in dorsal view; a.s., a.i., the first and
second antennae; m., the mandible in connexion with the upper lip; mx. 1, mxp., the first maxilla
and a maxilliped ; gn. 1, prp. 5, first gnathopod and fifth pera^opod ; pip., urp., a pleopod and one
of the uropods.

The mouth organs are more highly magnified than the other parts.

C. Rocinela oriental!*, Schiudte and Meinert. — n.s., lines indicating natural size of specimen figured at

the centre in dorsal view; n.s. $ , lines indicating natural size of a full-grown male specimen, from
which the figures marked $ are taken; a.s., a.s. $ , first antenna of each specimen; mxp., mxp. <$ ,
one maxilliped of the smaller specimen and both maxillipeds of the larger; gn. 1, gn. 2, prp. 5, first
and second gnathopods and fifth peraeopod from the smaller specimen; pip. 2, second pleopod of
the full-grown male; PI. urp., dorsal view of the pleon of the smaller specimen, much of the right
side omitted for want of space; PL <J, urp., telsonic segment and left uropod of the full-grown
male in dorsal view. The unsymmetrical right margin of the segment is seen through the figure
of the transparent pleopod placed above it for convenience.

The maxillipeds are more highly magnified than the other parts.

PLATE VII.

Sphvroma walked, n. sp. — n.s., n.s., curved line indicating natural size of partially rolled specimen figured
above in lateral view, crossed lines showing length and breadth of the same specimen unrolled and
figured below in dorsal view; a.s., a.i., first and second antennae; l.s., epistome and upper lip;
in., m., the mandibles, the palp of one separately figured on the right to display relative length of
the first joint; mx. 1, mx. 2, mxp., mxp., first and second maxilhe, and one of the maxillipeds
figured from the outer and the inner surface; gn. 1, gn. 2, the first and second gnathopods, with a
more enlarged figure of the terminal part of the first; pp. 1, 4, 5, first, fourth and fifth perneopods;
g.sp., gastric spines, more highly magnified than the other details, among which the mouth organs
are on a higher scale than the antennae and limbs.

(52 CEYLON PEAEL OYSTER REPORT.

PLATE VIII.

Ciliccea latrcillii, Leach. — n.s., outline indicating natural size of specimen figured below in partially bent
position and in lateral view; Pl.V., ventral view of the pleon, omitting thepleopods; Pl.D., dorsal
view of the pleon with seventh segment of the person; a.s., upper antenna; a.i, a.i, both members
of the lower pair of antennae, to show the casual want of symmetry ; l.s., upper lip and epistome
from the upper (inner) side; Li, lower lip; m., m,, the two mandibles; mx. 1, mx. 2, mxp., the first
and second maxillae and the maxillipeds; gn. 1, prp. 5, the first gnathopod and the fifth peraeopod;
pip. 1, 2, 4, 5, the first, second and fifth pleopods, and part of the fourth; int., a small piece of the
integument from side-plate of seventh peraeon segment.

The antennae, mouth organs, and limbs in detail are drawn to a uniform scale. The fragment of
the integument is more highly magnified.

PLATE IX.

A. Ciliccea whiteleggei, n. sp., £ . — n.s., line indicating natural size of specimen figured above in partially

bent position and in lateral view ; CD., dorsal view of head with first two segments of peraeon not
flattened out ; Pl.D., PI. V., dorsal and ventral views of pleon to the same scale as preceding figure ;
mx. 1, mxp., first maxilla and maxillipeds more highly magnified than the other figures, with one
exception; a.s., a.i, l.s., first and second antennae with epistome and upper lip; gn. 1, prp. 5, first
gnathopod and fifth peraeopod ; pip. 1, 2, 3, pip. 2, m.s., first, second and third pleopods, to the same
scale as the antennae and trunk limbs, but the separate male stilet of pip. 2 to the same scale as the
maxillipeds.

B. Ciliccea whiteleggei, n. sp., ? . — n.s., line indicating natural size of specimen figured above, much bent

and in lateral view ; PI., pleon in dorsal view, with last segment of peraeon and parts of the two
preceding segments; a.s., a.i., first and second antennae; mx. 1, mxp., first maxilla and maxillipeds.
These are more highly magnified than the other details, and more highly than the corresponding
parts of the male; gn. 1, prp. 5, pip. 1, first gnathopod, fifth peraeopod, and first pleopod.

C. Ciliccea sp., juv. — n.s., line indicating natural size of specimen figured above, slightly bent, and in

lateral view; mxp., maxillipeds magnified to the same scale as those in Plate IX., B; gn. 1, urp.,
first gnathopod and uropod.

PLATE X.

A. Ciliccea beddardi, n. sp. $ . — n.s., line indicating natural size of male specimen figured above in dorsal

view; a.s., a.i., first and second antennae; l.s., upper lip with epistome; Li., lower lip; »?,, m,,
complete mandible on the right of the plate, and on the left the cutting edges and spine row of its
companion; mx. 1, mxp., first maxilla and maxillipeds; gn. 1, 2, pip. 1, 2, 3, 4, 5, the first and
second gnathopods and the five peraeopods in lateral view, connected together; pip. 1, 2, 5, the
first, second, and fifth pleopods ; urp., uropod.

Of the details, the mouth parts are magnified on a higher scale than the other appendages.

B. Ciliccea beddardi, n. sp. $ . — n.s., line indicating natural size of female specimen figured above in dorsal

view and not quite flat. Some of the details are from another female specimen ; juv., dorsal view
of a young one taken out of the specimen of which the mouth organs and pleon are figured ;
a.s., a.i., first and second antennae; m., Li., mx. 1, 2, mxp., mandible, lower lip, first and second
maxillae, and maxilliped, rather more highly magnified than the antennae and limbs; gn. I, prp. 1,
first gnathopod and first peraeopod ; PL, telsonic segment and uropods.

ISOPODA. 63

C. Cymodoce bicarinata, Stebbing. — n.s., lines indicating natural size of specimen figured above in dorsal
view; PI., ventral view of pleon, the pleopods removed; g., part of the gastric apparatus; a.s., a.i.,
first and second antenna; Ls., l.i., m.rp., upper and lower lips and maxillipeds; gn, 1, prp. 5, first
gnathopod and fifth peraeopod ; g.p., genital papilla; from seventh person segment ; pip. 2, second
pleopod, with apical part of male appendix more highly magnified.

PLATE XI.

A. Amescptms richardscma; n. gen. et sp. — n.s. ? , line indicating length of body and second antenna; of the

female specimen figured above in lateral and below in dorsal view ; n.s. £ , line indicating length of
body of a young, probably male, specimen figured in dorsal view. C.T., a.s., a.i., cephalothorax
(head and first peraeon segment), with a first antenna and part of the second as far as base of fourth
joint; the flagellum shown separately; l.i., m., mx. 1, mx. 2, mxp., lower lip, a mandible (with part
more enlarged), first and second maxillae, and maxillipeds; gn. 1, gn. 2, first and second gnathopods.
All these details are from the female specimen, the mouth organs more highly magnified than the
other appendages, prp. 4 $ , pip. <$ , wrp. £ , fourth peraeopod, second pleopod, and uropod from a
male specimen 4 millims. long. The figures drawn to the same scale as that used for the limbs of
the female.

B. Jsfacilla amblyura, n. sp. — n.s., line indicating length of body and second antennae of the specimen

figured in lateral view ; ft, part of head ; PL, dorsal view of pleon, showing one of the valvular
uropods thrown open ; a.s., a.i., first and second antenna} ; m., mx. 1, mx. 2, mxp., niandible (with
part more enlarged), first and second maxilla;, maxillipeds; gn, 1, gn, 2, prp. 5, first and second
gnathopods and fifth peraeopod, with the finger of each more enlarged ; pip. 2, urp., second pleopod
and uropod.

C. Jaropsis curvicornis (NiCOLET). — n.s., line indicating length of specimen figured above in dorsal view ;

ft, dorsal view of head, with upper lip and one of the mandibles projecting in front; PL, ventral
view of pleon, without the pleopods ; one uropod more highly magnified ; operc, opercular plate
formed by the first pleopods; a.s., a.i., upper and lower antennas, with most of the lower antenna
more enlarged ; in., m., mx. 1, mx. 2, mxp., the mandibles, first and second maxilla;, and maxillipeds,
with the cutting plates of the mandibles and one palp of the maxillipeds more highly magnified ;
gn. 1, gn. 2, prp. 5, first and second gnathopods and fifth peraeopod, with the fingers of first
gnathopod and fifth peraeopod more enlarged.

All the detail figures are enlarged to the same scale, but are accompanied in some instances by
parts more highly magnified.

PLATE XII.

A. Stenrtrinm chiltoni, n. sp. — n.s., lines indicating natural size of specimen figured below in dorsal view;
C, a.s., a.i., dorsal view of head more enlarged, with the eyes, the first antenna of the left side,
part of that on the right, and parts of the second antennae, ending with the third joint on the left,
with the fourth on the right. PL, terminal part of pleon ; this, with the legs and pleopods, is
enlarged to the same scale as the preceding figure, while the mouth organs are more highly
magnified, and the spines of the first gnathopod still more highly. Ls., Li., upper and lower lips;
in., in., mx. 1, mx. 2, mxp., <>p., mandibles, first and second maxilla?, maxillipeds, with one epipod
detached; gn. 1, gn. 2, prp. 1, first gnathopod, with some of the spines of the palm and finger very
highly magnified ; second gnathopod ; first peraeopod ; pip. 1, 3, 4, 5, the first, third, fourth and fifth
pleopods ; a.i'., second antenna from a different specimen, to which the following parts also belong ;

64 CEYLON PEAEL OYSTER REPORT.

gv. V, first gnathopod, showing the marsupial plate; prp. 5', terminal part of fifth peraeopod ;
wrp.', the uropods.
B. Gnathia insolita, n. sp. — re.s., lines indicating natural size of specimen figured in dorsal view ;
C, cephalic region, showing the muscles belonging to the mandibles; PL, pleon in dorsal view,
with rudimentary seventh segment of person ; a.s., a.L, first and second antennae ; m., mxp., a
mandible and the maxillipeds ; r/n. 1, 2, first and second gnathopods ; pip. 1, pip., first pleopod, and
one of the following pairs.

All the details are magnified to the same scale.

CEYLON PEARL OYSTER REPORT.
A.

ISOPODA, PLATE I.

prp.i.

prp

gn.i ?

urp. ?

gn.i.l

D.

Del. T.R R Stebbinft

A.HETEROTANAIS CRASSICORNIS.n.sp B. LEPTOCHELIA MIRABILIS.nsp. C. L. LIFUENSIS, Stefbing

D.TANAIS GRACILIS. Heller.

J T Rennic Reid.LithEdirf

I EYLON PEARL OYSTER REPORT
A' — 4#0

ISOPODA, PLATE II.

Del r. RR Subbing

A. CONILORPHEUS HERDMANI, n £.etsp.

B. HANSENOLANA SPHAEROMIFORMIS (Hansen).

JTRenme Reid. Lich EdinT

CEYLON PEARL OYSTER REPORT
A.

ISOPODA, PLATE III

Del T R R. 5tebbm£

JTReraii; R«d

A.ARGATHONA NORMANI. n.g.et.sp. B. ClLICAEA LATREILLII, Leach, juv. C. jANIRAt?) NANA, n.sp.

CEYLON PEARL OYSTER REPORT

ISOPODA, PLATE IV.

Del T R.R Stebbin/;.

PI.

J.T.Eonnie nnu,

JEGA OMMATOPHYLAX, n.sp.

(FALUN PEARL OYSTER REPORT.
A.

ISOPODA, PLATE V

n.s.

Del. T.K.R Stubbing

J.T.Rennie Held 1.

A. /EGA OMHATOPHYLAX.n. sp. B. LANOCIRA ZEYLANICA, n.

sp

CEYLON PEARL OYS :ER REPOR1
A.

ISOPODA, PLATk VI

:- Seebbing

A.RHIOTHRA CALLIPIA, Schioate & Meinert . B . IRONA NANOIDE S, n. sp.
C. ROCINELA ORIENTALIS, Schiodte & Meinert.

( KYIA.LN PKAKL I lYSTER REPORT

IS OP 0 DA. PLATE \ II

Bel. TR R SicDbmg.

SPHAEROMA WALKERI, n sp

. emue Reid L:cn EOm-

0 * PEARL >■■ REP0R1

OPODA PLATE VIII

5tcM>tn£

'

CIL1 ■ \l \ I rREILLIl I ea.ch.

CEYLON PEARL OYSTER REPORT

1SOPODA, PLATE IX.

Del. T.R.R Sobbing.

J T.Reniuc Reid.Lith Edm1

A B ClLICAEA WH1TELEGGEI, n. sp- C ClLICAEA (?) sp. juv.

CEYLON PEARL OYSTER REPORT

A.

ISOPODA, PLATE X.

r?

7J

Del T R R Stebbrag. P2.

A.B. ClLICAEA BEDDARDI.n sp . C. CYMODOCE BICARINATA, Stebbi:

J. T Renme Beid. L-.tti- Edw

CEYLON PEARL OYSTER REPORT.

ISOPODA, PLATE XI.

B.

,T\

(jTL.S

prps

R R Storing

J.T.Renme He id, LitV E«W

A. AMESOP'OUS RICHARDSON^, n.sp. B. ASTACILLA AMBLYURA.n sp. C. JAEROPSIS CURVICORNIS, (Nicolet).

CEYLON PEARL OYSTER REPORT
A.

ISOPODA, PLATE XII

B.

Del T R R Stebiing

J TRcuJUe field Lilh Edin1

A, STENETRIUM CHILTONI, n. sp. B. GNATHIA INSOLITA, n . sp.

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

REPORT

ON THE

MACRURA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY
JOSEPH PEARSON, B.Sc,

DEMONSTRATOR AND ASSISTANT LECTURER IN ZOOLOGY, UNIVERSITY OF LIVERPOOL,

[With TWO PLATES.]

Professor Herdman's collection of Macrurous Crustaceans from Ceylon consists of
53 species, of which 4 are new to science.

With the probable exception of 3 species the collection is entirely typical of a
representative Indo-Pacific shallow-water fauna. Bithynis savignyi (Sp. Bate),
Athanas nitescens, Leach, and Urocans longicaudata, Stimpson, appear hitherto to
have had a known distribution limited to the Atlantic and the Mediterranean.

The following is a list of the species described in this Report :—

Tribe: Pen^eidea.

Family: Penaeidoe.

Penwus canaUcuIatus (Olivier).
„ monodon, Fabricius.
„ indicus, Milne-Edw.
Pampcnwus anchwalis (Sp. Bate).
„ iwhipes (Sp. Bate).

„ dalei, Rathbun.

„ mogiensis, Rathbun.

,, acclivis, Rathbun.

gallensis, n. sp.
Philonicus peciinatus, Sp. Bate.
Sii-i/nni'i hincifer (Olivier).
„ cristaia (de Haan).
„ sculpta, Milne-Edw.

Family : Sergestidse.

Antes indicus (Milne-Edw.).
Leudfer typus (Vaugh. Thomps.).

Tribe : Caridea.
Family: Atyidse.
Caradina viliensis, Borradaile.

Family : Pontoniidae.

I'ericlimrnes ritirnsis, BORRADAILE.

„ dance (Stimpson).
Conchodytes meleagrinw, Peters.
Anchistus inermis (Miers).

66

CEYLON PEARL OYSTER REPORT.

Family: Palaemonidse.

BUhynis savignyi (Sp. Bate).
Urocaris longicaudata, Stimpson.

Family: Latreutidae.

Nimticaris futilirostris, Sp. Bate.

,, unirecedens, Sp. Bate.

,, grandirostris, n. sp.
Latreutes ceylonensis, n. sp.

Family: Alpheid*.

Alpheus idiocheles, Coutiere.

„ phrygianus, Coutiere.

„ paraeuleipes, Coutiere.

„ parakyone, COUTIERE.

„ miersi, Coutiere.

„ pareudwirus, Coutiere.

„ bis-intisus, var. malensis, Coutiere.

„ „ „ stylirostris, Coutiere.

„ audouini, Coutiere.

„ inacrodadylus, Ortmann.

„ spongiarum, Coutiere.

„ Icevis, Randall.
Synalpheus gravieri, Coutiere.

Synalpheus laticeps, Coutiere.

„ biunguicalalus, Stimpson.

„ comatukrum, Haswell.

„ neomeris, de Man.

„ carinatus, de Man.
Athanas nitescens, Leach.
„ orientalis, n. sp.

Family : Crangonidee.
Mgmn cafaphradus (Olivier).

Family : Processidse.

Processa canaliculata, Leach.

Tribe : Scyllaridea.

Family: Scyllaridfe.

ScyUarus tuberculatus (Sp. Bate).
„ sordidvs (Stimpson).

Tribe : Thalassinidea.

Family : Callianassidie.

Callianassa rotundicavdaia, Stebbing.
,, maldivensis, Borradaile.

Upogebia intermedia (de Man).

In addition to the above species, two fresh-water forms — Biihynis Jar and Bithynis
granclimanus — were obtained from the River Gin Ganga. They are not included in
this Report.

List of Stations

at which Macrura were obtained, with the species collected in each :—

STATION I. — Five miles west and south-west of Negombo ; 1 2 to 20 fathoms ;
bottom coarse yellow sand with a few dead shells.
ParaperuBus mogiensis.

STATION III.— Two and a-half to four miles off Chilaw ; 9 to 14 fathoms ; bottom
coarse sand and small corals.

Penceus canaliculatus, Parapenceus anchoralis, Parapenceus dalei, ParapencBUs

mogiensis, Parapenceus acclivis, Pliilonicus pectinatus, Sicyonia lancifer, Anchistus
inermis, Perxclimenes dance, Alpheus miersi, Alpheus phrygianus, Alpheus audouini,
Synalpheus gravieri, Synalpheus neomeris, Processa canalictdata, Callianassa
maldivensis.

STATION VI.— Across Muttuvaratu Paar ; depth 6 to 9 fathoms ; bottom sand, with
hard patches of " rock " at intervals.

Leucifer typus, Bitlujjiis savignyi, Alpheus idiocheles, Athanas orientalis, Processa
canaliculata, Upogebia intermedia.

MACRURA. 67

STATION XVIII. — South-west part of Palk Bay, off Rameswaram Island and Adam's
Bridge ; 7 to 9 miles off shore ; bottom fine soft bluish-grey mud containing casts
of various Molluscan shells ; depth 7 to 8 fathoms.

Penoms monodon, Parapenceus mogiensis, Leucifer typus.

STATION XX.— North part of Back Bay, Trincomalee ; depth 11 to 13 fathoms;
bottom hard.

Parapenwus anchoraUs, Pcirapenceus dalei, Parapenceus acclivis, Sicyonia sculpta.

STATION XXXV.— Entrance to Galle Harbour ; depth 4£ to 7 fathoms ; bottom
coarse sand.

Nauticaris unirecedens, Nauticaris grandirostris, &gcon cataphractus, Scyllarus
tuhercidatus.

STATION XXXVI.— Galle Harbour, off Gibbet Island ; depth 2 to 4§ fathoms ;
bottom fine sand and mud.

Parapenceus dalei, Leucifer typus, Acetes indieus, Caradina vitiensis, Nauticaris
futilirostris.

STATIONS XXXIX. to XLIL— Deep water off Galle ; depth up to 100-fathom line ;
bottom sand, shells, nullipores.

Parapenceus dalei, Parapenceus mogiensis, Parapenceus acclivis, Parapenceus
gallensis, Synalpheus laticeps, Synalpheus bkmguiculatus, Synalpheus carinatus,
Alpheus Icevis.

STATION XLIII. — Six miles west of Kaltura ; depth 22 fathoms ; bottom hard sand
and nullipores.

Parapenceus anchoraUs, Philonicus pectinatus.

STATIONS XLVII. to XLIX.— Cheval Paar ; depth 6^ to 13 fathoms ; bottom sand,
nullipores and dead shells.

Penceus indieus, Parapenwus anchoraUs, Leucifer typus, Periclimenes vitiensis,
Periclimenes dance, Conchodytes meleagrince, Anchistus inermis, Alpheus plirygianus,
Alpheus paraculeipes, Alpheus miersi, Alpheus spongiarum, Alpheus bis-incisus, var.
malensis, Synalpheus gravieri, Synalpheus comatidorum, Athanas nitescens, Athanas
orientalis, Processa canalicidata, Callianassa rotundicaudata.

STATION LIIL — Ten to twelve miles north of Cheval Paar, and 12 miles west of
Vankali Church ; depth 7^ to 9 fathoms ; bottom muddy sand with dead shells.
Latreutes ceylonensis.

STATION LIV.— South of Adam's Bridge ; depth 4 to 40 fathoms ; bottom varied,
from sand to living coral.

Conchodytes meleagrince, Alpheus par eucheir us, Upogebia intermedia.

K 2

(58 CEYLON PEAEL OYSTEE EEPOET.

STATIONS LV. to LX.— Coral reefs and pearl banks, Gulf of Manaar ; depths
varying from 9 to 36 fathoms ; bottom nullipores and dead coral.

ParapencBus mogiensis, Sicyonia cristata, Conchodytes meleagrince, Alpheus miersi,
Alpheus bis-incisus, var. stylirostris, Alpheus audouini, Synalpheus gravieri,
Synalpheus comatidorum, Synalpheus laticeps, Synalpheus biunguiculatus, Athanas
orientalis, Scyllarus sordidus.

STATION LXL— Northern end of Periya Paar ; depth 12 to 14 fathoms ; bottom
sand, nullipores and coral.

Leucifer typus, Sicyonia cristata, Synalpheus gravieri, Athanas orientalis.

STATION LXIV. — From between South Modragam and Jagerboom paars along a line
south-west towards Kodramallai Point ; depth 4|- to 5|- fathoms ; bottom coarse
sand, with much fine green weed and small pearl oysters.

Periclimenes vitiensis, Processa canaliculata.

Aripu Reef; depth 5 fathoms.

Parapenceus acclivis, Parapenmus anchoralis, Philonicus pectinatus, Urocaris
longicaudata, Scyllarus tuberculatum,

STATION LXVIL— Off south end of Mutwal Island; depth 10 to 14 fathoms;
bottom dead coral and nullipore.

Synalpheus gravieri, Alpheus par alcy one, Alpheus miersi, Alpheus macrodactyhis.

STATION LXVIIL— From off Coppeluddi southwards to Navakaddua Paar ; depth
8 to 18-J- fathoms ; bottom nullipores, coral and muddy orbitolites sand.
Synalpheus comatidorum.

STATION LXIX.— North end of Chilaw Paar; depth 8 to 11 fathoms; bottom
yellow quartz sand with some coral fragments.

Synalpheus biunguiculatus, Synalpheus comatulorum, Nauticaris unirecedens.

MACRURA.

Tribe: PEN^EIDEA.

Family: PEN^EIDiE.
Penaeus, Fabricius, 1798.

Penaeus canaliculars (Olivier).

Palaemon canaliculatus, Olivier, 'Ency. Method.,' VIII., p. 660, 1811.
Penaus canaliculatus, Milne-Edwards, 'Hist. Nat. Crust.,' II., p. 414, 1837.
Penaeus canaliculatus, var. japonicus, Sp. Bate, '"Challenger" Macrura,' p. 245, 1888.

Locality: — Pearl banks, Gulf of Manaar (Station III.), 1 specimen very much
damaged.

General distribution :— Japan, Mauritius, Figi, Australia, Ceylon.

MACRURA. 69

Penaeus monodon, Fabricius.

Penaeus monodon, Fabricius, 'Suppl. Ent. Syst.,' p. 408, 1798.
Penaeus semisnlcatns, de Haax, ' Fauna Japonica,' p. 191, 1849.
Penaeus carinatns, DANA, ' U.S. Expl. Exp.,' p. 602, 1852.

Locality : — Palk Strait (Station XVIII.), several specimens.
Measurements of two males and two females : —

From end of telson to

Males. Females.

rom end ot telson to "] .... ....

„ > 162 millims., 131 millims. 195 nullims., 135 millims.

tip ot rostrum . . J

Length of carapace 1
and rostrum . . J

52 „ 41 „ 60 „ 43

Both male and female specimens possess a well-marked median groove extending
from behind the rostrum to the posterior end of the carapace. Spence Bate found
this groove absent in the single male which he examined. There is no doubt that
the position of the ventral rostral teeth relative to the upper rostral teeth is subject
to variation.

General distribution : — India, Ceylon, Singapore, Japan, Pacific and South Africa.

Penseus indicus, Milne-Edw. — Plate I., fig. 1.

Penaeus indicus, Milne-Edw., ' Hist. Nat. Crust.,' II, p. 415, 1837.

(?) Penaeus merguiensis, he Max, ' Journ. Linn. Soc.' (Zool.), vol. 22, p. 287, 1888.

Locality: — Gulf of Manaar (Station XLIX.), 5 specimens. Lengths varying from
14 millims. to 40 millims. All immature.

The rostrum in these specimens differs from the type species and has a formula

— *— . The anterior half of the rostrum bears no teeth dorsally. The rostrum

6 J

is very slender and extends in front of the antennular peduncle a distance equal to
half the length of the peduncle. In spite of these differences from the type species,
I have referred these specimens to the above species, because there seems to be little
doubt that the rostrum is subject to a great deal of variation in this form. The
" Challenger " specimens differed in the form of the rostrum from Milne-Edward's
species, and de Man's species, P. merguiensis, appears to differ in no important
respects from P. indicus.

An examination of the various forms grouped together under this species would be
valuable and instructive, and would throw some light on the value of the rostrum in
classification.

General distribution : — India, Philippines, Mergui (?), Ceylon.

70 CEYLON PEAEL OYSTER REPORT.

Parapenseus, Smith, 1885.

I have followed Smith'5' in separating certain species from the genus Penreus. The
characters of the genus Parapenceus, which distinguish it from the genus Penceus,
are: — (l) Endopodite of 1st maxilla is short and unsegmented ; (2) 3rd maxilliped
without an epipodite, and (3) the absence of branchiae from the last thoracic segment.

Parapenseus anchoralis (Sp. Bate).

Penaaus anchoralis, Sr. Bate, ' " Challenger " Macrura,' p. 258, 1888.

Localities: — Pearl banks, Gulf of Manaar (Station III.), 1 specimen (?); Trin-
comalee (Station XX.), 1 specimen (?) ; off Kaltura (Station XLIIL), 1 *$ ; Galle
(Station XXXVIL), 1 ? ; Aripu Reef (Station XLIV.), 1 S and 2 ?.

Male : — Total length 40 millims., carapace and rostrum 15 millims.

Female : — Total length 64 millims., carapace and rostrum 23 millims.

Rostral formula is ^ — <- •

0

The rostrum in the females appears to be slightly longer than in the males. The
female rostrum reaches to the end of the second antennular segment. In the male it
only reaches slightly past the first segment.

The dorsal groove of the telson appears to be deeper in the female than in
the male.

General distribution : — Pacific, Japan, Ceylon.

Parapenaeus incisipes (Sp. Bate).

Penaeus incisipes, Sp. Bate, ' " Challenger " Macrura,' p. 257, 1888.
Parapenseus incisipes, Rathbux, ' Proe. U.S. Nat. Mus.,' vol. xxvi., p. 38, 1902.

Locality: — Gulf of Manaar, 1 specimen, <$.

The rostrum is straight and not so deep as in the preceding species. It extends

• 9 + 1
to the middle of the 3rd joint of the antennular peduncle. Rostral formula is — — •

The flagella of the antennule are short and equal in length to the 2nd and 3rd
joints of the peduncle.

Two obliquely longitudinal grooves cross each side of the carapace. The anterior
groove becomes comparatively deep ventrally. The carapace tends to become much
shallower dorso-ventrally at the anterior end.

The meropodite of the 5th pereiopod is notched at its proximal end. There is a
tubercle present on the endopodite of the 2nd abdominal appendage. The 6th
abdominal segment is 1^ times as long as the 5th. The uropods are slightly notched
at their outer proximal margins.

General distribution : — Philippines, Japan, Ceylon.

* Smith, ' Proc. U.S. Nat. Mus.,' VIII., p. 170, 1885.

MACKURA. 71

Parapenaeus dalei, Rathbun.*

Parapenaeus dalei, M. Rathbun, ' Proc. U.S. Nat. Mus.,' vol. xxvi., p. 40, 1902.
Localities: — Pearl banks, Gulf of Manaar (Station III.), 1 J; Galle (Station
XXXVI. ), 2 cf ; south of Galle (Station XXXIX.), 1 6 ; Trincomalee (Station XX.),

1 <?.

These specimens appear to agree with Miss Rathbun's diagnosis of the species.
The antennular flagella are thicker than in the other species of the velutinus group.
Length of largest specimen ( 6* ) : —

Total length from end of rostrum to tip of telson . . 40 millims.

From tip of rostrum to end of carapace 11 ,,

Side length of carapace 9 ,,

Rostral formula .

0

General distribution : — Japan, Ceylon.

Parapenaeus mogiensis, Rathbun — Plate I., fig. 2.

Parapenaeus mogiensis, M. Rathbun, ' Proc. U.S. Nat. Mus.,' vol. xxvi., p. 39, 1902.
Localities : — Pearl banks, Gulf of Manaar (Station III.), 31 specimens ; west of
Negombo, hauls 1 to 4 (Station I.), 15 specimens; Palk Straits (Station XVIII. ),

2 specimens; south of Galle (Station XXXIX.), 1 specimen; Coral reef, Gulf of
Manaar (Station LIV.), 7 specimens.

The petasma agrees with the type, but the thelycum in all the specimens shows a
slight difference (see fig. 2). The appendages are more richly setose than is the
case with the other species of this group. The 3rd maxillipeds do not extend as far
as the tips of the antennal scales, but are situated behind them a distance equal to
the length of the distal joint of the 3rd maxilliped. The anteDnal scale is slightly
longer than the antennular peduncle.

Dimensions of males (3 specimens) : —

Total length 56 millims., 48 millims., 61 millims.

Length of rostrum and carapace . .18 ,, 15 ,, 19 ,,
Lateral length of carapace ... 15 „ 12 „ 14 ,,

Dimensions of females (3 specimens) : —

Total length 75 millims., 67 millims., 61 millims.

Length of rostrum and carapace . . 24 ,, 23 „ 20 ,,
Lateral length of carapace ...19 ,, 18 ,, 16 ,,

Rostral formula .

0

General distribution : — Japan, Ceylon.

* This and the three following species belong to the Parapenaeus velutinus group. This collection of
Ceylon Crustaceans has given me an opportunity of examining a large number of specimens belonging to

72 CEYLON PEARL OYSTER REPORT.

Parapenseus acclivis, Rathbun.

Parapenaeus acclivis, Rathbun, ' Proc. U.S. Nat. Mus.,' vol. xxvi., p. 41, 1902.
Localities : — Pearl banks, Gulf of Mauaar (Station III.), 1 specimen ; Trincomalee
(Station XX.), 1 specimen; south of Galle (Station XXXIX.), 1 specimen; Aripu
Reef (Station LXIV.), 1 specimen.

The 3rd maxillipeds do not extend as far as the ends of the antennal scales, but
they are longer than those of Parapenceus mogiensis. The antennal scale is slightly
shorter than the antennular peduncle.
Dimensions of two females : —

Total length 88 millims., 80 millims.

Length of rostrum and carapace .... 29 ,, 26 ,,
Lateral length of carapace 25 ,, 21 ,,

Rostral formula ^ ^ .

0

General distribution : — Japan, Ceylon.

Parapenaeus gallensis, n. sp. — Plate I., fig. 3.

Locality : — South of Galle (Station XXXIX.), many specimens.

This species is of the velutinus type, and possesses certain characters which
distinguish it from Dana's species and also from the species formed by
Miss Rathbun.

In the form of the rostrum this sjaecies resembles somewhat closely P. dalei. The
length of the rostrum, however, is slightly shorter than in the latter species. In
typical specimens the rostrum reaches to the end of the 1st segment of the antennular
peduncle. In P. dalei the rostrum reaches to the middle of the 2nd segment. The
rostrum in P. gallensis is generally less toothed than in P. dalei, having a formula of

^ — — '- . The anterior tooth is much smaller than the others, and in some cases is

hardly perceptible. But in the form of the petasma and thelycum it differs distinctly
from P. dalei, and approaches nearer to P. akayebi, showing, however, distinct
differences from the latter species. The rostrum is much shorter than in P. akayebi
and has fewer teeth. The left branch of the petasma is much more delicate and
slightly longer than the right branch, and ends in a few small denticles, which,
however, are only observed when the petasma is examined under a microscope. The
antennal scale reaches as far forward as the extremity of the antennular peduncle.
The 3rd maxillipeds are slightly longer than those of the three previous species,
reaching almost to the end of the antennal scale. There is a pair of well-developed

to the velviinus group, and I think Miss Rathbun is justified in separating from the old species certain
forms possessing various definite and distinctive characters. But it is doubtful whether these characters
are of sufficient importance to warrant the formation of new species, and I am not sure whether their
separation from Farapenmus velutinus merely as new varieties would not have been preferable.

MACRURA. 73

spines present between the bases of the feet of the second pair. These only appear
to be present in the female. The last four abdominal segments have a dorsal carina
which ends in a well-developed tooth at the posterior end of the 6th segment.
The sixth abdominal segment is about \^ times as long as the fifth and slightly
shorter than the telson. The telson is slightly shorter than the uropods and has the
usual number of spines, which is characteristic of the velutinus group. The anterior
pair of spines are much smaller than the posterior three pairs, and in some specimens
are only observed with difficulty.

The general surface of the body is smooth. The dimensions are : —

millims. millims. millims. niillims. millims. millims.
Total length along mid-dorsal line from] .„ ,„ ^ .„ .„ .„

tip of rostrum to end of telson . . J
Length of carapace and rostrum along lir 1C- ic -i a i~ i n. c

mid-dorsal line I

Side length of carapace to tip of] . „ i n i c, -, n i c q

antennular tooth J

Length of sixth segment of abdomen, 1 „ fi „ fi _

mid-dorsal line J

Sex . . ? ? ? ? ? <?

It would appear that this species is in an intermediate position between ParapencBus
akayebi and P. dalei.

The characters of the rostrum and of the petasma and thelycum appear to be the
most reliable characters on which to base the identification of the various species of
this group. Of these characters the former is not altogether trustworthy, as it is
subject to some variation, and it is not impossible to obtain a series based upon the
length of the rostrum and the number of rostral teeth which will connect all the
species of this group. Still, in a broad manner, Miss Bathbun's method of separation
holds good.

The form of the genital opercula appears to be much more constant, and each of
the species is quite distinct in this respect.

The comparative length of the sixth abdominal segment does not seem to be
constant enough to be of value as a basis of identification.

So that, as I have already suggested, it is, perhaps, placing too high a value upon
the distinguishing characters of the various forms of the velutinus group to give these
forms the rank of species. For the present, however, I follow Miss Bathbun.

Philonicus, S pence Bate, 1888.

Philonicus pectinatus, Sp. Bate.

Philonicus pectinatus, Sp. Bate, ' " Challenger," Macrura,' p. 279, 1888.
Localities: — Deep water off Galle (Station XL.), 1 specimen ?, 35 millims.; off

L

74 CEYLON PEAKL OYSTER REPORT.

Kaltura (Station XLTII), 2 ? , 35 millims. and 48 millims. ; Pearl banks, Gulf of
Manaar (Station III.), 1 ? , 48 millims. ; Aripu Reef (Station LXIV.), 1 <?, 35 millims.

All the specimens are much damaged, the exoskeleton apparently being poorly
calcined. In all the female specimens the thoracic legs are incomplete.

In the male the petasma is comb-like as described by Sp. Bate. The antennular
flagella, which were absent in the "Challenger" specimen, are typical of the genus.
Tliere are two long flagella on each antennular peduncle, the upper one being thin and
the lower being very broad. The antennal scale is well developed and is longer than
the antennular peduncle and twice as long as the antennal peduncle. The antennal
flagellum is long and is slightly thicker than the upper antennular flagellum. The
form of the rostrum and carapace agrees with Spence Bate's description. In the
" Challenger" specimen the 3rd maxillipeds and the thoracic legs were absent.

In the Ceylon specimens the 3rd maxillipeds are very long, reaching well in front
of the antennal scale. The distal end of the propodite is about on a level with the
end of the antennal scale. All the joints are richly covered with setae, some of
which are so robust as to have the appearance of long slender spines. Of the thoracic
legs the first is the shortest and only reaches to the end of the carpos of the 3rd
maxilliped. The second leg reaches to the end of the propodite of the 3rd maxilliped.
The third pair is missing. The fourth is very long and slender, reaching almost as
far forward as the end of the 3rd maxilliped. The last pair of thoracic legs are short,
being only slightly longer than the 1st pair and more slender.

Comparative lengths of appendages in male specimen : —

3rd maxilliped 12 millims.

1st thoracic leg 9 ,,

2nd „ „ 11-5 „

3rd (wanting)

4th thoracic leg 14 millims.

5th „ „ 10' „

General distribution : — Papua, Ceylon.

Sicyonia, Milne-Edw., 1830.

Sicyonia lancifer (Olivier).

Palaemon lancifer, Olivier, ' Encyclop.,' t. vi., p. 664.
Sicyonia lancifer, de Haan, 'Fauna Japonica,' p. 194, 1849.
Sicyonia lancifer, Sp. Bate, ' " Challenger" Macrura,' p. 297, 1888.

Locality : — Pearl banks, Gulf of Manaar (Station III.), 3 ? and 1 c?, average length
40 millims.

There is very little to add to Spence Bate's description and figures.

In the male the petasma is symmetrical, and there is a conspicuous triangular plate
between the bases of the last three pairs of thoracic legs, having the apex produced

MACRURA. 75

into a long pointed tooth anteriorly like the thelycum in the female. As in the
female, there is a pair of spines present at the base of each of the first two pairs of
walking legs. The 3rd maxillipeds are similar in both sexes, and are more massive
than the thoracic walking legs and extend slightly further forward than the tip of
the antennaJ scale. Thev are richly setose and have the joints somewhat flattened.
The first three pairs of walking legs are chelate and the carpopodite is only slightly
longer than the propodite. The fingers are slightly longer than the palm.
General distribution :— New Guinea, Japan, Indian Ocean.

Sicyonia cristata (de Haan) (?).

Hippolyte cristatus, de Haan, 'Fauna Japonica,' pi. xlv., 1849.
Sicyonia cristata, de Haax, 'Fauna Japonica,' p. 194, 1849.

Localities :— West of Periya Paar (Station LXL), 1 c?, 25 millims. ; off Dutch
Modragam (Station LVIL), 1 3, 25 millims.; Coral reef, Gulf of Manaar (Station
LIV.), 1 $, 35 millims.

I refer these specimens to the above species, although a little doubtful as to their
identity. I have not been able to see figures of de Ha ax's species, but these
specimens agree closely with the description. Sp. Bate evidently considered this
species to be identical with Sicyonia lancifer. The Ceylon specimens resemble the
latter species closely, but they possess characters differing from S. lancifer. The
carapace is more arched than in S. lancifer. The rostrum does not turn up at
the end and only reaches to the end of the eyes. The abdomen differs only from that
of S. lancifer in not having pleural spines on each segment. The three distal joints
of the 3rd maxillipeds appear to be more flattened than in S. lancifer, and the
hepatic spine on the carapace is not so well developed as in that species.

General distribution : — Japan, Ceylon.

Sicyonia sculpta, Milxe-Edw.

Sicyonia sculpta, Milxe-Edw., ' Ann. des Sci. Nat.,' ser. 1, t. 19, p. 339, 1830.

Locality : — Trincomalee (Station XX.), 1 ? , 17 millims.
General distribution : — Atlantic, Mediterranean, Ceylon.

Family: SERGESTID^E.

Acetes, Milxe-Edw., 1830.

Acetes indicus, Milxe-Edw.

Acetes indicus, Milxe-Edw., 'Ann. des Sci. Nat.,' t. 19, p. 350, 1830.
Locality :— Galle Harbour (Station XXXVI.), 1 specimen, 14 millims. long, in a
damaged condition, but evidently belonging to this species.

This species is generally found either in fresh water or in the brackish water of
estuaries.

Distribution : — Mouth of River Ganges, Ceylon.

L 2

76 CEYLON PEARL OYSTER REPORT.

Leucifer, Milne-Edw., 1837
(= Lucifer, Vaughan Thompson, 1829).

Leucifer typus (Vaughan Thompson).

Lucifer typus, Vaughan Thompson, 'Zool. Researches,' p. 58, 1829.
Leucifer typus, Milne-Edw., ' Hist. Nat. Crust.,' t. ii., p. 469, 1837.

Localities: — Muttuvaratu Paar (Station VI.); Palk Strait (Station XVIII.) ;
Galle Harbour (Station XXXVI.) ; south of Cheval (Station XLVIL); Cheval Paar
(Station XLIX.) ; Periya Paar (Station LXL). A large number of specimens from
the various localities. It is probable that most of these specimens were taken in the
tow-net, although the labels do not definitely say so, except in one instance.

Altogether there are some hundreds of specimens, and I am satisfied that they all
belong to this species. It is worthy of note that in the separate gatherings the
specimens are almost entirely of one sex.

The differences from Spence Bate's description are very slight. In the male the
eyes do not quite reach to the end of the 1st segment of the antennular peduncle.
Bate describes them as reaching nearly to the end of the 2nd segment.

The antennal scale is shorter than the 1st segment of the antennular peduncle and
about equal in length to the eye. In the females the eyes are slightly shorter than
in the male, and in the females the spines at the base of the abdominal appendages
are not so well developed as in Sp. Bate's figures.

General distribution : — North and South Atlantic, Pacific, Australia, Ceylon.

Tribe: CAKIDEA

Family: ATYID.E.

Caradina, Milne-Edw., 1837.

Caradina vitiensis, Borradaile — Plate I., fig. 4.

Caradina vitiensis, Borradaile, 'P.Z.S.,' 1898, p. 1003.
Locality : — Galle (Station XXXVL), 5 specimens, average length 13 millims.
These specimens appear to agree closely with Borradaile's description, but the

i g 20

rostrum is not quite so richly toothed, having a formula — - — . The ventral rostral

teeth are smaller than the dorsal ones. The anterior border of the eyes appears to
be slightly concave in all the specimens. The chelas of the first two pairs of
thoracic legs are typical of the genus. The distal joint of the last thoracic legs has
a large number of closely packed spines on its posterior border. The sixth abdominal
segment is almost twice as long as the fifth, and the telson is equal in length to the
sixth segment. The telson bears five pairs of small spines on its dorsal side. Each
corner of the posterior border bears a small spine, and there are four pairs of longer
spines arranged along the posterior border.

MACRURA.

77

The Ceylon specimens are marine. Borradaile's Figi specimens were obtained
from fresh water.

General distribution : — Figi and Ceylon.

Family: PONTONIIDiE.

Conchodytes, Peters, 1851.

Conchodytes nieleagrinae, Peters.

Conchodytes meleagrinae, Peters, ' Ges. naturf. Freunde, Berlin,' 1851.
Pontonia meleagrinae, Bate, ' " Challenger " Macrura,' p. 707, 1888.

Localities : — Cheval Paar (Station XLVIII.), four 6 ; Cheval Paar (Station LIV.),
two <$ and two ? within Pinna ; West Cheval (Station LVIII. ) two S and two ?
from shell of Pinna.

The mouth parts agree with Spence Bate's description of the thoracic legs, the
second pair are very massive and are longer than the body. The ischium, meros and
carpos are subequal in length, and the three together are shorter than the palm,
which is long and massive and more than twice as long as the fingers. The 1st, 3rd,
4th and 5th pairs are small, decreasing slightly in length from before backwards.
The distal joint of the last three pairs is triunguiculate.

Dimensions of male specimen : — -

Total length from rostrum to end of telson
Length of carapace and rostrum

1st thoracic leg

2nd

3rd

4th

5th

24 millims.

10

12

27

11

10

10

General distribution : — East Africa, Torres Straits, New Guinea, Pacific, Ceylon.

Anchistus, Borradaile, 1898.

Anchistus inermis (Miers).

Harpilus inermis, Miers, ' Zool. Coll. of " Alert," ' 1884.

Anchistns inermis, Borradaile, ' Ann. Mag. Nat. Hist.' (7), ii., 189S.

Localities: — Pearl banks, Gulf of Manaar (Station III.), three c? and two ?;
Cheval Paar (Station XLVIII), three <$ and three ? .

There is nothing to add to the original description. In most of the specimens the
sides of the carapace and abdomen are only very slightly calcified.

General distribution : — West Australia, Ceylon.

78

CEYLON PEAKL OYSTER REPORT.

Periclimenes, Costa, 1844.

Periclimenes vitiensis, Borradaile.

Periclimenes vitientis, Borradaide, ' Ann. Mag. Nat. Hist.' (7) ii., p. 383, 1898.
Localities: — Cheval Paar (Station XL VIII. ), 1 specimen ; south-east of Modragam
(Station LXIV), 1 specimen " on weed bearing oyster spat," length 18 millims.
General distribution : — Pacific, Ceylon.

Family : PALiEMONID^E.
Bithynis, Philippi, 18G0.

Bithynis savignyi (Sp. Bate).

Brachycarpus savignyi, Sp. Bate, ' " Challenger" Macrura,' p. 798, 1888.
Bithynis savignyi, Rathbun, ' Bull. U.S. Fish Comm.,' vol. 2, p. 124, 1900.

Locality: — Muttuvaratu Paar (Station VI.), 1 specimen, 12 millims.

7
Rostrum reaching to the end of the antennular peduncle and having a formula ~.

The antennular peduncle has the 1st joint broad and much flattened and equal in
length to the sum of the 2nd and 3rd joints, which are cylindrical. The fiagella are
slightly longer than the peduncle. The scale of the antenna is as long as the
rostrum. Thoracic legs are mostly missing, but the specimen agrees closely with
Bate's description.

Up to the present this species appears to have only been recorded from the
Atlantic.

General distribution : — Bermudas, West Indies, Ceylon.

Urocaris, Stimpson, 1860.

Urocaris longicaudata, Stimpson — Plate I., fig. 5.

Urocaris longicaudata, Stimpson, ' Proc. Ac. Nat. Sci. Phil.,' XII., p. 39, 1860.
Urocaris longicaudata, Rathbun, ' Bull. U.S. Fish Com.,' vol. 2, p. 126, 1900.

Locality : — Aripu Paar (Station LXIV.), 1 specimen. Female bearing eggs.
Dimensions : —

Total length from tip of rostrum to end of telson . .
Length of carapace and rostrum along mid-dorsal line .

„ ,, along mid-dorsal line

„ 1st abdominal segment along mid-dorsal line

2nd

3rd

4th

5th

6th
,, telson

32 millims.

8

>>

4-5

))

2

))

2

m

5

3

>)

3

J>

4

))

4

)3

MACRUEA. 79

The rostrum is straight and slightly arched, semi-transparent except on the ventral
side, which is strengthened by a thick ridge. There are nine dorsal teeth, the
posterior of which is a little remote from the others and is situated on the carapace.
There are two minute teeth at the tip of the rostrum on the ventral side. The
rostrum reaches almost to the end of the 2nd segment of the antennular peduncle.
The antennular peduncle has the 1st joint broad and flattened and equal to the sum
of the 2nd and 3rd. There are two fiagella, the outer of which is thicker and shows
signs of bifurcation at its distal extremity. The flagella are slightly longer than the
peduncle. The antennal peduncle is half as long as the 1st joint of the antennular
peduncle. The flagellum is about as long as the body. The antennal scale is slightly
longer than the antennular peduncle. The eye stalks are long and the eyes project
laterally. The carapace 1 tears on its anterior margin a well-developed spine below
the eye and also a smaller spine ventral to this. There is also a large hepatic spine.
Running along the carapace are two slight grooves. The dorsal groove starts behind
the antennal spine and extends half the length of the carapace. The ventral groove
starts at the anterior ventral border and traverses the entire length of the carapace in a
sinuous manner. The first two pairs of legs are chelate, the second pair being longer
and stouter than the first. The last three pairs are long and slender and have the
dactylos biunguiculate. The abdomen is more than three-fourths the length of the
body and is suddenly bent at right angles at the 3rd segment, the dorsal part of
which is much swollen. The dorsal side of the last three abdominal segments forms
a straight line. The (3th segment is long, being about one and a half times as long
as the 5th, and equal in length to the telson. The telson is slightly shorter than the
uropods and ends in two spines.

General distribution : — Atlantic coasts of North America, Ceylon.

Family : LATREUTID^.
Nauticaris, Sp. Bate, 1888.
Nauticaris grandirostris, n. sp. — Plate I., fig. 6.

Locality :— Galle (Station XXXV.), 2 males.

The carapace has a prominent antennal spine and a spine at the antero-lateral
border, as well as a spine on the anterior border half-way between these two spines.
The median dorsal surface of the carapace is occupied by well-marked teeth, which
are continued on to the rostrum. There is a well-developed rostrum, two-thirds as long
as the carapace. The rostrum is deep and is turned upwards at the tip. There are
six teeth occupying the whole of the mid-dorsal line of the carapace and the posterior
half of the rostrum. Of these teeth the posterior one is very small and not easily
made out. The others are well developed, the anterior tooth being slightly smaller
than the rest. Only the first two of these teeth are situated on the rostrum. The
anterior half of the rostrum bears no dorsal teeth. The extremity is marked by
three small teeth. On the ventral side there are six teeth, the posterior four being

80 CEYLON PEARL OYSTER REPORT.

exceptionally well marked and deep. The antennular peduncle is short, and is only-
half the length of the rostrum. Its proximal joint is the largest, being equal to the
sum of the other two. There are two branches to the flagellum. The inner branch
reaches slightly beyond the rostrum, and the outer branch, which is slightly shorter,
is much thicker and plumose, and shows signs of bifurcation at the tip. The
antennular scale is almost as long as the peduncle. The antennal peduncle is as long
as that of the antennule. The flagella are broken in both specimens. The scale is
stout and reaches to the end of the rostrum. The mouth parts are similar in all
important respects to those of Nauticai'is manonis.

The 3rd maxillipeds reach past the end of the rostrum and have the distal joint
tipped by about five well-marked spines. The proportionate lengths of the joints are
similar to those of N. marionis. There are only four joints, and the second bears a
large spine at its distal end. The 3rd joint is richly setose on its anterior face. The
legs are robust, excepting the 2nd pair. The 1st pair are chelate and reach a little
past the end of the antennal peduncle. The 2nd pair are long and slender and have
a multi-articulate wrist. They reach to the extremity of the rostrum, the 3rd, 4th,
and 5th pairs are similar to one another in form, decreasing slightly in length from
before backwards. The 5th leg reaches to the base of the antennal scale. The
carpos of each of the last three pairs bears a blunt process at the anterior distal
border. The dactylos ends in two larger spines, and bears several smaller spines on
its posterior border. The abdomen is robust and bent at right angles at the
3rd segment. There are two spines on the ventral side of each of the first four
abdominal segments, and on the ventral side of the 5th and 6th segments there is a
long median spine pointing backwards. The 6th abdominal segment has a well-
marked movable spine at each of its posterior lateral borders, and also bears two
well-marked spinous processes on its posterior border overhanging the telson. The
uropods are slightly longer than the telson. The 2nd abdominal segment has a small
transverse groove in the mid-dorsal line.

The telson tapers somewhat posteriorly and has a slightly grooved dorsal surface
carrying two pairs of spines. The posterior border bears two pairs of spines and
numerous long hairs.

Dimensions (measured along mid-dorsal line) : — ■

From tip of rostrum to end of telson . . . . 50 millims.

Rostrum and carapace 20 ,,

Carapace 11 ,,

1st abdominal segment 2 ,,

^nrl 4

3rd ,, ,, 8

4th „ „ 51 „

6th ,, ,, 3a ,,

Telson 5|- ,,

MACRURA. 81

Nauticaris unh'ecedens, Sp. Bate.

Nauticaris unirecedens, Sp. Bate, '"Challenger" Macrura,' p. 608, 1888.
Localities: — Galle (Station XXXV.), 5 specimens; Jokkenpiddi Paar, 1 specimen.
These specimens agree closely with Sp. Bate's description, but the rostral formula
is slightly different. The Ceylon specimens are not so richly toothed as the
" Challenger " specimens.

Rostral formula - — .

2

Average length 35 millims. (3 males and 3 females).
General distribution : — Hong Kong, Ceylon.

Nauticaris futilirostris, Sp. Bate — Plate II., fig. 8.

Nauticaris futilirostris, Sp. Bate, '"Challenger" Macrura,' p. 606, 188S.
Locality: — Galle Harbour (Station XXXVI.), 4 specimens taken in the tow-net.
Average length 1 1 millims.

These specimens agree with Spence Bate's description ; the posterior rostral tooth
is slightly more remote from the others than in Bate's figure.
General distribution : — Off Japan, Ceylon.

Latreutes, Stimpson, 1860.

Latreutes ceylonensis, n. sp. — Plate II., fig. 7.

Locality : — Cheval Paar (Station LIIL), 1 specimen, 8 millims. long.

The latero-anterior edge of the carapace is furnished with 4 spines at each side.
There is a deep rostrum, almost equal in length to the carapace. The carapace and
rostrum together ecpial half the length of the body. The rostrum dips slightly
downwards and bears dorsally two prominent teeth above the eyes and a tooth of
equal size slightly behind the pointed anterior extremity. In addition to these there
are about a dozen smaller teeth on the dorsal side which are only detected under a
microscope. The under side of the rostrum is smooth. Each antennular peduncle is
short, the proximal joint being as long as the sum of the other two. Each antennule
has two flagella which reach to the end of the rostrum, and are slightly longer than
the peduncle. The inner flagellum is slightly thicker and shorter than the outer.
The antennal peduncles are slightly longer than those of the antennules. The
antennal scale is large and broad, and extends slightly beyond the rostrum. Each
scale is furnished with half-a-dozen small spines along its outer border. The antenna!
flagella are missing.

The first two pairs of thoracic legs are chelate. The 2nd pair are longer than the
1st pair and have the wrist 3-jointed. The remaining three pairs of legs are more
strongly made, the meros and carpos being very broad. The dactylos terminates in
two very robust spines.

M

82 CEYLON PEARL OYSTER REPORT.

The 3rd, 4th, and 5th abdominal segments each has the posterior part of its dorsal
surface raised into a blunt keel. The abdominal segments are all subequal. The
telson is long and narrow, and is twice as long as the 6th abdominal segment. It
ends posteriorly in a blunt median spine, and is furnished with a lateral spine half-
way along each side. The uropods are as long as the telson.

Family: ALPHEID.E.

Synalpheus, Sr. Bate, 1888.

Synalpheus gravieri, Coutiere.

Synalpheus gravieri, H. Coutiere, 'Fauna of Maldives and Laccadives,' p. 870, 1905.

Localities : — Pearl banks, Gulf of Manaar (Station III.), 7 specimens ; Cheval
Paar (Station XLVIIL), 19 specimens; Coral reef, Gulf of Manaar (Station L1V.),
2 specimens; west of Periya Paar (Station LXL), 2 specimens, "commensal with
dendritic Alcyonarian ;" off Mutwal Island (Station LXVIL), 2 specimens.

The rostral and orbital spines are equal in length, the latter sometimes being
slightly divergent. The rostrum does not reach the end of the 1st joint of the
antennular peduncle. The 1st joint of the antennular peduncle is equal to the sum
of the other two, the 2nd joint is twice as long as the 3rd. The antennular scale is
a little longer than the 1st joint of the jDeduncle. The antennal peduncle is one and
one-third times as long as the antennular peduncle. The outer spine of the antennal
scale is as long as the antennular peduncle.

The 3rd pair of legs has a small spine on the dorsal side of the dactylos. The
meros has 4 spines on its ventral posterior border. The propodite has about 8 spines
on its posterior border.

General distribution : — Maldives, Ceylon.

Synalpheus laticeps, Coutiere.

Synalpheus laticeps, Coutiere, 'Fauna of Maid, and Lace.,' p. 874, 1905.

Localities: — Coral reef, Gulf of Manaar (Station LIV.), 1 specimen; deep water off
Galle (Station XL.), 4 specimens.

Orbital spines equal in length to rostrum, but slightly broader. Rostrum shorter
than 1st joint of antennular peduncle. First joint of antennal peduncle longest,
the 2nd and 3rd joints subequal. Antennular scale a little longer than the 1st joint
of the antennular peduncle. The antennal peduncle one and one-fifth times as long
as the antennular peduncle. Antennal scale as loug as antennal peduncle.

Third legs having the dactylos biunguiculate, posterior border of propodite armed
with about a dozen spines.

This species approaches closely to S. biung uiculatus.

General distribution : — Maldives, Ceylon.

MACRUBA. 83

Synalpheus biunguiculatus, Stimpson.

Synalpheus biunguiculatus, Stimpsox, ' Proc. Acad. Phil.,' p. 31, 1860.
Localities : — Deep water off Galle (Station XL.), 2 specimens ; Coral reef, Gulf of
Manaar (Station LIV.), 14 specimens; Chilaw Paar (Station LXIX.), 1 specimen.

Tin's species differs principally from the preceding species in (1) the rostrum and
orbital spines being shorter in comparison with the 1st joint of the antennular
peduncle ; (2) the antennal scale is shorter, and only reaches to the end of the
antennular peduncle ; (3) the posterior border of the telson has a deeper curve.
General distribution : — Maldives, Philippines, Pacific, Ceylon.

Synalpheus carinatus, de Man — Plate II., fig. 9.

Synalpheus carinatus, de Man, ' Arch. f. Naturg.,' I., 1887.
Synalpheus carinatus, Coutiere, ' Ann. des Sci. Nat.,' (8), t. ix., 1899.

Locality : — Deep water off Galle (Station XLL), 2 specimens.

The rostral spine is more than twice the length of the orbital spines and reaches
nearly to the end of the 1st segment of the antennular peduncle. The rostrum and
the orbital spines point slightly upwards. The 1st segment of the antennular
peduncle is slightly longer than the 2nd and twice as long as the 3rd. The
antennular spine reaches to the end of the 1st peduncular segment. The antennal
peduncle is slightly longer than the antennular peduncle. The antennal scale reaches
to the end of the antennular peduncle. The large chela is on the left side. The
carpopodite has a strong ventral spine. The dorsal side of the palm ends in a fairly
prominent spine, immediately in front of the digits. This spine seems more strongly
developed than in the type species.

In the 3rd pair of legs there is a spine on the ischiopodite, the meros is almost as
long as the carpos and propodite combined, and bears 8 well-developed spines on the
posterior border. The carpos bears a single spine at the distal end of its posterior
border. The propodite has 8 spines on the posterior border. The dactylos is bifid.

In the male each of the abdominal pleura is produced posteriorly into a spine. In
the female the pleura are rounded. The telson is as long as the uropods and tapers
slightly. It carries two pairs of spines on the dorsal side and there are 3 spines at
each of the posterior corners.

General distribution : — Indian Ocean.

Synalpheus comatulorum, Haswell.

Synalpheus comatulorum, Haswell, ' Cat. Austr. Crust.,' p. 187, 1882.
Synalpheus falcatus, Sp. Bate, ' " Challenger" Macrura,' p. 574, 1888.

Localities: — Navakaddua Paar (Station LXVIIL), 4 specimens; Gulf of Manaar
(Station LIV.), 3 specimens; Chilaw Paar (Station LXIX.), 1 specimen "on
Antedon" ; south end of Cheval Paar (Station XL VIII.), 1 specimen.

Rostrum twice as long as the orbital spines, and reaching to end ol the 2nd

M 2

84 CEYLON PEAEL OYSTER REPORT.

segment of antennular peduncle. Antennular scale reaching past the middle of the
2nd peduncular joint. The antennal peduncle is longer than the antennular peduncle.
The antennal scale is as long as the antennular peduncle.
General distribution : — Australia, Maldives, Ceylon, Pacific.

Synalpheus neomeris, de Man.

Synalpheus neomeris, de Man, ' Zool. Jahr.,' 9. Bd., p. 734, 1897.
Locality: — Gulf of Manaar (Station III.), 1 specimen.

I have referred the single specimen to the above species. It appears to differ but
slightly from Synalpheus gravieri.

General distribution : — Red Sea, Bay of Bengal, Madagascar, Australia, Japan,
Maldives, Pacific, Ceylon.

Alpheus, Fabricius, 1778.

Alpheus idiocheles, Coutiere.

Alpheus idiocheles, H. Coutiere, 'Fauna of Maid, and Lace.,' p. 883, 1905.

Locality : — Muttuvaratu Paar (Station VI.), 3 specimens.

Carapace deep. Rostrum short, and separated from the orbits by deep grooves.
The orbits are well formed and unarmed. In the antennule the 1st and 3rd segments
of the peduncle are subequal, each being half as long as the 2nd segment. The
antennular scale reaches to the middle of the 1st peduncular segment. The antenna
is short, only reaching three-fourths of the way along the 2nd joint of the antennular
peduncle. The antennal scale is nearly as long as the antennal peduncle. The large
chela is peculiar, having the dactylos portion of the digit hammer-shaped, and the
propodite part short as in Alpheus malleodigitatus. The 3rd and 4th legs are robust
and have the propodite armed with 5 spines, and end in a simple dactylopodite.

General distribution : — Maldives, Ceylon.

Alpheus phrygianus, Coutiere.

Alpheus phrygianus, Coutiere, 'Fauna of Maid, and Lace.,' p. 886, 1905.

Localities: — Gulf of Manaar (Station III.), 2 specimens; Cheval Paar (Station
XLVIII.), 2 specimens.

Somewhat similar to A. idiocheles. The eyes are not so prominent. The antennae ■
are comparatively longer, reaching to the end of the 2nd segment of the antennular
peduncle. The antennal scale reaches to the middle of the 2nd segment of the
antennular peduncle. The digits of the large chela are similar in shape to the
preceding species, but the palm is broader.

General distribution : — Maldives, Ceylon.

Alpheus paraculeipes, Coutiere.

Alpheus paraculeipes, Coutiere, 'Fauna of Maid, and Lace.,' p. 894, 1905.
Locality : — Cheval Paar (Station XLVIII.), 1 specimen.

MACRURA. 85

Rostrum poorly developed. Orbits unarmed. 1st and 3rd segment of the
antennular peduncle equal. 2nd segment twice as long as each of the others.
Antennular scale not reaching to the end of the 1st segment. Antennal peduncle
one and one-fourth times as long as the antennular peduncle. Spine of antennal
scale as long as antennular peduncle.

The 3rd legs have the ischiopodite unarmed. The posterior border of the mero-
podite is fringed with about 20 very delicate spines and ends distally in a strong
spine. The carpopodite has on its posterior external border 1 spine and about
5 hairs, and on its internal border about 15 long, fine spines. The propodite has
7 pairs of spines on its posterior border and is fringed with hairs anteriorly. The
dactylos is slightly biunguiculate.

General distribution : — Maldives, Ceylon.

Alpheus spongiarum, Coutiere.

Alpheus spongiarum, Coutiere, 'Fauna of Maid, and Lace.,' p. 895, 1905.
Locality : — Cheval Paar (Station XL VIII.), 1 specimen.

This species is very closely allied to A. paraculeipes, but differs in the form of the
3rd pair of legs. The meropodite is not so stout as in the latter species. Along the
posterior border of the meropodite there are 7 very long hairs with about 1 5 short
hairs between. The carpopodite has no spine on the posterior border and the
dactylos is not bifid.

General distribution : — -Maldives and Ceylon.

Alpheus paralcyone, Coutiere.

Alpheus paralcyone, Coutiere, 'Fauna of Maid, and Lace.,' p. 895, 1905.

Locality: — Off Mutwal Island (Station LXVII.), 2 specimens.

Rostrum is well defined and slightly carinated behind. The 1st and 3rd segments
of the antennular peduncle are subequal, and the 2nd segment is one and a half
times as long as each of the others. Antennular scale very small, antennal peduncle
one and one-third times as long as the antennular peduncle. The spine of the
antennal scale reaches past the end of the antennular peduncle. The palm of the
large chela is massive, narrowing distally, and the digits are very short.

In the 3rd pair of legs the ischiopodite is armed with a single spine, the meropodite
is large and ends distally at the posterior border in a large spine. The carpopodite
has about 4 spines on its posterior border, and the propodite has 8 pairs of spines.
The dactylos is bifid.

General distribution :— Maldives, Ceylon.

Alpheus miersi, Coutiere.

Alpheus miersi, Cuutiere, 'Fauna of Maid, and Lace.,' p. 903, 1905.

Alpheus rapax, var. miersi, Coutiere, ' Bull. Soc. Entom. de France,' No. 7, p. 166, 1898.

Localities : — Pearl banks, Gulf of Manaar (Station III.), 3 specimens; Cheval Paar

86 CEYLON PEARL OYSTER REPORT.

(Station XLVIIL), 3 specimens; Coral reef, Manaar (Station LTV.), 2 specimens ; oft1
Mutual Island (Station LXVIL), 1 specimen.

Rostrum well developed, reaching to the end of the 1st segment of the antennular
peduncle. The 2nd segment of the antennular peduncle is slightly longer than the
1st or 3rd, which are subequal. The antennal peduncle and scale are equal in length
and slightly longer than the antennular peduncle. In the Ceylon specimens the
large chela is massive and the palm is broad. The meropodite is spiny on its inferior
border.

The 3rd pair of legs has the ischiopodite with a single spine. The meros is smooth
except for a small spine at the posterior distal border. The carpos is smooth. The
propodite bears five pairs of spines. The dactylopodite is half the length of the
propodite and slightly curved.

General distribution : — Pacific, Japan, Maldives, Ceylon.

Alpheus pareucheirus, Coutiere.

Alpheus pareucheirus, Coutiere, ' Fauna of Maid, and Lace.,' p. 906, 1905.

Locality : — Haul 6, south of Adam's Bridge (Station LIV.), 1 specimen.

The antennules and antennse do not differ greatly from those of the preceding
species. Their peduncles are not so stout. The rostrum is only two-thirds as long as
the 1st segment of the antennular peduncle. The large chela differs in having the palm
grooved on both sides. The meropodite is smooth. In the 3rd legs the ischiopodite
does not bear a spine. The meropodite is smooth and not so robust as in the previous
species. The carpos is smooth and the propodite bears 8 long spines on the posterior
border. The dactylopodite is long and slightly curved.

General distribution : — Maldives and Ceylon.

Alpheus bis-incisus, var. malensis, Coutiere.

Alpheus bis-incisus, var. malensis, Coutiere, ' Fauna of Maid, and Lace.,' p. 910, 1905.

Locality : — Cheval Paar (Station LXIX.), 1 specimen.

The rostrum is triangular, and separated from the orbits by two well-marked
grooves. It reaches to the middle of the 1st segment of the antennular peduncle.
The 1st and 2nd segments of the antennular peduncle are subequal and slightly
longer than the 3rd. The antennular scale almost as long as the 1st peduncular
segment. The peduncle and scale of the antennal are equal in length and slightly
longer than the antennular peduncle.

I have placed this specimen in Coutiere's variety merely on the characters of the
rostrum and antennae. The large chela and most of the legs are missing.

General distribution : — Maldives and Ceylon.

Alpheus bis-incisus, var. stylisostris, Coutiere.

Alpheus bis-incisus, var. stylisostris, Coutiere, 'Fauna of Maid, and Lace.,' p. 911, 1905.
Locality : — Coral reef, Gulf of Manaar (Station LIV.), 1 specimen.

MACRURA. 87

This specimen appears to differ mainly from the variety malensis in the form of the
rostrum, which is much narrower in proportion to its length.
General distribution : — Maldives and Ceylon.

Alpheus audouini, Coutiere.

Alpheus edwardsi, Audouin (see Coutiere, 'Fauna of Maid, and Lace.,' p. 911).
Alpheus audouini, Coutiere, 'Fauna of Maid, and Lace.,' p. 911, 1905.

Localities: — Off Mutwal Island (Station LXVIL), 2 specimens; Pearl banks, Gult
of Manaar (Station III.), 6 specimens ; Coral reef, Manaar (Station LIV.), 3 specimens.

This species, which Coutiere has separated from Alpheus edwardsi, resembles the
latter species in the form of the rostrum and in the appearance of the antennae. The
large chela has the dorsal and ventral projections of the palm rounded, thus differing
from those of A. edwardsi, which are spiny.

General distribution : — Red Sea, Indian Ocean, Malay Archipelago, New Zealand,
Sandwich Islands.

Alpheus macrodactylus, Ortmann.

Alpheus macrodactylus, Orthann, ' Zool. Jahrb.,' V., p. 473, 1890.

Locality : — Off Mutwal Island (Station LXVIL), 1 specimen. Related to
A. edwardsi and A. euphrosyne.

The rostrum is well pronounced and more than half as long as the 1st segment of
the antennular peduncle. Of the joints of the antennular peduncle the 2nd is the
longest, being twice as long as the 3rd and nearly twice as long as the 1st. The
antennular scale is broad and reaches nearly to the end of the 1st segment of the
peduncle. The antennal peduncle and scale are equal in length and extend as far
forward as the end of the antennular peduncle. The large chela differs from that of
A. edwardsi in the absence of a dorsal spine and in the comparatively greater length
of the digits.

General distribution : — Australia, Ceylon.

Alpheus laevis, Randall.

Alpheus laevis, Randall, ' Journ. Acad. Nat. Sci. Phil.,' vol. viii., p. 141.

Localities: — Galle (Station XXXV.), 4 specimens; Coral reef near Galle (Station XL),
5 specimens.

Well-developed rostrum reaching to the end of the 1st segment of the antennular
peduncle. Orbits armed with two small spines. Segments of the antennular peduncle
subequal. Antennular scale slightly longer than the 1st peduncular segment.
Antennal peduncle and scale equal to one another and slightly longer than the
antennular peduncle. The large chela has a massive palm, laterally compressed, with
no dorsal or ventral notches. The carapace is deep.

General distribution : — Indian Ocean, Pacific, Australia.

88 CEYLON PEAEL OYSTER REPORT.

Athanas, Leach, 1813.
Athanas nitescens, Leach.

Athanas nitescens, Leach, ' Edin. Encycl.,' vol. vii., p. 432, 1813.
Athanas veloculus, Sp. Bate, ' " Challenger " Macrura,' p. 529, 1888.

Locality : — Cheval Paar (Station XLVIL), 1 specimen.

This specimen clearly belongs to the well-known species in which must be
included — according to Coutiere — Bate's species Athanas veloculus.

This record is of interest, inasmuch as the distribution of this species up to the
present has been limited, so far as I can ascertain, to the Atlantic and Mediterranean.

General distribution : — Atlantic coasts of America, Cape Verd Islands, north-west
Europe, Mediterranean, Ceylon.

Athanas orientalis, n. sp. — Plate II., fig. 10.

Localities : — Cheval Paar (Station XLVIIL), 2 specimens ; Muttuvaratu Paar
(Station VI.), 1 specimen; Coral reef, Gulf of Manaar (Station LIV.), 2 specimens ;
west of Periya Paar (Station LXL), 2 specimens.

This species is in many respects very closely allied to Athanas dimorphus,
Ortmann, and A. minikoensis, Coutiere, but there are differences in the form of the
extra- and infra-orbital spines, as well as in the form of the 1st pair of legs which
lead me to place it in a new species.

The rostrum extends as far as the end of the 2nd joint of the antennular peduncle.
The infra-corneal spine reaches slightly beyond the eye, and the extra-corneal spine
just reaches to the anterior end of the eye, so that it is difficult to make out in side
view. There is no supra-corneal spine.

The antennule has the 3 joints of its peduncle subequal, and its scale reaches as
far forward as the end of the 2nd peduncular joint and the tip of the rostrum. The
autennal peduncle reaches to the end of the 2nd joint of the antennular peduncle, and
its scale, which is very broad, reaches slightly beyond the end of the antennular
peduncle. In the small leg of the female the carpopodite and the propodite are about
equal in length, but the latter is more robust. The meropodite is one and a half times
as long as the carpopodite. The ischiopodite has a long delicate spine at the distal end
of its dorsal border, and there are five smaller spines along the same border. This
is the only specimen bearing the 1st pair of legs, so that it is not possible to
compare these appendages in the male.

This species differs from the two allied species in the length of the rostrum and
also in the relative lengths of the extra- and infra-orbital spines, as well as in the
length and robustness of the joints of the small leg. The other two species are
devoid of spines on the ischiopodite of that limb.

MACRURA. 89

Family : CRANGONIDiE.

iEgeon, Guerin-Meneville, 1835
(= Egeon, Risso, L816).

JEgeon cataphractus (Olivier).

Cancer cataphractus, Olivier, 'Zool. Adriatica,' p. 30, 1792.

Egeon loricatus, Risso, 'Crust, de Nice,' p. 100, 1816.

Crangon cataphractus, Milne-Edw., 'Hist. Nat. Crust.,' vol. 2, p. 343, 1837.

JEgeon cataphractus, Ortmann, 'Zool. Jahrb.,' vol. 5, p. 535, 1890.

Locality : — Galle (Station XXXV.), 1 specimen.

This single specimen agrees closely with the type species, and also with Henderson's
species, /Egeon orientalis,* in most respects. The latter species differs from the ahove
species only slightly with regard to the teeth present on the carapace, and in the
absence of the small hepatic groove on the sides of the carapace. Since the Indian
species was formed from the characters of a single specimen, it is by no means
improbable that this specimen merely represents an extreme variation of the parent
species. In all other characters, excepting the two above mentioned, /Egeon orientalis
agrees with /Egeon cataphractus.

General distribution : — Mediterranean, Senegambia, South Africa, Ceylon.

Family: PROCESSID^E.

Processa, Leach, 1815
(=Nika, Risso, 1816).

Processa canaliculata, Leach.

Processa canaliculata, Leach, ' Malac. Podoph. Brit.,' p. 641, 1815.
Nika edulis, Risso, 'Crustaces de Nice,' p. 85, 1816.
Nika canaliculata, Desmaret, 'Consid. gen. Crust.,' p. 231, 1825.
Nika hermudensis, Rankin, 'Ann. N.Y. Acad. Sci.,' XII., p. 536, 1900.

Localities: — Pearl banks, Gulf of Manaar (Station III.), 6 specimens; Muttuvaratu
Paar (Station VI.), 1 specimen; 10 miles west of Cheval (Station XLVIL), 2 speci-
mens; Cheval Paar (Station XLVIIL), 4 specimens; south-east of Modragam
(Station LXLV), 2 specimens.

None of these specimens appear to differ in any marked degree from the ordinary
characters of the species.

General distribution : — Madeira, Japan, West Indies, Gulf of Mexico, South Africa,
North-west Europe, Ceylon.

* J. R. Henderson, ' Trans. Linn, Soc.,' 2nd series (Zoology), vol. v., part 10, p. 446, and plate 40,
figs. 16 and 17, 1893.

N

90 CEYLON PEARL OYSTER REPORT.

Tribe : SCYLLAPJDEA.

Family: SCYLLARLD.E.

Scyllarus, Fabricitjs, 1793
(= Arctus, Dana, 1852).

Scyllarus tuberculatus (Sp. Bate).

Arctus tuberculatus, Sp. Bate, ' " Challenger" Macrura,' p. 70, 1888.

Localities : — Pearl banks, Gulf of Manaar (Station III.), 1 specimen, female ; Galle
(Station XXXV.), 3 females and 1 male ; Aripu Reef (Station LXIV.), 1 specimen.

Characterised by large tuberculations on the mid-dorsal line of the carapace and of
the first 3 pairs of abdominal segments. Those of the second abdominal segment are
very distinctive of the species.

In this genus all the legs of the male end in a simple dactylos, and in the female
there is a poor developed chela on each of the 5 th legs. The propodite digit is not
very well developed.

General distribution : — Australia and Ceylon.

Scyllarus sordidus (Stimpson).

Arctus sordidus, Stimpson, ' Proc. Acad. Nat. Sci., Phil.,' p. 8, 1860.

Locality : — Coral reef, Gulf of Manaar (Station LIV.), 3 females and 1 male.
General distribution : — Pacific and Ceylon.

Tribe: THALASSINIDEA.

Family : CALLIANASSIDiE.

Callianassa, Leach, 1813.

Callianassa rotundicaudata, Stebbing.

Callianassa rotundicaudata, Stebbing, 'Marine Invest, of S. Africa.' Crust., ii., p. 41, 1903.

Locality: — Cheval Paar (Station XLVIIL), 1 specimen, 18 millims. long.
General distribution : — South Africa, Ceylon.

Callianassa maldivensis, Borradaile.

Callianassa maldivensis, Borradaile, 'Fauna of Maid, and Lace.,' vol. ii., part 3, p. 753.

Locality: — Gulf of Manaar (Station III.), 1 specimen, 24 millims. long.
This specimen is imperfect, but it agrees closely with the above species.
General distribution : — Maldives and Ceylon.

MACRURA. 91

Upogebia, Leach, 1813.

Upogebia intermedia (de Man).

Gebiopsis intermedia, de Man, ' Journ. Linn. Soc.,' vol. 22 (Zool.), 1888.

Localities : — Haul 6, south of Adam's Bridge (Station LIV.), 2 males and 1 female ;
Muttuvaratu Paar (Station VI.), 1 male.

This species is characterised by the possession of a large number of denticulations
on the cephalic portion of the cai*apace, and also by the anterior thoracic legs being
richly clothed with long and very fine setse.

The anterior portion of the carapace covers the small eyes and projects almost to the
end of the antennular peduncle. The peduncle of the antennule is 3-jointed, the 3rd
being slightly longer than the 1st and two and a half times as long as the 2nd. The
3rd joint is more slender than the other two. There are 2 flagella one and a half times
as long as the peduncle. The inner flagellum is slightly longer and less robust than the
outer. The antennal peduncle is slightly longer than that of the antennule and is also
3-jointed. The 1st and 2nd joints are equal and slightly longer than the 3rd joint,
which is also less robust than the other two. The 1st and 2nd joints are richly
clothed with long fine seta;. The eyes are small and project nearly to the end of the
2nd antennular peduncle. The middle of the carpos of the 1st legs reaches to the
end of the rostrum. The appendages agree with de Man's description. The
abdomen is large, being twice as long as the carapace and proportionally broad.
The segments are subequal, the 2nd and 6th being slightly longer than the others.
The telson is broader than long and equal in length to the uropods.

General distribution : — Mergui, Ceylon.

N 2

92

CEYLON PEARL OYSTER REPORT.

EXPLANATION OF THE PLATES.
PLATE I.

Fig. 1. Penccus indicus, Milne-Edw., head, side view, x 4.

2. Parapenceus mogiensis, Rathbun, thelycum in Ceylon specimen, x 8.

3. „ gallensis, n. sp., head, side view, x 5.
3a. „ „ „ thelycum. x 4.
3b. ,, „ „ petasma. x 10.

4. Caradina vitiensis, Borradaile, rostrum, side view, x 12.

5. Urocaris Imgicaudata, Stimpson, head, from above, x 7.
5a. „ „ rostrum and carapace, side view, x 7.

6. Naulicaris grancliroslris, n. sp., side view, x 3.
6a. „ ,, rostrum, side view, x 3.
6b. „ „ head, from above, x 5.
6c. ,, ,, telson, from above, x 3.

PLATE II.

7. Latreutes ceyhnensis, n. sp., rostrum and anterior edge of carapace, x 65.
7a. ,, „ head, from above, x 65.

7b. „ ,, 3rd maxilliped. x 36.

7c. „ „ 2nd thoracic leg. x 36.

7d. „ „ 3rd thoracic leg. x 36.

7e. ,, „ telson, dorsal view, x 28.

8. Nauticaris fvMliroslris, Sp. Bate, 4th thoracic leg. x 30.

9. Synaljpheus cwinatus, de Man, head, from above, x 12.
9a. „ ,, 3rd thoracic leg. x 54.

9B. ,, „ telson and right uropod. x 20.

10. Athanas arientaiis, n. sp., head, from above, x 30.
10a. „ „ side of head, x 30.

10b. „ „ 1st thoracic legs; small chela, x 25.

CEYLON PEARL OYSTER REPORT

MACRURA- PLATE I.

FIG. I.

FIG.2.

FIG. 3b.

FIG. 6c-

J. P <M • E WUson.Cajnbndge

Fig. 1, Pen.ecs indicus, M. Edw. ; Fig. 2, Parapen.eus mogien'sis, Rathbun ; Fig. 3, Parapen.eus gallensis, n. sp. ;
Fig. 4, Caradixa vitiensis, Borradaile ; Fig. 5, Urocaris lonoicaudata, Stimpsou ; Fig. 6, Nauticaris grandirostris, n. sp.

CKYLON PEARL OYSTER REPORT

MACRURA PLATE II

FIG. 7

FIG. 7b.

FIG.IOb.

FlG. 7, LATREDTE8 CEYLONENSIB, II. s]>. ;

E Wilson .Cambridge

Fig. 8, Nauticaris futilirostris, Sp. Bate ; Fig. 9, Synalpheus carixatos, de Man
Fig. in, Athanas orientalis, n. *\>.

VKYLON PEARL OYSTER FISHERIES— 1905— SUPPLEMENTARY REPORTS, No. XXV.]

REPORT

ON THE

ANTIPATHARIA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY

Professor J. ARTHUR THOMSON, M.A., University of Aberdeen..

AND

JAMES J. SIMPSON, M.A., University of Aberdeen.

[With ONE PLATE.]

The collection of Ceylonese Antipatharians 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. The localities are more precisely referred to in
Professor Herdman's " Narrative" in Part I. of the General Report (1903).

The collection is a small one, including thirteen species, but nine at least of these
seem to be new. The list is as follows : —

Family: ANTIPATHID.E.
Sub-family: CIRRIPATHIN^.

*Cirripathes (?), n. sp.

Sub-family : ANTIPATHIN^.

* Antipathes gallensis, n. sp. Stichopathes gracilis, Gray, var. spiralis, nov.

* Antipathes gracilis, n. sp.t Stichopathes echinulata, Brook.
Antipathes abies, Gray. * Antipathella rugosa, n. sp.

* Stichopathes ccylonemis, n. sp. * Antipathella elegans, a. sp.

*Stichopathes contorta, n. sp. * Antipathella irregularis, a. sp.

* Stichopathes papillosa, n. sp. * Antipathella ceylonensis, n. sp.

* Those marked with an asterisk are reported as new.

t Non Antipathes gracilis, Gray (1860) = Antipathella gracilis, Gray;

Non Antipathes gracilis, Koch (1889) = Antipathes mediierranea, Brook (1889).

94 CEYLON PEAEL OYSTER REPORT.

Before proceeding to the systematic report, we would make a few general
observations : —

(a) In the detailed classification of Antipatharia much importance has been
attached to the form and distribution of the spines on the axis. But it does not
seem to have been sufficiently emphasised that there is considerable variation in both
of these characters in the different parts of the colony. Thus in the branched forms the
nature of the spines, the number seen from one aspect, and the arrangement of these
in spirals or longitudinal rows in no way correspond on the larger branches and on
the pinnules (Plate, fig. 2). In the simple colonies this is even more emphasised, e.g.,
in Cirripathes {?) (Plate, fig. 8) those at the base are arranged irregularly (a), those
about the middle of the colony have a distinct linear arrangement (6), while those
near the tip are disposed in whorls around the stem (c). This distinction has been
illustrated relative to the species described, in the figures on the Plate, where two
views, and in one case three, have been given. This is of great importance where
species are determined from fragmentary specimens.

(b) This difference between stem, branches, and pinnules is also borne out with
regard to the size and shape of the polyps. In many of the specimens, on the stem
and larger branches, the polyps are almost circular and disposed irregularly, their
diameter being less than that of the axis, while on the pinnules they are elongated
and rectangular, exceeding in breadth the diameter of the axis. Their distribution
also varies in the different portions of the colony ; they are often separated by
considerable intervals on the older parts, while on the pinnules they may be closely
apposed. The length and degree of transparency of the tentacles does not seem to
be a safe criterion, varying as it does with the degree of contraction and the mode of
preservation.

(c) Of some general interest and deserving further investigation is the extra-
ordinary modification shown in the spines of several species. The general Anti-
patharian spine is simple or papillose, but in some species they pass from an
elongated sinuous form through a series of gradations to an antler-like growth and
eventually simulate a tree-like structure (Plate, fig. 1, a-f). This has been
previously noted, e.g., by Carter for Antipathes spinosa (Carter), but no inter-
pretation has been suggested. It may be due to a pathological condition, and in
some cases where branched spines were observed it was noted that a sponge-like
growth surrounded the axis, or it may be the result of irregular regeneration of
broken spines. It is particularly well seen at the base of Antipaihes gracilis, n. sp.,
and on Antipathella rugosa, n. sp. In Antipathes abies, Gray, a forking of the spines
was occasionally observed.

(d) The polyps of Antipathes abies, Gray, which were unknown when Mr. George
Brook described the " Challenger " Antipatharia, are very well preserved in some of
Herdman's specimens, and we have therefore given a full descrij^tion and figures of
their external features (Plate, fig. 4).

