i O

i o
!o

REPORT ON THE MARINE ZOOLOGY OF
OKHAMANDAL IN KATTIAWAR

UJ

O

UJ

I-

z
O

DC

Q.

_T

a

z

u.

O

>-

O
O

O
O
N

u

z

D
O
CO

(8

00

73

C
(8

C
JS

73
(8

(0

Q

k IK:

rr

REPORT

TO THE GOVERNMENT OF BARODA ON THE

MARINE ZOOLOGY OF
OKHAMANDAL IN KATTIAWAR

BY

JAMES HORNELL, F.L.S.

Government Marine Biologist, Madras, and formerly Marine Biologist to the Government oj Ceylon

WITH SUPPLEMENTARY REPORTS ON SPECIAL GROUPS BY OTHER ZOOLOGISTS

PART II

LONDON

WILLIAMS AND NORGATE
14, HENRIETTA STREET, COVENT GARDEN, W.

1916

PRINTED IN GREAT BRITAIN BY

RICHARD CLAY AND SONS, LIMITED,

BRUNSWICK STREET, STAMFORD STREET, S.E.

AND BUNGAY, SUFFOLK.

CONTENTS OF PART II

DWARKA HEADLAND AND BAY, FROM THE SOUTH Frontispiece

PAOK

INTRODUCTION vii

ERRATUM : NOTE ON THE IDENTIFICATION OF THE EDIBLE OYSTERS OF OKHAMANDAL xi

REPORTS : —

I. — THE INDIAN CONCH (TURBINELLA PYRUM LINN.) AND ITS RELATION TO HINDU LIFE

AND RELIGION. By JAMES HORWELL, F.L.S. (Seven Plates and two Text-figures) 1

II. — THE CALCAREOUS SPONGES OF OKHAMANDAL. By ARIHUR DENDY, D.Sc., F.R.S.

(Two Plates) 79

III. — THE NON-CALCAREOUS SPONGES OF OKHAMANDAL. By ARTHUR DENDY, D.Sc.,

F.R.S. (Four Plates) 93

IV. — THE HYDROIDA OF OKHAMANDAL. By Miss LAURA R. THORNELY . . . .147

V. — NOTES ON SOME JELLY-FISHES FROM OKHAMANDAL. By EDWARD T. BROWNE, M.A.,

F.L.S. . ....... . 151

VI. — THE POLYZOA OP OKHAMANDAL. By Miss LAURA R. THOENELY (with six Text-
figures) 157

To facilitate reference, the list of reports contained in Part I is appended : —

CONTENTS OF PART I

PORTRAIT OF H.H. MAHARAJA SAYAJI RAO, GAIKWAR OF BARODA . . . Frontispiece

PACK

PREFACE v

I. — THE MARINE RESOURCES OF OKHAMANDAL. By JAMES HORNELL, F.L.S. (Eight Plates) 1

II. — THE INDUSTRIAL DEVELOPMENT OF OKHAMANDAL. By V. M. SAMARTH, B.A. .' . 35

III. — THE ANATOMY OF PLACUNA PLACENTA. By JAMES HORNELL, F.L.S. (Six Plates) . . 43

IV. — DESCRIPTION OF A NEW SPECIES OF PINNOTEEES. By JAMES HOHNELL, F.L.S., and

THOMAS SOUTHWELL, A.R.C.S., F.L.S. (One Plate) . . . . .99

V. — THE ANOMURA OF OKHAMANDAL. By THOMAS SOUTHWELL, A.R.C.S., F.L.S. (One Plate) 105

VI. — THE ALCYONARIANS OF OKHAMANDAL. By J. ARTHUR THOMSON, M.A., and GEORGE

CRANE, B.Sc. (One Coloured Plate and two Text-figures) 125

VII. — THE NUDIBRANCHS OF OKHAMANDAL. By SIB CHARLES ELIOT, with a note by

J. HOENELL on the presence of Symbiotic Algse in MELIBE RANGII, with Text-figures 137

INTRODUCTION

WHEN sending Part I. of these reports to the press, I hoped to be able to include in
the second part — that which is now sent forth — reports upon the whole of the remaining
portion of the collection made during my short visit of inspection to Okhamandal
in December 1905 and January 1906. Unfortunately the outbreak of the Great German
War has made this impossible, particularly in respect of the two most important of
the remaining reports — those on the Polychaeta and the Tunicata. Indeed, I am
doubtful if the report upon the Polychaet collection will ever be available, as the latter
was in the hands of Professor Ehlers of Gb'ttingen at the time war broke out.

In these circumstances, I decided that as I had in hand a number of valuable
papers, including two long and important reports on the sponges, together with others
on the Hydroida, Medusae and Poiyzoa, as well as a lengthy one on the Indian Conch
which has considerable economic importance on the shores of the Gulf of Kutch, it