ANTIPATHARIA. 95

(e) A fact which may yet prove to be of some importance is that in Stichopathes
papillosa, n. sp., belonging to a genus typically simple, a knob-like projection, about
6 centims. from the base, indicates, without doubt, the remains of a branch.

(f) The various specimens show a considerable number of epizoic animals: — e.g.,
Sponges, Polyzoa, Serpula-tubes, Spirorbis-tubes, Cirripede-galls, stalked Barnacles,
Corals, and in one case a young pearl oyster.

Cirripathes (?) n. sp. — Plate, fig. 8.

A very large simple colony, 135 centims. long, with a diameter varying from
3*75 millims. at the base to 0"75 millim. at the top. The basal portion, which is
attached to a stone, is expanded into a circular disc 16 millims. in diameter. The
stem is sinuous for the first 35 centims., but after that it is coiled into three distinct
spirals, with diameters of 10 centims., 9 centims., and 8 centims. respectively — the
corresponding heights being 10 centims., 8 centims., and 7 centims. The total height
of the colony is 65 centims. The colour of the axis at the base is jet black, changing
gradually to a golden brown. It is hollow, at least in the upper region, and is
covered with distinctly papillose spines, which are 0-l millim. in height near the
top of the colony, but shorter and thicker further down. They are arranged
irregularly near the base, where twenty can be counted from one aspect, but further
up a linear arrangement seems to predominate. Near the top they are disposed in
verticils round the stem, about one and a half to two lengths apart, and the
number from one aspect diminishes to nine. They are of a paler colour than
the stem.

As there are no polyps on the specimen, it is impossible to decide its position with
security, beyond saying that it is either a Stichopathes or a Cirripathes, but the
arrangement of the spines and the general nature of the colony would point towards
its being a new species of Cirripathes, which we would refrain from naming. The
specimen was trawled at Station XXIV., off Foul Point, outside Trincomalee,
46 fathoms.

Antipathes gallensis, n. sp. — Plate, fig. 15.

A fragmentary portion of a colony, 8 centims. high and 4 '5 centims. broad. The
branching is irregular, giving the whole a shrub-like appearance, suggestive of the
broom. The main stem is short and sinuous, but a large branch arises about midway
up and constitutes the greater part of the colony. The general colour is black
towards the base, but rusty brown in the smaller branches, which are long and
slightly flexed. The diameter of the axis is 1 millim. at the base and tapers very
gradually.

The spines on the main branch are low and conical, standing perpendicularly and
arranged irregularly, so that no definite number could be counted from any one
aspect. Those on the smaller branches are compressed and thorn-like — the upper

96 CEYLON PEARL OYSTER REPORT.

margin being sub-horizontal, while the lower is convex. They are -comparatively
short, about one-third the diameter of the branch, and are disposed in fairly steep
sinistrorse spirals and longitudinal rows, those in a row being about two lengths apart.
The rows do not consistently alternate, but a quincuncial arrangement is infrequent.
Five can be clearly seen from one aspect, while the tip of another is visible, seven
making a complete revolution.

The polyps on the stem are large and circular, with a low truncated oral cone
and prominent mouth opening, which is also circular. The tentacles are arranged
radially and slightly distant on the branches, the polyps are arranged in a single
longitudinal row and are elongated in the direction of the axis — this being specially
marked on the smaller branchlets. They are very large and prominent, measuring
T5 millims. in length. The projection bearing the circular oral opening is large and
cylindrical. The tentacles when fully expanded are very long, but in most of the
polyps they are contracted — being then thick set and conical. They are disposed
in three pairs, the sagittal pair being inserted low down in the polyp. On the larger
branches the distance between the polyps is equal to about one-half their length, but
this diminishes considerably on the branchlets, where the polyps are more elongated
though still of the same general character.

This species differs from any known form both in its mode of branching and in the
arrangement of the spines.

Locality : — Station XLI. , deep water oft' Galle.

Antipathes gracilis, n. sp. — Plate, figs. 7 and 14.

A small, complete, delicate whin-like colony, 6 centims. high and 1 centim. in
diameter, attached by a small expansion. It consists of a main axis, 1 millim.
in diameter at the base, tapering gradually to a fine point. The lower 13 millims. of
the stem are devoid of branches. On the next 2 centims. small branches bearing
pinnules arise sub-horizontally from three sides. These are almost straight and taper
to a point, the longest being 8 millims. The remainder of the stem bears branches
arising on all sides, but apparently in no definite order. The branches gradually
diminish in length towards the apex of the colony. The colour of the axis is golden
yellow when seen with transmitted light.

The spines on the bare part of the axis are slightly elongated, compressed, and
triangular, the upper margin being sub-horizontal, the lower convex. They are
arranged irregularly, about nine being visible from one aspect, at intervals of about
one length. Many of the basal spines show an antler-like or dendriform mode of
branching. On the upper part of the stem they are disposed in more regular
longitudinal lines, five being now visible from one aspect. On the pinnules they are
still of the same type, but more elongated, and with a greater slope towards the axis.
They seem to be arranged irregularly, but a closer examination reveals a hint of a

ANTIPATHARIA. 97

steep dextrorse spiral. They are about one and a half lengths apart, and seven is a
typical number from one aspect.

The polyps are all arranged so as to face in one direction, and it is worthy of note
that this is away from the bare portion of the axis mentioned before. On the stem
they are disposed irregularly and are somewhat circular, the tentacles being inserted
almost equidistant from the mouth opening. The oral projection is very prominent
and cylindrical, and the mouth opening is circular. On the branches and pinnules
they are very much elongated, and the tentacles are disposed in three pairs, the
sagittal pair being inserted slightly below the level of the others ; but apart from
this the structure is much the same as in those on the stem. On the branches
they are separated by intervals about equal to their breadth, but on the pinnules this
distinct demarcation disappears.

This species approaches most closely to Antipathes spinosa (Carter) (Hydraden-
drium spinosum, Carter), but differs from it both in the mode of branching and in
the character of the spines. In A. spinosa (Carter) the polyps had not been
observed when Mr. Brook described the " Challenger" Antipatharia.

Locality : — Deep water off Galle.

Antipathes abies, Gray — Plate, fig. 4.

Several very fine specimens of this species are included in the Ceylon collection.
All are of the bottle-brush, or, more correctly, fir-tree type.

One magnificent colony (A) is 65 centims. in height, the breadth varying at the
different parts. At 20 centims. from the base the diameter is 15 centims., but this
gradually diminishes to 10 centims. at 40 centims. from the base, and tapers almost
to a point at the top of the colony. It is attached to a stone, and for the first
1 1 centims. the axis is bare. Above this there are about a hundred principal branches
of varying sizes. The colour of the axis is black, but owing to the thin ccenenchyma
it presents a greyish surface. The branches have a superficial rusty or reddish-
brown tint, getting paler towards the top of the colony. The diameter of the axis at
the base is 5 millims. ; it gradually tapers upwards.

A smaller colony (B) is also very perfect and compact. It is 30 centims. high and
15 centims. in diameter about half way up. From that point it ascends in a
symmetrical cone. From the first 9 centims. of the stem the branches have been
broken off, but the knob-like ends have been quite overgrown by the ccenenchyma.

The axis is 7 millims. in diameter at the base and tapers to a point. The colour is
identical with that of (A).

The main stem is slightly curved and the branches are longest on the concave
surface, so that in contour the colony is symmetrical.

The mode of branching is by no means regular. At some places there are signs of
a spiral arrangement, but this is often interrupted by extra offshoots. The branches
arise very close together, often only 2 millims. apart. They are mostly in planes at

o

98 CEYLON PEARL OYSTER REPORT.

right angles to the main stem, but some are turned upwards and others downwards,
interlocking, so that no two overlap. At their point of origin they are about
2 millims. in diameter and present the crescentic shape characteristic of the species.
A typical branch from the concave side of the stem has a chord of 10 centims., a
perpendicular height of 2 centims., and a breadth of 5 centims. The branches bear
branchlets, and even secondary and tertiary branchlets, extending in a plane at right
angles to the long axis of the branch. The secondary branches arise in a distinctly
alternate manner, the planes bearing them 'enclosing an angle of 60°. The branches
do not all curve in one direction, but for the most part they diverge in pairs, so that
the tips of two approximate, enclosing an ellipse. The secondary branches arise in a
similar manner, so that the maximum of surface is exposed on the contour of
the colony.

The spines vary greatly in the different parts of the colony. On the black main
stem they are disposed very irregularly, and are very small and abundant. This is
also the case on the paler branches, and owing to the conical form the whole gives
the impression of a moss-rose stem. They are horny in colour and have a black
broadened part where they arise from the stem. The smaller branchlets are trans-
parent and hollow, being of a golden-brown colour with a faint reddish tinge. Here
the spines are much longer, being bluntly conical and pointing slightly upwards.
They are arranged in distinct longitudinal lines, which in reality are the result of
steep sinistrorse spirals, five or six being seen from one aspect. They are almost
equal to the radius of the pinnule and are about one length apart.

The polyps, which were unknown when Mr. Brook described the "Challenger"
Antipatharia, are of two kinds, according to the position in the colony. On the
main stem and larger branches they are scattered irregularly on the concave surface,
being thus within the general network of the circumference. They are visible to the
naked eye and appear as six-rayed stars. They are almost circular, the tentacles
being disposed on six radii. The tentacles vary considerably in size, according to the
state of retraction. On the pinnules the polyps are arranged on the convex surface
and thus all face outwards. They are more rectangular in shape than on the stem,
being elongated in the direction of the long axis of the pinnule. The tentacles are
disposed in three pairs, those in the sagittal axis being inserted very far down and
standing mostly perpendicular to the polyp. The distance between the polyps is
approximately the same as the length of a polyp, viz., 0'9 millim. In all cases the
mouth is situated on a prominent cylindrical projection, the oral cavity being
elongated in the direction of the sagittal axis.

Cirripede galls are of frequent occurrence, and these are overgrown with the mud-
coloured ccenenchyma, also bearing spines. Numerous barnacles are attached to the
larger branches.

Another almost complete colony (C), without the basal attachment, is 22 centims.
in height and 9 centims. in diameter at the widest portion. The main stem is bent

ANTIPATHARIA. 99

so as to form two arcs. The length of the branches on the concave surface of the
stem greatly exceeds that on the convex, so that the contour is symmetrical.

The spines are typical both as to size and arrangement. A few are bifurcated,
but as this is only of local occurrence it does not justify the dignity of a new species.
The polyps are also typical, but in some, owing to contraction, the tentacles are
very inconspicuous.

A beautiful complete colony (D), closely resembling a young larch tree, was
attached to a stone by a disc-like expansion. It is 30 centims. in height, the greatest
width being 10 centims. The first 10 centims. are bare, owing to the branches
having been broken off, and the next 6 centims. bear branches only on one side.
The diameter of the axis at the base is 2 millims. The spines and polyps are typical.
A great number of barnacles are attached to the branches.

Another complete colony (E) was more of the bottle-brush type. It is 13-5 centims.
high, with branches down to the very base — breadth 6 "5 centims. The diameter ot
the axis at the base is 1*5 millims., tapering to 0"5 millim. The branching is not so
regular as in the others, but in no case do the branches overlap. The spines and
polyps agree closely with the typical forms, but the colour of the branches is slightly
darker.

Localities : — Station LXIIL, west of Periya Paar, in the Gulf of Manaar,
40 fathoms ; and Station XXIV., off Foul Point, outside Trincomalee, 46 fathoms.

Stichopathes ceylonensis, n. sp. — Plate, fig. 9.

A small, complete, simple colony attached to a piece of stone by an almost spherical
expansion. It is 8 "5 centims. long, but only reaches a height of 6 centims. owing to
its spiral course.

The diameter at the base is 1 millim., and this diminishes to 075 millim. at the tip
of the colony, so that the tapering is very slight. The stem is translucent, golden
brown near the base, becoming paler upwards ; it is hollow throughout its entire
length. The first 4 centims. are straight, followed by two open sinistrorse spirals
1 "3 centims. in diameter.

The spines near the base are short, triangular, and compressed, standing at right
angles to the stem, disposed irregularly, but mostly one to two lengths apart. Four
may be seen from one aspect. On the upper half of the colony the spines are of the
same type, but considerably longer in proportion to the thickness of the stem, being
equal to about one-third of the diameter. They are arranged in steep sinistrorse
spirals and longitudinal rows about two to three lengths apart. Four can be
distinctly seen from one aspect, while the tips of two others are visible, so that there
are eight altogether in a spiral.

The polyps are typical and prominent. The tentacles are very long and transparent,
and the sagittal pair are inserted almost diametrically opposite. They are separated
by a distance of about one-half the length of a polyp. Towards the top of the colony

O 2

100 CEYLON PEARL OYSTER REPORT.

they are alternately large and small — the smaller forms being probably younger.
They are also separated by greater intervals.

This specimen comes nearest to S. pourtalesi, Brook, but cannot be identified with
it. It differs, for instance, in not having " crowded r polyps, and the arrangement of
the spines is also different.

Locality : — Station LV., outside the pearl banks, Gulf of Manaar.

Stichopathes contorta, n. sp. — Plate, fig. 3.

A simple slender colony, 40 centims. long, attached to a piece of rock. It is
very sinuous, growing first upwards, then coiling and turning downwards, again
twisting and starting on an upward course. Thus the total height is only 7 centims.,
and the growing point is but 3 centims. above the base. The diameter of the axis
is 1 millim. and is uniform throughout. The colour is blackish with a brown tinge,
the axis is hollow down to the disc of attachment.

The spines are of a pale horny colour, and are slightly but distinctly papillose.
They are arranged in longitudinal rows in the lower portion about two to two and
a half lengths apart, seven being seen from one aspect. Further up, a distinct steep
spiral arrangement may be seen — seven being required to form one revolution. Those
in one longitudinal row are about two lengths apart.

The polyps are arranged on one side of the axis at intervals of about 1 millim.,
which is also the length of a polyp.

The oral cone is prominent and the mouth opening circular. The tentacles are about
0-5 millim. in length even in a contracted state. Young polyps are frequent between
the larger older forms.

This species is nearest S. lutkeni (Brook), but differs from it both in the number
and arrangement of the spines.

Locality : — From off Galle and onwards up West Coast of Ceylon.

Stichopathes papillosa, n. sp. — Plate, figs. 6 and 13.

A complete, simple, robust colony attached to a piece of rock. It is 38 centims.
long and attains a height of 18 centims.

The first 4 centims. are almost straight, the remainder coiled into ten distinct
dextrorse spirals, 13 millims. in diameter and averaging 14 millims. high. The axis
is l-25 millims. in diameter at the base and tapers gradually to 0-5 millim. at the
top. At a distance of 6 centims. from the base there is a projection which indicates
the remains of a branch. The colour is black at the lower part, becoming lighter
towards the apex.

The spines are slightly but distinctly papillose, and vary in number in the different
parts of the colony ; thus at the base fourteen can be counted from one aspect, whereas
at the top only ten and points of two are visible. Those near the base are conical and
covered throughout their whole length with small papillse, but on the upper part of

ANTIPATHARIA. 1 0 1

the colony the papillae are confined to the apex of the spines, which in this region
are more flattened and triangular. They are arranged in distinct dextrorse spirals,
and are about one length apart.

The coenenchyma is very thick on the side devoid of polyps.

The polyps are about 1 millhn. in diameter and almost form a square — the directive
tentacles being inserted at the corners and the sagittal pair at a slightly lower level.
The tentacles are very long when expanded, but on contraction form low, broad
cones. The oral cavity is large and elliptical, being elongated transversely. The
oral cone is large and prominent. The polyps in the upper part of the colony are
close together, but lower down they are separated by a distinct groove. Smaller
polyps, probably young forms, are of not infrequent occurrence.

The lower part of the stem is covered with Polyzoa.

The distinctive features of this new species are : the thickness of the coenenchyma,
the papillose character of the spines, together with the dextrorse spiral arrangement,
and the distance between the spines as compared with their length.

Locality : — Deep water off Galle and onwards up West Coast of Ceylon.

Stichopathes gracilis, Gray, var. spiralis, nov.

A very slender, simple colony, 58 centims. in length. Owing to its sinuous and
spiral course, it only attains a height of 20 centims. It is attached by a broadened
basal expansion. The first 5 centims. are straight ; succeeding this there is a
sinuous portion 17 centims. long, followed by two distinct spirals 6 centims. in
diameter and about 6 centims. high. The diameter at the base of the stem is
1 minim, and this measurement scarcely diminishes even at the tip. The colour
appears black, but when viewed with transmitted light has a decided reddish-brown
tinge. The stem is hollow to the very base.

The spines near the base are very much damaged, but appear to be short, some-
what flattened cones, irregularly arranged, a comparatively small number (about 5)
being visible from one aspect. About the middle portion of the colony they are
short and triangular, standing perpendicularly to the axis, and disposed in irregular
longitudinal rows and dextrorse spirals. The spines in a longitudinal row are
separated by about two lengths. Six can be counted from one aspect.

Near the top of the colony the spines are much smaller and inclined to the axis,
the upper margin being concave and the lower convex. They are arranged in very
steep dextrorse spirals, nine in one spiral being seen from one aspect.

The distance between any two varies greatly — from two to four lengths. The
polyps are arranged in a single longitudinal row, and are very large and prominent,
measuring about 2 millims. in length. They are slightly elongated, the oral cone
being low but distinct. The tentacles are disposed in three pairs, the sagittal pair
being inserted at a considerably lower level than the others. They are very
long when fully expanded, but when contracted are low, broad and conical ; others are

102 CEYLON PEARL OYSTER REPORT.

like large spheres with a small filiform projection. They arise from somewhat
spherical bases. The polyps are separated by a very small interval, but an annular
constriction between each pair seems to pass round the stem.

This specimen agrees on the whole with Stichopathes gracilis, but as it differs in
some details regarding the spines, and conspicuously in having a spiral course, it has
seemed convenient to name a new variety.

Locality : — Deep water off Galle.

Stichopathes echinulata, Brook.

This species is represented by a simple colony, incomplete at the base, 26 centims.
long, very irregular and sinuous, so that the total height is only 8 centims. The
growing point is turned downwards and is only 5 centims. above the lowest portion.
The colony tapers very markedly from 1/5 millims. to 0"5 millim. The colour of the
axis is almost black.

The spines are very short, compressed and directed upwards. They are arranged
in very steep spirals, the distance between two rows being almost the same as that
between any two on a spiral, so that they show a quincunx grouping. The
distance between any two is equal to about four lengths of a spine.

The polyps are typical, but the sagittal tentacles are relatively distant from the
oral cone. They are separated by a distance about two-thirds of the length of the
polyp.

This species has been previously recorded from Mauritius.

Locality: — Station LX., outside the pearl banks, Gulf of Manaar.

Antipathella ragosa, n. sp. — Plate, figs. 5 and 11.

There are two specimens of this new species in the collection, both slightly
damaged. The larger of these is 19 centims. high and 13 centims. broad. It is
branched mostly in one plane, and consists of two main branches which arise
dichotomously from a short main stem 1 centim. long. One branch arises at 60°,
while the second after a short distance at right angles to the stem bends upwards
and runs closely parallel to it. The first is broken off about 7'5 centims. from its
origin, and the second at a slightlv lower level. At the point of fracture large
secondary branches are given off, and it is noteworthy that the angles of inclination
are the same as for the first two, viz., 60° and 90°. The primary and secondary
branches give off pinnae in a strictly alternate manner, and these again bear pinnules.
The pinnse converge slightly. The whole plane of branching is slightly curved and
the polyps arise on the convex surface.

Near the base the axis is opaque and black in colour, but this passes gradually
into a transparent horny yellow in the upper parts of the colony. The smaller
branches and pinnules are hollow.

The spines on the large branches are short and slender, tapering in a marked

ANTI PATH ARIA. 103

degree. They are slightly inclined to the axis, and are arranged irregularly, about
fifteen being visible from one aspect. Those on the pinnules are very thin and
delicate, fairly long, conical in shape, and inclined to the axis at an acute angle.
They are arranged in distinct longitudinal lines which are the expression of steep
dextrorse spirals. Five can be seen from one aspect. The distance between two in a
longitudinal line is equal to about two lengths.

On the main branches the polyps are arranged irregularly, and are almost circular,
the tentacles, which are moderately long, being equidistant from the oral cavity.
On the pinnules the polyps are arranged on the convex surface and are elongated in
the direction of the axis. The distance between the polyps varies. In some places
they are crowded together, while in others they are separated by a distance equal to
their breadth. In all cases the body of the polyp is large. The oral cavity is
circular and borne on a very prominent cylindrical projection. The tentacles are
large and inclined outwards, being very rugose in appearance, due probably to the
state of retraction. They are arranged in three rows of two each — the sagittal
tentacles beino- inserted far down.

The colony bears numerous epizoic animals : — Cirriped galls and stalked barnacles,
tubes of Spirorbis, several Polyzoa, a Sponge, and a young pearl oyster shell. It is
worthy of note that the majority of these are overgrown by the ccenenchyma and
bear both polyps and spines.

A second specimen of this species — also slightly damaged — is 14-5 centims. in
height and 8 centims. in breadth. It consists of a main stem with a basal attachment
from which three branches arise on one side at about 60°. These are slightly arched,
and the longest, which is 12 centims. in length, has a diameter of 1 millim.,
tapering to a point. The whole colony is flabellate. In its spines, polyps, and colour
it agrees with the other specimen.

This species should be included in Brook's Group A (''Challenger" Eeport), but
it does not approach closely to any of species already included in that group.

Localities : — Deep water off Galle, and Station VIII., deep water, in Gulf of Manaar.

Antipathella elegans, n. sp. — Plate, fig. 10.

A complete, very graceful colony, 13 centims. in height, with a maximum breadth
of 55 centims. at a distance of 9-5 centims. from the base, which is expanded into a
discdike attachment. At a distance of 3 centims. from the base the main stem
bifurcates and the two subsidiaries develop almost equally. The mode of branching
is not uniform. The general appearance is dichotomous, but this breaks down in
several places where three or four branches arise on one side. The stem and branches
are black at the base, gradually passing into golden brown near the apex. All the
branches and pinnules are hollow. The diameter of the branches varies very little in
the different parts, and the gentle tapering gives the whole colony a very graceful
appearance. The axis is 1 millim. in diameter at the base. The spines are short and

104 CEYLON PEARL OYSTER REPORT.

very much compressed, being somewhat triangular with a very broad base. The
upper margin is sub-horizontal, while the lower is convex. They are arranged in very
steep sinistrorse spirals and longitudinal rows, the distance between two in a row
being four lengths, while that between two in a spiral is one length. From one
aspect five can be seen quite distinctly along with the tips of other two, making in
all eight spines in a circumference. The spines near the base are shorter, smaller and
more conical. They are disposed in sinistrorse spirals and distinct rows, the distance
between two in a row being about four lengths.

The polyps are typical. They are situated in a single row on the branches and
branchlets, and are very much elongated in the direction of the axis. The distance
between the polyps varies in the different parts ; they are in some places close together,
in others separated by intervals equal to half their length. Very often a line passing
through the oral cones is a line of spines, and the polyps occupy a length corre-
sponding to four spines in a longitudinal row. The oral cone is very prominent and
the mouth opening is circular. The tentacles vary in different parts according to
their state of contraction. The sagittal pair are inserted rather far down, corresponding
to spines 1 and 5 on a circumference. In some cases the tentacles have spherical
terminations.

This species is chiefly distinguished by the nature and arrangement of the spines,
which are markedly different from those of other species, but also by the polyps,
which, though typical of the genus, nevertheless bear specific characters.

Locality : — Station LX. , outside pearl banks, Gulf of Manaar.

Antipathella irregularis, n. sp. — Plate, fig. 12.

This species is represented by a small complete colony and a fragment. The
former is 4 centims. in height and 6 centims. in breadth, the general shape being
sub-flabelliform. The branches arise mostly in two planes, but occasionally in a third,
leaving one quadrant bare. Frequent fusions occur.

The other specimen consists of a short main stem with a disc of attachment at the
base. The stem is only 2 centims. high, and the axis tapers from 1 millim. at
the base to a very fine point at the apex. The branches are longer than the main
stem and constitute the greater part of the colony. They arise on three sides at a
very large angle, so that the expansion is mostly lateral.

The colour of the axis near the base is black, but it passes through a dark amber
to a horny yellow in the branchlets. The secondary branches are somewhat elongate
and slender and appear slightly flabellate.

The spines near the base of the stem are short, conical, and irregularly disposed ;
but on the branches a definite arrangement can be traced. There they are compressed
and triangular in form, arranged sometimes in a dextrorse and sometimes in a
sinistrorse spiral. They are about two lengths apart and five can be seen from one
aspect.

ANTIPATHARIA. 105

The polyps are rather small and are disposed on one side of the stem and branches,
so that none appear on the quadrant devoid of branches. They are oval in shape,
being elongated in the direction of the axis. The oral cone is low but definite ; the
mouth is scarcely discernible. The tentacles are comparatively long and delicate,
having a broad base and tapering markedly. The distance between the polyps varies,
but in most cases there is a valley-like depression between them, giving the surface an
undulating appearance.

The distinctive features of the species are the irregular mode of branching, the
character and arrangement of the spines, and the nature of the polyps.

Locality : — Station XXIV., off Foul Point, outside Trincomalee, 46 fathoms.

Antipathella ceylonensis, n. sp. — Plate, fig. 2.

Of this species there are two specimens — one complete colony and a broken part of
another.

The former is a small, graceful, delicate colony, complete but for the tips of some
of the branches. It is 7 centims. in height and 5 centims. in breadth. The main
stem is 5 centims. long, and the axis is about 1 millim. in diameter above the disc of
attachment, which is 1 centim. broad. The branching is approximately in one plane,
and there are signs of fusion in three places. About 1 centim. from the base two
branches arise, almost of the same diameter as the main stem. One of these is
4 centims. long, the other is 6 centims. and bears a comparatively large secondary
branch. The branching is irregular, but nearly alternate. The branchlets are very
slender and arise at different angles, very seldom at an angle less than 60°, and most
frequently at right angles.

The spines near the base are short, conical, and distant ; those on the pinnules are
larger and thorn-like, the lower margin of the compressed triangle being convex.
They are equal to about half the diameter of the pinnule and are about three lengths
apart. The arrangement is in steep sinistrorse spirals, four being seen from one
aspect.

The polyps are disposed on one side of the branches and are elongated in the
direction of the long axis. The tentacles are short and are arranged in three pairs,
the sagittal pair being inserted at a level slightly lower than the others. The oral
opening is circular, and is elevated on a prominent cylindrical projection.

This species comes nearest to A. tristis (Duch.), but is distinguishable from it both
in spines and polyps.

Locality: — Station XXIV., off Foul Point, outside Trincomalee, 46 fathoms.

106 CEYLON PEARL OYSTER REPORT.

LITERATURE REFERRED TO.

1891. Bell, F. Jeffrey. — "Contributions to our Knowledge of the Antipatharian Corals." 'Trans.

Zool. Soc. Lond.,' vol. xiii., part ii., No. 7.
1900. Bourne, G. C— " Anthozoa" in E. Ray Lankester's ' Treatise on Zoology,' part 2.
1889. Brook, G. — " Antipatharia," ' " Challenger" Keport,' vol. xxxii.
1880. Carter. — ' Ann. Mag. Nat. Hist.,' ser. v., vol. 5.

1904. Forster-Cooper, C. — "Antipatharia" in 'The Fauna and Geography of the Maklive and

Laccadive Archipelagoes,' part iii., vol. 2.
1857. Gray, J. E.— ' Proc. Zool. Soc. Lond.,' p. 291.
1860. Gray, J. E— ' Ann. Mag. Nat. Hist.,' ser. iii., vol. 6, p. 311.
1877. Klunzinger, C. B. — ' Die Korallthiere des Rothen Meeres.'
1883. Ridley, Stuart O.—" The Coral-fauna of Ceylon," &c. ' Ann. Mag. Nat. Hist.,' ser. iv., vol. 11.

1905. Roule, L. — " Antipathaires et Cerianthaires." 'R&ultats Camp. Seient. Prince de Monaco,'

Fasc. xxx.
1896. Schultze, L. S. — "Beitr. z. Syst. der Antipatharien." ' Abh. Senckenberg. Nat. Ges.,' Bd. xxiii.
1899. Schultze, L. S. — 'Die Antipatharien der Deutschen Tiefsee-Expedition, 1898-99.'
1869. Verrill, A. E. — ' Transactions of the Connecticut Academy,' vol. i.

EXPLANATION OF THE PLATE.

Fig. 1. Antipathella rugosa, n. sp., dendriform spines, «-/, different stages of growth.
„ 2. Antipathella ceylonensis, n. sp., arrangement of spines.

(a) Main axis ; (b) Axis of pinnule.
„ 3. Stichopathes contorta, n. sp., arrangement of spines.

(a) Near base of the colony ; (b) Towards the tip of the colony.
,, 4. Antipathes abies, Gray, polyps.

(a) On main branches ; (b) On axis.
,, 5. Antipathella rugosa, n. sp., polyps on pinnules.
„ 6. Stichopathes papillosa, n. sp., polyps near the tip of the colony.
,, 7. Antipathes gracilis, n. sp., complete colony. Nat. size.
,, 8. Cirripathes (?), n. sp., arrangement of spines.

(a) Near base of colony ; (b) Middle of colony : (c) Near tip of colony.
„ 9. Stichopathes ceylonensis, n. sp., arrangement of spines.

(a) Near base of colony ; (b) Towards top of colony.
,, 10. Antipathella elegans, n. sp., arrangement of spines.

(<f) Lower part of axis ; (b) Near top of axis.
,, 11. Antipathella rugosa, n. sp., arrangement of spines.

(a) Near base of main stem ; (b) Part of a pinnule.
,, 12. Antipathella irregularis, n. sp., arrangement of spines.

(a) Main stem and branch ; (/>) Pinnule.
,, 13. Stichopathes papillosa, n. sp., arrangement of spines.

(a) Near the base of the axis ; (5) At the tip of the axis.
,, 14. Antipathes gracilis, n. sp., arrangement of spines.

(a) On main stem ; (b) Part of a pinnule.
„ 15. Antipathes gallensis, n. sp., arrangement of spines.

(a) Main stem and branch ; (b) Pinnule.

CKYLON PEARL OYSTER REPORT

ANTIPATHARIA PLATE

E. Wilson, Cambridge.

Fig. 1, Antipathella rugosa, n. sp. ; Fia. 2, Antipathella ceylonensis, u. sp. ; Fig. 3, Stichopathes contorta, n. sp. ;

Fig. 4, Antipathes abies, Gray; Figs. 5 and 11, Antipathella rugosa, n. sp. ; Figs. 6 and 13, Stichopathes papillosa, n. sp.

Figs. 7 and 14, Antipathes gracilis, n. sp. ; Fig. 8, Cirripathes, n. sp. ; Fig. 9, Stichopathes ceylonensis, n. sp. ;

Fig. 10, Antipathella elegans, n. sp. ; Fig. 12, Antipathella irregularis, n. sp. ; Fig. IS, Antipathes gallensis. n. sp.

[CEYLON PEARL OYSTER FISHERIES— 1905-SUPPLEMENTABY REPORTS, No. XXVL]

REPORT

ON THE

POLYZOA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY
LAURA ROSCOE THORNELY

(ZOOLOGY DEPARTMENT, UNIVERSITY OF LIVERPOOL).

[With ONE PLATE.]

INTRODUCTORY.

There are at least 116 species of Polyzoa iu this collection from Ceylon. Of these,
31 had already been found in Indian seas, 32 of the remainder have been reported
from Australian waters, 13 from the China Sea, and several from neighbouring but
outlying waters to the east, west, and south of the Indian Ocean. Of the rest,
some have not been, I believe, recorded from nearer than Florida (7 species), the
Queen Charlotte Islands (l), and the Mediterranean Sea (2). Several are cosmopolitan
in their distribution, and 19 are British species. Finally there are 16 species and
1 variety which I consider to be new, and which will be described below as : —

OnychoceUa cucullata, n. sp. Lepralia nitida, n. sp.
Sch'r.oporella arirularis, n. sp. ,, ceylonica, n. sp.

„ viridis, n. sp. „ fissa, n. sp.

„ collaris, n. sp. ,, triangula, n. sp.

Rhyncopora incisor, a. sp. Smittia trispinosa, var. proteda, n.

„ corrugata, n. sp. Phyladella spiralis, n. sp.

Gemellipora prolrusa, n. sp. Retepora podlhim, n. sp.

Lepralia mtdtidentata, a. sp. Cellepora compada, n. sp.
„ purpurea, a. sp.

Polyzoa are well distributed round the coast of Ceylon. No less than 89 species
were found in the Gulf of Manaar, and 32 off Galle. Many of them were obtained in

P 2

108 CEYLON PEARL OYSTER REPORT.

several distinct localities and in great abundance, while other species are represented
by one colony each. From the large number of specimens obtained, and the manner
in which the colonies are crowded together on the foreign bodies to which they are
attached, it may be inferred that the Ceylon seas are a favourable locality for the
Polyzoa ; but at the same time, from the multiplication of avicularia and of spinous
processes of various sorts which characterise these Ceylonese specimens, the impression
is derived that there is severe competition and that the colonies have a struggle to
hold their own. It is possible that the avicularia, spines and other roughnesses on the
surface of the zoarium may protect such species from being smothei'ed by overgrowths
of colonies of their own kind ; and this seems a very necessary protection in this
case, and the need may account for some of the extraordinary calcareous outgrowths
of the zocecia which I have to describe below.

Comparatively few collections of Polyzoa have been made in Indian seas. Hincks
(16) reported in 1887 upon a collection made by Dr. Anderson in the Mergui
Archipelago. Kirkpatrick (22) described in ] 895 a collection made by Mr. Thurston
in the Gulf of Manaar. Among Hincks's series of papers, entitled " Contributions
towards a History of the Marine Polyzoa," in the ' Annals and Magazine of Natural
History ' (8), there is a report upon a small collection from Indian seas, and here and
there among other papers of this series may be found descriptions of a specimen or
two from Ceylon. In all, perhaps, 45 species have been previously described from
the seas around Ceylon. Of these, 31 are represented in the present collection ; and
85 additional species, including 16 new to science, are now recorded for the first time.

The work has been carried out chiefly in Liverpool, but it is a pleasure to record
the help that has been freely given to me in Cambridge and in London. Dr. S. F.
Harmer has kindly allowed me to consult his private collections of specimens and
notes, and Mr. P. Kirkpatrick has helped me with those at the British Museum ; to
both gentlemen I am indebted for advice, and suggestions. In conclusion, I should
like to express my gratitude to Professor Herdman for the privilege of being allowed
to handle and name his valuable collection of Polyzoa.

Order: ECTOPROCTA.

Sub-order : CHEILOSTOMATA.

Family: .ETEID/E.
JEtea anguina, Linn.

Localities: — Station I., off Negombo, 12 to 20 fathoms; and Station XLVI., off
Mount Lavinia, 7 to 12 fathoms.

Family: CATENARIID/E.
Catenaria lafontii (Auu).

Locality: — North of Cheval Paar, 7 to 10 fathoms.

POLYZOA. 109

Catenicella elegans, Busk (1).
Locality : Gulf of Manaar (attached to floating oyster cages).

Family: CELLULARIID.F,.
Scrnpocellaria cervicornis. Busk (2).
Locality : — Navakaddu Paar, Gulf of Manaar.

Scnvpocellaria diadema, Busk (2).

Localities: — North of Cheval Paar, in Gulf of Manaar, 7 to 10 fathoms; and
Station LXIIL, west of Periya Paar, 17 to 55 fathoms (large quantities).

Scrupocellaria scrupea, Busk (2).
Locality : — Gulf of Manaar.

Caberea retiformis (Pourtal^s).

The peculiar, unequal development of the fornix, mentioned by Miss Phillips (25),
is a characteristic of these Ceylon specimens. The upper half is developed into a
long, sharp process, while the lower is scarcely produced below the stem by which it
is attached to the zooecium. Avicularia are present both along the median line and
on the ocecia.

Locality : — Off Galle in deep water.

Family : BICELLAPJID/E.

Diploecium simplex, Ktrkpatrick (20).
Localities : — North of Cheval Paar, 7 to 10 fathoms ; oft* Manaar; and off Galle.

Bugula neritina, Linn.

There is one specimen in the present collection which differs from the others, being
lighter in colour, less rigid in growth, and with avicularia. Hincks (12) mentions
these last as present on some of his specimens, therefore there does not seem sufficient
reason for separating the two forms. The avicularia are large, placed on the outer
sides of the zooecia, and have long mandibles. Ooecia are present in the position usual
for the species.

Localities : — Palk Bay ; and Gulf of Manaar.

Beania mirabilis, Johnston.
Locality : — Cheval Paar.

Family : CELLARIID^E.
Cellaria johnsoni, Busk.

Localities : — Gulf of Manaar ; off Galle, deep water ; and Station XL., 10 miles oft'
Watering Point, 34 fathoms.

110 CEYLON PEARL OYSTER REPORT.

Nellia oculata, Busk (2).
Locality : — Gulf of Manaar.

Family : TUBUCELLARILLm

Tubucellaria cereoides (Ellis and Sol.).

Localities : — Off Galle and onwards up the West Coast of Ceylon, deep water ;
Navakaddu Paar, Gulf of Manaar (several colonies).

Family : MEMBRANIPORIDiE.

Membranipora favus, Hincks (8).

There are tubercles between some of the zocecia in the colonies of this species which
otherwise agree with Hincks's description of the species.
Locality : — Gulf of Manaar (on Conus shells).

Membranipora irregularis, D'Orbigny.
Locality : — Gulf of Manaar.

Membranipora hastilis, Kirkpatrick (21).
Locality : — Gulf of Manaar.

Amphiblestrum cervicorne, Busk (2).
Locality : — Off Manaar Island (several colonies).

Amphiblestrum papillatum, Busk (1).

Localities: — Station XLVL, off Mount Lavinia, 25 to 30 fathoms; Navakaddu
Paar, Gulf of Manaar.

Amphiblestrum granuliferum (Hincks, 3).
Localities : — Gulf of Manaar ; oft* Galle ; and oft" Mount Lavinia (several colonies).

Amphiblestrum marginella (Hincks, 8).

The lateral avicularia of these specimens are more pointed than in Hincks's
description, and are directed upwards instead of downwards. The large avicularia
occupying a whole zocecial area are present.

Locality : — Gulf of Manaar.

Amphiblestrum delicatulum (Busk).

The Ceylon specimens show the serrated denticle of this form and have no avicularia.
The zocecia are not quadrangular but of a diamond shape, resembling one of the zocecia
in Hincks's figure (3, plate xi.) of this species. There are no knobs such as Hincks
describes on his specimens. Ocecia are present, not described before ; they are sunk

POLYZOA. Ill

below the membrane of the zocecium, above the one to which they belong, are finely-
punctured and have a calcareous arch above the zocecial orifice. The colony is of a
brownish colour and adheres closely to the bivalve shell on which it is growing.
Locality : — Gulf of Manaar.

Siphonoporella bursaria (MacGillivkay, 23).

The Ceylon colonies of this species have avicularia with spatulate mandibles,
interspersed here and there among the zocecia, on separate areas. In old specimens
the front wall of the zocecium is distinctly punctured. With these slight differences
the specimens have all the appearance of Membranipora rossellii (Aud.) which
MacGilliyray (23, plate xxvi., fig. 4) later named Amphiblestrum bursarium.
There is even an indication, in his drawing of one of the zocecia, of a siphon, which is
clearly present in the Ceylon specimens. I believe the specimens undoubtedly belong
to MacGillivray's species.

Localities : — Off Galle ; off Trincomalee ; and from the Gulf of Manaar (several
good colonies).

Family : ONYCHOCELLID.E.

Onychocella antiqua (Busk).
Locality : — East Cheval Paar, Gulf of Manaar.

Onychocella abyssicola (Smitt, 26).
Localities :— Gulf of Manaar ; and off Galle at Station XL., 10 miles off Watering-
Point, 34 fathoms.

Onychocella cucullata, n. sp. — Plate, fig. 1.

Zoarium incrusting or erect, branching and bilaminar. Zocecia large and irregularly
oval, raised towards the upper end, where the margin is often coarsely beaded.
Operculum large and horse-shoe shaped, bent forward along with the raised portion
of the zocecium. Cryptocyst coarsely granular and punctured on the front wall,
descending as in Steganoporella lateralis, MacGillivkay (24), to the basal wall of
the zocecium and pierced by a tubular orifice which has an everted rim. Avicularia
curved over the top of the zocecium, usually branched on one side and having the tips
of the branches more or less forked. A triangular swollen area, probably ocecial, is
seen above some zocecia, having an oval opening covered by a membranous operculum,
situated in the centre of the beaded upper margin of the zocecium to which it belongs.

Localities : — Gulf of Manaar ; and off Trincomalee.

This species appears to be closely allied to Steganoporella in the form of zocecium,
but has a smaller operculum and avicularia of the Onychocella type. The two forms
present in this collection, incrusting and erect, have the same zocecial and avicularian
characters, and cannot therefore be separated. The erect form is represented by

112 CEYLON PEARL OYSTER REPORT.

broken specimens of about 2 centims. in height, flat and branched in yarious ways
and of a grey colour. The incrusting form has all the appearance of a Steganopurella
to the naked eye.

Family : MICROPORID^E.

Stegunoporella buskii, Harmer (17).

There is a large colony of this species measuring about 10 square centims. It is
growing in a loose honeycomb-like form and is of a remarkably light and brittle
texture. Another colony, the same in detail, is incrusting a seaweed and is much
smaller in every way.

Locality : — Cheval Paar, Gulf of Manaar.

Steganoporella sulcata, Harmer (17).
Locality : — Gulf of Manaar.

Steganoporella simplex, Harmer (17).
Localities : — Gulf of Manaar ; and oft" Trincomalee.

Thalamoporella rozieri (Aud.), form indica, Hincks (3).
Localities : — Gulf of Manaar ; Palk Bay ; and oft Galle.

Thalamoporella rozieri (Aud.), form falcifera, Hincks (3).
Locality : — Generally distributed round the Ceylon coast.

Family: CRIBRILINID.E.

Cribrilina radiata, Moll.

Localities : — Off* Galle ; Gulf of Manaar ; off Mount Lavinia ; and Station XL. ,
10 miles off Watering Point, 34 fathoms.

Family : MICROPORELLID^E.

Microporella violacea, Johnston, form plagiopora, Busk.
Locality : — Gulf of Manaar.

Microporella ciliata (Pallas).
Locality : — North of Cheval Paar, 7 to 10 fathoms.

Microporella ciliata, var. personata, Busk (2).

The avicularia on the Ceylon specimens have wing-like membranous extensions of
the sides of the mandible. Two of the colonies are of a pink colour.

Localities: — Gulf of Manaar; and Station XXXIII.. south-east of Ceylon,
18 fathoms.

POLYZOA. 113

Microporella decorata, Reuss.
Locality : — Gulf of Manaar.

Berenicea prominens, Lx. [= Chorizopora brongniartii (Aud.)].

Localities : — Gulf of Manaar ; off Galle, deep water ; and Station XL., off Watering
Point, 34 fathoms.

Family: PORINID^.

Porina magnirostris (MacGillivray, 23).
Locality : — Gulf of Manaar.

Gigantopora fenestrata, Ridley.

On the most perfect of the few, small, incrusting colonies of this form there are
little branched spines, in between the perforations of the front wall of the zocecia.
The secondary, tubular orifice has a 4-toothed margin and avicularia are very long,
reaching from the base of the tube and extending beyond its margin. Where an
avicularium is absent the tube has a fissure in its place. Ooecia are present, low
down at the back of the tubular orifice, and perforated like the front wall of the
zocecia.

Localities : — Gulf of Manaar ; and off Galle.

Lagenipora spinulosa, Hincks (8).
Locality : — Gulf of Manaar.

Lagenipora tuberculata, MacGillivray (23).

There are several colonies corresponding to MacGillivray's description of this
species, but there are also among them some that have the hollow tubercles very
much lengthened, ending in points, or being jagged and irregular in outline. These
specimens have also simple or branched spines round the margin of the much raised
peristome. The front walls of the zocecia have punctures in between the spines ; the
peristome is granular. There are no avicularia, and no ocecia are present.

Locality :— Off Galle.

Family : MONOPOEELLID^.

Monoporella albicans, Hincks (6).

An interesting point in the present example of this form is that the ocecia are in
most cases set a little " awry," a fact that Hincks thought merely a peculiarity of
the colony in his collection. There are none of the large avicularia present here.

Locality :— Gulf of Manaar.

Monoporella lepida, Hincks (5).
Locality : — Station LIIL, north of Cheval Paar, 7 to 10 fathoms.

Q

114 CEYLON PEAEL OYSTER REPORT.

Family: MYEIOZOIDiE, Smitt.
Schizoporella spongitis (Pallas).
Localities : — Gulf of Manaar ; and off Mount Lavinia.

Schizoporella ampla, Kirkpatrick (20).
Localities : — Gulf of Manaar ; and Navakaddu Paar.

Schizoporella argentea, Hincks (6).

In the single small colony of this species the sinus is wider than Hincks descrihes
and there are no oval avicularia, but in their place the elongated kind as in some of
his specimens. There are the two spines above and the granular, silvery surface of
the zooecia with foramina between the granulations.

Locality : — Gulf of Manaar.

Schizoporella aperta, Hincks (6).

There are two spines on the upper margin of some of the orifices of these specimens
not described before. No ocecia are present, but in other points the characters agree
with those of the species.

Localities : — Gulf of Manaar ; and Navakaddu Paar.

Schizoporella cecilia (Atjd.).

The one small colony of this sjaecies has raised, irregular-shaped ridges of calcareous
matter on the ocecium, and a decided arch above the orifice at the base of the ocecium,
also a screen-like process on the front of the zooecium.

Locality : — Gulf of Manaar.

Schizoporella unicornis, Johnston.

The present specimen has the peculiar long avicularia of the form longirostris,
Hincks (11), but not the more important feature of that variety, namely the loop-
shaped sinus.

Locality : — Navakaddu Paar.

Schizoporella nivea, Busk (1).

There are numerous colonies of this species, loosely attached, being, as it were,
folded round the stems of the large zoophyte, Campanularia juncea, and having the
opposed edges of the zoarium adhering to one another. (See Part II., p. 115, fig. 2.)

Locality : — Gulf of Manaar.

Schizoporella circinata, MacGillivray (23).

The zocecia of these specimens have the ridge-like mucro of Hincks's (10) species,
but not the avicularium. There is a single row of large punctures at the edge of the
zooecium. No ocecia are present on the two small colonies in the Ceylon collection.

Localities : — Off Mount Lavinia : and off Galle.

POLYZOA. 115

Schizoporella sanguinea (Norman).

Avicularia are usually present on these specimens. They are long and pointed
and raised, a pair on each zocecium, situated at their extreme upper angles, above the
orifice and directed downwards and inwards.

Locality :— Off Galle.

Schizoporella magnifica, Hincks (11).

The peculiar distinctive ocecia of this species are not present on the one colony in
this collection, which is deep red in colour. The long pointed sinus, the pairs of
upward pointing avicularia, and the reticulate front wall of the zocecium, without a
raised margin, accord with Hincks's description of the species.

Locality :— Off Galle.

Schizoporella depressa, Phillips (25).

There are some colonies of a pink colour corresjionding with the figure and
description of this species. The orifice of the zocecium is deeply sunk, the front wall
rising into almost an umbo below it, and calcareous ridges radiate from this to the
margin, leaving large, loop-shaped areolations between them. The small rounded
avicularia, on one or both sides of the orifice, are sunk below the surface, together
with the orifice. Ocecia are smooth and hyaline or ridged, and have an oval area on
either side of them.

Localities : — Gulf of Manaar ; and off Mount Lavinia.

Schizoporella triangula, Hincks (5).

The specimens of this form in the present collection have the triangular orifice, the
raised margins and punctured surface of the zocecium, with small avicularium pointing
downwards from just below the orifice, although this last is smaller and never raised
on an elevation. There is a second avicularium lying transversely on a separate area
above the orifice. Busk (1) mentions, but does not figure, a second avicularium with
a slender, spear-shaped mandible which corresponds with these, with the exception
that his are described as lying vertically and these are transversely placed. Where
ocecia are present, they take the place of these. They are large, covering nearly the
whole of the zocecium above the one to which they belong, as described by Busk, but
are sub-immersed and have no nodules, being punctured like the zocecium.

Localities : — Gulf of Manaar ; and East Cheval Paar.

Schizoporella subsinuata (Hincks, 9).

There are a few colonies of a grey-white colour, incrusting pieces of broken shell,
having the zooecial characters of this form with the addition of numerous avicularia,
which in Hincks's specimens did not exist and which were rare in MacGillivray's (23).
They have long, pointed mandibles, and are usually pointing downwards, like those in

Q 2

116 CEYLON PEAEL OYSTER REPORT.

MacGillivray's figure (plate cxxxviii., fig. 5), from an upper angle of the zooecium ;
but are sometimes below the orifice, transversely placed, or they may be on separate,
raised areas.

Localities : — Gulf of Manaar ; and Navakaddu Paar.

Schizoporella avicularis, n. sp. — Plate, fig. 2.

Zoarium adnate, of a pink colour ; zocecium punctured, with slightly raised margins.
Orifice with a broad sinus and umbo below it. Avicularia here and there large and
spatulate, starting from above the umbo, covering over the oral aperture and resting
on an extended portion of the zooecium above. Ocecia large, more finely punctured
than the zooecium and with a calcareous arch over its summit.

Locality : — Gulf of Manaar.

I do not know of any species having an avicularium in the position described above.
When the mandible is removed, the zocecial orifice, with its operculum, appears beneath,
in its usual form. There is a specimen, in this collection too, of S. triangula, which
has apparently a similar arrangement of the avicularium and zocecial orifice, but it will
be necessary to examine more material before coming to a definite conclusion as to the
exact relations of these parts.

Schizoporella collaris, n. sp. — -Plate, fig. 4.

Zoarium incrusting, of a dull white colour. Zocecia rhomboidal, punctured, the
upper portion slightly narrowed to a neck and bent forward. Orifice with a broad
shallow sinus, peristome thickened and forming a triangular bracket-like process in
front. Ocecia punctured like the zocecia, the sides extending to form with the
peristome a collar round the orifice.

Locality :— Station XL VI., oft* Mount Lavinia.

Schizoporella pulcherrima (MacGillivray, 23).

There is a strong resemblance between the " small colony," described by
MacGillivray, of this species and the large specimens from Ceylon, although the
former has no punctures on the zooecium, nor is it coloured, while the latter have
punctures as well as radiating lines and are coloured a deep red, excepting for one
colony, probably old, which is white. The raised margins of the broad zocecia, the
shallow sinus and avicularia on either side of the orifice are the same ; these are
sometimes quite up to the edge and in an angle of the zocecium. Ocecia, not described
before, are large, covering almost entirely the zooecium above and embedded in it,
punctured like the zocecium. The orifices of the zocecia which bear ocecia are larger
than those of the others. It is a very striking species.

Localities : — East Cheval Paar ; off Galle ; off Mount Lavinia.

Schizoporella viridis, n. sp. — Plate, fig. 3.
Zoarium incrusting and forming very large colonies, extending to nearly two feet
(50 centims. or upwards) across.

TOLYZOA. 117

It has a coarse looking, roughened surface and is of a greenish colour, somewhat
obscured by a brownish surface layer or membrane ; zooecium prostrate, irregularly
oval, ventricose and punctured closely all over the front wall ; often with an umbo in
its centre, or to one side, which may become tall and massive. Orifice with a deeply
rounded sinus, peristome raised above.

Avicularia small and pointed, one or two transversely placed below the orifice, and
a few long and sword-shaped, on separate areas, scattered over the zoarium.

Locality : — Coral banks, Gulf of Manaar.

This fine species forms thick massive colonies of many superimposed layers, and
spreads over other objects to form very large masses ; one colony measures upwards
of 53 centime, in length by 18 centims. in breadth and 23 millims. in thickness. The
verdigris-green colour is a striking feature, best seen in the thickness of the super-
imposed layers at the edges of broken pieces ; a brown membrane which envelops
the zoarium conceals it, somewhat, on the surface. The large ventricose zooecia and
ocecia can be seen by the naked eye as small pimples covering the irregularly undu-
lating surface of the zoarium. The zocecia are always prostrate, but are heaped and
turned in various ways and are seen at various levels.

Schizoporella incrassata, Hincks (6).

The great variety in appearance of the zooecia in different parts of one colony is
quite as marked in the present specimens as it was in Htncks's, and, although
the variations do not always agree with his, the differences seem to belong to
unimportant characters. The primary orifice has, here, sometimes from 2 to 4 spines
on its upper margin. A large process, bearing an avicularium, is sometimes to be
seen on both sides of the orifice instead of on one side only. The surface of the
zooecium is usually grooved, the grooves radiating towards the centre. Avicularia on
raised processes are scattered irregularly over the zooecia. The peristome forms
sometimes a spinous collar, open in front, or the points of the edges of the opening
join and leave the opening below like an oval pore. The ocecium agrees with that of
Hincks's species, but that the flat plate covering its aperture is of a dead white,
contrasting with the glassy appearance of the rest of the zoarium.

Localities :— Gulf of Manaar ; Station XL., 10 miles off Galle, 34 fathoms.

Mastigophora dutertrei, var. pes-anseris (Smitt, 26).
Locality : — Station XL., 10 miles off Galle, 34 fathoms.

Rhyncopora bispinosa, Johnston.

The primary orifice of this species varies in one specimen from the usual shallow
sinus form to one with a deeper, narrower sinus, which variety has also a crenulated
margin as in Rhyncopora crenulata, Waters (27), a species which has, however, no
sinus.

Localities : — Gulf of Manaar ; and Station XL VI., off Mount Lavinia.

118 CEYLON PEARL OYSTER REPORT.

Rhyncopora corrugata, n. sp. — Plate, fig. 5.

Zoarium incrusting, yellow-white in colour. Zooecia large, distinct throughout, with
the secondary orifice prolonged into a tube, widening from the base up and with an
uneven margin. An uncinate process at the base of the tube a little to one side of
the centre, and an avicularium of small size on the corresponding other side, its beak
pointed and forming with the uncinate process a loop-shaped sinus.

Locality : — Gulf of Manaar.

The raised secondary orifice of this species gives it the appearance of a Lagenipora,
but that the tube is irregularly fluted in outline. It is a good deal smaller than
Rliyncop>ora incisor, n. sp., although at first taken to be the same when the two were
growing together.

Rhyncopora incisor, n. sp. — Plate, fig. 6.

Zoarium incrusting, of a white colour. Zooecia crowded, hexagonal, with a deeply
areolated margin, a long tubular peristome and even rim.

Primary orifice orbicular with a transversely placed avicularium a little to one side
of the centre, an uncinate process on the opposite side becoming very long and
pointed. Ocecia behind the tubular orifice, smooth with a circular area on either side.

Localities : — Gulf of Manaar ; and off Galle, 34 fathoms.

The chai'acters of this species are very simple and constant ; the most striking
feature is the elongated uncinate process which sometimes projects almost across the
orifice, so as to bar the entrance to the tube, and has a curved needle-like point.
There is no avicularium to be seen when this stage is reached.

One perfectly preserved colony has the appearance to the naked eye of the pile of
white velvet ; the zooecia almost approach those of Lagenipora tuberculata in size and
appearance.

Hippothoa flagellum, Manzoni.
Locality : — Chilaw Paar.

Gemellipora glabra, form striatula, Smitt (26).
Localities : — Gulf of Manaar ; Trincomalee ; and off Mount Lavinia.

Gemellipora lata (Smitt, 26).

The Ceylon specimens of this form have the yellow colour of the colony, the dark
colour of the opercula, more brown than green in these specimens, the conspicuous
pores of the zooecia, the form of orifice and peculiar fold of the peristome below this,
as figured by Smitt (26, plate vii., fig. 157).

Avicularia are distributed among the zooecia, on separate areas, but are much larger
than he describes and spatulate, not oval.

Ocecia, not described before, are more broad than high, perforated like the zooecia,

POLYZOA. 119

with, sometimes, an umbo. The orifices of fertile zooecia are much larger than those
of the others.

Locality : — Gulf of Manaar.

J

Gemellipora protrusa, n. sp. — Plate, fig. 7.

Zoarium incrusting, of a pale brown colour. Zocecia rhomboidal, smooth or slightly
roughened, front wall much raised, punctured round the margin. Orifice of the usual
Gemellipora form, sometimes much elongated ; peristome often raised above and
irregularly lobed. A large avicularium on the front of the zocecium, directed across it,
supported on a raised process ; mandible wide at the base, becoming long and pointed ;
a small avicularium, with rounded mandible, on a raised jjrocess on one or both sides
of the orifice. Ocecia granular and minutely punctured, open in front, and leaning
forward over the orifice.

Locality : — Gulf of Manaar (numerous colonies on Nullipore balls).

There are sometimes two avicularia, in place of the one large transversely placed
one ; they point outwards to either side of the zocecium. There is some resemblance
in the appearance of this species to Schizoporella ampla, Kirkpatrick (20).

Family: ESCHARHLE.

Lepralia robusta, Hincks (8).

The central circular pore, alluded to by Hincks, is quite evident on the present
specimens, and is sometimes multiplied to three or more of irregular shapes, as seen
in an old, worn specimen.

Locality : — Gulf of Manaar (in large quantity on broken shells).

Lepralia poissonii, Audouin.
Localities: — Navakaddu Paar ; off Galle and onwards up West Coast, deep water.

Lepralia mortoni, Haswell (19).
Localities : — Gulf of Manaar ; off Trincomalee ; north end of East Cheval Paar.

Lepralia triangula, n. sp. — Plate, fig. 8.

Zoarium incrusting. Zocecia punctured all over, covered by a yellow membrane.
There are usually scattered spinous processes on the front wall of the zocecium and also
two curved horn-like processes, one on either side below the orifice, which may be
avicularia. Orifice longer than broad, with a much raised, thin, often irregularly
pointed peristome, forming a collar round it. Operculum with a triangular
excrescence having its base attached to the base of the operculum. No ocecia.

Locality : — Gulf of Manaar.

It is possible this species may be a variety of L. pallasiana, Moll., as it resembles
L. canthariformis, Busk, which is probably a variety of L. pallasiana, but the

120 CEYLON PEARL OYSTER REPORT.

peculiar processes on the cell wall and that on the operculum give it a distinctive
character.

Lepralia turrita, Smitt (26).

The fresher looking colonies of this species are pink in colour and have pointed
tubercles round the margin of the orifice ; with age they become white and the
tubercles worn down into blunt knobs. There are large spatulate avicularia on the
sides of large massive tubercles and small oval ones on more slender tubercles, and
also scattered over the zocecia.

There are very few ocecia, and I have not been able to see perforations on them.
They have a semicircular, marked area in front.

Localities: — Gulf of Manaar ; Station XL., 10 miles off Galle, 34 fathoms; Nava-
kaddu Paar.

Lepralia multidentata, n. sp. — Plate, fig. 9.

Zoarium incrusting, white. Zooecia very small, rotund, irregularly placed, usually
lying flat, but occasionally standing upright, surface roughened. Orifice arched
above, sides widening downwards, base slightly convex, six long slender spines on the
upper margin and a long, pointed rostrum below. Sometimes avicularia with rounded
mandibles are present on either side of the orifice. Ooecia granular above, smooth in
front, with au arched rib between the rough and the smooth portions.

Localities : — Gulf of Manaar ; and off Trincomalee.

Lepralia cucullata, Busk (2).

Localities : — Cheval Paar, Navakaddu Paar and elsewhere in Gulf of Manaar
(common on Ascidians and pearl oyster shells).

Lepralia depressa, Busk (2).

Avicularia in various forms, long and seta-like, or thick and spear-like, or the
mandible branched, looking like the leg and foot of a bird.
Localities : — Gulf of Manaar and off Galle.

Lepralia gigas, Hincks (10).

Localities: — Station LILT., north of Cheval Paar, 7 to 10 fathoms; Palk Bay;
Trincomalee (many large colonies, up to 3 inches across) ; Welligam Bay ; and
various parts of Gulf of Manaar (growing on pearl oyster shells).

Lepralia purpurea, n. sp. — Plate, fig. 13.

Zoarium forming purplish grey patches, incrusting. Zocecia small with thick walls,
sub-immersed, diamond-shaped, occasionally heaped and upright, smooth and shining
or slightly roughened. Orifice arched above, widening downwards, with an almost
straight lower lip, peristome slightly thickened with five marginal spines and a sub-

POLYZOA. 121

oral mucro below the orifice. The tip of this, the bases of the spines and the
peristome are of a purple tint. Sometimes an avicularium on a raised process at
one side of the zocecium, varying in size, spatulate and large or small and pointed.
Ocecia small, narrow, open in front, two of the spines showing in front.
Locality : — Gulf of Manaar.

Lepralia nitida, n. sp. — Plate, fig. 10.

Zoarium closely adhering, with a shining surface, pale yellow in colour. Zooecia flat
but rising to a prominent umbo situated below the orifice, granular, the granules
radiating from the umbo to the margin of the zocecium which is punctured there.
Orifice arched above, narrowing, then widening, and with a convex lower lip. A
small round avicularium on either side on a level with the lower margin of the orifice.
Ocecia small, rising to a prominent umbo, orifice of the fertile zocecium much larger
than the usual orifice.

Locality : — Gulf of Manaar.

This is a small neat species with little variation of form. The ocecia are merely
like a triangular extension of the zooeciurn above the orifice.

"to'

Lepralia adpressa, Busk (2).

The surface of the zooeciurn in the present specimen is granular, punctured round
the margins only ; the upper portion of the zocecium is considerably raised ; the lateral
lumps carry small round avicularia ; when in old specimens these are absent the
lateral processes are seen to be hollow tubes. Sometimes two or three of these
processes are present above the orifice and one below in addition.

Locality : — Off Galle, deep water.

Lepralia feegeensis, Busk (2).

The ocecium of this species has not been described before. There is one well
preserved ocecium on the Ceylon specimen. It is large and irregular in shape, the
front wall is marked with large circular pitted areas, punctured in their centres,
giving the ocecia a much coarser appearance than the zooecia, the reverse of what
is seen in Lepralia gigas, Hincks (10). The orifices of fertile zooecia are broader at
their bases than those of others. Avicularia of the usual type, and in the usual
position for L. feegeensis, are often present in pairs, but usually singly, or there may
be none.

Locality : — Gulf of Manaar.

Lepralia cleidostoma, Smitt (26).

There is one small colony of this species. It has no ooecia to show the characteristic
striae as a help to identification, and the avicularia are always directed upwards, not
outwards. On one of the zooecia, on the outside of the colony, jointed spines are to be
seen, as in Smitt's (26), plate xi., fig. 217. A larger, loosely incrusting colony from

R

122 CEYLON PEARL OYSTER REPORT.

another locality resembles this so nearly as to make it probable it is one of the same
species. The zocecia are larger, avicularia are smaller and turned sideways, as in
Smut's figures. Ocecia are present here, but are punctured, not striated, and they are
often half buried in the calcification of the zoarium. There is still a third variety,
smaller than either of the above, and with a smooth and glistening surface, with
avicularia turned sideways and ocecia faintly striated, but with a small arched area in
front. As all these three specimens have the same key-shaped orifice and pointed
avicularia I am inclined to think them varieties of the one species.
Localities :— Gulf of Manaar ; and off Galle.

Lepralia ceylonica, n. sp. — -Plate, fig. 11.

Zoarium adhering, of a yellowish-white colour. Zocecia in linear series closely set,
becoming upright in places, with glistening front walls, pitted and punctured and
lumpy. Orifice arched above, with straight sides and slightly concave lower lip, six
long sharp spines above and several bosses below the orifice, carrying small avicularia
with pointed mandibles on their summits. Ocecia roughened like the zocecia and
having scattered raised avicularia upon them.

Locality : — Gulf of Manaar.

Lepralia fissa, n. sp. — Plate, fig. 12.

Zoarium incrusting, forming brown patches. Zocecia finely punctured at the outside
of the colony, becoming coarsely ridged further in and having a large limbo occupying
most of the area, and a large avicularium on its outer side, with the mandible directed
upwards. Orifice broader than long, with a pouting lower lip. Ocecia rounded,
smooth, with a large cone-shaped fissure in front.

Locality : — Oft' Galle, deep water.

Lepralia subimmersa, MacGillivray (23).

There is one colony, greyish-white in colour, and covering the shell of a univalve
mollusc inhabited by a hermit crab. Another colony is of a deep red colour, from a
membranous covering still adhering to the zoarium.

Locality : — Gulf of Manaar.

Escharoides verruculata (Smitt, 26).
Localities : — Gulf of Manaar ; and off Mount Lavinia.

Porella malleolus, Hincks (8).
Localities : — Gulf of Manaar ; oft' Galle ; Navakaddu Paar.

Smittia trispinosa (Johnston).
There is great variety in the form of zocecia and avicularia among the Ceylon

POLYZOA. 123

specimens of this species. It is widely distributed around the coast, and apparently
very abundant.

Localities : — Gulf of Manaar ; off Galle ; off Mount Lavinia ; and Navakaddu Paar.

Smittia trispinosa, var. spatulata (Smitt, 26).
Locality: — Station XLVL, off Mount Lavinia.

Smittia trispinosa, var. protecta, nov.

This large variety has the peristome raised and in front, below the orifice, produced
into a long spout-like extension. Avicularia, not on every zocecium, are of enormous
size, reaching from above the orifice and bending and spreading so as to cover the
whole front wall of the zocecium, and having a blunt extremity.

There are sometimes two small raised avicularia, one on either side of the orifice.
The finely punctured ocecium has a narrow, prominent arched rib across the front.

Locality : — Gulf of Manaar.

Smittia tubula, Kirkpatrick (20).

The present specimens of this form are of a pink colour. There are usually two
avicularia, one on either side of the tubular secondary orifice, pointing upwards. Two
of the oral spines, of which there are six, remain in front of the ocecium when present.

Locality : — Gulf of Manaar.

Smittia rostriformis, Kirkpatrick (20).

Avicularia on these specimens are sometimes pointing downwards, as figured by
Kirkpatrick, but sometimes upwards with the appearance of a spine on either side
of the orifice. In both cases the peculiar serrated edge of the beak is plainly visible.
Oral spines vary in number from two to six. In other respects the specimens agree
with the original description of the species.

Locality : — Station XLVL, off Mount Lavinia.

Phylactella spiralis, n. sp. — Plate, fig. 14.

Zoarium forming small pink patches on shells. Zooecia arranged in radiating lines,
smooth or slightly roughened, areolated round the margin, with a much raised
tubular peristome and an avicularium raised to the margin of this on a semi-spiral
tube. Primary orifice, having a very wide denticle with sharp lateral points.
Secondary orifice, with from two to four spines above. Ocecium behind the tubular
peristome, smooth or slightly roughened like the zocecium.

Localities : — Gulf of Manaar ; off Mount Lavinia ; and off Galle.

This species approaches most nearly to Pliylactella gcometrica, Kirkpatrick (21),
in form, but the one broad denticle instead of three and the difference in the position
of the avicularium, which resembles that of Lagenipora nitens (MacGillivray, 23),
makes it impossible to mistake the two species.

R 2

124 CEYLON PEARL OYSTER REPORT.

Mucronella coccinea, Abild.
Localities : — Gulf of Manaar ; and off Galle.

Mucronella tubulosa, Hincks (3).

In the Ceylon specimens, which I believe to belong to this species, there is an
enormous development of the central mucro, which is here long and spinous ; there is
also an avicularium on the inner side of this, either at its base, lying in front of the
primary orifice, or at varying heights up this process, always transversely placed
and with a sharp curved beak, but varying in size.

These points add to the resemblance between M. tubulosa and Rhyncopora longi-
spinosa, which Miss Jelly regards as synonymous, but there is missing still, in these
specimens, the uncinate process of Rhyncopora, unless the curved beak of the
avicularium takes the place of this. In the faint indication of a sinus of the primary
orifice, the spinous mucro and markings of the ooecia, there is a likeness to a species
of quite another genus, i.e., Cellepora longirostris, MacGillivray, as described by
Miss Philipps (25).

Localities : — Off Trincomalee ; off Mount Lavinia ; Navakaddu Paar and elsewhere
in the Gulf of Manaar.

Mucronella thenardii, Aud. — Plate, fig. 15.

The cross-shaped process, situated below the orifice of the zocecium in this species,
is greatly developed (see fig. 15). Its upright portion is often occupied by a large
spatulate avicularium, and, where this is so, one of the arms of the cross is missing,
giving a one-sided appearance to the process. Sometimes the arms are duplicated,
one pair below the other, and they are always much branched, each branch bearing a
small, rounded aviculariurn on its summit. Slender, spinous processes, resembling the
branches in size, are scattered over the front wall and round the margin of the orifice
of the zooecium, and there are sometimes ordinary spines, from two to five in number.

There is a strong resemblance between the characters of this species and those of
M. aviculifera, Hincks (14). The slender, spinous, aviculiferous processes are present
there, but the cross-shaped mucro is only represented by a small simple mucro, and
there are large, lateral avicularia. There is, however, much variation in form and
position of the avicularia and of the processes which carry them, even in one small
specimen of the present collection, so that it seems possible that the differences
represent various stages in the development of one and the same species.

Localities : — Gulf of Manaar ; and off Galle.

Mucronella vultur, Hincks (7).

There are in the present collection specimens having all the important character-
istics of M. vultur, as described by Hincks (7) and MacGillivray (23) with this
exception, that the avicularium on the central mucro has a rounded, instead of a

POLYZOA. 125

pointed, mandible. There are sometimes small, raised, rounded avicularia scattered
over the front wall of the zooecium in great profusion. Some of the specimens have
spinous knobs in front of the ooecia and also one on the summit of this.

Altogether, with the addition of the usual six long marginal spines of the orifice,
the colony has a formidable appearance, especially in specimens where the zooecia are
crowded together. The largest colony is partly incrusting and partly sends off free
expansions. It measures 8 centims. by 4 centims. and is of a dull pale brown colour
and of a very brittle and light substance.

Locality : — Gulf of Manaar (growing over a colony of Lepralia gigas, Hincks).

Retepora tubulata, Busk (1).
Locality : — Gulf of Manaar.

Retepora simplex, Busk (1).
Locality : — Off Galle and onwards up west coast of Ceylon.

Retepora apiculata, Busk (1).
Locality : — Ceylon seas.

Retepora pocillum, n. sp. — Plate, fig. 16.

These colonies correspond with Retepora avicularis, MacGillivray (23), in size, and
the zooecia in having triangular teeth within the orifice, and below this a loop-shaped
fissure, but there are some other additional marked characteristics in these specimens.
The zooecia are rhomboidal in form and have slightly tubular necks crowned by about
six spines, which are beautifully jointed in the Equisetum-like form described for
B. monilifera, MacGillivray (23). These are often broken, but their bases, which,
united, form the tubular neck, are visible, and when ooecia are present two are to be
seen in front of it, not always of the jointed form. The ooecia have a fissure faintly
visible. There is generally a small avicularium with a rounded mandible below the
orifice, more or less in the middle of the zooecium ; sometimes there are two of these,
one below the other, or rarely one large spear-shaped one, lying across the zooecium
and pointing upwards.

Locality : — Gulf of Manaar.

Family : ADEONIDJi.
Adeonella subsulcata (Smitt, 26).

The largest colonies of this species in the present collection are about 8 centims. in
height. They correspond entirely with Smitt's description, but that there is also a
serrated denticle within the orifice, corresponding to that of A. pectinata, Busk (1).

Localities: — Station I., off Negombo, Gulf of Manaar; Station XXIX., Trin-
comalee ; Station XLVL, off Mount Lavinia, 25 to 30 fathoms ; Station XL., 10 miles
off Watering Point, 34 fathoms ; and off Galle, deep water.

12G CEYLON PEARL OYSTER REPORT. -

Family : CELLEPORID.F.

Cellepora albirostris (Smitt, 26).

Localities : — Gulf of Manaar ; off Mount Lavinia ; off Kaltura ; and off Galle, deep
water. There are large quantities of this species.

Cellepora megasoma, MacGillivray (23).

There is usually a raised avicularium to one or both sides of the orifice on the
present specimens and large scattered spatulate ones here and there. The zocecia are
smooth or slightly ridged. Ocecia thickly punctured all over, with no marked area.

Localities : — Off Galle and onwards up west coast of Ceylon, deep water ; Gulf of
Manaar (several colonies on worm tubes and stems of Zoophytes) ; Station XXXTL,
off south-east coast of Ceylon (on Sponge).

Cellepora rota, MacGillivray (23).

Localities: — Station LIIL, north of Cheval Paar, 7 to 10 fathoms; and elsewhere
in Gulf of Manaar.

Cellepora cidaris, MacGillivray (23).

There is a large quantity of material resembling the description of this form
excepting that the columnar processes in between the zocecia are solid instead of
hollow. C. albirostris, Smitt, has, in the " Challenger " collection, occasional, solid
columns of this sort on the older parts of a colony, but, although the two species
C. albirostris and C. cidaris of the present collection resemble each other pretty
closely in some points, neither the long pointed rostrum, nor the broad one with
serrated beak, nor the dark operculum, characteristic of C. albirostris, are present on
the specimens I have considered to be C cidaris. Ocecia, said in C. cidaris to be
globular and immersed, have an arched area in front, which, being often absent, leaves
a cave-like space. There are occasional large sjmtulate avicularia to be found on
raised areas in between the zocecia, but these are the only points of difference to be
seen betAveen the present specimens and MacGillivray's description of C. cidaris.

Localities: — Off Mount Lavinia; Station XL., off Galle, 34 fathoms; and in Gulf
of Manaar.

Cellepora compacta, n. sp. — Plate, fig. 17.

Zoarium incrusting, white or purplish in colour. Zocecia small, upright, rounded,
smooth, with a few marginal punctures.

Orifice rounded, with a loop-shaped sinus, and below, rather to one side, a large
thick rostrum, sometimes pointed above, having a long pointed avicularium on its
side ; often other slenderer jjrocesses round the orifice and long hollow columns in
between the zcecia. Ocecia standing upright, smooth and shining, with a narrow
arched ridge in front.

Locality : — Gulf of Manaar.

POLYZOA. 127

Sub-order: CYCLOSTOMATA.

Family: CRISIIDiE.

Crisia holdsworthii, Busk (2).

Localities: — Station LIIL, north of Cheval Paar, 7 to 10 fathoms; Station I., and
elsewhere in Gulf of Manaar ; and Palk Bay.

Family .- TUBULIPORID^E.

Idmonea milneana, D'Orb.

Localities:— Navakaddu Paar; and Station XL, 10 miles off Watering Point,
34 fathoms.

Family : LICHENOPORID^E.

Lichenopora liispida, Fleming.

Localities : — Navakaddu Paar, and elsewhere in Gulf of Manaar.

Lichenopora novae-zelandiae, Busk (2).
Locality : — Navakaddu Paar.

Sub-order III. : CTENOSTOMATA.

Family : ALCYONIDIID^E.
Alcyonidium mytili, Dalyell.
Locality : — Gulf of Manaar.

Pherusa tubulosa, Lx.

The largest colony among the few present in the present collection is about
3 centims. in height, and its branches spread to about the same in width. Zooecia are
on both sides of the branch, not as in the original description only on one surface.

Locality : — Station LIIL, north of Cheval Paar, 7 to 10 fathoms.

Family : AKACHNIDIIDiE.
Arachnidium fibrosum, Hincks (15).
Locality : — Gulf of Manaar.

128 CEYLON PEARL OYSTER REPORT.

Family : VESICULARIID^E.
Amathia distans, Busk (1).
Locality : — Gulf of Manaar.

Farrella atlantica, Busk (1).
Locality : — North of Cheval Paar.

Family: BUSKIIDJE.
Buskia setigera, Hincks (16).
Localities : — Station LIIL, north of Cheval Paar, 7 to 10 fathoms; and Palk Bay.

Family : CYLINDRCECIID^E.

Cylindrcechun dilatatum, Hincks (15).

Localities : — Station LIIL , north of Cheval Paar, 7 to 1 0 fathoms ; oft Galle ; and
Station I., off Negombo.

Family: VALKERIID.F.
Valkeria uva, Linn.

Locality : — Station LIIL, north of Cheval Paar, 7 to 10 fathoms.

Order: ENTOPROCTA.

Family : PEDICELLINID^E.

Ascopodaria discreta, Busk (1).

One small colony of a reddish colour, growing on a sponge.
Locality : — Navakaddu Paar.

POLYZOA.

129

LIST OF LITERATURE CITED.

(1.)

(2.)
(3.)
(4.)

(5.)
(6.)

(7.)
(8.)
(9.)
(10.)
(11.)
(12.)
(13.)
(14.)
(15.)
(16.)
(17.)
(18.)
(19)
(20.)
(21.)
(22.)
(23.)
(24.)
(25.)
(26.)
(27.)

Busk. — 'Voyage H.M.S. "Challenger" Reports,' xxx. and 1.

' British Museum Catalogue.' Polyzoa.

HlNCKS. — 'Annals and Mag. Nat. Hist.,' ser. 5, vol. vi.

5,

, vu.

5,

. viii.

5,

, ix.

5,

, X.

•5,

, xiii.

5,

. xiv.

5,

. XV.

5,

. xvii

•5,

, XX.

6,

, v-

6,

, vii.

' British Marine Polyzoa.' 1880.

'Journal Linnean Society,' Zoology, vol. xxi., 1889.

Harmer, S. F. — " Revision of the Genus Stegamoporella." ' Quart, Journ. Micros. Sci.,' vol. xliii.

. "Morphology of the Cheilostomata." 'Quart. Journ. Micros. Sci.,' vol. xlvi.

Haswell. — "The Queensland Polyzoa." 'Proc. Linn. Soc, New South Wales,' vol. v.
Kirkpatrick, R. — 'Annals and Mag. Nat. Hist.,' ser. 6, vol. i.

,, ,, ,, ,, t», ,, \ .

In Thurston's "Fauna of Gulf of Manaar." 'Madras Gov. Mus. Bull.,' No.

MacGillivray. — In McCoy's " Prodromus." ' Zoology Victoria.' Dec, i.-xx.

'Transactions Royal Society Victoria,' vol. iv., 1895.

Philipps, Miss.— Willey's ' Zoological Results,' part iv., p. 439, 1 900.

Smitt. — 'Floridan Bryozoa,' 1873.

WATERS. — 'Annals and Mag. Nat, Hist.,' ser. 5, vol. iii.

130

CEYLON PEARL OYSTER REPORT.

EXPLANATION OF PLATE.

Onychocella cucullata, n. sp.
Schizoporella aviculam, a. sp.

„ viridis, n. sp.

„ coUa/ris, n. sp.

Rhyncopora corrugata, n. sp.

„ incisor, n. sp.

th' nullipara protrusa, n. sp.

8. Lepralia triangula, n, sp.

9. ,, multidentata, n. sp.

10. Lepralia nitida, n. sp.

11. ,, ceylonica, a. sp.
12 „ ,/w.sfl, n. sp.

13. „ purpurea, ti. sp.

14. Phyhctella spiralis, n. sp.

15. Mucronella thenardii, Aud.

16. Ret&pora pocUlum, n. sp.

17. Cellt'pora compacta, ii. sp.

Sketch-map of the Ceylon coast, showing the principal localities from which specimens were collected.
C, Chilaw Paar; Ch., Cheval Paar; K., Karativo Paar; M., Modragam Paars; Mu., Muttuvaratu
Paar; P., Periya Paar; P.K., Periya Paar Kerrai.

CEYLON PEARL OYSTER REPORT

POLYZOA- PLATE

r~\

>

: .-...•

.• *

3.

Wm

5

I

(•- ~~ of.

V

:

3

\

s

a'

!

9 j

.

, v ^ <. - ' . '

8.

6.

11.

\

12.

13.

14.

17.

LR.T.del

in, Cambridge

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

REPORT

ON THE

MEDUSA

(HYDROMEDUS^E, SCYPHOMEDUS.E and CTENOPHORA)

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY
EDWARD T. BROWNE, B.A.,

ZOOLOGICAL RESEARCH LABORATORY, UNIVERSITY COLLEGE, LONDON.

[With FOUR PLATES.]

INTRODUCTION.

The collection of Medusae brought back by Professor Herdman from Ceylon and
another forwarded by Mr. James Hornell were kindly sent to me by Professor
Herdman for examination. So far as I know there are no previous records relating
to the medusas of Ceylon, except for the Siphonophora which have been specially
studied on the spot by Haeckel. Unfortunately none of his beautiful species occur
in the collection.

The specimens come chiefly from two places — Galle, at the south of Ceylon, and
the Pearl Banks in the Gulf of Manaar. It is distinctly a littoral collection from
shallow water, and the littoral character is shown by the number of Anthomedusse
and Leptornedusse. The collection has not brought to light any new genera, nor are
the species in any way very remarkable, considering that they live within a tropical
region, in a sea at about 80° F. From a geographical point of view the collection is a
valuable one, as it has increased our knowledge of the distribution of genera.

Some of the specimens had been splendidly preserved and were a pleasure to work
with, but many were in bad condition, more or less broken up, and these gave me
endless trouble. I have endeavoured, as far as possible, to give such details as I hope

s 2

132

CEYLON PEAEL OYSTER REPORT.

will lead to the recognition of the species again, but should any doubt arise when
comparing a description or figure with living specimens, I would suggest that it
would be better to amend my work rather than describe another species.

[After this report had been sent to the printer 1 received intimation from
Professor Otto Maas, of Munich, that he was preparing for publication his account of
the Hydromedusaj of the Dutch "Siboga" Expedition. As our respective reports
were likely to be published about the same time we agreed to exchange proofs. I
wish here to express my thanks to Professor Maas for his kindness in allowing me to
make use of his work, which must be regarded as a valuable contribution towards our
knowledge of the Hydromedusze.

A few of the species in the Ceylon collection were also taken by the "Siboga" in
the East Indies, namely : — Irenopsis hexanemalis, Octocanna polynema, Mesonema
'pensile, Liriope tetraphylla, and Solmundella bitentaculata.]

The following is a classified list of the species described in this report :—

Dipurena sp. t

Laodice indica, n. sp.
Mitrocomium assimile, n. sp.
Eutima curva, n. s]5.
Odorchis orientalis, n. sp.
Irene ceylonensis, n. sp.

HYDROMEDUS.E.

Anthomedus^.

Cytceis herdmani, n. sp.
Proboscidactyla minima, n. sp.

Leptomedus.e.

Irene palkensis, n. sp.
Irenopsis hexanemalis, Goette.
Octocanna polynema (Haeckel).
yEquorea conica, n. sp.
,, parva, n. sp.
Mesonema pensile (Modeer).

Gmionemus homelli, n. sp.

Trachomedus^e.

Olindias sp. 1
Liikipe tetraphylla (Cham, et Eys.).

Nakcumedus.e.
Sohmmdella bitcntacidata (Quoy et Gaim.).

SlPHUNOPHUKA.

Diphyes cliamissonis, Huxley.
( 'iijuilila sp. 1

Charybdea sp. 1
Nausithoe punctata, KdLL.

Agalmopsis sp. 1
Physalia utriculus, Esch.

Forpita sp. ?

SCYPHOMEDUS^E.

Pelagia sp. 1

Crambessa sp. 1

OTENOPHOKA.

l'l< urubcarhw, ijlobosa, M.0J3ER, var. ccylonensis, n. Beroe Jleminyi (E.SCIi.).

MEDUSA 133

A Comparison between the Medus.e of Ceylon and those of the

Maldives.

It is somewhat hazardous to draw a comparison between the medusoid fauna of
Ceylon and that of the Maldives. About that of Ceylon I feel that we at present
know but little, in fact, we have just made a beginning in our observations. We
know, perhaps, a little more about the medusae of the Maldives, which have recently
been visited by two expeditions — the hrst under the leadership of Mr. Stanley
Cardiner, and the second, which soon followed the first, under the guidance of
Professor Alexander Agassiz.

Leaving out the oceanic medusa1, which have usually a wide geographical range,
and limiting the comparison to the Anthomedusa? and Leptoniedusas, Mr. Gardiner's
collection contained 5 genera and 5 species (3 of which were new species). Professor
A'i.vssiz's collection contained 10 genera (1 new genus) and 11 species (6 of which
were new species and 3 were not named). With the possible exception of one of the
iiEquoriicUe, the genera and species were quite distinct in the two collections. Taking
the two collections, without any limitations, there is a well-marked difference between
them, which is quite as great as if they had come from localities a thousand miles
apart. It is difficult to assign a correct reason for such a difference, but probably a
different season of the year and the localities visited had much to do with it.

The two Maldive collections contain altogether 14 genera and 15 species of
Anthomedusse and Leptomedusa?. The Ceylon collection contains 12 genera and
14 species (10 of which are regarded as new species). A comparison between the
Maldive and Ceylon collections shows that 6 genera [Dipurena?, Proboscidactyla,
Irene, Trenopsis, Mesonema and yEquorea) are common to both, but only 2 species
(Irenopsis hexanemalis and Mesonema pensile ; the former also occurs at Zanzibar
and the latter in the Red Sea), I think that these figures show clearly that the
medusoid fauna of Ceylon is quite distinct from that of the Maldives.

HYDROMEDUSJL

Order : ANTHOMEDUSSE.

Family: SAKSIID^E, Forbes, 1848.

Dipurena, McCrady, 1858.

Dipurena, sp. ? — Plate II., figs. 1 and 2.

Description. — Umbrella about as high as broad. Manubrium extending beyond
the margin of the umbrella. Mouth circular. Gonads arranged in several large

134 CEYLON PEARL OYSTEE REPORT.

isolated clusters upon the manubrium. Four tentacles with globular basal bulbs, and
with numerous small clusters of nematocysts. Ocellus upon the outer side of each
basal bulb.

Size :— Umbrella about 3 millims. in width and height.

Locality : — Galle Bay, one specimen on June 12 and two on August 25.

Notes. — The umbrella of the three specimens is so badly out of shape that a figure
of it cannot be given. The manubrium (fig. 2) is a long thin tube, about two to
three times the length of the cavity of the umbrella, with an apical knob in the jelly
of the umbrella, and with a large terminal stomach which has a circular mouth. The
gonads are situated on the manubrium in large roundish clusters, and their size makes
them conspicuous. One specimen has four clusters of gonads, another has two
clusters. The third specimen has lost its manubrium. The stomach itself also
appears to be surrounded with generative cells. The tentacles (fig. 1) are long and
flexible, and are closely studded with clusters of nematocysts, except for a short
distance next to the basal bulb. The upper half of the basal bulb is globular and
embedded in the jelly of the umbrella, the lower part is like a circular band round
the tentacle, broader on the inner side than on the outer side. In this band, on the
outer side, the ocellus is situated. The ocellus is circular, of a yellowish colour in
formalin, and has a small lens.

I have placed this species in the genus Dipurena and follow at present Haeckel's
classification, though 1 think that this species and Dipurena ophiogaster will
ultimately have to be removed to another genus. All the other species have the
nematocysts on the tentacles confined to a large conspicuous terminal knob, and some
in addition have a few large swellings containing nematocysts just above the terminal
knob. Dipurena ophiogaster has the nematocysts on the tentacles arranged in quite
a different manner. They form numerous semi-circular or three-quarter spiral loops
round the tentacle when it is in a contracted or semi-contracted state. When the
tentacle is fully expanded the nematocysts form minute globular clusters which give
a moniliform appearance to the tentacle. The terminal cluster of nematocysts is very
small and inconspicuous. The arrangement of the nematocysts of this species from
Ceylon is similar to that of Dipurena ophiogaster and to that found in the genus
Sarsia.

The specimens from Ceylon are very much like Dipurena ophiogaster which belongs
to the British fauna, and it is not possible to point out a character in the structure of
the tentacles or the manubrium with its gonads by which they can be specifically
separated. As the specimens are in bad condition, 1 think it is best to leave the
specific name in abeyance. They may differ from the Dipurena ophiogaster in the
shape of the umbrella, in colour, and perhaps in other details.

Bigelow (1904) found at the Maldives a Dipurena which he was not able to
clearly distinguish from Dipurena frugilis, Mayer, belonging to the fauna of the
Tortugas in the West Indies.

MEDUSAE. 135

Family: MAKGELIIDjE, Haecket,, 1877.

Cytaeis, Eschscholtz, 1829.

Margeliida? with four single perradial marginal tentacles, and with unbranched oral
tentacles.

Cytaeis herdniani, n. sp. — Plate I., fig. 1 ; Plate IV., fig. 12.

Description. — Umbrella somewhat bell-shaped, about as high as broad, with fairly
thick walls. Velum narrow. Stomach large, about as long as wide, situated on a
short peduncle, and extending a little over half-way down the cavity of the
umbrella. Mouth circular (expanded) with about 50 to 60 oral tentacles, which are
unbranched and evenly distributed. Four broad radial canals. Gonads forming
four large perradial swellings, and extending the whole length of the stomach. Four
thick tentacles, perradial, with very large basal bulbs, which are somewhat triangular
in shape.

Size : — Umbrella about 3i millims. in width and height.

Locality : — One specimen from Chilaw Paar on March 20, and one from Cheval
Paar on November 1 1 .

Notes. — Of these two specimens, one is an adult and the other is an intermediate
stage. The stomach is cross-shaped in transverse section, and the gonads occupy the
sides of the cross. The oral tentacles have a small round terminal cluster of
nematocysts. The four marginal tentacles are thick and have a dark central band
of pigment (perhaps a brilliant colour in the sea). A transverse section (fig. 12)
shows that the pigment granules are confined to the endoderm cells, which form a
solid central band of cells along the tentacle. The pigment granules form a dense
layer round the periphery of the endoderm, and are also scattered along the walls of
the cells. There is a fairly thick layer of mesoglcea and an extra thick ectoderm
which contains an enormous number of nematocysts, closely packed together. The
basal bulbs are very large and extend some way up the umbrella. There is a layer
of dark pigment along the inner side of the bulb, and a thick whitish mass of cells
on the outer side. Sections were not cut of the basal bulb, but the dark pigment
granules would probably denote the endoderm and the external whitish cells the
ectoderm. The specimen at the intermediate stage in development is about 2 millims.
in width and height. It has four marginal tentacles, fewer oral tentacles, and smaller
gonads than the adults.

This species is nearly related t<> Gytoeis nigritina and Cytteis macrogaster of
Hakckel. It differs from them in having many more oral tentacles, in the shape
of the basal bulbs, and in the structure of the tentacles.

Family: WILLIIILE, Forbes, 184-8.
Proboscidactyla, Brandt, 1835, ex Browne, 1904.
Williidse with four radial canals leaving the stomach.

136 CEYLON PEARL OYSTER REPORT.

Proboscidactyla minima, n. sp. — Plate II., fig. 3.

Description. — Umbrella fairly thin and probably hemispherical in shape. Stomach
divided into four longitudinal lobes. Mouth with a sinuous margin. Four radial
canals with lateral branches. Gonads upon the lobes of the stomach. Tentacles
short, about 16 to 20, with globular basal bulbs.

Size : — Umbrella about 1 millim. in diameter.

Locality : — Marichchukaddi, on the Gulf of Manaar, in February.

Notes. — There are 12 specimens, but their condition does not permit the production
of a figure. I was unable to trace out the branching of the radial canal system,
owing to the umbrella of all the specimens being more or less contracted, and to
their fragile condition. The stomach is divided longitudinally into four lateral lobes
and appears cross-shaped in transverse section. The gonads form lateral swellings
upon the sides of the lobes, which do not extend along the top of the umbrella. The
tentacles (fig. 3) are very short, about 0*25 millim. in length, and have a globular
basal bulb situated in the jelly of the umbrella. On the ex-umbrella, not far from
the margin, there are circular clusters of nematocysts, one between every two
tentacles. The velum is very narrow. A few of the specimens are about
075 millim. in diameter and have 10 to 12 tentacles. Others, a little older, have
14 to 16 tentacles. One specimen has 18 tentacles. A full-grown adult has probably
20 tentacles. This species may be distinguished by the smallness of its size and the
position of the gonads on the stomach.

Order: LEPTOMEDUS^.

Family: THAUMANTIIDiE, Gegenbaur, 1856,

Laodice, Lesson, 1843.

Laodice indica, n. sp. — Plate I., fig. 5 ; Plate IV., figs. 7 to 11.

Description. — Umbrella slightly curved, about two to four times as broad as high,
with moderately thick walls. Stomach cross-shaped, fairly large. Mouth with four
short lips, having a slightly folded margin. Gonads extending from the stomach to
about half-way or close to the margin of the umbrella, forming a large hollow sac
upon each of the four radial canals. Tentacles about 60 to 80, with blackish basal
bulbs, and without spurs. A large black ocellus upon the inner side of nearly all the
basal bulbs. A single cordylus between every two tentacles. Cirri present.

Size : — Umbrella up to 6 millims. in diameter.

Locality : — Off Mutwal Island, West Coast of Ceylon, 12 specimens on March 19 ;
and Galle, 3 specimens, on July 15.

Notes. — The collection contains 1 5 specimens which are about 5 millims. to 6 millims.
in diameter and about the same age. The distance to which the gonads extend from
the stomach along the radial canals varies in different specimens. Some have the

MEDUS.E. 137

gonads on the proximal half of the canals, whilst in others the gonads reach nearly to
the margin of the umbrella. The gonads first arise in the proximal part of the canals
quite close to the stomach, and grow outwards towards the periphery of the umbrella.
Owing to the great extension of the walls of the radial canals in the region occupied
by the gonads, it is difficult in some specimens to mark the spot where the stomach
ends and the canals hegin. The gonads look as if the)' were situated upon lobes of
the stomach. The length of the gonads is independent of the development of the
generative cells, as a gonad extending over only half the radial canal has large ripe
ova. Sections show (fig. 10) that the ova at an early stage in their development are
among the endoderm cells, and that later on they move outwards to the ectoderm.
The section figured shows an ovum leaving the ovary and breaking through the
ectoderm.

The tentacles are closely packed together round the margin of the umbrella, and
apparently form two alternating series, one projecting outwards and the other hanging
down. Although similar in structure, only those belonging to the former have a
conspicuous blackish basal bulb with a conspicuous black ocellus. The latter series
have a smaller basal bulb, either colourless or slightly pigmented, and either without
an ocellus or with a very small one. The basal bulb is on the inner side of the
tentacle ; it is a semi-circular thickening containing nematocysts and granules of
pigment, which cover the exterior of the bulb and also extend in radiating lines into
the interior. There is no spur-like projection at the base of the tentacles. The cirri
(fig. 9) are capable of extending to a great length. There is probably one between
every two tentacles, but very few were seen on the specimens. The free end terminates
in an oval knob containing large nematocysts (fig. 11). The ocellus is situated on the
margin of the basal bulb just below the velum. It is of an intense black colour,
spherical in shape, with a circular pit penetrating nearly to the centre (fig. 8).

The cordylus is very small and club-shaped (fig. 9). The interior of club is
composed of endoderm cells which are connected with endoderm cells of the circular
canal (fig. 7). The structure of the cordylus resembles that of Laodice calcarata
(see Brooks, 1895).

The specimens from Ceylon come nearest to Laodice calcarata which inhabits the
North Atlantic. They differ from it in having no spur at the base of the tentacles,
in having larger ocelli, and perhaps in colour and in size.

Family : EUCOPID^E, Gegenbauk, 1856.

Mitrocomium, Haeckel, 1879.

Mitrocomium assimile, n. sp.- — Plate I., fig. 3.
Description. — Umbrella fairly thick, a little broader than high. Velum narrow.
Stomach short, with a quadrangular base. Four radial canals. Gonads upon the
outer half of the radial canals, forming large oval sacs. Four perradial tentacles.

T

138 CEYLON PEARL OYSTER REPORT.

About 5 to 7 marginal bulbs in eacb quadrant; the central, interradial, bulb being
much larger than the others. A cluster of cirri adjacent to and on each side of the
tentacles. About five marginal sensory vesicles, each with two (occasionally three)
otoliths, in each quadrant of the umbrella.

Size : — Umbrella 2 millims. in width and 1^ millims. in height.

Locality : — Cheval Paar, in February.

Notes. — There is only one specimen of this little medusa in the collection. It is
rather opaque with a yellowish stain, and its margin is badly curled inwards. The
umbrella is somewhat contracted, so that it may not be quite so highly arched as
figured. The stomach is contracted back. The mouth is wide open and quadrangular
in outline, but has probably four lips when closed. The gonads (male) are very large
for the size of the medusa. Each gonad is divided into tAvo by a median longitudinal
line. The tentacles have large basal bulbs and transverse bands of nematocysts. The
cirri are more or less contracted, and have a small terminal cluster of nematocysts.
The cirri are apparently confined to the proximity of the tentacles and none were
seen scattered along the margin of the umbrella.

As there is only one specimen, I place the species provisionally in the genus
Mil rocomium. It bears a resemblance to Mitroeomium cirratum in having cirri
clustered at the base of each tentacle. According to Haeckel's definition of the
genus it should have 8 tentacles and 16 sense-organs.

Eutima, McCrady, 1858.

Eutima curva, n. sp. — Plate III., figs. 1 to 3.

Description. — Umbrella probably hemispherical, nearly twice as broad as high,
moderately thick. Peduncle of the stomach long, quadrangular in transverse section,
and with a conical base. Stomach small, about twice as long as broad. Mouth with
four small lips, and sinuous margin. Four radial canals. Gonads along nearly the
whole length of the peduncle, one on each radial canal, beginning a little way below
the conical base of the peduncle and terminating not far from the stomach. Four
perradial tentacles, with long tapering cone-shaped basal bulbs, which are laterally
compressed and curve over the margin of the umbrella. About 30 to 35 marginal
bulbs in each quadrant of the umbrella, and alongside each bulb usually one,
occasionally two cirri. Eight adradial marginal sensory vesicles, each with about
8 to 10 otoliths, which are arranged in a semicircle.

Size : — Umbrella 10 millims. in width and 6 millims. in height. Peduncle about
10 millims. in length.

Locality : — Off Mutwal Island, West Coast of Ceylon, on March 19.

Notes. — The single specimen, although in a good state of preservation, has the
umbrella badly compressed and folded, so that it is impossible to figure the whole
medusa. It is an adult, as the gonads contain large ova. The basal bulbs of the

MEDUSAE. 139

tentacles are attached to a slight thickening of the umbrella and curl over the margin.
The nematocysts along the tentacle are arranged in transverse bands which do not
quite meet on the inner side of the tentacle, so that a shallow groove is formed along
the inner side, running the whole length of the tentacle. The marginal bulbs have
a patch of blackish pigment at their apex.

This species comes nearest to Euthna mira and Eutima insignis, but is distinguished
from them by the shape of the basal bulbs.

Octorchis, Haeckel, 1864.
Octorchis orientalis, n. sp. — Plate III., fig. 4.

Description.— -Umbrella probably hemispherical, a little broader than high, and
moderately thick. Peduncle of the stomach long, quadrangular in transverse
section, and with a broad roundish base. The length of the peduncle is about twice
the diameter of the umbrella. Stomach small, about as long as broad. Mouth with
four short lips and a deeply folded margin. Gonads upon the peduncle of the
stomach and also upon the sub-umbrella. The gonads occupy the greater length of
the peduncle, extending along the radial canals, beginning a little way below the
base of the peduncle and terminating close to the stomach. The gonads upon the
sub-umbrella usually occupy the central third of the radial canals, or the outer half,
but do not reach to the margin of the umbrella. Four long perradial tentacles, with
long tapering cylindrical basal bulbs. About 18 to 20 marginal bulbs in each
quadrant of the umbrella, each one with a lateral cirrus. Eight marginal sensory-
vesicles.

Size : — Umbrella about 5 milium to 6 millims. in diameter.

Locality : — Galle Bay, one specimen on June 5, seven on June 12, and two on
August 21.

Notes. — None of the specimens are in good condition, the umbrella being so
flattened out and crumpled that it is not possible to draw a figure of it. Some of
the largest specimens have the gonads upon the peduncle in a series of folds (fig. 4),
but it is possible that the folding may be due to the contraction of the peduncle.
The gonads upon the peduncle are much larger and longer than those ujdoii the
sub-umbrella, the latter forming merely a thin narrow band along the radial canals.
The marginal bulbs are very small and inconspicuous. The cirri are very slender and
have a small terminal cluster of nematocysts. The sense-organs are situated near
the tentacles. They are very small and globular in shape, their otoliths not visible.

Notes on Intermediate Stages.

(a.) Umbrella about 2 millims. in diameter. Peduncle about 5 millims. in length,
with gonads just appearing upon it. Four tentacles. Cirri present. About 9
marginal bulbs in each quadrant. Eight sense-organs.

T 2

I id ci;\ LON PEARL OYSTEB REPORT.

(I>.) [Jmbrella about .'! millims, in diameter. Peduncle about 6 millims, in length.
Gonads jusi appearing upon the pedunole and sub-umbrella. About L2 marginal
bulbs in eaoh quadrant .

'Tins speoies docs not agree with EIaeokel's definition of the genus Octorch%s, since
ii has only I instead of 8 tentacles, bu1 in other respeots it conforms to the generio
oharaoter. Octorchxs gegenonwi lias been frequently taken l>v me m British seas.
Early and intermediate stages (the latter often with gonads) have I tentacles,
whereas the fully developed adult has 8 tentacles. I think it would be better to
enlarge the generic oharaoter so as to include speoies with I and 8 tentaoles, than to
establish a Dew genus for speoies which have only I tentaoles.

Haeokel, in his monograph, mentions two speoies of Octoi'chis 0. gegenooywn and
0, ccvmpnwulcitus, both ooourring in the Mediterranean, but probably there is only one
species. The Bpeoimens from Ceylon arc distinguished from the Mediterranean
species by the greater length of the gonads on the pedunole.

Irene, Ksciisciioi.t/., I S-J!'.

Irene eeylonensis, n. sp, Plate III., figa 9 to I I.

Description. Umbi'ella probably watohglass-shaped, much broader than high,
with thin walls. Velum narrow. Stomaoh short, situated upon a long cylindrical
pedunole. Mouth with Pour lips, winch have a folded margin, Four radial canals.
Gonads linear, extending from the base of the peduncle to near the margin of the
umbrella. Tentaoles about too. Cirri absent. Sensory vesioles, one between every

two tentacles, each vesicle with a Single otolith.

Size; Umbrella up to aboul 25 millims, in diameter.

Locality: (Jalle Bay, one specimen on .1 ulv I .» ; I'heval l'aar, five ID November,
Notes, The collection contains six specimens differing in age and size, the smallest

being aboul .> millims. m diameter, All the specimens are more or less damaged.

They are in a fair state of preservation, but are stained dead black, probably Owing

to the use of osmio acid,

The umbrella is llat and thin, hut is no doubt slightly curved when the medusa is
alive. Only one specimen show s I he peduncle fairly well, the others have either lost
it 01 have it twisted up. The gonads form thin narrow hands, either Btraight or
sinuous, extending along the radial canals over the sub-umbrella. In the largest
specimens the ova are large and clearly visible. Some o( the specimens have a

marginal l>ulh between some <>t' the tentacles, and these bulbs I believe to be the

origin of new additional tentacles, and not warts or tubercles, which Ao not develop
tentacles. 1 am doubtful about the presence of excretory pores at the back of the
basal bulbs, as there were no indications >>f papilla', but they may be contracted.

MKPUS.K. 141

The sensory vesicles have one otolith, but occasionally a vesicle was seen with two
otoliths, which may have been caused by twinning.

(a) Umbrella about 5 millims. in diameter. About 28 tentacles.
('') » " 7 » » ,, 36 „

(<0 » „ 15 „ „ „ 72

Irene palkensis, n. sp. — Plate III., figs. 12 to 16.

Description. — Umbrella watchglass-shaped, about four times as broad as high.
Velum narrow. Stomach short, situated upon a long cylindrical peduncle. Mouth
with four short lips, which have a folded margin. Four radial canals. Gonads linear,
extending from the base of the peduncle to near the margin of the umbrella. Tentacles
about 50. Usually two or three marginal bulbs between every two tentacles.
Excretory pores opposite the basal bulbs of the tentacles and all the marginal bulbs.
Sensory vesicles about 2 to 4 between every two tentacles, each vesicle with two
otoliths (variation 1 to 4).

Size : — Umbrella up to 20 millims. in diameter.

Locality : — Palk Bay, north of Ceylon, five specimens on March 1G.
Notes. — The five specimens are all in a damaged condition, especially as to the
gonads and the margin of the umbrella. The smallest is about 1 5 millims. in diameter
and the largest about 20 millims. The gonads are upon the sub-umbrella along the
radial canals. One specimen has the gonads extending from near the margin of the
umbrella up to the peduncle, and for a short distance down the peduncle. The
number of tentacles is only given from an estimation, as not one specimen has even a
quadrant of the margin of the umbrella in a perfect condition. The basal bulb of the
tentacle is somewhat globular when the tentacle is contracted, and more cone-shaped
and tapering when the tentacle is expanded. On the inner side of the basal bulb
just above the velum there projects an excretory pore. These pores are conspicuous
and clearly visible when expanded, but almost invisible when contracted. Cirri were
specially searched for, but none were seen. The marginal bulbs are small, and to
judge from their appearance in one of the specimens, I think that some are capable of
developing tentacles. Their number between every two tentacles is variable, usually
two or three, sometimes only one. All these bulbs have excretory pores, similar to
the pores opposite the basal bulbs of the tentacles. The sense-organs are closed
vesicles with generally two otoliths (rig. 14), occasionally three to four otoliths, rarely
one. The otoliths possess well-marked eccentric zones, which are conspicuous in
specimens which have been apparently killed with a re-agent containing osmic acid.

At first sight Irene palkensis and Irene ceylonensis look very much alike, but after
an examination of the organs on the margin of the umbrella I came to the conclusion
that they were distinct species. Irene ceylonensis has about twice as many tentacles
without a series of marginal bulbs in between them, and there is a difference in the
shape of the basal bulbs of the tentacles, but I attach more importance to the sense-

142 CEYLON PEARL OYSTER REPORT.

organs as a better means of distinguishing between the two species. Irene ceylonensis
has only one sense-organ between every two tentacles, each sense-organ with a single
otolith. Irene palkensis has two to four sense-organs between every two tentacles,
and each sense-organ usually has two otoliths, occasionally three to four, and rarely
one. There is also a difference in the structure of the otoliths.

Ireuopsis, Goette, 1886.

Eucopidre with numerous sensory vesicles, and with numerous tentacles. Six gonads
in the course of six radial canals. Stomach upon a peduncle.

Irenopsis hexanemalis, Goette, 1886. — Plate I., fig. 4 ; Plate III., figs. 5 to 8.

Irenopsis hexanemalis, Goette, 1886, p. 832; Chun, 189G, p. 5.
Phialidium tenue, Browne, 1901, p. 730, plate liv., fig. 4; plate Ivii., fig. 16.

Description. — Umbrella like an inverted basin in shape, with a flattened top, about
twice as broad as high. Velum narrow. Stomach short, with six lateral lobes,
situated upon a short, broad, cone-shaped or semi-globular peduncle. Mouth with
six lips, having a deeply-folded margin. Six radial canals. Gonads linear, on the
distal part of the radial canals, close to the margin of the umbrella. Tentacles, about
30 to 40. Marginal bulbs about three or more between every two tentacles. Excretory
pores opposite the basal and marginal bulbs. Sensory vesicles, usually one, sometimes
two, between every two bulbs, each vesicle containing a single otolith (occasionally
about two to four). Cirri absent.

Size : — Umbrella up to about 1 8 millims. in diameter.

Locality : — Palk Bay, 18 specimens on March 1G ; Cheval Paar, 9 specimens.

Notes. — The collection contains about two dozen specimens and nearly all are in
bad condition. The smallest is an intermediate stage measuring 5 millims. in diameter.
The stomach is situated upon a short peduncle which is about 2 millims. to 4 millims.
in length. The peduncle is variable in shape. In some specimens it is conspicuous,
but in others hardly noticeable. When semi-globular, or like a broad inverted cone,
it is quite recognisable. In some of the specimens the peduncle is flattened out
(whether this is natural or due to preservation I am unable to say), and in this
condition the roof of the sub-umbrella appears convex, and the top of the umbrella is
very thick. The stomach (fig. 8) is divided into six lobes, and its base seen aborally
is like a six-rayed star. It is very short, about 1 millim. in length, and about twice
as broad as long. The mouth has six conspicuous lips, which are continuous with the
lobes of the stomach, and the margin of the lips is deeply and closely indented with a
series of folds. In some specimens the stomach and its peduncle are within the cavity
of the sub-umbrella, but those specimens which have an extra thick umbrella may
have the stomach projecting a little way outside the cavity.

MKDUSiE. 143

When a medusa lias normally six radial canals, a variation in number may be
expected. Medusae with six radial canals have been derived from a form with four
canals, and are much more liable to variation than those with fQiir canals. There are
altogether 27 specimens of Irenopsis, and six show a numerical variation in the radial
canals, their numbers being as follows ; — 4, 7, 8, 8, 9, 11. The number of gonads also
varies with the radial canals. The gonads vary very much in size, and are always
situated upon the distal or outer half of the radial canals. Most of the specimens
have very short linear or spindle-shaped gonads, about 1 millim. or little more in
length, and situated near the margin of the umbrella. Three specimens have the
gonads extending over nearly the whole of the distal half of the canals, but not quite
reaching to the margin of the umbrella.

The tentacles vary in number according to the size and age of the specimens. The
exact number in any one specimen could not be ascertained, as all the specimens have
the margin of the umbrella more or less damaged. As a rule, in the largest specimens,
there are about five or six tentacles (one specimen has six or seven) between every
two radial canals. I estimated the number of tentacles in several large specimens to
be about 36, and in one specimen at about 40. About the exact shape of the basal bulbs
of the tentacles I am uncertain. In a contracted state they look somewhat globular,
but are probably more conical when the tentacle is expanded. The marginal bulbs
between the tentacles are very minute and their number is variable. Usually about
three are present, but occasionally only one between every two tentacles. There are
excretory pores opening above the velum, opposite every basal and marginal bulb. In
nearly every specimen these pores are so contracted that their presence is not
noticeable. In a few specimens they are well expanded (fig. 5) and form long papillae.
The marginal sense-organs (fig. 7) are closed vesicles, usually with a single otolith,
but occasionally with two or three otoliths, rarely with four. There is generally only
one ' between every two marginal bulbs, or about two to four between every two
tentacles.

The genus Irenopsis was established by Goette for Irenopsis hexanemalis, found
at Zanzibar. The original description is rather brief and there is no figure. Chun,
however, has given a fuller account of some specimens taken at Tumbatu, off
Zanzibar. The genus clearly belongs to the sub-family Irenidae, and is readily
distinguished by the presence of six radial canals. As the specimens from Ceylon
agree with the descriptions given by Goette and by Chun, I have presumed that
they are Irenopsis hexanemalis, though I should have liked to see a figure for
comparison.

After seeing these specimens of Irenopsis I again examined Phialidium tenue,
which was described by me as a new species in the Report on the Hydromedusae of
the Maldive Islands. The description of this species, based upon a single specimen,
was given as follows : — " Umbrella watch glass- shaped and thin. Stomach small,
quadrangular in shape, and situated on a semi-globular thickening of the umbrella.

144 CEYLON PEARL OYSTER REPORT.

Mouth with four lips and a siuuous margin. Four gonads extending over the outer
half of each radial canal. Tentacles 25 in number. One or two minute marginal
bulbs between every tivo tentacles. Sense-organs numerous, one or two, rarely three,
between every two tentacles, with a single otolith. Umbrella 15 millims. in
diameter." I clearly pointed out that I did not regard the thickening of the
umbrella as a definite peduncle, and consequently placed the species in the genus
Phialidium instead of in Irene. The result of the second examination, with specimens
of Irenopsis for a comparison, leaves no doubt that the thickening of the umbrella
must be regarded as a peduncle, so that the species does not belong to the genus
Phialidium. It resembles Irenopsis in the shape of the peduncle, in the position of
the gonads, in the number of tentacles, marginal bulbs and sense-organs. The basal
bulbs of the tentacles are slightly larger. But it has only four radial canals, four
gonads, and a mouth with four lips. If the specimen had been in this collection I
should certainly have considered it to be an abnormal Irenopsis, having four instead
of six radial canals. With four radial canals one would expect to see a mouth with
four lips. I think that Phialidium tenue had better be regarded as an abnormal
Irenopsis.

Octocanna, Haeckel, 1879.

Octocanna polynema (Haeckel) — Plate II., figs. 8, 9, 10.

Description. — Umbrella about twice to three times as broad as high, and thick.
Stomach flat, octagonal base with eight lateral lobes, about 2 millims. in diameter.
Mouth with eight small lips. Eight radial canals. Gonads linear, extending over
the outer half of the radial canals and nearly reaching to the margin of the umbrella.
Sixteen tentacles. About three to four marginal bulbs between every two tentacles,
each having an excretory pore. One marginal sensory vesicle (seldom two) between
every two bulbs, each vesicle with two otoliths (rarely with one or three).

Size : — Umbrella up to 12 millims. in diameter.

Locality: — Palk Bay, one on March 16; off Mutwal Island, one on March 19;
Galle, one on August 25.

Notes. — The umbrella of two specimens is plano-convex in shape, fairly thick, and
its margin is curled inwards. The third specimen has a very thick umbrella, which
is more highly curved than those of the other two specimens, and the cavity of the
umbrella is very shallow. The stomach has eight lobes, from which run the radial
canals. The mouth is expanded in all the specimens and has eight small lips,
corresponding in position to the radial canals. The gonads, in two of the specimens,
are on the outer half of the radial canals, but in the third specimen they are more
central, occupying the central third of the radial canals. They are linear in shape,
increasing in thickness towards the distal end, and show fairly large ova. There are
eight tentacles in the smallest specimen (8 millims. in diameter), one opposite each

MEDUSAE. 145

radial canal, aud eight large marginal bulbs, one midway between every two tentacles.
A few of these bulbs are just beginning to develop tentacles. The tentacles are long
and slender, and their basal bulbs are somewhat globular. The excretory papillae are
plainly visible, and project out just above the velum. All the basal bulbs and the
small marginal bulbs have excretory pores. The small marginal bulbs are more or
less conical in shape, and some look as if they were capable of developing tentacles.

I place this species provisionally in the genus Octocanna, as it does not possess
all the characters according to Haeckel's definition. There are two species of
Octocanna, both of which were described, without figures, by Haeckel, and have
not since been recorded.

Octocanna octonema has 8 tentacles. Gonads reaching along the whole length
of the radial canals. Sixteen sense-organs, each with a single otolith. Umbrella
10 millims. in diameter. Red Sea.

Octocanna polynema has 32 tentacles. Gonads not along the whole length of the
radial canals. 60 to 80 sense-organs, each with two otoliths. Umbrella 15 millims.
in diameter. Singapore.

Both the above species have four very long oral lips, which Haeckel includes in
the generic characters. The Ceylon specimens have eight small lips. They also
possess marginal bulbs and excretory pores which are not mentioned by Haeckel.

[In the report upon the Hydromedusse of the "Siboga" Expedition, Professor Maas
describes under the name of Octocanna polynema, Haeckel, some medusae which
appear to me to be identical with the specimens in the Ceylon collection. These
specimens I had described in manuscript as a new species of Octocanna. As Maas
has emended Haeckel's description and transfers the genus from the ^Equoriidse to the
Eucopidae, he has prevented me from introducing a superfluous new species. I quite
agree with him as to the desirability of the removal of the genus to the Eucopidae and
have adopted the classification here.]

Family : JEQUORIIDJE, Eschscholtz, 1829.

iEquorea, Peron et Lesueur, 1809 ; ex Browne, 1904.

/Equoriidae with numerous simple unbranched radial canals. Stomach circular, with
the lower wall fully developed. Mouth capable of closing up.

iEquorea conica, n. sp. — Plate I., fig. 2 ; Plate II., figs. 16, 17, 18.

Description. — Umbrella somewhat cone-shaped, with a rounded summit, a little
higher than broad, and very thick. Velum narrow. Stomach flat and circular,
about half the diameter of the umbrella. Oral lips about 16 in number, long and
slender. About 16 radial canals. Gonads upon the proximal half of the radial
canals, very much laterally compressed. Tentacles about 26 to 30, small and slender ;
their basal bulbs small and somewhat cone-shaped, Between every two tentacles a

U

146 CEYLON PEARL OYSTER REPORT.

very minute marginal bulb and two sensory vesicles (sometimes only one), eacb with
two small otoliths .

Size : — -Umbrella up to 7 millims. in width and 8 millims. in height.

Locality : — Pearl banks, Gulf of Manaar.

Notes. — The collection contains six specimens, which are mostly about the same
size (5 millims. in width and 6 millims. in height) and age. Some are males and
others are females having gonads with large ova. The oral lips have an external rib,
with an internal groove which is probably ciliated. In this species the gonads are
confined to the proximal half of the radial canals, and hang down as laterally
compressed sacs. It is upon the position and shape of the gonads that I base the
specific character. Excretory pores along the circular canal are not visible. Four
of the specimens have 16 radial canals and 16 oral lips, one specimen has 15 canals
and another 18 canals.

JEquorea parva, n. sp. — Plate II., figs. 5, 6, 7.

Description. — Umbrella plano-convex in shape, a little broader than high, very
thick. Velum of moderate width. Stomach flat and circular, about one-third the
diameter of the umbrella. Oral lips 13 to 16 in number, of moderate length and
width. Radial canals 13 to 16. Gonads sac-like, in the central third of the radial
canals. Four (perhaps eight) tentacles, with large basal bulbs. About 12 or more
marginal bulbs between every two tentacles. About 10 or more marginal sensory
vesicles between every two tentacles, or usually one between every two bulbs ; each
vesicle with two small otoliths.

Size : — Umbrella up to 6 millims. in width and 4 millims. in height.

Locality : — Galle Bay, one on June 5 and two on June 12.

Notes. — The three specimens are about the same size and age. One is a female
and the other two are males. The stomach is about 2 millims. in diameter, and its
lower wall about 1 millim. in width ; the oral lips do not exceed 1 millim. in length.
The gonads have lost their original shape, as they have been crushed down by the
folding in of the margin of the umbrella. They occupy the central part of the radial
canals, and are slightly nearer to the margin of the umbrella than to the stomach.
The gonads hang down as sacs, somewhat laterally compressed. The female has large
ova. One specimen has 13 radial canals, gonads and oral lips; the other two have
16 radial canals, gonads and lips. Two specimens have only four tentacles, but the
third specimen has one interradial bulb which is just developing a tentacle. The
interradial bulbs are much larger than the other bulbs and probably have tentacles in
a fully developed specimen. The marginal bulbs, which are very variable in size, are
somewhat cone-shaped and contain nematocysts. Some of the bulbs have an
excretory pore opening on the sub-umbrella just above the velum. One specimen is
badly infested with Cercaria.

MEDUSA. 147

This little JEquorea differs from the other species of the genus in the small numbers
of its tentacles, and in the shape and position of the gonads upon the radial canals.

Mesonema, Eschscholtz, 1829 ; ex Browne, 1904.

^Equoriidse with numerous simple, unbranched radial canals. Stomach circular, with
lower wall quite rudimentary. Mouth nearly as large as the diameter of the
stomach and cannot be closed.

Mesonema pensile (Modeer), 1791 — Plate II., figs. 11 to 15.

Medusa sp., Forskal (1776, p. 9, tab. xxviii. B.).
Mesonema ccelem pensile, Modeer (1791, p. 32).
Mesonema pensile, Haeckel (1879); Browne (1904, p. 733, pi. lv., fig. 4; pi. Ivii., figs. 2-9).

In my Report upon the Hydromedusae of the Maldive Islands I gave a description
of Mesonema pensile (Modeer). In this Ceylon collection there are fragments of a
specimen which I believe belongs to this species. The specimen is from the Cheval
Paar, Gulf of Manaar, and is broken up into about twenty-five pieces, which together
represent only a portion of the whole medusa. Fortunately some of the fragments
contain all the organs of the medusa, and it is possible, within certain limits, to give a
description and to identify the species.

This medusa is so peculiarly constructed that all the organs lie close to the margin
of the umbrella. The umbrella is rather like a plano-convex lens in shape and of
great thickness. Around its periphery lie the mouth, stomach, radial canals, marginal
tentacles, and sense-organs. These organs are all close together, the distance from
the oral lips to the margin of the umbrella is only about 20 millims. To judge from
the curvatures of the stomach and the margin of the umbrella on the three largest
fragments (the largest fragment which contained all the organs measures 35 millims.
in length), the diameter of the umbrella should be much larger than that of the
largest Maldive specimen, which measured about 60 millims. in diameter. I think
that this medusa when alive was probably about twice the size of the largest Maldive
specimen.

The stomach (fig. 14) is rudimentary, and its lower wall is about 4 millims. in
length. The margin of the mouth is furnished with a large number of long narrow
lips, which are strengthened by an external rib. The length of the longest lips is
about 4 millims. Among the lips there are many small ones in the course of
development. In structure and shape the oral lips are exactly like those of the
Maldive Mesonema, but they are a little longer. The lower wall of the stomach is
also longer, twice the length.

The radial canals are very numerous, and very short ; the distance from margin of
the stomach to the circular canal is about 9 millims. The radial canals usually run
straight from the stomach to the circular canal, and in one fragment the canal
system is quite normal, but some fragments show that the short portion of the canals,

U 2

148 CEYLON PEARL OYSTER REPORT.

between the termination of the gonads and the circular canal, has a strong tendency
to curve and to send out lateral branches, which occasionally unite with lateral
branches from an adjacent canal, or the union of two or three canals may occur, so
that just near the margin of the umbrella the radial canal system appears to be very
irregular.

The gonads are situated upon the radial canals and extend almost from the stomach
to within a short distance of the circular canal, the distance from the termination
ot the gonad to the circular canal being about 2 millims. to 3 millims. The gonads
are arranged in a lateral band along each side of the radial canals. At first a radial
canal is merely a narrow, slender, inconspicuous tube (fig. 14, B.), then when the gonads
begin to develop, the wall of the canal becomes thicker and increases in size. In this
specimen the gonads are much larger than in the Maldive specimens. They have the
appearance of cylindrical sacs, about 6 millims. in length and 1 millim. in diameter,
with the wall slightby crumbled. Between the canals bearing the fully-developed
gonads there are, here and there, canals which are of much later growth showing
gonads in various stages of development. Some of these canals are at about the same
stage as those in the Maldive specimens, showing that the Maldive specimens had
not reached their full development.

The tentacles (fig. 12) belong to the same type as those of the Maldive specimens,
but the basal bulbs have not such a long lateral extension along1 the margin of the
umbrella. I have again examined the tentacles of the Maldive specimens, and find
the extension along the margin to be slightly variable. The tentacles are also much
longer and larger than those in the Maldive specimens, but they have the nemato-
cysts arranged in the same manner. The nematocysts are in large clusters, which are
laterally situated, on both sides, along the whole of the tentacle (fig. 13).

The marginal bulbs, like the basal bulb, at first sight, as shown by the figures in
this Report and in the Maldive Report, do not appear to be similar, but I believe
that the difference in general appearance is due to a lateral contraction of the margin
of the umbrella of the specimen in this collection. The bulbs are closely packed
together, touching one another, and the sense-organs are squeezed out on to the
inner margin of the umbrella (fig. 15). This lateral contraction would also explain
the shortness of the basal bulbs of the tentacles upon the margin. In the genus
JEquorea, excretory pores are present upon the inner side of the circular canals, one
opposite each tentacle or bulb. In my description of this species in the Maldive
Report I did not mention the excretory pores, for the simple reason that I could not
see any. But I have now cut a series of sections of a marginal bulb and found the
pore in the usual place just above the velum. There is no trace of any external
papilla or swelling, but simply a slender, narrow tube running from the circular canal
to the exterior. It is just like a slit in the wall of the circular canal.

The sense-organs (fig. 15) are on the inner side of the margin of the umbrella, and
are arranged in groups These groups are placed midway between the marginal

MEDUSAE. 149

bulbs. Between two tentacles I counted the number of bulbs and sense-organs, and
found that there were 10 bulbs and 20 sense organs. The latter were arranged in
numbers thus: 1.1.2.3.1.1.1.2.2.2.2.2. From the examination of other groups of
sense-organs it may be said that there are either one or two, rarely three sense-organs
between every two bulbs. A sense-organ contains two otoliths. The figure (11)
shows the shape of the vesicle and the position of the otoliths, but the minute
details of structure are somewhat diagrammatic.

It is impossible to estimate the number of tentacles, radial canals, &c, which the
specimen should have, as the fragments are only a portion of the whole medusa.
The tentacles are about 5 millims. to 8 millims. apart, and between them there are
about 8 to 12 marginal bulbs, and about 4 to 8 radial canals.

Distribution : — Indian Ocean.

Order: TRACHOMEDUS^E.

Family: OLINDIID^E, Haeckel, 1877; ex Browne, 1904.

Gcmionemus, A. Agassiz, 1862.

Gonionemus hornelli, n. sp. — Plate I., fig. 6 ; Plate II., fig. 4.

Description. — Umbrella hemispherical, with moderately thick walls, about twice as
broad as high. Velum fairly broad. Stomach cross-shaped, having four perradial
lobes, situated upon a short, broad, cone-shaped peduncle. Mouth with four short
lips. Four broad radial canals, upon which are situated the gonads. Gonads small
in size, deeply folded and lobed, extending laterally from the canals and close to the
velum. Tentacles about 70, arranged in 16 groups, and all have an adhesive disc
about half-way down. Sixteen internal sense-organs, oval in shape, with a single
otolith.

Size : — Umbrella 6 millims. in width and 3 millims. in height.

Locality : — Pearl Banks, Gulf of Manaar.

Notes. — The single specimen is in an excellent state of preservation and in perfect
condition. The gonads are not papilliform, but are deeply folded and extend outwards
on both sides of the radial canals. They are about twice as broad as high, and contain
ova of a fan size. On one of the radial canals there is an additional gonad, smaller in
size, and not far from the stomach. It may be regarded as an abnormal growth, as
the other three canals show no signs of a gonad in that position.

The tentacles are arranged in 16 groups, but the grouping is not so well marked as
in the genus Gossea. The tentacles forming a group are not of the same size, which
is due to development. The perradial and interradial groups each contain five
tentacles, the adradial four tentacles. The central tentacle in each group is the
largest, the tentacles on each side of the central one come next in size ; the two
outside tentacles vary very much in size, one is always very small. The attachment

150 CEYLON PEARL OYSTER REPORT.

of the basal part of the tentacle to the ex-umbrella varies in length according to the
age of the tentacle. It proceeds furthest up the umbrella in the oldest tentacles and
less far in the other tentacles, showing well the arrangement of the tentacles in
groups. There is a semi-globular basal bulb on the inner side of each tentacle, and
for a short distance the base of the tentacle is attached on its outer side to the margin
of the ex-umbrella, being partly embedded in a groove. The tentacles are covered
with nematocysts, which are arranged in transverse bands. The adhesive disc is on
the outer side of the tentacle, forming a slightly raised elongated loop, and as it
extends about half-way across the tentacle it is easily seen. All the tentacles of this
specimen are contracted, and in this condition the adhesive disc is about half-way
down the tentacle.

The sense-organs are inside the margin of the umbrella, adjacent to the circular
canal, and their position is between the groups of tentacles.

Olindias, F. Muller, 1861.
Olindias, sp. ?

There is only one specimen, which is in bad condition. The umbrella is about
6 millims. in diameter. The stomach is fairly large and cross-shaped. The mouth
has four lips and its margin is slightly folded. Four perradial canals, and about three
centripetal canals in each quadrant. The gonads extend over the outer half of the
radial canals and are arranged in papilliform clusters. The margin of the umbrella is
torn and damaged. There are two kinds of tentacles ; the primary tentacles have a
few spiral bands of nematocysts and a horseshoe-shaped terminal cluster, the secondary
tentacles have numerous bands of nematocysts. Upon the margin of the umbrella
there are a number of large bulbs which look like the basal bulbs of the secondary
tentacles which have been broken off, and also a number of small bulbs. An internal
sense-organ lies at the base of some of the primary tentacles, but this could only be
seen here and there, owing to the opaqueness and damaged condition of the margin.

The specimen may be Olindias singularis, found at the Maldives, but it is not in a
condition suitable for an accurate determination of the species. It was found amongst
sea- weed at Galle, on February 17.

Note on the Olindiidse.

In my Report on the Hydromedusse of the Maldive Islands I revised the genera of
the Olindiidse. but did not know till too late that Professor Seitaro Goto had
published a paper on " The Craspedote medusa Olindias and some of its Natural
Mlies " in the ' Mark Anniversary Volume.' It was not until several months after
the publication of my paper that I was able to obtain a copy of the volume, and later
on Professor Goto kindly sent me a reprint of his paper.

Goto has also revised the Olindiidse, but excludes from the family the genera

MEDUSA. 151

Aglauropsis and Gossea, which have not an adhesive disc on the tentacles. About
the genus Olindias we differ, and it is quite likely that I may be in the wrong. We
both examined specimens sent out from the Zoological Station at Naples. I came to
the conclusion that the primary (ex-umbrellar) tentacles had not a terminal adhesive
disc, but Goto has expressed an opposite opinion. It is an important point in the
classification and could, no doubt, be quickly settled by watching the habits of
Olindias in the aquarium at Naples.

Goto has investigated the development of the sense-organs of OUndioides formosa,
Goto, and has come to the conclusion that they are entirely derived from the ectoderm.
On the ground that the sense-organ is ectodermal, Goto transfers the Olindiidaa from
the TrachomedusaB to the Leptomedusre and places them under the Eucopidse. In
this Report I have left the Olindiidse in their old place for convenience sake, not that
I dispute Goto's account of the development of the sense-organs, but rather that I
am doubtful about their being true Leptomedusse.

In 1901, when I was examining the medusae brought back from the Falkland
Islands by Mr. Rupert Vallentin (I regret that the report on the collection is still
unfinished, but hope to finish it next year), I cut some sections of the sense-organs of
Aglauropsis concmbii. The sense-organ lies in a corner, formed on one side by the
ectoderm containing nematocysts on the margin of the umbrella, and on the other
side by the endoderm of the circular canal. It is a globular vesicle containing an
otolith upon a short stalk. The wall of the vesicle is composed of a single layer
of cells which are in contact with the ectoderm, but isolated from the endoderm by
what looks like a layer of mesoglcea. As this layer took a definite shape and stained
a much deeper colour than the mesoglcea seen elsewhere, I, not knowing its origin,
was doubtful about its really being mesoglcea. I was puzzled for a time over the
sense-organ, not being sure whether the cells of the vesicle were ectoderm or
endoderm, but finally came to the conclusion that the deeply-stained layer between
the vesicle and the endoderm had some connection with the sense-organ and regarded
the whole sense-organ as endodermal. As the sections showed that the preservation
was not suitable for histological work (the specimens were preserved in formalin), I
did not attempt to trace the development of the sense-organ.

After reading Goto's description of the development of the sense-organ of
OUndioides, I again examined the sections of Aglauropsis. I am now inclined
towards the view that the vesicle is ectodermal, and that it is cut off from the
endoderm by mesoglcea, but before coming to a definite conclusion I should like to see
earlier stages in development.

Family : GERYONIID.E, Eschscholtz, 1829 ; ex Maas, 1893.

Trachomedusae, with four or six radial canals, in the course of which are situated
leaf-shaped gonads. Blind centripetal canals. Stomach on a long peduncle.
Internal sensory vesicles.

152 CEYLON PEAEL OYSTER REPORT

Liriope, Lesson, 1843 ; ex Maas, 1893.
Geryoniidse, with four radial canals and with four or eight tentacles.

Liriope tetraphylla (Chamisso et Eysenhardt), 1820.

Geryonia tetraphylla, Chamisso et Eysenhardt (1820, p. 357, plate xxvii.).

Liriantha tetraphylla, Haeckel (1879).

Liriope tetraphylla, Vanhoffen (1902, p. 82, taf. x.); Browne (1904, p. 738, pi. liv., fig. 3).

The collection contains 19 specimens; only a few are in fairly good condition
There are a few early and intermediate stages, but their condition is not satisfactory
for a description. The largest specimens are similar to a figure given by Vanhoffen.

When I wrote the ' Report on the Hydromedusse of the Maldive Islands ' I was not
quite certain about the correctness of the identification of a Liriope which I called
L. tetraphylla (1904, plate liv.). I have again examined this specimen (there was
only one in that collection) and have come to the conclusion that it must be regarded
as Liriope tetraphylla. I have failed to find a character by which it could be
specifically separated from those in the collection from Ceylon.

Notes on the Largest Specimens. — The shape of the umbrella is similar to that in
the figure given by Vanhoffen, and is not so thick or so rounded as in that figured
by me in the Maldive Report. The peduncle of the stomach is long and tapering ;
its length in the largest specimen is about 13 millims. Along the peduncle run four
interradial, longitudinal muscle bands, which bifurcate at the base of the peduncle
and the two ends curve outwards. The stomach is large and sac-shaped. The
gonads vary very slightly in shape. They resemble Vanhoffen's figure, and
measure 7 millims. in width and 5 millims. in length. The space between the gonads
(measured from the upper margins) is about 2 millims. The radial canals are fairly
broad, and that part of the canal between the gonad and the circular canal is much
broader than as figured by Vanhoffen and myself. Most of the specimens have
three centripetal canals in each quadrant. They are broader and less tapering than
those shown in the figures mentioned above. One of the specimens has only one or
two centripetal canals in each quadrant. A few of the specimens have eight tentacles,
but the majority have only the four perradial tentacles.

Size : — The largest specimen measures 15 millims. in width and 7 millims. in height.

Locality : — Cheval Paar, off Mutwal Island and Chilaw Paar, various dates in
March and November ; Galle Bay, in June and July.

Distribution : — Atlantic and Indian Oceans.

Order: NARCOMEDUS.E.
Family: iEGINID^E, Gegenbaur, 1856; ex Maas, 1904.
Solmundella, Haeckel, 1879; ex Maas, 1904.
iEginidae with two tentacles and with a stomach having eight pouches.

MEDUSAE. 153

Solmundella bitentaculata (Quoy et Gaimard), 1833 — Plate IV., figs. 1 to 6.

Charybdea bitentaculata, Quoy et Gaimard (1833, tome v., p. 295, plate xxv., figs. 4 and 5).

JEginella bitentaculata, Haeckel (1879).

Solmundella bitentaculata, Browne (1901, p. 711, plate lvi., fig. 3).

Description of the Adult. — Umbrella cone-shaped, usually a little broader than
high. Stomach circular and flat, nearly as wide as the umbrella, having eight lateral
poucbes which are rectangular in shape and about twice as broad as high. Mouth
circular, with an everted rim. Gonads on the inner wall of the pouches and also
extending over the outer half of the lower wall of the stomach, forming a continuous
band. Two opposite tentacles, which are situated above the stomach, and are about
two to three times longer than the diameter of the umbrella. Peronial bands and
grooves present. Sense-organs 24, perhaps more, usually three in each octant.

Size : — Umbrella up to 9 millims. in height and width.

Locality : — Galle, in February and August ; Modragam Paar and Cheval Paar, in
November ; and Trincomalee.

Distribution : — Australasian seas ; Amboina Island (Quoy et Gaimard). Singapore
(Bedford ; in Coll. E.T.B.). Indian Ocean ; Maldive Islands (Bigelow, 1904, p. 261,
under the name of JEginella dissonema ; and Browne).

Notes. — The collection contains 39 specimens ; only a few are in good condition, and
most of them are about 3 millims. to 5 millims. in diameter.

The umbrella is cone-shaped and nearly as high as broad. There is a slight
variation in its shape, as the apex is more rounded in some specimens than in others.
All the specimens have the apex of the umbrella more or less battered down so that
it is impossible to note its exact shape, but it is not so pointed as that shown in the
figure given by Quoy and Gaimard. The peronial groove below each tentacle is
very deep, and goes right back to the wall of the sub-umbrella. The stomach is
circular and flat and has eight lateral pouches. The upper w7all of the stomach is
either flat or slightly convex. The lower wall is also flat, with a circular mouth in
the centre. The mouth, when fully expanded, is almost as wide as the diameter of
the stomach. Its natural size is apparently about one-third to one-quarter the
diameter of the stomach, but when closed the opening is very small. The margin of
the mouth has an everted rim, and it does not usually hang down so low as in the
specimen figured by me in the Maldive Report.

Haeckel, in his description of JEginella dissonema, and also Mayer (1900, p. 66,
plate xiv.) state that there are four double perradial canals, each canal being divided
into two by a longitudinal septum (called by Haeckel the peronium). The
appearance of a double radial canal was seen in the two Maldive specimens, and also
very plainly in some of the specimens in this collection, especially when the umbrella
had been lightly stained. Transverse sections, however, do not confirm the presence
of radial canals, and, after cutting several complete series, I have come to the
conclusion that they are a delusion.

x

154 CEYLON PEAEL OYSTEE EEPOET.

Description of the Peronia. — It is in the perradii, without the tentacles, that the
appearance of a double canal is best seen, and transverse section in this position
shows the " septum" but no canals (fig. 1).

In the two perradii, which have the tentacles, there is a longitudinal groove, the
peronial groove, running from the margin of the umbrella up to the tentacle. This
groove is very deep, running back to the wall of the sub-umbrella, cutting the wall of
the sub-umbrella nearly in two (fig. 2). At the bottom of this groove is the
peronial band (figs. 2 and 6), which runs from the margin of the umbrella to the
base of the tentacle. The peronial band is a solid cord of ectoderm cells, nearly
circular in transverse section, and surrounded by mesogloea, except on the side facing
the peronial groove. In the lower wall of the stomach there are two little funnel-
shaped pockets, one under the root of each tentacle. Sections show that the ectoderm
of the lower wall of the stomach, at the apex of the pocket, unites with the peronial
band, and is continuous with the ectoderm of the tentacle. In the ectoderm of the
tentacles there are large round nematocysts. These nematocysts form a conspicuous
band along the under or lower side of the tentacle near its base (fig. 6) and then, a
little further along, spread all round the tentacle. I have found similar nematocysts
in the ectoderm of the pockets in the lower wall of the stomach and scattered among
the generative cells (fig. 2) adjacent to the pockets. They are also in the strand of
ectoderm between the apex of the pocket and the tentacle, but not in the peronial
band, which is between this point and the margin of the umbrella. It seems to me
that the nematocysts develop in the lower wall of the stomach in the neighbourhood
of the pockets, then migrate into the ectoderm of the pocket and pass along the
strand to the ectoderm of the tentacle.

The "septum" in the perradii, without tentacles, has the same structure as the
peronial bands connected with the tentacles, but there is no peronial groove and the
band (" septum ") is completely surrounded with mesoglcea. It starts from the
margin of the umbrella, runs up the side of the wall of the sub-umbrella, and at the
level of the lower wall of the stomach it curves outwards and passes through the
jelly to the ex-umbrella. In its passage through the jelly it tapers out almost to a
point, and in some specimens stops a little way short of the ex-umbrella. Its presence
marks the former existence of a tentacle, and shows that Solmundella is descended
from a medusa which had four perradial tentacles.

The appearance of radial canals on each side of a " septum " is, in my opinion, due
to the transparent mesoglcea in the short interval between the gastric pouches.

Sections across the margin of the umbrella do not show the existence of a definite
circular canal.

Gonads. — Some of the specimens have the gonads confined to the inner wall of the
gastric pouches, where they lie in the ectoderm (figs. 1 and 2). The gonads may
extend over the lower half of each gastric pouch or over the whole pouch. Some
of the large specimens have the gonads not only over the gastric pouches but

MEDUSAE. 155

also over a part of the lower wall of the stomach, forming a continuous ring round
the lower wall of the stomach just like the genital ring of a Solmaris. One
specimen has the outer half of the lower wall of the stomach covered with ova,
which are large and clearly visible ; other specimens have only one quarter or one
third of the wall of the stomach occupied with gonads. It appears from the specimens
that the gonads first start developing at the bottom of the pouches, and then spread
upwards and finally reach the lower wall of the stomach. The smallest specimens
have the gonads confined to the pouches, but it is only in the largest specimens that
the tronads are on the wall of the stomach.

Tentacles. — My figure of Solmundella in the Maldive Report shows that the base
or root of the tentacles is curved outwards towards the ex-umbrella. This I now find
is not the normal position, but the position occasionally taken when a specimen is in
a contracted condition. As a rule the root of the tentacle points towards the centre
of the umbrella (fig. 6), and in specimens which do not show signs of contraction it
is sometimes clear of the upper wall of the stomach and the curve is scarcely visible.
The tentacles have numerous internal transverse septa (fig. 5) which are connected
in the centre by an elongated endoderm cell, containing usually two nuclei. The
lower part of the tentacle (fig. 4) is somewhat triangular in shape ; along this portion
there is a longitudinal muscle band.

Sense-organs. — The smaller specimens have two sense-organs and the largest ones
three and perhaps more in each octant. In certain octants I have seen extra bulbs
without sense-organs, and these may be the bases of sense-organs which have lost the
otolithic part through injury.

A few of the specimens are infested with a Cercaria.

SIPHONOPHORA.

Order : CALYCOPHOJLE, Leuckart.

Family : DIPHYID^E, Eschscholtz, 1829.

Diphyes, Guvier, 1817.

Diphyes chamissonis, Huxley, 1859.

Diphyes chamissonis, Huxley (1859, p. 36, pi. i., fig. 3) ; Browne ^1904, p. 742, pi. liv., fig. 6).

The collection contains eleven anterior nectophores, some of which are in very good
condition. The specimens are similar to those which were described and figured by
me in the ' Report on the Hydromedusae of the Maldive Islands.'

One specimen is from Galle, in July, but all the rest were from the Gulf of Manaar,
mostly in February and March.

The nectophores measure about 8 millims. to 11 millims. in length. The somatocyst

x 2

156 CEYLON PEARL OYSTER REPORT.

shows considerable variation in length and thickness. Some of the specimens have the
somatocyst similar in shape and size to that shown in my figure of the species, whereas
in other specimens it is longer (the length varies from 2 millims. to 3^ millims.) and
much thinner. The length of the hydrcecium is also variable, about one-third to half
the length of the umbrella.

Order : PHYSOPHOILE, Eschscholtz, 1829.

Family : AGALMIDiE, Brandt, 1835.

Cupulita, Quoy et Gaimard, 1824.

There is one small specimen of a Cupulita, from the Cheval Paar, which is very
much broken up. I am unable to determine the species.

Agalmopsis, Sars, 1846.

There are two small specimens, from the Cheval Paar, both of which are badly
contracted and broken. The nectophores have all disappeared with the exception of
a few minute buds, and only one damaged bract remains. The tricornuate tentilla
are large and in excellent condition.

Family : PHYSALIID^, Brandt, 1835.

Physalia, Lamarck, 1801.

Physalia utriculus, Eschscholtz, 1829.

Physalia utriculus, Huxley (1859, p. 101, pi. x., pi. xii,, fig. 12); Browne (1904, p. 744).

Two small specimens were caught off Watering Point, Galle. The float is about
1 5 millims. in length. There is one main tentacle and several very small secondary
tentacles. The gonophores are beginning to develop.

Family : PORPITIDyE, Brandt, 1835.

Porpita, Lamarck, 1801.

Porpita is represented by the remains of a single float, obtained on the Pearl Banks,
Gulf of Manaar, and measuring about 35 millims. in diameter. The upper surface of
the float has numerous radial rows of stigmata on the back of prominent ridges. It
resembles the float of Porpita umbella, which is figured by Haeckel (l£
plate xlv., fig. 5).

MEDUSAE. 157

SCYPH0MEDUSJ1.

CHARYBDEIDA.

Charybdea, Peron et Lesueur, 1809.
Charybdea, sp. ?

There is a single specimen in the collection from the pearl banks, and it is not in a
first rate condition. The umbrella has become soft and limp, consequently it has
collapsed and lost its natural shape. The umbrella measures about 75 millims. in
length, and is probably cone-shaped. The stomach is very short and flat ; the
mouth has small lips. The gastric filaments appear to be perradial in position (the
top of the umbrella is damaged and crushed in). Each of the four groups is
composed of about six tufts of filaments packed so close together as to form a
continuous row. The sense-organs are about 10 millims. away from the margin of
the umbrella. There are four ocelli on the inner side of each tentaculocyst. The
principal ocellus is very large and semi-globular in shape. Above it, a little nearer
the base of the tentaculocyst, is a transverse ocellus, forming a narrow pigmented
band. The other two ocelli are more lateral in position, and situated between the
semi-elobular and the transverse ocelli. The ocelli are of a reddish brown colour in
formalin. The velarium contains seven unbranched canals between every two
tentacles. The gonads form very narrow bands, and appear to be quite immature.
The pedalia are about 20 millims. in length and 1 5 millims. in width. The shape of
their wings and the tentacle resemble the figure of Charybdea grandis (Agassiz and
Mayer, 1902, plate vi.).

This may be an immature specimen of Charybdea grandis, but I remain uncertain.
A second specimen would have been an advantage for comparison.

CORONATA.

Family: NAUSITHOID^l, Haeckel, 1879; ex Vanhoffen, 1902.

Nausithoe, Kolliker, 1853.

Nausithoe punctata, Kolliker, 1853.

Nausithoe punctata, Vanhoffen, 1892, p. 13, Taf. iii., figs. 8 and 9; Mayer, 1900, p. 67,
plate xxiii., figs. 67 and 68, plate xxvi., figs. 87 and 88; Vanhoffen, 1902, p. 29;
Bigelow, 1904, p. 263, plate vi., fig. 21.

Description. — The umbrella is somewhat hemispherical in shape. At the top of
the umbrella there is a distinct hemispherical crown which is separated off from the

158 CEYLON PEARL OYSTER REPORT.

rest of the umbrella by a conspicuous circular furrow. Just below the circular
furrow the radial furrows begin. There are 16 deep radial furrows on the
ex-umbrella, one midway between every tentacle and sense organ, terminating at
the base of the marginal lobes. The bottom of each furrow is attached to the wall
of the sub-umbrella by a septum which divides the distal portion of the stomach
into 16 pouches (8 ocular and 8 tentacular). The septum is continued for a little
way down the middle of each marginal lobe, separating the prolongation of the
stomach in each lobe into two parts. But as the septum does not proceed along the
whole length of the gastric prolongation, two completely isolated pouches are not
formed. The whole of the ex-umbrella, including the marginal lobes, is closely
granulated.

Tbe gastric filaments are arranged in four distinct groups, which are isolated from
each other by the four basal angles of the cross-shajjed mouth. There are about 10
to 12 filaments in each group arranged in a single row. Each group occupies the
whole space between the angles of the mouth.

The mouth is large and cross-shaped, about 3 millims. in length and width.

The gonads vary in shape, and, looked at from the sub-umbrella, appear circular or
oval. The largest are about 1 millim. in length and 0'75 millim. in width. Three
specimens in one bottle have rose-red gonads, and two in another bottle are of an
orange colour. All the specimens are in formalin. One is a male and four are
females with large ova.

There are 16 marginal lobes, which are about as broad as long (2 millims.), and
have a rounded edge. Between these lobes are the eight tentacles and eight sense-
organs, which alternate with each other. The tentacles are of moderate length
(about 5 millims.), stiff, and taper to a fine point. The sense-organs have an otolithic
sac and a circular reddish pigmented ocellus.

The collection contains five specimens, three of which are in splendid condition. Two
were from off Mutwal Island on March 19, and three from Muttuvaratu Paar on
March 29. The largest measures 9 millims. in width and 7 millims. in height. Two
specimens are 9 millims. in width and 5 millims. in height. The others are slightly
smaller.

I have compared these specimens with Nausithoe punctata obtained from the
Zoological Laboratory at Naples, and feel certain that they belong to this species ;
in fact, they agree in every detail except in the shape of the ocellus. The Naples
specimens have a circular pigmented ocellus on a semi-circular or convex bulb,
whereas in the Ceylon specimens the ocellus forms a pigmented ring on a bulb with a
flat surface.

The results obtained by the " Valdivia " and " Siboga " Expeditions show that
Nausithoe punctata has a very wide geographical distribution. It occurs in all the
oceans. It was taken by the "Valdivia" off the east coast of Ceylon, and by
Bigelow at the Maldives.

Pelagia, sp. ?

MEDUSAE. 159

DISCOPHORA.

SEM^OSTOMATA.
Pelagia.

There are nine very young stages, the smallest 4 inillims. in diameter and the
largest 8 millims. They have eight tentacles and eight sense-organs. These specimens
are too immature for me to identify, as they have not long passed through the Ephyra
stage. They are all from the Cheval Paar, Gulf of Manaar.

RHIZOSTOMATA.

Family : LYCHNOKHIZID^E, Maas, 1903.

Crambessa.

Crambessa, sp. ?

The collection contains two specimens, both from Galle Bay, June and August ; one
is in fairly good condition and the other is damaged.

Umbrella. — The umbrella is semi-globular, about twice as broad as high, and
measures about 75 millims. in width and about 40 millims. in height. The ex-umbrella
looks smooth, but a close examination with a lens shows that the surface is closely
covered with very minute papilla?, which give it a granulated appearance. The
ex-umbrella of one specimen has fine markings which look like a pattern produced by
pressure against a tow-net. The pattern forms a network with a mesh of about half
a millimetre.

Canal System. — There are eight ocular canals and eight adradial canals. The
ocular canals run to the sense-organs, but the adradial canals stop at the circular
canal and do not proceed to the margin. The circular canal, which is broad and
conspicuous, is situated about 10 millims. from the margin of the umbrella. Between
the circular canal and the margin of the umbrella the canal system forms a network
of fine meshes. The ocular canals pass through this network and anastomose with it.
On the inner side of the circular canal and between the radial canals there is a very
coarse network of canals. This network is in communication with the circular canal,
but not with the stomach. In one specimen there is a slight anastomosis of the inner
network with some of the radial canals, but in the other specimen there is no union.

Margin of the Umbrella. — Some of the velar lobes are about as long as broad,
somewhat quadrangular in shape, with rounded corners, and some are narrow and
more pointed. There are about eight velar lobes between every two ocular lobes.

160 CEYLON PEARL OYSTER REPORT.

Sense-organs. — Eight sense-organs are present. The outer sensory pit is triangular
in outline and its surface is folded. The principal folds radiate outwards from the
bottom of the pit. The tentaculocyst is apparently without an ocellus, as there is no
trace of any pigment. The ocular lobes are much smaller than the velar lobes, and
are pointed.

Sub-umbrella Muscles. — In one specimen the sub-umbrella muscles have become
detached and a clear view of the canal system is obtained. In the other specimen
the muscles are present, and they form a circular band between the periphery of the
oral disc and the margin of the umbrella. The circular muscle band is continuous and
is not radially interrupted.

Sub-genital Cavity. — The four sub-genital ostia open into a common continuous
cavity. The ostia are very large, forming long but narrow slits, about 20 millims. in
width, and about as wide as the columns. The entrance is partly blocked in the
centre by a large triangular gelatinous knob on the sub-umbrella, and just inside
there is another median knob and also two small lateral ones.

Oral Arms. — In a normal specimen there should be eight oral arms of equal length,
but in both of these specimens the oral arms are abnormal in number and in length.
One specimen has ten oral arms, the four columns bearing respectively 2.2.3.3. arms.
The arms show a great difference in size, the largest is about 80 millims., and the
smallest about 25 millims. As one arm is much longer than the others, which are
all of different lengths, it is probable that the medusa received an injury in the oral
arms, and regeneration has followed. The upper arm is very short and is somewhat
laterally compressed. In the arm, measuring 80 millims. in length, the upper arm is
about 15 millims. and the lower arm about 65 millims. The lower arm has three thin
wings bearing oral mouths along the outer edges down to the distal end, which does
not bear a gelatinous knob. The oral mouths on the ventral wings are continued
along the upper arms to the oral disc, where they meet and form a cross-shaped
pattern. There are no special appendages of any kind upon the arms or the oral disc.
In the second specimen the arms are broken off close to the arm disc and there are
stumps of nine, possibly ten, arms.

Stomach. — The stomach is cross-shaped. The gastric filaments run round the
margin of the stomach and also curve downwards and inwards, forming a loop in the
base of the columns. It is at the end of the loop that the canal from the oral arm
enters the stomach.

The gonads are immature.

The specimens are of a whitish colour in formalin.

So far as I can make out, these specimens belong to the genus Crambessa, but I am
not able to determine the species. As they are immature they are probably at an
intermediate stage in growth and may develop into a species which has already been
described.

MEDUSAE. 161

CTENOPHOEA.

Order : CYDIPPIDEA, Lesson.

Family : PLEUROBEACHIIDiE, Chun, 1880.

Pleurobrachia, Fleming, 1822.

Pleurobrachia globosa, Moser, var. ceylonensis, nov.

Pleurobrachia globosa, Moser (1903, p. 7, taf. i., figs. 1-4).

The collection contains about 900 specimens, varying in size from about 2 millims.
up to 8 millims. in length. The largest specimen measures 8 millims. in length and
7 millims. in width.

A few are preserved in formalin and these have been used for examination, as they
have retained their shape better than those in alcohol. The specimens in alcohol
suffer more or less from contraction and shrinkage. When contraction or shrinkage is
considerable, the position of the various organs changes so much that it would be
quite possible to make two or more species out of a hundred specimens.

Locality : — A few were obtained on the Cheval Paar in March and Modragam Paar
•in November ; the great majority were from Galle Bay in June, July and August.

Description. — The body is egg-shaped, sloping towards the oral pole, and almost
circular in a transverse section. There are eight rows (costse) of ciliated plates, of
moderate length, extending over half, or a little more than half, of the meridional
surface, beginning and terminating at about ecpial distances from the aboral and oral
poles respectively. Each row contains about twenty narrow ciliated plates (combs).
The meridional canals are just as long as the costae. The two tentacles and their
sheaths lie above the level of the stomach. The base of the sheath is in the first fork of
the gastrovascular canals, on a level with the funnel. The sheath is like a long cone,
tapering from the base, and lying at an angle of about 45 degrees from the
perpendicular axis of the body and pointing towards the aboral end of the body. The
opening of the sheath on the surface of the body is just under the aboral boundary
line of the costae. At the bottom of the sheath is the base of the tentacle, which is
somewhat concave. The tentacles have lateral filaments, but no eolidiform appendages
were seen. The transverse canals from the funnel to the meridional canals slope
slightly in the aboral direction and join the meridional canals in the middle of their
length, slightly above the level of the funnel. The base of the tentacle lies a little
way from the funnel, but in a large number of specimens, owing to contraction or
shrinkage, the base of the tentacle is adjacent to the funnel. It has contracted back
on to the funnel and the top of the stomach.

On comparing the specimens with the figures of Pleurobrachia globosa (Moser,
1903, taf. i., figs. 1-3), I find that they differ mainly in the length of the costse, which

Y

162 CEYLON PEARL OYSTER REPORT.

are about half as long again. There is a slight difference in the shape of the body,
the specimens from Ceylon taper more towards the oral pole. The position of the
tentacular sheath is identical and so also is its sheath opening. As the chief
difference lies in the length of the costse, I hesitate to add a new species to the genus,
but prefer to mark the difference by establishing a new variety.
The " Siboga" specimens were obtained in the Malay Archipelago.

Order: BEftOIDEA, Lesson.

Family: BEROLLLE, Eschscholtz, 1829.

Beroe, P. Browne, 1756.

Beroe flemingi (Eschscholtz), 1829.

Pandora flemingii, Eschscholtz (1829, p. 39, taf. ii., fig. 7).

Beroe pandora, Moser (1903, p. 23, taf. ii., figs. 8 and 9; taf. iii., figs. 9 and 10).

There are about a dozen specimens in the collection, four of which are in fair
condition and the others in fragments.

Description. — The body is conical, compressed in the funnel (transverse) plane, a
little longer than wide, and rounded at the aboral end. The mouth is wide and has
a fairly thin margin. The costse are of unequal length ; the sub-transversal costse
are about twice or nearly twice as long as the sub-ventral costse. The meridional
canals do not unite with the stomodseal canals. The lateral canals of the meridional
canals meander without uniting in the smaller specimens, while in the larger
specimens they unite with those from the adjacent meridional canals, forming a coarse
irregular network in the outer wall of the body. Short blind canals also proceed
from the circular canal around the mouth. The meridional canals on their outer
surface are sparsely sprinkled with minute reddish-brown spots of pigment. The
gonads are along the walls of the meridional canals, male and female on opposite sides
of the canals.

Size: — 8 millims. long and 6 millims. wide, 10 millims. long and 10 millims. wide,
1 2 millims. long and 8 millims. wide. Larger specimens broken into fragments.

Locality: — Off Mutwal Island, March 19, twelve specimens; Galle Bay, July 15,
one specimen.

Distribution : — N. Pacific, east of Japan. Malay Archipelago.

Miss MosKR,in her 'Report on the Ctenophora of the "Siboga" Expedition,' has revised
the Pleurobrachiidse and Beroidas, and has given a useful key for the identification of
the species. I have tried to identify these specimens with the aid of the key, but
remain somewhat doubtful about the result. The difficulty of the identification is no
doubt increased by my want of experience in the group and by the fact that the best
specimens are early stages. Taking the unequal length of the costa? and the absence

MEDUSJE. 163

of cilia round the mouth as a guide, the specimens come nearest to Beroe flemingii.
They do not, however, quite agree in shape, and the lateral canals from the meridional
canals do not communicate with the stomodseal canals. Miss Moser states that the
Sihoga specimens are identical with Pandora flemingii of Eschscholtz. This species
Eschscholtz named after the English zoologist John Fleming. Miss Moser,
however, has changed the specific name to pandora, wishing to retain the generic
name Pandora as a specific name in honour of Eschscholtz. The changing of the
specific name of this species is certainly contrary to the International Rules on
Nomenclature, and consequently the specific name pandora is invalid.

REFERENCES TO LITERATURE.

Agassiz, A., and Mayer, A. G. — 1902. "Keports on the Scientific Results of the Expedition to the

Tropical Pacific .... in the 'Albatross.' — Part iii. Medusae." 'Memoirs Mus. Comp. Zool.

Harvard,' vol. xxvi., pp. 139-176, plates i.-xiii.
BlGELOW, H. B. — 1904. "Medusa? from the Maldive Islands." 'Bull. Mus. Comp. Zool. Harvard,'

vol. xxxix., pp. 245-269, 9 plates.
Brooks, W. K. — 1895. "The Sensory Clubs or Cordyli of Laodice." ' Journ. Morphol. U.S.A.,' vol. x.,

pp. 287-304, plate xvii.
Browne, E. T. — 1904. " Hydromedusae, with a Revision of the Williadae and Petasidse." ' Fauna and

Geogr., Maldive and Laccadive Archipel.,' vol. ii., pp. 722-749, plates liv.-lvii. Cambridge.
Browne, E. T. — 1905. "Scyphomedusre." 'Fauna and Geogr., Maldive, &c, Archipel.,' vol. ii.,

pp. 958-971, plate xciv. Cambridge.
Chamisso, A. de, and Eysenhardt, C. G. — 1820. "De Animalibus quisbusdam e Classe Vermium

Linneana in circumnavigatione Terra; .... annis 1815- 1818 peracta, observatis." 'Nova Acta

K. Leop. Carol. Deutsch. Akad. Naturforseher,' Band x.
CHUN, C. — 1896. "Beitriige zur Kenntnis ostafrikanischer Medusen und Siphonophoren nach den

Sammkuigen Dr. Stuhlmann's." ' Mitth. natur. Mus. Hamburg,' Band xiii., pp. 1-19, 1 Taf.
Eschscholtz, F. — 1829. ' System der Acalephen.' Berlin.
Forskal, P. — 1776. 'Icones Rerum Naturalium quas in itinere orientali.' 4to, 11 pp., 43 plates.

Hauniae.
Goette, A. — 1886. " Verzeichniss der Medusen, welehe von Dr. Sander .... gesammelt wurden."

'Sitz. Akad. wiss. Berlin,' Band xxxix., pp. 831-837. Reprinted in ' Math.Naturw. Mitth.,'

Band vii., pp. 597-603.
UuTn, S. — 1903. "The Craspedote Medusa Olindias and some of its Natural Allies." 'Mark Anniversary

Volume,' pp. 1-22, 3 plates, 4to. New York.
Haeckei., E. — 1879-1880. 'Das System der Medusen.' Jena.

Haeckel, E. — 1888. "Report on the Siphonophorse." ' "Challenger" Reports,' vol. xxviii.
Huxley, T. H.— 1859. "The Oceanic Hydruzoa." 'Ray. Soc. London.'
Maas, O. — 1893. "Die craspedoten Medusen." ' Ergebnisse der Plankton-Expedition' ("National"),

Bd. ii, K.c, 107 pp., 8 Taf.
Maas, O. — 1903. " Die Scyphomedusen der Siboga-Expedition.' 'Siboga Expedition.' Monogi-aph, xi.

91 pp., 13 Taf. Leiden.

Y 2

164

CEYLON PEARL OYSTER REPORT.

Maas, 0. — 1904. "Revision des Meduses appartenant aux Families des Cunanthidse et des iEginidae, et
groupement nouveau des genres." ' Bull. Mus. Oceanographique de Monaco,' No. 5, 8 pp.

Mayer, A. G. — 1900. "Some Medusae from the Tortugas, Florida." 'Bull. Mus. Comp. Zool. Harvard,'
vol. xxxvii., pp. 11-82, plates i.-xliv.

Moser, Fanny. — 1903. "Die Ctenophoren der Siboga-Expedition." ' Siboga Expedition.' Monograph,
xii., 34 pp., 4 Taf.

Quoy, J. R., et Gaimard, P. — 1833. " Voyage de decouvertes de 1' ' Astrolabe.' " ' Zoologie,' tome iv.
Paris.

Vanhoffen, E. — 1892. ' Die Akalephen der Plankton-Expedition (" National "),' Band ii., K.d., pp. 1-30,
4 Taf.

Vanhoffen, E. — 1902. "Die aeraspedoten Medusen der deutschen Tiefsee-Expedition 1898-1899
(' Valdivia ')." ' Wissen. Ergebnisse,' Band iii., pp. 1-52, Taf. i.-viii. " Die craspedoten
Medusen der deutschen Tiefsee-Expedition (' Valdivia ') — I. Trachymedusen." ' Wissen. Ergeb-
nisse,' Band iii., pp. 53-86, Taf. ix.-xii.

DESCRIPTION OF PLATES.

All the figures were drawn from specimens in formalin or alcohol ; all sense-organs
and sections were drawn with a camera lucida.

Reference Letters.

A. Adhesive disc.

B. Basal bulb.
CC. Circular canal.
Ec. Ectoderm

En. Endoderm.

Ex. Ex-umbrella.

G. Gonad.

M. Mesoglcea.

N. Nematocyst.

Oc. Ocellus.

Ov. Ovum.

Or. Oral lip.

P. Peronium.

PG. Peronial groove.

R. Radial canal.

S. Sense-organ.

Sm. Septum.

St. Stomach.

Sip. Stomach (gastric) pouch.

Sub. Sub-umbrella.

T. Tentacle.

V. Velum.

PLATE I.

Fig.

1.

1)

2.

)»

3.

J)

4.

)J

5.

»J

6.

Cyttzis herdmmi, n. sp. (p. 135). Lateral view, x 30.
sEquorea arnica, n. sp. (p. 145). Lateral view, x 10.
Mitrocomium assimile, n. sp. (p. 137). Lateral view, x 20.
Irenopsis hexanemalis, Goette (p. 142). Lateral view, x 10.
Laodice indica, n. sp. (p. 136). Oral view, x 15.
Gonionemus Iwrnelli, n. sp. (p. 149). Lateral view, x 10.

MEDUSAE. 1<>5

PLATE II.

Tentacle of Dipwrma sp. 1 (p. 133). Lateral view, x 35.

Manubrium of Dipwrena, showing the gonads and stomach. (Slightly contracted.) x 20.

Tentacle of I'rolmscidnrtyln minium, n. sp. (p. 13G). Inner view, x 60.

Tentacle of Gonionemus hornclli, n. sp., showing the position of the adhesive disc (A.) Outer

view. Enlarged.
Mqwrea parva, n. sp. (p. 1-tG). Oral view, x 10.
Portion of the sub-umbrella of Mqwrea parra, showing the oral lips (Or.), the lower wall of the

stomach (St.), the gonad (<?.), and a tentacle, x 20.
Sense-organ of Mqwrea parva. Optical section, x 400.
Odocanna polynema (Haeckei.) (p. 144). Oral view, x 5.

Tentacle of Odocanna polynema, and also marginal bulbs and sense-organs. Outer view, x 40.
Sense-organ of Odocanna polynema. Optical section, x 400.
11 to 15. Mesonema pensile, Modeer (p. 147).
Sense-organ. Optical section, x 400.
Tentacle and marginal bulbs. Outer view, x 10.
Clusters of neniatocysts on the tentacle, x 80.
Portion of the stomach and sub-umbrella, showing the oral lips (Or.), the lower wall of the

stomach (St.), the radial canals (li.) and gonads (67.).

R and E" are radial canals at an early-stage development, x 5.
Marginal bulbs and sense-organs, x 16.
Portion of the sub-umbrella of Mqwrea arnica, showing the oral lips (Or.), the lower wall of the

stomach (St.), the gonads (G.) and tentacles, x 10.
Gonad (female) of Mqwrea arnica. Lateral view, x 15.
Sense-organ of Mqwrea conica.. Optical section, x 400.

PLATE III.

Fig. 1. Manubrium of Eutima currn, n. sp. (p. 138), showing the peduncle with gonads and stomach, x 9.
„ 2. Portion of the margin of the umbrella of Eutima curva, showing the basal bulb of a tentacle, the

marginal bulbs and cirri. Outer view, x 50.
,, 3. Sense-organ, marginal bulbs and cirri of Eutima currn. Oral view, x 100.
,, 4. Manubrium of Octorchis orientalis, n. sp. (p. 139), showing the peduncle with gonads and stomach.
xlO.
Figs. 5 to 8. Irenopsis hexanemalis.

Fig. 5. Tentacles and marginal bulbs, showing the excretory papilla? fully expanded. Outer view, x 30.
„ 6. Tentacles, marginal bulbs, and sense-organs (papillae contracted), x 30.
„ 7. Sense-organ. Optical section, x 400.
,, 8. The stomach and its peduncle. Lateral view, x 10.
„ 9. Irene ceylonensis, n. sp. (p. 140). Oral view of the umbrella, showing the stomach and its peduncle,

the position of the gonads, and the tentacles in one quadrant of the umbrella, x 3.
,, 10. Sense-organ of Irene ceylonensis. Optical section, x 400.
,, 11. Tentacles and sense-organs of Irene ceylonensis. Outer view, x 30.
,, 12. Irene palkensis, n. sp. (p. 141). Oral view of the umbrella, showing the stomach and its peduncle,

the position of the gonads, and the tentacles in one quadrant of the umbrella, x 4.
„ 13. Tentacles, marginal bulbs, and sense-organs of Irene palkensis. x 40.
Figs. 14 to 16. Sense organs of Irene palkensis. x 400.

Fig.

1.

»»

2.

'J

3.

J»

4.

))

5.

TJ

6.

JI

7.

»)

8.

))

9.

>J

10.

Figs

i. 11

Fig.

11.

»»

12.

)>

13.

))

14.

JJ

15.

))

16.

J>

17.

))

18

166 CEYLON PEARL OYSTER REPORT.

PLATE IV.

Figs. 1 to 6. Solmundella bitenfaculala.

Fig 1. Transverse section of the peronium (P.) in the radius without tentacles, also showing the gastric
pouches (Stp.) and the gonads (G.). x 260.
„ 2. Transverse section of the peronium (P.) in the radius with tentacles, showing the peronial

groove (P. G.) x 270.
„ 3. Transverse section across the gastric pouches (Sip.), showing the position of the peronium (P.)

in the radius without tentacles, x 20.
„ 4. Transverse section of a tentacle not far from the base, x 100.
„ 5. Longitudinal section of a portion of a tentacle not far from the distal end. x 270.
„ 6. Sketch showing the peronium (P.) running from the margin of the umbrella to the tentacle.
* Plane of section, fig. 2. ** Plane of section, fig. 4.
Fis;s. 7 to 11. Laodke indica, n. sp.
Fig. 7. Longitudinal section of a cordylus. x 350.

8. Longitudinal section through the basal bulb of a tentacle, x 150.

9. Tentacle, cirrus, and cordylus. Inner view, x 85.

10. Transverse section of an ovary, showing the escape of an ovum, x 80.

11. Terminal cluster of nematocysts of a cirrus, x 380.

12. Cytceis herdmani, n. sp. Transverse section of a tentacle, showing the nematocysts in the
ectoderm and the pigment granules in the endoderm. x 40.

CKYLON PEARL OYSTER RETORT

MKIII S V, PLATE I

2.

( KYLO.V PEARL OYSTER REPORT

MEIHS.E PLATE II

■v
cc

3.

10.

- r

Or

S!

12

17.

13.

r2 »'

R

14.

18.

16.

15

.

( KYLON PEARL OYSTER REPORT

MEDUSAE -PLATE 111

4.

■

10. %

'?,

hjr

15.

■n

,5§9

9.

X

14.

16

12.

CEYLON PEARL OYSTER REPORT

/' M

Ex

Sep

• «

-
t

.-'

MEDUSA -PLATE IV

2.

PG ,

M

Stp

Ec.

» -

En
Af

Ov

Ec

Ex

Or

10.

., -

p Ec

'■

■

'

PG

Sm

>

•

cc

'

Oc

JV

Ti

'".'' B

8
T

""•■.■

,

Ec

1/

En

:

:-s.

%s.

V- :-.,*.

12.

.. .*,.

•..

-

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

APPENDIX TO THE REPORT

ON THE

ALCYONARIA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY

Professor J. ARTHUR THOMSON, M.A., University of Aberdeen.
[With ONE PLATE and TWO TEXT-FIGURES.]