would be better to publish these at once, and relegate any further reports that may come
to hand at a future date either to a supplementary issue or contribute them to some
other Indian Zoological publication.

The remarkable faunistic richness of the Okhamandal coast is demonstrated most
emphatically by the wonderful variety of sponge, hydroid and polyzoon life catalogued
in the valuable papers contributed by Professor A. Dendy, F.R.S., and Miss Laura
R. Thornely. In this respect it must be remembered that the collections were made
single-handed in the face of many difficulties and within the limited period of five
weeks, during which my attention was in the main directed to work of another nature,
faunistic collection being of subsidiary importance.

The Government of His Highness the Maharaja Gaikwar have rain most generously
undertaken to defray the whole cost of production of the present volume, thereby
making the zoological world in general, and that of India in particular, their debtors
for this generous and liberal expenditure.

That the work done during my inspection of the sea bottom around Okhamandal
has borne useful fruit is evidenced by (a) the establishment of a highly remunera-
tive Placuna pearl-fishery on the north-east coast, (6) the development of the
local chank-fishery, and (c) the appointment of fishery officers charged to organise

viii INTRODUCTION

and develop these and other marine industries. With regard to the first-named,
it is to be noted that within a year from the inception of the enquiry the
attention drawn in my first report to the Government of Baroda on the fishery
potentialities of this outlying portion of His Highness's dominions, resulted in a com-
mencement being made in the exploitation of the large beds of window-pane
oysters (Placuna placenta) which were found .during my visit of enquiry lying to
the east of Beyt Island and Poshetra peninsula. The beds first exploited were
largely great deposits of many generations of the shells of this species which had
lived and died successively in shallow water in this locality. The pearls that had
formed during the life of these shells were still in situ between the valves and
could be obtained with ease by washing the shells and sieving the muddy sediment.
The pearls thus found, although duller in lustre than those obtained from live
shells, readily found a remunerative market, and large profits, I am pleased to
learn, have accrued to the Baroda Government. The method of working these deposits,
employed to date, has been to lease the right to collect the shells and extract the pearls
for an annual rental.

The amounts so obtained are as follows for the years named, viz :—

Revenue.
Year. Rs.

1907-08 216

1908-09 325

1909-10 251

1910-11 355

1911-12 4,000

1912-13 12,100

1913-14 25,300

1914-15 16,001

1915-16 11,600

Total Rs. 70,148

From what I know of the local conditions, I am strongly of opinion that the beds
of dead Placuna are of very great extent and that the portion exploited to the present
is an insignificant proportion of the whole. What is wanted is to devise means for the
efficient working of beds lying in deeper water, where the collectors at present find it
impossible to work. Largely on this account, I understand that of late the lessees
have turned their attention to the beds of living Placuna, from which they are getting
a good outturn of pearls.