A revision of the collection of Ceylonese Alcyonarians has enabled me to add to the
list given in Supplementary Report No. XX., by Mr. W. D. Henderson and myself.*
The additions are : —

Spangodes cervicornis, Wright and Studer. Caligorgia versluysi, n. sp.

,, involute,, Kukenthal. Bebryce indica, n. sp.

,, putteri, Kukenthal. Acampiogorgia gracilis, n. sp.
„ microspiculata, Putter. ,, rubra, n. sp.

„ „ var. ceylonensis, now Muricella rubra, n. sp.

Chironephlhya indica, n. sp. Virgularia elegans, Gray.
Suberogorgia kottikeri, Wright and Studer, var. ,, calycina, n. sp.

ceylonensis, nov. ,, indica, n. sp.

,, rubra, n. sp. Fusticularia herclmani, n. gen. et sp., Simpson.

One of these [Caligorgia versluysi) is hardly an addition, since it replaces Primnoa
ellisii, von Koch (= Caligorgia ellisii, = C. verticillata). For this correction I have to
thank Dr. J. Versltjys, Amsterdam. The last on the list, Fusticularia herdmani,
represents an interesting new genus, which has been separately described by
Mr. J. J. Simpson, M.A. ('Annals and Magazine of Natural History,' June, 1905).

* By an oversight the two species of Xenia (Part III., pp. 271, 273) were misplaced; they should come
first in the order Alcyonacea. On p. 289, the title of the family Muriceidse was omitted in front of
Acanthogorgia.

I(j8 CEYLON PEARL OYSTER REPORT.

Family : NEPHTHYID7E.

Spongodes cervicornis, Weight and Studer.

A much weathered specimen, expanded in one plane, 10*5 centims. in height, by
10 centims. in breadth, and 1*8 centims. in thickness. The general colour is yellowish-
white, but the polyp stalks and the polyps are covered with orange-red spicules. The
polyps themselves are white in colour, but there appear to be minute orange-red
spicules on the tentacles.

The polyp stalk is about 2 millims. in length and is covered by longitudinally
disposed spicules. At the base of the polyp, which stands at right angles to the
Stittzbundel, there is a double ring of transverse spicules, from which there arise
eight converging triangles, each composed of two pairs of spicules, the basal pair
converging at an obtuse angle.

As the colony is markedly flattened, with an irregular outline, with leafdike lower-
most branches, the specimen must be referred to Kukenthal's cervicornis group, and
among those species in which the polyps have long stalks.

The specimen agrees, on the whole, with Sp. cervicornis, Wright and Studer,
which seems to be a very variable species, including, according to Kukenthal,
Spongodes rhodosticta, Wright and Studer,

Locality : — Ceylon seas.

Spongodes involuta, Kukenthal.

A beautiful divaricate colony, without the stalk, but otherwise almost complete.
The general surface of the ccenenchyma is white, near the polyps the colour is
strengthened by golden-yellow spicules ; the spicules supporting the white polyps are
deep rose-pink. The dimensions of height, maximum breadth and average thickness
are 45, 40, and 12 millims. As the divaricate polyparium has no regular outline and
is markedly flattened, the specimen is referable to Kukenthal's cervicornis group.
As the polyps are borne on stalks 1'5 millims. to 2 millims. in length, the specimen
must be included near Sp. cervicornis, Wright and Studer. The eight double rows
of polyp spicules consist of five pairs of spindles, and thus our attention may be
restricted to Kukenthal's Dendronephthya involuta*

Our specimen agrees with this species in the following features : — (1) there are
five pairs of polyp spicules in each double row ; (2) one of the uppermost in each row
projects above the polyp ; (3) the polyp stalk is about 2 millims. in length.

Locality : — Ceylon seas.

Spongodes putteri, Kukenthal.

A divaricate form, with a regular outline, a slightly flattened polyparium, and

* Kukenthal has referred all typical Spongodes species to his new genus Dendronephthya, but we see
little advantage in this.

APPENDIX TO ALCYONAEIA. 169

foliate lower branches, is referable to Kukenthal's rigida group, and closely
approaches Sp. piitteri. Thus the polyparium is a slightly flattened cylinder ; the
lateral branches are almost cylindrical and sometimes dichotomous ; the groups of
polyps, usually seven in number, are borne by stalks about 2 millims. to 3 millims. in
length ; the polyps are supported by eight pairs of converging spicules, one of which
projects farther than the others; the longest Stiitzbundel spicule exceeds 3 millims.,
is covered with minute thorns, and projects for a little over a millimetre. The
spicules closely resemble those of Sp. piitteri in form and size, but are of a bright
orange colour around the polyps and on the polyp stalks. It seems warrantable to
regard this specimen as a colour variety of Sp. piitteri-.
Locality : — Ceylon seas.

Spongodes microspiculata, Putter.

This species is represented by a very beautiful small colony, — 6 5 centims. in
height and 5 "5 centims. in maximum breadth. The stalk by which it is attached
is 2"5 centims. long and 1*3 centims. in diameter. The lower branches are foliaceous,
but above this there are three main branches, one central and two lateral, in length
3 centims., 2*5 centims., and 2 centims. respectively, with an average diameter of
0"8 centim. From these, smaller branches arise on all sides almost perpendicularly,
the lateral ones being the longest, so that the whole colony presents a somewhat
flattened appearance, the tips so arranged that the contour is regular. The general
colour of the colony is orange-red.

The polvps are in clusters of five to eight, seven being the most typical number.
They arise from tertiary branches, except on the lower foliaceous part, which is the
expanded portion of primary branches. The polyps are almost globular, standing at
right angles to a short stalk 1 millim. in length. The Stiitzbundel is fairly well
developed, one long spindle projecting about 0"8 millim. beyond the insertion of the
polyp. On the outside of the anthocodia the spicules have a very definite arrange-
ment, consisting of eight double rows of 6 to 7 converging pale yellow spicules, the
terminal pairs projecting a little beyond the anthocodia. Between these groups
small colourless spindles are scattered irregularly. On the tentacles the spicules are
arranged in a biserial manner, approximately at right angles to the long axis.

The spicules are warty spindles of very varied proportions, some long and slender,
others short and thick. On the main stem they are of the second type, the following
being typical measurements : — 1"2 millims. x 0-5 millim. ; 1*3 millims. x 0-4 millim. ;
2 millims. x 0'4 millim. On the secondary branches they become more elongated
and slender, and have a slight yellow tinge. Further up the colour becomes more
marked and there is a distinct orange-coloured core. The following measurements
were taken: — 2"5 millims. x 0-l millim.; 19 millims. x 0-l millim. Scattered
irregularly amongst these there are small elongated thin spindles of a bright red

170 CEYLON PEARL OYSTER REPORT.

colour of the following sizes : — 0"3 millim. x 0'02 millim. ; 0"35 rnillim. x 0"02 millim.
The spicules of the polyps are very minute and are pale yellow or colourless,
0-15 millim. x 0-01 millim.
Locality : — Ceylon seas.

Spongodes microspiculata, var. ceylonensis, now

Another colony, 8 "5 centims. high and 5 centims. hroad, belongs to the same species,
but as it differs markedly in general form and colour it seems advisable to record a
distinct variety. The chief points of difference are : —

(1) It does not extend so markedly in one plane, one of the branches arising more
or less at right angles to the £>lane of expansion ;

(2) The general contour is more rugged ;

(3) The colour approaches purple, even on the main branches, owing to the presence
of faintly purplish spicules ;

(4) The spicules are arranged on the main stem and branches more transversely ;

(5) The spicules are long, slender, warty spindles, with a light purple tint in the
older parts, becoming darker on the smaller branches which bear the polyps. Those
of the polyps are transparent and colourless. Measurements : — On the main stem,
2*2 millims. x 0"2 millim. and 1*2 millims. x 0-2 millim. ; on the smaller branches,
1*4 millims. x O'l millim., 1"5 millims. x 0*1 millim. ; around the polyps, 0'6 millim.
x 0-01 millim.

The architecture of the polyps is essentially the same as in the former specimen, so
that this form must be classed in the species 7nicrospiculata, though the characters
enumerated above justify its position as a new variety.

Locality : — Ceylon seas.

Family: SIPHONOGORGIID^.
Chironephthya indica, n. sp. — Plate, figs. 1 and 14.

A specimen of a reddish-brown colour, consisting of two small branches, about
4 centims. in height and 4 millims. in thickness.

The polyps occur on all sides and a frequent interval is 1 millim. Each is about
0-5 millim. in diameter. The anthocodiee are almost without exception nearly flush
with the general surface of the branch.

The opercular covering consists of eight triangular portions or " points " converging
over the tentacles, and at the bases of the triangles there are numerous (5) rows of
horizontal spicules forming a circlet or " crown." Each triangular "point" consists of
three diverging pairs of spicules arranged en chevron, the two outermost enclosing the

APPENDIX TO ALCYONAPJA. 171

others. The spicules of the " poiuts " and of the "crown" are reddish -brown to
orange-red in colour.

The whole ccenenchyma is covered with large rough spindles, mostly curved in a
slightly S-shaped manner, some reddish-brown and others very light in colour.
Similar internal spicules constitute the rigid substance of the branch around the
longitudinal canals, of which four were seen in a cross-section. The knobbed tubercles
often present a spiral arrangement.

The following measurements were taken in millimetres : — (a) Long stout spindles,
2-3 x 0-2, 25 x 0-175, 27 x 0'3 ; (b) slender smaller spindles, l'l x 0-03, 1'3 x 0-04,
l-5 x 0-025 ; (c) very small spindles, 0-2 x 0*025 ; (d) canal-wall spicules, uncoloured,
with relatively long spines, 0-2 x 0"01, 0-16 x 0-01, 0'14 x 002, including the spines.

It seems difficult at present to distinguish between the genera Chironephthya and
Siphonogorgia, if indeed they are not one. Hickson has suggested that the name
Chironephthya be retained for species or facies with a form and mode of branching
like Nephthya, with anthocodise rarely retracted, and with four principal spicules
en chevron in the points of the anthocodise ; and that the name Siphonogorgia be
retained for species or facies of more massive Gorgonia-like form of growth, with
anthocodiae capable of complete retraction within the general coenenchyma, and with
spicules irregularly placed or arranged in a fan-like manner in the points of the
anthocodiaj (' Alcyonaria of the Maldives,' Part I., p. 491).

If we apply these distinctions to the present specimen, we find that it agrees with
Chironephthya in having triangular opercular coverings, but disagrees in having
almost all the anthocodiae completely retracted, and in having, as far as we can judge,
a more massive mode of growth. As the minute architecture of the polyps is probably
the most distinctive feature, we have referred the specimen to Chironephthya.

Locality : — Ceylon seas.

Family : SCLEROGORGIID^.
Suberogorgia kollikeri, Wright and Studer, var. ceylonensis, nov.

Several fragments, including a basal piece, of a yellowish colony or colonies with a
sclerogorgic axis. The stem is 3 millims. in diameter at the base, and 1 millim. in
the thinnest branch of the chief specimen. The greatest length is 8'5 centime.

Verrucae may arise on all sides, but they are, for the most part, lateral. In the
smaller fragments they are altogether lateral and regularly alternate. Their diameter
is about 1 millim. Here and there the aperture shows an eight-rayed figure in the
fully retracted state of the polyp.

The thin ccenenchyma is marked on opposite sides by two shallow winding grooves.

The spicules include the following forms : — (a) Numerous warty spindles,
0'175 millim. x 0-075 millim., 0'225 millim. x 0'075 millim ; (6) a few very slender

z 2

172 CEYLON PEARL OYSTER REPORT.

spindles, O'l millim. x 0"02 millim. ; (c) almost orbicular forms, 0-l millim. x (PI millim. ;
and (d) small spindles (0-l millim. x (P03 millim.) on the aboral surface of the tentacles,
forming in retraction an opercular covering with eight points.

This form closely approaches S. kollikeri, Weight and Studer, but there the
verrucee have a diameter of 2 millims. to 3 millims., the polyps are large and
prominent, the verructe have eight-rayed margins, and the colour is yellowish-brown.
But as the differences are hardly more than quantitative, we rank the Ceylonese
specimens simply as a variety of the " Challenger " species.

Locality :— Ceylon seas.

Suberogorgia rubra, n. sp. — Plate, fig. 4.

Half-a-dozen fragments of a red colony, with a sclerogorgic very horny axis
(1'5 millims. in diameter), with a nutrient canal on each side, with thin friable
ccenenchyma, and with close-set, lateral, alternate verrucse, about 1 millim. in
diameter. The completely retractile polyps are white with a slight yellowish sheen,
and are supported by groups of small colourless spicules. It is difficult to make out
with certainty what the precise arrangement of these spicules is, but in two or three
cases eight triangular groups were seen on the polyp wall. In the fully retracted
state of the polyps the verruca appears as a rounded hillock beset with somewhat
blunt spindles.

The spicules of the ccenenchyma are warty spindles of very varied dimensions.
The following measurements were taken in millimetres : — {a) yellow spindles,
0-45 x 0-1, 0-3 x 0-1 ; (b) colourless spindles, 0-175 x (P07, (P2 x 0-08.

The colour when first examined was a bright red, but it has since become paler and
shows a tint of orange. We note this change because it is unusual in Alcyonarians.

The fragments in question most closely approach & kollikeri, Wright and Studer,
var. ceylonensis, n., but the polyps are much more crowded and decidedly smaller,
and both colour and spiculation are different. One of the fragments bore a very
minute pearl oyster.

Locality : — Ceylon seas.

Family: PRIMNOID^.

Caligorgia versluysi, n. sp. — Plate, figs. 6 and 15.

The beautiful form which we recorded (Part III., p. 289) as Primnoa ellisii, von
Koch, appeared to us to agree with the description given by yon Koch, and so up to
a certain point it does. But Dr. J. Versluys has been good enough to point out to
me that yon Koch's description is not sufficiently minute to enable one to
discriminate between Primnoa (or better Caligorgia) ellisii and other species which
have since been defined off. Dr. Versluys was kind enough to examine the

APPENDIX TO ALCYONAEIA.

173

specimen from the Ceylonese collection and another from the Indian Museum
collection (see Part III., p. 289). He regards them as representatives of two new
species, distinct from von Koch's, and closely related to a new species (C. similis)
which will be described in Versluys' memoir on the Primnoidae of the "Siboga"
Collection. In the "Siboga" material he found that the number and arrangement
of the sclerites in the polyps afforded a constant and reliable basis for specific
diagnosis.

We have therefore to withdraw the remarks we made on the geographical
distribution of Primnoa ellisii (= Caligorgia verticillata), and we have to record from
Ceylon the new species Caligorgia versluysi, the particular features of which are
discussed in the following note which Dr. Versluys has generously supplied : —

Note by Dr. Versluys.

" I find that the Primnoid collected by Professor Herdman at Ceylon is not referable
to Primnoa ellisii, von Koch. As it seems to me that von Koch was not justified in
separating his species from Caligorgia verticillata (Pallas), I shall use the older
name in comparing it with Heruman's species.*

The Ceylonese species is decidedly more delicate than C. verticillata, and the type
of ramification is different. The more important differences are stated in the following
table : —

1. Characters of Herdman's Species.

(a) The colony is dichotomously branched.

(/<) The polyps are arranged in whorls of 3 and

occasionally of 2 ; no higher number was

observed.

(') On a centimetre of the twigs about 8 whorls of

polyps are found.
(il) The length of the contracted polyps is less than

0-75 millim.
{") The distance between two successive whorls

generally varies between 0 ■ 6 millim. and

0 ' 75 millim. ; sometimes, however, it rises

up to 1 millim.

Of the 8 longitudinal rows of scales, which formed the covering of the polyps in
the typical primitive Primnoinse, viz., a pair of adaxial, inner lateral, outer lateral
and abaxial rows, only four rows, the abaxial and outer lateral pair, are well developed
in both species. The inner lateral rows are each reduced to a single large distal

* Compare Versluys, ' Primnoidae ; " Siboga " Expeditie, Monograph XIII,' second part, which is in
the press.

2. Characters of Caligorgia verticillata.

(«.) The colony is pinnately branched.

(b) On the thinnest twigs the polyps are arranged

in whorls of 3, very rarely of 2 ; on the
thicker branches the whorls mostly consist
of 4 polyps.

(c) On a centimetre there are only 5 or 6 whorls.

(d) The length of the contracted polyps is 1 millim.

to 1 • 25 millims.

(e) This distance is 1 millim. to 1"5 millims.

174 CEYLON PEARL OYSTER REPORT.

scale in Herdman's species, against two large distal scales in C. verticiUata. The
outer lateral rows are somewhat reduced in Herdman's species, as they number only
5 scales, against 7 (or 6 if the most basal scale is considered as belonging to the
coenenchyma) in C. verticiUata. The abaxial rows in this last species consist of
9 scales each, all of nearly the same rounded form, and not extending over the sides
of the polyps, which are entirely covered by the scales of the well-developed outer
lateral rows. In Herdman's species the abaxial rows are formed by 10 or 11 scales
of rather diverse form, as 2 or 3 are somewhat prolonged laterally over the sides of
the polyps between the scales of the somewhat reduced outer lateral rows. The
upper margin of the polyp scales in Herdman's species is not strongly toothed, but
these teeth are well developed and more numerous in C verticiUata. In this last
species the operculum forms a higher and, consequently, more pointed cone on the top
of the polyp. The scales in the coenenchyma are of the same type, polygonal scales
with strongly toothed irregular borders, and with the outer surface covered with
radiating and anastomosing prominences.

The species collected by Professor Herdman may also be easily distinguished from
all the previously described dichotomously branched species of Caligorgia, viz.,
C. ventilabrum, modesta and compressa* It is more delicate, with smaller polyps,
and none of these three species has so few polyps in each whorl, even on its thinnest
twigs.

There are, however, two new species in the collection made by the " Siboga "
Expedition in the Malay Archipelago, which in their habit, the dimensions of their
polyps, and the small number of polyps in each whorl very closely resemble
Herdman's species. They will be described in my paper on the Primnoidse of the
" Siboga " Expedition ; in this note I can only point out the more important differences
between these two species and Herdman's.

One of them, C minuta, is easily distinguished by the much less numerous and
proportionately much larger scales in its polyps. The abaxial rows consist of only
5 scales, of which the 4 proximal ones extend over the sides of the polyps, where they
replace the outer lateral rows, of which only one large distal scale in the upper margin
of the polyps is left.

The other new form, C. similis, is more closely allied to Herdman's species. But
while its polyps are arranged in whorls of 3, very rarely 2, on the thinnest twigs, on
the thicker branches the whorls number 4, perhaps even sometimes 5 polyps. On one
centimetre length of the twigs the same number of whorls (8) is found. But the
polyps are somewhat larger, measuring from 075 millim. to 0-8 millim. in length,
sometimes even 1 millim., against 075 millim. or less in Herdman's species, and
consequently the distance of the successive whorls is on an average somewhat less.
The most important differences, however, are found in the polyps ; they are shown in
the following table : —

* C. elegans, Gray, is insufficiently described and a doubtful species.

APPENDIX TO ALCYONARIA.

175

2. Polyps of Caligorgia similis.
(a) The abaxial rows are formed by 7 scales.

(/*) Of the outer lateral rows only one large distal

scale is left,
(r) No scale of the inner lateral row is visible when

the polyps are seen from the side.

('/) The 5 proximal abaxial scales are produced
laterally and replace the missing outer lateral
scales.

(e) No prominences are developed ; the teeth are
few and feeble.

1. Polyps of Herdman's Species.

(«) The abaxial rows are formed by 10 (or 11)

scales each.
(!>) The outer lateral rows are but little reduced

consisting of 5 scales each.
(') When the polyps are seen from the side, the

distal scale of the inner lateral row is clearly

visible.
(i/) The abaxial scales show no well-developed

extensions over the sides of the polyps,

though such an extension is clearly indicated

in some of them.
(e) The outer surface of the scales of the polyps is

covered with prominences radiating from the

nucleus and partly ending in teeth at the

upper border of the scales.

In both species the outer surface of the scales of the coenenchyma is covered with
anastomosing prominences, many of which end in a tooth at the border of the scales.
These prominences give a typical appearance to the scales of the coenenchyma ; they
are more strongly developed in Herdman's species than in C. similis. They are also
found in some pinnately branched species of Caligorgia, for instance, C. verticillata
and C. sertosa.

From this comparison it may be concluded that the Caligorgia in the collection
made by Professor Herdman is a new species, recognisable by its dichotomous
ramification, its delicate habit, the small number of polyps in each whorl, its very
small polyps, and by the comparatively large number of scales covering the polyps.

I have to thank Professor Herdman and Professor J. Arthur Thomson for their
kindness in sending me some material to study this species. In many respects it
resembles C. verticillata. The description given of this last species under the name
Primnoa ellisii, by von Koch, though excellent and valuable in many respects, does
not give many morphological details, especially in regard to the arrangement and
form of the scales in the polyps. A renewed investigation was necessarv to make
out how far it is different from Herdman's specimen. This investigation has shown
conclusively that the two are different species, but it is readily comprehensible that
Professor Thomson was led by von Koch's description to consider the species as
identical."

Locality : — Deep water off Galle.

Family : MURICEID^E.

Bebryce indica, n. sp. — Plate, fig. 3.

The collection included several specimens of a Muriceid, which we have, with some
hesitation, referred to the genus Bebryce. It agrees with this genus in its general
features, but differs considerably in its spiculation from B. mollis, von Koch,

176 CEYLON PEARL OYSTER REPORT.

B. studeri, Whitelegge, B. philippi, Studer, and B. hicksoni already described by
us (Part III., p. 294). In our account of B. hicksoni we referred to what might be
regarded as varieties of that species, but some additional specimens which we have
studied cannot so be dealt with.

All are dark-coloured irregularly branched colonies, disguised by a growth of
monaxonial siliceous sponge, and with one exception spreading in one plane. The
exception has seven alternate branches, terminally clavate, on an average 2 millims.
in diameter, and is 8 centims. in height by the same in maximum breadth. One of
the specimens attained a height of 1 8 centims.

The verrucse are prominent truncated cones, usually on all sides of the main stem
and branches, but sometimes almost restricted to the sides in the plane of branching.
They are about 1 -5 millims. in height and breadth, and are separated along one line
by intervals of 1*5 millims to 2 millims. Except in a few cases, the polyps are
completely retracted and the verrucae are thickly beset with sponge spicules.

The axis is non-calcareous, light brown in colour, very soft and flexible, and
traversed by twisted longitudinal grooves whose depth has doubtless been increased
by shrinkage of the core. The diameter of the axis is about T25 millims.

When the sponge spicules are carefully removed, the surface of the ccenenchyma
exhibits the characteristic Bebryce appearance. There is a coherent mosaic of
interlocked tuberculate discs, and large pieces can be separated off after slight heating
without any loosening of the component spicules. The average thickness of the
ccenenchyma is about 0"3 millim.

On the few polyps which could be satisfactorily seen, there were relatively large
tuberculate spindles.

The following types of spicules occur : — ■

(a) Almost regular tuberculate quadriradiate forms, with more or less pronounced
cruciate arms, and sometimes at least with an internal boss at right angles to the
expanded disc, O'l 5 millim. x O'l 25 millim. in length and breadth of the disc.

(b) Irregular quadriradiate forms, with the arms to one side longer and stronger
than those to the other side; 0-25 millim. between the tips of the longest arms,
0"15 millim. between a long and a short arm, O'l millim. between the two short arms.

(c) Approximately square forms, suggestive of amphiccelous vertebrae, with slightly
prolonged corners and slightly concave sides, 0'175 millim. x 0'175 millim. ; and
transitional forms connecting these with the pronounced quadriradiate types.

(d) Tuberculate "capstans," with a very slightly marked middle zone, separating
two equal portions; and a variety of this "double-club" type with the part to one
side of the waist much smaller than the other — a feature prominent in the spicules of
some other species of this genus ; 0'125 millim. x O'l millim. in height and breadth.

(e) Small tuberculate forms, more like " double wheels" than capstans ; O'l millim.
x O'l millim. in length and breadth, and 0'04 millim. across the " waist."

(/) Bent, warty spindles from the polyps, including (1) short, thin, rough forms

APPENDIX TO ALCYONARIA. 177

with relatively long, irregularly disposed tubercles, e.g., (P35 millim. in length by
(P075 millim. at the broadest part, and (2) longer, smoother forms with relatively
shorter, more regular tubercles, 0*5 millim. to 0-6 millim. in length by 0'05 millim. in
maximum breadth.

It is unsatisfactory that we have not been able to study the polyps of this species,
but there is no doubt that some of the spicules are flattened tuberculate discs with an
internal boss, approaching the so-called " scales" of Bebryce. If we are right in our
diagnosis, B. indica is nearest B. hicksoni, but there the "scales" were numerous and
unmistakable, and quadriradiate or cruciate forms, which are here characteristic, were
not seen.

Locality : — "West of Periya Paar, Gulf of Manaar.

Acamptogorgia gracilis, n. sp. — Plate, figs. 12 and 13.

A weathered and yet distinctive specimen, 6 centims. in height by 4 centims. in
breadth. From a short distance the branches appear pinkish red, and the verrucas
stand out like white papillae. Closer examination with the lens shows an exquisite
mingling of red and colourless spicules, the former producing a somewhat characteristic
punctate appearance.

Tbe verrucas may occur on all sides of the branches, but are for the most part
lateral in position ; they stand out almost at right angles to a height of about
1 millim. ; they are sometimes opposite, sometimes alternate, and the branch terminates
in a pair.

The ccenenchyma presents a rough surface owing to the projecting sharp points of
the continuous layer of spicules. The characteristic spicules of the ccenenchyma are
continued without marked change up the sides of the verrucas, the projecting tips
being all directed upwards. At the top of the verruca there is a ring of horizontal
spicules in two rows, and on this is based an opercular covering of eight parts, each
composed of two curved foliaceous spindles.

The axis is brown in colour, non-calcareous, 1'5 millims. in diameter at the base,
and rather less than 0'5 millim. in the delicate twigs, where it becomes much paler in
colour, almost approaching yellow. It shows at places the chambered appearance
seen in some other species of this varied genus.

The spicules include the following types : —

(a) Curved warty spindles with a bidentate or otherwise toothed foliaceous
expansion from the middle of the curve, 0'3 millim. between the tips by 0*1 millim. at
the broadest part ;

(h) Clubs with irregularly expanded divaricate ends, 0"25 millim. in extreme length
by 0-15 millim. at the broadest part ;

(c) Small irregularly stellate forms, 0*1 millim. x O'l millim. ; and

(d) Forms with four or more rays, 0-3 millim. x0'2 millim.
Locality : — Ceylon seas.

2 A

178

CEYLON PEARL OYSTER REPORT.

■-..

Acamptogorgia rubra, n. sp. — Plate, fig. 5, and text-fig. 1.

A deep crimson, incomplete colony, 5 centims. in height, giving off in one plane
three lateral branches which have an average thickness of about 1 niillim. There is

also a fragment about 2 centims. in length, with four
twigs arising almost at right angles from the main branch,
which is slightly over 1 millim. in breadth.

The verrucas, 0'5 millim. in height by 1 millim. in
breadth, are alternate or sub-opposite, and arise at right
angles to the axis. Two occur side by side at the end of
a branch. The coenenchyma is thin and presents a very
prickly appearance, due to the projecting folia and spines
of the beautiful crimson spicules. The axis is yellowish ,
0*3 millim. to 0-4 millim. in diameter.

The spicules show considerable diversity of form : —

(a) Straight, warty spindles, 0-6 millim. x 0*1 millim. ;

(b) Narrow, curved spindles, smooth terminally, warty
about the middle, 0"6 millim. x 0'05 millim.

(c) Large triradiate forms in which the shortest, almost
smooth ray projects externally and often bears a foliaceous
expansion, while the two larger rays are denticulate or
branched to a very varied degree ; 0'6 millim. in breadth

between the tips of the longest rays and 0"4 millim. in height between a line joining
these internal tips and the tip of the external ray.

(d) Smaller and often simpler triradiate forms, G"45 millim. in breadth and
0"25 millim. in height, and much smaller.

(e) Small, very warty spindles with a foliaceous expansion about the middle,
0'2 millim. x 0"1 millim.

Locality : — Ceylon seas.

Fig. 1.
Acamptogorgia rubra, n. sp. x 15

Muricella complanata, Wright and Studer.

Reference has already been made (Part III., p. 303) to varietal forms of this species,
which seem to vary considerably in detail. Another form deserves to be recorded.
It differs from the type in the coloration of the spicules, large yellowish spindles
covering smaller transparent ones, and others still smaller which are rose-coloured.
Thus the rubbed base of the specimen is bright rose, while the upper parts are
ochreous yellow — a difference which seems to be wholly due to the degree of abrasion.

The verruca? show the characteristic longitudinal grouping of small spindles in eight
triangles, but it may be noted that in some cases they occur opposite one another,
instead of alternately as in the type, and that the diameter of their base can hardly
be said to exceed a millimetre, while that in the type was 2 millims.

APPENDIX TO ALCYONARIA. 179

Muricella rubra, n. sp. — Plate, figs. 2 and 7.

A minute fragment of a bright-red colony with a relatively thin coenenchyma filled
with long warty spindles.

The verruca? are crowded and alternate, low and subcorneal, arising at right angles
to the surface, and covered with spindles smaller than those of the ccenenchyma. The
polyps are white and stand out conspicuously. The axis is light yellow, non-
calcareous, flexible, 0"5 millim. in diameter.

The general spicules of the ccenenchyma are (a) large warty spindles, straight or
slightly S-shaped, l-2 x 0'2, 0"95 x (P175 millims. ; and (b) small warty spindles, 0'5 x (VI,
0'4x0"075 millim. In the polyps the spicules are transparent, slightly warty
spindles, 0-25x0'04, 0"2x0-03 millim.

Locality : — Ceylon seas.

Family: V1RGULAMIILE.
Virgularia loveni, Kolliker.

We have already recorded the occurrence of representatives of V. loveni in this
collection, but another specimen has come to hand which deserves notice, especially
as that on which Kolliker based his species was very imperfect.

This is also an incomplete specimen, but very well preserved, and presenting
several features which make us hesitate in referring it to V. loveni unless Kolliker's
diagnosis be somewhat modified. It is about 90 millims. in length, 5 millims. to
9 millims. in breadth, and is altogether pinnule-bearing, except 15 millims. at the
broken basal part. The lower pinnules are closely crowded without visible intervals ;
those on the upper portion are separated on the prorachidial surface by intervals of
1 millim. to 1'5 millims., and overlap on the metarachidial surface. There is a bare
tract, 2'5 millims. in breadth, along the prorachidial surface, and up the middle of
this there runs a well-marked narrow groove.

The free margin of the pinnule curves like an elongated S, and its prorachidial
insertion is higher than the other. Most of the pinnules are turned markedly
downwards. There were 30 polyps on the pinnules counted, and a single row of
zooids extends transversely in the interspace between the pinnules. The polyps are
very distinct, but the indentations between the calices are shallow.

The axis, which measures 2"25 millims. in diameter at the base, is cylindrical,
whitish, and covered with remarkable indentations and irregular ridges.

This form agrees with V. loveni in having up to 30 polyps on the pinnules, an
indented ridged axis, the zooids in a single row, and so on ; but in Kolliker's
specimen the pinnules were 3 millims. to 4 millims. apart, the shape was approxi-
mately fan-like, and the calices were scarcely distinct.

Locality : — Ceylon seas.

2 A 2

180 CEYLON PEARL OYSTER REPORT.

Virgularia elegans, Gray.

To this imperfectly described species we refer an imperfect specimen, and if the
reference be correct we can make Gray's diagnosis a little more definite.

The rachis bears 106 pairs of pinnules; it is 145 millims. in length, including
1 5 millims. of exposed axis at the broken basal end ; its breadth is 2 millims. to
3 millims., and that of the axis 1'25 millims.

The pinnules have a breadth and height of 2-75 millims. x 0'5 millim., 275 millims.
x 2 millims., 2 2 5 millims. x 075 millim., at the base, middle portion, and top of the
rachis. The interval between them is 0'25 millim. at the base and 2 millims. at the
top. They bear 18 polyps in a single row.

Locality : — Ceylon seas.

Virgularia calycina, n. sp. — Plate, figs. 8, 9, 10, 11.

An incomplete specimen, 57 millims. in length by 3 millims. to 4 millims. in
breadth, altogether pinnule-bearing, except 5 millims. of bare axis at the broken
basal end.

The lower pinnules are closely crowded, with no intervals between them, but in the
upper portion of the rachis they become distant, being separated by intervals of
175 millims. There is a bare streak, about 075 millim. in breadth, along the
prorachidial surface.

The pinnules are substantial and distinct, though relatively narrow, and their
prorachidial insertion is very markedly higher than the other. The curve of their
free margin is a crescent. On each of those counted there were 18 very distinct
calices, with conspicuous longitudinal grooves. When the polyps are retracted, the
calices are ovoid in shape, about 1 millim. in height, with a minute circular aperture.
The polyps appear to be arranged, except the first three or four at the prorachidial
side, in two alternating rows, but this is simply due to the fact that they are
alternately shunted to opposite sides along the margin of the pinnule. The inferior
surface of the pinnule is strongly marked by eighteen parallel ridges, each corre-
sponding to a calyx or polyp. On the upper surface there are similar markings, but
less pronounced. Some of the polyps contain ova. Here and there a few minute
lateral zooids were seen in a row about halfway between two pinnules.

The axis is white and cylindrical, with somewhat delicate and complex markings.

This form should be ranked among those species of Virgularia which Kolliker
described as having distinct pinnules, distinct calices, zooids in one row or in two
rows, with the ventral insertion higher than the dorsal, with non-transparent,
crescent-shaped pinnules, — that is to the artificial section including V. mirabilis
(0. F. Muller) (with 6 to 9 polyps), V. multifiora, Kner (with 11 to 15 polyps),
and some uncertain forms. But from V. multifiora this new form is at once
distinguishable by the markedly separate calices, and also by their number, &c, In
V. juncea, again, the pinnules are very slight, and the calices are not separate ; in

APPENDIX TO ALCYONARIA. 181

V. reinwardtii there are 18 polyps, but the calices are not separate. The present
specimen appears to be nearest V. elegans, Gray, but in that species the pinnules
are much less distinct, the number of calices is 14 to 24, and the calices are "scarcely
separate at the margin."
Locality :— Ceylon seas.

Virgularia indica, n. sp.

Several incomplete specimens, about 12 centims. in length by 5 millims. to 6 millims.
in breadth, bearing over 120 pairs of pinnules. About a fourth of these are well
developed and are separated by intervals of 075 millim. On the upper region of the
prorachidial surface there is a bare streak 2 millims. in breadth, which becomes a
narrow line in the lower half of the colony. On the lower half of the metarachidial
surface there is a deep channel left up the middle between the two rows of pinnules,
but higher up this is lost, since the opposite pinnules meet or regularly overlap. There
were 15 polyps on all the pinnules counted and the siphonozooids occur in equal
number in a row midway between two pinnules or near the base of the upper pinnule.
The ovoid calices are distinctly marked off from one another and the translucent
pinnule shows grooves and ridges corresponding in number to the polyps. Several
ova were seen in the pinnules.

The axis is cylindrical, about 1 millim. in diameter, and covered by a reddish-brown
investment. As regards the number of polyps on a pinnule, this form resembles
V. multiflora, Kner, which Jungersen regards as merely a variety of V. mirabilis
(O. F. Muller). But the pinnules are much closer together, they often show an
S-shaped curve, and they are translucent ; the calices are distinctly separated from
one another ; the axis is flexible and covered with a reddish investment.

It therefore seems necessary to establish a new species.

Locality : — Modragam Paar, Gulf of Manaar.

It may seem remarkable that Professor Herdman should have found in a short time
within a limited area no fewer than six species of Virgularia. This is the more
extraordinary since the rich collection of deep-water Alcyonarians made by the
'Investigator" in the Indian Ocean does not include a single representative of the
genus.

It seems extraordinary, almost suspicious, that the handful of Vivgulavice before us
should include three new species, in spite of our endeavours to unite these with others
previously described. We would therefore note that : —

(1) V. tuberculata is conspicuously characterised by its red tuberculated axis and
by the six, relatively large, very distinct, barrel-shaped polyps on each small pinnule ;

(2) V. calycina is conspicuously characterised by its 18 very distinct ovoid calices
on each prominently ridged pinnule ;

(3) V. indica approaches V. multiflora, but differs in having the pinnules much

182 CEYLON PEARL OYSTER REPORT.

closer, often S-shaped, translucent, with distinctly separated calices, and in other
characters.

Family : CAVERNULARIID^E.

Fusticularia herdmani, Simpson — Plate, figs. 16 to 22, and text-fig. 2.

I entrusted to Mr. J. J. Simpson, M.A., my private assistant, a small club-like
specimen which had been overlooked in the first study of Professor Herdman's
collection. It had, indeed, so much resemblance to a corticate sponge that it was

originally sent for examination to Professor Dendy. The
accompanying text-figure (fig. 2) is a reproduction of a
drawing of the colony made by Professor Dendy.

Mr. Simpson has published an account of this interesting
form in the ' Annals and Magazine of Natural History,' xv.
(1905), pp. 561-5, 1 plate, and has named it appropriately
Fusticularia herdmani, gen. et sp. n.

The specimen is a small free-living sponge-like colony,
37 centime, in length, 1*7 centims. in breadth, and 1 centim.
thick. It consists of a flattened ovoid stock separated by a
constriction from a comparatively slender trunk 1 "2 centims.
long and 0"6 centim. in breadth.

The general colour is a dark brown, approaching
chocolate.

The zooids are dimorphic, the smaller siphonozooids being
scattered irregularly among the larger autozooids, which
Pig. 2. Fusticuiaria herdmani, ai'e separated by distances varying from 1 millim. to
.Simpson, x 2. 3 millims. The zooids are completely retractile into pit-

like depressions, about 0*5 millim. in diameter. The length
of a fully expanded autozooid is about 0'75 millim. to 1 millim. and the tentacles
measure 07 millim.

Three longitudinal canals traverse the colony throughout its entire length.
The coenenchyma is densely spiculose. The spicules, which vary greatly in form
and size in the different parts of the colony, are arranged in bundles supporting the
polyp cavities. All are hyaline and smooth and the majority bear blunt digitiform
terminations which are often marked by characteristic annulations. The most
frequent types are the following : blunt spindles, cylinders, clubs, double barrels and
palmate forms. They vary in size from 0-3x0-05 millim. in the cortical layer of the
stock to 0-45x0-025 millim. in the trunk. Among the characteristic features of this
genus the following are most noteworthy : (a) the minuteness of the zooids ; (b) the
absence of an axis ; (c) the broadly palmate spicules ; (d) the elliptical trunk ; (e) the
constriction between stock and trunk; (f) the number (3) of canals in the stock ;
(g) the small number of autozooids.

APPENDIX TO ALCYONAMA. 183

The diagnosis reads : — " A somewhat sponge-like cavernnlarid, with a flattened
ovoid stock separated by a constriction from a comparatively slender sterile trunk;
with dimorphic retractile polyps, the antozooids not exceeding 1 niilliin. in length, the
much smaller siphonozooids scattered irregularly among the autozooids ; with
abundant densely spiculate ccenenchyma, traversed by three longitudinal central
canals passing down into the trunk ; with smooth hyaline spicules bearing peculiar
digit iform terminal processes and showing very characteristic annulations, especially
near the ends."

Locality : — Cheval Paar, Gulf of Manaar, 6 fathoms.

Note. — Further inquiry into the history of the specimen listed by Professor
Herdman and Mr. Hornell in their diaries as a possible Corallium sp. has shown
that this tentative identification was mistaken. We therefore withdraw the para-
graph referring to this specimen on p. 289 of Part III. of this Report. — J. A. T.

184

CEYLON PEARL OYSTER REPORT.

Complete List of the Species of Alcyonaria reported on : —

Order I. : Stolonifera.
Family : Clavulariidte.
*Clavularia margaritifero?, n. sp.

Order II. : Alcyonacea.

Family : Xeniida;.

Xenia ternatana, Schenck.
„ wmbellata, Sav.

Family: Alcyoniidse.
*BeUoneUa indica, n. sp.

Family : Nephthyidae.

Nephthya chabrolii, Aud., var. ceylonensis, n.
„ lobuNfera, Holm.

* „ ceylonensis, n. sp.
*Eunephthya purpurea, n. sp.
* Paraspongodes striata, n. sp.
*Capnella manaarensis, n. sp.

Spongodes bicolor, Wright and Studer.

„ „ „ „ var. ceylon-
ensis, n.

,, ,, „ „ \ar. dubia, n.

„ rosea, Kukenthal.

,, armata, Holm., var. ceylonensis, n.

„ dendrophyta, Wright and Studer.

„ splendens, Kukexthal.

,, cerviconiis, Wright and Studer.

,, involuta, Kukenthal.

„ piitteri, Kukexthal.

,, microspicnlata, Putter.

„ „ var. ceylonensis, n.

* „ pulckra, n. sp.

* ,, awantiaca, n. sp.

Family : Siphonogorgiidse.

*Paraniphthya pratti, n. sp.
Ckironephthya variabilis, Hickson.

* „ iiulica, n. sp.

Siphonogorgia pustulosa, Wright and Studer.
„ viiniacea, Kukenthal.

„ lullikeri, Wright and Studer.

Order III. : Pseudaxonia.
Family : Briareidse.

Solenocaidon torluosum, Gray.

Family : Sclerogorgiidse.

Keroeides gracilis, Whitelegge.
Suberogorgia verriadata, Esper.

„ kbllikeri, Wright 'and Studer, var.

ceylonensis, n.

* ,, rubra, n. sp.

Order IV. : Axifera.
Family : Primnoidse.
*C'aligorgia versluysi, n. sp.

Family : Muriceidse.

Acanthogorgia miiricata, Verrill, var. ceylonensis, n.

* ,, media, n. sp.

* ,, ceylonensis, n. sp.
*Aslromuricea ramosa, n. sp.

Echinomuricea indo-malaccensis, Ridley.

* ,, ceylonensis, n. sp.
Echinogorgia pseudosasappo, Kolliker.

„ multispinosa, n. sp.

* Heterogorgia verrilli, n. sp.
*Bebryce hicksoni, n. sp.

* „ indica, n. sp.
Acamptogorgia spinosa, Hiles.

„ ,, „ var. ceylonensis, n.

* „ atra, n. sp.

,, gracilis, n. sp.

* ,, rubra, n. sp.
Acis orienlalis, Ridley.

* „ indica, n. sp.

* ,, alba, n. sp.

* „ ceylonensis, n. sp.

„ ,, ,, var. imbricata.

Muricella nitkla, Verrill.

„ cmnplanata, Wright and Studer.

* ,, ramosa, n. sp.

* „ ceylonensis, n. sp.

* „ rubra, n. sp.

APPENDIX TO ALCYONARIA.

185

Family: Plexauridae.

Plemwra prcelonga, var. typica (Ridley).
„ „ ,, elongata, n.

„ antipathes, Klunzinger, var. flemosa, n.

Family : Gorgoniid;r.

Lopkogorgia lidkeni, Wright and Studer.
„ rubrotincta, u. sp.

„ irregularis, n. sp.

Leptogergiaj australiensis, Ridley, var. flavotincta.

„ perflam.
(?) sp.
*Stenogorgia ceylonensis, n. sp.
Gorgonia capensis, Hickson.
Rhipidogorgia sp.

Family : Gorgonellidse.

* Scirpearella aurantiaca, n. sp.

* „ divisa, n. sp.
„ sp. a.

sp. /3.
sp. y.
Srirjit'dria sp.

Juncella gemmacea, VALENCIENNES.
,, juncea, Pallas.
,, fragilis, Ridley.
„ ,, „ var. rubra, n.

* ,, trilineata, n. sp.

Verruccllafli citosa, Kluxzinger, var. aurantiaca, n.
,, ,, ,, „ gallensis, n.

* „ rubra, n. sp.

Order V. : Stelechotokea

Section I. : Asiphonacea.

Family: Telestidse.

Telcsto rubra, Hickson.
„ (Oarijoa) trichostemma, Wright and Studer.

Section II. : Pennatulacea.
Family : Umbellulidfe.
Umbdlula sp.

Family : Virgulariida3.

Virgularia multiflora, Kner.
„ loveni, Kolliker.
„ elegans, Gray.

* „ tuberculoid, n. sp.J

* ,, cahjcina, n. sp.

* „ inclica, n. sp.

Family : Pennatulidas.

Halisccptrwn gustavianum, Herklots.

„ periycnse, n. sp.

Pteroeides lacazti, var. spinosum, Kolliker.

Family : Veretillidae.

Cavernularia obesa, Valenciennes.
*Stylobelcmnoides herdmani, n. gen. et sp.
*Fusticularia herdmani, n. gen. et sp., Simpson.

* There are thus forty-two new species described in this report.

t Ridley has changed this title to Gorgonia austra.lie.nsis, ' Journ. Linn. Soc, Zool.,' xxi. (1888), p. 238.
I Non Virgularia hdierculata,, Marshall, 1883 (= Virgularia cladiscus, Jungersen, 1904).

2 B

186 CEYLON PEARL OYSTER REPORT.

EXPLANATION OF PLATE.

Fig. 1. Chironephthya indica, n. sp. ; two groups of anthocodial "point" spicules.
,, 2. Mur'n-i I la r ultra, n. sp. ; showing texture of the surface, x 2.
,, 3. Bebryce indica, n. sp. ; spicules.

„ 4. Suberogorgia rubra, n. sp. ; showing texture and verrucas, x 10.
,, 5. Acamptogorgia rubra, n. sp. ; showing branching and the alternate verrucas, x 2.
„ G. Caligorgia versluysi, n. sp.
„ 7. Muricclla rubra, n. sp. ; spicule.
,, 8. Virgularia cahjcina, n. sp. ; the entire specimen.
,, 9, 10, 11. Virgularia cahjcina, n. sp. ; pinnules at various levels.
„ ] 2. Acamptogorgia gracilis, n. sp. ; showing spinose texture of the ccenenehyma, and the eight

triangles of anthocodial spicules, x 10.
,, 13. „ „ types of spicules.

„ 14. Chironephthya indica, n. sp. ; a large curved spindle with tubercles.
„ 15. Caligorgia versluyi, n. sp. Drawn by Dr. Versluys. op., operculum; o., abaxial row; l.} outer

lateral row; i., inner lateral row; sc, basal scale of abaxial row, or belonging to general

ccenenehyma.
,, 16. Acamptogorgia rubra, n sp. ; types of spicules.
,, 17. Ftisticularia herdmani ; spicules of the body parenchyma.

„ 18. „ transverse section of the stock, showing three central canals.

,, 19. ,, portion of the surface enlarged, showing different stages of retraction of

autozooids.
,, 20. ,, entire colony. Natural size.

,,21. ,, spicules of the cortical layer.

,, 22. ,, spicules of the trunk.

CEYLON PEARL OYSTER REPORT

ALCYONARIA SUPPLEMENTARY PLATE.

'

m

%*s&

;fij

W

vV

f.;-.

13

1

19.

20

21

E.Wils.

*

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

REPORT

ON THE

SOLITARY CORALS

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY

GILBERT C. BOURNE, M.A., D.Sc, F.L.S.,

FELLOW AND TUTOR OF NEW COLLEGE, OXFORD.

[With FOUR PLATES and FIVE FIGURES in the Text.]

The following paper contains an account of a small but remarkably interesting
collection of Solitary Corals from Ceylon, which Professor Herdman has kindly
entrusted to me for identification and description. The collection includes five
species of Turbinolidse, of which two are apparently new to science, three species of
Flabellidae, six species of Fungiidse and seven species of Eupsammiidse. There are, in
addition, some specimens of Cyathohelia, sp. incert., and other colonial corals, but as it
is my intention to confine myself to a description of the solitary corals and to species
closely related to them, these compound forms will not be included in the present
memoir.

As every student of corals knows, the classification and identification of the
Turbinolidas and Eupsammiidse is attended with no inconsiderable difficulties. The
greater number of the genera and species described by Milne-Edwards and Haime
(35 and 36) were fossil and the type specimens have been inaccessible to me. The
researches of the last forty years have added a considerable number of recent, as well
as fossil species, some of which are undoubtedly identical with forms described by
older authors ; others are undoubtedly new, but many must be regarded as doubtful,
for there is ample evidence that varieties have frequently been classed as sjjecies, and
the identity of recently discovered living genera and species with previously described

2 B 2

188 CEYLON PEARL OYSTER REPORT.

fossil species has in several cases been overlooked. Hence one cannot deal with any
collection of Turbinolida3 and Eupsammiidse, however small, without entering upon a
criticism of these families, and, unfortunately, my claims to undertake such a
criticism are limited. A short study of the literature of the subject is sufficient to
convince one that no sufficient degree of exactitude can be attained without a study
of the original specimens on which the genera and species were founded. But these
are largely contained in foreign museums, which I have been unable to visit, or were
included in private collections which have been dispersed and are no longer traceable.
The British Museum is rich in the possession of the " Challenger" collections and in
duplicates of species collected by Pourtales, but it is singularly poor in Eupsammiidae,
and a careful search among the named specimens of this family failed to bring to light
anything that was of assistance in disentangling the difficulties attendant on its
classification. I have, therefore, had to rely chiefly on published descriptions and
illustrations, and these are, in many cases, too vague and inaccurate to be of material
assistance.

I wished, when I undertook this piece of work at Professor Herdman's recpiest, to
make a thorough examination of the anatomy of the soft parts as well as of the
coralla of the specimens he entrusted to me, but this I found was impossible, as
several of the more important species were represented only by a single specimen and
it was necessary to dissolve away the soft parts in order to identify the coralla.
T have, however, made a tolerably thorough investigation of the anatomy of three
forms, Heterocyathus cpquicostatus, Heteropsammia rnichelinii, and Dendrophyllia
gracilis. As a matter of convenience this paper will be divided into two sections, the
first dealing with the systematic description of the coralla contained in Professor
Herdman's collection, the second with the details of the anatomy of the three above-
mentioned species.

PART I.-SYSTEMATIC.

Family : TUEBINOLID^E (pars.), M. Edwards and Haime*

M. Edwards and Haime divided the Turbinolidae into the two sub-families
Caryophyllince. with one or more crowns of pali, and Turbinolinae, without pali. The
Caryophyllinse were further subdivided into Caryophylliacese, with a single circle of
pali, and Trochocyathacese, with more than one circle of pali. The Turbinolinse were
subdivided into the Turbinoliacea?, destitute of epitheca, and Flabellacefe, in which
" the wall is completely covered by a pellicular epitheca." Martin Duncan, in his
revision of the families and genera of the Madreporaria (13), abolished the sub-families

* 'Ann. des Sci. Nat.,' 3e ser., t. ix., 1848, and 'Hist. Nat. des Corall.,' t. ii., p. 7, 1857;
P. M. Duncan, "Revision of the Madreporaria," 'Journ. Linn. Soc.,' vol. xviii., "Zoology," 1885.

SOLITARY CORALS. 189

of M. Edwards and Haime depending on the presence or absence of pali, and created
in their stead three sub-families Turbinolidsa simplices, Turbinolidas gemmantes and
Turbinolidae reptantes. The first of these includes nearly the whole of the Caryo-
phyllinae and Turbinolinoe of M. Edwards and Haime and is divided into seven
alliances, viz.,Smilotrochoida, Flabelloida, Placotrochoida,Turbinoloida, Trochocyathoida,
Discocyathoida and Haplophylloida. Want of space and want of material forbid my
entering upon a criticism of Duncan's classification, and I must content myself with
remarking that it has been of no great assistance to systematists, and that the older
classification of M. Edwards and Haime has been generally preferred to it by recent
writers. Both classifications agree in placing the genus Flabelhmi and its allies as a
subordinate group of the Turbinolkke, and the most recent writer on the genus
Flabelhnu, Mr. J. Stanley Gardiner (22), adopts the classification of Duncan,
giving, however, a definition of the "Flabelloida," which I cannot find anywhere in
Duncan's paper. As I shall have to enter more fully into the structure and
classification of Flabelhmi and its allies in a subsequent part of this paper, I need
only say here that I am of the opinion that they differ from all other Turbinolidse in
several important characters, but chiefly in the fact that their wall is a persistent
prototheca (Bernard, 4), which is not thickened externally by a secondary deposit
of calcareous substance laid down by an "edge zone" or " perisarc," this latter
structure being, indeed, absent. They must, therefore, be classified apart in a family
Flabellidre, and the Turbinolidae, after the removal of the Flabellidae, may be defined
as follows : —

Corallum simple, or forming colonies by gemmation from the wall or from
stoloniform basal outgrowths. The wall is thickened externally by the secondary
deposit of calcareous tissue formed by the edge-zone, and is solid, variously
ornamented with costas, granules and spines, the spaces between the costas sometimes
filled up by the deposit. The upper part of the wall commonly formed by the union
of the enlarged outer ends of the septa. Epitheca, when present, pellicular or
lamellar, closely adherent. The septa commonly exsert and solid. Septal loculi open
to the base. Pali may or may not be present. Columella present or absent ; when
present it may be essential or parietal and of very various shape and composition.

Paracyathus, M. Edwards and Haime.*

This genus is easily recognised by its numerous lobate pali, scarcely distinguishable
from the prominences of the columella. The pali, however, should more correctly be
described as paliform lobes, as they are clearly thickenings of the inner ends of the
septa, and the outermost of the several crowns, at any rate, do not extend to the
base. The original definition of the genus given by M. Edwards and Haime, in

* 'Ann. des Sci. Nat. Zool.,' 3e s<5r., t. ix., p. 318, 1848; 'Hist. Nat. des Cor.,' t. ii., p. 52, 1857;
1'. M. Duncan, " Rev. Madrep.," 'Journ. Linn. Soc.,' vol. xviii., "Zool.," p. 24, 1885.

190 CEYLON PEARL OYSTER REPORT.

1848, was amended in the ' Hist. Nat. des Coralliaires ' ; in the former the columella
was described as " tres developpe"e," the pali as " tres dlevds," but both these
characters are exceedingly variable, and in the ' Hist. Nat. des Coralliaires ' the
definition of the structures in question runs thus : "La columelle est formde des
tigelles qui paraissent naitre du bord interne et inferieur des cloisons, et qui sont
d'autant plus elevees qu'elles sont plus dxtdrieures ; sa surface est papilleuse et
concave. Les palis paraissent se detacher egalement de la partie infdrieure des
cloisons, et se distinguent a peine des tigelles columellaires." The species of the
genus are not easy to determine with certainty. Lacaze Duthiers (33) gives
excellent photographs of P. striatus, and Duncan (11) gives several species from
different parts of the world, but all these are obviously different from the specimens
in Professor Herdman's collection. It should be observed that the initial and
therefore solitary corallites of the arborescent Cyathohelia closely resemble Para-
cyathus, the septal pali and columella being almost identical, and I am inclined to
think that several of the described species of Paracyathus are nothing more than
young Cyathohelise.

Paracyathus stokesi, M. Edw. and H.

There are three specimens of Paracyathus in Professor Herdman's collection ;
they were taken at different times from different localities and differ from one
another in details, but their resemblance is sufficiently close to lead me to refer them
all to P. stokesi, M. Edw. and H. (35, plate x., figs. 7 and 7a).

Specimen a is from Trincomalee. The corallum is 17 millims. in height ; the calice
elliptical, its longer axis measuring 13 millims., its shorter axis 8 millims. The calice
is depressed at either end of the long axis. The septa correspond very closely with
M. Edwards and Haime's description of P. stokesi, but the fifth cycle is incomplete,
and the outer ends of the primary and secondary septa are scarcely, if at all,
thickened. The costae differ from those of P. stokesi, in being so slightly developed
that they are scarcely distinguishable at a short distance below the level of the
calice.

Specimen b is from Galle. Height of corallum 14 millims. ; longer axis of the
calice 12 millims. ; shorter axis 9 millims. Septal arrangement as in specimen a, but
the septa are rather thicker externally. The calicular fossa is deeper, the columella
less develojjed, and the septal pali stouter and more prominent than in a. The costse
agree exactly with M. Edwards and Haime's description of P. stokesi; they are
rather broad, distinct to the base, and thickly covered with granules.

Specimen c is from deep water near Galle. Height of corallum 1 4 millims. ; longer
axis of calice 7 millims. ; shorter axis 6 millims. This specimen resembles P. pro-
cumbens, M. Edw. and H, from the Eocene of Hauteville, in the following characters.
The corallum is curved and the calice inclined to one side ; the calice is subcircular
and rather deep ; the septa are somewhat exsert and their external ends project

SOLITARY CORALS. 191

beyond the lip of the calice ; the costae are somewhat slender and form sharp ridges
projecting unequally at different levels, but never prominently. On the other hand,
it resembles P. caryophyllus, from the Eocene of Sheppey, in having only four cycles
of septa. (It seems very probable that these two Eocene species are identical, the
presence or absence of a fifth cycle of septa not being sufficient to distinguish them.)

I have no hesitation in referring a and b to P. stokesi, but I am more uncertain
about c. If it cannot be referred to this species, it cannot be anything else than
P. procumbens; but a comparison of M. Edwards and Haime's figm-es (35, plate x.,
figs. 6 and 7) lead me to believe that the latter species is only a variety of P. stokesi,
and that the three Ceylonese specimens are local varieties of that species.

Paracyathus striatus, Philippi.

Cyathina striata, Philippi, ' Arch, fur Natur.,' 1842, vol. 1, p. 48.

Paracyathus striatus, M. Edwards and Haime, 'Pol. foss. des terr. palseoz.,' 1851, p. 25.

A single specimen, from deep water off Galle, which only came into my hands after
this paper was written, must clearly be referred to this species.

Rhodocyathus, n. gen.

Corallum simple, free, saucer-shaped, with signs of former adherence. Calice sub-
circular, wide, shallow. Columella essential, well developed, composed of numerous
oblique thin overlapping lamellae whose upper edges are produced into numerous
flattened spines. Septa in six systems and five cycles, the last cycle incomplete.
The primary and secondary septa subequal, very exsert, arched at their outer ends
and sloping inwards to join the columella ; their upper margins furnished with blunt
spines ; their outer margins produced into sharp and prominent spines where they
pass into the costse ; the surfaces of the septa marked with broad radiating ridges
corresponding to the marginal spines, the whole surface finely granular. Tertiary
septa similar to the primaries and secondaries, but smaller and less exsert. The
quaternary much smaller than the tertiary septa, and bending inwards to join the
latter near the columella. The quinary septa mostly very small and free at their
inner ends ; where they are longer they generally become attached to an adjacent
septum, usually to a quaternary, but in some cases to a primary, secondary, or
tertiary; the inner ends of the quinary and quaternary septa are very thin and
cribriform. The costse corresponding to the first four cycles of septa form distinct
fairly prominent ridges covered with fine granulations ; those corresponding to the
fifth cycle very small ; the costae of opposite ends meet below and cover in the basal
scar of attachment.

Rhodocyathus ceylonensis, n. sp. — Plate I., figs. 1 and 1a.
The characters are those of the genus. A single specimen from Trincomalee.
The dry corallum is of a yellowish colour, and measures about 18 millims. in height

L92 CEYLON PEARL OYSTER REPORT.

from the rounded base to the top of a primary septum. The calice is subcircular,
measuring 35 millims. in its longer and 33 millims. in its shorter diameter. The
columella is oval, measuring 10 millims. X 6 millims., and has a horizontal surface.
The specimen was evidently fixed in early life and has become detached, the scar of
detachment being healed over in a peculiar manner by the continuation of sixteen
costas round the bottom of the corallum, to meet and fuse with their fellows of the
opposite side. These costse belong to the septal systems at either end of the longer
axis of the calice ; the costse of the lateral systems converge below and unite to form
two somewhat irregular and not very conspicuous lateral basal prominences. The
specimen was completely covered by soft tissues, but these had to be removed to
facilitate the study of the corallum.

I have been unable to refer this specimen to any known genus, and have therefore
made the new genus Rhodocyathus to contain it. The name is descriptive of the
shape of the calice, which bears a resemblance to a conventional rose. It differs from
all the members of Duncan's Turbinoloida in the characters of the columella. It is
distinguished from the majority of the Trochocyathoid alliance by the absence of pali,
and its shape and the character of its septa and costse mark it off from the somewhat
vague and all-embracing genus Ceratotrochus, as defined by Duncan. It cannot be
placed in any genus of the Discocyathoid alliance, nor yet among the Haplophylloida.
The presence of a columella separates it from all the Smilotrochoida ; the character
of the columella prevents its being placed among the Placotrochoida ; and it certainly
is not a Flabelloid. The genus shows most affinity to Ceratotrochus and Delto-
cyathus, and may be described as a Trochocyathoid without pali.

Cyathotrochus, n. gen.

Corallum simple, free, without a trace of adherence, forming a short, laterally
compressed cone. Costse moderately prominent near the lip of the calice, but scarcely
distinguishable below and indistinguishable near the bluntly pointed base ; the
primary and secondary costse somewhat more prominent than the tertiaries, which in
turn are rather more prominent than the quaternaries ; the costse corresponding to
the primary septum at each end of the longer axis of the calice much more prominent
than the rest and continued as sharp ridges about half-way down the corallum, but
not forming aliform expansions. Calice elliptical, subplane. Columella essential,
with a papillary upper surface, projecting slightly in the calice. Septa in six systems
and four regular cycles ; the primary and secondary septa subequal, moderately
exsert, arched above, with nearly vertical inner margins, their surfaces ornamented
with rows of relatively large granules, arranged nearly parallel to the inner margins.
Paliform lobes, separated from the septa by a deep but narrow notch, stand in front
of the first three cycles of septa and connect them with the columella ; those in
front of the tertiaries are the largest, but they do not form chevrons or deltas.

SOLITARY CORALS. 193

Cyathotrochus herdmani, n. sp. — Plate L, figs. 2 and 2a.

Height of corallum, 7 millims. ; longer axis of calice, 10 millims. ; shorter axis,
7 5 millims. The characters are those of the genus. The septa and the bluntly
pointed basal part of the corallum are white, the rest of the wall of a dull Indian-red
colour. The wall and costee are covered with close-set but very fine granules, which
can only be distinguished by the aid of a strong lens. A single specimen from the
west of Periya Paar.

After some hesitation I have founded a new genus to receive this species, which
does not correspond exactly with any described genus of the Turbinolidse. It is
undoubtedly closely allied to Troehocyathus, from which it differs only in the form of
the pali, which should perhaps be described as paliform lobes rather than true pali,
us they are evidently the thickened inner continuations of the septa, separated from
the latter by a notch which, although fairly deep, does not extend to the bottom of
the septum. These pali, or paliform lobes, form a single crown, not two crowns as
in Troehocyathus, and they do not form chevrons as in Trojndocyathus. The
characteristic feature of the species is the keel-like projection of the costse at each
end of the long axis of the corallum, suggesting an affinity to Tropidocyathus
(Troehocyathus) lessoni. But in this latter species, as Michelins' (34) and
Alcock's (3) figures show very clearly, the corresponding costa? are very much
enlarged and form aliform expansions extending round the base of the corallum,
whereas in Cyathotrochus they are but slightly enlarged to form ridges extending
scarcely half way to the base.

Heterocyathus, M. Edwards and Haime (35).

Stephanoseris, M. Edwards and Haime (38).

Heterocyathus aequicostatus, M. Edw. and H.

Stephanoseris rousseaui, M. Edw. and H., 'Hist. Nat. des Cor.,' t. iii., p. 56, 1860.

H. philippinensis, Semper, 'Zeit. Wiss. Zool.,' xxii., p. 254, 1872.

H. parasiticus, Semper, loc. cit.

H. pulchellus, Rehberg, ' Abh. Ver. Hamb.,' xii., p. 8, 1892.

H. oblongatus, Rehberg, loc. cit.

H. aequicostatus, J. C. Gardiner, 'Mar. Invest, in South Africa,' "Turbinolid Corals," 1904.

Professor Herdman's collection contains numerous specimens of this interesting
species, of all sizes, and apparently collected from all the localities in which he dredged.

Mr. Stanley Gardiner has recently shown how very variable this species is,
and has absorbed the various species of Semper and Eehberg. The Ceylonese
specimens also show great variety, and, though I have not as large a collection for
comparison as Mr. Gardiner, I have no doubt of the identity of these specimens
with //. aequicostatus, and further, I have no doubt that the forms first described by

2 c

194 CEYLON PEARL OYSTER REPORT.

M. Edwards and Haime as H. roussetmi and afterwards as Stephanoseris rousseaui
are identical with H. cequicostatus.

The genus Stephanoseris was created, M. Edwards and Haime tell us, to
contain the species first described by them as H. rousseaui, but subsequently found
to possess synapticula, the presence of which involved its removal to the Fungiidae.
Subsequent authors, including Duncan, have accepted the genus Stephanoseris,
the last-named remarking that he places it with much doubt among the Fungiidse, for
he could not find synapticula in any of his specimens. Omitting for the moment
any reference to synapticula, we find that S. rousseaui differs from H. cequicostatus
(1) in having costee of unequal size, the tertiaries being particularly large; ('2) in
having four instead of five cycles of septa ; (3) in the fact that the septa do not project
externally.

Among the specimens in Professor Herdman's collection there are individuals
with all the characters of H. cequicostatus, and every intermediate grade between
these and individuals identical with S. rousseaui. Further, I have found well-marked
synapticular structures in every individual that I have examined, so there can be
no doubt that the forms described as Stephanoseris rousseaui are only varieties of
the forms described as Heterocyathus cequicostatus, and for reasons that I will explain
in the latter half of this paper 1 prefer to retain the latter name and to place the
genus among the Turbinolidse. Synapticula have, been found to be common to so
many different kinds of corals that their presence is no longer a reason for including
any given form among the Fungiidse. It should be mentioned that the synapticula
of Heterocyathus cannot be seen on mere inspection of the corallum, as they are
hidden by the swollen upper ends of the septa. In order to see them it is necessary
to make sections or to grind down the corallum below the level of the exsert septa,
when they are at once apparent, as is seen in fig. II. Thus one can easily under-
stand how both M. Edwards and Haime and Duncan, as well as other authors,
have failed to recognise their presence. The genus Psammoseris is described by
M. Edwards and Haime as resembling Stephanoseris (Heterocyathus) in almost
every respect, except that it has no pali. Not having access to the type specimens
of Psammoseris, I cannot speak with certainty on this subject, but I am inclined to
think that the genus was founded upon a variety of Heterocyathus, in which the
pali are so slightly developed as to be indistinguishable from the papilliform columella.
An inspection of Gardiner's excellent photographs of Heterocyathus (23, plate iii.,
figs. 13 to 19) shows that such a reduction of the pali is not uncommon. Psamruo-
seris, like Heterocyathus and Heteropsaminia, fixes itself in the young state on a
gastropod shell, which it subsequently envelops, and in the adult the shell, and the
spiral basal cavity continuous with it, is tenanted by a Sipunculid of the genus
Aspidosiphon. I shall refer to this interesting association in a subsequent part of
this paper, and I give in the second part a detailed account of the anatomy and
minute structure of the corallum of Heterocyathus.

SOLITARY CORALS. 195

Family : FLABELLID7E.

Simple corals, multiplying asexually by transverse fission from a fixed nurse-stock.
Corallum more or less compressed and flabelliform or cuneiform. Calice elongated ;
elliptical or angular at the extremities of the long axis. Septa numerous, increasing
in number during the growth of the corallum, chiefly by the addition of new septa in
the systems contiguous to the dnective septa, in such a manner as to appear to be
arranged in a variable number of ternary systems. The columella may be essential
and lamellar, or parietal and formed by the union of spines or trabecular projecting
from the lower ends of the principal septa. The wall is protothecal, and increases in
thickness only by addition to the inside surface. There is no edge-zone. Costa?
rudimentary or absent. Protothecal spines commonly present. Genera : — Flabellum,
Lesson, and Placotrochus, M. Edwards and Haime.

The Flabellida?, as characterised above, differ from all the other Turbinolidse,
among which they have hitherto been placed, in the absence of an edge-zone, that is, of
soft tissues external to the wall. The wall, as von Koch pointed out (30), is an
" epitheca," or, if we adopt the more exact nomenclature of Bernard, a " proto-
theca," that is to say, it is a direct upward continuation of the primitive basal cup
common to all Madreporarian corals, of which the development has been carefully
described by von Koch (28), de Lacaze Duthiers (33), and Duerden (10). As
there are no soft tissues external to the prototheca, it can only increase in thickness
by addition from within, and a section of the wall of Flabellum shows that the "dark
line of growth," which lies in or near the middle of the theca of other corals, is here
on the outside. As the wall does not increase in thickness externally, the costas are
very feebly developed, and there is an absence of the ridges, spines, and other
external ornamentations formed by the deposit of additional calcareous matter on the
outside of the wall by the activity of the calicoblasts of the edge-zone. The spinous
processes and rootlets found in most members of the Flabellida? are hollow, and only
become solid by the secondary deposit of calcareous matter within. Thus they differ
from ordinary costal spines, and their mode of formation, by local extensions of the
lip of the calice, has been indicated by de Lacaze Duthiers (33). In all the
Turbinolidse, as here limited, there is a distinct and usually well-developed edge-
zone. Hence the wall increases in thickness, both externally and internally, and
there are well-developed costal ridges, variously ornamented with granules, tubercles,
or spines, or sometimes the furrows between the costal ridges may be filled up by a
secondary deposit of an " epithecal " character, formed by the calicoblastic layer of
the edge-zone. Moreover, in the typical Turbinolidse, the septa grow in height well
beyond the limits of the protothecal cup, and are either prominently exsert or their
peripheral margins are thickened and become attached to one another, forming the
so-called pseudotheca of Ortmann. The marked difference in the structure and
growth of the wall seems a sufficient reason for separating Flabellum and its allies

2 c 2

196 CEYLON PEARL OYSTER REPORT.

from the rest of the Turbinolidse, and placing them in a separate family, Flabellidae,
and the septal arrangements and the peculiar mode of multiplication by transverse
fission, characteristic of the family, are additional reasons for keeping it apart.
Gardiner (23) has shown that the genera Blastotrochus, M. Edw. and H., and
Rhizotrochus, M. Edw. and H., must be absorbed. I have not been able to examine
Duncan's fossil genus Thysanus, and I do not include it in the family Flabellidse, as
here defined, because it has well developed granular and minutely spined costse,
which seem to indicate an external thickening of the wall.

The genus Placotroehus was not included by Duncan (13) in his alliance Flabel-
loida, but was placed, along with Sphenotrochus, Nototrochus, Plaeoeyathvs, and
Platytrochus in an alliance Placotrochoida, characterised by the presence of an
essential lamellar columella. For this I can find no justification whatever. Placo-
trochus has no edge-zone ; its wall is protothecal and devoid of costas ; it has the
compressed flabelliform or cuneiform shape characteristic of Flabellum ; it reproduces
itself asexually by transverse fission from a nurse-stock, and the truncated free forms
have a basal scar exactly like that of Flabellum ; it has protothecal spines ; its septal
arrangements are those of Flabellum, The only point of difference is the essential
lamellar columella, but this cannot outweigh the other characters, and the well-
preserved specimens of this genus in Professor Herdman's collection show that the
general anatomy of the polyp is the same as that of Flabellum, though 1 have
not yet completed my study of its microscopic structure. I have no hesitation in
placing it in the family Flabellidse.

Flabellum, Lesson (1831).

A detailed criticism of this genus has recently been given by Gardiner (22), to
which the reader is referred. The subdivisions Subpedieellata truncata and Ji.ra, of
M. Edwards and Haime, were shown by Semper (49) to be purely artificial, and
Duncan's subdivisions of the genus (13, p. 13) are still more arbitrary and unnatural.
Semper, and more particularly Gardiner, have shown that most of the living species
enumerated in the ' Hist. Nat. des Coralliaires,' must be regarded as varieties of a
few distinct species, and the latter author has reduced most of M. Edw. and Haime's,
as well as Semper's species to varieties of either F. pavoninum, Lesson, or F. rubrum,
Quoy and Gaimard, the latter sj)ecies being apparently protean in its characters. I
speak with all diffidence, for I have not had the opportunity of comparing a large
number of specimens, but I am inclined to think that Gardiner has gone too far.

Flabellum crassum, M. Edw. and H. — Plate I., figs. 3 and 3a.

There are in Professor Herdman's collection two specimens of Flabellum, from the
pearl banks in the Gulf of Manaar, which agree in almost every respect with Semper's
description and figures of F. irregulare. Both specimens were preserved in spirit ;
one I have decalcified, the corallum of the other when cleaned and dried measures

SOLITAKY COKALS. 197

30 millims. in height. The longer axis of the calice measures 16 milium, the shorter
axis 8 millims. ; the small scar at the hase 4 millims. x 2 millims. ; thus the ratio of
the axes is 2:1, agreeing with Semper's description. The corallum is compressed,
with rounded directive edges, whose sides form an angle of 23° in the dry specimen,
but a considerably wider angle in the decalcified specimen. The wall is constricted
at tolerably regular intervals and there are three short and stout spines at varying
heights on the directive faces, and two short and corroded spines near the basal scar.
The costse, especially those corresponding to the principal septa, are distinct and are
recognisable nearly down to the basal scar. The corallum is tall, and the directive
margins of the calice are nearly level with the lateral margins. The fossa is very
deep (6 millims.), narrow and slit-like. The parietal columella is formed by the union
of trabeculae given off from the lower ends of the principal septa. There are in all
sixty-six septa ; the primaries and the secondaries are of equal size and unite in the
columella. The twelve tertiaries are nearly equal in size to the primaries and
secondaries and also join the columella. The fourth cycle septa are short, excepting
those nearest the two directive septa at opposite ends of the long axis of the calice ;
these are nearly as large as the tertiaries and join the columella. Fifth cycle septa
are developed in the chambers between the primary and secondary septa on both
sides of the two directive septa, and are equal in size to the lateral tertiary septa.
In one chamber nearest to a directive, two sixth-cycle septa are developed, one on
either side of a quinary septum (see fig. 3a). (This interpretation of the septal
succession differs from Semper's, who explained an exactly similar arrangement by
supposing that additional second-cycle septa were developed in each of the terminal
systems, but it is clear that what he describes as additional secondary septa are
nothing more than third-cycle septa, which have grown to the same size as the
primaries and secondaries.) The result of this arrangement is that there appear
to be sixteen ternary systems, and, in addition, two small septa in one of the
systems. The principal septa have very slightly arched, or nearly horizontal, upper
edges ; their inner edges descend vertically to the columella, and they are distinctly,
though slightly, notched near their insertions on the wall. The surfaces of the septa
are thickly covered with fine spinose granules, which have an obscurely radial
arrangement. The dry corallum is brilliantly white in colour, and the lower five-
sixths of the wall is encrusted by a secondary calcareous deposit, which, on
decalcification, proves to be formed by interlacing algal filaments, as described by
Fowler for F. patagonichum (14). It is noteworthy that the costae are more
conspicuous in the region of this secondary deposit than they are at the upper end of
the corallum.

These Ceylonese specimens are so similar in all respects to the figures and
description of Semper's F. irregulare that there can be no doubt of their identity.
According to Gardiner (22), F. irregulare, Semper = F. ruhrvm, Quoy and
Gaimard, and there is certainly some resemblance between his (not wholly satis-

198 CEYLON PEARL OYSTER REPORT.

factory) figs. 26, 30, 31 and the Oeylonese specimens, while the latter can certainly
he included in the amended definition of F. rubrum given by Gardiner on p. 28.
It is further stated that F. va/riabile, Semper, is connected with F. rubrum by
individuals, Kemper's species being identified by Moseley with F. stokesi,
M. Edwards and H., which in turn is identical with F. oweni, aculeatum, spinosum,
debile, sumatrense and candeanum. I must take leave to doubt whether these
identifications are correct. That F. variabile, Semper, is identical with Flabellum
(Turbinolia) rubrum, Qrov and Gaimard, I have no doubt. The characters of the
corallum are similar and a comparison of Semper's figure of the expanded polyps
(49. plate xviii., fig. 1) of F. variabile with Quoy and Gaimard's figure of
F. rubrum (46, plate xiv., figs. 5 to 9) show a very close correspondence in the
coloration of the living animal ; in both there are the same six alternating radial
bands of deeper red and lighter red or yellow. On the other hand, F. irregulare, as
described by Semper, is red with as many radial white lines as there are principal
septa. The colour, no doubt, is variable to some degree, but it must be taken into
account, and what I wish to point out here is, that Semper is the only author who
has paid special attention to the colour of the living polyp ; that he had at least as
many specimens for comparison as Gardiner ; that he found a constant colour-
difference between his F. variabile and F. irregulare, and that the colour of the
former, and not that of the latter, corresponds to Quoy and Gaimard's figure of
F. rubrum. It follows, therefore, that F. variabile = F. rubrum, and, as Semper
showed beyond all cavil, F. stokesi, oweni, aculeatum, spinosum, debile and
sumatrense, all M. Edw. and H., are synonyms of F. variabile.

The Oeylonese specimen, which I identify with F. irregulare, Semper, agrees very
closely with F. crassum, M. Edw. and H., especially in the characters of the septa,
which are notched near their insertion on the wall (38, plate viii., fig. 8a), and
have thickened and rugose internal borders. Semper observes that the two species
are very closely allied, differing chiefly in the height of the corallum. There can be
no doubt that they are identical ; the specific name crassum has the priority, and I
therefore identify the Oeylonese specimen above described as F. crassum, M. Edw.
and H.

Flabellum rubrum, Q. and G. — Plate I, fig. 4.

There are three other specimens of Flabellum in Professor Herdman's collection :
a and b (locality not recorded), and c from deep water near Galle. These are all
truncate forms with a fairly large basal scar ; their measurements correspond closely,
their calices are elliptical, their calicular fossa? wider, and the septa are thinner, and
their inner margins less vertical than those of F. crassum, and the spines on the
surfaces of the septa are longer, further apart, and definitely arranged along radial
thickenings of the septa. Their walls are thinner than in F. crassum, the costse are
of ecpial size and very slightly developed, and the corallum is not constricted at

SOLITARY CORALS. 199

intervals as in that species, but only exhibits a succession of somewhat sinuous lines
of growth parallel to the curved margin of the calice. The three specimens differ
from one another in colour, in the characters of the lower edges of the septa, and of
the trabecular forming the parietal columella, but their resemblances are so great that
I must refer them to one species, viz., F. rubrwn, Quoy and Gaimard.

Specimen a. — Height, 12 milium ; calice, 15 millims. x 8 '5 millinis. ; scar,
G millims. x 35 millims.; depth of calice, 5 millinis. Angle formed by the directive
faces of the wall, 37°. The dry corallum is white and has a small directive spine at
each end of the scar. The margin of the calice is entire, the directive ends of the
calice 2 millims. lower than the sides. There are eighty-two septa, whose character
and arrangement are shown hi fig. 4. The lower margins of the septa are furnished
with sharp spines, but are not sinuous ; the septal trabecular uniting to form the
columella are large and spiniform. The character of the trabecular columella
approximates this specimen to F. profundwn, M. Edw. and H., which is a variety of
F. rubvum.

Specimen b. — Height, 13 millims.; calice, 17 millinis. x 8*5 millims.; scar,
7 millims. x 3 5 millims. ; depth of calice, 4*5 millims. Angle formed by directive
faces of wall, 45°. The dry corallum is of a brownish-grey colour and bears short
protothecal spines at either end of the scar. The calicular margin is crenulate, the
directive ends of the calice about 2 millims. lower than the sides. There are eighty
septa arranged in twenty apparent ternary systems. Eight of the tertiary septa are
equal in size to the primaries and secondaries and unite with them in the columella,
but four of the laterally placed tertiaries are of smaller size, and either barely reach,
or do not reach, the columella. The principal septa are thin, their inner margins
slope obliquely into the fossa, their lower edges are expanded and slightly sinuous
and terminate in nodular trabecular, which unite to form the columella. The septal
surfaces bear distinct radial ridges ornamented with spiniform granules. This
specimen is clearly identical with F. crenulatum, M. Edw. and H., which is a variety
of F. rubrum.

Specimen c. — Height, 13 millims.; calice, 15 millims. x 9 millims.; scar, 8'5 millims.
x 4 millinis. ; depth of calice, 5 millims. Angle formed by the directive faces of the
wall, 35°. The wall in the dry corallum is of a deep reddish-brown colour, and the
same colour extends half way across the septa, a peculiarity which Semper
(49, p. 250, footnote) says is confined to F. pavoninum and F. distinctum. The
concentric lines of growth on the wall are distinct, the costar are scarcely recognisable,
and there are no protothecal spines. The calicular margin is entire, the directive
ends of the calice are about 1^ millims. below the lateral margins. There are
seventy-two septa, the tertiaries equal in size to the primaries and secondaries and
meeting them in the columella. The fifth cycle is complete in the terminal systems,
but incomplete in the lateral systems. The septa resemble those of specimen b, but
their inner edges are more vertical and their lower ends more decidedly sinuous.

200 CEYLON PEARL OYSTER REPORT.

The calicular fossa is rather narrow, and the septal trabecular forming the columella
si tort and bluntly spiniform. This specimen might be referred either to F. rubrum
or F. stokesi, M. Edw. and H. ; but it cannot be either F. pavoninum or dLstiactum,
the colour of the septa notwithstanding. These three specimens, therefore, are
varieties of F. rubrum, QuOY and Gaimard, which is synonymous with F. variabile,
Semper, but is, in my opinion, distinct from F. crassum, M. Edw. and H. (= F. irrc-
gulare, Semper).

I differ, with some hesitation, from Mr. Gardiner, but I feel bound to point out
that his reasons for uniting nearly all the described species of Flabellum under
F. rubrum are not satisfactory. He relies very much on certain numbers and
measurements, and he appears to think that if one form is connected with another by
some individuals, the two must be reckoned as constituting one species. Now, in the
first place, much depends on the numbers or measurements that are chosen for
comparison. Two of his characters, viz., the number of septa fusing by trabecule
and the total number of septa, are characters depending on the age of the coral, and
are therefore of little value. His third character, the relation of the length to the
breadth of the calicle, is tolerably constant at all ages after maturity, and is therefore
better. But whatever characters are chosen, it is not sufficient to set out the results
of the measurements in a simple table and to say that, since the average measure-
ments of a number of unequal groups of individuals can be arranged in a continuous
series, all the specimens measured must belong to one species. To deal with the
statistics of a number of individual forms, proper statistical methods must be
employed, or systematic zoology will be thrown into confusion. The extreme
measurements — the characters that vary most widely from the mean — of two closely
allied species may be expected to overlap, but the fact of their overlapping does not
break down the distinction between two species whose mean is different. We have
at present no data for determining the range of variation in the species of Flabellum
by statistical methods, and until such data are available I prefer to place reliance as
much upon such characters as the shape of the columellar trabecular, the shaj>e and
thickness of the septa and the kind and arrangement of the granules on their
surfaces, the visibility or otherwise of the costas, &c, as upon measurements which
have been shown to vary very widely according to the age and condition (whether
recently liberated or not) of the individual.

Placotrochus, M. Edwards and Haime (1848).
The reasons for placing this genus in the family Flabellidse have been given above.

Placotrochus laevis, M. Edw. and H. — Plate I., fig. 5.

Three specimens from Periya Paar are referable to this species. Two were
preserved with the polyps partly extended, one is dead and corroded. The three

SOLITARY CORALS. 201

specimens vary between the following limits : Height 16 millims. to 12 millims.
Calicle from 16 millims. x 7 millims. to 14 millims. x 6 millims. Scar from 6 millims.
x 2 millims. to 7 millims. x 2*5 millims. (the smallest specimen has the largest scar).
Angle formed by the directive faces of the wall 33° to 40°. They correspond exactly
with M. Edwards and Hatme's description, except that the directive faces of the
wall show only the merest traces of wing-like expansions, in this respect agreeing
with Semper's specimens. This appears to be a very rare species. It is remarkable
among the Flabellida? for the perfectly regular development of the septa, the first,
second, and third cycles are equal in size, and appear to form with the intermediate
quaternary and quinary cycles twenty-four regular systems (Plate I., fig. 5).

Family : FUNGIID^E, M. Edwards and Haime.
Fungia (pars.) Lamarck.

Fungia danai, M. Edwards and Haime (1851).

Two specimens from the pearl banks, Gulf of Manaar, are referable to this species.
The larger specimen is subcircular, measuring 18 centims. x 19 centims., and the
central part of the upper surface is raised into a prominent convexity. The smaller
specimen has no central prominence, but is evenly convex above and concave below.

Fungia dentig-era, Lettckart, 1841.

A single specimen, measuring 16'5 centims. in length, 11 centims. in breadth, and
5"5 centims. in height. According to Doderlein this is a variety of F. scutarta,
Link, but Gardiner (20) has given a sufficient criticism of this author's varieties.

Locality : — Gulf of Manaar.

Cycloseris, Michelin (}3ars.) (34).

Cycloseris cyclolites, Lamarck

Fungia cyclolites, Lamarck, 'Hist, des Anim. sans Vert.,' ii., p. 236, 1816.
Cycloseris cyclolites, M. Edw. and H., 'Ann. des Sci. Nat.,' 3e se>., xv., p. 112, 1851.

Professor Herdman's collection contains thirty-two specimens of this species,
twenty-five of which come from the pearl banks in the Gulf of Manaar, four from
south of Modragam, one from off Mutwal Island, two from Trincomalee. The range
of variation in this species is considerable. In the thirty-two specimens the diameter
of the corallum varies from 42 millims. to 27 millims., the height from 19 millims.
to 10 millims., the depth of the basal concavity from 9-5 millims. to 1 millim., the
length of the fossa from 15 millims. to 6 millims. There is also a considerable amount
of variety in the granulation of the costae, but the septal characters are closely similar
in all the specimens.

2 D

202 CEYLON PEARL OYSTER REPORT.

Cycloseris tenuis, Dana.

Fungia tenuis, Dana, ' Zoophytes,' p. 290, 1846.

Cycloseris sinensis, M. Edw. and H., 'Ann. des Sci. Nat.,' 3e ser., xv., p. 112, 1851.

Cycloseris hexagonalis, M. Edwards and Haime, loc. cil.

Cycloseris hexagonalis, Gardiner, Willey's 'Zool. Results. Solitary Corals,' 1899.

Professor Herdman's collection contains seven specimens, of which the largest
measures 40 millims. in diameter, while the smaller specimens vary from 22 millims.
to 9 millims. in diameter. The smaller individuals, among which are two somewhat
hexagonal forms, agree in all respects with the definition of C. hexagonalis, M. Edw.
and H, and, allowing for individual variation, with Gardiner's excellent photographs
of this species (20, plate xx.). Two of them have definite basal scars, not yet filled
up by secondary deposit, and the costse reach to the edges of the scars ; in the other
specimens the costse, and the extent of the central basal granulation, differ to a
considerable extent, but the septal characters are quite uniform and none of the
septa are fenestrated. In the largest specimen the two last cycles of septa are
fenestrated, a character which would attach it to C. sinensis, M. Edw. and H. This
species however is synonymous with Cycloseris [Fungia) tenuis, Dana, and the latter
name has the priority. Except for the fenestration of the lower cycles of septa there
is absolutely no difference between the largest and the smaller specimens in Professor
Herdman's collection, hence I am of opinion that hexagonalis and sinensis are merely
varieties of Dana's species tenuis, and must be absorbed into it.

Diaseris, M. Edwards and Haime (37).

Quelch (45, p. 121) has expressed his opinion that the specimens referred to this
genus are nothing more than broken and distorted individuals of Cycloseris tenuis,
but I cannot agree with him. Professor Herdman's collection contains a specimen of
C cyclolites, which has been broken and repaired, and shows considerable distortion
and re-arrangement of the septa, but its characters are widely different from Diaseris.
It is true that the structure of the wall, the septa, and even of the rudimentary
columella of Diaseris are very similar indeed to that of Cycloseris, as M. Edwards
and Haime pointed out, but the former genus is sufficiently distinguished by its
peculiar method of reproduction by radial division, whence the corallum has the
appearance of being composed of a number of lobes with rounded edges, or is divided
into wedge-shaped fragments. This method of reproduction has been so sufficiently
and clearly described by SfiMrER (49) that I can add nothing to his account, which is
fully borne out by a study of the Ceylon specimens.

Diaseris distorta, Michelin.

Fungia distorta, Michelin, ' Mag. de Zool. V. Zooph.,' pi. 5, 1843.
Diaseris distorta, ' Ann. des Sci. Nat.,' 3e se>., xv., p. 118, 1851.

Numerous specimens and cuneiform fragments from off Mutwal Island, and some
small fragments and specimens from deep water off Galle and from Kaltura.

SOLITARY CORALS. 203

Diaseris freycineti, M. Edwards and Haime (37).

A number of petaloid fragments from the Gulf of Manaar and from off Kaltura
appear to belong to this species. As described by M. Edwards and Haime, the septa
are of less height, much more closely packed, and have thicker edges than in D. distorta.
I may add that all the septa are more or less fenestrate. The spirit specimens are of
a rich red-brown colour, the upper surface exhibiting black spots near the inner ends
of the septa. These spots probably indicate the position of the tentacles, but the
specimens are so contracted that 1 cannot speak with certainty on this point.*

Family : EUPSAMMIID.E.

As Semper has remarked (49, p. 256), the difficulty of determining the species and
even the genera of the Eupsammiidse is very great. In order to relieve himself of the
difficulty, he founded the genus Khodopsammia to include his Philippine specimens,
and the genus has been accepted by most subsequent authors, even by Duncan in
his critical revision of the genera of the Madreporaria, and this in spite of the fact
that, as I shall show, Semper's genus was confessedly provisional.

Semper's definition of the genus Khodopsammia runs as follows: — " Polypary
simple or with lateral buds, free or attached, sometimes cylindro-conical, sometimes
compressed (not, as Duncan gives it, ' Corallum simple or colonial, free or attached,
with lateral buds'). Epitheca absent or rudimentary. Costa? simple, visible from
the base upwards, similar, thickly granulated. Calicular fossa rather deep, with a
more or less prominent columella consisting of curled leaflets (aus gewundenen
Blattchen). The septa narrow, with sharp edges, scarcely rising above the lip of the
calice ; those of the first cycles equal, extending right down to the columella ; those
of the third cycle smaller, and also united to the columella ; the septa of the remain-
ing— often irregular — cycles much narrower and invariably united to those of the
preceding cycle."

This definition differs from M. Edwards and Haime's description of any genus of
Eupsammiidae, as it includes both single and compound, free and attached, forms, but
it embraces all members of the genera Eupsammia, Balanophyllia, Leptopsammia,
and Endosammia, and possibly some species of the genus Dendrophyllia as defined
by the authors of the ' Histoire Naturelle des Coralliaires,' and it is quite clear, from
what Semper says, that he meant at least the first four of these genera to be included
in his new genus. Thus (lot: cit., p. 256) he says: " Da mir nun leider das fossile
Material fehlt, welches nothig ware, urn diese Iucongruenzeu grundlich ausgleichen zu
konnen, so ziehe ich es vor, hier die von mir bei Bohol aufgefundenen 8 Arten ohne

* Professor Hekdman 's "Field-notes' contain the following colour record of this species when .alive: —
Station XLIII., off Kaltura, 22 fathoms, " also a living group of four ' flabellums ' having upper side dark
purple on the outer margin and dark green within, mottled with paler yellowish-green, grey, and deep
drab-purple, the septa especially being of the latter colour — the under side is white." The accompanying
sketches show that the Coral is a Diast ris,

2 D 2

204 CEYLON PEARL OYSTER REPORT.

Riicksicht anf ihre Basis als Species einer neuen Gattuug zu beschreiben, wobei ich es
Anderen ttberlasseu muss, die schon beschriebenen lebenden und fossilen Eupsammidse
nach den durch die philippinischen Form en sicb ergebenden Andeutungen zu unter-
suchen, und mit diesen in systematischen Zusainmenhang zu stellen." Notwith-
standing Skmper's hope that one of his successors would show the proper systematic
relation between his Philippine specimens and previously described fossil and living
forms, Duncan, in his 'Revision of the Madreporaria,' retains Semper's genus
Rhodopsammia, but removes it from the " alliance' Balanophylloida, to which, on its
author's own testimony, it belongs, and places it apart as a genus which cannot be
included in any alliance !

The Eupsammiidse in Professor Herdman's collection include free and fixed, solitary
and colonial forms, and most of them resemble one another very closely in such
distinctive characters as septal arrangement, columella, and costse. The solitary
forms (with the exception of Heteropsammia) must, without doubt, be referred to a
single genus, and they agree in all respects with Semper's definition of the genus
Rhodopsammia. But if we agree with Semper and ignore, as we must after the
study of a sufficient number of specimens, such variable characters as attachment or
its contrary, the presence or absence or the completeness or incompleteness of a fifth
cycle of septa, or the relative thickness of the septa, we must recognise that the
genus Rhodopsammia, Semper, includes Eupsammia, M. Edwards and Haime ;
Balanophyllia, Searles Wood ; Leptopsammia, M. Edw. and Haime ; Endo-
psammia, M. Edw. and Haime, and it becomes a question as to what name shall be
used to denote the single genus into which all the other genera are absorbed. The
genus Balanophyllia was founded by Searles Wood in 1844, and there is no
objection to be taken to it on the score of iudetiniteness. Therefore, by the rules of
nomenclature, it has the priority over the genera founded by M. Edwards and
Haime in 1848, and a fortiori over Semper's genus Rhodopsammia founded in 1872.
Hence I suggest the amendment of the definition of Balanophyllia in the terms of
Semper's definition of Rhodopsammia, the latter name being dropped and
Eupsammia, Leptopsammia, and Endopsammia merged into Balanophyllia.

It is a characteristic, though not a peculiarity of the Eupsannniidie, that the septa
of the later cycles curve towards and are usually united to those of the preceding
cycle, and it commonly happens that the septa of what on the ordinary system of
reckoning would be called the last cycle are larger than those of the preceding cycle.
Semper, many years ago (49, p. 25y), called attention to the inapplicability of
M. Edwards and Haime's law of septal sequence to the Eupsammiidpe as well as to
other Madreporaria described in his well-known memoir, but until recently there has
been no satisfactory explanation of the apparent irregularity in the septal sequence
of these forms, and it has been necessary, in spite of its obvious disadvantages, to
retain M. Edwards and Haime's system of notation in describing the septa of all
Madreporaria. But as long ago as 1871, Pourtales (43) gave a description of the

SOLITARY CORALS. 205

development of Balanophijllia Jloridana which, if it had received the attention it
deserved, would have given a full explanation of the peculiarities of the septal
arrangement in this genus. Pourtales' description is as follows : " The youngest
individuals observed have the shape of a truncated cone attached by the base. The
wall is (piite smooth, imperforate, and the septa, twelve in number, equal and not
quite extending to the centre, where the rudiments of the columella are already

visible The next step is the formation of costee on the upward prolongation of

the wall. They first appear in the shape of sharp points grouped about the origin of
the septa. At the same time an opening appeal's on the border and rather outside of
the calicle, opposite each of the secondary septa, which gradually widens inwards,
apparently dividing the septum in two. The two borders of that opening become the
tertiary septa ; the secondary septum is gradually pushed inwards and is replaced by
a new one growing out on the same radius from the wall, and but loosely connected

with the jointed tertiaries and original secondary The interior part of the

tertiary septum is now to all intents and purposes a palus As the growth

proceeds the point of junction of the tertiaries and secondaries moves further into the
calicle until it reaches the columella. At this period the older or internal part of the
secondary septum has nearly entirely disappeared and the same process of growth goes
on with the septa of the fourth cycle which become joined to those of the third." In
these few sentences Pourtales anticipates the discoveries of several recent authors.
The imperforate wall is clearly the hasal plate of von Koch (= prototheca, Bernard).
The description of the formation of costre on the upward prolongation of the proto-
theca agrees exactly with de Lacaze Duthier's description and figures of the
development of these structures in Balanophijllia regia (33, plate x., tigs. 20
and 21). The description of the bifurcation of the peripheral ends of each secondary
septum, and the formation of a new secondary septum on the same radius, is in
exact agreement with Duerden's (9 and 10) account of the septal sequence in
Siderastrcea radians. But, although the credit of priority must rest with Pour-
tales, it is the last-named author who has given a full and perfectly intelligible
account of this mode of septal sequence, and has shown in detail the relation hetweeu
the order of appearance of the septa and that of the mesenteries. In his valuable
and beautifully illustrated memoir, Duerden has shown that in Siderastrcea radians
the six septa comprised in the first cycle appear simultaneously, and are situated
within the entocceles of the six primary pahs of mesenteries. The six septa of the
second cycle make their appearance later and occupy the six exoccelic chambers, thus
alternating with the six primaries. As growth proceeds the peripheral ends of all
the septa become bifurcated, as a consequence of the continuous addition of skeletal
nodules to their outer ends, but the angle formed by the bifurcating limbs is much
larger in the secondary (exoccelic) than in the primary (entoccelic) septa. The bifur-
cations become filled up and disappear in the primary septa, but in the secondaries
they continue to extend, and presently a second mesenterial cycle is developed, each

206

CEYLON PEARL OYSTER REPORT.

mesenterial pair being situated within the angle formed by the bifurcated peripheral
extremities of the secondary septa. After a time a new cycle of entocoelic septa is
formed, each septum within a mesenterial pair of the second cycle, and as growth
proceeds these secondary entoccelic septa grow ceiitripetally, and fuse with inner
portions of the septa which originally constituted the exoccelic cycle. Thus a stage

' X'

1 x'

j 2 3

Fig. I. Diagram illustrating the development of the septa, and their relation to the mesenteries in Balano-
phyllia flmidana (Pouiitalks), and Siderastrwa radians (Duerden). The primary entoccelic septa are
shaded with cross lines; the exoccelic septa are dotted ; the secondary entoccelic septa are black; the
tertiary entoccelic septa are handed black and white. In fig. 1 the outer ends of the six exoccelic
'septa have bifurcated to form the secondary exoccelic septa A'2; within each bifurcation a new
mesenterial cycle of the second order has been formed, embracing an entocoelic septum of the second
order En.2. In fig. 2 the secondary entoccelic septa have grown inwards and fused with the inner
limbs of the first formed exoccelic septa, thus forming the apparent second cycle of septa En? + A"1.
In fig. 3 the secondary exoccelic septa A"- have in turn bifurcated, and in each of the twelve
bifurcations a mesenterial pair of the third order is developed, embracing a tertiary entoccelic
septum En?. The primary entoccelic septa are denoted by 1, 1 in all the figures.

is reached in which there are three cycles of septa ; two cycles, namely those which
are ordinarily reckoned as the first and second, are entoccelic, the remaining cycle,
which would ordinarily be reckoned as the third, is exoccelic. But it is clear that
each apparently secondary entoccelic septum is a compound structure, its inner end
formed of a septum which was actually second in order of appearance, its outer end of
a septum which was actually third in order of appearance. And it is further clear
that the exoccelic septa, which appear to constitute the third cycle, are really the
forked peripheral ends of the septa that were second in order of appearance. The
formation of the next cycle of mesenteries and septa follows the same rule. The
exoccelic septa become bifurcated at their extremities ; new mesenterial pairs of the
third cycle are formed within the bifurcations ; a new cycle of entoccelic septa is
formed within the pairs of third-cycle mesenteries, and these grow ceiitripetally and
unite with the inner limbs of the apparently third-cycle septa of the previous

SOLITARY CORALS. 207

stage, the bifurcated outer ends of these same septa constituting the apparent fourth
cycle. The further development has not been followed, but it probably continues on
the same plan. Thus it appears, as is shown in the diagram above, that at any
given stage in the growth of the coral, all the entocoelic septa, after the first cycle,
are composite structures, but all the exocoelic septa, forming the apparently last
cycle, are the derivatives of the original exocoelic septa which were second in order of
appearance. One can scarcely imagine anything more at variance with M. Edwards
and Hatmk's law of septal sequence. Dukrdkn has further shown that there is a
regular dorsi-ventrality in the mesenterial, and therefore in the septal succession of
Siderastrcea radians. New mesenterial pairs appear Hist in the dorsal member of
the two exocoeles of each system. This fact may be <>t' assistance in explaining the
irregularities in the apparently last cycle of septa which are so common in the
Eupsammiidaa and other corals, but, as I shall show, the same sequence does not
seem to obtain in Balanophyllia as in Siderastrtea, and it is quite probable that the
dorsi-ventral order of the appearance of the mesenteries varies as much in the
Kcleractinias as it does in the Malakactiniae.

The convergence and union of the inner ends of the lower orders of septa is a much
more common phenomenon than is generally supposed. Tn the collection of corals
forming the subject of this paper, such septal unions occur in Rhodocyathus, Hetero-
cyathus, and Paracyathus among the Turbinolidse. They are characteristic of the
genera Cycloseris and Diaseris, and, as I have shown elsewhere (6), they are well
marked in the anthoblast of Fungia. But they are above all characteristic of the
Eupsammiidae, in which the septal arrangement of the adult appears to retain its
embryonic character, and it may be inferred from the septal characters of the adults
that the septal sequence in the course of development has followed the law of
Pourtales and Duerden. It is obvious that, if this inference is correct, all the
adult septa of the apparently last cycle will be exocoelic, while those of the apparently
penultimate cycle will be entoccelie and contained within the mesenterial pairs of the
last formed, and therefore smallest, mesenterial cycle. In the second part of this
paper I shall show that this is the case, not only in Hetero'psammia and a species
of DendrojyJii/Uia, among the Eupsammiidse, but also in Hetfrocyathus among the
Turbinolidse.

It might seem desirable to amend Milne-Edwards and Haime's method of
enumerating the septa, and to adopt a system more consistent with the facts of
developmental sequence as now ascertained. But after several attempts to invent a
new notation, I have decided to retain the old method. After all, we have only
Pourtales' and Dtjerden's accounts to go by, and we are by no means certain that
what is true of the species thev describe is true of all Madreporaria. Future researches
may bring to light considerable differences in the septal sequences of different groups,
and it would be premature to invent a system of notation that might prove to be
inapplicable to new cases. I have therefore retained the old system in describing the

208 CEYLON PEARL OYSTER REPORT.

Eupsammiidse in Professor Herdman's collection, leaving it to be understood that the
words " secondary," " tertiary," &c, indicate only the size of the septa, and not their
order of succession. And in this connection it may be noted that de Lacaze
Duthier's account (33) of the septal succession in Balanophyllia regia differs widely
from that of Pourt ales' account of B. jloridana, and therefore from Dtjerden's. I
am inclined to think, however, that de Lacaze Duthiers missed the actual
succession of the development of the septa. An inspection of his figures (Plate x.,
figs. 23 and 29) will show that they are not inconsistent with Dtjerden's and
Pourtales' accounts. In fig. 23, the septa marked 2 are clearly bifurcated at their
distal ends, and the sclerite lying within each bifurcation gives every appearance of
having been formed independently and having become secondarily united with the
inner limb of the bifurcated secondary. In fig. 29 the septum marked 3 is equal in
size to, and clearly continuous with, the inner end of the septum marked 2, the latter
thinning out very much towards its point of union with the angle of the Y. And in
the same figure the " tertiary " septum occupies the position of a " quaternary," and
the septa that occupy the positions of " tertiaries " are described as subsequent
formations. They are clearly the new entoccelic septa formed between the bifurcated
peripheral ends of the exocoelic septa marked 3.

Balanophyllia parallela, Semper — Plate I., figs. 6 and 6a.

Rhodopsammia parallela, Semper, ' Zeit. fur Wiss. Zool.,' xx., 1872.
Ehodopsainmia parallela, Fowler, ' Quart. Jour. Mier. Sci.,' xxv., new ser., 1885.

A single specimen from the pearl banks, Gulf of Manaar. Height of corallum
29 millims. ; longer axis of calice 18 millims. ; shorter axis 13 millims. ; depth of
calice 8 millims. In this specimen the outline of the calice is not so hexagonal as in
the Philippine form described by Semper, and the base is narrow and pedicellate,
whereas the Philippine forms are free and pointed below. Septa in six systems and
five cycles, the fifth cycle being incomplete. The primary and secondary septa are
equal in size, similar and somewhat exsert ; their outer ends are thickened and
porous, their inner ends sharp, entire, and descend nearly vertically to unite with the
columella ; their sides are ornamented with radiating rows of very small granules ;
their upper ends more or less fenestrate. The tertiary septa are smaller and their
inner ends become thickened and trabecular and pass into the columella ; each is
joined at two-thirds of its length from the lip of the calice by the quaternary septa.
There is an incomplete cycle of quinary septa, which can best be described by saying
that the peripheral ends of the quaternaries bifurcate and a small septum is developed
between their bifurcated ends in each directive moiety of the four primary systems
adjoining the directive septa, and the same arrangement is present in all but one of
the four outer moieties of the same systems. The quaternary septa are not bifurcated
in the lateral primary system on one side of the coral, and on the other side only one
is bifurcated and contains an entoccelic septum in each of the moieties of the system,

SOLITARY CORALS. 209

In other words, the formation of a new cycle of entoccelic septa has progressed more
rapidly at the two directive ends than at the sides, and the sequence is therefore
different from that ohserved by Duerdex in Siderast/rcsa radians.

This specimen is in many respects intermediate between Semper' s Rhodopsammia
carinata and R. parallela, and I am inclined to think that these two and R. amema
are only varieties of one species. The Ceylonese specimen agrees with R. parallela
in having vertical inner edges of the primary and secondary septa, in the costse being
of equal size, in the presence of a fifth incomplete cycle of septa, and in the depth of
the fossa. In all these respects it differs from carinata, but such characters are
clearly variable. The calice is less compressed and more regularly oval than in
Semper's figures of parallela, but Fowler (14) gives a figure of the calice of this
species in which the difference between the longer and shorter axes is even less than
in the specimen here described, and the septal arrangement is almost identical.

Balanophyllia cumingi, M. Edwards and Haime (36) — Plate II., figs. 7 and 7a.
Rhodopsammia ovalis, Semper, ' Zeit. fur Wiss. Zool.,' xxii., p. 262, 1872.

There are three specimens in Professor Herdmax's collection, of which one, from
deep water oft' Galle, so closely agrees with M. Edwards and Haime's description
and figure of B. cumingi that T have no hesitation in referring it to this species. It
is somewhat larger than the type specimen (the height being 12 millims., the calice
19 millims. x 12 millims.) and the fifth cycle of septa is well developed and nearly
complete. The second specimen is from Trincomalee and is a small colony, or rather
an aggregation consisting of three dead and decayed corallites, one small and one
large living corallite. Of the living corallites the larger measures : height 30 millims.,
calice 19 millims. x 12 millims., depth of calice 12 millims. Smaller corallite : height
12 millims., calice 8 millims. x 6 millims., depth of calice 5 millims. The larger
corallite is identical with the specimen of B. cumingi, from Galle, in septal
characters, the columella, and the costee, differing from it only in being longer and
turbinate in shape with a narrow base, whereas the Galle specimen is short with a
broad base. The septal arrangement is shown in fig. 7a, and it agrees very exactly
with Semper's description of Rhodopsammia ovalis. The tertiary septa are continued
below into the columella, but, as the figure shows, their lower ends converge distinctly
towards the secondary septa. The " quaternaries " are inserted very low down on
the tertiaries and the quinary cycle is nearly complete, but in the two lateral systems
three of the quaternaries adjoining the secondary septa are not bifurcated, nor is the
quaternary adjoining a secondary in one of the remaining systems. The smaller
corallite in the Trincomalee specimen has a nearly circular calice. The six primary
septa are conspicuously larger than the others and pass straight to the columella.
The secondary septa are joined just above the columella by the " tertiaries," and the
latter bifurcate in all but two instances and a new cycle of entoccelic septa is formed
in the bifurcations. This shows that the secondary septa have been formed in the

2 E

210 CEYLON PEARL OYSTER REPORT.

maimer described by Pourtales and Duerden, and that it is only at a later stage of
growth that the extension of the cohunellar trabecular overspreads the union of the
tertiaries with the secondaries and causes the latter to appear as if they sprung
directly from the columella. The primary and secondary septa are much thickened
and porous towards their thecal ends.

Balanophyllia socialis, Semper. — Plate II., figs. 8 and 8a.

Rhodopsammia socialis, Semper, 'Zeit. fur wiss. Zool.,' xxii., p. 260, 1872.

A beautiful cornuate specimen from the pearl banks, Gulf of Manaar, clearly belongs
to this species, as is shown by its subcircular calice, deep fossa, well-developed and
projecting columella, and four fully developed cycles of septa. Its measurements are :
height 27 millims. ; calice 10 minims, x 9 millims. ; depth of calice 6 millims. There
are no buds, but a small outgrowth of the theca 3 millims. below the lip of the calice
on the opposite side to that drawn in fig. 8 is evidently the commencement of a bud.
The primary and secondary septa are equal in size, as described by Semper, and
their surfaces are covered with very fine granules ("ausserst fein gekornelt"), but
they are only very slightly thickened peripherally, and in this respect resemble
Rhodopsammia affinis, which, as Semper himself says, is probably nothing more
than a variety of R. socialis. R. dubia and R. incerta, both Semper, again are
almost certainly nothing more than varieties of the same species. Semper describes
the quaternary septa as uniting with the tertiaries close to the columella. In the
Ceylonese specimen, as is shown in fig. 8a, the quaternaries do not invariably unite
with the tertiaries, and indeed they follow M. Edwards and Haime's law very
closely, the quaternaries nearest the primary septa being, as a rule, the best
developed. The infrequent union between the third and fourth cycle septa, however,
does not indicate that the septa have not been developed according to Pourt ales'
and Duerden's law, for the latter author shows (10, p. 103, fig. 12, c) how the
bifurcated external limbs of the lower cycles of septa commonly become detached
from the inner limb.

A dead and corroded specimen from deep water off Galle seems to belong to this
species. The calice was broken so much that its characters were indistinguishable,
hut a section taken lower down shows that it is peculiar in having a tetrameral
instead of a hexameral arrangement. There are four systems : the primary septa
are the largest, the secondaries are smaller, but pass direct to the columella, the
tertiaries also join the columella, but converge in a marked manner towards the
secondaries, and there is a complete fourth cycle uniting with the tertiaries. There
is no indication of this arrangement being derived from an originally hexameral
arrangement, and it must be regarded as a remarkable variation from the normal.

Balanophyllia taprobanse, n. sp. — Plate II., figs. 9 and 9a.
Height of corallum L5 millims. ; calice 0 millims. x 5 millims. ; depth of calice,

SOLITARY CORALS. 211

3 minims. The six primary sepia are much larger than the others, considerably
thickened at their thecal ends, and they, and the two septa adjacent to them, are
exsert, forming a crown of six prominent points round the edge of the calice. The
edges of the primary septa slope rather steeply inwards and then descend vertically
into the depth of the fossa without joining the columella ;. the inner half of the
surface of each primary septum is nearly smooth, the outer half hears a number of
spinose granules. There are five septa of lower orders in five of the systems and
seven in the sixth system (adjoining one of the directive septa), those adjacent to the
primary septa being the largest and most exsert ; the smaller septa join the columella
very deep down in the fossa. Costa? of the six primary septa larger and more
prominent than the rest, forming ridges extending nearly half-way down the corallum ;
the remaining costse subequal, extending to the base, the perforations between the
costse numerous and relatively large. Columella oval, very prominent, spongy in
texture. Buds are formed on opposite sides a little way below the margin of the
calice, each bud being astride of a primary costa.

A single specimen from dee]) water off Galle.

I have founded this species to receive a small and remarkably beautiful coral,
which resembles B. rediviva, Moseley, in the costse and arrangement of the septa,
but differs from it in size, in colour (B. rediviva is " reddish coloured," B. taprobance
a brilliant white), and in the fact that the secondary septa are not exsert. These
differences can hardly be due to immaturity, as the specimen of B. taprobanm bears
two fairly advanced lateral buds.

The septal characters are particularly interesting as being wholly unintelligible
unless interpreted in the light of Pourtales' and Duerden's work. In fig. 9a it is
evident that, in the lateral system on the left hand of the drawing, the inner ends of
the septa marked X.2, X.3, are the bifurcated arms of the exoccelic septa of the cycle
second in order of appearance ; these arms unite together and with the columella
very low down in the calice. The septum marked En.2 is clearly the entocoelic
septum of the cycle third in order of appearance. In the upper half of the system
the exoccelic septum has bifurcated to form the secondary exoccelic septa A".3, X.3,
and a new entocoelic septum En.3 is formed in the bifurcation. A precisely similar
arrangement obtains in four of the remaining systems, and in each it is the upper
member of the exoccelic septa that has bifurcated, giving a dorsiventral arrangement
entirely consistent with Duerden's description of the septal succession in Siderastrcea
radians. In the left-hand lower terminal system both exocuelic members have
bifurcated and an entocoelic septum is present within each bifurcation. In B. rediviva
all the systems are similar to this single system in B. ta/probanm, and the arrange-
ment would obviously be that which is normal in the Eupsammiidse, if it were not for
the preponderant size of the apparent quaternaries, whose outer ends become more
or less intimately united to the primary septa. The condition in B. taprobanm is a
step towards the septal arrangement in B. verrucaria, Pallas, and B. corrni,

2 e 2

212 CEYLON PEARL OYSTER REPORT.

Moseley. Semper (49, p. 263), in discussing the validity of his genus Rhodopsammia,
makes reference to the difference in the septal characters of the former species as
shown in Milne-Edwards and Haime's figure (36, plate i., figs. 6, 6a), but it
requires no great amount of ingenuity to show that the difference, great as it may
appear at first sight, is really nothing more than a slight variation of the grouping of
septa developed according to the sequence established by Pouktales and Duerden.

It is possible that the specimen here referred to Balanophyllia, is nothing more
than the initial individual of a colony of Dendrophyllia, and the two lateral buds
lend some support to this view. On the other hand, the buds, both from their
position and character, might very well be similar to those of B. socialis or parallela,
and as the specimen is a solitary one, I have |)referred to refer it, provisionally, to the
same genus.

Lobopsaminia, M. Edwards and Haime (36).

The small colony shown in Plate II., fig. 10, agrees very closely with the definition
of this genus, hitherto represented only by fossil forms. The structure of the
corallites, however, is so similar to that of Balanophyllia cumingi that it is hard to
draw any distinction between them ; the septa and columella are almost identical,
and the most that can be said is that the costse are smaller and the theca decidedly
thinner than in the latter species. If this specimen must be separated from Balano-
phyllia on account of its colonial habit, it must be placed with Lobopsammia rather
than with Ccenopsammia because of its elliptical calices and because the mode of
aggregation of the colony suggests that it has been formed by fissiparity rather than
by gemmation. Moreover, its likeness to L. cariosa, Goldfuss (25, taf. xiii., fig. 7),
is sufficient to establish its generic position.

Lobopsammia robusta, n. sp. — Plate II., figs. 10 and 10a.

Colony consisting of a few corallites borne on a very thick and short stem. The
individual corallites short, radiating outwards from the stem. Calices oval, somewhat
irregular in outline ; calicular fossae deep ; columella spongy, with a flat upper surface,
not projecting into the calice. Five cycles of septa ; the primaries and secondaries
subequal, not exsert, with nearly vertical inner edges. The quaternaries are united
with the tertiaries near the columella ; the fifth cycle is complete except in the lateral
systems, and the quinary septa unite with the quaternaries high up in the calice ; the
lower orders of septa fenestrate with serrated or denticulated margins. Costse fine, of
equal size, extending to the base of the stem.

Height of colony 27 millims. ,■ the largest calice measures 15 millims. x 11 millims. ;
depth of calice 7 millims. ; smallest calice 11 millims. x 8 millims.

A single specimen from deep water off Galle.

SOLITARY CORALS. 213

Dendrophyllia, de Blainville ('Diet. des. Sci. Nat.,' lx., p. 319, 1830).

Dendrophyllia gracilis, M. Edwards and Haime (36).

A portion of a colony, from deep water off Galle, appears to belong to this species.
The fragment only included three mature corallites, two of which, with the buds
attached to them, were decalcified for the study of the soft parts. The remaining
corallite is somewhat broken, but the size, septal characters and mode of budding
appear to be those of D. gracilis.

Dendrophyllia minuscula, n. sp. — Plate II., figs. 11 and 11a.

Colony ai-borescent ; the trunk and branches slender ; the lateral corallites disposed
in alternate and opposite pairs. Costae of equal size, rather prominent, finely
granulated. Calice circular, with a fairly deep fossa. Columella formed of a few
calcareous trabecular, moderately prominent in the fossa. Septa in six systems and
three cycles, with traces of a fourth. The primary septa exsert, forming a crown of
six points at the edge of the calice. The apparent tertiary septa next in size closely
applied to the primaries at their outer ends, their inner ends converging and uniting
deep in the calice just before joining the columella, The apparent secondaries short,
usually not united to the inner ends of the tertiaries. Height of colony 25 millims. ;
diameter of calices 2 millims.

A single small colony and a fragment of this very elegant species from deep water
off Galle. The large exsert primary septa recall those of Balanophyllia laprobance,
but the corallites are much smaller, the primary costaa are not prominent, and there
are three cycles of septa with traces of a fourth in one system only, in the corallite
depicted in fig. 11a. This figure shows very clearly that the apparent tertiaries are
the exoccelic bifurcated ends of the original secondary cycle, and that the apparent
secondaries are the entoccelic septa formed in the angles of the secondaries, and,
therefore, are the third cycle in order of appearance.

Heteropsammia, M. Edwards and Haime (36).
Heteropsammia michelini, M. Edwards and Haime (36).
Numerous specimens from nearly all stations. The septal arrangement and anatomy
of this genus are fully described in the second part of this paper.

PART H — AN ATOMY.

1. Heterocyathus sequicostatus, M. Edwards and Haime.
(Plates III. and IV., figs. 12 to 21.)

Gardiner (23) has given a short account of the anatomy of this species, but he
does not deal with the minute structure of the corallum, and I am able to supplement
his description of the anatomy of the polyp in many particulars.

214 CEYLON PEARL OYSTER REPORT.

The C o r al 1 u m. — Tn the majority of specimens there are forty-eight septa, which,
according to Milne-Edwards and Haime'k system of notation, would be described as
being regularly arranged in six systems and four cycles. It will become obvious,
however, that the septa have been developed according to Duerden's and Pourtales'
law, and that what are apparently the quaternary septa are really exoccelic septa
belonging to the second cycle in order of appearance. In some few specimens the
normal regularity of the septal arrangement is disturbed by the development of one
or two additional septa in one or more of the systems. These additional septa are
always inserted between the apparent tertiaries and the apparent quaternaries
adjacent to them, and suggest the commencement of a fifth septal cycle, but in every
system in which they occur there is so much irregularity in the septal arrangement
that I attribute their presence to a process of re-growth and repair rather than to the
formation of a new cvcle. The normal arrangement of the septa is shown in fig. IT.
There are three cycles of endosepta. The six primaries are prominently exsert and
their inner ends bear large paliform lobes just before they unite with the columella.
The secondary endosepta are likewise prominent and exsert, and there are two or
three prominent vertical pali at the inner extremity of each. The innermost of these
pali can with difficulty be distinguished from the vertical upgrowths of the columella,
and the outermost is fused to the inner free margin of the septum, so that there is a
transition between " true pali" and " paliform lobes." I am inclined to think that the
sharp distinction drawn between these paliform structures is artificial and untenable.
The inner ends of the third cycle of endosepta converge towards the secondaries and
unite with them just outside the columella through the intervention of palial
upgrowths. In specimens that have been rubbed down, the pali in front of the
secondary and tertiary septa are seen to unite to form a chevron, similar to the
chevrons characteristic of Deltocyathus.

Between every two endosepta there is an exoseptum. and in the deeper parts of the
corallum, i.e., below the level of the exsert portions of the septa, these exosepta are
seen to be united to the adjacent endosepta by tangential bars which, although they
do not exactly correspond to the similarly named structures of the Fungiidee, must be
described as synapticula. Their arrangement is very regular and characteristic.
Each primary endoseptum is joined to the exoseptum on either side of it by a
synapticulum which curves downwards and inwards to the fossa. These primary
synapticula, as they may be called, are equidistant from the centre of the coral, and
are situated nearer the centre than the others, so they form a sort of inner
synapticular ring. Each secondary endoseptum is similarly joined to the two adjacent
exosepta by synapticula which are situated somewhat further from the centre than
the primary synapticula, and similarly the tertiary endosepta are joined to the adjacent
exosepta by synapticula still further from the centre and forming an outer ring.
Thus there is only one synapticular bar in each interseptal loculus. In longitudinal
section the synapticula are seen to be narrow curved bars or partitions, in shape and

SOLITARY CORALS.

215

Fig. II. Diagram of a section through the corallum of Heterocyatkus ceguicostatus taken a short distance
above the Aspidosiphon chamber. The septa and synapticula are black, the stereoplasm is shaded
with dots. 1, 2, 3, endosepta of three cycles, ex., exosepta. sir., stereoplasm. syn.\ synapticulum
connecting a primary endoseptum with an exoseptum. syn.*; vyn.3, synapticula connecting secondary
and tertiary endosepta with adjacent exosepta.

structure resembling auy single synapticulum of Fungia. Their lower ends pass into
the secondary calcareous deposit in the neighbourhood of the Aspidosiphon chamber ;
their upper extremities end in free margins at heights varying according as they
connect a primary, a secondary or a tertiary endoseptum with an adjacent exoseptum.
The exosepta, in addition to this synapticular union, converge towards, and their
inner ends unite with, the tertiary endosepta.

216 CEYLON PEARL OYSTER REPORT.

In the upper parts of the corallum the synapticula form the only union between
adjacent septa, but a little distance lower down, and at some height above the
Aspidosiphon chamber, the interseptal loculi become largely filled up by a secondary
calcareous deposit, which, being of a dark colour from the presence of an abundant
brown pigment, stands in sharp contrast to the white septa and synapticula in a
rubbed down specimen. This deposit is crystalline in structure, and as it appears to
have been formed independently of and subsecpient to the original septa and
synapticula, it has all the characters of a " stereoplasm," i.e., a thickening on either
side of every septum filling up the interseptal loculi. It should be observed that
this deposit is most abundant and extends furthest towards the centre in the
interseptal loculi adjacent to the primary endosepta ; it is abundant, but does not
extend so far towards the centre in the loculi adjacent to the secondary endosepta ;
it is least abundant in the loculi adjacent to the tertiary endosepta (fig. II.).

The structure of the septa is somewhat complicated, and different to anything
that has hitherto been carefully figured or described. Each septum is very thin at
its inner end and gradually thickens towards its outer or costal extremity. As is
shown in Plate III., fig. 12, its inner edge is produced into a number of irregularly
shaped teeth or paliform lobes, which merge into the columella, and the impossibility
of distinguishing between these paliform sej)tal offsets and true pali has been alluded to
above. The laces of the septa are covered with small spiniform granulations which
seem to radiate from a centre situated just within the lower end of the synapticulum.
These spiniform granules are shown by longitudinal, horizontal and tangential sections
to be the extremities of the radiating trabecular (using this term in Pratz's sense) of
which the septum is composed. Plate III., fig. 13, is a careful drawing of a horizontal
microscopical section through the middle of the corallum, above the level of the
Aspidosiphon chamber. The thin inner ends of the septa are shown to consist of a
series of trabecular, standing in a single line, but at intervals diverging from the
centre to end in a spiniform granule on the septal surface. Passing from the centre
towards the periphery, the thickening of the middle and outermost parts of each
septum is seen to be due to the multiplication of the trabecular, which no longer
form a single line but stand in rows, first two, then three, and finally, at the costal
extremity, five or six deep. It can be seen clearly enough in the horizontal section
that these trabecular radiate not only towards the central and the peripheral ends of
each septum, but also outwards towards the two septal surfaces, on which they
emerge as the spiniform granules already described. The septum, then, is formed of
a number of trabecular which radiate in all directions from an imaginary centre
situated low down and nearer to the inner than the outer (costal) margin of the
septum, and it is apparent that the centrifugal growth of the septum is effected by
the addition of new trabecular on the outside of those first formed, these new
trabecular being at first in single series, but in the more peripheral parts becoming
arranged in curved rows, two, three, and more abreast. The curved growth lines,

SOLITARY CORALS. 217

showing the successive additions to the septum, and their relations to the rows of
trabecule, are well shown in the figure, as well as the way in which the outermost
trabecule in each row from place to place bend sharply towards the surface and end
in a spiniform granule.

It will be observed that there is no dark central line and no dark " centres of
calcification." The centre of each trabecula is clearer than the surrounding calcareous
deposit, and the latter, which appears white by reflected but dark by transmitted
light, seems to have a very fine fibrous rather than a crystalline structure, but I
could not make out the details clearly in the thinnest sections that I was able to
prepare. In longitudinal sections each septum is clearly seen to be made up of a
number of trabecula? radiating like a fan from a centre, and each trabecula appears
to be made up of a number of growth segments joined end to end, each segment
formed by the fascicle of diverging fibres described by Pratz and Miss Ogilvie as
characteristic of the minute structure of the Madreporarian corallum.

But if no definite crystalline structure can be discovered in the septa themselves,
or in the synapticula, the case is different for the " stereoplasm " in the interseptal
loculi. This stereoplasm, except where it is heavily charged with brown pigment,
appears dark by reflected and light by transmitted light, and, as is indicated in fig. 13,
it is clearly made up of coarse crystalline fibres. The orientation of these fibres should
be carefully studied, as it affords a proof that the stereoplasm is a secondary structure,
and not a simple addition to the thickness of the septa. At the thin innermost ends
of the septa the secondary fibro-crystals are disposed at right angles to the long axis
of the septum (as seen in transverse, i.e., horizontal section), and here we seem to have
the characteristic structure of the Madreporarian septum with a middle dark line or
" centre of calcification," &c. But where the stereoplasm is thick and completely fills
the interseptal loculus, it displays a number of curved lines of growth, generally
emphasized by the deposition of curved bands of dark brown pigment, and it should
be noted that, whereas the curved growth-lines of the septa have their convexities
directed outwards, the reverse is the case with the growth-lines of the stereoplasm.
The fibro-crystals of the latter, as is shown in fig. 13, are arranged in diverging
bundles, conformably to the curved lines of growth, in such a manner as to appear to
diverge from a stereoplasmic " centre of calcification " which is more apparent in
tangential than in horizontal sections. It is obvious that, after the septa and
synapticula have been formed, the soft tissues in the deeper parts of the corallum
shrink away from the septal and synapticular surfaces both externally and internally,
and as they shrink away the calicoblastic layer again enters into activity and deposits
the coarsely fibro-crystalline stereoplasm that eventually fills up the interseptal loculi
to a greater or less extent.

Concerning the brown pigment, I have very little to say. It appears to be
deposited in the form of minute granules between the fibro-crystals, but I found it
impossible to make sections sufficiently thin to admit of an accurate study of it.

i F

218 CEYLON PEARL OYSTER REPORT.

I have entered at some length into this question of secondary deposit or
" stereoplasm," as there appears to be unquestionable evidence of its existence in
Heterocyathus, and the minute structure of the corallum of this genus may prove
serviceable in the interpretation of the structure of some fossil corals. Miss Ogilvie
(41, pp. 93-99) denies the existence of a "stereoplasm" in both recent and fossil
corals, and she was perfectly right as regards the corals that she describes. There
is, however, a close analogy between the secondary thickening or stereoplasm described
by von Koch (29) in Pholidophyllum and that of Heterocyaihus, and it is possible
that the structure of Lonsdaleia indica, as described and figured by Waagen (51),
may be referred to the same type, but in neither case are the authors' figures suffi-
ciently detailed to enable one to speak with certainty. The description given by
Frech (19) of Idastrcea profunda, and quoted at length by Miss Ogilvie (41, p. 99),
might be applied with very little correction to Heterocyathus, but in this case again
the author's figure is on too small a scale to enable one to judge whether the structure
is identical or not with that here described.

General Anatomy of the Polyp. — As described by Gardiner (23),
the corallum is completely invested by the tissues of the polyp, the latter being
interrupted only at the mouth of the Aspidosiphon chamber. It must be borne in
mind that the young Heterocyathus is attached to and grows round a gastropod shell
(usually a Cerithium) tenanted by an Aspidosip>hon, and that the growth of the coral
and the Sipunculid proceed pari passu till we get the intimate association between
the two forms characteristic of their adult condition. The shell, which served for the
original habitation of the Sipunculid and the surface of attachment for the coral, is
completely overgrown and eventually greatly exceeded in size by the latter. It is
evident, then, that as the coral grows, its soft tissues must be folded down all round
the edge of the cup to form a " perisarc," using that term in the restricted sense
given to it by Bernard (4, p. 21). As the corallum grew round the shell, the
perisarc must have kept pace with it ; indeed, it would be more correct to say that
the soft tissues enveloped the shell, and that their innermost ectodermic or calicoblastic
layer secreted the corallum which eventually enclosed the shell, and, growing beyond
its limits, formed the Aspidosiphon chamber. Eventually the edges of the perisarc,
growing round the shell on all sides, meet and unite below, excepting in the region of
the mouth of the shell. Thus we get a basal union of the perisarc similar to that
observed in Fungia, and as in the last-named genus we find that the mesenteries are
prolonged into the cavity of the perisarc and divide up the extra-thecal ccelenteron
into as many entoccelic and exocoelic chambers as there are mesenteries. The
mesenteries, however, do not extend to the centre of the base of the coral. They may
be traced as far as the costse, but die out where the costre pass into the central
irregular basal granulations, and here the soft tissues appear to be supported on
the granulations in the manner described by Fowler (16). Above the level of
the synapticula, which not only physiologically replace but are in some respects

soUTAKY CORALS. 219

morphologically equivalent to a true theca, the external portions of the exocoeles and
endocoeles are continued into the corresponding internal chambers, and when the
polyp is fully expanded the depth of these intermesenterial chambers must be
considerable, owing to the relatively great height of the exsert septa. Spirit specimens
are, of course, very much contracted, and the soft tissues are everywhere stretched
over and down in between the septa, pali and columella to an extent that makes the
interpretation of sections a matter of considerable difficulty.

The Tentacles. — These, as stated by Gardiner, are twenty-four in number,
one corresponding to each endocoele. Owing to contraction it is somewhat difficult to
determine the position of the tentacles in an expanded polyp, but it appears that the
twelve tertiary tentacles form an external circlet situated quite at the edge or even
outside the margin of the calyx. The six primary and six secondary tentacles form
an inner circle within the margin of the calyx ; they alternate with one another and
the primaries are nearer to, but still at some considerable distance from, the mouth.
The tentacles are thickly covered with knob-like batteries ot nematocysts, which are
conspicuous in sections. Each tentacle is attached at its base to the two members of
a mesenterial pair, and the longitudinal muscle fibres of the mesenteries are continued
up into and may be traced to the tip of the tentacle. In contraction the tentacles
are completely introverted by the action of the muscle fibres, and in spirit specimens,
owing to the excessive contraction of the mesenteries themselves, the invaginated
tentacles are doubled over the endoccelic septa and each appears in sections to be
prolonged downwards into two pockets, one in each mesentery of the pair to which it
belongs (fig. III.).

The Peristome and Mouth . — Owing to the distance of the tentacles from
the mouth, the peristome is of considerable extent. It is closely contracted against
the numerous pali and paliform lobes at the inner edges of the septa, and its contours
are scarcely distinguishable in spirit specimens. The mouth is an oval aperture
whose size varies very much in spirit specimens ; in some it is contracted to a
narrow oval, in others it is widely open, giving a clear view of the soft tissues
investing the columella. The mouth, according to all received ideas on Actinian
anatomy, should open into a stomodaeum, but in Heterocyatlius there is no definite
stomodaeum, in the sense of a longer or shorter tube lined by a modified epithelium.
The ectoderm of the peristome is very thin, and at the lips of the mouth it is some-
what thickened, and, as described below, there is evidence of a distinct sphincter oris
and a distinct dilator oris (radiating fibres) muscle in this region. Immediately
within the lips, but not uniting to form a conrplete tube, are the large gutter-shaped
" filaments " of the six primary mesenterial pairs which undoubtedly perform the
functions of a stomodaeum, and may be described as twelve discontinuous portions of
the stomodaeum, united only by the thickened but not modified ectoderm of the lips.
There is no trace of a sulcus or sulculus (gonidial grooves). The reduction of the
stomodaeum seems to be clearly due to the great development of the pali and the

2 F 2

220

CEYLON PEARL OYSTER REPORT.

ent.m

Fig. III. Semi-diagrammatic representation of six successive horizontal sections through the same sextant
of Heterocyathus cequicostatus. The corallum is shaded with dots; the mesogloea is represented by a
black line, the endoderm by a contour line following the mesogloea; the ectoderm is blocked.
A passes through the exsert portion of the septa, B through the mouth, C and D are sections about
0-25 millim. apart and below the mouth ; the interseptal loculi in these sections are not reduced by
the formation of secondary calcareous deposit or " stereoplasm." In E a considerable amount of
stereoplasm bas been deposited in the loculi adjacent to the primary endoseptum. In F all the loculi
are reduced by the deposit of stereoplasm and the mesenteries contain ova. ent.l., ent.U., endosepta
of the first and third cycles ; portions of the second-cycle endosepta may be seen at the edges of each
sextant, ex., exosepta. fil., band-like primary mesenterial filaments, (jr., groove leading from the
mouth to one of the primary mesenterial filaments, m.f., coiled mesenterial filaments. /.', t.", /.'",
introverted tentacles of three cycles doubled over the corresponding endosepta. col., columella.

columellar upgrowths. There is not, in fact, room for a stomodseum in the contracted
polyp.

The Mesenteries. — These are of the normal Actinian character and are

SOLITARY CORALS. 221

forty-eight in number, consisting of six primary, six secondary, and twelve tertiary
pairs. The muscle banners are well developed, and the directive pairs, as is usual,
have the muscle banners on their outer faces. The general arrangement of the
mesenteries and their relations to the septa, mouth, and peristome are shown in the
diagram, fig. III., A, B, C, D, E, F, which represent six sections of the same sextant oi
a decalcified specimen of Heteroeyathus taken at different levels. The six pairs of
primary mesenteries extend further inwards than the rest, and they alone reach the
mouth. The six secondary pairs are intermediate in length, and the twelve tertiary
pairs are the shortest of all. As stated above, the free edges of the primary
mesenteries at and below the lip of the mouth are broadened out to form a T-shaped or
Y-shaped "filament" (fig. III., D,JIL). The inner face (i.e., the face directed towards
the central cavity of the polyp) is covered with a very thick ciliated epithelium which
passes somewhat abruptly into the thinner but still thick epithelium of the lips.
Below the level of the mouth the arms of the T- or Y-shaped filament become free
from adjacent tissues and may be traced as broad ciliated bands, which in section
appear T-shaped, or Y-shaped, or W-shaped, for a considerable distance below the
level of the mouth into the gastro-vascular cavity. As long as they retain this size
and shape, they are composed almost exclusively of very long attenuated ciliated
epithelial cells, whose nuclei stain deeply and are closely crowded together ; there are
few glandular elements interspersed among the attenuated cells, and no nematocysts.
Towards the bottom of the gastrovascular cavity these bands become smaller, and
gradually assume the normal shape of a mesenterial filament ; gland cells become
more abundant and large nematocysts make their appearance. At the bottom of the
cavity the filament is thrown into a complex coil, and is loaded both with gland cells
and nematocysts of the large type. It is impossible to say, from a study of sections
alone, whether the coiled masses at the bottom of the intermesenterial chambers are
acontia, i.e., free offsets of the edges of the mesentery, or simply coiled filaments.
Gardiner (21) describes acontia in Cceuopsammia, but I am inclined to the opinion
that in Heteroeyathus the structures that might be mistaken for acontia are only
coiled mesenterial filaments.

The filaments of the secondary and tertiary mesenteries do not reach the axial
gastrovascular cavity, and are easily distinguished from the primary filaments. They
are much smaller, especially in the upper part of their course, and have the usual
kidney-shaped outline in section. Like the primary filaments they appear to consist
almost exclusively of ciliated cells in the upper part of their course, but they soon
show glandular cells and large nematocysts (Plate III., fig. 15). At the bottom of
the gastrovascular cavity the filaments of the secondary and tertiary mesenteries are
thickened and coiled like those of the primaries.

The absence of a definite stomodaeum and the extent and importance of the ciliated
bands forming the upper ends of the primary mesenterial filaments are features which,
though peculiar, are readily explained by a consideration of their relations to the

222 . CEYLON PEARL OYSTER REPORT.

palial and columellar structures. As may be seen in fig. III., C, the hammer-shaped
palial jii'ocess at the inner end of the primary septum cuts off the filaments of the
primary mesenteries from the axial space, and in D the palus is seen to have contracted
unions with the columellar pillars, cutting off the interseptal loculi from the axial
space. It is evident that the broad ciliated filaments, extending far down in these
nearly isolated loculi, are the chief if not the only agents in maintaining the circulation
in the deeper parts of the loculi. Fig. III., F, shows how much the interseptal loculi
become narrowed and isolated in the deeper parts of the coral in consequence of the
abundant secondary thickening or stereoplasm.

All the mesenteries are fertile. In female jDolyps the ripe ova are large, and filled
with granules of deutoplasm. I have been unable to make out details to my
satisfaction, but the ova, when young, appear to become enclosed in the mesoglcea,
and as they increase in size they project from the sides of the mesenteries, still
enclosed in a thin mesoglceal envelope, outside of which is a layer of endoderm, forming
a sort of follicle. The ripe ova are pyriform and hang in bunches from the sides of
the mesenteries, each ovum attached by a slender stalk of mesogloea and surrounded
by its follicle of endoderm cells.

The relation of the mesenteries to the septa affords a strong presumption in favour
of the view that these structures have been formed according to Pourtales and
Duerden's law. As has been shown, the septa are alternately exoccelic and
endocoelic. The exoccelic septa, which on the usual system of notation would be
called quaternaries, are larger than the endocoelic tertiaries, aud, as is shown in
fig. III., B and C, their inner ends meet and unite in front of the latter. Further than
this, in each system the exoccelic septa adjacent to the primaries converge and meet
together in front of the secondaries, forming the more or less distinct chevron-shaped
pali described above. This union can best be seen in fig. II. As the tertiary
endocoelic septa are enclosed within the smallest and therefore the most recently
developed mesenterial pairs, and as they are themselves the smallest and least exsert
of all the septa, and only unite with the Y-shaped figures formed by the exoccelic
septa low down in the corallum (fig. III., B, C, and D), the evidence that they are the
most recent in point of formation, aud that they and the mesenterial pairs embracing
them originated between the diverging Y-shaped outer ends of the exoccelic septa, in
the manner described by Pourtales and Duerden, is sufficiently convincing.

The little canals or tubes, running inwards from the lateral walls of the coral and
opening into the Aspidosiphon chamber, have been described and figured by several
authors, but their minute characters have not yet been investigated. They are
almost exactly like the similar tubes in Hetaropsammia, and the transverse section,
Plate IV., fig. 24, of a tube of the latter genus serves equally well for Hetcrocyathus.
The resemblance is the more striking because in Heteroeyatlms, an imperforate
Turbinolid, there are endodermal canals, usually twelve in number, closely attached
to the whole length of the tube, which is itself lined by an invagination of the

SOLITARY CORALS. 223

ectoderm. In Heteropsammia these canals form part of the system of endodermic
canals characteristic of perforate corals, but their presence in Heterocyathus is
remarkable. The transverse section has a curious resemblance to a section through a
young Actinian with twelve mesenteries and a very wide stomodseum. The ectoderm
lining the central tube is curiously modified at its inner end, as will be described
below. The tubes vary in number and position. There may be from five to nine of
them, and they are not, as a rule, in the same plane, but in the majority of specimens
they are more numerous and more closely set together on the side furthest from the
mouth of the Aspidosiphon chamber. Tangential sections of the corallum show that
the tubes are interseptal, and that the stereoplasm filling up the interseptal loculi is
interrupted by their presence. Being interseptal, their external openings are always
between the costse.

Histology. — Though Professor Herdman's specimens are exceptionally well
preserved, histological details are, as is usual in corals, difficult to determine to one's
satisfaction. In what follows I do not profess to give a complete account of the
histology of the different tissues, but will confine myself to such details as I have
been able to make out to my own satisfaction.

The E c t o d e r m . — As is shown in fig. III., the ectoderm of the body-wall is more
or less deeply infolded between the costre in spirit specimens. It is thinner where it
is stretched over the edges of the costse and thicker in the furrows between, and this
does not seem to be due to contraction in spirit and the consequent stretching of the
tissue over the costal edges, but to a differentiation of the ectoderm, which is not
only thicker, but more glandular and more richly provided with nematocysts in
longitudinal stripes, corresponding to the attachments of the mesenteries. In the
thinner stripes of ectoderm corresponding to the costse the tissue consists almost
entirely of columnar or cubical epithelial cells. There are very few gland cells, and
few, if any, nematocysts. In the thicker stripes opposite the attachments of the
mesenteries the epithelial cells are longer, and there are numerous gland cells and
nematocysts of the kind shown in Plate III., fig. 18. The gland cells are of an
elongated goblet form, with a comjjressed nucleus at the base of the goblet and a,
very thin protoplasmic stalk passing from the nucleus to the mesogloea. They
contain a number of yellowish-brown highly refractive granules, which do not stain
with any of the ordinary aniline dyes, or with hematoxylin. Similar gland cells are
very abundant in the ectoderm of the tentacles. The nematocysts, two of which are
shown everted in fig. 16, are elongate oval in shape with a somewhat coarse thread
coiled loosely within. The everted threads are covered with long barbs disjiosed in a
spiral. Before eversion these nematocysts contain a flocculent substance which stains
bright blue in picro-indigo-carmine, a7id therefore probably belongs to the class of
hyalogens, all of which stain similarly with this dye.

The ectoderm lining the outer moieties of the tubes leading into the Aspidosiphon
chamber is of the same character as that of the thickened stripes of the body-wall,

224 CEYLON PEARL OYSTER REPORT.

but it is remarkably modified in tbe inner moieties of tbese tubes. It becomes much
thicker (Plate III., fig. 17) and has a vacuolated appearance, and in place of the
gland cells described above we find elongated pyriform gland cells filled with round
granules, which stain deeply in hematoxylin. But the most striking feature is
afforded by the nematocysts. These are very numerous, closely crowded together,
are ovoid in shape and very large, with a thick thread covered with barbs arranged
in a double spiral. They contain coarse, closely-packed granules, which stain
intensely blue in picro-indigo-carmine. In spite of their much larger size (figs. 16
and 17 are drawn to the same scale) and their different shape, I regard these as
modifications of the ordinary nematocysts of the ectoderm of the body-wall. But
they are certainly very strikingly modified, and I am at a loss to explain their
function, situated as they are at the deeper end of tubes whose function is also
problematical.

The ectoderm of the tentacles is raised into a great number of the well-known
knob-like " batteries," crowded with small nematocysts of the usual form with a
closely coiled spiral thread. Interspersed among these are larger nematocysts
resembling those of the body-wall, but usually much longer and narrower. The
ectoderm of the peristome is extremely thin, and consists of a cubical epithelium with
very few gland cells, and, as far as I could ascertain, no nematocysts. At the lips of
the mouth, however, the ectoderm is thickened, and shows some special features not
visible in the ectoderm of the body-wall. There are very few gland cells, those
which are present containing coarse granules staining blue in picro-indigo-carmine.
I could not detect any nematocysts. Muscle fibres, which were scarcely distinguishable
in the peristome, are here well developed and arranged radially so as to form a more
or less distinct dilator oris muscle. In many sections the epithelio-muscular character
of the ectoderm cells becomes evident. The layer of nerve fibres at and between the
bases of the ectoderm cells is distinct, and among the epithelio-muscular cells very
small attenuated, spindle-shaped cells may be distinguished which may be interpreted
as sense cells. This thickened ectoderm passes insensibly into the endoderm at the
lower margin of the lips, except at the places where the twelve primary mesenteries
reach the mouth. Here it is continued into the very thick ectoderm of the Y- or
T-shaped filaments of these mesenteries. In these filaments, as mentioned above,
the ectoderm consists almost exclusively of very long attenuated, ciliated, epithelial
cells, whose character is sufficiently indicated in fig. 14. There are very few
glandular elements and no nematocysts. The coiled filaments of the bases of the
primary mesenteries, and nearly the whole extent of the filaments of the secondary
and tertiary mesenteries, consist of closely-packed, attenuated, epithelial cells, among
which numerous gland cells and nematocysts are wedged in. The gland cells are of
two kinds : (l) ovoid vesicular cells with clear contents ; (2) pyriform cells containing
coarse granules staining deeply in picro-indigo-carmine or hsematoxylin. The
nematocysts are all of the type shown in fig. 18. They are very large, measuring

SOLITARY CORALS. 225

(V06 millim. in length, and the central thread stains intensely blue in picro-indigo-
carmine. They are clearly of the same type as, but somewhat larger than, the
mesenterial nematocysts described by Gardiner (21) in Ccenopsammia.

As regards the layer of calicoblasts and the desmocytes, they are of the usual
character, and I have nothing to add to what I have published concerning these
structures in a previous paper (7).

The En do derm. — As is usual in corals, the highly vacuolated endoderm cells
are so badly preserved that nothing very definite can be said of their structure.
They differ, however, in different regions of the body. The endoderm cells covering
the muscle-banners and the extrathecal continuations of the mesenteries, and also
those of the tentacular endoderm, are very long and columnar, and are crowded with
Zooxanthellse. As a rule there is a similar modification just within the mesenterial
filaments, this modification being most pronounced in the primary mesenteries.
Elsewhere the endoderm consists of a rather low columnar or cubical epithelium, and
Zooxanthellae are more scarce or, in some places, absent. Gardiner (21) lays great
stress on the absence of glandular elements in the endoderm, but my observations
do not support his conclusions. Glandular elements, it is true, are few or altogether
absent, not only in the elongated endoderm above referred to, but in the whole or
the greater part of the endodermic investment of the mesenteries. But in Hetero-
cyathus the endoderm of the axial gastrovascular cavity, that is to say the tissue
investing the pali and columellar upgrowths, is invested by a moderately thick
cubical endoderm in which there are few Zooxanthellse, but numerous ovoid or bean-
shaped cells containing clear refractive granules which do not stain with any of the
aniline dyes used, nor with hematoxylin. These cells are shown in Plate IV., fig. 19.
It is, of course, possible that they may be modified nematocysts. Similar cells are
found, though not so abundantly, in the endoderm wherever it is opposite a layer of
calicoblast, and this would seem to suggest that the function of the cells in question is
to elaborate material which is passed through the thin mesoglceal lamina to the layer
of calicoblasts and converted into calcareous tissue by the agency of the latter. This
view has a certain probability, because the calicoblasts form so thin a layer, and are
themselves so retrograde in structure that it is difficult to believe that they are the
only agents in the active growth of the corallum. On the other hand, I have
observed fragments of copepods and diatoms in the intermesenterial chambers, at
some distance from the filaments, and as these have evidently been or are being
digested, it is equally possible that these glandular-looking cells in the endoderm may
secrete a digestive fluid. At all events the facts do not warrant so sweeping an
assertion as that of Gardiner, that the endoderm is excretory but not glandular, < ti-
the conclusions as to the homology of the Anthozoan layers that he has founded
on it.

In some, but not in all my series of sections, the endoderm, in addition to the
glandular elements described above, contained a number of large amoeboid cells of

2 g

226 CEYLON PEARL OYSTEE REPORT.

various shape, two of which are shown in fig. 20. They were very abundant in one
series, but entirely absent in another, and in a third they were rare. They are filled
with large refracting granules, which generally stain crimson in picro-indigo-carmine,
but in some cases they stain a deep indigo blue. They can be nothing else than
amoebocytes. and are probably excretory in function. Their abundance in one
specimen and their rarity or absence in others is probably attributable to the different
conditions of nutrition of the polyps in question. It is well known that the
endodenn of Ccelenterates, e.g., of Hydra, presents very different appearances
according as the animal has been recently fed or starved, and I have found the most
diverse appearances, particularly in the matter of the presence or absence of
endoderm cells loaded with granules, in the endoderm of Hydra, according as they
had been fed abundantly or starved.

Finally, it may be noted that at the bottom of the axial cavity there are large
spaces, and frequently there is a single large central space in the axial gastro-
vascular cavity. The endoderm lining these spaces or space is invariably devoid of
Zooxanthellse and gland cells, and has the simple columnar form depicted in fig. 21.

As regards the occurrence of nematocysts in the endoderm, I occasionally found
large nematocysts, of the second type characteristic of the tentacular batteries, in
the tentacular endoderm, but invariably in close proximity to one of the batteries.
I conclude, therefore, that these nematocysts were not formed in the endoderm, but
have been forced through the mesoglcea into the endoderm during the violent
contraction produced by the action of reagents. Elsewhere I could find no trace of
nematocysts of any kind in the endoderm.

2. Heteropsammia michelini, M. Edw. and H.
(Plate IV., figs. 22 to 25.)

The remarkable analogies between this coral and Heterocyathus CBquicostatus have
been commented on by several authors. Both are built up round a gastropod shell
tenanted by an Aspidosiphon, both have an exactly similar spiral Aspidosiphon
chamber in the adult state. In both there is a minute Lamellibrauch commensal
with the Aspidosiphon within the chamber, and in both there is a number of minute
tubes lined by ectoderm leading from the lateral walls of the coral into the chamber
in question. The shape and general appearance of the two corals is closely similar,
but whereas Heterocyathus is always a simple coral, Heteropsammia michelini
generally exhibits two calices produced by fission of the parent calicle, and H. multi-
lobata exhibits several calicles. The anatomy of the last-named species has been
described by Fowler (18), who notes the following features. The external soft
tissues rest on the echinulations of the cienenchyme ; the tentacles are simple, without
nt'inatocyst batteries, and are apparently both exoccelic and endocoelic ; exosepta and
endosepta are present, and certain of the septa fuse centrally, as in Rliodopsammia

SOLITARY CORALS. 227

(Balanophyllia) ; the number of mesenterial pairs is very variable and there are no
directive mesenteries ; the tubes leading into the Aspidosiphon chamber are lined by
ingrowths of the body-wall consisting of ectoderm, mesoglcea, and endoderm. No
account is given of the histology.

IT. mvltilobata is a well-marked colonial form with several calices, whereas
//. michelini usually lias two calices only, and some of the differences between my
observations and Fowler's are no doubt attributable to the difference in habit of the
two species. In older specimens of H. michelini the septal arrangement is sometimes
very irregular, but in young specimens with a single undivided calicle the septa are
arranged in a normal manner in six systems and four complete cycles, and their
relations are practically identical with those described in Balomophyllia, and with a
little trouble one may select a considerable proportion of adult specimens, hi which
two calicles have been formed by fissiparity, in which the septal arrangement differs
very little from the normal.

The most interesting results, however, are obtained from specimens with a single
elongated but as yet undivided calicle. I made a series of sections through one such
specimen and another series through an obviously regular calicle of a specimen in
which fissiparity was complete. The sections show that the septa are alternately
exocoelic and endocoelic. In the first specimen with an undivided calicle there are
two pairs of directive mesenteries defining the primary septa at each end of the long
axis of the calicle. A portion of one side of the specimen was destroyed by the
ravages of a boring sponge, but in the complete half I was able to count three
systems and twenty-six endosepta arranged in four complete cycles, and a single
septum of a fifth cycle in the chambers adjoining the directive septa. I have given a
diagram of this specimen in fig. IV. The diagram is carefully constructed with
reference to a camera drawing of the actual section, but the complexity of the actual
drawing is so great, owing to the porous nature of the corallum, and the details are
60 minute that it is impossible to get them all into any figure of reasonable size other
than a diagram.

It will be observed that the arrangement of the septa and their relations to the
mesenteries are practically identical to what has been described in Heterocyathus.
But particular attention should be paid to the exocoelic septa, which are shaded with
lines to distinguish them from the endosepta. Nothing can be more clear than the
fact that their peripheral ends have been thickened ; that chambers have been formed
in the thickened ends, whereby the septa became Y-shaped. And, finally, that the
quaternary endosepta have been formed between the mesenterial pairs arising in the
forks of the endosepta, and in some cases they have and in some they have not united
with the inner ends of the exosepta. Moreover it is clear, from the manner in which
the inner ends of the exosepta unite in the columella, that the tertiary endocoeles
must have been formed in exactly the same manner as the quaternaries obviously
have been. There could not be a more striking demonstration of the validity of

2 0 2

228

CEYLON PEARL OYSTER REPORT.

' 4

3

JD 1

Fig. IV. Diagram of a specimen of Hdempmmmia michelini in which the calice has not yet divided by
transverse fission. The endosepta are cross-shaded, the exosepta, the porous theca, and the columella
are black. The mesenteries are indicated in the upper half of the figure but not in the lower. D, D,
directive mesenterial pairs. 1, 2, 3, endosepta of the first, second, and third cycles. 4, 4, fourth-
cycle endosepta, each enclosed in a chamber formed by the forked peripheral ends of an exoseptum.
In several cases the inner ends of the exosepta may be seen to unite in front of a tertiary endoseptum
and to be prolonged beyond the jtoint of union into the columella. 5, an endoseptum of the fifth
cycle formed in a bifurcation of an exoseptum adjoining one of the directive endosepta.

Pourtales and Duerden's law in the case of Eupsammid corals. It should be
noted, however, that the quinary mesenterial pairs do not arise according to the
sequence observed by Duerden in Siderastrcea, but are contiguous to the sulcar
(dorsal) and sulcular (ventral) primary septa.

My second series of sections, through a polyp which had been formed by transverse
division from the original single parent polyp, gave some interesting results. A
tracing of a section taken a little above the level of the stomodeeum is given in
fig. 22. The arrangement of the septa and mesenteries is remarkably regular.
There are apparently six systems and three cycles of endosepta, with an additional
quaternary septum in the chamber on the right side of the directive septum, that
is to say, there is just half the number of endosepta that there was in the elongated
and undivided specimen. The endosepta may be classified according to size and
relations as primaries, secondaries, and tertiaries, but they cannot have made their
appearance in this order, as they were derived from the pre-existing septa of the
parent polyp. The number of parental mesenteries has evidently been halved in the
process of division. No new septa have been added, but the quaternaries of the
parent have become the tertiaries of the offspring.

SOLITARY CORALS. 229

Another singular fact is the presence of only one pair of directive mesenteries, the
absence of the second pair being obviously explained by the division of the parent
at right angles to the long axis of the calicle, whereby one pair of directives
remained in the one, and the other pair in the other of the offspring. The
existence of the single pair of directives in a coral which has doubled itself by a
single act of fissiparity confirms the conclusion I arrived at in an earlier paper (5),
that the absence of directives in many corals is to be explained by the fact that they
have multiplied by fission.

To deal briefly with other points in the anatomy of the polyp. The external body
wall, as Fowler describes, rests upon the echinulations of the ccenenchyma, and
there are no peripheral continuations of the mesenteries. The tentacles are all
endoccelic, and therefore correspond in number to the endosepta. In contraction they
are introverted, and in this condition are doubled over the inner edges of the
endosepta. Thus they are situated nearer the mouth than in Heterocyathus, and the
peristomial area is correspondingly reduced. The primary and secondary tentacles
form a circle nearest to the mouth, the remaining orders form circles at greater
distances from it, and it follows from their relations to the endosepta that the
tentacles of different orders alternate with one another. In H. michelini the
tentacles are covered with well-developed batteries of nematocysts ; in H. multilobata,
according to Fowler, they are not.

There is a short, but distinct stomodasum, and I think that I found traces of a
sulcus (gonidial groove), but of this I cannot be certain, as both my specimens
suffered partly from the attacks of boring sjjonges, and partly from the fact that
grains of quartz sand were lodged in the angles of the mouth and produced imper-
fections in my sections in those regions. However this may be, the circulation of
water in the complex chambers in which the lower parts of the mesenteries are
lodged is provided for by means similar to that described in Heterocyathus. The
filaments of the primary and secondary mesenteries form broad sinuous bands in the
upper part of their courses, and it is only deep down in the coral that one meets
with the characteristic kidney-shaped sections of mesenterial filaments. The number
and arrangement of the mesenteries has been sufficiently described. As far as I
could determine, all the mesenteries are fertile, but in the undivided specimen of
which I made sections the quaternary mesenteries, especially those in the lateral
chambers, were in advance of the remaining orders as regards the maturity of ova
contained in them. The ova, and as far as I could make out in a series of loncd-
tudinal sections the testes also, are embedded in the mesoglcea of the mesenteries in
the manner figured by the Hertwigs for Actinia, and do not hang from the sides of
the mesenteries in follicles as in Heterocyathus.

Histology. — The external tissues were very well preserved, but the reagents
had not penetrated well, and the endoderm and mesenterial filaments were in
consequence macerated and of little use for histological examination.

•230 CEYLON PEAEL OYSTER REPORT.

The E c t o d e r m (fig. 23) of the body-wall is relatively much thicker than in
Heteroeyathus, and is richly supplied with gland cells and nematocysts. The
character of the epithelial cells is well shown in the figure. The gland cells are of
three kinds: — (l) Goblet cells with a central nucleus and filiform internal ends, the
wider external end filled with small granules staining deeply in hematoxylin or
picro-indigo-carmine. (2) Flask-shaped cells with broad internal ends and a narrow
neck opening at the surface ; these cells are filled with coarse granules which stain in
the same way as those of the first variety. (3) A few large vacuolated sac-shaped
cells whose contents stain with eosin, but remain colourless in picro-indigo-carmine.

The nematocysts are of two kinds. The small spiral-thread variety is fairly
abundant, and here and thei*e are large torpedo-shaped nematocysts with barbed
threads, similar to those described for Heteroeyathus.

The ectoderm lining the small canals leading into the Aspidosiphon chamber is not
modified as in Heteroeyathus, but contrariwise, it loses its glandular character and
consists almost exclusively of ciliated columnar cells, among which nematocysts of the
larger variety with barbed threads are to be found. In some of my sections the
lumina of the tubes are packed with everted nematocysts of this kind, which have
evidently been discharged when the polyp was killed, and this suggests that these
canals serve in some manner as special batteries of nematocysts protecting the
commensal Sipunculid. In fig. 25 I have given a representation of the manner in
which these chamber canals pass through the general mass of ccenenchymal canals into
the Aspidosiphon chamber. It will be observed that the endoderm canals, generally
twelve or thirteen in number, surrounding them are specially related to the chamber
canals and lie parallel to them.

The tentacular ectoderm is raised into large and broad nematocyst batteries
containing numerous small spiral nematocysts and a lesser number of those of the
larger type. In the peristomial region the ectoderm becomes thinner, less glandular,
and contains but few nematocysts. At the lips it again becomes thick, and in this
region the nervous layer is thicker than elsewhere, and I was able to observe a few
large pale nuclei embedded in the layer of nerve fibrils which appear to belong to
gland cells. In this region, as in Heteroeyathus, the radially disposed muscular fibres
of the ectoderm cells are very conspicuous, and one can equally well recognise the
circular layer of endodermic fibres forming a sphincter oris.

The ectoderm of the lips passes without any abrupt change into the stomodaeum.
Here the same elements may be recognised, but in different proportions. The
epithelial cells are elongated filiform, and their deeply staining elongated nuclei are
closely crowded together. Nematocysts of the spiral and barbed types are fairly
abundant, especially the latter variety. The glandular cells are much increased in
number, especially the eosinophile cells of the third type. The layer of nerve fibrils
at the bases of the cells is relatively thick. This structure is continued without much
change into the broad filaments of the primary and secondary mesenteries, but in the

SOLITARY CORALS. 231

lower part of their courses the spiral nematocysts and the basophile gland cells become
less numerous, and the number of eosinophil gland cells is correspondingly increased.
The coiled mesenterial filaments were too much macerated to admit of accurate
observation. 1 was only able to satisfy myself that they contain numerous large
barbed nematocysts, but none of the spiral variety; that the epithelial cells are
attenuated and closely packed together in groups, and that there are numerous very
large eosinophile gland cells between the groups.

The endoderm was too much macerated to admit of careful study. The cells are
evidently highly vacuolated, and their nuclei are unusually small. Zooxanthellee are
abundant in the most external ccenenchymal canals, but are scantily distributed in
the deeper parts. In one of my specimens the endoderm was full of granular
amcebocytes similar to those described in Heterovyathus.

3. Dendrophyllia gracilis, M. Edw. and Haime.
(Plate IV., figs. 26 to 28.)

The anatomy and histology of D. ramea has been fully described by von Heider (26),
whose observations are so careful and accurate that I have little to add to them.
He has given a very full account of the relations of the perisarc (edge-zone or
" Randplatte"), and has fully realized the importance of the exosepta and the
manner in which the tertiary mesenterial pairs, and the exosepta embraced by them,
are formed in the Y-shaped peripheral extremities of these exosepta. The full
significance of this observation of course escaped him, as he was unacquainted with
Pourtales' account of the development of Balanophyllia, and his memoir was many
years anterior to Duerden's recent work.

Dendrophyllia is an arborescent genus propagating by lateral buds. The soft
tissues extend for a considerable distance below the lip of the calicle, forming a well-
marked edge-zone or perisarc. The septa and theca are thin and fragile, and there
are well-marked external costee corresponding to the septa. Correlated with the
presence of costse is the existence of peripheral continuations of the mesenteries in
the perisarc, as has been correctly described and figured by von Heider, There is
practically no difference between the septal arrangement of D. gracilis and D. ramea.
In both there are three cycles of endosepta, and the exosepta alternate with the
endosepta, forming an apparent quaternary cycle. Von Heider has given an
excellent account of the manner in which the apparent quaternary exosepta meet
and unite in front of the short tertiary endosepta, and his diagram (loc. tit.,
plate xxxi., fig. 7) leaves nothing to be desired in clearness and accuracy. It
is evident that the septa and mesenteries are formed in strict accordance with
Pourtales' and Duerden's law, and I am able to add this much in confirmation. I
made a series of sections through a small lateral bud measuring about 2 millims. in
diameter. In this bud there are only twelve pairs of mesenteries. Six pairs, of

232 CEYLON PEARL OYSTER REPORT.

which two are directives, are well developed and reach the stomodseum ; they embrace
the primary endosepta, the latter being tolerably well developed as narrow, ridge-like
projections in the calicle, but with branched peripheral ends passing into the network
of trabecular forming the coenenchyme. Alternating with the endosepta are six
exosepta of approximately the same size, whose peripheral ends bifurcate and enclose
chambers in which are the mesenterial pairs of the second cycle. These secondary
mesenteries are unequally developed, but not, as far as I could determine, according
to any regular sequence. Each pair embraces an endoseptum, which in some cases is
very rudimentary, but in other cases has grown centipetally and has united with the
inner limb of the Y-shaped exoseptum (fig. V). Thus we see that in the bud the

Fig. V. Semi-diagrammatic section through a young bud of Dendrophyllia gracilis. 1, 2, the primary and

secondary cycles of endosepta embraced by the corresponding mesenterial pairs. 1>, l>, the two pairs
of directive mesenteries, ex, an exoseptum with forked peripheral ends, within which a pair of
secondary mesenteries enclosing a secondary endoseptum is developed. /', t2, tentacles of the first
and second cycles corresponding to the endosepta.

secondary mesenteries, and the endosepta embraced by them, are formed between
the bifurcated outer ends of the exosepta, and show the same relations to the latter
that the tertiary endosepta show in the adult. This is a clear proof of the nature ol
the septal sequence.

SOLITARY CORALS. 233

Vox H eider counted eighteen tentacles in one-half of a specimen of D. ramea.
My specimens of D. gracilis were too much contracted to enable me to count with
certainty, but I found, both in the bud and in the adult, indications of a cycle of
exotentacles alternating with the primary and secondary endotentacles. The exo-
tentacles appear to be simply introverted, and do not enter into close relations
with the mesenteries. The endotentacles, as in Heteropsammia, are introverted and
doubled over the inner edges of the endosepta. As far as I could determine, there
ai*e twelve exotentacles in the adult, alternating with the primary and secondary
endotentacles. Thus there would be thirty-six tentacles of both kinds, and my
observations agree with von Heider's.

The stomodajum is relatively longer than in Heteropsammia, measuring fully
1 millim. in length in the contracted spirit specimen, and in the expanded polyp it
is probably much longer. I could find no trace of sulcus or sulculus. The stomodseal
ectoderm extends some little way down the free edges of the primary and secondary
mesenteries, as in Heterocyathus and Heteropsammia, but not so far as in these two
genera, and eventually it gives place to a normal mesenterial filament.

As regards the histology, I have not very much to add to what has been published
by vox Heider. In D. gracilis the ectoderm of the body-wall is peculiar, and
unlike anything that I have seen in any other coral. It is difficult to obtain a clear
idea of its structure in sections, but it appears, as shown in fig. 26, that the cells are
large and vacuolated, and form a thickened cell-wall, which, from its staining
properties, seems to be of the nature of an intercellular substance. The walls of
adjacent cells cohering together give a semi-cartilaginous consistency to the whole
tissue, which no doubt forms an efficient protection against the numerous sponges
and other organisms that infest the majority of corals. Embedded in the ectoderm
cells are nematocysts of three kinds : ( 1 ) the common spiral-thread nematocyst
measuring 0"02 millim. in length, more or less ; these are very scantily distributed.

(2) Medium-sized elongated nematocysts, about 0'028 millim. long and 0"005 millim.
broad (fig. 27b) with a loosely and irregularly coiled thread ; these are abundant.

(3) Large elongate oval nematocysts about 0'035 millim. long and O'Ol millim. broad.
According to Gardiner (21) the medium-sized nematocysts are developing
stages of the spiral 0-02-millim. nematocysts, and this may be the case. It is not
easy to speak with certainty on this point from a study of sections only, but there is
some reason to think that the medium-sized nematocyst, as I have figured it, is
really a third variety. In the first place they are very abundant in the ectoderm of
the body-wall, whereas the small spiral nematocysts are very scantily represented
there ; in the second place they occur in the stomodaeum in which no other
nematocysts are to be found. Von Heider found two kinds of nematocysts in the
ectoderm of D. ramea, the small spiral 0'02-millim. variety, and the large elongate
oval variety, measuring in this case 0-05 millim. in length. He describes the latter
as tilled with a coarsely granular material, but without a thread, and a large

2 H

234 CEYLON PEARL OYSTER REPORT.

proportion of these nematocysts in D. gracilis present this appearance ; but others
may be found in which a thick thread loosely wound in a spiral of few turns is
clearly distinguishable, and it is scarcely open to doubt that the forms filled with
granular material are simply unripe. In fig. 27c I have drawn a nematocyst of this
variety which seems to throw some light on the development of the thread. In the
centre of the capsule is the pointed end of the thread (much longer than in
Coenopsammia) surrounded by a granular sheath which seems to be differentiating to
form the eversible portion of the capsule. At the base of the capsule the granular
sheath widens out and forms a rounded mass in which an irregularly and imperfectly
formed coil can be distinguished : this I take to be the middle portion of the thread
in course of formation. Near the point of the thread, and outside the granular
sheath surrounding it, is a spiral of five turns closely wound round the granular
sheath : this must be the terminal portion of the thread differentiated from the
granular sheath at the same time that the latter gave rise to the eversible portion
of the capsule. The capsule itself is lined by a rather thick granular layer. It
should be observed that in these nematocysts the terminal part of the thread is
wound round the eversible sheath near the aperture for the extrusion of the latter.
Gardtner (loc. cit., plate xxxiv., fig. 14) figures it at the opposite end of the
capsule, but in Coenopsammia the eversible sheath is only one-third the length of the
capsule. In DendrophylUa, moreover (von Heider agreeing with me in this), there
is a distinct spearhead-shaped tip to the thread, as in Euphyllia, whereas Gardiner
found no such armature in Coenopsammia.

In the "batteries" of the tentacles there are, as usual, very numerous and closely
crowded nematocysts of the small spiral variety, and among them a considerable
number of the second variety described above, which, if Gardiner is right, are to be
regarded as early stages of the spiral variety. I think, however, that they are really
a different form of nematocyst. I was able to distinguish a fine fibril passing inwards
from the bases of many of the small spiral nematocysts, and in some cases I could
observe that this fibril passed into a fine layer of protoplasm surrounding an oval
nucleus, and that from this a fine branching fibril passed into the layer of nerve fibres
overlying the mesoglcea (fig. 27a). The ectoderm of the peristome is very thick as
compared with Heteropsammia and Heterocyathus, and contains a large proportion of
gland cells and the same nematocysts as the ectoderm of the body wall. The
stomodaeal ectoderm, as von Heider remarked, is composed almost exclusively of
elongated cells of the columnar type ; they are almost certainly ciliated, but the cilia
had been destroyed by the action of alcohol. There are few gland cells, those that
are present being of the flask-shaped finely granular type, in the stomodeeum and very
few nematocysts. All the nematocysts that I was able to recognise belong to the
second variety described above.

The mesenterial filaments are crowded with gland cells and nematocysts, the latter
all of the same medium-sized variety as those of the stomodseum, and this fact leads

SOLITARY CORALS. 235

me to think that Gardiner was in error in describing this form as a stage in the
development of the small spiral variety. The mesenterial filaments are relatively
small, and the absence of the largest ectodermal nematocysts is explained by the fact
that there is no room for bodies of such size.

The endoderm is remarkable chiefly for the fact that Zooxanthellse are very
sparingly distributed in the mesenterial epithelium, and the epithelio-muscular cells
are for the most part clearly defined and but little vacuolated. The muscular processes
of the cells are remarkably well developed. Scattered through the endoderm, but
somewhat sparingly, are minute nematocysts, about 0-01 millim. long, containing a
loosely and irregularly coiled spiral thread. Each is contained in a transparent cuticle,
and there is a flattened nucleus to one side of and outside the nematocyst itself. In
the somewhat similar endodermic nematocysts of Flabellum, Gardiner (22) figures
the nucleus inside the capsule of the nematocyst. The mesogloea of D. gracilis
is relatively thick, no doubt in correlation with the unusually great development of
the musculature, but I could find no trace of structure in it. The desmocytes,
described by von Heider as calicoblasts. are well developed and form tassel-like
groups at the points of attachment of the mesenteries to the theca, but otherwise
present the usual features.

It is evident that the three forms whose anatomy is described in this memoir do not
differ in any important points of anatomical or histological structure from the normal
Actinian type, which has been shown by many authors to be characteristic of the
polyps of Madreporarian corals. There are, indeed, minor characters, both anatomical
and histological, which have a certain interest, but none of them can be regarded as
having any classificatory value. One's attention is arrested by the presence of both
endotentacles and exotentacles in some genera and of endotentacles only in others.
But when we see that in the Eupsammiidge, a well-defined family, only endotentacles
exist in Heteropsrxmmia while Dendrophyllia possesses both endo- and exotentacles,
this character does not appear to be of much value. Moreover, owing to the great
difficulty experienced in counting and localising the tentacles in spirit specimens, the
information we possess on this point is not altogether trustworthy, and before any
attempt is made to use the tentacles as an aid in determining the affinities of different
genera of corals, it will be necessary to accumulate a large number of facts based on
the study of living or well-preserved expanded polyps.

But a much more promising field is offered to the future investigator by the study
of the relations of the septa to the mesenteries, especially by the developmental
sequence of the endosepta and their connections with the exosepta, if present. My
chief object in this paper has been to show that it is possible, by a study of the
relations of hard and soft parts in the adult corals, to determine whether the septal
sequence follows the rule established for Balanophyttia by Pourtales and Siderastrcea
by Duerden, and I have given sufficient evidence to show that the peculiar septal

2 H 2

236 CEYLON PEARL OYSTER REPORT.

arrangement of the Eupsammiidee, which has been commented upon by many authors,
is due to the sequence in question. I have further shown that the septal arrangement
in Heteroeyathus, an imperforate coral, can only be interpreted upon the same
principle, and Duerden's account of the development of Siderastrcea is of itself
sufficient evidence that a similar mode of septal sequence prevails both in perforate
and imperforate corals.

On the other hand, it is clear from von Koch's (32) and de Lacaze Duthier's (33)
account of the development of Caryophyllia, that Pourtales and Duerden's rule is by
no means applicable to all corals, but that, contrariwise, there is another and distinct
mode of septal sequence which is either independent of the formation of exosepta, or
in which the part played by the exosepta is very different.

It is tempting to suppose that there are two principal modes of septal development
among corals, typified by the Eupsammiidse, Heteroeyathus and Siderastrcea on the one
hand, and Caryoj)hyUia on the other hand, and that the presence or absence of
exosepta will indicate the group to which any given coral belongs. But while I think
that it is very possible that a solution of many classificatory difficulties may be found
by extensive investigations on the lines laid down by Duerden, I must recognise
that it would be premature to make any positive statements in the present state of
our knowledge. There can, I think, be very little doubt that forms like Mussa,
Euphyllia, and Galaxea, which have endosepta only, cannot have been developed on
the same plan as Siderastrcea, but must have followed the mode described by
von Koch for Caryophyllia. But this statement cannot be extended to all forms
known to possess endosepta only. Gardiner (22) has shown that in Flabellum the
new cycles of septa are at first exoccelic, and that in the course of growth a pair of
mesenteries is formed in connection with each exoseptum, in consequence of which
the exosepta become endosepta. I have shown (6) that the same rule holds good
for the anthoblasts of Fungia. Though our evidence is not complete upon this point,
I believe that this will be found to be the usual mode of septal formation in all corals
which have only endosepta in the adult condition. It must be borne in mind that a
coral in the course of growth increases not only in height but in diameter. As a
consequence of its constantly expanding perimeter, the peripheral ends of the original
radial structures (the primary and secondary septa) become further and further
removed from one another, and the polyp forms new radial folds between which new
radial calcareous structures are formed to fill up the gaps between the diverging
septal extremities. These new radial structures are the exosepta or their equivalents.
In those cases in which the exoseptal formations acquire a very intimate union at the
time of their appearance with the adjacent endosepta, further peripheral growth is
accompanied by a bifurcation or splitting of the peripheral ends of the exosepta, and
a newer set of radial folds with corresponding calcareous structures is formed within
their bifurcated extremities, as in the Eupsammiidse, Siderastrcea, &c. In other cases
further peripheral growth is provided for by the formation of new radial folds on

SOLITARY CORALS. 237

cither side of instead of between the forked peripheral ends of the exosepta, and
these latter, as a consequence of the formation of new mesenterial pairs embracing
them, become in their turn endosepta. But, whichever the mode of peripheral
growth, it is obvious that the so-called " theca," about which so much discussion has
taken place, is really the result of the formation of radial structures in connection
with peripheral growth, and is not a circumferential structure. An examination of
the figure in the text, fig. IV., will make this clear. Here the different stages of
peripheral growth are clearly marked by the different distances from the centre of
the insertions of the primary, secondary, tertiary, and quaternary mesenteries. The
wall or theca must at each stage have been formed by the trabecular offsets of the
peripheral ends of the exosepta united to the endosepta, and it is clear that the
perforate " theca " of the adult is nothing more than a network of trabeculee formed
by the peripheral ends of the exosepta and endosepta. In Heterocyathus the wall is
formed by the intervention of the synapticula, which must be regarded as being
formed simply by very short radial folds between the outer ends of the septa, and are
therefore quasi-septal structures, as has been pointed out for Fungia by Miss Ogilvie.
In many corals the wall is formed by simple apposition of the thickened peripheral
ends of the septa, giving rise to the so-called " pseudotheca." In others it is formed
by the so-called "euthecal" pieces, which are really nothing more than very short
radial structures intervening between the peripheral ends of the septa, but not
produced internally so as to form septa. In the present state of our knowledge it
would be unprofitable to pursue the subject further, but I may repeat that the whole
tendency of recent investigation has been to show that a true theca, that is to say, a
circumferential structure independent of the radial growths, or septa, is always to be
referred to the prototheca, as defined by Bernard, and that all other so-called
"thecal" structures are in reality radial growths, formed in radial folds of the
polypal walls, between previously existing radial structures. Further, it would seem
from the most recent embryological investigations, which are confirmed by a study of
adult structure, that there are two ways in which the new radial growth may take
place : (1) by the simple formation of new radial folds between every pre-existing
radial fold (Caryophyllia, &c); (2) by the bifurcation of every alternate radial fold
and the formation of a new radial fold in every space formed by such bifurcation
(Siderastrcea, BalanophyUia, Heterocyathus, &c).

It is possible, but it would be rash to make a positive assertion at present, that
these two modes of peripheral growth will be found to be of primary value to the
systematist.

It may lie worth while, in conclusion, to observe that the following corals have
been found to have endosepta only : — Mussa, Euphyllia, Madrepora durvillei,
Madracis aspenda, Duncania, Galaxea, the lateral polyps of Madrepora variabilis.
The following have been found to possess both exosepta and endosepta : — Astroides,
Stephanotrochus, BalanophyUia, Hetei-opsammia, Dendrophylha, Heterocyathus,

238 CEYLON PEARL OYSTER REPORT.

Madrepora aspera, the apical polyps of Madrepora variabilis, Lophohelia, Seriato-
pora, Amphihelia, Stephanophyllia, Stephanaria, Sphenotrochus, Caryophyllia,
Trochocyatlms, Bathyactis.

In Flabellum and Fungia there are both endosepta and exosepta, but the latter
eventually become endosepta on the formation of mesenterial pairs embracing them.
The list is incomplete, but it suffices to show that little reliance can be placed on the
mere presence or absence of exosepta. On the other hand, I believe that it is of
importance to discover in every coral whether the new endosepta are formed in the
forked peripheral extremities of the exosepta, or simply by the formation of a new
pair of mesenteries in connection with a radial upgrowth.

SOLITARY CORALS. 239

LIST OF LITERATURE.

(1.) Alcock. — "Newly Recorded Corals from Indian Seas." 'Jour. Asiatic Soc. Bengal,' lxii., p. 138.
(2.) Alcock. — "Natural History Notes H.M. Indian Marine Survey Steamer 'Investigator,'" ser ii., 9.

'Jour. Asiatic. Soc. Bengal,' lxii., p. 169.
(3.) Ai.COCK. — 'Deep Sea Madreporaria of the Siboga Expedition.' Leiden, 1902.

(4.) Bernard, H. M.— " The Prototheca of the Madreporaria." ' Ann. Mag. Nat. Hist.,' ser. 7, xiii. 1904.
(5.) Bourne, G. C. — "On the Anatomy of Mussa and Euphyllia." 'Quart. Jour. Micr. Sci.,' N.S.,

xxviii. 1888.
(6.) Bourne, G. C. — "The Post-embryonic Development of Fungia." 'Trans. Roy. Dublin Soc.,'

ser. 2, v. 1893.
(7.) Bourne, G. C. — " On the Structure and Formation of the Calcareous Skeleton of the Anthozoa."

' Quart. Jour. Micr. Sci.,' N.S., xli.
(8.) Dana, J. D.— ' Zoophytes. U.S. Exploring Expedition.' 1846.

(9.) Duerden, J. E. — " The Morphology of the Madreporaria. V. Septal sequence." ' Biol. Bulletin,'
vii. 1904.
(10.) Duerden, J. E. — ' The Coral Siderastrwa radians and its Post-larval Development.' Washington, 1904.
(11.) Duncan, P. M.—" Madreporaria of the Deep Sea." 'Trans. Zool. Soc.,' viii. 1871.
(12.) Duncan, P. M.— "Deep-sea and Littoral Corals." ' Proc. Zool. Soc. Lond.' 1876.
(13.) Duncan, P. M. — "Revision of the Madreporaria." ' Journ. Linnean Soc.,' xviii., Zoology. 1885.
(14.) Fowler, G. H. — " The Anatomy of the Madreporaria. I. Flabelhim, Mhodopsammia." ' Quart.

Jour. Micr. Sci.,' N.S., xxv., p. 577. 1885.
(15.) Fowler, G. H. — "The Anatomy of the Madreporaria. II. Madrepora durvillei and M. aspera."

'Quart. Jour. Micr. Sci.,' N.S., xxvii., p. 1. 1887.
(16.) Fowler, G. H. — "The Anatomy of the Madreporaria. III. Twbinaria, Lophohelia, Seriat<rpora,

Pocillopora." 'Quart. Jour. Micr. Sci.,' N.S., xxviii., p. 17. 1888.
(17.) Fowler, G. H. — " The Anatomy of the Madreporaria. IV. Madracis, Jmphihelia, Stephmophyllia,
Sphenotrochus, Stephanaria., Pocillopora nobilis, Seriatopora tenuicornis." ' Quart. Jour. Micr. Sci.,'
N.S., xxviii., p. 413. 1888.
(18.) Fowler, G. H. — "The Anatomy of the Madreporaria. V. Duneunia., Madrepora, Galaxea, Hete.ro-

psammia, Bathyactis." 'Quart. Jour. Micr. Sci.,' N.S., xxx., p. 405. 1890.
(19.) Frech, F.— " Die Korallenfauna der Trias." ' Paheontographica,' vol. 37. 1890-1891.
(20.) Gardiner, J. S.— "On the Solitary Corals collected by Dr. A. Willey." 'Willey's Zool.

Results.' 1899.
(21.) Gardiner, J. S. — " On the Anatomy of a Supposed New Species of CWnopsammia from Lifu."

' Willey's Zool. Results,' p. 357. 1899.
(22.) Gardiner, J. S. — " South African Corals of the Genus Flabellum." ' Mar. Investig. in S. Africa.'

II., Cape Town, 1904.
(23.) Gardiner, J. S.— " The Turbinolid Corals of South Africa." ' Mar. Investig. in S. Africa.' III.,

Cape Town, 1904.
(24.)* Gardiner, J. S. — 'The Fauna and Geography of the Maldive and Laccadive Archipelagoes,'

vol. ii., Suppl. i., " Madreporaria," Parts iii. and iv.
(25.) Goldfuss.— 'Petref. Germ.,' i. 1826.
(26.) Heider, A. R. von. — " Korallenstudien : Adroules calycularis u. Bendrophyllia ramea." ' Arbeit a. d.

Zool. Inst, zu Graz,' i., p. 153. 1886.
(27.) Hertwig, O. and R. — 'Die Actinien.' Jena, 1879.

* This memoir did not come into my hands till my MS. had been sent to the press.

240 CEYLON PEARL OYSTER REPORT.

(28.) Koch, G. von. — "Uber die Entwicklung des Kalkskelettes von Astroides calyailaris." 'Mitth. a. d.

Zool. Stat. Neapel,' iii. 1882.
(29.) Koch, G. von. — "Mitth. fiber die Structur von Pholidophyllum loveni, &c." 'Pakeontographica,'

vol. 28, p. 216, 1882.
(30.) KOCH, G. VON.— " Uber Flahellwm." ' Morph. Jahrb.,' xiv. 1888.

(31.) Koch, G. von. — " Das Skelett der Steinkorallen." ' Festschrift fiir Gegenbauer,' ii. 1895.
(32.) Koch, G. von. — "Entwicklung von Caryophyllia ryatlms." 'Mitth. a. d. Zool. Stat. Neapel,' xii.,

p. 755. 1897.
(33.) Lacaze Duthiers, H. de. — "Faune du Golfe du Lion." 'Arch, de Zool. exper. et g^n.,' 3e se>., v.

1897.
(34.) Michelin.— ' Mag. de Zool. V. Zoophytes.' 1843.

(35.) Milne-Edwards, H. and Haime, J. — " Mon. des Turbinolides." ' Ann. d. Sci. Nat.,' 3C ser., ix. 1848.
/36.) Milne-Edwards, H. and Haime, J. — " Mon. des Eupsammides." ' Ann. d. Sci. Nat.,' 3e ser., x. 1848.
(37.) Milne-Edwards, H. and Haime, J.—" Mon. des Fongides." ' Ann. des Sci. Nat.,' 3e ser., xv. 1851.
(38.) Milne-Edwards, H. and Haime, J. — ' Histoire naturelle des Coralliaires.' 3 vols. Paris, 1857-1860.
(39.) Moseley, H. N.— "Deep-Sea Corals." '" Challenger " Reports. Zoology, II.' 1881.
(40.) Ogilyie, M. M. — 'Die Korallen der Stramberger Schichten. Palseont. Stud. ii. d. Jura- u. Kreide-

Formation der Karpathen, Alpen u. Appeninen.' Stuttgart, 1897.
(41.) Ogilvie, M. M.—" Microscopic and Systematic Study of Madreporarian Types of Corals." 'Phil.

Trans.,' vol. 187.
(42.) Ortmann, A. — " Beobachtungen an Steinkorallen von der Siidkiiste Ceylons." 'Zool. Jahrb.,'

Band. iv. SuppL, p. 493. 1889.
(43.) Pourtales, L. F. de. — 'Illus. Cat. Mus. Comp. Anat. Harvard.' Cambr., 1871.
(44.) Pratz, E. — " Uber die verwandtschaftlichen Beziehungen einiger Korrallengattungen, mit haupt-

saehlicher Beriicksichtigung ihrer Septalstructur." ' Pakeontographica,' vol. 29.
(45.) Quelch, J. J.— " Reef Corals." '" Challenger " Reports. Zoology, XVI.' 1886.
(46.) Quoy and Gaimard.—' Voyage de 1'" Astrolabe." ' 1834.
(47.) Rehberg. — " Neue unci wenig bekannte Korallen." ' Abh. Ver. Hamburg,' xii.
(48.) Reuss, A. E. — "Uber einige Anthozoen a. d. Tertiarschichten des Mainzer Beckens." 'Sitz. d.

kais. Akad. d. Wiss. in Wien.,' xxxv. 1859.
(49.) Semper, C. — " Generationsvvechsel bei Steinkorallen." 'Zeit. fiir wiss. Zool.,' xxii. 1872.
(50.) Studer, T. — "Ubersicht der Steinkorallen (Eupsammina u. Turbinaria), welche auf der Reise S.M.S.

'Gazelle' um die Erde gesammelt wurden." 'Monatsb. k. preuss. Akad. Wiss. Berlin.' 1877.
(51.) Waagex, W— "Salt Range Fossils." ' Palaeontographica Indica,' ser. 13, vol. i. 1887.

SOLITARY CORALS. 241

EXPLANATION OF PLATES I., II., III., and IV.

Fig. 1. Rhorluri/athxs o'l/Iimrnsis, n. sp. Lateral view of the corallum. x 2.
„ ].\. „ Diagram illustrating the arrangement of the septa. 1, 1, primary

septa. 2, secondary septa. 3, 4, 5, tertiary, quaternary and quinary septa.
„ 2. Cyathotrochus herdmani, n. sp. A view of the corallum showing the paliform lobes, p, and one of

the carinated directive cost*, c.
,, 2A. Gyathotrochus herdmani, n. sp. Calice from above. The numerals indicate the cycles of septa.
„ 3. Flabellwm erassum, M. Edw. and H. Profile view of the corallum. b.s., basal scar. ( x 2.)
„ 3.V. „ „ The calice viewed from above. The numerals indicate the

different cycles of septa.
,, 4. Flabellum rubrum, Quoy and Gaimard, var. profumdum, M. Edw. and Haime. The calice viewed

from above, showing the different cycles of septa 1 to 5, and the parietal columella formed by

spiny outgrowths from the lower ends of the septa.
,, 5. Plttro/iYirliiis Urns, M. Edwards and Haime. The calice viewed from above, showing the regular

arrangement of the septa and the essential lamellar columella.
,, 6. Bcdanophyllia parallda, Semper. Profile view of the corallum. x 2.
„ Ga. ,, The calice viewed from above.

„ 7. Balanophyllia rumiwji, M. Edwards and Haime. View of a colonial aggregate of corallitcs.

Nat. size.
,, 7a. Balanophyllia cumingi. Diagram of the septal arrangement. The numerals indicate the cycles

of septa.
„ 8. Balanophyllia socialix, Sempeu. Profile view of the corallum.
„ 8a. „ Diagram of the septal arrangement.

,, 9. Balanophyllia taprobana; n. sp. Profile view of the corallum, showing the enlarged primary costse

and one of the lateral buds, b.
,, 9a. Balanophyllia taprobance, n. sp. Calice viewed from above. 1, 1, the enlarged primary entocoelic

septa. X.2, the secondary exoccelic septum of the left-hand lateral system. A'.3, Ar.3, tertiary

exoccelic septa of the same system. En.'2, secondary entoccelic septum; and En.3, tertiary

entocoelic septum of the same system.
„ 10. Lobopsammia robusta, n. sp. Lateral view of the colony.
,, 10a. „ ,, Diagram of the septal arrangement. The numerals indicate the

cycles of septa.
,, 11. Dendrophyllia minuscula, n. sp. Profile view of the colony, x 2.
,, 11a. ,, „ A calice viewed from above. 1, 1, primary entoccelic septa.

En.2, secondary entoccelic septum. A'.'-, A'.2, secondary exoccelic septa (largely magnified).
„ 12. Side view of a primary septum of Heterocyathus xquicostatus, showing the septal trabecular

emerging on the surface as spiniform granulations. A.c, Aspidosiphon chamber, p., pali and

paliform lobes, sir., stereoplasm represented by cross shading, syn., synapticulum.
,, 13. Portion of a horizontal section through the corallum of Heterocyathus mquicostatus, showing the

diverging trabecule of which the septa are composed and the pigmented stereoplasm largely

filling up the interseptal loculi. S., S., septa, sir., stereoplasm external to, and sir.1, stereo-
plasm internal to syn., the synapticula.
,, 14. Section through a band-shaped primary mesenterial filament of Heterocyathus, some little way

below the lip of the mouth, c.e., ciliated epithelio-muscular cells of the filament, en.,

endoderm. mg., mesoglcea. Zeiss. Obj. D, Oc. 4.
,, 15. Section through a convoluted mesenterial filament of Heterocyathus. en., endoderm. gl1., gl2.,

two varieties of gland cells. »., nematocyst. Zeiss. Obj. D, Oc. 4.

2 I

242 CEYLON PEARL OYSTEE REPORT.

Fig. 16. Section through the ectoderm of the body-wall of Heterocyatkus. gl., gland cells, mg., mesoglcea.

n., n., nematocysts. nv., nervous layer. Zeiss. Imm. TV. Comp. Oc. 4.
„ 17. Section through the inner end of one of the canals leading into the Aspidosiphon chamber in

Heterocyatkus. gl., gland cells. n., «., ovoid nematocysts with granular contents. Zeiss.

Imm. jlj. Comp. Oc. 4.
„ 18. A nematocyst O06 millim. in length from the convoluted mesenterial filaments of Heterocyatkus.
,, 19. Endoderm covering the septa of Heterocyatkus showing gl., glandular looking cells, which may be

degenerate nematocysts. An isolated cell is shown to the right of the figure, zo., zooxan-

thellae. Zeiss. Imm. yV. Comp. Oc. 4.
„ 20. Endoderm of Heterocyatkus showing am., amoeboid cells full of granules, cal., calicoblasts.

Zeiss. Imm. TV. Comp. Oc. 4.
„ 21. Columnar endoderm from the axial gastro-vascular cavity of Heterocyatkus.
„ 22. Semi-diagrammatic representation of a horizontal section through Heteropsammia micJwlini a

little above the stomocheum. The corallum is black. The mesenteries, peristomial, tentacular

and external ectoderm are represented by red lines. 1, 2, 3, the several cycles of endosepta.

4, a single endoseptum of the 4th cycle. D., the single pair of directive mesenteries. t.\ t?,

t.3, the several orders of tentacles, introverted and doubled over the endosepta.
,, 23. Ectoderm of the body-wall of Heteropsammia michelini. en., endoderm. gl.l, goblet-shaped

gland cells with granular contents, gl.l, flask-shaped gland cell, mg., mesoglcea. n.\ spiral

nematocysts. n.2, large nematocyst. zo., zooxanthellre. Zeiss. Imm. TV. Comp. Oc. 4.
,, 24. Transverse section through a canal leading into the A sjridusiphon chamber in Heteropsammia.

The histological details are indicated diagrammatically. ec, ectoderm lining the canal.

en.c, endoderm canals surrounding the central ectodermic canal, mg., mesoglcea. Zeiss D.

Oc. 2.
,, 25. Modified ectoderm of the body-wall of Dendrqphyllia gracilis, showing the thickened walls of the

ectodermic epithelium, mg., mesoglcea. n2, medium sized nematocysts, 0'028 millim. in

length n3, large nematocysts, 0-035 millim. in length. Zeiss. Imm. -jV-. Comp. Oc. 4.
„ 26. Nematocysts of D. gracilis. Zeiss. Imm. TV. Comp. Oc. 4.

„ 26a. A spiral 0-02-millim. tentacular nematocyst, its inner end prolonged into a fine nerve fibril.
,, 26b. Nematocyst from the stomodajum or mesenterial filament.
,, 26c. Developing nematocyst from the external ectoderm (see text).
„ 26d. A ripe nematocyst from the external ectoderm.
„ 26e. The same diagrammatically represented.
„ 27. Four endoderm cells from a mesentery of D. gracilis, showing n, an endodermic nematocyst.

m.p., muscular processes. Zeiss. 1mm. tV. Comp. Oc. 4.

CEYLON PEARL OYSTER REPORT.

SOLITARY CORALS- PLATE I.

6s

6a.

4656 .

CEYLON PEARL OYSTER REPORT.

SOLITARY CORALS -PLATE II

FIG. 7.

FIG. 7a.

' -4.

FIG.IIa.

;

-JJm

FIG. 8.

FIG.IOa.

FIG. 10.

4

\ -

*m

FIG. 9.

FIG. II.

E .Wilson ■ ■

CEYLON PEARL OYSTER REPORT.

SOLITARY CORALS -PLATE III.

fig. 15.

FIG. 17.

CEYLON PEARL OYSTER REPORT

SOLITARY CORALS -PLATE IV.

*J

\#

\ Q*>

- ' % t)

Fig. 19

3

i

•-

Fig. 24.

my.

:9l>?

gt'

71.3

Fig. 2 5.

; I"% I ft N «,

ft

i

*

®

Fig. 23.

J«'

26a.

26b.

26c.

26d.

26e.

E. Wilson, Cambridge-

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

REPORT

ON THE

POLYCH^ETA

COLLECTED BY

Professor HERDMAN, at CEYLON, in 1902.

BY

ARTHUR WILLEY, F.R.S.,

DIRECTOR OF THE COLOMBO MUSEUM.*

[With EIC4HT PLATES.]

Thts collection of Polychseta is the most extensive which has been brought together
from the coast of Ceylon. There are only three older collections of any magnitude
from Ceylon upon which reports have been published, namely, that of Schmarda,
worked out by the traveller himself and published in 1861 (' Neue Wirbellose
Thiere'); a small series gathered by Mr. Holdsworth and described by Grube in
1874 ('P. Zool. Soc, London'); lastly, another small series collected by Dr. Hans
Driesch and described by Michaelsen in 1892 ('J. B. Hamb. Anst.,' ix. 2). Of
these older collections the most important was the Schmarda collection, which
included the discovery of the remarkable genera Gastrolepidia and Bhawania, the
latter not being represented in Professor Herdman's collection.

The material was handed over to me at Professor Herdman's suggestion by
Mr. James Hornell in September, 1904. A further consignment which had been

* Editorial Note. — In order to save delay, I have, with Dr. Willey's consent, undertaken to see
this Report through the press without sending proofs to Ceylon. I have to thank Mr. Arnold Watson
and Mr. Cyril Crossland, both of whom are familiar with the group, for their kindness in reading the
proofs along with me. The only change of any importance we have had to make is the name of the new
genus on p. 251. Dr. Willey had proposed Hulolepulia, but we regarded this as being practically the same
as Moore's name Hololepida ('Proc. Acad. N.S. Philad.,' 1905) applied to an allied but distinct Polychaite,
so we altered the title of Willey's genus to Ilolulqndella. — W. A. Herdman.

2 I 2

244 CEYLON PEARL OYSTER REPORT.

taken to England, consisting mostly, but not entirely, of duplicates of some of the
species contained in the first set, was sent out to me by Professor Herdman. These
had also passed through the hands of Mr. Arnold T. Watson, who kindly forwarded
to me such notes and drawings as he had made, some of the latter being reproduced
on Plate VIII. A species of Polydora which attacks the pearl oyster was subsequently
forwarded by Mr. Horn ell.

Some new facts of systematic importance relating to previously described species
are recoi'ded here. Among the species described as new are some of considerable
interest, e.g., Autolytus orientalis, Branchiomma quadrioculatum, Ceratonereis
falcaria, Grymcea cespitosa, Halosydna zeylanica, Leprea inversa, Paramarphysa
orientalis, Serpula watsoni, TJialenessa stylolepis.

The occurrence of Onupliis conchylega and the recovery of Harmothoe dictyophora
are also noteworthy features of the collection.

Family: AMPHINOMID^E.

Chloeia flava (Pallas) — Plate I., figs. 1 and 2.

Chloeia ceylonica, Grube, 1874, "Ann. Ceylon," 'P. Zool. Soc,,' p. 326.
Chloeia flava (Sav.), Grube, 'Ann. Semp.,' 1878, p. 10.

Four rather small specimens from Station LVL, Dutch Modragam Paar, 9 fathoms ;
average length about 28 millims., with 24 setigerous segments. Another smaller
specimen is also in the collection.

Examples of dorsal and ventral bayonet setae, the former barbed, are shown in
Plate I., figs. 1 and 2. The branchife commence on the fourth segment ; mediad of
the violet tinted dorsal cirrus of the first four segments there is an accessory cirrus,
that on the fourth segment being very slender. The caruncle of this well-known
species is attached to the first two segments, and is produced backwards over the
next two ; it consists of an upper and lower series of lamellae, those of each series
united together in a zigzag manner by their lower ends ; those of the upper series are
further united together in couples along the crest of the caruncle.

The Amphinomidfe are not, I believe, as a rule, rapacious Annelids, but swallow
sand and small stones, &c. Probably they are preyed upon by the Aphroditidoe, since
the small specimen (No. 64) was penetrated by some of the enormously long barbed
spines of Hermione.

A small worm taken in 24 fathoms, at Station LXIIL, west of Periya Paar, seems
to be a form of the same species ; it is 9 millims. long, 20 segments. The tentaculum
impar is nearly as long as the caruncle, and, like the dorsal cirri and the stems of the
gills, deep purplish crimson in colour. No colour marks were observed along the
hack, and the crimson cirri stood out prominently from the midst of the very long and
delicate setae. Branchias commence as usual on the fourth segment ; accessory dorsal
cirrus observed on the first three segments only, as described by Grube ; the folds of

POLYCHiETA. 245

the caruncle are more open, less closely set than in older worms. No barbed seta?
were found. Unlike Hesione ceylonica (q. v.), young individuals of Chloeia Jtava do
not possess the full number of segments characteristic of the adult.

Eurythoe complanata (Pallas).

Several examples of this species, whose distribution coincides with that of coral
reefs, were obtained from Aripu Reef and at Galle, ranging in length up to
200 millima and in breadth up to 17 millims. over the setae. Branchiae commence on
the second setigerous segment and there is one dorsal cirrus to each parapodium.
The caruncle is inserted into the first three setigerous segments.

Eurythoe latissima (Schmarda, op. cit., p. 141) is a synonym of this species.

Eurythoe longicirra (Schmarda).

(See Schmarda, 'Neue Wirbellose Thiere,' ii., 1861, p. 142.)
Whether or not this is an exceptional form of E. complanata I am unable to
decide, but it is certain that the caruncle is inserted upon the first four setigerous
segments and overlaps the fifth.* The branchiae commence on the second setigerous
segment, and there is one dorsal cirrus to each parapodium.

Length 40 millims., width 5-5 millims. One specimen, from Aripu Reef.

Family: APHRODITID^E.

Hermione malleata, Grube — Plate I., figs. 3 and 4.

Hermione malleata, Grube, 'Ann. Semp.,' 1878, p. 17.

Hermione ridgewayi, HOKNELL, 'Ceylon Pearl Oyst. Rep.,' Part I., 1903, pp. 16 and 74.

Two specimens from old Dutch Modragam Paar, 9 fathoms, Station LVI. ; one
specimen from Aripu Reef, 18th March, 1902.

The malleiform processes described by Grube are clearly not definite morphological
structures, but merely dermal folds associated with the elytrophores and branchial
tubercles, the latter being transversely elongate.

This species is the Oriental form representing the Mediterranean species Hermione
hystrix. I have satisfied myself on this point by actual comparison of the material
from Ceylon with specimens of II. hystrix procured from the Stazione Zoologica at
Naples for the special purpose of this investigation. The only serious divergence in
Grube's description relates to the paljas, which he describes as being smooth. His
sjiecimen only measured 16 millims. in length, and the palps, when examined under
low magnification by transmitted light, appear smooth, though in reality they are
beset with minute papilla?. The Neapolitan specimens of H. hystrix are larger and
darker than the Ceylon worms ; the elytra thicker and more opaque.

A complete Ceylon worm measures 30 millims. long by a width of 12 millims. over

* For further remarks on this point see Ehlers, ' Florida- Anneliden,' 1887, p. 30.

246 CEYLON PEARL OYSTER REPORT.

the neuropodia ; fifteen pairs of elytra covering the back ; length of dorsal glochideal
spines up to 10 millims. (Plate I., fig. 3). Eye-peduncles rounded. Cirrophores and
neuropodia thickly covered with minute rounded papillae ; neuropodia 3 millims. long.
Ventral surface beset with similar papillae. Tentaculum impar, tentacular cirri,
dorsal cirri and first ventral cirrus with distal clavate tip articulated to the main
shaft, which is expanded at this point. Palps 13 millims long, beset with six
longitudinal rows of minute spiniform papilla?. The elytral segments carry on each
side a nabellum of curved smooth-tipped setae radiating dorsad and a backwardly
directed fascicle of long brown glochideal setae ; below the dorsal setae there is a tuft
of fine silken threads. The neuropodia carry furcate setae with or without an
accessory tooth (Plate I., fig. 4). The portion of the notopodium from which the
flabellum arises is adnate to the elytrophore.

Pontogenia indica, Grube — Plate L, fig. 5.

Grube, 'Ann. Semp.,' 1878, p. 18; Herdman, "Palmyra ami/era," 'Ceylon Pearl Oyster
Report,' Part I., Narrative, 1903, p. 75.

The close resemblance between Pontogenia indica and Palmyra aurifera (see
Grube, op. cit., pp. 13-14) is one of the remarkable facts of Polychaet taxonomy.

The specimen was obtained from a living coral block in 6^ fathoms, one mile north
of Muttuvaratu Paar, Station LIX. It measures 20 millims. in length and has
45 segments. The head is retracted within the anterior segments and beneath the
anterior elytra, the second pair of elytra overlapping the ommatophores. The
ceratophore is marked off from the frontal border of the prostomium and the
ommatophores extend to this level. The palps are beset with longitudinal rows of
delicate recurved papillae. By pressing the ceratophore back a tuberculum faciale
with granulated surface may be seen extending from the prostomium to the anterior
border of the mouth. Each ommatophore carries two eyes, but as the pigmented
areas overlap there appears to be only one eye in certain lights (Plate I., fig. 5).
The paleae which form flabella projecting over the elytra are utrinquedentate, as
described by Grube. The dorsal setae which occur in addition to the paleae are long,
delicate, colourless and numerous, and constitute a tela tomentosa over the elytra,
concealing the latter, but not felted together so as to form an inextricable tangle.
The two rows of distantly placed denticulations with their points directed towards
the apices of the paleae are not always visible in one view, and the paleae then appear
to be denticulate along one side only.

Family: POLYNOID^E.
Iphione muricata, Savigny — Plate I., fig. 6.

Polynoe peronea, Schmarda, 1861 ; ' Neue Wirbellose Thiere,' ii., p. 157.
Locality : — Gulf of Manaar. One specimen of the typical yellowish brown colour ;
length 17 '5 millims., breadth over the setae 8'5 millims. The head is withdrawn

POLYCH^ETA. 247

between the anterior segments to such an extent that four pairs of elytra had to be
removed in order to expose it (Plate I., fig. 6). The antennae converge from the
anterior pinnacles of the head towards the middle line, and then run side by side
close together ; one of them may sink down at a lower level, and then one only
remains in view. They are subglabrous, being very sparsely and minutely papillose.
Schmarda observed one antenna only, which he described as a tentaculum impar
erroneously. There is no doubt as to the identification of the present specimen with
Schmarda's species ; the only question is whether it is co-specific with Savigny's
I. muricata (Grube, ' Ann. Semp.,' 1878, p. 21).

An elytron from the mid-body shows a concave anterior border, a large gold-
coloured outer surface and a smaller pale inner portion directed obliquely forwards
and overlapped by neighbouring scales. The whole surface of the elytron is divided
up into polygonal, mostly hexagonal areas, and these again into numerous secondary
areoles. The main areolation resembles that of Harmothoe dictyophora, but differs in
some details ; the areas are largest on the inner (mesial) portion of the scale, smaller
along the outer and posterior borders ; the secondary areolation is highly charac-
teristic and was noted by Schmarda. Focussing through the superficial secondary
areolation, another reticulum, which may be called the interstitial reticulum, comes
into view, the meshes of which do not coincide with the former. The interstitial
mesh-work shows nodal dilatations with a refringent body in each, like a nucleus,
very clear in caustic potash. Proceeding to the pigmented portion of the scale, the
surface enclosed by the secondary areoles becomes elevated to form low papillae,
which, near the outer border, assume the form of inclined spines. Posteriorly, certain
of the primary areas become elevated to form long stout aculeate spines terminating
distally in a pair of prominent horns and covered by numerous secondary spines, as
figured by Schmarda. The secondary spines become larger distally, so that the
main spine sometimes presents a more or less trifid appearance at the summit. The
dorsal setae are excessively numerous and bipinnate. The ventral setae show a well-
marked subterminal dilatation, a verticillate tract, and a smooth curved apical
portion.

As for Iphione muricata, Grube makes no mention of the secondary areolation of
the elytra, a highly remarkable feature ; and he states that the outer and posterior
margin of the scale is fimbriated, which is not so in /. peronea. Across the major
diameter of the scale I counted about 30 primary meshes, and across one of the latter
about nine sharply defined secondary areoles. There are 29 segments, 13 pairs of
elytra. The anus is dorsal and is bordered by the last pair of elytra. Schmarda
says the last segment has " zwei kleine Endfortsatze," but I cannot see them. The
dorsal and ventral cirri, the palps and tentacular cirri are papillose. The tentacular
cirri are concealed below the palps in dorsal view.

A smaller specimen, 5 millims. wide, shows a wide membranous fringe round the
outer and posterior border, not yet areolated ; there are no fimbriae ; about 1 7

248 CEYLON PEARL OYSTER REPORT.

primary meshes across the scale. The antenna? adhered together and broke away
from their peduncles.

Lepidonotus carinulatus, Grube (1869) — Plate L, figs. 7 to 11.
(See Gkube, 'Ann. Semp.,' 1887, p. 26.)

Localities : — South-west Cheval Paar, one specimen, broken in half, and several
specimens mostly fragmentary; Chilaw Paar, Station LXIX., one specimen.

The antennae and dorsal cirri show a very slight distal dilatation followed by a
nagelliform terminal process. The cirri and palps are smooth ; dorsal seta? more slender
than ventral, 20 to 30 in number, disposed in three concentric arcs, the seta? of the
dorsal arc being shorter than the rest. Ventral setse about 25, bidentate and fringed
(Plate!, fig. 11).

Elytral papilla? carinulate and spheroidal on the surface, echinulate and stellate
near the fimbriated border ; in some elytra the echinulate papilla? extend over the
surface to the region of the scar. Elytra deciduous and body fragile ; twelve pairs of
elytra. Exserted pharynx with fringe of nine dorsal and nine ventral marginal
papilla?. Some elytra become narrower towards the inner side than at the outer side,
others are nearly equally wide throughout, with concave anterior border. Patches of
dark brown pigment are scattered over the surface. The fimbria? of the outer border
are densely placed, those at the posterior border are sparser. Some scales are much
less papillose than others ; in a highly tuberculate scale the carinulate papilla? occupy
the anterior and inner (mesial) portions ; the echinulate papilla? occur at the posterior
border, extending thence over the scar, this region being somewhat elevated ; the
echinulate papilla? near the outer border are smaller and less hirsute than those of
the posterior border, sometimes presenting a more or less stellate appearance (Plate I.,
fig. 10). The prostomium (cephalic lobes) with eyes and antennary bases (cerato-
phores) is shown on Plate I., fig. 7.

Observations on a larger specimen, in which some of the elytra were better
preserved in position, show that the dominant macroscopic character of the species
lies in the difference between the elytra of the anterior region and those of the
middle and posterior regions. The anterior scales (only those of the fourth and fifth
segments are present in the specimen) are much smaller than the rest, they are
placed subtransversely and their surface appears verrucose under low magnification.
The more posterior scales show a finely granulose surface, they are about twice the
size of the anterior scales (excluding the scales of the first pair, which are generally
small and round, and are absent from this specimen), and they are longitudinally
elongate, the anterior end narrower. All scales show a large opaque whitish patch
in the region of the scar, and all are fimbriated externally. The length of the
specimen is 17 millims. ; total width over the seta? 5 millims. Along the posterior
border of the elytra there are small saucer-shaped elevations which appear to be the
bases of deciduous fimbria?.

POLYCHyETA. 249

Microscopic examination shows that the verrucose appearance of the anterior scales
is due to the presence of large numhers of spheroidal echinulate papillae (Plate I.,
fig. 9). These papilla? do not occur on the posterior scales, which owe their granulose
appearance to the ordinary carinulate papilla? (Plate I., fig. 8), small stellate papilla?
and smooth globoidal papilla?.

The elytra of this species show an analogy with those of Marenzeller's Lepido-
notus pleiolepis from Japan, which, however, possesses fifteen pairs instead of the
usual twelve pairs. Some individuals show a narrow black ring round the lower part
of the subtermmal dilatation of the dorsal cirri. In a specimen with extruded
proboscis the outline of the prostomium was nearly circular.

This species appears to be the most abundant and typical representative of the
genus Lepidonotus on the Ceylon pearl banks.

Lepidonotus cristatus, Grube.

Grube, 'Ann. Semp.,' 1878, p. 27 ; Gravier, 'Ann. Mer Rouge,' 1901, p. 212.

A. fine example of this species, 43 "5 millims. in length, 17 millims. wide over the
seta?, was taken on the Galle Reef under a boulder, 7th June, 1902.

The smooth-bordered elytra show a large tumid bilobed transverse crest. The
ventral seta? have the usual laciniate fringes on the region of the subtermmal
dilatation and end in a smooth curved tip. The distal portions of the antenna?,
tentacular cirri, dorsal cirri and first ventral cirrus, up to the subtermmal bulb, are
coloured black.

Gravier describes for the first time the modified ventral seta? of the second
segment, which are characterised by the possession of a very long verticillate tract,

Lepidonotus trissochsetus, Grube.

Grube, 'Ann. Roth. Meer.' (Ehrenberg coll.), 1869; 'Ann. Semp.,' 1878, p. 25.

Locality: — South-east Cheval Paar. Length of the specimen 12 millims., breadth
over the seta? 5 millims.

This species is distinguished by the possession of two kinds of seta? in the dorsal
ramus of the parapodium, and these seta? are very numerous. There are rather short,
stout, transversely spinulose seta? of a common type, and enclosed by these are numerous
fine smooth capillary seta? ending in a point, a short distance below which there is a
delicate dilatation, as in a spear-head without barbs ; these may be called hastate
seta?. In the anterior half of the body the hastate seta? .do not project beyond the
spinulose seta? which surround them, and cannot therefore be seen without adopting
special measures. In the posterior half the hastate seta? project far beyond the short
spinulose seta?, which occur like a sheath at the base of the bundle ; they project here
as far outwards as the ventral seta?. The resulting difference between the anterior
and posterior dorsal fascicles as seen under low magnification is very pronounced. It

2 K

250 CEYLON PEARL OYSTER REPORT.

is, however, probably due to a temporary protrusion of the setae, but it may serve to
account for the alleged difference between L. trissochatus, Gr., and L. indicus,
Kinberg. The hastate setae can clearly be retracted and protruded at will. In one
segment they are projecting on one side, retracted on the other. The elytra are
sparsely covered with smooth obtuse pustules of varying sizes, the smaller and more
numerous occupying a submarginal position. The margin of the elytron is quite
smooth.

As indicated by Grube in 1869, this species presents a cryptocephalous condition,
as in the later described species L. cryptocephalus, the head being concealed below a
projecting collar formed by the second segment. The dorsal cirri are smooth, with a
terminal flagellum and a subterminal swelling. The palps are beset with conical
papillae, and terminate in a smooth attenuate extremity (seen in one of the palps
only). The ventral setae have simple hamulate extremities (not bidentate) with the
usual subterminal fringes.

Halosydna zeylanica, n. sp. — Plate I., figs. 12 and 13.

Total length about 15 millims., width 2 millims. to 2 -5 millims. Commensal on
Astropecten, Ceylon seas.

Body much flattened, extremely fragile, elytra smooth, colourless, covering the
dorsum, leaving parapodia exposed, lightly attached to the elytrophores. The
prostomium is divided by a shallow linear groove into two halves, upon which no
eyes were observed, and the bases of antennae and tentaculum are somewhat concealed
below the frontal border of the head, though arising at one level, as seen in frontal
view. The number of segments may be as many as 50, but this will depend upon age
and variation, and the same applies to the number of the elytra, the highest number
observed being 24 pairs.

The chief character is presented by the distribution of the elytra on the following
segments :— II, IV., V., VII., IX., XI, XIII, XV, XVII, XIX, XXI, XXIII,
XXVI, XXIX., XXXII, XXXIII, XXXV, XXXVII, XXXIX., XII, XIIII,
XIV., XLVII, XIIX. The peculiarity here is the occurrence of successive elytra on
segments XXXII and XXXIII. a point which I have verified on three specimens.
In the specimen of which the elytral formula is given above, the elytra were lost
from a number of the posterior segments but present on No. XIIX.

The dorsal ramus of the parapodium is very small, and the setae which issue from it
are few in number, plain and hyaline. Foreign particles adhere to all the setae
individually throughout their exposed portions. The ventral setae have the form
shown in Plate I, fig. 13, a plain apex slightly curved, and at some distance removed
from it a dilatation with a projecting semilunar cusp. The number of ventral setae
varies, but is about 12.

The ventral cirri are shorter than the ventral ramus, arising from a prominent base.
The dorsal cirri are plain structures occurring on the segmenta nuda. arising from a

POLYCH^ETA. 251

projecting base which lies obliquely outwards and backwards when seen from above.
The dorsal cirri are highly deciduous. They exceed the length of the parapodium
(Plate I., fig. 12). The dorsal ramus and elytrophore of a foot from an elytra-bearing
segment appear to carry long vibratile cilia. No tubercles were observed on the
segmenta nuda. The ventral ramus is richly furnished with decussating muscles.
The elytra are orbicular, glabrous, with few " veins " ramifying out from the scar.

Harmothoe dictyophora (Grube) — -Plate I., figs. 14 to 16.
Polynoe dictyophorus, Grube, ' Ann. Semp.,' 1878, p. 44.

Locality : — East side of Cheval Paar. Length 12 millims., breadth over the setse
475 millims.

Prostomium normal, anterior eyes placed in the centre of the lateral border
(Plate I., fig. 15). Antennae ciliate ; palps beset with numerous minute blunt
papilla? ; dorsal cirri densely ciliate, the long filiform papillae ceasing at the base of
the terminal filament ; no dilatation below the terminal filament. Dorsal setae
numerous, verticillate spinulose, but the whorls are not complete, only occupying
three parts of the circumference of the setae ; the shorter setae of the dorsal bundle
are distinctly stouter than the ventral setae ; ventral setae without exception con-
spicuously bidentate, dilated at a point variously remote from the apex and spinulose
thereafter (Plate I., fig. 16). The spinulose tract of the superior ventral setae is much
longer than that of the inferior.

Thirty-five segments ; fifteen pairs of elytra covering the back. Exposed portions
of the elytra divided into polygonal areoles carrying chitinous spines and filiform
papillae ; some of the areoles are densely pigmented dark brown. Some of the spines
are bifurcated, such spines being particularly prominent on the round elytra of the
first pair. The outer fimbriae of the elytra are longest and densest, these are followed
posteriorly by a group of papillae with globular tips, and these again by shorter
filiform papilla? (Plate I., fig. 14). It should be noted further that the ventral cirri
also carry short blunt papillae, scattered and not very numerous.

Grube founded the species upon a single elytron.

Hololepidella, n. gen.

A Polynoid ; antennae arising at a lower level than the tentaculum impar ; segments
and elytra numerous.

Hololepidella commensalis, n. sp. — Plate I., figs. 17 to 20.

Station I., off Negombo, 12 to 20 fathoms, on Clypeaster humilis.
This species appears to be allied to Polynoe venosa, Grube (' Ann. Semp.,' p. 43),
in which however there were only 18 pairs of elytra and 42 segments in the single

2 K 2

252 CEYLON PEAEL OYSTER REPORT.

individual examined. The number 18 is a generic character of Acanthicolepis, McInt.
(= Dasylepis, Mgn.), but it seems likely, though not certain, that a polymeric
Polvnoid with more than 18 pairs of elytra will pass through a stage with 18 only.
At any rate, Polynoe venosa, Gr., does not belong to the genus Acanthicolepis.

Those polymeric Polynoidse in which the paired antennae arise at a lower level than
the tentaculum impar appear to be inadequately classified. Malmgren's genera
Nemidia and Enipo are at most only sub-genera of Schmarda's Hemilepidia.

All Polynoidae (s. str.) in which the antennae arise at a lower level than the
tentaculum are placed in the sub-family Harmothoina. The genus Polynoe (s. str.) is
polymeric, having more than 45 segments in the fully formed condition. It comprises
two sections: — (1.) Hemilepidia, Schmarda, in which there are 15 pairs of elytra
restricted to the anterior region of the body. This section comprises the following
species : — Polynoe scolopendrina, Hemilepidia erythrotcBnia, Nemidia torelli, and
Enipo kinbergi. (2.) Hololepidella, n.g., in which the elytra are not so restricted.
Very possibly Grube's Polynoe venosa belongs here in spite of its few segments. The
species now described also belongs to this section.

In Hololepidella commensalis the small antennae clearly arise at a lower level than
the tentaculum (Plate I., fig. 20). The elytra are pale, delicate, translucent, smooth,
large, orbicular, covering the back ; the first pair with central insertion, the others
with excentric insertion near the anterior margin becoming submarginal in the
posterior region ; there is an indication of nervures radiating out from the scar of
insertion, as in P. venosa. The elytra are inserted upon segments II., IV., V., VII., IX.
* * * * * XXL, XXIII. , XXVI, XXIX., XXXI. , XXXIV, XXXVI, XXXVIII.,
XL., XLIL, XLV. ; segment XLIV. has an elytrophore on the left side, a branchial
tubercle and cirrophore on the right side. There are 46 segments present, incomplete
behind ; in a fragment of the posterior end the alternation of cirrophores and
elytrophores, the former sometimes skipping one segment, sometimes two, seems to be
continued to the end of the body, but it is not always easy in this region to distinguish
between a cirrophore and an elytrophore.

Head hexagonal ; ground colour of body nearly black (becoming brown after a
length of time in spirit) with a pale ridge across each segment between elytrophores
and branchial tubercles respectively. Palps, antennae, tentacle and cirri smooth ; first
ventral cirrus long, the rest short, with swollen basal portion and terminal flagellum.
Dorsal cirri dark brown, tapering to a pale blunt point which may be slightly swollen,
but without subterminal dilatation. The setae are pale ; dorsal setae, about 8, broad
parce-serrulate, much shorter than the ventral setae ; superior ventral setae obliquely
laciniate towards apex, giving the appearance of alternate serrulations, inferior ventral
setae with subterminal dilatation, curved simple tip and normal fringes (Plate I.,
figs. 17 to 19).

This species is one of those which easily undergo fragmentation, so that the length
cannot be given ; the width is 2 millims.

P0LYC1LETA. '253

Gastrolepidia clavigera, Schmarda.

(SCHMARDA, 1861, 'Neue Wirhellose Thiere,' ii., p. 159.)

Gastrolepidia arablyphyllus, Grube, 'Ann. Semp.,' 1878, p. 46.

Gastrolepidia clavigera, Herdman, ' Ceylon Pearl Oyster Report,' Part I., 1903, pp. 29, 79.

Localities :— Station XVIII. , off" Bameswaram, 7 to 8 fathoms; Station LXL, oft
Periya Paar, 12 to 14 fathoms.

The lateral portions of the sterna of the segments are produced into conspicuous
imbricating lamellae, arching helow the bases of the parapodia. They are semilunar
folds, hut if a segment is detached from the body and looked at from behind, they
appear to be subcordate, as figured by Schmarda. The dorsal elytra are inserted
upon segments II., IV., V., VII. **##*## XXIII., XXVI. , XXIX,
XXXIL, XXXV., XXXVI, XXXVIII. , XXXIX, XLL, XLIIL, XLV., this enumeration
being based upon one specimen from which all the elytra had fallen off'; the other
specimens were smaller and showed fewer elytrophores, so that the peculiar distribution
of the posterior elytra could not be confirmed. Dorsal setae curved, stout, shorter
than the ventral seta?, with as many as forty serrulations along the convex border ;
superior ventral seta? with not less than fifteen laciniate fringes ; inferior ventral setae
stronger than the rest, with curved simple tips, subterminal dilatations and about
eight fringes. The antenna? are inserted at a slightly lower level than the tentaculum.
Dorsal cirri clavate, mostly lost. Beyond the club there is a terminal flagellum shown
by Guube, omitted by Schmarda.

Family: ACOETID^.
Panthalis, Kinberg.

This is a genus of Acoetidae characterised by the reversed imbrication ot the
anterior scales.

The family to which it belongs agrees with the Aphroditidae (s.str.) in having
pedunculate eyes ; with the Sigalionidaa in the polymeric body ; with Polynoidae in
having only setae simplices ; with Iphionidae in having serrated jaws. It is clearly a
family of composite affinities and its members possess an extraordinary interest,
regarded from bioiiomical and anatomical points of view. They manufacture a felted
tube woven from chitinous silken fibres which appear to be homologous with the
fibres composing the dorsal felt of Aphrodite, that is to say, modified setas which issue
from the dorsal ramus of the parapodium and are contained within a convoluted sac
intruding into the body cavity. The tube-forming habits of Panthalis cerstedi have
been admirably described by Mr. Arnold T. Watson in the ' Transactions of the
Liverpool Biological Society,' vol. ix., 1895, pp. 169 to 188.

Two species of Panthalis were described by Grube from the Philippines, and very
fortunately both of these are represented in Professor Herdman's material from
Ceylon, so that I am able to supplement Grube's diagnoses with additional facts and
figures.

254 CEYLON PEARL OYSTER REPORT.

Panthalis melanonotus, Grube — Plate I., figs. 21 to 27.

Grube, ' Ann. Semp.,' 1878, p. 48 ; also Part I., ' Ceylon Pearl Oyster Report,' p. 71.

Locality : — Station LIV., south of Rameswaram, 40 fathoms. One specimen.

A typical cirrus-bearing parapodium from the anterior region has the following
structure : — (i.) A short dorsal cirrus consisting of a stout basal portion and an
acuminate apical portion ; (ii.) the dorsal ramus supported by a flexible acicula and
penetrated by the tomentose seta? which are employed in weaving the tube ; (iii.) the
ventral ramus with four kinds of setae, dorsally a small bundle of very fine serrulate
seta?, then a bundle of long delicate penicillate setse, then a central group of stout so-
called aristate setse ; lastly, a ventral bundle of long spinulose setae (Plate II., figs. 24
to 27). At the tip of the aristate setae there is an alveolus. From the 28th segment
the peduncle upon which the dorsal cirrus is inserted acquires a geniculate appendix,
containing a prolongation of the internal tissues, possibly a diverticulum of the gut,
but of this I cannot be certain without transverse sections. A few segments farther
back an ovate or subcylindrical appendix appears on the squamiferous segments, in
the same position as that occupied by the cirriform branchia of the Sigalionidae, also
containing the same kind of tissue as that which occurs in the geniculate process
above described. The surface of these processes is not ciliated (Plate I., figs. 22
and 23).

Panthalis melanonotus appears to be an Indo-Pacific representative of the Mediter-
ranean and Atlantic P. oerstedi. The accounts given of the eyes of the latter are
rather puzzling. Von Marenzeller* denies the existence of eyes, although the
ocular peduncles are present. Mr. Watson, in the article referred to above, has
omitted to note the presence or absence of black pigment, t stating that the sight of
the animal is good, each reddish-coloured eyestalk being faced with a clear lens, and
having at its tip a rounded papillated appearance.

In Panthalis melanonotus the stalked eyes are very large and provided with
abundant black pigment, in the centre of which is a lens ; the eyes are subspherical,
occupying the extremity of the clavate peduncles. In addition there is, as described
by Grube, an eye-spot on each side of the prostomium behind the ocular peduncle ;
between and slightly behind the sessile eye-spots the median tentacle arises. The
antennae are small and largely concealed by the eye-stalks and were not seen by
Grube (Plate I., fig. 21).

The extruded pharynx bears distally thirteen papillae above and the same number
below, forming a terminal crown ; the median dorsal papilla is enlarged and the
median ventral is lost from the specimen. The palps show pigment spots and appear
glabrous with a simple lens, although minute papillae can be found with the microscope,
especially towards the apex, but there is no marked distinction between apical and

* V. Marenzeller, 'Polychseten des Grundes,' Vienna, 1893, p. 28.

t Mr. Watson informs me that no black pigment is visible in the eyes of the specimens of P. cerstedi
now in his possession. — W. A. Herdman.

POLYCH^TA. 255

*

basal portions, such as occurs in the next species. The body is incomplete behind,
upwards of fifty segments being preserved ; total width of anterior region 6 millims.
The specimen is well preserved.

Panthalis nigroinaculata, Grube — Plate I., figs. 28 to 32.
(Grube, 'Ann. Semp.,' 1878, p. 50.)

Locality : — Station II., North of Negombo, 9 fathoms. One specimen.

Antennae arising from the frontal margin of the prostomium ; the tentaculum impar
of about the same length, from the occipital margin. Only one pair of eyes observed
on the slightly protuberant lateral borders. Palps short and stout, the anterior half
densely fimbriate. Dorsal cirri short, ventral cirri subarticulate at tip. Each foot
(after the first few pairs) is provided with a silk-producing gland which is apt to
remain iu the body after the extraction of the parapodium, the silken strands being
drawn through the orifice of the notopodium.

The anterior elytra are inversely imbricate, inserted upon the elytrophores near
the posterior border ; the third elytra touch in the middle line behind the head
(Plate I., figs. 28 to 31). The elytra were observed on segments II., IV., V., VII.
and all subsequent odd segments up to XXXI., after which the body was quite
flaccid and collapsed, full of the gold-coloured glistening silken coils and firmly
attached to the tube at the posterior end.

The anterior portion of the tube was mostly free from the body of the worm,
thougb slightly connected therewith by loose silken strands. The elytra are sub-
translucent, colourless. The extruded proboscis carries a crown of marginal papillae,
eleven above, thirteen below ; the median dorsal papilla is barely larger than the
others. The setae differ from those of P. melanonotus considerably. The dorsal
group of setae in the ventral ramus consists of small spinulose setae, the spinules
arranged in three remote whorls, below which the seta is slightly dilated ; next a few
fringed setae ; then the strong aristate setae in two groups above and below the
acicula ; finally the ventral group of spinose setae (Plate I., fig. 32). There are no
penicillate setae.

Length of anterior portion, including the proboscis, 13 millims. ; width about
3 millims. Length of the posterior flaccid portion about 13 millims.

Family : SIGALIONID.E.
Psammolyce zeylanica, n. sp. — Plates I. and II., figs. 33 to 43.

Locality : — Ceylon Seas. One specimen, incomplete.

A fragment comprising about thirty setigerous segments, length 17 millims., body
width 5 millims., 8 millims. over the setae. The dorsum is completely encrusted with
fine particles, chiefly calcareous debris, leaving the setae exposed at the sides. The
elytra are thin, transparent and triangular, the base of the triangle directed forwards.
The basal area is clear, but the hinder area bears a luxurious growth of club-shaped

256 CEYLON PEARL OYSTER REPORT.

and capitate papillae to which the foreign particles adhere. The anterior pair of
elytra probably meet or overlap, but in general the elytra are confined to the sides of
the body, leaving the mid-dorsum free. Nevertheless the latter is covered with sand
grains which adhere to dermal capitate papillae (Plate I., fig. 33).

The prostomium, carrying two pairs of eyes, is overshadowed by the massive
ceratophore. There are no paired antennae on the prostomium, these having become
adnate to the cirrophores of the buccal segment which carry the tentacular cirri and
bundles of plumose setae (Plate I., fig. 34). The lower side of the ceratophore
carries a pair of dermal folds or lobes, which project from the sides below and
inwards. The ventral surface is not encrusted, but is thickly beset in each segment
with transverse rows of not very long acuminate papillae which are continued upon
the ventral surface of the parapodia. The ventral middle line is depressed. Amongst
the bases of the acuminate papillae are scattered minute globular papillae reminding of
those present in Hermione. The cirrophores (appendages of buccal segment) lie
between the palps and the head, so that it is difficult to remove a cirrophore without
the palp of the same side ; the palps are glabrous.

The simple plumose setae of the buccal segment have been mentioned above. In
the second segment in addition to the plumose setae there are compound setae of the
form represented in Plate II., fig. 36, with filiform apex of the appendix, the shaft
being plumose. The ventral cirrus of the second segment is stout and long, nearly
twice as long as the parapodium, and arises from a distinct basal joint. In the third
segment compound setae of more usual form appear, the shafts being stout and
smooth except towards the extremity where they are squamose (Plate II., fig. 37) ;
the appendices are lost. The dorsal cirrus of this segment has the articulated
structure shown in Plate II., fig. 35. In the fourth segment the shafts of the
compound setae are still squamose distally ; the appendix is elongated, curved at
apex with a slight truncation jutting out below the apex. In a parapodium of a
normal body-segment (anterior third of body) there is a dense dorsal fascicle of finely
plumose capillary setae, followed by a superior ventral group of moderately stout
compound falcigerous setae with subelongate appendices ; then a central group of
stout compound setae with shorter appendices ; finally an inferior ventral group
of slender setae with elongate appendices (Plate II., figs. 38 and 39). Sometimes
the apex of the appendix is clearly bidentate, varying from this condition to smooth
(Plate II., fig. 40). The ventral border of the parapodium is beset with acuminate
filiform papillae and some spheroidal papillae (Plate II., fig. 41). The elytra are
notched and lobed, especially at the inner border (Plate II., fig. 42).

Psammolyce rigida, Grube — Plate II., figs. 44 to 47.

GRUBE, 1868, ' Ann. Roth. Meer.' (Frauenfeld) ; ' Verh. zool.-botan. Ges. Wien,' p. 631, 1878 ;
' Ann. Semp.' (Philippines), p. 55.

Locality: — Station LIIL, 10 miles north of West Cheval Paar. Three specimens,

POLYGH^TA. 257

one complete, 115 millims. long, tapering gradually behind; 8 millims. wide over
the seta?.

This species has the same general appearance as the Psammolyce arenosa of the
Mediterranean. It is probably distinct in minor points, though I cannot specify what
these points are without comparing actual specimens.

It differs from the preceding species in the following superficial respects :—

Ps. seylanica.

1. Colour fuscous.

2. Calcareous grains predominate.

3. Venter beset with capillary papillse, i.e.,

Venter hairy.

4. Terminal portion of dorsal cirrus of third

segment slenderer and shorter than its
peduncle.

Ps. riffida.
Colour fulvous.
Quartz grains predominate.
Venter beset with globular papillse, i.e., Venter

tuberous.
Terminal portion of same cirrus tapering and

longer than its peduncle.

The parapodial armature of the two species shows such close correspondence that
the slight differences which are observable could easily be attributed to individual
variation, but the differences noted above under Nos. 3 and 4 would seem to preclude
the possibility of regarding them as co-specific. Unfortunately I have no information
as to the precise locality of the Psammolyce zeylanica. In Professor Herdman's
' Narrative ' of his expedition, Psammolyce is recorded from two stations, LII. and
LIU. The latter is that from which the present species was dredged in 7|- to
9 fathoms, " bottom muddy sand with some dead shells ; " the former station may be
the locality of the other species, "between north of Cheval Paar and Vankali reef;
depth 3 to 6 fathoms ; bottom sand."

The difference noted above in respect of the adventitious coating of sand-grains is
only of local significance as indicating a difference of habitat. The tuberculation of
the ventral surface of the body is the most obvious character of this species. The low
rounded dermal tubercles are placed close together without reference to segmental
limits and not in rows, forming an even, elastic sole on each side of the depressed
neural tract. The dermal tubercles are continued upon the parapodia, especially in
the posterior two-thirds of the body, where they are thickly covered. Besides these
spheroidal dermal tubercles there are tufts of filiform papillse like those shown on
Plate II., fig. 41. These occur on the base and summit of the parapodia, at the base
of the ventral cirrus, and there is a ventro-lateral tuft on each segment between the
tubercular tract and the parapodia.

Other substantial differences between the two species are shown in the figures.
The elytra are formed upon the same model in both, and they are not safe objects for
comparison, since they vary from segment to segment. Those of Ps. rigida have a
concave anterior border, those of Ps. zeylanica are straight. The scar of insertion is
short in the former, elongated in the latter (Plate II., fig. 46). The setse of the
second segment have a shorter appendix (Plate II., fig. 47). An important difference

2 L

258 CEYLON PEARL OYSTER REPORT.

appears in the stout setae from the centre of the ventral fascicle. In Plate II.,
figs. 43 and 44, two of these setae, one from the right sixteenth foot of each species,
are shown side hy side. In Ps. zeylanica the declivity of the articular surface of the
shaft is greater, and below the apex of the shaft there is a very distinct semilunar
cusp which is not present in the seta of the same order in Ps. rigida. The other setae
show close correspondence, and here, as there, some of them are distinctly bidentate,
but the difference noted appears to be constant.

Sthenelais zeylanica, n. sp. — Plate II., fig. 48.

A fragment with extruded proboscis, 15 millims. long, 5 millims. wide, in company
with Hyalinoecia camiguina, was dredged off Foul Point, Trincomalee, Station XXV.,
8 fathoms.

This species is closely allied to Sthenelais boa, but differs in several well-marked
points. The prostomium carries four eyes in front, two on either side of the stout
ceratophore with its aliform lobes (characteristic of the genus) ; the tentaculum impar
borne upon the ceratophore is not very long, about as long as the ceratophore and the
prostomium together ; the palps are long, nearly as long as the extruded proboscis.
The free margin of the latter carries both dorsally and ventrally a row of eleven
evenly disposed papillae.

The scales are carried on the usual segments II., IV., V., VII., IX

* * # * XXVII. , XXVIII. , XXIX., &c, becoming consecutive at segment
XXVII. In the condition of extruded proboscis the dorsum of the third segment
hardly shows, and the first two scales appear to occur on consecutive segments until
they are pressed apart. The general character of the scales (their pronounced reniform
shape, the tuberculation of the surface and the fimbriation of the margin) resembles
that described and figured by Professor McIntosh for Sthenelais boa (' Ray Soc.
Mon.,' 1900) ; the fimbriae of the outer margin are acuminate, others which occur on
the posterior margin are blunt ; the tubercles are thickly scattered over the whole
surface except near the anterior border of the inner lobe and near the corresponding
border of the outer lobe ; brownish pigment occurs all over the exposed portion of
the scale.

The characters upon which I rely for specific differentiation are presented by the
parapodia of the body-segments. In Sthenelais boa there are four principal groups of
setae, the dorsal plumose capillary setae, the superior ventral plumose spiniform setae,
the mid-ventral compound bidentate falcigerous setae, and the inferior ventral com-
pound setae with articulate bidentate appendix (cf. McIntosh, op. cit., 1900). In
Sth. zeylanica instead of the group of superior ventral spiniform setae we have,
composing this group, a few of the same kind of slender compound setae with
articulate appendix as occur in the inferior ventral group, and the spiniform setae are
not represented in the ventral ramus of the parapodium. The setae themselves are
not sensibly different from those of the corresponding forms in Sth. boa (Plate II.,

POLYCrLETA. 259

fig. 48). The ciliated cushions or cteuidia (McIntosh) on the upper surface of the
foot beneath the cirriform branchia are the same as in Sth. boa, but the disposition ot
stvlodes is different, particularly as regards the two long stylodes which proceed from
the base of the ventral cirrus in Sth. zeylanica (Plate II., fig. 48).

Sthenolepis, n. gen.

The genera Thalenessa and Leanira are characterised by the presence of a very
small teutaculum impar inserted directly upon the prostomium, not borne upon a
ceratophore. The compound setaa of Thalenessa are a variety of the falcigerous type
with bidentate appendices; those of Leanira are spinigerous.

Sthenelais and Sthenolepis, n. gen., are characterised by the presence of a long
tentaculum impar borne upon a ceratophore which is provided with a pair of
spatulate appendages. The compound setas of Sthenelais are falcigerous like those
of Thalenessa ; those of Sthenolepis are spinigerous as in Leanira.

All the species of Leanira described by Professor McIntosh in the " Challenger"
collection are to be ranged in the genus Sthenolepis. In this genus the scales are
generally smooth.

Sthenolepis japonica (McIntosh) — Plate II., fig. 49.

Leanira japonica, McIntosh, ' "Challenger" Polychseta,' 1885, p. 154.

A headless posterior fragment of a Sigalionid worm, labelled " L,eanira sp.," was
dredged at Station XXXV., off Galle, 7 fathoms. An anterior fragment of the same
species with same locality and date was contained in another bottle in company with
< rlycera lancadivcp.

The different preservation of the two pieces gives them a different aspect, but the
structure of the parapodia shows that they belong to the same species. The thin
translucent smooth elytra embrace the body so closely as to be inconspicuous ; in
front they meet in the middle line and overhang the head ; further back they leave
the mid-dorsum exposed and then approximate again.

The long tentaculum, accompanied by a pair of equally long tentacular cirri, projects
straightly and stiffly forwards ; the spatulate appendages are nearly as long as the
ceratophore ; at the base of the latter on each side of the prostomium a small eye is
visible from above. The ends of the parapodial rami are furnished with a rich growth
of stvlodes ; the dorsal setce are very numerous, long, slender, and transversely
fringed ; from the dorsal end of the ventral ramus a compact tuft of about 20
spinulose simple setse issues ; the rest of the ventral setae are compound spinigerous,
the appendices are not very long and present a peculiar intrinsic laminated structure.
Between the cirriform gill and the dorsal fascicle are three ctenidia. The ventral
cirrus has a rounded protuberance some distance beyond its insertion (Plate II.,
fig. 49). The body is slender, the anterior region having a width of 2-5 minims.
across the body, 4 millims. across the setaa.

2 L 2

2(50 CEYLON PEARL OYSTER REPORT.

A much smaller complete specimen from the same locality is also contained in the
collection and shows to perfection a feature which is present in the larger individual.
The anterior setae of the buccal segment are exceedingly long and slender, as long as
the tentaculum and tentacular cirri, and they embrace these appendages in such a
manner that the whole complex bears the appearance of a compact flabellum.

There is a great quantity of mucus surrounding the specimens which was
apparently produced by the Glycerids with which they were preserved, and this no
doubt binds the setae and tentacles fortuitously together, but the extremely delicate
porrect fascicles of long hair-like setae are highly characteristic. Another point which
should be noted is the greater length of the appendices of the compound setae in the
anterior segments. In the smaller specimen there are fewer spinulose or verticillate
setae in the superior fascicle of the ventral ramus, and the stylodes appear to be less
numerous ; in fact, in a parapodium from the mid-body I only see one verticillate seta
as in Leanira japonica, McInt. Although the Japanese worm retained no scales,
yet it seems probable that the Ceylon specimens are co- specific. In these the scales
adhere firmly to the body.

Thalenessa digitata, McIntosh — Plate II., figs. 50 to 52.

Thalenessa digitata, McIntosh, " Polychoeta," '"Challenger" Rep.,' 1885, p. 140 (off the

Admiralty Islands in 16-25 fathoms).
Thalenessa im-tlnirni, Hornell, ' Ceylon Pearl Oyster Report,' Part I., 1903, pp. 16, 49, and 52.

Taken off Galle (Station XXXVIII.) and off Panadure (Station XLV.) down to
25 fathoms.

Three small frontal antennae ; the lips of the anterior four pairs of setigerous
parapodia are produced into conspicuous membranellai and also carry numerous
stylodes ; the first five setigers contain setae with long, jointed appendices, such setae
being numerous in the first four feet, scarce in the fifth, and thereafter absent, only
setae with unjointed appendices being found (Plate II., fig. 50). All these points
correspond closely with Professor McIntosh's description and figures.

The elytra do not cover the back ; the first two elytra are smaller than the rest,
and the first has a smooth margin, destitute of fimbriae ; a typical elytron is sub-
triangular in shape, the outer border provided with about a dozen compound fimbriae,
more frequently with four divisions each (Plate II., fig. 51), but there may be as
many as six branches, or only three, two, or one. The elytra are attached to seg-
ments II., IV., V, VII. , IX., ****** * XXV, XXVII. , XXVIII. ,
XXIX., &c, becoming continuous at segment XXVII. Cirriform branchiae are
present in all setigerous segments, two clear ctenidia on the dorsal surface of the foot,
one attached to the peduncle of the gill. Ventral cirrus long, projecting beyond the
foot. No papillae on the ventral division of a typical foot from the middle region.

Tlmlenessa oculata, McIntosh, is apparently another form of digitata, slightly
differing from the type. It is interesting to take note of the differential shifting of

POLYCH^TA. 261

parts which has taken place, for example in such genera as Thalenessa and Iphione.
In the former the palps are concealed below the cirrophores of the tentacular cirri
(Plate II., fig. 52) ; in the latter the cirrophores of the tentacular cirri are concealed
below the palps (Plate I., fig. 6).

Thalenessa stylolepis, n. sp. — Plate III., figs. 53 to 56.

Localities: — Station LVIL, llj to 36 fathoms; and Station LVL, Dutch
Modragam Paar, 9 fathoms ; specimen incomplete behind ; taken out of coral block.

This must be the worm referred to on p. 74 of the ' Ceylon Pearl Oyster Report,'
Part I., as Sigalion mathildee.*

The chief characters, which typify a new section of the genus Thalenessa, are the
possession of a pair of small frontal antennae and an equally small median occipital
antenna ; the insertion of the elytra upon high peduncles ; and, in the anterior region,
the absence of the cirriform branchiae from the scaleless segments.

Prostomium large, flattened, shield-shaped, with three notches or emarginations,
two at the frontal border from which the paired antennae arise, one at the occipital
border from which the tentaculum impar arises (Plate III., fig. 53). Only one pair
of eyes was observed occupying a central position behind the frontal emarginations.
The cirrophores of the buccal segment are porrect, sub-median and adnate to the
prostomium. The line of division between the frontal border and the cirrophores is
at the same time the line of concrescence of these structures, so that the antennae are
virtually inserted into the base of the cirrophores, a condition leading to that found
in Sthenelais, Sthenolepis and Psammolyce. The palps are long and smooth.

The elytra are pedunculate, the elytrophores rising like stout pillars from the
dorsum, as in Eulepis ; the elytra are firmly attached to the elytrophores ; proceeding
outwards from the latter below each elytron is a rather long cirriform appendage
which does not occur on the intervening segments (segmenta nuda) ; the latter show
a small tubercle in the line of the elytrophores. The elytra are placed on the usual
segments up to the 27th, when they begin to be continuously successive. They
carry 12 to 13 (or fewer) plumose fimbriae (Plate III., fig. 56), which project straight
outwards from the outer border, followed by a few simple filiform papillae. The
elytra of the first pair are smaller than the following, rounded, not accompanied by
cirriform appendage. The remaining elytra cover the dorsum, their inner borders
meeting so as to form a tunnel along the length of the body ; on the inner side of
each elytrophore a small ctenidium projects into the tunnel, a pair of these ctenidia
being inclined towards one another in each body-segment (Plate III., fig. 54). As
mentioned above, the cirriform branchiae of Thalenessa stylolepis are confined to the
elytrophores. The dorsal ramus of the parapodium (Plate III., fig. 54) carries a
bundle of numerous long simple fringed setae, the fringes appearing in side view as a

* The plumose fimbriae of the elytra exactly resemble those of Sigalion Mathilda' as figured by Professor
McIxtosh (1900).

262 CEYLON PEARL OYSTER REPORT.

series of projecting overlapping scales ; mediae! of the dorsal fascicle the simple setae
are shorter and hyaline. The ventral ramus carries a superior bundle of simple
whorled seta?, the verticillate tract short. All the remaining ventral setae are of the
compound falcigerous type, bearing very long, many-jointed, tapering, bidentate
appendices, the number of joints being as many as ten. The compound setae of the
middle group are characterised by the possession of squamose (fringed) shafts, while
the more slender setae of the inferior group have plain shafts. A parapodium from the
posterior region (near the 60th segment) shows the dorsal ramus projecting clearly
beyond the ventral, and in the ventral fascicle, in addition to the compound setae with
the long, jointed, flexible appendices, there are two stout setae with short unjointed
appendices provided with a gaping beak (Plate III., fig. 55). Between 50 and
60 segments are present in the imperfect sjsecimen described above, which is
35 millims. long with width of 3 millims. without the setae, 4 millims. over the setae.

Family: PHYLLODOGID^.

Anaitis zeylanica, n. sp. — Plate III., figs. 57 to 60.

Locality : — South of Manaar Island, 8 to 9 fathoms. One specimen, proboscis
retracted.

The dorsal phyllodes are broadly ovate (cordate-lanceolate) and pedunculate, as they
are in a dozen other species. Head rounded ; eyes large ; tentacular cirri normal,
elongate. Proboscis (dissected) consists of two well-separated portions, a thin-walled
proximal or adoral portion densely crowded with papillae not serially disposed ; a
thick-walled distal portion with six prominent rows of large subtriangular papillae, six
or seven in a row.

Setae, ten in a parapodium, the shaft terminating in a triangular apex fringed at
the sides and articulating on one side with a long flagelliform strongly serrulate
appendix (Plate III., fig. 57). Anal cirri (one preserved) of moderate length,
acuminate with stout basal portion. Body slender, length 38 millims. ; total width
under 2 millims.

The setae appear heterogomph in side view, homogomph in frontal view (Plate III.,
fig. 58). I am afraid they offer no reliable diagnostic character, unless it is their size.
In this, however, as in other points they resemble the setae of Phyllodoce sancti-
josephi. The shape of the head and of the phyllodes is shown on Plate III., figs. 59
and 60.

Carobia castanea, Marenzeller.

Marenzeller, 'Siidjapan. Ann.,' i., 1879, p. 127 (p. 19 of reprint).

This species is distinguished from other Oriental Phyllodocidae by its deep red
colour.

The red dorsal phyllodes are cordate and the setae are distinguished from all others

POLYCH/ETA. 263

in this collection by their rather short appendices destitute of serrulations and the
wide, nearly horizontal articular end of the shaft, as figured by VON Marenzeller.

Notophyllum laciniatum, n. sp. — Plate III., figs. 61 and 62.

Station V., oft' Chilaw Paar, Gvdf of Manaar, 1 I fathoms. One specimen.

This Phyllodoceid worm is characterised by a relatively short and thick body and
by the possession of broad closely imbricating scales or phyllodes (foliaceous dorsal
cirri) ; the anterior phyllodes conceal the occipital region of the head, meeting across
the middle line, the rest leave the mid-dorsum exposed. There are as many as ninety
segments. The specimen measures about 20 millims. in length and 3 millims. in
breadth.

The head is simple, with two large eyes in the posterior half, between which arises a
brown-tinted tentaculum impar. The somewhat fusiform antennas arise between the
eye-region and the anterior border, and the fusiform palps, with rather long terminal
flagellum, arise nearer the middle line from the under side of the frontal region. The
left antenna was lost from the specimen. The four pairs of tentacular cirri are closely
aggregated in the cephalic region, only two of them are seen from above, the second
and fourth, the latter being the longest. The first cirrus is more massive thau the
others.

The great character of the species is given by the presence of three cirriform
occipital or nuchal lappets, which hang backwards on each side from the back of the
head (Plate III., fig. 61). This feature also occurs in Phyllodoce multicirris, Grube
('Ann. Semp.,' p. 100), which is clearly a closely allied species, differing in the
structure of the head, especially in the decided possession of four eyes. The seta? and
phyllodes also exhibit minor differences. Each occipital lappet has a pale border and
a brown centre. The dorsal phyllodes have the same shape and arrangement as
figured by Grube for Ph. multicirris, but they do not show the round surface
markings of that species. The ventral phyllodes also have the same peculiar
disposition, inserted high up near the extremity of the parapodium and lying out-
spread behind the pharetra setarum.

In the seta? the articular end of the shaft shows a character of its own. There are
as many as 21 setaa in a ventral fascicle from the anterior region; most of them are
broken. In optical section the ends of the shafts appear deeply excavate, the
articular fossa bounded by stout refringent walls (Plate III., fig. 62). The dorsal
ramus is represented by the lobe upon the summit of which the dorsal phyllode is
inserted ; it is penetrated by a single slender acicula. The appendices of the
compound setae are rather long and minutely serrulate.

Phyllodoce dissotyla, n. sp. — Plate III., figs. 63 to 66.

Station V., oft* Chilaw Paar, Gulf of Manaar, 11 fathoms. One specimen, 25 millims.
long by 1 millim. broad ; proboscis not extruded.

264 CEYLON FEARL OYSTER REPORT.

The head is longer than broad ; the antennae do not stretch back to the eyes ; the
eye shows a clear lens (Plate III., fig. 63). There are four pairs of long tentacular
cirri of normal form. The dorsal phyllodes are rounded, not lanceolate, and strongly
pedunculate. The setae, 18 in a fascicle, are heterogomph, the articulation of the
appendix distinct; the appendix with serrulate edge (Plate III., figs. 64 and 65).

The dominant specific character was ascertained by removing and opening the
retracted proboscis, the adoral portion of which is beset with longitudinal rows of
rounded normal papillae ; in two of the rows, probably median dorsal and median
ventral (since they are separated from one another by normal rows), there are three
large triangular papillae placed one behind the other, with normal papillae in front
and behind in the same rows. These modified papillae are denser than the rest and
are noticeable under low magnification with a simple lens (Plate III., fig. 66).
They are quite definite, two sets of three on opposite sides of the proboscis.

The phyllodes aud setae are also characteristic, but the determining character is
given by the papillae of the proboscis.

Phyllodoce foliosopapillata, Hornell — Plate III., figs. 67 to 69.

'Ceylon Pearl Oyster Report,' Part I., 1903, p. 16 and p. 28, " Yellowish green Phyllodocid."

Station XVII., outside Periya Paar, 11 fathoms.

The specimen, which is incomplete behind, has a length of 135 millims. (141 millims.
with extruded proboscis) ; about 207 segments ; ventral width between the parapodia
2 millims. ; over the parapodia 5 millims. The proboscis is firm and sexangulate ;
towards the base it is traversed by transverse rugae, and behind these, at the level of
the head on each side, there are six rows of transversely elongate, foliate papillae,
5 or 6 in the first row, 8 or !) in the second, 10 in the third, and again decreasing
below; at the free margin I counted 14 or 15 rounded papillae (Plate III., fig. 67).
The foliaceous dorsal cirri, which may be called phyllodes to distinguish them from
the elytra of the Aphroditidae, are borne upon broad peduncles to which they are
strongly adherent. In this specimen the phyllophores are filled with ova (Plate III.,
fig. 68).

There are as many as 25 compound setae in a parapodium. The long flexible
spiniform appendices are serrulate along one border, but in some of them the serru-
lations are obscure or even obsolete. The setae differ from those of other species in
the presence of a long stout spur at the head of the shaft, which sometimes projects
considerably beyond the neighbouring denticulations, even more so than in the
example figured (Plate III., fig. 69).

Closely apposed to the sides of the prostomium is a pair of lateral nuchal organs
(Plate III., fig. 67). The prostomium is broader than long, deeply emarginate behind,
and in the notch there is an occipital papilla, such as occurs also in Phyllodoce
madeirensis and other species.

P0IATH7ETA. '265

The insertion of the four pairs of tentacular cirri is the same as in Ph. madeirensis,
namely, on each side : —

Segment I., one tentacular cirrus ;
,, IT., two tentacular cirri ;
„ III., one tentacular cirrus and a cirrus ventralis foliaceus.

Tlu1 first appearance presented is that of two cirri in the first segment, but closer
examination showed that this appearance is deceptive. The longest is the dorsal
cirrus of the second segment, stretching back over eleven segments.

Phyllodoce macrolepidota, Schmarda — Plate III., figs. 70 and 71.

Schmarda, 'Neue Wirbellose Thiere,' ii., 1861, p. 83, Taf. xxix., fig. 229.
Phyllodoce tenuissima, Grube, 'Ann. Semp.,' 1878, p. 95.

Locality : — Galle Habour, on oyster cage. One specimen.

Body very long and slender, 200 millims., with a width of 3 millims. in front,
over the setae ; more than 350 segments. The proboscis is only half extruded, and
this portion is inflated and soft ; it carries six longitudinal rows of broad, flattened,
arcuate, dark-coloured papillae on each side in front of the head-region, 8 to 10
papillae in a row, and, in addition, a median dorsal row of six papillae. The general
arrangement of the papillae and the presence of the median row are features which
characterise Ph. madeirensis, Langerhans, of which this may be an Oriental repre-
sentative, but it has nearly three times the length and about twice as many segments
as in the type-species of Langerhans. There are 17 setae in a parapodium
(Plate III., fig. 70) ; they do not offer any striking character (Plate III., fig. 71).
The head is about as long as it is broad, with rather large eyes in the centre of
the posterior cephalic convexity on each side ; it is acutely notched behind, but the
specimen does not show a nuchal papilla. There are two rather short, stout, subulate
anal cirri.

Phyllodoce sancti-josephi, Gravier — Plate III., figs. 72 and 73.

(Iravier, "Ann. Polychetes de la Mer Rouge," 'Arch. Mus. Paris (4 ser.), ii., 1900, p. 196.
Station V., off Chilaw Paar, Gulf of Manaar, 11 fathoms. Two sj)ecimens.
These worms differ from the type in the shape of the head and in the apparent
absence of an occipital papilla. The widening of the head is correlated with the
extrusion of the proboscis, and the occlusion of the papilla may be due to the same
cause. One of the specimens has a length of 55 millims. and a total width in front
of 2 millims.

The proboscis in its general consistency and in the presence of a posterior rugose
portion resembles that of Ph. foliosojMpillata, but the papillae, of which there are
six rows on each side, extend over the rugose region in front of the head (Plate III.,
fig. 72). In the second specimen the proboscis shows six prominent angulations in
front of the rugose portion.

2 M

2G6 CEYLON PEARL OYSTER REPORT.

The head itself, under the condition of extruded pi'oboscis, is about as broad as
long and not deeply notched behind. On each side of the prostomium there is a
prominent lateral nuchal papilla. The phyllodes resemble those of Ph. macrolepidota
in shape and proportion, the dorsal phyllode being distinctly pedunculate and its axis
occupied by a rete mirabile, from which radial vessels proceed to the margins. The
parapodium carries twelve setae of normal type (Plate III., fig. 73), though, as shown
by M. Gravier, the number varies within narrow limits in different regions of the
body. The number which I have given holds good for the mid-region. A foot from
the posterior region has sixteen setae.

Pterocirrus ceylonicus, Mtchaelsen.

This species is well characterised by the five long slender subequal prostomial
appendages, longer than the prostomium. The remarkable winged ventral cirrus
of the second segment which, seen from above along its thickened dorsal border,
looks like an ordinary tentacular cirrus, is a generic feature of first importance.
The dorsal phyllodes are more lanceolate than that shown in Dr. Michaelsen's
figure, "Polychiiten von Ceylon," ' Jahrb. Hamburg. Wiss. Anstalt,' 1892, ix., 2, p. 103,
otherwise I have nothing to add to his description. The colour of the preserved
specimens is dark greenish-brown. Locality : — Ceylon Seas.

Family : HESIONID/E.
Hesione ceylonica, Grube.

Grube, "Ann. Ceylon" (Holdsworth coll,) ' P. Zool. Soc. London,' 1874, p, 327.

Locality : — Numerous specimens from various stations.

This is probably a geographical form of Hesione splendida, Savigny, which is
characterised by the possession of one pair oi frontal antenna? ; eight pairs ot
tentacular cirri ; sixteen pairs of uniramous parapodia ; four distinct eyes ; pigment
disposed more or less in longitudinal bands or streaks. Savigny was in error regard-
ing the minute antennae, and the illustrations do not correspond with the text in
certain particulars, hence confusion has arisen.

The Ceylon specimens range in length of body from 9 millims. to 3G millims. ; in
the latter case the ventral width between the parapodia is 4 millims., over the setae
11 millims. In all cases the full number of setigerous segments, namely 16, is
present. The extruded proboscis shows a median dorsal callosity or hard papilla
(not seen in the smallest specimen) a short distance from the base, like that figured
by Audouin and Milne-Edwards (1834) in 11. pantherina. The compound
falcigerous setae have bidentate appendices with a spiniform guard arising below
the denticulations. The guard may be worn away.

A specimen of 14 millims. length had a ventral width between the parapodia
of 2 millims., over the setae nearly 5 millims.

I'OLYCILKTA. 267

Hesione genetta, Grtjbe. (See Grube, 'Ann. Semp.,' 1878, p. 104.)

One specimen from Chilaw Paar, Station LXIX.

This species is distinguished from the members of the splendida group by its very

characteristic and enduring pigmentation. At the junction of the tirst and second
setigerous segments there is a broad dark transverse band passing from one side to
the other across the convex dorsum. Behind this band the segments are marked by
transverse brown spots which are sometimes confluent. The cuticle is highly
iridescent ; the body flexed ventrally and the ventro-lateral borders elevated, leaving
a median neural groove between them. The prostomium is quadratic ; two frontal
notches from which the minute antennae arise ; eyes barely visible. The body is
shaped like that of H. reticulata, von Maeenzeller (South Japan), and resembles
the body of a caterpillar, as remarked by vox Marenzeller. Length 13 millims.,
inclusive width 3 millims.

Irma limicola, n. sp. — Plate III., figs. 74 to 76.

This species is founded upon a single example which occurred in one of the mud-
tubes of Loimia montagui from Palk Bay. It is not very well preserved in
consequence of having been trapped in such a narrow recess. Its length is
20 millims., width 3 millims., with upwards of 50 segments.

The head is transverse, the eyes widely separated, the anterior and posterior eyes
of each couple crescentic and closely approximated. From the frontal border of the
prostomium on each side two equal cephalic appendages (antennas and palps) curve
backwards, lying over the eyes and sides of the prostomium. The biannulate
proboscis is extruded. No tentaculum impar could be observed. Six tentacular
cirri are present on each side of the buccal segment, placed in a dorsi-ventral manner,
the upper and lower of the group being shorter than the intermediate.

The dorsal ramus of the parapodium is intimately connate with the cirrophore of
the dorsal cirrus into which an acicula extends, and from a slight protuberance on
the lower side of the cirrophore a bundle of very fine capillary setae emerges
(Plate III., fig. 74). The compound setae of the ventral ramus are numerous, about
fifty in each fascicle. They are falcigerous, carrying very long appendices curved at
the tip and bidentate, the apical tooth frequently connected with the minor tooth by
a limbus. The superior setae of the ventral ramus have appendices about one-half
shorter than the average ; the inferior setae, i.e., the most ventrally placed seta?, carry
quarter-sized appendices (Plate III., figs. 75 and 76). The relations of the dorsal
and ventral cirri and the lip of the parapodium are shown in the figure (Plate III.,
fig. 74). Repeated examination failed to reveal any trace of a tentaculum impar in
Irma limicola. In this respect it approaches the genus Castalia, Sav., as amended
by M. Sars ('Christ. Vid. Selsk. Forh.,' 1861, p. 88), but there are no coronary
papilke on the proboscis and no jaws.

2 m 2

2 (58 CEYLON PEARL OYSTER REPORT.

The definition of Irma should be modified as follows : —

Proboscis nuda ; palpi biarticulati ; antennas 3 (interdum 2 observatse) ; oculi 4 ;
cirri tentaculares 12 ; pedes quasi biremes ; ramus dorsalis vestigialis cum articulo
basali cirri dorsalis connatus, setas dorsales capillares gerens aut nullus ; seta? ven-
trales composite falcigerse ; acicula? binse.

The fact of an acicula passing into the cirrophore indicates the primitive biramous
nature of the foot in those species where dorsal capillary setse do not occur, exactly
as in the analogous case of Podarke, Ehlers. The absence of a median antenna in
/. limicola is puzzling, and will remain so until more material is obtained. The
species described by me as Oxydromus aucklandicus (' "Southern Cross" Collections,'
London, 1902, p. 281) should be transferred to the genus Irma as now defined.

Family: SYLLIDiE.

Typosyllis taprobanensis, n. sp. — Plate III., figs. 77 and 78.

Localities : — East side of Cheval Paar ; and South-west Cheval Paar, Gulf of
Manaar.

This is a variable species, one specimen seemed to resemble T. variegata, the next
T. krohni, the differences being in part substantial, in part due to the fact that the
pharynx of the one was on the point of being projected, that of the other was
retracted completely.

Antennae and cirri moniliform ; in the first specimen (length 19 millims., width
1'5 millims.) the paired and unpaired antennae were about equal in length to each
other and to the palps, the paired rather slenderer than the hnpar, shorter than the
d< irsal cirri ; the latter, not deciduous, longer alternating with shorter and slenderer ;
the first ventral tentacular cirrus slender, not much longer than the palps ; head
transverse, partially concealed below the buccal segment ; pharynx, on the point of
projection from the mouth, lined by thick dark-brown cuticle (showing dark purplish-
black through the skin) carrying a subtermmal dorsal tooth coloured the same as the
chitinous intima ; proventriculus, with about 37 rows, finely tessellated, glistening
whitish with delicate flush.

In the second specimen the median antenna was long, showing 36 joints ; most of
the dorsal cirri were lost, that of the 5th setiger showed 42 joints, that of the
64th setiger upwards of 30 joints ; the pharynx and proventriculus, seen by trans-
parency, exactly resembled those of the first specimen and decided the specific identity,
but were differently placed, being retracted ; the pharynx extended from the 7th to
the 15th setigerous segment, the proventriculus from the 15th to the 25th setigerous
segment ; colour of the first 24 setigerous segments : — two brown transverse bands in
each segment; total number of segments 150 ; length 30 millims.

The setas of both corresponded in general characters, differing in a manner shown
in the figures (Plate III., figs. 77 and 78). In the anterior segments there are about

POLYCH^ETA. 269

12 to 16 compound falcigeroua bidentate setae lying below 5 acicula? ; the aciculae
become reduced to 3 and posteriorly to 2 ; the setae also undergo changes, the shaft
becoming stouter and the appendix shorter ; the shaft further assumes a characteristic
dilatation of varying intensity below the joint (Plate III., figs. 77 and 78).

The anal cirri showed 24 joints ; no median caudal process was observed. The
head of the second specimen was normal, rounded, not transversely elongate, the
latter condition apparently resulting from the commencing extrusion of the pharynx.
The palps are quite separate, with distinct subfrontal insertions, a wide proximal
half and a narrower distal hah.

In Typosyllis rarityata, according to von Marenzeller," the palps are fused
together at the base. The Ceylon species would appear to be related to T. hrohni,
as described by Langerhans. t The difference noted above in respect of the shafts
of the setae may be correlated with the difference of age and size of the specimens of
T. taprobanensis.

Haplosyllis spongicola (Grube) — Plate III., figs. 79 and 80.

Gkube, 'Arch. Naturg.,' 1855, p. 104; Marion et Bobketzky, 1875, 'Ann. Marseilles,' p. 2-i.
Syllis haniata, Claparede, 1868; Langerhans, Madeira, 1879, ' Z. f. w. Z.,' xxxii., p. 527;

St. Joseph, ' Dinard,' 1886, part i., p. 142; 1895, part iv\, p. 185.
Syllis violaceo-flava, Grube, 'Ann. Semp.,' 1878, p. 115; see also Langerhans, Madeira,

part iii., ' Z. f. w. Z.,' xxxiv., p. 128.

A small specimen, 4 '5 millims. long, was takeu from between the whorls of the
operculum of a large Serpulid [Pomatostegus actiuocc7vs^, off Panadure, Station XLV.,
25 fathoms.

Over 40 setigerous segments ; setae brittle, many of them with the tips broken off.
The proventriculus is long, occupying segments 13-25, but the body is contracted.
On examining the specimen in toto. it is common to see two hamate unguarded
acicular setae projecting from a parapodium, one of which is stronger than the other ;
both of these seta? have a bidenticulate apical tooth as figured by Marion and
Bobretzky, and by Langerhans. (In Haplosyllis djiboutiensis,\ Gravier found, in
one individual, the apical tooth of the stronger seta simple. ) The pharynx is armed in
front with a single conical tooth surrounded by a crown often soft papilla? (Plate III.,
fig. 80). The head is shown in Plate III., fig. 79.

Syllis gracilis, Grube (1840).

(Of. St. Joseph, "Ann. de Dinard," part i., 'Ann. Sei. Nat.' (7) i., 1886, p. 158; part iv.,
t. xx., 1895, p. 190.)

Locality : — South-west Cheval Paar. Three specimens.

1. About 180 segments ; length 45 millims. ; body contorted ; diameter uniform,

* ' Adriat. Ann.,' ii., 1875, p. 147.

t •' Wurmfauna Madeira," i., ' Zeit. f. wiss. Zool.,' xxxii., 1879, p. 529.

| Gravier, "Ann. Mer Rouge," 1900, 'Arch. Mus. Paris' (4) ii., p. 117.

270 CEYLON PEARL OYSTER REPORT.

about 1 milliin. without feet, 2 millims. inclusive. Anterior setae compound, falcigerous,
bidentate, in about 30 segments ; thereafter simple furciform setae ; in the posterior
segments some of the furciform setae become quasi-compound. Dorsal cirri of about
40 anterior segments long, with a tendency to alternate longer and shorter ; afterwards
they become shorter, subfusiforni, still alternating. Tentaculum impar with 27 joints ;
paired antennae, about 10 joints; dorsal tentacular cirrus, 18-20 joints; first dorsal
cirrus, 34 joints ; posterior dorsal cirri, 8-12 joints. Pharynx black in segments 3-20 ;
proventriculus in segments 20-28.

2. Length 15 millims., 83 segments followed by about a dozen regenerating
segments showing small compound setae of the usual type. Compound setae in anterior
28 segments.

3. Length 22 millims. ; segments 118 ; compound setae in 28 anterior segments; in
segments 29-90 simple furciform setae ; in segments 91-118 the falcigerous bidentate
setae re-appear. Pharynx brown in segments 1-13 ; proventriculus 14-20. Tentaculum
impar moniliform, 18 joints; dorsal cirrus of first setiger longest, with 20 joints.

Autolytus orientalis, n. sp. — Plate IV., figs. 81 to 84.

Locality: — Found in plankton, 2nd and 3rd March, 1903, North-east Cheval.
Several stoloniferous specimens; total length 10 to 20 millims.

In the example figured (Plate IV., fig. 81) there are 29 setigerous segments in the
anterior or parent individual. The parapodia contain two aciculae and numerous,
upwards of 20, compound falcigerous setae ; the head of the shaft is laciniate and the
appendix is minutely bidentate and minutely fringed (Plate IV., tig. 84). The dorsal
cirri are rather short, lanceolate, petaloid, with strong basal articulation. The second
dorsal cirrus, i.e., the cirrus of the first setiger, is the longest. The rounded reduced
palps, joined together in the middle line along their own length, are only visible from
below (Plate IV., fig. 83). The pharynx is long and has a sigmoid flexure ; it is
armed in front with a circle of 44 denticles, larger and smaller irregularly alternating
(Plate IV., fig. 82). The proventriculus shows 28 glandular rows.

Family: NEREID^E.
Nereis indica, Kinbekg.

KlNBEKG, ' Ainmlata nova — Nereidum dispositio nova,' ' Ofv. Akad. Forh.,' 1865, p. 169.

Locality : — Galle, from buoy ; one female containing ova.

Length 50 millims., inclusive width 4"5 millims. Tentacular cirri not long, reaching
to third segment ; antennae f length of prostomium ; palps long with narrow terminal
appendix ; eyes large, equal, the posterior rather closer together. Body incomplete
behind, 77 setigerous segments.

Feet subequal ; dorsal cirri of anterior segments about twice as long as the blunt
dorsal ligule ; ventral cirrus not reaching the trp of the ventral ligule ; in the middle
segments the neuropodia and the bases of the cirri become membranous, the dorsal

P0LYCH7ETA. 271

ligule becomes elongate and acuminate,* the dorsal cirrus exceeding it by about one-
fifth only ; in the posterior segments the dorsal cirri are upwards of four times the
length of the dorsal ligule. Dorsal setae homogomph spinigerous ; superior ventral
setae homogomph spinigerous and heterogomph falcigerous ; inferior ventral setae
heterogomph spinigerous and heterogomph falcigerous. The shafts of all the setae are
striated and all the appendices are strongly fringed (setulose) ; the falciform appendices
art' entire at the apex, not bidentate.

The proboscis was retracted and was dissected out: — I., 1; II., 12-12 (oblique
acervi) ; TIL, 10 (triangular acervus) ; IV., 15-15 (three unequal rows); V., 0; VI.,
6-6 (irregular); VII. and VITL, biserial, anterior row large alternating paragnaths ;
paragnaths of the posterior row half as large and one and a half times as uumerous
as those of the front row.

Nereis unifasciata, n. sp. — Plate IV., tigs. 85 to 88.

Locality : — South-east Cheval Paar.

A small Nereid (length 11 millims., incomplete behind) characterised by the
presence of a dark brown collar across the whole of the dorsum of the second
setigerous segment. The buccal and the first setigerous segments are pale brownish
at the sides, but there is no band across them. Commencing from the intersegmental
groove between the fourth and fifth setigerous segments there are about twelve pairs
of small intersegmental brown spots on the back. The feet are equal and resemble
those of Ceratonereis pectinifera (p. 272) in the general proportions of parts
(Plate IV., figs. 85 and 86). The two worms were collected at the same time and
place.

The distribution of the setae in the foot is as follows : — Dorsal, homogomph
spinigers ; superior ventral, homogomph spinigers and heterogomph falcigers ; inferior
ventral, heterogomph spinigers and falcigers. The heterogomph spinigers are not
completely heterogomph, but, at least in the middle and posterior segments, present a
condition which might be described as hemigomph (Plate IV, figs. 87 and 88). This
small detail affords the only real distinction between the armature of the feet of this
species and that of Ceratonereis pectinifera. The proboscis was not extruded, and,
although as a rule I do not attempt the description of small Nereids unless the
proboscis is extruded, I was anxious to test the specific value of the colour- markings
and of the setae. The dissected proboscis showed paragnaths in the adoral division,
those of the order VI. in small acervi, of VII. and VIII. in a single row as in Nereis
trifasciata, Grube. It is therefore not a Ceratonereis.

I cannot state any differences between the armature of the proboscis of this species
and that of Nereis trifnsci<tt<<.'\ The tact that such a minute detail as the presence

* The inferior lip of the notopodium is also produced as a ligule, the inferior dorsal ligule, resembling
the superior ligule throughout, but rather shorter,
t Except that the maxilke show seven teeth.

272 CEYLON PEARL OYSTER REPORT.

of hemigomph instead of homogomph spinigers in the inferior ventral fascicle is
sufficient to indicate fundamental divergence is interesting. The third tentacular
cirrus stretches back to the end of the seventh segment as in N. tmfasciata.

Nereis ehlersiana, Grube. (Grube, 'Ann. Semp.,' 1878, p. 71.)

Locality : — Station LXIX., Chilaw Paar.

The anterior portion of the body comprises 15 segments (14 setigerous) ; the dorsal
cirri of the first seven setigerous segments are modified, unequally elongate and
foliaceous. A normal parapodium from the anterior region shows reduced rounded
ligules, moderate dorsal cirrus much longer than its ligule, short ventral cirrus ;
dorsal seta? homogomph spinigerous ; superior ventral homogomph spinigers and
heterogomph falcigers ; inferior ventral heterogomph falcigers.

Ceratonereis falcaria, n. sp. — Plate IV., fig. 89.

Locality : — South-west Cheval Paar, two specimens.

About 107 segments with an even diameter of 2'5 millims. up to the region of the
55th segment, then rapidly attenuating ; length 27 millims. Frontal border notched
so that the antenna? are borne upon ceratophores as in C. tentaculata, Kbg., and
C. excisa, Gr, Eyes large, in a rectangle. Tentacular cirri moderately long ; dorsal
cirri more than twice as long as the dorsal ligules ; ventral cirri of the two middle
quarters of the body with basal lobules.

The distinctive character of the species is afforded by the setre. In the first
sixteen segments the setse of the dorsal fascicle are homogomph spinigerous ; of the
superior ventral fascicle homogomph spinigerous and heterogomph falcigerous ; in the
inferior ventral fascicle heterogomph spinigerous and heterogomph falcigerous ; the
shafts are striated, the appendices fringed and the falciform appendices are simple,
not bidentate. After this the dorsal seta? undergo a change, becoming homogomph
falcigerous with a deeply set, strongly bidentate appendix without guard (Plate IV.,
fig. 89). Transitional forms between the two kinds of dorsal setse are to be observed,
namely, seta? in which the appendix is shorter and stouter than in the normal
spinigerous setse. Paragnaths : — I., 0; II., 5-5; III., 7 (in one row); IV, 10-10.
Maxilke with about six denticulations. In the character of the paragnaths this
species approaches C. vulgata, Ktnberg, from Honolulu. The change of dorsal setse
from spinigerous to falcigerous is remarkably diagnostic.

Ceratonereis pectinifera, Grube — Plate IV, figs. 90 and 91.
Grube, 'Ann. Semp.,' 1878, p. 66.

Locality : — South-east Cheval Paar.

A small Nereid, 17 millims. in length, almost the only one in the collection with
the proboscis extruded. The systematic importance of the paragnaths of Nereidse is
well known, and a particular illustration of it is afforded by this species in comparison

POLYCrLETA. 273

with Nereis unifasciata (q. v.). Antennae shorter than proboscis, eyes in a wide
trapezium or subrectangle ; third tentacular cirrus reaches back to seventh segment.
Posterior segments with brown glandular tracts on the dorsal sides of the segments
and on the bases of the parapodia (Plate IV., fig. 91). Feet equal, dorsal cirrus
longer than dorsal ligule ; in the anterior feet there are prominent labia pharetrarum
(Plate IV., fig. 90).

Dorsal setae homogomph spinigerous ; superior ventral setae homogomph spinigerous
and heterogomph falcigerous ; inferior ventral setae homogomph spinigerous and
heterogomph falcigerous. The paragnaths do not quite agree with Grtjbe's formula,
but I think nearly enough to determine the species : — I., 0, II., a double oblique row,
3 or 4 larger in the hinder row, 4 or 5 smaller in the front row; III., 1 ; IV, an
acervus of 8, two very small in front, three rather larger in the middle and three still
larger behind. The actual length of the third tentacular cirrus is about 2 millims.

The Ceylon specimen seems to agree with Grube's C. pectinifera in all characters
except the mouth-parts and with C. lapinigensis in respect of the mouth-parts. The
maxillae show 5 blunt teeth. Perhaps the two forms are co-specific. The arrangement
of the setae in the foot is not described by Grube in either of his species. The
peculiarity here is the occurrence of homogomph spinigerous setae as well as hetero-
gomph falcigers in the inferior ventral group. In Ceratonereis costce, according to
Ehlers (' Borstenwiirmer,' p. 525), there are only falcigerous setae in the inferior
ventral group. It is uncommon to find homogomph spinigers in the inferior ventral
fascicle.

Platynereis bengalensis, Kinberg — Plate IV, figs. 92 to 94.

Muttuvaratu Paar, March 29. One specimen, incomplete behind, 45 setigerous
segments, width over setae 4 millims.

Description of a foot from anterior region (9th foot of right side) : — Dorsal cirrus
normal, 2|- times the length of the dorsal ligule, which is short, rounded, and bulky ;
inferior dorsal ligule like the superior; ventral ligule like the dorsal ligules; ventral
cirrus not reaching the end of the ventral ligule. Dorsal setae homogomph spinigerous ;
superior ventral setae homogomph spinigerous and heterogomph falcigerous (appen-
dices simple, fringed, short) ; inferior ventral setae heterogomph, spinigerous and
falcigerous.

Description of the 44th foot of right side : — A large brown glandular tract divided
into two portions at base of dorsal cirrus ; ligules obtusely pointed, more than half the
length of the dorsal cirrus (ventral ligule rather shorter than the two dorsal ligules
and narrower at its base). Dorsal setae homogomph spinigerous and falcigerous ;
superior ventral homogomph spinigerous and heterogomph falcigerous ; inferior ventral
heterogomph spinigerous and falcigerous. The appendix of a dorsal falciger is guarded
and unfringed ; its extremity is boldly hooked, and at the vertex of the hook there is
a small tooth in front of which the guard makes another slight projection (Plate IV.,

274 CEYLON PEAEL OYSTER REPORT.

fig. 92). The appendix of a superior ventral falciger in this foot is fringed along its
lower portion, guarded along its upper half (Plate IV., fig. 93). The appendices of
the other ventral falcigers of this foot and of the other feet, and also in many other
species of Nereidae, are minutely guarded quite at the apex beyond the fringe ; a
double contour at this point indicates a rudimentary guard, as may be realised by
comparison with the structure of the superior ventral setae here described (Plate IV.,
fig. 93).

The parapodium (44th) from the middle region of the body described above closely
resembles that figured by Gravier for his Platynereis insolita (' Arch. Mus. Paris '
(4), iii., 1901, p. 198), and, in fact, could not be distinguished from it. The only
difference seems to lie in the structure of the appendices of the dorsal falcigerous setae
which have a smooth rounded vertex without any prominence in P. insolita (loc. cit.,
p. 198, fig. 20G). The same setae with tooth on the vertex occur also, two in number,
in the 1 9th foot of the Ceylon specimen.

The tentacular cirri are long, the third pair extending over 16 segments. The
proboscis was retracted and had to be dissected out : — Paragnaths of the orders L, II.,
III., and V. absent; of the order IV. (Plate IV., fig. 94) in several more or less
complete rows (as in the type). Antennae as long as the palps, eyes in a trapezium.

Family: ONUPHID^.

Diopatra amboinensis, Aud. & M.-Edw. — Plate IV, figs. 95 to 97.

Station V., off Chilaw Paar, 11 fathoms. — One specimen with tube.

Another empty tube was taken from the Muttuvaratu Paar, 29th March, 1902.

The tube is characteristic, encrusted with unequal pieces of coarse broken shells
which stand out at right angles to the wall of the tube, being attached thereto by
one edge, giving the impression of pieces of shell threaded together ; some of the
fragments are nearly entire shells, as much as f inch long, the majority are smaller ;
seaweed may grow on the shelly covering ; the tube is from 3 to 3|- inches long, and
presents a tabulate appearance owing to the disposition described above.

The body of the specimen, which is incomplete behind, is flattened, 4 millims.
wide. The dorsal side is brown coloured, with a dark haemal line and a shining
cuticle ; the lateral parapodial tracts show up whitish in the preserved state ;
ventrally there is a pale neural tract flanked by two broad brown submedian bands,
followed on the outside by the series of broad whitish ventro-lateral tori or cushions
which resemble the uncinigerous tori of a Terebellid without the uncini, and doubt-
less serve to facilitate the passage of the worm up and down its tube. The first five
setigers are much larger than the rest, with a ventral and pc-rrect inclination, and
carry ventral cirri instead of tori. The spiral penicillate branchiae commence on the
sixth segment (i.e., fifth setigerous segment). Towards the fortieth segment they
become reduced and finally cease in the region of the sixtieth ; in the mid-region
they are more than twice the length of the stout, subulate dorsal cirri.

POLYCH.ETA. 275

The prostomial tentacles are somewhat pigmented, especially at the base, where the
long flagelliform portion is inserted into the 10- or 12-ringed basal peduncle. The
frontal antennas have a stout pigmented basal portion, followed by a narrower apical
portion, the whole as long as the peduncle of the lateral tentacle. The tentacular
cirri arising from the border of the single apodous buccal segment are as long as the
peduncles of the submedian tentacles, or slightly longer. The eye-tracts occur behind
the submedian tentacles. The broad rounded palps have a teat-like apex. In the
protruding jaws I counted 8 teeth on the right saw, 7 on the left, about 10 on the
right arc, 8 or 9 on the left, and 5 teeth on the left iinpar. The lower jaw-plates or
laminae ventrales have thin white calcareous end-pieces with simple border and
rounded divergent apices ; the mandibles have the usual form (Plate IV., fig. 95).

This species is closely related to Diopatra neapolitana. Its differentiation from any
other species of the genus will probably depend upon the character of the setae in the
anterior modified parapodia, about which our information is rather deficient, so far as
I am able to judge from the literature of Diopatra. It is equally closely related to
Diopatra semperi, Grube (' Ann. Semp.,' p. 282 = D. luzonensis, p. 138), but the
latter has very short frontal antennae and the calcareous caps of the lower jaw-plates
are trilobate ; the anterior setae are not described. The identification of the present
species cannot be considered final until fresh material for comparison is obtained from
Amboina or adjacent parts. A preparation of the first left foot of D. amboinensis
from Ceylon shows about 18 setae projecting. There is a dorsal group of about six
long, simple, curved acuminate, non-limbate capillary setae ; the rest have a bidentate
apex beyond which the pointed guard projects (Plate IV., fig. 96-97). The anterior
appendages possess other characteristic features besides the setae which they carry.
The lips of the orifice of the pharetra setarum are triangulate, an anterior short
truncate border followed by a long stout cirriform posterior median lobe and a
smaller ventral ligule ; I do not know of any other species of Diopatra in which this
third ligule has been described. Sometimes it is ajjpressed against the median lobe
and so concealed from view on the slide, but it can be found with a simple lens on
the body of the worm. In Grube's description of D. semperi (= D. luzonensis) the
diagnostic features of the anterior modified parapodia are not given. From the
parapodia of the mid- and hind-body two aciculae project with bifid apices as
figured by Ehlers (' BorstenwiArmer,' Taf. xii., fig. 15) for D. neapolitana. The
scalprate setae of the posterior segments have numerous fine denticulations (over 20),
whereas according to Ehlers they are few (about 8) in D. neapolitana. The fila-
ments of the branchiae are long.

Onuphis basipicta, n. sp. — Plate IV, figs. 98 and 99.

Station XXXV., Galle, 7 fathoms. One specimen in company with Glycera
lancadivce. Length 23 millims., width 1 millim.

Patches of brown pigment on the bases (ceratophores) of the antennae ; submedian

2 N 2

276 CEYLON PEARL OYSTER REPORT.

antennae longer than the median (tentaculnm impar) ; ceratophore of the median
antenna normal, nearly one-third the total length of the antenna, less than one-half
the length of the ceratophores of the submediau antennae which are nearly one-half
the total length of these appendages ; ceratophores of the lateral antennse about
three-fourths the length of the submedian ceratophores, about two-thirds of the total
length of the lateral antennae.

The above proportions, the moderate length of the median and the considerable
length of the submedian and lateral ceratophores are characteristic. The submedian
antennae stretch back over ten segments. The tentacular cirri were lost, one was
observed detached. Setae of the body -segments comprise 6 to 8 simple capillary with
two bidentate guarded acicular setae and three acuminate aciculas ; a small bundle of
fine capillary setae passes into the dorsal cirrus, sometimes extending far along the
cirrus. Dorsal cirri normal, subulate. Branchiae commence as a simple filament on
the first setiger and remain simple for 7 or 8 anterior segments, then becoming-
pectinate with 5 or 6 processes.

The first setiger contains the dorsal bundle of setae for the dorsal cirrus and a
ventral fascicle comprising one simple capillary seta, six characteristic tridentate
compound setae and an acicula (Plate IV., figs. 98 and 99). The second setiger
resembles the first, but contains in the ventral fascicle three simple setae and five
compound setae of the same type as in the first. The posterior lingule of these
anterior parapodia is long, acuminate, cirriform.

This species is determined by the combination of characters afforded by the
ceratophores, setae, and branchiae. It occupies an intermediate position between
0. longissima, Gr., and 0. teres (Ehlers). The shape of the frontal antennae is also
important ; these are broad, almost foliaceous, divergent, subtriangular appendages,
neither filiform, fusiform, nor subulate. Their shape is liable to vary in different
states of contraction and only the combination of characters can be relied upon. The
setae give the best indications, the peculiar setae of the anterior parapodia and the
simple setae of the body -segments.

Onuphis conchylega, Sars (1835).

Localities : — South of Manaar, 8 to 9 fathoms ; and from Station LIL, Cheval
Paar, 3 to G fathoms.

It is somewhat surprising to find this species which is common at Plymouth
occurring also off the coast of Ceylon. Its flattened tube, covered with large hori-
zontally placed shell fragments, is so characteristic that it is impossible to name it
differently from the type. This species should be the type of the unfortunate genus
North ia, of Johnston. I have carefully unravelled the confusion which this genus
has caused, and the result is that I agree with the Baron de St. Joseph that it should
be dropped. The species has wide distribution (see Ehlers, ' Florida- Anneliden,'
1887, p. 73) and is synonymous with Diopatra eschrichti ((Erst.).

POLYCtL-ETA. 277

The shelly tube is 40 millims. long by 10 mill'ims. wide. The tentaculum impar
stretches back over 12 segments; the prostomial tentacles are smooth, the basal
joints short, subequal, pauci-annulate. The tentacular cirri are slender, elongated,
-acuminate, much longer than the narrow buccal segment, about as long as the
prostomium. Branchiae simple, commencing on the 8th setiger ; stout, acuminate,
appearing flattened towards the base. Anal cirri long, slender, filiform. First three
setigers porrect, the first dorsal cirrus shorter than the following ; the fourth setiger
slightly porrect. Ventral cirri of first and second setigers subulate ; of the third,
shorter, obtuse, and fleshy ; of the fourth reduced to a slight projection of the first
torus ventralis.

A large female measures 30 millims. long by 3"25 millims. wide; a smaller specimen
is 17 millims. in body -length (excluding the antennae), and the branchiae are flattened.
The frontal antennas are short, ovate. In the first foot there are no setae associated
with the dorsal cirrus ; the setae of this foot comprise two acuminate aciculae, three
bidentate acicular setae which project, and one or two bidentate quasi-compound setae.
A gill-bearing foot carries six limbate capillary setae, two aciculae, and two bidentate
acicular setae, the prongs of which are gaping and subequal ; there are no setae
passing to the dorsal cirrus.

Onuphis dibranchiata, n. sp. — Plate IV., fig. 100.

Locality : — Galle lagoon, low-tide, 3rd August, 1902. Several fragmentary specimens.

The annulate bases of the prostomial tentacles are well developed, nearly equal, the
submedian slightly exceeding the others, and the median less than the rest ; the
submedian peduncles are less than one-third of the total length of the appendage ;
the median peduncle is barely one-fourth of the total length of the tentaculum impar,
and the lateral peduncles are about two-thirds the total length of the lateral antennae.
The submedian peduncles show twelve annulations in addition to the distal collar-like
portion into which the flagellum is inserted.

In the above proportions it will be noted that the lateral antennae resemble those
of Onuphis basipicta. The frontal antennae are short and fusiform. The subulate
tentacular cirri are inserted behind the submedian antennae and slightly exceed the
length of the buccal segment. The branchiae commence as a simple filament on the
first foot and continue simple on the first 17 parapodia, becoming bifid thereafter and
considerably longer than the dorsal cirri. The first dorsal cirrus is tumid at the base,
rather shorter than the first gill-filament, equal in size and shape to the posterior
liffule of the first foot.

The setae of the first foot do not afford such satisfactory distinction when compared
with the corresponding setae of 0. basipicta as might have been hoped. There is a
bundle of internal setae associated with the dorsal cirrus ; then three stout aciculae
ending in narrow flexible points ; next several slender quasi- compound setae like those
of O. basipicta, but in some of them the third tooth is absent. Besides these setae

278 CEYLON PEARL OYSTER REPORT.

there are two stout setae, tridentate (one damaged), Avhich are not compound
(Plate IV., fig. 100). These clearly represent the acicular setae in the first foot. A
parapodium from the anterior half of the body contains three aciculae, a bundle of
simple capillary setae, two bidentate acicular setae, and a small bundle of internal
capillary setae associated with the dorsal cirrus, passing nearly to the apex of the
cirrus. The width of the worm is 3 millims. ; the median antenna stretches back
over 7 segments.

Onuphis holobranchiata, Marenzeller — Plate IV., fig. 101.

Station V., off Chilaw, Gulf of Manaar, 11 fathoms.

One specimen contained in a membranous tube coiled up between the valves of a
Lamellibranch, adhering to one of the valves (Plate IV., fig. 101). The length of
the coiled mass was 13 millims. ; calcareous particles adhered to the borders of the
tube ; the whitish end of the tube projected beyond the posterior edge of the shell,
like a siphon, to the length of 21 millims. ; diameter 2-25 millims., a few shell-
fragments and other calcareous ddbris attached. No part of the worm protruded
from the mouth of the tube. After extraction of the anterior portion of the worm,
which broke away from the rest of the body, it proved to be an Onuphid allied to
Onuphis holobranchiata, Marenzeller.

The submedian ceratophores show about 10 annulations and a terminal collar ;
they are less than one-fifth the total length of the appendages ; half as long again
as the median ceratophore and equal to the lateral. The lateral ceratophores are half
the total length of the appendages to which they belong. The median antenna
stretches back over 21 segments, but the body has a dorsal flexure in the preserved
state ; the submedian antenna? are longer than the median. The frontal antennae are
obtusely subulate, coloured brown at the base. The tentacular cirri are shorter than
the buccal segment, which is rather longer than the prostomium, and has a concave
anterior dorsal border.

The first four parapodia are porrect, decreasing in size to the fourth, and occupying
a ventral position. After the fourth, the line of parapodia bends up in a conspicuous
arc to a more dorsal position. The first foot is not greatly enlarged, not longer than
the buccal segment ; the dorsal cirrus and posterior ligule are equal and similar and
longer than the bulk of the foot. The unifilar branchiae commence on the first foot.
The first foot contains a small bundle of fine setae passing to the dorsal cirrus, two
aciculae and four tridentate compound setae. A parapodium from the region between
the 40th and 50th segments contains, besides the internal setae of the dorsal cirrus,
four or five capillary setae, three scalprate or comb setae, four aciculae, and two
bidentate acicular setae.

There are two discrepancies between the above description and that furnished by
Marenzeller (' Siidjapan. Ann.,' i., 1879, p. 24), and, in addition, the peculiar
character of the habitaculum has not been described before. According to

P0LYC1LETA. 27 U

Marenzeller, the buccal segment is three times as broad as long, and the tentacular
cirri equal it in length. This feature may be attributed to the method of preserva-
tion. The second point relates to the lateral ceratophores which according to the
author of the species are nearly twice as long as the submedian. This may be an
individual variation, an abnormality, or the distinction of a local race. Unfortunately
the length of the lateral ceratophore, in proportion to the total length of the antenna,
is not stated. The calcareous ends of the laminse ventrales of the jaws show an
emarginate frontal border, agreeing closely with von Marenzeller's figure.

Hyaliiioecia camiguina, Grube.

Grube, 'Ann. Semp.,' 1878, p. 142.

C. Crossland, " Maldive Polychajta," 'P. Zool. Soc. London,' 1904, p. 281.

Localities : — Several specimens, Station XXV., f mile west-north-west of Foul
Point, Trincomalee, 8 fathoms. Station XLIIL, off Kalutara (Kaltura) ; one specimen
in company with Pectinaria tubes. Station XLV. , off Panadure ; several specimens.

The tubes, 65 millims. to 70 millims., are faintly ringed at intervals of about
2 millims., and sometimes the annulation shows the double crescentic arrangement
described by Grube. In the Ceylon specimens, the branchiae commence almost
constantly on the 22nd foot (23rd segment) ; once I found the first gill on the right
side on the 21st foot. 1 only observed aciculas with simple blunt rounded apices
protruding from the first modified parapodium, but in the second foot may be found
bidentate aciculae and a compound seta as figured by Crossland, as well as acuminate
forms. This species is closely allied to H. tubicola, of which it is clearly a local
representative. A full account of the hyaline tube which is guarded internally by
valves has been given by Mr. Arnold T. Watson (" On the habits of Onuphidse,"
' Trans. Liverpool Biol. Soc.,; vol xvii., 1903, p. 303).

Family: EUNICLLVE.
Eunice afra, Peters (1855).

Crossland, " Polychceta of Zanzibar," 'P. Zool. Soc. London,' 1904, vol. i., p. 289.
(1.) Station V., off Chilaw, Gulf of Manaar, 11 fathoms. In this specimen
branchiae commence on the 19th foot, becoming successively 3-filar, 4-filar, and 6-filar,
decreasing behind and leaving about 30 abranchiate posterior segments ; they are
inconspicuous, barely stretching halfway to the mid-dorsal line ; body flattened and
ribbon-like behind the anterior branchiferous region; length 75 millims., width
5'5 millims.

(2.) Galle, 14th February, 1902. Fragment of anterior region; gills begin as a
bud on the 16th foot and rapidly attain the maximum of 5 filaments ; calcareous caps
of laminse ventrales present ; pale collar on the fourth setigerous segment ; prostomial
tentacles with violaceous annulations ; body-colour of preserved specimen dark olive-
green ; width over the setae, 5 millims.

280 CEYLON PEARL OYSTEE REPORT.

Eunice antennata, Savigny.

Crossland, "Polychajta of Zanzibar," 'P. Zool. Soc. London,' 1904, vol. i., p. 312.

Locality : — South-east Cheval Paar, on pearl oysters, very numerous.

The pectinate branchiae commence on segment VIII. ((5th foot), the first gill is
multifilar ; they diminish towards the middle and then increase again both in size
and number of filaments posteriorly. In some specimens small branchiae commence
on the 5th foot, in others on the 7th foot, very rarely on the 3rd and 4th foot.
Antennas and cirri moniliform. The acicular setae show a third smaller tooth over the
two main prongs ; the compound setae of the posterior feet are characterised by the
presence of a third denticulation in a similar position to that in the acicular setae.

Eunice coccinea, Grube.

Crossland, "Polychseta of Zanzibar," 'P. Zool. Soc. London,' 1904, vol. i., p. 297.

Station V., off Chilaw, Gulf of Manaar, 11 fathoms, two specimens.

The differentiation of this species from E. afra depends chiefly upon examination
in the fresh condition. The colour is leucostict, but E. leucosticta is regarded as a
synonym of E. afra. There is a pale collar on the 6th segment, so there is in
E. collaris, which is another synonym of E. afra. Branchiae commence on 13th foot,
biramous on 15th, triramous and greatly exceeding the dorsal cirrus from 20th to
50th, long unifilar on 60th, decreasing and ceasing about the 90th, leaving upwards
of 70 posterior segments without branchiae and with rudimentary dorsal and ventral
cirri. Another broken specimen had quadrifilar branchiae. Complete specimen small,
coiled, about 45 millims. long, width of anterior gill-bearing region 3 millhns. ;
segments about 170 ; compound setae normal; acicular setae bidentate.

All characters are misleading except the shape of the body ; the head is narrow,
the body widening out markedly in the anterior region ; the hind-body is rounded,
not flattened ; the intersegmental grooves are coloured brown. The general form of
the body decides me to refer the present specimens to E. coccinea. E. coccinea
apparently bears the same relation to E. afra that E. murrayi does to E. antennata.

Eunice indica, Kinberg.

Gravier, op. cit., 1900, p. 242; Crossland, op. cit., 'P. Zool. Soc. London,' 1904, p. 318.

Station V., off Chilaw, 11 fathoms, one small specimen, incomplete behind.

Prostomium seen from above with frontal border entire ; tentacles smooth ;
compound setae with projecting guards, best seen in the branchial segments ; acicular
setae from the postbranchial segments with trifid tip, the apex ecpually bidentate, and
a large subapical tooth.

Another specimen was associated in the trawl with Phyllochcetopterus ramosus from
Galle, Station XXXVIIL, 17th February, 1902, depth 22 fathoms.

POLYCH^ETA. 281

Eunice martensi, Grube — Plate IV., figs. 102 to 104.

(tRube, "iMitth. iiber Euniceen," 'Schles. Ges.,' 1877; Breslau, 1878, p. 24.

Locality : — One specimen, labelled " Purple violet Eunice," taken from Buoy, Galle.

The dorsum of the 6th segment is pale ; segments streaked longitudinally
violaceous; antennae and cirri banded violaceous and white. Segments 178, incom-
plete behind; length 100 millims. ; width (ventral) over the feet 8 millims., over the
seta? 10 millims. Antennas nearly smooth, quasi-articulate, cirri smooth ; tentaculum
impar and submedian antennae equal to seven segments, but the former ends bluntly
and has probably been longer ; dorsal cirri long throughout and smooth.

Branchiae begin as a bud on the 7th segment (5th foot), plurrfilar on the 8th, in
subsequent segments acquiring 14 filaments; in the 17th segment the branchial stipe
is shorter than dorsal cirrus, in the 28th it is equal to the cirrus, and in the 48th
longer than the cirrus (Plate IV., fig. 102). The general distribution of the branchiae
and the number of filaments will be understood from the following enumerations, the
Roman numerals giving the number of the segment : —

VII, VIII. , XVII. , XXVIII. , XL VIII. , LX, LXXL, CXIV., CXL., CLXX.
1. 4. 13. 13. 14. 14. 14. 12. 10. 5.

In the anterior segments there are 8 or 9 dorsal capillary setae and about 20
compound falcigerous bidentate setae in the ventral fascicle (Plate IV., fig. 103) ; the
two-pronged ventral acicular setae commence at the 35th foot (Plate IV., fig. 104);
comb-setae (scalprate setae) more numerous posteriorly, with about 12 denticulations.

Jaws : — P. II.— 6 ; L. II.-6 ; L. III. (the unpaired sinistral jaw-piece)-7 ; R. IV.-9 ;
L. IV. -4.

This species belongs to that section of Eunice characterised by multifilar branchiae
and long dorsal cirri extending from end to end of the body. The type of this section
may be taken to be E. tentaculata, Quatrefages (1865, i., p. 317), from South
Australia (Bass Strait, Tasmania). Another example is the original E. clseyi, Baird
(1870), from Queensland (not E. clseyi, McIntosh), which is synonymous with
E. acquahilis, Grube, 'Schles. Ges.,' 1877, Breslau, 1878, p. 24, from Cape York.

Eunice murrayi, McIntosh.

Crossland, "Polychajta of Zanzibar," op. tit., 1904, p. 310.

(1.) Station XL V., off Panadure (Pantura), depth 25 fathoms, inhabiting a tube on
the under side of a Spondylus valve ; a polynoid was in the same tube.

Prostomium wide, frontal border deeply emarginate ; antennae moniliform ; the
tentaculum impar longe-articulate, stretching back over 18 segments; gills with
12 pinnae; guards of compound setae slightly projecting; total width in anterior
region, 4 millims. The tube which accompanies the specimen is encrusted with sand
and Foraminifera.

(2.) South-east Cheval Paar ; occurring with Eunice antennata.

2 o

282 CEYLON PEARL OYSTER REPORT.

Branchiae commence unifilar on segment V. (3rd foot), tri filar on VI. and VII.,
sexfilar on VIII. , then multifilar ; they decrease rapidly after the 30th foot and cease
after the 43rd. Antenna? snbmoniliform, impar stretching over 11 segments. Setae
normal ; acicular setae trifid, as many as four in the posterior segments.

Eunice siciliensis, Grube.

Graviek, " Ann. Mer Rouge," ' Arch. Mus. Paris ' (1 ser.), ii., 1900, p. 261.
Ckossland, 'P. Zool. Soc. London,' 1904, p. 323. [= Eunice valida, Gkavier.]

Station V., off Chilaw, Gulf of Manaar, 11 fathoms. Also from South-west Cheval
Paar, 13th November, 1902.

Branchiae commence on the 62nd setiger and remain simply filiform throughout.
This well-known and widely distributed species can be recognised by the incurved
rami of the lower jaw-plates.

Eunice tubifex, Crossland.

Station LIX., Muttuvaratu Paar, without tube.

This is a worm of large size, one specimen measuring 120 millims. by 5 millims.
It is distinguished by the occurrence of spinigerous setae in the anterior segments and
falcigerous setae in the posterior region. I have nothing essential to add to the
excellent description given by Mr. Crossland (' P. Zool. Soc. Lond.,' 1904, p. 303).

The branchiae commence as a small bud on the 20th segment and continue as a
single filament, increasing in length to the 45th segment, after which a second filament
appears. By the 100th segment there are still only three filaments; by the 200th
there are four long filaments. Further back the number of branchial filaments
increases to five, all arising from a very short axis ; filaments subequal, much longer
than dorsal cirrus ; the latter increases somewhat posteriorly.

Some statements in a footnote on p. 308 of Mr. Crossland's paper require to be
modified. Schmarda's account of Eunice depressa from Auckland, New Zealand,
which also possesses the two kinds of compound setae, is fuller than many of his
descriptions ; this worm is, however, a Marphysa, since, as shown and stated by
Schmarda, tentacular cirri are absent. Grube named two species of Eunice with
spinigerous setae from the Philippines, namely, E. impexa and E. megalodus, but
records no distinction between the anterior and posterior setae.

There is a headless fragment over 150 millims. long accompanying the specimen
above described.

Marphysa chevalensis, n. sp.

Locality : — South-east Cheval Paar, Gulf of Manaar.

This species belongs to the same group as Marphysa depressa (Schmarda, ' Neue
Wirbellose Thiere,' ii., 1801, p. 127, New Zealand) and M.fallax, Marion et Bobretzky

POLYCH/F.TA. 283

(Ann. Marseilles/ 1875, p. 13). Its interest lies in its occurrence here, not in its
supposed differences from its two allied forms, since there are no differences which are
figurable. The character of the group depends upon the presence of two kinds of
compound setae, spinigerous and falcigerous, in the parapodia. The special feature
of the Ceylon representative depends upon the distribution of the branchiae.
Length 25 millims., width 275 millims. ; 104 setigerous segments.

The frontal border of the prostomium is emarginate and the division is continued
as a shallow median groove along the dorsal surface of the head. Branchiae com-
mence as a small filiform process on the 11th foot, uniramous from 11th to 19th foot,
biramous from 20th to 25th, triramous from 26th to 68th, decreasing and ceasing at
the 85th foot. Dorsal cirri fusiform, decreasing in size posteriorly. Anal cirri
subulate, smooth, with ring of brown pigment at base ; ventral anal styles minute
filiform. Compound setae, falcigerous and spinigerous, the spiniform appendices
longer and shorter ; both kinds of setae occur together from the first parapodium, but
in the specimen now being described only the spinigerous form was observed in the
posterior segments.

In the anterior parapodia a dorsal fascicle of limbate capillary setae, a central
group of three aciculae and the ventral fascicle of compound dimorphic setae ; in the
posterior parapodia the dorsal fascicle contains limbate and scalprate setae ; there
is a central group of two aciculae, a ventral fascicle of compound setae and a ventral
bidentate acicular seta. A second specimen, darker coloured, anal cirri black, showed
both kinds of setae in the posterior feet ; branchiae from 10th to 55th foot ; maximum
number of three filaments observed on one foot only ; length 9 millims. ; width
nearly 2 millims ; segments about 73.

Paramarphysa orientalis, n. sp. — Plate IV., fig. 105.

Locality : — South-east Cheval Paar, Gulf of Manaar. Several specimens.

Length 21 millims., width 1 millim. ; 91 setigerous segments.

Prostomium almost entire, a shallow groove only dividing the front ; antennae
short, about the length of the prostomium, tentaculum impar somewhat longer.
There is an eye-spot on each side of the head between the bases of the submedian
and lateral antennae. Bidentate ventral acicular setae begin on the 26th foot ; a
segment or two behind this the main aciculae become dark coloured ; there may be
two acicular setae in the posterior parapodia. The superior fascicle of capillary setae
shows slight difference in anterior and posterior regions, the limbate tract being
shorter and more convex behind ; comb-setae with long marginal laciniae occur as far
forwards as the 9th foot. The compound bidentate falcigerous setae are normal, not
markedly different in front and behind, distinguished by the slight inflated appearance
of the end of the shaft (Plate IV., fig. 105).

Only one other species of Paramarphysa has been described to my knowledge,
namely, P. longula, Ehlees ('Florida Ann.,' 1887, p. 99).

2 o 2

284 CEYLON PEARL OYSTER REPORT.

Lysidice collaris, Ehrenberg and Grube.

Grube, 'Ann. Semp.,' 1878, p. 166; Gravier, 'Arch. Mus. Paris' (4 ser.), ii., 1900, p. 272.
Crossland, ' P. Zool. Soc, London,' 1904, p. 284.

Station V., off Chilaw, Gulf of Manaar, 11 fathoms. One specimen; length
about 15 millims. Several other specimens in the collection.

Nematonereis unicornis, Schmarda.

Schmarda, ' Neue Wirbellose Thiere,' ii., 1861, p. 119. (Atlantic Ocean.)

Locality : — South-west Cheval Paar, Gulf of Manaar. Two specimens.

In the smaller specimen the head is entire and rounded ; in the larger slightly
emarginate ; a pair of large eyes at the hack of the head and between them a small
occipital tentaculum impar about the same length as the prostomium. There are no
tentacular cirri ; no branchiae ; two achaetous buccal segments. The larger specimen
is a mature female.

In the parapodia the dorsal cirrus is short, subulate, about as long as the pharetra
setarum ; a central acicula separates the setae into superior and inferior groups,
the former consisting of simple capillary setae, the latter of compound, bidentate,
falcigerous seta?. In the posterior bundles there is also a guarded bidentate acicular
seta, the subterminal tooth much larger than the apical tooth. The jaws are pale
and normal : — II. 5-4, III. 4.

Family : LUMBRICONEREIDiE.
Aglaurides fulgida (Sav.) — Plates IV. and V., figs. 106 and 107.

Gulf of Manaar. Several specimens.

The largest measured as much as 7 inches in length, a few segments missing
behind ; dorsum very convex ; width 5 millims. ; width including setae nearly
9 millims. ; segments narrow and very numerous ; dorsal cirri large, foliaceous ; head
rounded to triangular in shape.

The characterisation of the cirrobranchiate Eunicea is at present somewhat obscure,
and it is not rendered less so by the fact that the present species agrees identically
with Schmarda's GZnone diphyllidia from Jamaica in respect of the size and shape
of the dorsal cirri, the buccal and the parapodial armatures.* (Plate V., fig. 107.)

The buccal segment is of double nature, as in Lumbriconcreis, the intersegmental
groove being plainly visihle at the sides and continued forwards to the edge of the
mouth, so that the processus oralis of the second segment borders the mouth helow.
Dorsally the buccal segment lies over the back of the prostomium like a collar, the
groove between the collar and the head deepening into a profound sinus, in which
three minute papillifbrm occipital tentacles are occluded. In front of the tentacles
there are four eyes, two large lateral eyes and two small submedian eyes, the latter

* Cf. Ehlers, ' Florida-Anneliden,' 1887, p. 109, Taf. 34.

POLYCH/TCTA. 285

sometimes simple, sometimes multiple (Plate IV., fig. 106). The arrangement of the
eyes is the same as in CEnone diphyllidia, and they can easily he found hy raising
the margin of the collar. The presence of the antenna? or occipital tentacles is the
one feature in which Aglaurides is helieved to differ from CEnone, and in view of the
ahsolute identity of the other parts this divergence appears remarkable.

In Aglaurides, especially if the material he well preserved, the mere pulling hack
of the collar will not always suffice to discover the prostomial tentacles. I have
found it necessary to divide the collar hy a longitudinal incision before the antennae
come into view. They are extremely fugitive structures, not too easily found even in
half-macerated examples, so deep do they lie within the recesses of the nuchal sac.
In Aglaurides erijthrfeensis recently described by Gravier ['Arch. Mus. Paris' (4th
ser.), t. ii., 1900, p. 278] from the Red Sea the character of the jaw-pieces seems to
differ from that of the corresponding parts in A. fulgida. The reader is left with
the impression, that while " les pieces correspondantes sont plus developpe'es a droite
qu'a gauche," yet on the whole they are subequal, and only one half is figured on
the plate. Otherwise the species is indistinguishable from Aglaurides fulgida, the
shape of the head varying as mentioned above. The author does not mention the
radices maxillarum, which in A. fulgida are very long, longer than the rest of the
upper jaw, as in Maclovia and Holla, not short as they are in Lumbriconereis and
Lysarete.

In addition to the characters presented by the antennae and the details of the
jaws as described and figured by Ehlers (' Borstenwurmer,' 18G8, p. 408) the genus
Halln (= CirrobrancJiia, Ehlers) differs from Aglaurides in having two entire
apodous segments behind the head, a peculiar processus oralis being furnished by the
first segment. A. fulgida has been recorded from Ceylon by Michaelsen, ' Jahrb.
Hamburg. Wiss. Anst.,' ix., 2, p. 99).

Aracoda obscura, n. sp. — Plate V., figs. 108 to 112.

Station XLIIL, off Kaltura. A single specimen, incomplete behind, 16 millims.
long, L millim. wide, in an empty Pectinaria tube.

Colour nearly black, with strongly iridescent tough cuticle. Prostomium pear-
shaped or conical, nearly equal to the four next segments (Plate V., fig. Ill) ; two
distinct achtetous segments in front, without processus oralis of the second segment.
The usual long posterior ligule on the parapodia. There are six or seven simple
curved limbate setae in the parapodium ; in some the limbus is nearly plain, in others
denticulate as shown in Plate V., fig. 112. The dilated portion of the setas is finely
and obliquely striated.

The only way to identify the species of this and allied genera of Lumbriconcreida-
is by the character of the jaws. These are black and overlap, consequently they are
difficult to see in themselves and not easy to prepare from such a small specimen.
The principal feature in this case is provided by the jaw-pieces of the first pair which

286 CEYLON PEART- OYSTER REPORT.

are subequal and without a special terminal hamulus distinct from the dentate portion

(Plate V., fig. 108). The jaw-pieces of the second pair are the most difficult to

make out satisfactorily ; one side, so far as I could ascertain, showed three large

teeth, and a long edentulous shaft reaching back to the radices maxillarum (Plate V.,

fig. 110) ; on the other (probably the left) side the second jaw-piece, which was

fractured near the apex, showed a strong apical tooth and a long posterior dentate

portion (Plate V., fig. 109). There are, as usual, five pairs in all, the fifth being a

simple hamulus. As stated above, the diagnostic character is the absence of a

pronounced diastema between the apical tooth and the succeeding teeth of the first

pair of jaws.

Family: GLYCERID^E.

Glycera lancadivae, Schmarda — Plate V., figs. 113 to 116.

Schmarda, op. cit., 1861, p. 95 ; Michaelsen, ' Jahrb. Hamburg. Wiss. Anst.,' ix., p. 102.

Locality: — Galle, 14th February, 1902. Two specimens.

This species is a representative in these waters of the Atlantic Glycera capitata,
and is characterised by the absence of branchiae and by the biligulate anterior lip
of the parapodium, the posterior lip being rounded and slightly emarginate. It
appears to present a dimorphism analogous to that existing between G. capitata
and the variety G. setosa. In the typical form the parapodia are longer than in
the second form ; we may refer to these as A and B respectively. In both, the body
segments are biannulate, and there are two kinds of compound setae differing in the
structure of the articular end of the shaft ; in one kind the transparent guard
projects beyond the edges of the cup, in the other the guard is level with the cup
and shows fine sulcations (Plate V., figs. 113 and 114). In form A, the ligules are
fusiform and approximately ecpual in bulk, the ventral ligule rather longer (Plate V.,
fig. 115). In form B, the ligules are obtuse and the ventral ligule, besides being
longer, has about twice the bulk of the dorsal ligule (Plate V., fig. 116). If these
specimens had been taken at different times and places they would probably have
been described as separate species. The extruded proboscis (form B) is covered with
minute papillae arranged irregularly, and of two kinds, acuminate and rounded, the
former greatly predominating. Length of specimen of form A (incomplete behind),
upwards of 60 millims. ; form B, complete with subulate anal cirri, 45 millims. long,
including the proboscis (7 millims.).

Family : SPIONID^l. [See also p. 325.]
Polydora hornelli, n. sp. — Plate V., fig. 117.

From pearl oysters, Gulf of Manaar.

It is hard to say whether this species is really distinct from Polydora ciliata ; the
differences, so far as can be judged from preserved material and published records, are
very slight, but such as they are, they seem to afford ground for separation when
taken in conjunction with the geographical distribution and the character of the host.

POLYCH^ETA. 287

The general appearance of the anterior end resembles that figured by Carazzi* for
/'. ciliata. The ground colour is chocolate brown, and this pigment occurs upon the
peristomium and upon the greater part of the caruncle; the latter projects beyond
the front of the peristomium, terminating in two rounded lobes, divided from one
another by a shallow notch. The proportions of the caruncle are characteristic ; it is
inserted user the length of the peristomium and over the first two segments of the
trunk ; in the region of the first setigerous segment there is a slightly dilated pale
area causing an interruption of the brown pigment ; there is a patch of pigment in
the centre of this area which I shall call the ocular area, although I could not
distinguish definite eye-spots; on each side of the ocular area on the dorsum of the
first setigerous segment, midway between the caruncle and the margin of the body,
there is a very small cirrus ; at the outer margin of the same segment, on each side,
there is another cirriform appendage, below which occurs a capillary fascicle ; the
latter appendage is shown in Carazzi's figure, but not the former ; the portion of the
caruncle in front of the ocular area is longer than the portion behind it. The
remaining segments of the body carry notopodial and ueuropodial setae, the 2nd, 3rd,
4th and 6th segments with capillary seta? only. The notopodial setae of the 5th
setigerous segment are modified, consisting of an oblicpie row of eight large brown
acicular seta? with two not fully formed in reserve, each acicular seta being
accompanied by a delicate colourless spatulate seta. The acicular setae of the 5th
segment differ from those figured by Carazzi for P. ciliata; in P. hornetti these
seta? are not toothed, but a convex limbus or vane occurs below the curved apex
(Plate V., tig. 117). [See Mr. Arnold Watson's "Note" on p. 325— W. A. H.]

From the 7th segment the neuropodial series of capillary setae are replaced by a
single row of about a dozen bidentate guarded acicular setae not differing in structure
from those of P. ciliata as figured by Professor McIntosh in 1868.1 The branchiae
also commence on the 7th setigei'ous segment, but the material is so fragmentary
that I cannot say where they end. The diameter of a large specimen is 1'5 millims.
No other dorsal setae beyond the modified seta? and their spatulate attendants were
observed in the 5th segment, but a bundle of very delicate neuropodial setae occurs at
their base. The 5th segment is firm and the fore-body porrect. The 2nd, 3rd, 4th,
and 6th setigerous segments carry a short cirriform notopodial ligule behind the
dorsal fascicle. The long peristomial tentacles are lost from most of the specimens.

Family : CAPITELLID^E.
Notomastus zeylanicus, n. sp. — Plate V., figs. 118 and 119.
Locality : — East Cheval Paar, 8 fathoms.
Anterior fragment of small individual. The first two segments are achaetous ; the

* D. Carazzi, " Revisione del genere Polydora, Bosc, e cenni su due specie che vivone siule ostriche,"
' Mitth. Zool, Stat. Neapel,' xi., 1893, see Taf. ii., fig. i.

t See also T. Whitelegge, " Report on the Worm Disease affecting the Oysters on the Coast of New
South Wales," ' Rec. Austral. Mas. Sydney,' 1890, vol. i., No. 2.

288 CEYLON PEARL OYSTER REPORT.

eleven following segments are biannulate and carry dorsal and ventral fascicles of
capillary setae. The unciniform seta? commence on the 14th segment, the dorsal
torus containing a single row of 19 setse, the ventral torus a single row of 34
(Plate V., fig. 119). From the mouth projects a median tongue-like lobe. The. head
is half withdrawn into the buccal segment, but by pressing the latter backwards ac
acervus of eye-spots at the side of the head can be seen (Plate V., fig. 118).

Family: OPHELILD.E.
Armandia lanceolata, n. sp. — Plate V., fig. 120.

Locality : — South of Manaar Island, 8 to 9 fathoms.

One specimen about 17 millims. in length ; 29 pairs of parapodia, the last three or
four of which are very small ; 22 pairs of cirriform branchiae commencing on the 2nd
setigerous segment ; eleven pairs of lateral eyes situated in front of the jwapodia
from segments VII. to XVII. inclusive ; no cephalic eyes observed. This is a slender
worm 1 millim. in diameter ; body pale.

Ophelioids of this group show a striking resemblance to Amphioxus in shape, size,
habits, distribution, consistency, translucency, and movements. This has been
remarked previously by Lo Bianco (1893) in the case of Armandia polyopldhalma
and by me (1896) in New Guinea waters.* Although infinitely removed from each
other in morphology, Armandia and Amphioxus are closely approximated in bionomics.
It is a case of true homoplasy ; there is no question of affinity, nor of mimicry, nor of
parallel evolution.

The ventral musculature in this species is so arranged as to cause two prominent
longitudinal ventro-lateral ridges, like metapleural folds, extending uninterruptedly
from the mouth to the base of the anal siphon, leaving a narrow deeply depressed
median ventral tract, the neural groove, between them ; and a lateral groove above
on each side of the body in which the parapodia lie. The branchiae can be kept
lengthwise lodged within the lateral groove. The anal siphon is a narrow mem-
branous tube terminating in a circular orifice fringed by not less than 12 subulate
papilla?, subequal.

In Armandia leptocirrus, Grube ('Ann. Semp.,' p. 194) there are 33 pairs of gills, t
a pair to each of 34 segments excepting the first ; the lateral eyes are said to
commence on the fifth setigerous segment and to occur on the ten following segments.
Sometimes the eyes may be absent or lost from some of the ommatophorous segments.
In this specimen of A. lanceolata they are complete on the right side, but on the left
side missing from the 11th to 13th segments. The first eye occurs on the 7th
setigerous (6th branchiferous) segment, the last on the 17th setigerous (16th
branchiferous) segment. The anal siphon has a marked ventral flexure in the

* Cav. Lo Bianco, 'Atti Ace. Napoli,' v., 1893; A. Willey, 'Quart. Journ. Micro. Sc.,' vol. 39,
August, 1896, p. 219.

t Given as 22 pairs by a misprint.

POLYCH.ETA. 289

preserved specimen. The simple capillary setae issue in two bundles, a dorsal bundle
of five long setae and a ventral bundle of about eight shorter setae.

The anterior end of the body is produced as a transparent rostrum which
terminates in a white appendix (Plate V., fig. 120). The mouth lies below on the
level of the base of the rostrum, and the paired nuchal organs occur behind at a
distance from the mouth, equal to the latter's distance from the frontal extremity.
The long simple gills are traversed by blood-vessels.

Arniandia leptocirris, Grtjbe. ('Aim. Semp.,' 1878, p. 194.)

Locality : — Cheval Paar, Gulf of Manaar.

This species is twice as bulky as A. lanceolata. The essential differences are not
easy to specify by the examination of preserved material only. Length 21 "5 millims.,
diameter 2 millims. ; 37 setigerous segments ; 27 gills, commencing at the second
setigerous segment; 20 marginal anal papillae, larger and smaller intermixed; 14
lateral eyes, commencing in front of the 6th gill. Nuchal organs lie in front of the
mouth ; metapleural ridges commence at the sides of the mouth. The most striking
feature of the specimen here described is the occurrence of several small cirriform
papillae protruding from the buccal orifice. Notopodia and neuropodia minute,
contiguous.

Another example from Modragam Paar does not show the buccal papillae
protruding.

Polyophthalmus australis, Grube — Plate V., fig. 121.

Two specimens from east side of Cheval Paar, one measuring 15 millims. long by
1 millim. wide; the other 8 millims. by 0'5 millim.

The setae are extremely delicate and not obvious at first examination. The head is
rounded, the body with characteristic pigment markings figured by Grube (' Ann.
Semp.,' p. 196), and the anus is surrounded by a circlet of papillae. The body has the
consistency of that of an Ophelia. When placed in fluid of different density from
that in which it had been kept, it twitched convulsively like a Nematode.

In the smaller specimen there are 26 setigerous segments, capillary setae in separate
dorsal and ventral fascicles, the latter issuing from a convex crateriform elevation.
The ciliated apparatus of the head, not figured by Grube, is shown on Plate V.,
fig. 121. The lateral eyes were not distinct.

Family : FLABELLIGERID^E.

Stylarioides parmatus, Grube — Plate VIII., fig. 5.

Stylarioides parmatus, Grube, 'Ann. Semp.,' 1878, p. 199.

Stylarioides iris, Michaelsen, ' Jahrb. Hamburg. Wiss. Anst.,' ix., p. 108.

Muttuvaratu Paar. Four specimens, one very small.

The distinction between & iris and S. parmatus is based upon the number of long

2 P

290 CEYLON PEARL OYSTER REPORT.

and gorgeously iridescent setae of the second segment. These are disposed round the
head in four bundles ; each bundle may contain up to 10 setae. Grube found only
six setae in each bundle. The setae of the third segment are much more slender and
shorter than those of the second, but equally numerous, though the number of setae
in both these segments varies, probably in many cases through accidental loss. The
caudal extremity of the body may be recurved as in Pallasia, as noted by Mr. Watson,
but, unlike Pallasia, the segments are all setigerous (Plate VIII., fig. 5).

Family: MALDANID^.

This family may be divided into three sub-families : — Clymeninas, Rhodininae, and
Nicomachinae. For the identification of the worms belonging here it is essential that
the whole body should be preserved. There is a fragment of a Nicomachine Maldanid,
of which only the six anterior setigerous segments have been preserved, obtained in
14 fathoms to the west of the northern end of the Periya Paar at Station LXI.
In the absence of the anal segment it is unfortunately impossible to assign it to its
proper genus with any certainty.

Nicomache truncata, n. sp. — Plate V., figs. 122 and 123.

Its head is extremely characteristic, the front vertical with a median crest widening
out below into a transverse ridge which overhangs the mouth (Plate V., fig. 123).
The peristome is indistinguishably fused with the prostomium, and the two together
compose a remarkably short head. This is followed by a short first setigerous segment
with the capillary fascicle in the centre ; next a slightly longer segment with the
capillary fascicle nearer to the anterior than to the posterior border ; thereafter the
segments become elongated, the setae and tori occupying an anterior position in each
segment. Below the first and third capillary fascicles there is a large acicular seta ;
below the second there are three acicular setae on the left side, one only on the right
(Plate V., fig. 122).

Family : AMMOOHARID^.

Ammochares orientalis, Grube — Plate V., figs. 124 and 125.
Grube, 'Ann. Semp.,' 1878, p. 204.

Station LIV., south of Adam's Bridge, 4 to 40 fathoms. One specimen.

Total length 28 millims., width 1*5 millims., tapering to less than half this size
behind. Length of the laciniate gills, 2 millims. ; length of first segment 2 '5 millims.,
of second 3 millims., third 375 millims., fourth 3-25 millims., fifth 2-5 millims. Total
number of setigers 22 pairs ; the last eight segments are quite short and were detached
from the rest of the body. The first uncinigerous torus occurs below the fourth
capillary fascicle. The oesophageal nerve-collar becomes confluent with the ventral
nerve-cord in the region between the first and second fascicles (Plate V., fig. 124).

The first segment is a composite one, comprising three pairs of capillary fascicles

POLYCH^ETA. 291

unaccompanied by uncinigerons tori ; of these the third pair has not been noticed by
Grube (Plate V., fig. 125). On the dorsal side in the middle line, slightly in advance
of the first pair of fascicles, there is a peculiar nuchal organ flush with the surface of
the body (Plate V., fig. 125). The uncini are arranged in a polystichous manner in
the tori ; as many as 20 rows of peculiar minute uncini of a type best defined as
ammocharine ; when examined in situ they look like the separated teeth of multidentate
uncini. The capillary setae are linear, non-limbate, finely squamose towards the
acute apex.

Tube tough, mucoid, encrusted with sand grains which are arranged with considerable
regularity in an imbricating manner. Wider in front, tapering behind, the tube is
much longer than the contained worm, being about 60 millims. The lips of the anterior
and posterior orifices are approximated at their respective ends, so that the openings
of the tube appear closed. The specimen was taken in company with Panthalis
mgromaculata. The structure and habits of the Ammocharidae form the subject of a
paper by Mr. Arnold T. Watson in ' Journ. Linn. Soc.,' Zoology, vol. xxviii., 1901,
pp. 230-260, plates 23-25.

Family : CHvETOPTEKlD^E.
Chaetopterus appendiculatus, Grube — Plate V., fig. 126.

Grube, "Ann. Ceylon" (Holdsworth coll.), 'P. Zool. Soc. London,' 1874, p. 328.
Herdman, 'Ceylon Pearl Oyster Report,' Part I., 1903, p. 80.

Locality : — Station LXII., near Periya Paar, 7 to 13 fathoms.

There are ten thoracic segments, the last two carrying uncinigerous tori and the
last bearing a large aliform notopodium. The penultimate thoracic tori are separated
by a deep median ventral notch ; the last thoracic tori are confluent across the
middle line. Uncini of the last thoracic torus mostly 6-dentate, some 5-dentate ;
uncini of a ventral abdominal torus 8 -dentate.

In Grube's original description, which was based upon a large specimen 124 millims.
long, the tube measuring 364 millims. by 22 millims. in diameter, the uncini of the
posterior ventral tori are said to have nearly 20 teeth. This is a large number for
Chaetopterus, twice the normal number,* and might conceivably be due to the uncini
lying appressed, end to end, as they frequently do.

Of the thoracic parapodia the first, fourth, and ninth are the shortest in the
specimen under examination. The fourth contains seven broad modified setae of the
type characteristic of the family. In this case these setae show a finely crenulate or
beaded distal border (Plate V., fig. 126). The fore-body of the specimen measures
5 "25 millims. long by 3 "5 millims. broad, hence much smaller than Marenzeller's
Ch. cautus from Japan, where the corresponding measurements were 20 millims.
by 10 millims., and there were 20 to 30 chaetopterine setae in the fourth parapodium.

* Cf. Crossland, "Maldive Polychajta," 'P. Zool. Soc. London,' 1904, vol. i., p. 276.

2 P 2

292 CEYLON PEARL OYSTER REPORT.

The tube is pergamentaceous, finely fibrous, thinly encrusted with mud and calcareous
minutiae. The posterior notopodia adhered firmly to the inner wall of the tube and
were greatly elongated.

Phyllochaetopterus herdmani (Hornell) — Plate V., figs. 127 to 132.

Spioclwlopterus herdmani, Hornell, ' Ceylon Pearl Oyster Report,' Part I., 1903, p. 16.

Locality : — Galle shore, under stones.

Narrow cylindrical tubes encrusted with relatively coarse sand-grains and hard
fragments of all kinds, including Foraminifera. The worms outside their tubes are
soft and convoluted, and the hinder poi'tion of the body is very fragile. The three
regions of the body behind the buccal region are the thoracic region, with ten (in one
case nine) setigerous segments ; the branchial region, consisting of two segments ;
finally the abdominal region with as many as 40 to 50 segments. The ventral wrall
of the thoracic segments is thickened to form a large glandular cushion, and the
lateral wall of each abdominal segment is marked by a long narrow brown tract
between the dorsal and ventral divisions of the parapodium on each side ; sometimes
this pigment tract is very dark. Above its dorsal end occurs the clavate notopodium,
and below its ventral end the neuropodium, subdivided into two narrow uncinigerous
lobes. The head is surrounded by an incomplete collar, open dorsally (Plate V.,
fig. 127) ; above and behind the dorsal ends of the collar occurs the second pair of
tentacles, supported by a bundle of three long delicate internal setae. The long
s})irally coiled tentacles of the first pair are inserted immediately in front of the
second pair, and are still retained in some of the specimens.

The first parapodium contains about twenty spatulate or vane-tipped setae, arranged
in general in a single row, forming a dorsiventral monostich. The length and width
of the vane varies ; sometimes it terminates in a point, generally its border is broken
up into shreds (Plate V., fig. 131). The second parapodium resembles the first,
and so does the third as a rule, but in one specimen there are five modified setae
(resembling those which usually occur only in the fourth parapodium) intercalated
between the normal spatulate setae (Plate V., fig. 129). The fourth parapodium
contains, in addition to the spatulate setae, eight or nine modified flattened setae
(Plate V., fig. 128). The remaining thoracic parapodia contain the usual setae. A
notopodium from the second branchial segment contains 16 or 17 long internal setae
in the inner division of it. The plan of a branchial segment is shown on Plate V.,
fig. 130. The borders of the notopodium below the distal expanded bifid portion
are densely ciliated. The dorsal body-wall of the branchial region appears to be
glandular.

Next to the notopodium comes a gill, a thin membranous expansion folded upon
itself in the preserved specimen ; this passes below into a wing-like dermal fold with
free surfaces and a thickened outer border, which is the uncinigerous torus. The
unciui in a neuropodium of a branchial segment have 9 to 11 denticulations ; when

POLYCH.ETA. 293

9 or 10 occur they are distinct, when 11 are present the apical tooth is very small.
The abdominal notopodia carry a bundle of four spatulate setae. The abdominal
uncini show 10 or 11, perhaps sometimes 12 denticulations. All the uncini
show the peculiar ribbed structure noted by previous authors in other species
(Plate V., fig. 132).

This species is closely related to PhyllochcBtopterus acicidigerus, described by
Crossland from the Maldive Group (' P. Zool. Soc. Lond.,' 1904, p. 278). The
aspect of the anterior region varies greatly according to the state of the preserved
material ; when well protracted and hardened, it resembles the condition figured by
Crossland for P. elioti from Zanzibar (cf. Plate V., fig. 127), except for the absence
of eyes in P. herdmani.

In nine specimens there are ten pairs of thoracic parapodia, in one specimen nine
pairs, the latter being the number given for P. aciculigerus. The principal
substantial difference appears to be due to the presence in the last-named species
of a glandular cirrus of uncertain homologies arising from the centre of the neuro-
podial tori of tbe first branchial segment.

Phyllochaetopterus ramosus, n. sp. — Plate V., figs. 133 to 136.

Locality : — A thick cluster of slender brown translucent ramifying tubes (Plate V.,
fig. 133), overgrown and welded together by foreign incrustations of other tubicolous
and colonial organisms and vegetable growths, trawled off Galle, at Station XXXVIII.

A fairly complete worm measures 25 millims. in length, its tube twice this length,
and l-5 millims. in diameter. The first region of the body consists of 15 segments;
the fourth parapodium contains one, sometimes two broad flattened modified setae
(Plate V., figs. 134 and 135). The second or branchial region comprises about
16 segments. As the worms were all preserved inside their closely investing tubes,
the branchial region is not preserved sufficiently well to show details clearly.
Practically it may be distinguished from the third or abdominal region by its long
styliform notopodia, each supported by not more than four internal slender setae.
The abdominal notopodia are clavate papillae, each supported by a single seta. The
long tentacles are coloured with brown patches which give them a banded
appearance ; they are followed by the small second pair of tentacles. The pro-
stomium carries a pair of elongated eye-spots.

This species represents P. pictus, Crossland (' P. Zool. Soc, London,' 1903, p. 174),
from which it differs in the character of the modified setae of the fourth parapodia,
the number of branchial segments and the character of the tubes. In the anterior
branchial region the uncinigerous tori encircle the sides of the body, appearing like
silver bands by reflected light. The uncini are excessively numerous in these bands,
and are disposed in a polystichous manner, nine or ten closely packed rows. Each
unoinus shows about twenty minute teeth, the lower end terminating in a blunt,
curved process (Plate V., fig. 136).

294 CEYLON PEARL OYSTEE REPORT.

Family : CIRRATULID^.

Cirratulus cylindricus, Schmarda — Plate VI., figs. 139 and 140.
Schmarda, 1861, ' Neue Wirbellose Thiere,' ii., p. 59.

Localities : — Galle, East Cheval Paar, &c. Associated with Eupomatus heteroceros,
Eunice antennata, &c.

The paired dorsal acervi of tentacular cirri occur on the fifth setigerous segment, a
double row on each side separated by an interval across the middle line. The body
is short and thick, simply flexed, not twisted, and equal at both ends ; length
16 millims., diameter upwards of 3 millims. ; setigerous segments 130. The head is
round and short, crescent-shaped, the mouth gaping below. Capillary setae commence
on the fourth segment, the ventral acicular setse appear on the 19th setigerous
segment, the dorsal acicular seta? on the 23rd ; the capillary seta? are continued to
the end of the body.

In C. dasylopkius, Marenz., the tentacular acervi occur on the dorsum of the third
and fourth setigerous segments ; the ventral aciculse begin on the 29th, the dorsal on
the 43rd. In C. comosus, Marenz., the tentacular acervi occur on the seventh
setigerous segment; the ventral aciculse begin on the 42nd, the dorsal on the 85th.
Both these species are from South Japan. The acicular setae afford no diagnostic
character in themselves.

Another, and perhaps more typical, individual of 30 millims. (from Galle, 4th
August, 1902), which I assign to this species, shows the same form of head, but the
dorsum of the anterior segments is elevated, and the tentacular acervi extend over
more than one segment (Plate VI., fig. 139). The gill-filaments are inserted high up
on the body, remote from the dorsal fascicles ; the ventral acicular setae commence on
the 11th setiger, the dorsal about the 26th. The frontal border shows pigment
specks (Plate VI., fig. 140).

In another specimen of 35 millims. from East Cheval Paar, 8th November, 1902,
8 fathoms, the ventral acicular seta? commence on the 9th setigerous segment, as
many as six in one fascicle in the anterior segments, becoming less numerous and
stouter behind ; the dorsal acicular setse are more slender than the ventral, and commence
on the 25th setigerous segment. The first lateral branchial filament appears on the
first setigerous segment, the transverse acervi occur in the region of the 7th and 8th
body-segments (i.e., 4th and 5th setigerous segments). Behind the acervus the
segments carry one pair of branch ise inserted at slightly different levels, lateral and'
more dorsal.

Cirratulus complanatus, n. sp.

Station XLV, off Panadure, 25 fathoms. Three specimens.

The length of the ribbon-shaped body is 33 millims., greatest width 5 millims.
The body is much flattened, giving a flattened rectangle in section ; the head pointed,

POLYCH^ETA. 295

elongate conical. The tentacular acervi occur on the dorsum of the second setigerous
segment, i.e., the fifth body-segment, the first three segments being as usual achsetous.
In one specimen a small gill occurs on the third achsetous segment. Both dorsal and
ventral acicular setse are present in the 36th setigerous segment. Gill-filaments only
occur in the anterior region of two complete individuals. The gill-filaments and
tentacular filaments give the appearance of a dense tangle in front, the rest of the
body being bare.

The species differs from C. dasylophius, Ma.renzeller, in the position of the
tentacular cirri, the distribution of the gill-filaments and the shape of the body.

Heterocirrus typhlops, n. sp. — Plate V., fig. 138.

Locality : — South-west Cheval Paar, Gulf of Manaar.

A very small worm, total length 10 '5 millims., diameter less than half a millimetre.
Capillary non-limbate setae in both fascicles ; dorsal and ventral acicular setse
commence at the first setigerous segment ; they resemble those of Cirratulus ; the
ventral acicular setae are two in number, more curved and thicker than the dorsal.
The disposition of such cirriform appendages as remain is shown on Plate V., fig. 138.

Family : AMPHICTENIDiE.
Pectinaria panava, n. sp. — Plate V., fig. 137.

Locality : — Ceylon Seas.

The specific name is the vernacular word for comb, referring to the comb of
palese. Nuchal disc subcircular, carrying a transverse series of twenty palese, divided
by an interval into two linear groups of ten each ; the palese terminate in curved
setiform apical processes. At each ventro-lateral angle of the disc, at the outer ends
of the paleal series, there is a tentacular cirrus ; a similar pair of cirri occurs on the
second segment. In front of and below the paleal crown there is a wide semilunar
membrane carrying not less than 32 lacinise frontales. Below the frontal membrane
are the grooved tentacles surrounding the mouth. Third and fourth segments
branchiferous ; ventral portions of 3rd, 4th, 5th and 6th segments elevated, the
crests traversed by transverse muscles ; the crest of the 4th segment is produced on
each side as a blunt lateral lobe below the gill ; on the posterior surface of the lobe
there is a pinhole aperture. Seventeen capillary fascicles commencing from the 5th
segment ; uncinigerous tori from the 8th ; uncini with about eight teeth, reduced
towards the base (Plate V., fig. 137). Scapha acetabuliform, with terminal supra-anal
semilunar valve and two rows of dorsal palese at the base, 7 palese on each side.

Tli is species is chiefly characterised by the number of palese, which however is
likely to vary, and by the large number of the lacinise frontales, which are filiform
processes along the edge of the frontal membrane.

2% CEYLON PEAEL OYSTER REPORT.

Family: SABELLARIID^.

Pallasia pennata (Peters) — Plate VIII. , figs. 1 and 2.

Sabellaria pennata, Peters, "Ueber die Gattung Bdella Sav. und die in Mossambique
beobachteten Anneliden," 'Arch. f. Naturg.,' xxi., 1855, p. 42 ; also ' Monatsber. Berlin.
Akad.,' 1854, p. 613 (quoted from Gritbe, 'Ann. Semp.,' 1878, p. 220).

Hermella bicomis, Schmarda, 1861, ' Neue Wirbellose Thiere,' p. 24.

Locality : — This worm occurs off the west coast of Ceylon either singly or in large
colonies, and builds a tube of coarse sand-grains cemented together, so as to form a
sandstone of excessive hardness which is capable of withstanding the full force of the
waves during the South-west Monsoon. It is evidently an important factor in the
preservation of the coast line. Sometimes the tubes are still further protected by a
Nullipore covering. The internal diameter of a large tube is about 6 millims.

The specific name given by Schmarda refers to the presence of a pair of large
brown hooks bent towards the middle line, placed at the dorsal ends of the peristomial
lobes which, with their armature, constitute the paleal crown. There is usually only
one pair of hooks, but I have seen a specimen with three hooks (two on one side),
and Peters describes four hooks (two. pairs) in the single specimen examined by him.
Each peristomial lobe terminates anteriorly in a truncate dorso-ventral crescent-
shaped area, from the inner and outer arcuate borders of which the modified bristles
or paleae arise. The number of paleae in the inner row varies roughly between 20 and
30 ; the number in the outer row between 30 and 45. The paleae of the outer row
are long curved spines, slightly widened and strongly serrated along the distal three-
fifths of their length ; towards the apex of the palea the serrations are larger on the
concave border than on the opposite border ; all are directed towards the apex of the
palea. The palese of the inner row are acuminate, smooth, and more slender than those
of the outer row. In side view a series of 8 to 12 subulate papillae occurs below the
outer row of paleae, the most dorsal papilla being placed behind the hook on each
side, as shown in Mr. Watson's drawing (Plate VIII. , fig. 1).

Dorsally the peristomial lobes are confluent across the middle line up to the level
of the hooks ; ventrally they are separated down to the level of the mouth. The
inner surface of each peristomial lobe is beset with numerous tentacular cirri arranged
in rows, the columns being indicated externally by crenulations of the ventral border
of the lobe (Plate VIII. , fig. 2). On each side of the mouth occurs the neuropodial
cirrus of the buccal segment, and at the base of this a bundle of capillary setae. The
second segment is distinct dorso-laterally, merging into the peristome below ; it
carries a bundle of neuropodial setae, above and adjoining which there is a triangular
lappet ; farther up the side follows another triangular lappet and dorsally the first
cirriform branchia (Plate VIII., fig. 1). The third, fourth, and fifth segments carry
on each side a neuropodial fascicle of long, rather narrow spatulate setae with laciniate
tips ending in a point, and a laterally placed notopodial fascicle of similar setae with a

POLYCH^ETA. 297

wider spatulate portion, disposed in a single series ; in both fascicles the spatulate
setae are accompanied by more slender limbate setae ; dorsally these segments carry
stout cirriform branchiae. Next follow the uncinigerous segments, as many as 50 in
number, after which the recurved achsetous caudal extremity terminates the body.

The first 20 to 30 uncinigerous segments carry dorsal branchiae ; all carry ventral
fascicles of fine capillary seta?, and the first three are provided with ventral ligules
adjoining the capillary fascicles ; the capillary setae carry projecting laciniate scales
or minute thecae, overlapping like the cups of a Sertularian Hydroid. Adjoining the
mouth, between the latter and the neuropodial cirri or ligules of the buccal segment,
there is a pair of fleshy labial processes, which Mr. Watson identifies as tube-building
organs (Plate VIII., fig. 2).*

It is possible that the Ceylon form may be varietally distinct from the Mozambique
type, and should be known as P. pennata, var. bicomis.

Family: TEREBELLID^E.
Leprea inversa, n. sp. — Plate VI, figs. 141 and 142 ; Plate VII., fig. 197.

Station V., off Chilaw, Gulf of Manaar; 11 fathoms. One specimen.

The body-length, exclusive of tentacular cirri, is from 30 to 40 millims. ; width
in front 2 '5 millims. The segments are numerous and the body much contorted.
The anus is surrounded by a funnel-shaped pygidium. The branchiae are inserted
without reference to particular segments, the third gill occurring high up on the
dorsal surface over the third or fourth setigerous lobe, the second over the first
setiger and the first gill in front. Between the first and second gill there is a
papilla. The first gill occupies the most ventral position and is the smallest in size,
not more than one-third of the bulk of the second gill, the latter having about one-
half to two-thirds the bulk of the third. There are eight ventral thoracic scutes
behind the tumid labium ; the first shield following after the labium is deeply
grooved transversely; the scutes end sharply in the region of the 8th and 9th
uncinigerous tori. The uncinigerous tori and the capillary fascicles are continued to
the posterior extremity of the body; from the 7th torus (11th segment) backwards
the uncini are biserially disposed. The tentacular cirri are not deciduous ; no eyes
were observed. The capillary setae of the anterior segments are very narrowly
limbate, with a slight twist towards the apex, but cpiite simple ; those from the
middle and posterior segments are geniculate (Plate VI., figs. 141 and 142).

In the anterior region, behind the ventral scutes, the median ventral tract is
depressed between the tori. In the posterior abdominal segments the uncini
(Plate VII., fig. 197) are still biserial. The proportions of the branchiae, the differ-
ence between the anterior and posterior capillary setae and the number of ventral

* A note upon the nature of the structures which, in the Sabellariidae, surround the mouth, appeared
in the 'Journal of the Marine Biological Association' (Plymouth), New Series, vol. vii. (1904), p. 301.

2 Q

298 CEYLON PEARL OYSTER REPORT.

scutes are the principal features which differentiate this species from Leprea ehren-
berr/i (Grube), Marenzeller (' Sudjapan. Ann.,' part ii., 1884, p. 5 of reprint).

Polymnia labiata, n. sp. — Plate VI., figs. 143 to 145.

Locality : — One specimen from the pearl banks, Gulf of Manaar ; another from
Aripu Reef. The former inhabited a membranous tube encrusted with coarse
unequal sand-grains and calcareous fragments (including an entire small Naticoid
shell) ; the latter has a large test-like tube encrusted with foreign particles, mostly
very small and chiefly calcareous. It appears that the incrustation of the tube of
Terebellids affords no trustworthy evidence of specific identity except in particular
instances. The tube in question is nearly six inches long, with a diameter of half an
inch. The abdominal region of the contained worm had undergone fragmentation,
but the other specimen from the pearl banks is complete ; length about 115 millims.,
width of thorax 7 millims., segments upwards of 120. The transition from the
torigerous to the pinnigerous region is not very abrupt, the diameter of the body
gradually decreases and the anterior pairs of pinnae may be rather wide and set low
in the integument, resembling the tori. The well-marked scuta ventralia end between
the ninth and tenth pairs of tori, leaving six clear pairs of tori behind the scutigerous
region. When the ventral wall of the thorax is much contracted there is an
appearance of ill-defined scuta being continued to the posterior end of the thoracic
(torigerous) region. The tori of the right and left sides are always widely separated,
not approximating in the mid- ventral line.

The diagnostic characters by which this species is to be distinguished from its
congeners (in the absence of information concerning colour) are afforded by the
structure and proportions of the uncini. The uncinus consists of three principal parts,
the hook, the shaft or neck, and the manubrium ; the relative dimensions of these
parts appear to be constant for the species. The uncini of P. labiata are noticeable
on account of the length and slenderness of the neck ; the basal angle of the
manubrium, which lies deepest in the integument, is produced into a short guber-
naculum resembling the " Muskelfortsatz " of the uncinus of Pista (v. Marenzeller,
' Adriat. Ann.'). The denticulations of the uncinus offer no tangible distinction and
present considerable variety both in number and dimensions. Generally the main
hook is surmounted by a pair of smaller hooks of the second order, followed by
another row of three denticulations of the third order (the middle tooth usually much
longer than the lateral) occupying the vertex of the uncinus. Occasionally accessory
denticles may be observed on the brow of the uncinus (Plate VI., figs. 144 and 145).

No Polymnia was described by Grube from the Philippines. One species,
P. congruens, v. Marenz., is known from South Japan, differing clearly from
P. labiata in the shape of the uncini, and in the number of scuta ventralia. Grube's
Terebella sarsii from the Philippines ('Ann. Semp.,' p. 223) has three pairs of
arborescent gills but no lateral lobes, conforming to Nicolea in the latter respect.

P0LYC1LETA. 299

The description was based upon a single indifferently preserved specimen. If
confirmed it will probably form the type of a new genus or sub-genus intermediate
between Polymnia and Nicolea.

The labium and lateral lobes of the third segment are of the same nature as the
corresponding processes in Loimia (Plate VI., fig. 143).

Polymnia socialis, n. sp. — Plate VI., figs. 146 to 148.

Locality : — Narrow sand- encrusted tubes of a small Terebellid were adhering to
the tubarium of PhyllochoBtopterus ramosus, trawled off Galle at Station XXXIX.,
16 to 30 fathoms. One of the larger tubes measured about 50 millims. in length,
2 millims. in diameter. Three pairs of arborescent gills, the first (longest) extending
forwards beyond the upper lip. Seventeen pairs of capillary fascicles, sixteen pairs of
thoracic tori, uncini uniserial in the first six tori, biserial and opposite in the rest,
uniserial again in the abdominal pinnules, which are supported by fine sustentacular
setae. Capillary setae simple, narrowly limbate. The band of scuta ventralia is well
set off from the surrounding parts, rounded in front, attenuate behind, ending as a
whitish streak in the region of the 13th-14th tori, but continued behind this point as
a colourless median streak into the anterior abdominal region ; behind the 9th torus
the band of scuta becomes paler, whitish.

The dorsal surface of the worm is smooth and convex, not showing segmental
divisions in the region of the thorax. The first segment which forms the lower lip is
long below and deeply cleft, the right half slightly overlapping the left (Plate VI.,
fig. 146). The lateral lobe of the second segment, seen in side view, appears as a
sub-elliptical or semilunar, symmetrical free dermal fold. The uncini are remarkable
for the number of denticulations on the vertex, which exceed the narrow limits
suggested by von Marenzeller for the genus. The vertex of the uncinus, when
seen from above, shows a rosette of twelve denticulations, and attentive examination
shows that these are arranged in arcs across the vertex. The general formula for an
uncinus of this species, according to Marenzeller's system of notation ('Adriat.
Ann.,' hi., 1884, p. 163, or p. 13 of reprint) would be as follows :— 1, 22, 333, 4444,
55555. The rosette which appears in vertical view is formed by the teeth of third,
fourth and fifth orders (Plate VI., figs. 147 and 148).

The limbus of the capillary setae is, at least in some of them, wider over a sub-
terminal tract than more distally, spreading like a pair of narrow fins on either side
of the seta. The terminal portion presents a very narrow obscurely striated border.
This is merely the continuation of the limbus, and is not like the laciniate plumose
structure of the terminal filament in Amphitrite. In some setae from one of the
specimens the terminal portion is marked off by a shallow constriction from the
bulk of the shaft. This is not a constant feature however, and the setae of the
larger individual end normally in a simple point.

Length of fore-body without the tentacular cirri (which are present knotted

2 q 2

300 CEYLON PEARL OYSTER REPORT.

together and calling for no remark) 13 millhns., width 2 millims., total length of body
20 millims.

Polymnia triplicata, n. sp. — Plate VI., figs. 149 to 152.

Locality : — Galle, 4th August, 1902. Several specimens.

This is another tribranchiate Terebellid of small size, with fragile body and highly
deciduous tentacular cirri. Length of a specimen, incomplete behind, 35 millims. ;
thorax barely wider than rest of body, 3 millims. in diameter, with 17 setigerous
segments ; abdominal portion with 34 pinnigers. Another abdominal fragment had
80 segments.

The second, third, and fourth segments (the fourth being the first setigerous
segment) carry lateral lobes of approximately ecpjal size (Plate VI., fig. 149). The
scuta ventralia are 14 in number, counting from the second segment; they end
sharply on the 13th setigerous segment, leaving four setigerous (thoracic) segments
without ventral shields. The central portions of the scuta are slightly marked off
from the anterior and posterior borders, but otherwise they are simple, not biannulate
nor areolated.

The arrangement of the anterior thoracic segments seems to differ considerably
from the condition observed in P. labiata. As shown on Plate VI., fig. 143, there is,
in P. labiata, a large lower lip with free sinuous border, considered as belonging to
the second segment (i.e., the first branchiferous segment) and followed by another
acheetous segment (the second branchiferous segment) carrying prominent lateral
lobes. In P. triplicata the lower lip is followed by two achsetous segments carrying
lateral lobes, and also bearing the first and second pairs of gills. It appears, therefore,
that in P. labiata and in the species of Loimia the much larger lower lip is a product
of the fused first and second segments, while in P. triplicata the first and second
segments are distinct below, the former alone carrying the lower lip. In this and in
other respects (number of scuta, &c), the present form offers a striking analogy to
Lanice triloba described by Fischli from Ternate,* and were it not for the existence
of the generic distinction, which is not apparent at the surface, the two species would
probably be ranged together.

The enumeration of the anterior segments and the nature of the prostomium of
Terebellidae are matters of some obscurity. On Plate VI., fig. 149, the thin half-
collar which overhangs and protects the mouth like an epistome or upper lip is seen
in side view ; at the back of it are the scars of insertion of the tentacular cirri,
followed by a clear surface of a segment showing numerous eye-spots. Then follow
the gills on the second, third, and fourth segments. The uncini are much smaller than
in P. labiata, and usually only one denticulation of the third order is to be observed
(Plate VI., figs. 151 and 152). The dorsal portion of the thoracic segments becomes
biannular, commencing from the region of the eighth torus (Plate VI., fig. 149).

* H. Fischli, "Polychiiten von Ternate," ' Abh. Senckenberg. Ges.,' xxv., 1900, p. 122.

P0LYC1LETA. 301

The branchire (as in Lanice triloba, Fischli) show very clear-cut dichotomous
ramifications, the main stem of each gill dividing near the base into two main
branches. In the biserial tori the uncini of the two rows are alternate and opposite,
(i.e., base to base), those of the anterior row progressive, of the posterior row
retrogressive. In Lanice triloba, the uncini of the biserial tori are inverted
(i.e., vertex to vertex), as in Loimia, though in structure they resemble the uncini of
Polymnia.

Loimia annulifilis (Grube) — Plate VI., figs. 153 and 154.
Terebella annulifilis, Grube, 'Ann. Semp.,' 1878, p. 225.

Locality : — Galle shore, under stones. Eight specimens.

Tube encrusted with coarse, rather large shell and coralline fragments up to 8 millims.
or 9 millims. in diameter. In most specimens the body-wall of the dorsal thoracic
region has ruptured, and a coil of the gut is protruding. Diameter of thorax up
to 9 millims.; length of body about 150 millims.; abdomen coiled; more than
200 segments.

Uncini from the first torus, 5-dentate ; from the third torus, 5-dentate (Plate VI.,
fig. 153); from the sixth, 5-dentate, the fifth tooth becoming smaller and in some
uncini obsolescent ; from the ninth torus, 4-dentate in both rows ; from the twelfth,
4-dentate ; from the sixteenth, 4-dentate ; from the twenty-seventh torus (which is
borne upon the eleventh abdominal pinnule), 4-dentate. From these enumerations
it follows that the uncini of the first six uniserial thoracic tori are 5-dentate, of the
last ten biserial thoracic tori and of the uniserial abdominal pinnules 4-dentate. No
such distinction is recorded by Grube among the five species of Loimia described by
him from the Philippines. In some of the uncini from the anterior tori there seem to
be indications of a sixth tooth, and in one uncinus from the first torus there are six
distinct teeth.

The first pair of arborescent branchiae greatly exceeds the second and third in size,
quite overshadowing them. Seen with low magnification, they present a densely
racemose or finely tufted appearance, due to the fact that the digitations arise in
groups from the main stem and thicker branches, curving inwards like the half-closed
fingers of a hand.

According to the accepted interpretation of the anterior complex of the Terebellidse,
the buccal segment is destitute of appendages and sense-oigans (segmentum buccale
nudum), but is produced ventrally into a large free labial fold, which closes the
ventral opening of the horse-shoe-shaped collar or epistome formed by the prostomium
surrounding the mouth above and in front. The hinder portion of the prostomium
looks uncommonly like a cephalic segment and carries the tentacular cirri, which are
ringed with purplish colour as in Grube's Loimia annulifilis. The buccal segment is
the first body-segment, and is here practically non-existent above, though represented
below by the prominent labium. The second body-segment is the first branchiferous

302 CEYLON PEARL OYSTER REPORT.

segment, well-marked dorsally, obsolete ventrally. The third body-segment or second
branchiferous segment is evident above and below, and is produced outwardly into
two ear-shaped lateral lobes, between which there is a narrow rim representing the
ventral portion of* the segment. Behind this follows ventrally the first ventral
scute belonging to the fourth body-segment, which is at the same time the third
gill-bearing segment and the first setigerous segment. The fifth body-segment is
the first uncinigerous segment, and the uncini are arranged uniserially in the first
six pairs of uncinigerous tori ; the ten following tori carry uncini in two rows. The
twentieth body-segment is the last thoracic segment. Then follows the abdominal
region, which is destitute of capillary setae, the uncini being borne in single file upon
the abdominal pinnules. The first scutum differs from those which follow by its
large size and corrugated surface. There succeed to this nine plain scuta ventralia,
becoming narrower and longer behind. After this again the scuta become sub-
divided by transverse grooves, finally ceasing between the penultimate thoracic tori.
The posterior subdivided scuta present a dark colour in the preserved material
(Plate VI., fig. 154). There are 17 fascicles of capillary setae, which have a fibrous
internal structure, the fibres separating out when the seta is crushed or broken.
Some of the setae, however, which were presumably broken before death, show a clean
fracture. It may be noted in conclusion, that in one specimen the scuta between the
eighth and ninth uncinigerous tori are divided by a cross furrow.

Loimia crassifilis (Grube).

Locality : — East side of Cheval Paar.

One very small individual of 20 millims., including the tentacular cirri ; the latter
are banded as in L. armulifilis. There are ten clear scuta ventralia, after which these
become quite obscure, but the tori remain separated by a wide decreasing median
ventral tract to the end of the thorax. Uncini from the fourth torus 4-dentate, with
an occasional rudimentary fifth tooth at the apex ; from the biserial tori 4-dentate,
the fourth tooth small ; from the abdominal pinnules 5-dentate, namely, four clear
teeth and a rudimentary apical tooth. The above corresponds with Grube's
description ('Ann. Semp.,' p. 226), "uncini pectiniformes dentibus tororum 4,"
although his specimen was a very large one. The abdominal region is slender and
even ; the thoracic region tumid ; about 67 segments.

Loimia medusa (Savigny) — Plate VI., figs. 155 to 159.

Savigny, ' Syst. des. Ann.,' p. 86 (quoted from Grube, ' Ann. Semp.,' p. 228).
(1.) Three specimens from Aripu Coral Reef, about 80 segments, width of thorax
10 millims. ; total length nearly 90 millims., of which 30 millims. belong to the
thorax; abdomen subcylindrical, varicose, simply flexed (not spirally coiled), narrower
than thorax (Plate VI., fig. 155). Uncini from the first torus 6-dentate ; from the
third some 5-dentate, some 6-dentate ; from the fifth 5-dentate ; from the biserial
tori and abdominal pinnules 4-dentate. There are five pairs of tori behind the scuta

POLYCH/ETA 303

ventralia. The last plain scutum corresponds with the seventh pair of tori ; the
rest are subdivided, terminating after the eleventh tori. The subdivided scuta are
dark-coloured. The three pairs of branchiae are subequal.

(2.) Two more small worms which I assign to this species were taken from the
South-west Cheval Paar. These also have five pairs of tori behind the scuta, but
the junction of the undivided and subdivided (or anterior and posterior) scuta occurs
at the level of the eighth tori, clearly so in one, approximately so in the other
specimen. One has 52 segments with a length of 36 millims., the abdominal
segments being closely compressed ; the other has about 64 segments with a length
of 45 millims. Uncini from the first torus 6-dentate, from the third also 6-dentate ;
fi'om the fifth 5-dentate ; from the biserial tori and abdominal pinnules 4-dentate.
Part of a tube is encrusted with small shell-fragments of various sizes.

(3.) There are two other well-preserved worms taken from the Aripu Coral Reef,
which afford the peculiarity of an accessory or eighteenth pair of capillary fascicles.
The uncinigerous tori which accompany these fascicles are narrower than the pre-
ceding, thus affording a graduated transition from the tori to the pinnules. This is
particularly so in one of the specimens where the accessory tori are much narrower
than in the other. The tubes constructed by these worms are encrusted with coarse
sand-grains with a scanty admixture of calcareous fragments.

The scuta ventralia are much narrower than in the first or typical variety, an
appearance which may be partly due to the state of protraction of the body and the
method of preservation (Plate VI., fig. 156). On account of this fact, together with
the difference of the tubes and the presence of eighteen pairs of capillary fascicles, I
will call this variety angustescutata , in order to facilitate future reference. The
uncini from the first torus are mostly 6-dentate (Plate VI., fig. 157), the sixth tooth
frequently reduced and sometimes obsolete ; in the fourth torus the uncini are
5-dentate (Plate VI., fig. 158); from the biserial tori and abdominal pinnules,
4-dentate (Plate VI., fig. 159). The body of these worms resembles that of the
typical form, though somewhat more slender ; 83 segments, length about 100 millims.,
abdomen varicose, width of thorax 8'5 millims. The lateral lobes of the anterior
segments resemble the corresponding structures in L. annulijtlis (q. v.).

Loimia montagui (Grube) — Plate VI., figs. 160 to 163.

Terebella montagui, Grube, ' Ann. Semp.,' 1878, p. 224.

Loimia montagui, Marexzeli.er, ' Siidjapan. Ann.,' ii., 1884, p. 205.

Not Terebella niontagni, Quatrefages, ' Hist. Nat. Ann.,' ii., 1865, p. 361.

A large number of cylindrical worm tubes are labelled " No. 32, Terebella, Palk
Bay." Many of the tubes are in short lengths, the more complete measure upwards
of 100 millims. (4 inches) in length, with a diameter of 9 millims. They are coated
with a thick dense layer of fine mud presenting a smooth surface, the lips of the
terminal orifices approximated, so that no sign of the contained worm appears
externally. I found two worms only belonging to the tubes.

304 CEYLON PEARL OYSTER REPORT.

The body is narrowed anteriorly (in the region of the scuta ventralia) and
attenuated behind ; in the middle region it is inflated and somewhat varicose ;
segments about 100, length about 85 millims., width of thorax in front 4 millims.,
behind the scuta 5 millims. (Plate VI., fig. 160). Counting the composite sternite of
the 2nd, 3rd, and 4th segments as the first scutum, there are eight clear scuta
followed by a rather ill-defined ninth. After this there is no trace of further scuta,
the median ventral tract separating the tori uncinigeri is wide and smooth, and the
length of body between the successive tori gives the worm a remarkable appearance.
In the first torus the uncini are mostly 6-dentate, the sixth tooth sometimes large,
sometimes rudimentary, and many uncini in the same torus are 5-dentate (Plate VI.,
figs. 161 and 162). In the fourth torus not one 6-dentate uncinus was observed, only
5-dentate uncini with a sprinkling (about 5 per cent.) of 4-dentate uncini. In the
biserial tori only 6-dentate uncini were found in both rows (Plate VI., fig. 163). An
abdominal pinnule contained 7-dentate uncini, the seventh tooth usually small and
occasionally obsolescent, but clear enough to characterise the region.

In ventral view (Plate VI., fig. 160) this worm presents distinguishing features
depending upon the size and limited number of the scuta ventralia, the absence of
subdivided scuta behind the main series, and upon the divarication of the meta-
thoracic tori. In the character of the uncini the species corresponds closely with
Loimia montagui, Gr., as described by Marenzeller (' Siidjapan. Ann.,' 1884, ii.,
p. 9, reprint), who, however, does not describe a definite succession of uncini such as
I have noted above, namely, the 6-dentate uncini of the first torus, followed by a
reduced denticulation in the succeeding uniserial tori, and this again by an increased
denticulation in the biserial tori and a further increase in the 'abdominal pinnules. It
agrees with L. montagui again in the triangular shape of the lateral lobes of the
third segment and in the presence of a whitish glandular tract on each side of the
dorsal surface over the capillary fascicles, extending from the gill-region to the eighth
setigerous segment, where it ends in a point.

Although von Marenzeller records eleven scuta ventralia for Loimia montagui
from Japan, and Grube ten for the example described by him from the Philippines ;
the former author specially notes that the most sharply circumscribed scutes are
those from the second to the eighth inclusive, which accords with the variety from
Palk Bay. Neither Grube nor von Marenzeller refers to the tube, which, in the
case of Terebellidse, may or may not be a matter of importance. Nor is any
information given as to the nature of the locality. Both of these points, in the
present instance, have a special interest, bearing upon the identification.

Loimia varieguta (Grube).

Terebella variegata, Ehrb., ' Monatsber. Ak. Berlin,' June, 1869, p. 30 of reprint; Grube,
1878, 'Ann. Semp.,' p. 227.

One specimen in a thin membranous tube encrusted with shell-fragments and

POLYCILETA. 305

sand-grains was present in the same bottle with Eupomatus heteroceros, Eunice
antennata, etc. ; no locality being given : probably South-east Cheval Paar.

In the second of the above-quoted publications Grube makes the important
correction that the uncini are not 4-dentate, as previously stated, but 5-dentate. This
species resembles L. medusa in the number of scuta ventralia, and consequently in
the occurrence of five pairs of tori behind the scuta. These posterior tori touch one
another in the middle line in their respective segments.

The specimen in the preserved state has a nearly uniform pale flesh tint, only the
branchiae, which are matted together, showing up dark. There is no great contrast
between the diameter of the thorax and that of the abdomen ; the latter is sub-
cylindrical, smooth, somewhat varicose. The total length is about 80 millims,, and
the number of segments about 60, the maximum width 7 millims., thus presenting a
close numerical correspondence with Grttbe's original type. The scuta ventralia are
narrow as in the angustescutate variety of L. medusa. This species is said by Grube
to differ in colour from L. medusa, but in the preserved state the chief characteristics
are the nearly even diameter of the body, the uncini of the biserial tori and abdominal
pinnules which are 5-dentate, to which may be added the pale flesh colour. The fifth
denticulation at the vertex of the uncinus is small but distinct. The average width
of the thorax (excluding the capillary fascicles) may be estimated at 6 millims., that
of the abdomen at about 5 millims. At the same time, if it were not for the
quinquedentate uncini, I should probably have placed the worm under L. medusa.

In the living condition more fundamental differences may come to light. The worm
was closely invested by its tube. The preserved specimens of L. medusa have a dark
neutral colour in marked contrast with the pale flesh colour of this worm. A second
smaller specimen shows very clearly the quinquedentate uncini of the abdominal
pinnules. In this case the abdominal region is more attenuated, and the scuta ventralia
end rather sooner, so that there are six pairs of tori behind them.

Grymaea, Malmgren, 1865.

Three pairs of acervi of filiform branchiae.

( !apillary fascicles commencing from the second segment (first branchiferous segment)
and extending to the posterior abdominal region.

Tori uncinigeri commencing on the fourth (? or fifth) setigerous segment. Uncini
avicular, uniserial.

Malmgren says the tori commence on the fifth setigerous segment, but in a
specimen of the type species, Grymcea bairdi, Malmgeen, collected off Norway by the
Rev. Canon Norman, which I had the opportunity of examining at the British
Museum some years ago, they commenced on the fourth setigerous segment.

Grymsea cespitosa, n. sp. — Plate VII., figs. 164 and 165.
One specimen, incomplete behind, was obtained, in 40 fathoms, south of Rameswaram
Island, at Station LIV.

2 R

306 CEYLON PEARL OYSTER REPORT.

The fragment comprises twenty-two setigerous segments, total length about
30 millims. ; width of thoracic region 3 to 4 millims. ; thorax straight, porrect,
abdominal region bent, with smooth, convex, turgid dorsum. The pharetrse setarum
are high, upwards of 1 millim. in front, decreasing gradually backwards. The simple,
narrowly limbate setae issue in two principal bundles from between the oblique lips.
The ventral portions of eleven segments from the second are flattened, forming bi'oad
ventral shields subdivided by transverse grooves. There are nine torigerous segments
in the thoracic region, the posterior thoracic segments being nearly as long as broad.
The thoracic tori graduate insensibly into the abdominal tori, which are sessile, not
pinnuliform. The uncini have the form shown in Plate VII., fig. 165, and the general
formula 1-22-3. In side view the three tiers of teeth show clearly.

The branchial filaments are numerous, forming dense coils. Those of the first
acervus commence low down near the ventral surface. They break away easily from
the body in groups. The tentacular cirri, plainly grooved on the lower side, are much
stouter than the branchial filaments (Plate VII., fig. 164). The three branchiferous
segments also carry capillary fascicles, the first uncinigerous torus occurring on the
first post-branchial segment. The branchiae arise from the anterior parts of the
segments in front of the corresponding capillary fascicles.

Family: SABELLID^E.
Branchiomma acrophthalmos, Grube — Plate VII., figs. 166 and 167.

This is the oriental form of the well-known B. vesiculosum.

Locality : — East Cheval Paar, Gulf of Manaar.

Baron de St. Joseph has shown that the character of the limbate setae of the
European species varies at different ages, being spatulate in the young.

The Ceylon specimen is a fragment comprising 16 segments, 21 millims. long
(including the gills), 3 millims. wide; 15 radioles* in each gill; gills banded about
six times purple and whitish. Each radiole carries a subterminal eye, the two dorsal
eyes being the largest. The definition of Branchiomma given by de St. Joseph
(1894, 'Ann. Sci. Nat.,' xvii., p. 249) states " Soies dorsales d'une seule sorte au
thorax." In the Ceylon specimen the dorsal fascicles of the thorax appear to contain,
and in fact do contain, two distinct kinds of setse, normal limbate to the number of
eleven, and spatulate (Plate VII., fig. 166) to the number of nineteen. The thoracic
tori are biserial, containing a row of avicular uncini and a row of cuspidate setae.

Grube describes ('Ann. Semp.,' 1878, p. 258) the collar as trilobate, apparently
overlooking the ventral incisure which divides the two acuminate ventral lobes.
These are in close juxtaposition and give the appearance of the simple wide triangular
lobe described by Grube. The dorsal portions of the collar show a characteristic
form, the free border deeply emarginate externally and reflected round the base of
the gill-supports (Plate VII., fig. 167). There are eight thoracic segments.

* See definition of this term on p. 308.

POLYCH^ETA. 307

Branchiomma quadrioculatum, n. sp. — Plate VII., figs. 168 and 169.

Locality : — Aripu Coral Reef. One specimen, apparently incomplete behind,
48 segments.

Thorax 4-5 millims. long, 2 millims wide. Abdomen 13 millims. long. Eight
thoracic segments. Gills 6 millims. long, banded with three or four purplish vittse ;
14 radioles in each. Collar with median incisura ventralis, a deep median dorsal
notch and a still deeper submedian dorsal notch on each side. The four most dorsally
placed gill-rays carry subterminal eyes, of which the dorsal pair are larger than the
subdorsal pair (Plate VII., fig. 168). Seen from below, pressing the gills aside,
laciniaj buccales are brought into view, including a flattened median tongue which
rises up straight between and below the antennas, and on either side of it an arcuate
membranous fold. The buccal seta? are normal, capillary, limbate ; thoracic setae
normal limbate and subspatulate ; thoracic uncini biserial, an anterior row of cuspidate
uncini "' and a posterior row of avicular uncini (Plate VII., fig. 169). Abdominal
seta? 15, limbate; abdominal uncini uniserial, avicular.

Hypsicomus phaeotaenia (Schmarda).

Sabella phaeotaenia, Schmarda, ii., 1861, p. 35.

Hypsicomus phaeotaenia, Marenzeller, ' Siidjapan. Ann.,' ii., 1884, p. 16 of reprint, p. 212 of

the 'Wiener Denkschr.'
Sabella pyrrhogaster, Grube, 'Ann. Semp.,' 1878, p. 250.

Localities: — Station V., off Chilaw, Gulf of Manaar, 11 fathoms, several specimens;
and South-west Cheval Paar.

Marenzeller notes the resemblance between Sabella pyrrhogaster and Hypsicomus
phcBOtcmia, but adds that in the former only one kind of setae was observed by
Grube in the thoracic tori. According to Grube also the dorsal setigerous lobes of
the thorax only contain paleae, whereas in phaeotamia there are a few lanceolate setas
in addition to the paleas. The latter discrepancy is easily accounted for by the fact
that the lanceolate setae are frequently broken and hence cannot be observed. The
former discrepancy with respect to the tori in which the armature is biserial and
dimorphic in phceotcenia and alleged to be uniserial and uniform in pyrrhogaster, is
undoubtedly due to the circumstance that the row of cuspidate uncini was overlooked.
Neither does Grube mention the branchial eyes, the pigment of which was perhaps
dissolved out of his material. This species has also been described by Grube as
Sabella fuscotcmiata in 'P. Zool. Soc, London,' 1874, p. 328.

A small though complete specimen measures 25 millims. long without the branchiae,

which add another 8 millims. From the posterior end of the basal lamella to the

anterior border of the interfilar membrane or web measures 3-5 millims. The eyes

occur on each side of the distal portions of the gill-stems, 14 to 17 in number, not in

pairs. The terminal filament of the gill-stem is long, somewhat flattened and tapering

to a point.

* " Soies en pioche " ; " Piekelborsten."

2 R 2

308 CEYLON PEAEL OYSTER REPORT.

The macroscopic characters by which the species may be recognised are the
occurrence of an arcuate oblique double row of modified setae in the first segment and
the rich dark-brown colour of the ventral shields of the abdominal region. The short
collar still retains a band of violet below ; it is entire, its dorsal border being slightly
concave, while its ventral border is produced forwards to a median point. The
cuspidate uncini which accompany the avicular uncini are called " Pickelborsten " by
von Marenzeller.

Accompanying the typical individuals described above there is another specimen in
which the whole collar is violet except at the anterior border. It is in a state of
protraction, and the modified setae of the first segment are disposed in a slightly
undulating longitudinal ridge which stands out pale upon the violet background.
This specimen further shows violet vittse and scattered spots on the gills, but no eyes
are to be found ; the terminal filaments into which the cartilaginoid axes extend are
exceptionally long. The apparent absence of eyes is remarkable. The brown colour
of the abdominal scutes is missing ; only the anterior abdominal region is preserved.

Dasychone cingulata, Grube — Plate VII., figs. 170 to 173.

Two specimens in bottle with Loimia variegata, Eu-pomatus, &c. [? Cheval Paar].

Total length 26 millims., gills 9 millims. to 10 millims., thorax 4 millims., width
3 millims. ; about 60 segments. Eight thoracic segments with eight fascicles and
seven tori, the capillary limbate setae of the first fascicle not different from the others.
In each gill 19 to 21 radioles, each provided with about 14 pairs of dorsal sty lodes
and eyes ; radioles subarticulate ; terminal filaments of moderate length. In a
thoracic fascicle there are 36 limbate setae (Plate VII., fig. 170), and in an abdominal
fascicle about half that number ; the setae issue in two bundles, upper and lower,
those of the latter are shorter and (especially in the abdominal region) have a
pronounced curvature, a broader limbus and a shorter flagellum (terminal filiform
portion). It may be explained that the term " radiole," as applied to a single shaft
or rhachis of the gills, is used here in the sense in which it has been employed by
Professor McIntosh. It corresponds to Grube's filum branchiale. The actual
branchial filaments which are borne upon a rhachis are the pinnae, to which Grube
applies the term radioli. Grube and McIntosh, therefore, use the term " radiole" in
different senses. Grube ('Ann. Semp.,' 1878, p. 259) describes the radioles (sensu
McIntosh) as longearticulate, which corresponds with the Ceylon specimen, and he
remarks further that there are about ten pinnae (branchial filaments) to each joint ;
this is also in accord with my observations (Plate VII., fig. 172). In the Ceylon
specimen the articulation of the radioles is faintly indicated and best seen under low
magnification.

Dasychone, like Phyllodoce, occurs singly in collections, and nearly every tropical
specimen has proved to be a distinct species. Grube says that in Dasychone decora,
Sars, the number of radioles in each trill varies from 20 to 36. The uncini of

POLYCH^ETA. 309

Dasychone are striated parallel to the curvatures and are somewhat characteristic for
the species. There are long-shafted and short-shafted forms and others, as in
D. baircli, McInt. (" Challenger") and D. picta, McLntt., of special form. Examples
of long-shafted uncini are those of D. wyvillei, McInt., and D. orientalis, McInt. ;
short-shafted uncini are in D. cingulata, Grube, D. japonica, McInt., D. nigro-
maculata, McInt.. and D. maculata, Fischli (" Polychiiten von Ternate," ' Abh.
Senckenb. Ges.,' Band xxv., Frankfort, 1900, p. 125).

The radioles of the Ceylon specimen of Dasychone cingulata agree in arrangement
and form of stylodes and eyes with the figures of D. japonica, McInt., and
D. maculata, Fischli. The colour has vanished except for the minute dark pigment
spots between the fascicles and the tori of the abdominal segments. The denticulations
on the vertex of the uncinus occur in transverse rows (about three in a row) as in
many of the Terebellacea ; two appear above the main hook in side view ( Plate VII. .
fig. 171).

Eurato porifera (Grube) — Plate VII., fig. 173.

Sabella porifera, Grube, 'Ann. Semp.,' 1878, p. 252.

Eurato, St. Joseph, 'Ann. Dinard,' part iii. ; ' Ann. Sci. Nat.' (7 ser.), xvii., 1894, p. 219.

" Large Sabellid from centre of coral block, Muttuvaratu Paar ; yellow-olive body,
scarlet and yellow plumes." Total length about 90 millims., of which the gills occupy
30 millims. and the thorax 10 millims. ; eight thoracic segments ; width 8 millims.

The colour of the preserved specimen is sandy-yellow without any trace of pigment
spots or markings. The only relief from the yellow ground colour is afforded by the
characteristic spongy glandular tract which occurs on the anterior dorsal thoracic
region. This has a brown colour and appears divided into right and left portions by
the dorsal thoracic groove.

Limbate capillary setae of different lengths throughout ; uncini uniserial, avicular
with multidenticulate (cross-hatched) vertex, not different from those figured by
von Marenzeller for Laonome japonica.

Branchial stems in a single row, a few of them only occurring at a slightly deeper
level than the rest, about 51 on each side. The length of the laciniae tentaculares is
about 7 millims.

The dorsal glandular tract by examination under a lens appears to consist of
anterior and posterior portions, the former of a looser texture and thicker ; but in
the specimen under description both regions have the same brownish tint. Another
smaller individual which accompanied the one described above shows the glandular
area much more clearly. There is a posterior pale porous tract with a labyrinthine
structure, the sulci running, in general, transversely ; in front of this there are two
dark-coloured convex porous cushions. In this specimen the general colour is duller
and the gills rather olivaceous. In both cases the ventral fsecal groove bends up

310 CEYLON PEARL OYSTER REPORT.

between the ninth and tenth ventral shields and passes round to the dorsum obliquely
between the eighth and ninth capillary fascicles (Plate VII., fig. 173). The total
length of the smaller worm is about 70 millims., of which nearly 7 millims. go to the
thorax, 20 millims. to the gills.

Eurato notata (Grube) — Plate VII., figs. 174 to 176.
Sabella notata, Grube, 'Ann. Semp.,' 1878, p. 256.

A small worm in a finely encrusted closely investing tube was associated with
Phyllocluetopterus ramosa from Galle. The worm could only be removed from its
tube with difficulty and not without damage, the gills adhering closely to it and to
each other. The gills, with about 12 radioles in each, are banded with purplish
colour, but the bands are all obscure in the preserved state with the exception of the
basal band formed of a row of elongate dense purple spots, one on each radiole placed
immediately in front of the level at which the radioles pass into the basal membrane
which supports them. This is the first feature to meet the observer on removing the
animal from its tube. The next distinction is afforded by the presence of a pair of
subtriangular or pear-shaped pigment aggregates on the dorsal side of the first
segment above the capillary fascicle, the apex directed forwards ; this no doubt
corresponds with the C-shaped mark which Grube noticed on his specimen. A
pigment patch occurs on each side of the succeeding segments between the groups of
setae and the uncini, i.e., between the fascicles and the tori.

In this specimen the thoracic segments were abnormal, fewer than normal and
fewer on one side than on the other. Similar variations have been described in
several Sabellidae by the Baron de St. Joseph. On the left side there are three
thoracic tori, on the right side four. In side view of the uncini three rows of teeth
are to be seen on the vertex (Plate VII., fig. 174). The capillary setae are all limbate,
some long-limbate, others short-limbate and some rather wide-limbate (Plate VII.,
fig. 175). The limbus of these setae is a kind of guard upon one surface embracing
the setae. From some points of view the seta appears bilimbate, but when the shaft
is broken across, the limbus may be left projecting far beyond the broken edge
(Plate VII, fig. 176).

The total length of the worm is about 125 millims., of which 2 "5 millims. belong to
the gills ; width 1 millim., length of tube nearly 20 millims. The collar of this
species is only conspicuous ventrally, where it is represented by a pair of ventro-
lateral lobes, broad lappets subacute in front at the outer angles ; these are what
Grube called the laciniae ventrales. Grube's was a larger example and the propor-
tion of the length of gills to body-length differs from that of the Ceylon representative.
Another fragment of a larger individual was obtained on east side of Cheval Paar,
and also showed clearly 4 thoracic tori on the right side, 3 on the left. The marks
on the pronotum are C-shaped, with the concavity directed forwards and inwards.

POLYCH^TA. 311

Sabellastarte indica, Sav. — Plate VII., fit>'. 177a.

Cf. Quatrefages, 1865, ' Hist. nat. des. Ann.,' ii., p. 432.

A specimen from the pearl banks, upwards of 40 millims. long (without the gills)
by 7 millims. wide, inhabited a tube nearly twice its own length encrusted with sand.
The deciduous branchial crown together with the tentacular lacinise was present, but
thrown off. Projecting beyond the collar are two collapsible buccal lobes, evidently
protruded by fluid pressure from within. At the ventral sides of the thoracic tori
purple streaks occur, and similar streaks occur towards the dorsal border of the tori.
In the abdominal region these streaks are replaced by minute spots. The gills are
banded. The thorax is composed of eight segments with seven pairs of tori ; the
setse, of all the fascicles alike, are limbate capillaries.

Another specimen labelled "No. 80, Sabella fusca, Grube, orange-bodied Sabella,
on under side of boulder, Galle lagoon, off Breffit, 7th June, 1902," is larger and
darker; 100 millims. long (without the gills), 10 millims. wide; body subcylindrical
in front, more flattened behind. Gills very dark ; inner surface of collar with black-
purple band and a patch of the same colour on the outer surface of the dorsal
portions of the collar.* In frontal view of the buccal crown the two buccal lobes
described above are seen, though not protruding to the same extent (Plate VII.,
fig. 177a).

Sabellastarte indica, var. quinquevalens, nov. — Plate VII., fig. 177.

Another specimen, brought up by the divers from Muttuvaratu Paar, inhabited a
membranous tube coated with fine mud. Total length 100 millims., of which
35 millims. belong to the branchiae; more than 140 segments. The thorax is
composed of six segments with five pairs of uncinigerous tori. The uncini are
uniserial and show numerous denticles above the main tooth, presenting, in frontal
view, a finely cross-hatched appearance, in side view a series of 8-10 minute rows on
the vertex of the hook (Plate VII., fig. 177). Width of the thorax 8 o millims., of
the anterior abdominal region 7 millims. Branchiae banded with purple, the radioles
or stipes in two rows, but the bases of all show through the basal membrane. The
radioles of the two rows are roughly alternate, though the outer row contains more
than the inner. Terminal filaments of gill-stipes short ; gill-filaments biserial.
Collar with broad purple submarginal band on inner surface, as in the typical
examples. It seems likely that S. indica is co-specific with Grube's S. spectabilis
('Ann. Semp.,' p. 253) and even with Marenzeller's Laonome japonica (' Siidjapan.
Ann.,' ii., 1884, p. 16). See also St. Joseph, "Ann. de Dinard," 'Ann. Sci. Nat,,'
xvii., 1894, p. 249. The number of thoracic segments is known to be subject to
individual variations. The collar is notched below by a median incisura ventralis.

* Similar colour-markings occur on the collar of the first example.

312 CEYLON PEARL OYSTER REPORT.

Jasmineira caducibranchiata, a. sp. — Plate VII. , figs. 178 and 179.

Locality : — East side of Cheval Paar.

A small slender worm of nearly even diameter, gently tapering behind, 22 millims.
long, 1*5 millims. wide; eight thoracic segments, 36 abdominal segments with anal
groove passing obliquely across the first abdominal (9th body-segment) to the right
side and on to the dorsal surface. The thoracic segments do not differ macroscopically
from the anterior abdominal segments ; there are 8 dorsal capillary fascicles, and
7 ventral tori carrying a single row of rostrate uncini with long manubrium
(Plate VII., fig. 178). The abdominal uncini are avicular (Plate VII., fig. 179).
The buccal segment carries the first capillary fascicle near its hinder border and
towards the dorsal side ; the collar is rounded and slightly projecting forwards below,
with a median notch or incisura dividing the two low rounded lobes, and a shallow
impression on each side of the notch. On the dorsal side the collar stands out at
right angles to the body and is then inflected at an acute angle, to be inserted on
either side of the anal groove in this region. The gill-radioles are lost, but their
carriers are retained and show the scars of about a dozen radioles each. Inside the
gill-crown there is a pair of broad, pinkish-white lacinise, and below these there is a
group of about six slender tentacular cirri attached to the lower ends of the gill-
carriers.

Family: SEEPULID^E.

Eupomatus albiceps, Ehrb., Grube — Plate VII, figs. 180, 180a, and 181.
Grube, ' Monatsber. Berlin. Akad.,' 1869, p. 520 (p. 40 of reprint).

One specimen inhabiting a quadrilateral tube winding round a tube of Phyllo-
chcBtopterus ramosus from Galle.

The worm only measured 7 millims. in length including the branchiae, and a
fraction of a millimetre in diameter. Grube's example was rather larger. The
operculum gives the character of the species. In this case there are eight nearly
erect, slightly curved virgulse and a laterally compressed ovate lamina dorsalis, the
latter being a direct continuation of the columella and bearing a pair of broad dorso-
lateral chitinous hamuli. The marginal teeth of the opercular disc are blunt.
Thoracic uncini, as stated by Grube, show about nine teeth.

Another specimen, growing upon the tube of a larger species from South-east
Cheval Paar, has length of about 20 millims. by width of 1*5 millims. There are
seven virgulse on the operculum, and the lamina dorsalis is quite flattened except at
the back (Plate VII., figs. 180 and 180a). The uncini from the last thoracic torus
show seven teeth. The bayonet setse of the first thoracic fascicle show two clear
spines at the base of the terminal process (Plate VII., fig. 181).

Eupomatus exaltatus, Marenz. — Plate VII., fig. 182.
VON Marenzeli.er, 'Siidjap. Ann.,' ii., 1884, p. 217.
Station V., off Chilaw, Gulf of Mauaar, 11 fathoms; one specimen without tube.

POLYCH^ETA. 313

Differs from E. heteroceros in the structure of the operculum, other characters
being approximately the same. The branchiae are rather short, the stems stout and
close together in a digitate manner, about 16 stems in each gill; terminal filaments
slender, flagelliform. Thoracic uncini with six teeth, some with five only. The
opercular style is stout, slightly flattened ; the margin of the opercular cup carries
23 denticulations ; from the centre of the cup a stout columella rises carrying a circle
of eight large hooks, of which the dorsal one is larger than the rest and somewhat
scythe-shaped, with a long hook directed ventrally and mesially ; the seven smaller
hooks have curved extremities directed outwards and are destitute of accessory
processes (Plate VII., fig. 182). The operculum is associated with the right gill ; no
rudiment could be found on the left side. The first fascicle contains besides a few
simple capillary sctse a bundle of strong unequally bifid setae, the terminal process
long and acuminate, the subterminal process very short and obtuse ; these are called
bayonet setse (von Marenzeller).

The Ceylon specimen varies from the type, in which there is a crown of nine hooks
instead of eight, and the thoracic uncini show seven or eight denticulations. In spite
of these differences, which are not outside a possible range of variation,* the character
of the crown of hooks is so distinctive as to leave no doubt as to the identification,
notwithstanding the fact that von Marenzeller unfortunately omitted to state
explicitly that the hooks of the crown, except the dorsal hook, are directed outwards ;
this is left to be inferred, and it is an important inference, since it decides the species.

Eupomatus heteroceros, Grube.

Grube, 'Verh. zool.-bot. Ges. Wien,' 1868, p. 639, Taf. 7, fig. 8.

Locality : — Several specimens from South-west Cheval and East Cheval, 8 fathoms.

Tube round, except flattened surface of attachment, doubled upon itself, U-shaped,
or slightly convoluted and faintly ridged ; it shows coarse growth rings and is
overgrown with Bryozoa and pearl-oyster byssus. Thoracic uncini sexdentate. The
opercular disc carries a crown of seven hooks turned inwards, of which one is larger
and plain, the others equal among themselves and each provided with a pair of
accessory lateral hamuli, and an inwardly directed basal process. The lateral hamuli
are inserted lower down the shafts of the coronal hooks than in Grube's figures, and
the basal hamules appear less upturned. The latter can hardly be seen without
bisecting the operculum. The bayonet seta? of the first fascicle have bicuspidate
processes as figured by Grube. The uniserial thoracic uncini, not observed by Grube,
show five and six teeth, rarely seven. The marginal denticulations of the opercular
disc end in rounded spatulate expansions.

The collection contains many examples of this species. In one series of seven
individuals from the South-west Cheval, 13th November, 1902, one specimen has
eight hamuliferous hooks on the operculum, in addition to the great hook instead of

* Compare E. albiceps.

2 s

314 CEYLON PEAEL OYSTER REPORT.

the typical six. Two others have seven such hooks in addition to the simple hook.
Meristic variations affecting a specialised organ composed of a limited number of parts
are of considerable interest, especially in view of the analogy presented by Eupomatus
exaltatus in this respect. Compare also the thoracic segments of Sabellastarte indica.
Examples from East Cheval show six, seven and eight hamuliferous hooks respectively,
in addition to the main hook.

The collar-margin is plain ventrally, not projecting forwards between the gills.
The anterior free flap of the collar (seen from below) can be distinguished by a
transverse groove from the posterior portion of the buccal segment, and at the sides
of this groove there is a pair of thoracic organs analogous to those observed in
Pomatostegus actinoceros. The orifice is guarded by a small triangular papilla which
occurs at the level of the incisura lateralis of the collar.

Eupomatus minax (Grube).

Locality : — South-west Cheval Paar.

Small involved round tubes showing coarse growth-rings ; pearl-oyster byssus
sometimes attached to the tube. The marginal spines of the opercular disc vary
between 19 and 21. The columellar spines are more erect than in Grube's figure
('Ann. Semp.,' 1878, p. 269): the large dorsal spine is vertical and has a strong
recurved hook and two lateral accessory hooks.

Pomatostegus actinoceros, Morch — Plate VIII., figs. 3 and 4.

Morch, Otto A. L., 'Revisio critica Serpulidarum,' Copenhagen, 1863, p. 400, plate xi.,
fig. 16 ; Grube, 'Ann. Semp.,' 1878, p. 271.

This is an extremely variable species in regard to the structure of the operculum,
and on the other hand it is the exact Indo-Pacific counterpart of the Antillean species
Pomatostegus stellatus, Abildgaard (see Ehlers, ' Florida-Anneliden,' 1887, p. 296).

This species and Eupomatus hetc.rocerus are the most abundant Serpulids in
Professor Hekdman's collection, and they must play an important part in consolidating
the pearl banks as well as serving as a base for the attachment of the pearl oysters.
This is proved by the byssi which still cling to many of the calcareous tubes secreted
by these worms. The tube is massive, coarsely rounded or trigonal, rugged and
overgrown. Sometimes the tube is provided with a high laciniate keel. The branchiae
are spirally rolled and banded. Their appearance varies according to the preservation.
The transverse striation sometimes indicated is due to contraction, which causes
wrinkling of the strong cuticle which covers the outer surface of the gills. The
radioles have an internal septate structure, but the superficial wrinkles are not related
to it, and the surface may be quite smooth. In one specimen from the Muttuvaratu
Paar, brought up by the divers, I counted 36 radioles in one of the gills. The thoracic
uncini showed as many as twelve teeth in addition to the basal, transversely expanded,

POLYCTLF-TA. 315

scalprate process which von Marenzeller calls the " Meisselzahn." In the last
thoracic torus the teeth of the uncini were not 80 numerous as in the preceding
segments, not exceeding nine.

The collar is approximately as long as the rest of the thorax, entire below and at
the sides, open above and continuous with the thoracic membrane on each side.
The operculiferous style carries a broad wing-like membrane on each side which
terminates above in a short free simple lobe. This membrane can embrace and
protect the gills when retracted within the tube.

The disc of the operculum is covered with a chitinous cuticle and carries a long
narrow columella arising excentrically near the dorsal side and furnished with circlets
of spines at intervals (Plate VIII., fig. 4). The circlets of spines may be webbed or
free. In the former case the chitinous membrane which they support projects beyond
the spines and forms a stage or disc equal in diameter to the actual basal disc of the
operculum. In the specimen before me the columella carries four stages above the
opercular disc and two circlets of spines above the topmost stage. Another from the
same locality shows two stages above the disc and three free circlets of spines above.
A third has one stage above the disc and six circlets of free spines beyond. A fourth
has three stages above the disc and two circlets beyond. A fifth has one stage above
the disc, then two free circlets, then two more, narrower, stages followed by a terminal
circlet. The abdominal seta?, as mentioned by Grube, have the apical portion slightly
marked off. They closely resemble the abdominal setae of Omphalopoma langerhansii
figured by Marenzeller (' Siidjap. Ann.').

In another example taken from the Muttuvaratu Paar the opercular style retains a
roseate flush in the preserved state. The tube has a low dorsal keel. The columella
of the operculum carries two corneous stages over the disc, surmounted by five circlets
of spines. The radioles of the gills are disposed in a simple spiral, about 27 in each
gill ; they are slender and the surface is smooth. Thoracic uncini from the sixth (last
thoracic) torus show 7 to 9 teeth in addition to the scalprate process. A specimen
from the South-west Cheval Paar shows three stages above the opercular disc
followed by three circlets. Others were obtained from the East Cheval Paar in
8 fathoms.

The buccal setae of this species are delicate, forming a small, frequently incon-
spicuous bundle of about a dozen setae, among which may be found some slightly
distinguished as bayonet setae. On each side of the buccal segment, on its ventral
aspect, there is a large orifice bounded by prominent lips (Plate VIII., fig. 3).
The collar-margin projects forwards as a tongue between the gills ventrally as in
Spirobranchus cervicornis.

An individual was taken off Panadure, Station XLV., 25 fathoms, measuring
20 millims. in side view from the free edge of the collar to the posterior extremity of
the body; the collar and thorax 7 millims. ; the operculum projects 7 '5 millims.
beyond the collar; width of thorax 4 5 millims. The brachysomatic condition of the

2 s 2

316 CEYLON PEARL OYSTER REPORT.

abdomen may be due to regeneration. Between the whorls of the operculum a small
Annelid was found by Mr. Hornell (see Haplosyllis spongicola).

Protulopsis palliata, n. sp. — Plate VII., figs. 183 to 185.

Associated in the trawl with PhyllochcBtopterus ramosus from Galle, Station
XXXIX., 16 to 30 fathoms.

A round, slightly curved calcareous tube, attenuated behind, widened in front. The
body without the gills is 19 millims. long, 3 millims. maximum diameter. Gills with
between 30 and 40 radioles on each side, rolled inwards in a single spiral turn ; no
operculum. Thorax consists of seven segments carrying seven pairs of capillary
fascicles, but I do not find the uncini commencing before the fourth segment. The
first fascicle contains numerous limbate capillary setae of the usual form. In the
fourth fascicle, in addition to the ordinary setae, there is a group of salmacine seta?
(Plate VII., fig. 183). There are about twelve of this kind of setae in the fascicle,
characterised by a short normally limbate tract followed by a clear striated border
extending to the tip of the seta. The striated border is not limited by a refringent
edge ; the latter ceases suddenly at the upper end of the limbus. The uncini show
about twenty equal denticulations ; the last tooth at the base of the uncinus is
followed by a long characteristic spur (Plate VII., fig. 184). The abdominal setae
occur to the number of five in a fascicle (Plate VII. , fig. 185). The collar, in com-
bination with the thoracic membrane, shows a remarkable development dorsally, where
it is produced forwards on each side into a wide lappet, which is rolled upon itself
and is probably able to follow the branchial spire to its termination ; the ventral
portion of the collar is marked off from the dorsal portion by an incisura lateralis,
and its ventral border is slightly concave.

Serpula granulosa, Marenz. — Plate VII., figs. 186 and 186a.

Von Marenzei.ler, ' Siidjapan. Ann.,' ii., 1884, p. 215 (p. 19 of reprint).

Locality : South-west Cheval Paar ; several specimens.

Tube round, subcristate to cristate ; fila branchialia (radioles) about 26 on each
side* ; operculum shallowly concave, with 46-52 rays which project as denticulations
at the margin. The grooves which separate the rays do not all reach to the centre
of the disc ; they are the superficial indications of dissepiments which project
vertically with a free inner border into the substance of the operculum (Plate VII. ,
fig. 186a). Seven thoracic segments ; modified setae of the first segment unequally
bifurcate ; capillary setae of the other thoracic segments simple, limbate ; thoracic
uncini 5-dentate. Serpula granulosa differs from S. yervaisii, Qfg., by its cristate
tube and the shallow cup of the operculum. The degree of concavity can be varied,
sometimes the disc is nearly flat, sometimes slightly convex, but it does not seem
likely that it could be deepened to the extent which characterises S. gervaisii.f

* Vo\ Marenzeller gives 35 for the type.

t Cf. Grtjbe, 'Ann. Roth. Meer.' (Frauenfeld), 1868, p. 640.

POLYCH^ETA. 317

In a typical example such as that figured on Plate VII., fig. 186, the total height
of the tube in the region of the orifice is about 6 millims., of which the crest occupies
1 millim. The operculum is sometimes dextral, sometimes sinistral in position, and
at the corresponding point on the other side, as noted by von Marenzeller, there is
usually a rudimentary operculum. Minute tubercles are sparsely distributed on the
concave opercular disc.

Serpula watsoni, n. sp. — Plate VII., fig. 187, and Plate VIII., fig. 6.

Locality : — Trincomalee, February, 1902.

I have much pleasure in dedicating this species to Mr. Arnold T. Watson, whose
drawing illustrates it (Plate VIII., fig. 6). It is characterised by the great length of
the ampulla of the operculum, 3 millims., slightly exceeding the length of the style,
the total length of the operculum being 5-5 millims. The ampulla is about twice the
length of that portion of the style which rises above the collar. The disc of the
ampulla is traversed by about 25 rays. There are about 30 radioles in each gill.
The collar is entire below, divided on each side by an incisura lateralis. Uncini
from the last thoracic torus with 5 and 6 teeth (Plate VII., fig. 187).

Spirobranchus cervicornis, n. sp. — Plate VII., figs. 188 to 192.

One specimen in bottle with Loimia vaiiegata, Dasyckone cingulata, &c. [? Gheval].

Total length 22 millims., made up as follows : — Gill-apparatus (including operculum)
7 millims., thorax 5 millims., abdomen 10 millims. Width of thorax 4 '5 millims.
The gills diverge outwards ; the gill-rays (radioles) are rolled inwards at the top and
are disposed in a spiral of one turn and a half, about 30 in each gill. The thorax
consists of seven setigerous segments with six pairs of uncinigerous tori. The first
segment is produced into a capacious collar open above, its dorsal ends overlapping
and covering the expanded style of the operculum, nearly reaching to the disc
(Plate VII., fig. 188). Ventrally the collar-membrane is produced forwards as a
tongue-like process between the gills (Plate VII., fig. 189). The buccal seta? (of the
first fascicle) are of two kinds, long, slender, capillary setse fringed with hair-like
striae though without a definite limbus ; towards the tips of the setae the marginal
hairs project; secondly, stouter bayonet setse with pilose extremities (Plate VII. ,
fig. 191). The remaining thoracic setse are of the common limbate type and call for
no remark ; there are two groups of different sizes, more slender and stouter, in each
fascicle. Thoracic uncini with fifteen teeth and a basal T-shaped mucro (Plate VII.,
fig. 192). The abdominal setse are of the kind called " Dutenborsten " [geniculate
setse] by von Marenzeller (' Siidjap. Ann.,' ii., Taf. iv., fig. 46).

In S. cervicornis the horns of the operculum resemble a pair of antlers, of which
the dorsal tine is half the length of the main tine (Plate VII., fig. 190). They only
differ in proportion from the horns of Spirobranchus giganteus as figured by Morch
(1803) and Ehlehs (1887).

m 8 CEYLON PEARL OYSTER REPORT.

It is not easy to define the limits of the genera Pomatoceros, Phtlippi (1844) and
Spirobranchus, Blainville (1818). Grube uses Pomatoceros in an extended sense
and appears to disregard the prior claims of Spirobranclius ; the latter is retained by
Ehlee*s (1887) and St. Joseph (1894). The genus Pomatoceros s. str. comprises the
species crucigera, Grube (Ked Sea, 1869), helicoides, Marenzeller (Japan, 1884),
triqueter, L. (Europe, Morch, 1863, St. Joseph, 1894), and bucephalus, Morch
(Philippines, 1863). Grube's Serpula quadricornis ('Ann. Semp.,' p. 275) appears
to me to be probably co-specific with Morch's Spirobranchus semperi (' Kevisio
Serpulidarum,' 1863, p. 405). The species of Pomatoceros named above practically
resolve themselves into two main groups represented respectively by the European
P. triqueter and the Oriental P. bucephalus.

Spirobranclius semperi, Morch.

Morch, 'Revisio Serpulidarum,' Copenhagen, 1863, p. 405.
Serpula quadricornis, Grube, 'Ann. Semp.,' 1878, p. 275.

Opercular style with wing-like expansions as in the preceding species, opercular
disc flat, carrying four distinct horns. Uncini from the last thoracic torus with
12 to 13 teeth.

Spirobranchus semperi, var. acroceros, no v. — Plate VII., fig. 193.

This variety is represented by several specimens apparently only differing from the
typical form in the fact that the opercular disc is cone-shaped, carrying the horns at
the top (Plate VII., fig. 193). Length 12-5 millims. (up to nearly 20 millims.) ; about
50 abdominal segments ; gill with 1 8 radioles.

Spirobranchus tricornigerus (Grube).

Serpula tricornigera, Grube, 'Ann. Semp.,' 1878, p. 273.
A small specimen from the pearl banks. The great feature of this species is the
horizontal branching of the horns of the operculum, the main branches very slightly
elevated and the ultimate ramifications lying approximately in one plane. The horns
are very pale, calcareous, slightly chitinised. It is possible to distinguish three main
branches proceeding from a common centre, but not so equilateral as in Grube's
figure. Grube assigns 20 teeth to the larger thoracic uncini. I only see the usual
1 2 teeth in the uncini of the last thoracic torus. The buccal setae forming the first
thoracic fascicle are of two kinds, simple capillary and bayonet setae, both kinds with
serrulate border. The gills are traversed by a strong, nearly black fascia a short
distance in front of the basal membrane.

Vermilia pygidialis, n. sp. — Plate VII., figs. 194 to 196.

Locality : — South-west Cheval Paar.

Distinguished by the long, brown, horny, ringed, conoidal operculum, and by the
obtuse posterior end of the body, which is furnished dorsally with an oval purplish-

POLYCH^ETA. 319

crimson cushion and long hair-like setae (Plate VII., figs. 194 and 196). Ten radioles
in the left gill ; nine and the operculum in the right. Uncini from last thoracic torus
with eleven teeth and a scalprate process (Plate VII., fig. 195). Capillary setae of
first thoracic segment not distinguished from the rest ; simple limhate, accompanied
by slender non-limbate setae. Mid-abdominal region with dorsal uncini and ventral
fascicles containing three geniculate setas. Total length of operculum 5 millims., the
style 2'5 millims., the ampulla and columella 2'5 millims. Length of gills 4 millims.,
of thorax 4 millims., of abdomen 1T5 millims. The terminal filaments of the gill-
radioles are clubbed ; in one specimen the clubs have a pinkish colour (Plate VII.,
fig. 194). The closeness of the intersegmental grooves gives the pygidium a foliate
appearance. The tube is coiled horizontally upon itself, widens out in front, and
presents 4 or 5 low longitudinal keels.

EXPLANATION OF THE PLATES.

PLATE I.

Fig. 1. Chlceia flam. Dorsal barbed bayonet seta. Zeiss 3 C.
„ 2. „ Ventral furcate seta. Z. 3 C.

„ 3. Hermione malleata. Dorsal glochideal seta ; the point is broken. Z. 3 C
„ 4. „ Ventral furcate seta showing accessory tooth. Z. 3 0.

„ 5. Pontogenia indica. Anterior end.

„ 6. Iphione muricata. Anterior end.

„ 7. Lepidonotus carinulatus. Prostomium.

„ 8. „ Carinulate papillae of elytron arising from clear areoles.

„ 9. „ Echinulate papilla.

„ 10. ,, Stellate papilla.

11. ,, Ventral seta. Z. 3 C.

„ 12. Halosydna zeylanica. Parapodium.
„ 13. „ Ventral seta. Z. 3 D.

„ 14. Harmothoe dictyophora. Elytron ; the superficial filiform papillae occurring on the larger shields

are omitted.

„ 15. Harmothoe dictyophora. Head.
„ 16. „ Ventral seta. Z. 3 C.

„ 17. Hololepidella commeiisalis. Dorsal seta. Z. 3 C.
fi 18. „ Superior ventral seta. Z. 3 C.

n 19. ,, Inferior ventral seta. Z. 3 C.

,. 20. „ Head and proboscis.

„ 21. Panthalis melanonotux. Head and proboscis.
n 22. „ Fifty-fourth parapodium. a, geniculate appendix; b, dorsal cirrus;

c, silken threads (tomcntose setse) ; d, notopodium ; e, notopodial acicula ; /, neuropodium ;

g, neuropodial acicula ; h, ventral cirrus ; i, coil of threads inside the body.

320 CEYLON PEARL OYSTER REPORT.

Fig. 23. Panthalis melanonoius. Lateral processes of four successive segments after the 28th, showing
dorsal cirri with geniculate appendices, alternating with elytrophores and ovate processes ;
seen from above.

„ 24. Panthalis melanonoius. Superior ventral seta.

25. ,, Penicillate seta.

26. „ Aristate seta.

,, 27. ,, Inferior ventral seta.

,, 28. Panthalis nigromaculata. Anterior end.

_,, 29. „ Head.

„ 30. „ Jaws.

,,31. „ Third elytron.

„ 32. ,, Parapodium.

,, 33. Psammolyce zeylanica. Two of the dorsal dermal papillae

„ 34. „ Head and portion of the anterior segments.

PLATE II.

Psammolyce zeylanica. Dorsal cirrus of third segment. Z. 3 a.*

,, Compound seta from second segment. Z. 3 C.

„ Shaft of compound seta from third segment. Z. 3 C.

,, Seta from central fascicle of neuropodium.

„ Inferior ventral seta.

„ Appendices of bidentate seta?.

„ Lower border of parapodium with ventral cirrus.

„ An elytron.

,, Seta from central fascicle of 16th foot with cusp on shaft.

Psammolyce rinida. Seta from central fascicle of 16th foot with plain shaft.

,, Dorsal cirrus of third segment. Z. 3 a.*

,, An elytron.

,, Compound seta from second segment. Z. 3 C.

Sthenelais zeylanica. Parapodium: — v.c, ventral cirrus with its associated stylodes; 1, 1, 1,
superior dorsal seta? (numerous, forming a dense tuft) ; 2, 2, inferior dorsal seta?, slender,
striated ; 3, superior ventral seta ; 4, seta of the central group ; 5, inferior ventral seta.
Sthenolepis japonica. Parapodium.
Tkalenessa digitata. Superior ventral seta ; the upper end of the shaft is faintly fringed. Z. 3 D.

,, One of the digitate fimbria? of an elytron.

,, Anterior end : — e., first elytrophore ; p., palp ; x., semilunar process of the

second foot arching over the prostomiuni. The porrect cirrophores carry no visible seta?.

PLATE III.

Thalenessa stylolepis. Head.

,, Parapodium showing the ctenidium on the medial surface of the elytro-

phore and the cirriform branchia below the elytron ; the marginal fimbria? of the latter are
concealed in this preparation.
Thalenessa stylolepis. Compound seta from the 60th segment. Z. 3 C.
,, Marginal plumose fimbria of an elytron. Z. 3 A.

Fig.

35.

J J

36.

3)

37.

11

38.

11

39.

11

40.

11

41.

11

42,

11

43.

11

44.

11

45.

11

46.

11

47.

11

48.

11

49.

11

50.

U

51,

11

52

Fig.

53.

)J

54.

11

55.

11

56.

POLYCPUETA. 321

Anaitis a ylaaiea. Seta in side view.

„ Articular portion of seta from above. Z. 3 D.

„ Prostomium.

„ Dorsal phvllode.

NiotophyUwm laciniatum. Head, as seen after removal of the anterior phyllodes.

„ End of shaft of seta showing the articular fossa.

I'hiillodoce dissotyla. Head.

,, Parapodium.

„ Portion of seta. Z. 3 D.

„ The two kinds of papilla? on the proboscis.

PhijUodori ftilinsiijHijiillafa. Head and proboscis.
,, Parapodium.

Seta. Z. 3 D.
Phyllodoce macrolepidota. Parapodium.

Seta. Z. 3 D.
Phyllodoce sanctijosephi. Head.

Seta. Z. 3 D.
Iinia limicola. Parapodium; the bundle of fine dorsal seta? proceeding from the cirrophore is

seen crossing the basal portions of the superior ventral seta;. Z. 3 a.*
Trma limicola. Average compound seta.

,, Inferior ventral seta.

Typosyllis taprdbanensis. Seta from posterior region of larger specimen. Z. 3 D.

,, Posterior seta of smaller specimen. Z. 3 D.

Haplosyllis spongicola. Head.

„ Pharyngeal orifice and tooth.

PLATE IV.

Fig. 81. Autolytus orientalis. Magnified about 1 1 times.

„ 82. „ Fore-gut removed from body.

„ 83. ,, Palps, mouth and tentacular cirri from below.

„ 84. „ Seta. Z. 3 J. water imm.

., 85. Nereis unifasciata. Outline of 8th parapodium of right side seen from behind.

„ 86. „ Similar outline of 31st parapodium.

„ 87. „ Heterogomph spinigerous seta from 8th foot. Z. 3 D.

„ 88. „ Hemigomph seta from 18th foot. Z. 3 D.

„ 89. Ceratonereis falcaria. Dorsal seta from 27th foot. Z. 3 D.

„ 90. Ceratonereis pectinifera. Outline of 8th foot of right side.

„ 91. ,, Outline of 32nd foot.

„ 92. Platynereis bengalensis. Dorsal falcigerous seta of 44th foot. Z. 3 D.

„ 93. „ Superior ventral falcigerous seta of same foot. Z. 3 D.

„ 94. „ Acervus of paragnaths of group IV., and maxilla.

„ 95. Diopatra amioimensis. Jaws protruding, seen from below. 1, 1, mandibles or forcipate jaws;

2, 2, serrse or saws; 3, impar; 4, 4, arcs; 5, 5, lamina; ventrales; b.s., buccal segment.

„ 96. Diopatra amboim n.<i*. First parapodium. Only a few seta; are indicated.

„ 97. „ Simple bidentate seta; from first foot. Z. 3 D.

„ 98. Omiphis basipicta. First right foot. Z. 3 A.

,, 99. „ Compound setn of first foot. Z. 3 D.

2 T

Fig.

57

>»

58.

)»

59,

JJ

60.

H

61,

)J

62.

»J

63.

»1

64.

a

65,

)»

ijG.

J)

67.

j»

$$.

)»

69,

)j

70.

'J

71.

)»

72.

n

73.

yy

74.

jj

75.

)j

76,

77.

>>

78.

>»

79

80.

322 CEYLON PEARL OYSTER REPORT.

Fig. 100. Onuphis dihanchiata. Acicular seta from first foot. Z. 3 D.

„ 101. Onuphis hohh ranch lata. Tube about natural size.

„ 102. Eunice martensi. Right 46th foot.
„ 103. „ Compound seta from 168th foot.

„ 104. „ Acicular seta from same foot.

„ 105. Paramarphysa mientalis. Compound seta from anterior region.

„ 106. Aglaurides fulgida. Head removed, showing submedian eyes in groups, two large lateral eyes
and three occipital antenna?.

PLATE V.

Fig. 107. Aglaurides fulgida. Right set of upper jaw-pieces from the side.
,, 108. Aracoda obscura. A jaw-piece of the first pair (left side).

„ 109. ,, Jaw-piece of the second pair (left side); fracture at the point marked *.

,, 110. ,, Right set of jaw-pieces.

,,111. ,, Anterior end.

,,112. „ Ventral seta. Z. 3 C.

,, 113. Glycera Iwncadivce. Articular ends of shafts of compound setae, of form A. Z. 3 D.
,, 114. ,, Portion of a seta of form B. Z. 3 D.

,, 115. ,, Parapodial ligules of form A. Z. 3 A.

,, 116. ,, Parapodial ligules of form B. Z. 3 A.

„ 117. Polijdura hornelli. One of the modified acicular setae of the fifth segment. Z. 3 D.
,, 118. Nbtomastus zeylanicus. Anterior end in left side view showing the half -retracted prostomium,

the " tongue " below it, the first and second achastous segments, and the first setigerous

segment.
„ 119. Notornastus zeylanicus. Unciniform or acicular seta. Z. 3 J. water iuim.
,, 120. Armandia lanceolata. Anterior end from above; 1, tentaculum impar ; 2, nuchal organ;

3, rostrum ; 4, anterior portion of metapleural fold.
121. Pohjophthalmus australis. Anterior end from above showing pigment tracks and nuchal organs.
„ 122. Nicomache truncata. Anterior fragment from the left side. Actual length 47 millims. ; width

5 millims.
,, 123. Nicomache truncata. Frontal view of head.
,, 124. Ammochares mientalis. Anterior end from below.
,, 125. ,, From above (dorso-lateral), showing three capillary fascicles in front of

the first pair of tori.
,, 126. Ghatopferus appendiculatus. Modified seta from 4th foot. Z. 3 C.
,, 127. Phyllocha'topterus herdmani. Anterior end from above. First pair of spirally coiled tentacles

are lost.
,, 128. Phyllocha'topterus herdmani. Modified seta from 4th foot. Z. 3 C.

„ 129. ,, Modified seta from 3rd foot of an aberrant individual. Z. 3 C.

,, 130. ,, Plan of one of the branchial segments.

,, 131. „ Spatulate seta from first parapodium. Z. 3 C.

,, 132. „ Abdominal uncinus. Z. 3 D.

,, 133. Phyllocha'topterus ramosus. Branching tube.

,, 134. ,, Modified seta from 4th foot in side view. Z. 3 A.

„ 135. ,, Modified seta from 4th foot of another specimen in back view. Z. 3 A.

,. 136. „ Lower portion of an uncinus from a branchial segment.

., 137. Pectinaria panava. Uncinus from a posterior torus.
,, 138. Hetcrocirrus typhhps. Anterior region from above.

POLYCTLETA. 323

PLATE VI.

Pig. L39. Cirrahihts cylindricus. Anterior region from above.

,. 140. „ Head and mouth from below.

., 141. Leprea inversa. Capillary seta from second setiger.

„ 142. ,, Geniculate seta from 45th setiger.

.. 143. Polymnia labiata. Anterior end from below ; the bases of the tentacular cirri are indicated in

front of the labium (= 1st and 2nd segments); the lateral lobes of the 3rd segment

project beyond.

„ 144. Polymnia labiata. Thoracic uncinus in front view. Z. 3 D.

,, 145. „ Uncinus from 9th torus in side view. Z. 3 D.

., 14G. Polymnia socialis. Anterior region from below ; the epistome projects in front ; tentacles omitted.

„ 147. ,, Front view of a thoracic uncinus.

,, 148. „ Side view of thoracic uncinus. Z. ■"> 1).

,, 149. Polymnia triplicata. Anterior region from the right side; ep., epistome ; lab., labium.

„ 150. „ Epistome (ep.), labium (lab.), and mouth in frontal view.

„ 151. „ Uncinus in three-quarter view. Z. 3 J>.

,. 152. „ Uncinus in side view. Z. 3 D.

„ 153. Loimia ammilifilis. Uncinus from third torus. Z. 3 D.

„ 1 r» 1 . „ Anterior region from below, tentacles omitted; in front is the epistome,

then the labium, then the lateral lobes.

,, 155. Loimia medusa. Anterior region from below.

,, 156. ,, var. angustescutata, showing additional tori and capillary fascicles.

,, 157. ,, Uncinus from first torus. Z. 3 D.

„ 158. ,, Uncinus from fourth torus. Z. 3 D.

„ 159. ,, Abdominal uncinus. Z. 3 D.

,, 160. Loimia montagui. Anterior region from below.

„ 161. „ Quinquedentate uncinus from first torus. Z. 3 D.

,, 162. „ Sexdentate uncinus from first torus. Z. 3 D.

„ 163. „ Uncinus from 13th torus. Z. 3 D.

PLATE VII.

Fig. 1G4. Grynuea cespitosa. Anterior end from the right side. The small circles in the first branchial
acervus are the scars of branchial filaments.
,. 165. Grynuea cespitosa. Uncinus. Z. 3 D.
„ 166. Branchiomma acrophthahnos. Spatulate seta from thorax.

„ 167. ,, Anterior end from above. In front the gill-bases are shown on

each side, the radioles being omitted.
/ '■' ■■!.. ■■■■''/ ■■■■ithitinn. Buccal segment and cephalic complex from below.

„ 169. „ Uncinus. Z. 3 D.

„ 170. Dasychone cingulaia. Average superior thoracic capillary seta. Z. 3 C.
„ 171. „ Thoracic uncinus. Z. 3 D.

,,172. „ Portion of a gill-radiole showing stylodes and eyes and the bases of the

gill-filaments.
„ 173. Ewato porifera. Anterior thoracic region from above.

2 T 2

324 CEYLON PEARL OYSTER REPORT.

Fig. 174. Ewrato notata. Thoracic uncinus. Z. 3 D.
,, 175. „ Thoracic capillary, broadly limbate seta. Z. 3 D.

„ 176. „ Thoracic capillary seta with the shaft broken and the limbus projecting.

„ 177. Sabellastark indica. Thoracic uncinus. Z. 3 C.
„ 177«. ,, Frontal view of cephalic crown; rf., dorsal side; /., antenna or tentacular

lacinia.

,, 178. Jasmineira cadudhranchiata. Thoracic rostrate uncinus. Z. 3 C.
,, 179. „ Abdominal uncinus. Z. 3 D.

„ 180. Enpomatus aibirepn. Operculum in side view.
„ 180«. ,, Dorsal view of operculum.

„ 181. „ Bayonet seta from first thoracic fascicle.

182. Enpniititliix rsaltutiix. Oblique lateral view of operculum.

„ 183. Protulopsis paUiata. Salmacine seta from fourth thoracic fascicle.
,,184. ,, Base of uncinus.

,, 185. „ Abdominal seta.

,, 186. Serpula granulosa. View of orifice of tube with operculum in situ.
„ 186a. ,, Operculum bisected.

,, 187 Serpula watsoni, Uncinus from last thoracic torus. Z. 3 D.

,, 188. Spirdbranch/as cervicornis. Cephalic crown in dorsal view: c, collar flap turned aside.
,, 1S9. ,, Cephalic crown obliquely from below.

,, 190. ,, One of the opercular horns in side view.

,, 191. ,, Bayonet seta from buccal segment.

192. „ Base of uncinus in side view and in frontal view.

193. Spiivbranrhus semperi, var. aeroceros. Operculum from below.
,, 194. Vermilia pygidialis. Anterior region from dorsal side.

,, 195. „ Uncinus from last thoracic torus.

,, 196. ,, Posterior end.

„ 197. Leprea inversa. Posterior abdominal uncinus. Z. 3 J. water imm. From one of the last three
or four segments.

PLATE VIII.

The figures on this plate were drawn by Mr. Arnold T. Watson, to whom the identification of the
" building organs " of Palladia pennata is due.

Fig. 1. Pailasia pennata. Side view of anterior region. /., retracted tentacles; 6, white portion of
peristome; c, brown portion; </, brown with white zebra-like stripes, x 19. a, internal
bristles crossing when head contracts.

,, 2. Pailasia pennata. Ventral view of anterior region ; b.o., building organ, x 19.

,, 3. Pomatostegus actmoceros. Ventral view; op.c, opercular cavity; .«., thoracic orifice, x 1G.

,, 4. ,, Dorsal view, x 16.

„ 5. Stylarioides parmatus. Showing recurved abdomen; the fore-body is seen from below, x 17.
a, uncini.

,, 6. Serpula watsoni. Operculum, x 18.

i EYLON PEARL OYSTER REPORT

POLYCHAETA PLATE 1

CEYLON PEARL OYSTER REPORT.

i'OI.YCHAKTA-PLATEll

■

(KYI, (>NT PEARL OYSTER REPORT.

POLYCHAETA PLATE 111

.

CEYLON l'KANI, OYSTER REPORT

POLYCHAETA I'l.A I'K IV

pooc . .;;"■; o ".uco
doooooooooooo

300000 j 3C 0 ■ -
3QOOOOOOOOOOC
3 ) JOOOO 3QOOOC
,n 300000C : "- 3C

ioooouooooooc

,(jC 1

■

CEYLON PEARL OYSTER REPORT

POLYCHAETA- PLATE V

129

CEYLON PEARL OYSTER KEI'OKT

IMU.YCHAKTA PLATE VI

160

CEYLON PEARL OYSTER REPORT

POLYCHAETA- PLATE VH

E Wilson , Cembndqe

(KYI, ON' PEARL OYSTER REPORT.

I'OI.YHIA ETA -PLATEV1II.

CD

@

■

y

^

t 325 ]

NOTE ON POLYDORA ARM ATA, Lnghs.*

BY
ARNOLD T. WATSON, F.L.S.

The specimens referred to in this note were found living commensally with a sponge,
Aulospongus tubvlatus, which is very common on the pearl banks, and is mentioned
by Professor Dendy in his Report upon the Sponges in this series (' Ceylon Pearl
I »\ ster Report,' Part III., p. 176).

A fuller description, with a figure showing the tubes piercing the sponge radially
and the worms in situ, had previously been given by Dendy in his " Report on
Sponges from the Gulf of Manaar" ('Ann. and Mag. Nat, Hist.' (6), hi., p. 73, 1889).

The general characteristics of the species are well described by Mesnil (' Bull. Sci.
France et Belgique,' tome xxix., 1896, p. 203), and I have but few points of difference
to note. Of these the most important, perhaps, relates to habitat.

The specimens described by Mesnil were obtained from Lithothamnion, through
which they had pierced and in which they had formed tubes of calcareous sediment ;
while Carazzi and Lo Bianco report these worms as living in the shells of Venus
and therein forming U-shaped tubes. A similar variation in habitat occurs in another
closely allied species of this genus, Polydora cceca, which sometimes lives commensally
with the sponge Microciona plumosa, as described by Hornell (' Nature,' vol. 47,
1892, p. 78).

Owing to the difficulty of separating the worms uninjured from the sponge, it has
only been possible to secure one or two fairly perfect specimens. The following notes,
therefore, may be incomplete.

The Ceylon worm is apparently a smaller form of the species Polydora arrnata,
Langerhans. The length varies from 2 to 3 millims. and the number of setigerous
segments from 22 to 26. The branchiae, of which I find four pairs, commence, as
usual, on the 7th setigerous segment. They are comparatively broad and sometimes,
but not always, of equal length. In one instance they gradually lengthen, the first
being only half the length of the fourth. Eye-spots may or may not be present ;
most frequently they are invisible, but, after special treatment, in one case, I
detected one pair. The tentacles are fairly long, reaching, in one specimen (though
bent and twisted), to the 7th setigerous segment,

The setae of the 5th segment, each with a characteristic hook, arising from an
upright collar, which terminates on either side in a prominent pointed process
projecting in the same direction as the hook (the whole having somewhat the
appearance of an articulated seta), correspond well with Mesnil's figure. They
number two, or sometimes three, in each parapodium.

* Langerhans, ' Zeitschr. f. Wiss. Zool.,' 34, 1880, p. 93.

32G

CEYLON PEAEL OYSTEE REPORT.

Each parapodium of the 7th and following segments bears two or three hooded
ventral uncinigerous setae. The general outline of the worm is noticeable. The
caruncle is slightly bilobed in front and extends backwards to about the 2nd

/!

Polydora armata, Lnghs. — Posterior part in dorsal
view, x 75 ; part of a dorsal posterior fascicle ;
and seven dorsal seta? from a posterior fascicle,
x 375.

setigerous segment : the anterior ten or
eleven segments (except the 5th) are
nearly equal to one another in length and
breadth ; they are followed by seven or
eight segments, which are much larger,
being nearly double the length and
breadth of the former ; in the hinder
part of the worm the proportions of the
segments become somewhat similar to
those at the anterior end. The anal
segment is broader than the preanal (see
fig. in text). It is somewhat reniform
in outline, the indent being dorsal. The
segmental divisions in the anterior part,
although fairly clear, are not deep, but
commencing with about the 13th seti-
gerous segment the separation becomes more and more distinct, until the last four or
five preanal segments, in which it is very marked. This marked separation is further
accentuated by the presence in the parapodia of these last segments of a fascicle of
from 15 to 18 stout brown setae, which vary greatly in form, size and proportion,
a fact which was not shown by Mesnil. The forms comprised are acicular, lanceolate
and scimitar-like (see tig. in text).

These setae, which are more numerous than in Mesntl's examples, have usually
their points drawn together in preserved specimens, forming a hollow, subspiral,
truncate cone ; but in several cases I have found the fascicle opened out, the points
of the setae being widely directed outwards, an arrangement extending over three
parts of a circle, the central convexity of which is directed antero-laterally. By
dissection the fascicle can be unrolled, when the setae of which it is composed are seen
to be arranged symmetrically side by side, the longest in the centre and the smallest
at the outer edges. Probably in life the fascicle is expanded with a sweeping action.
The forms of the setae seem admirably adapted to the function described, the swollen
parts, some distance from the base of the seta, doubtless acting as fulcra and
rendering mutual support and aid in the action I have suggested.

The distribution of the species as noted by Mesnil is : Atlantic Ocean (Madeira),
Mediterranean (Naples), British Channel (Manche : " Anse St. Martin"), to which we
now add the Gulf of Manaar.

HARRISON AND SONS, PRINTERS IN ORDINARY TO HIS MAJESTY, ST. MARTINS LANE.

•

r

Ymii i*H<" ••IBKARY

WH IflTZ

loqcj