The Baroda Government are very much alive to the importance of developing
and safeguarding this industry. To this end they have appointed a promising young
man to the charge of the local pearl-fishing industry, and he, after receiving instruction
in general methods at my hands, is now stationed in Okhamandal to carry on experi-
mental pearl-oyster culture and general fishery investigation. He is now engaged

INTRODUCTION ix

in studying the life-history details of the window-pane oyster as exhibited locally,
and has laid down considerable numbers on different bottoms in several selected
localities to test comparatively the suitability of the different locations and the
conditions which conduce to the rapid production of pearls.

A recent letter received from Mr. S. R. Gupte, the officer charged with this work,
records the interesting fact that he has found vast rubbish heaps on the mainland
opposite Beyt composed of waste from some ancient and extensive chank-bangle factories.
The fragments are largely the discarded apical and oral pieces cut off and thrown away
in the sawing of shells for bangle working ; they are similar in appearance and form
to those figured in my " Indian Conch Report " on PI. VI. (fig. 3). With them are
associated many broken fragments of unfinished bangles — the whole being typical
waste from a shell-bangle factory. That the industry was conducted on a very extensive
scale is evidenced by the great abundance of these remains ; the Government lessee
of the present-day chank-fishery is now exploiting the deposit, presumably for lime
burning, selling the material at Surat at the rate of Rs. 13 to Rs. 14 per khandi of 20
Bombay maunds (560 lb.). Strangely enough, the industry of shell-bangle manu-
facture has now entirely died out in this, locality, not a single shell bangle being manu-
factured at the present day in Kattiawar. Owing to the war having put an end to the
importation of glass bangles from Austria, the demand for shell bangles has greatly
increased in Bengal ; the result is a great appreciation in the value of conch shells,
amounting in the case of several qualities to as much as 100 per cent, increase. This
has reacted favourably upon the various conch or chank shell fisheries, and that of
Kattiawar has benefited with the rest.

It is appropriate to note here that substantial progress has been made in
the development of the local chank-fishery since I first drew attention to the
importance of this subject in my -original recommendations. In former years
barely 3,000 shells were fished annually, and most of these were disposed of to
pilgrims visiting Beyt. The greater attention now given to the industry through
the efforts of the Director of Commerce and Industry has already had gratifying
results ; from a catch of 3,000 shells the production has now risen to 16,000 per
annum, with promise of very considerable further increase. Direct relations have
been established with some of the most important of the Calcutta chank merchants,
whereby enhanced prices have been obtained to the profit alike of the Baroda
Government and the lessee.

An interesting minor shore industry not touched upon in my first report is
that of cuttle-bone collection. During the south-west monsoon the " bones " of
cuttlefish are cast ashore in great quantity on the Okhamandal coast, and as these
are in considerable request in European markets at remunerative rates, the export
is considerable. Mr. Gupte informs me that about 300 maunds are exported to
Bombay annually, say 75 cwt. at 28 lb. per maund. Seeing that recent quota-

x INTRODUCTION

tions which I have received from a leading London firm are £2 16s. per cwt.,
c.i.f. London, for selected pieces from 8 to 12 inches long, and £l 8s. per cwt.
for mixed sizes, by advice and assistance in the disposal of this product, the
Department of Commerce and Industry should be able to render it more profitable
to the collectors, thereby encouraging them to greater and more sustained efforts
and to the organisation of the trade upon systematic lines in place of the present
haphazard and intermittent method of collection.

With the enlightened encouragement of a sympathetic central Government anxious
to develop the marine resources of the State, I am confident that the fisheries of
Okhamandal have a bright future, though it is likely that progress will be slow, owing
partly to difficulties in regard to transport which render access to good markets costly
and slow, and partly to the conservatism and religious scruples of a coastal population
hitherto with little interest in fishery matters.

In conclusion I am glad to avail myself of this opportunity to thank the various
specialists who have been so kind as to furnish reports upon the collections, for giving
so freely and ungrudgingly of their time and knowledge to make their contributions
exhaustive and complete.

To the Government of His Highness the Maharaja Gaikwar I desire to express
my great appreciation of the honour done me by entrusting the original enquiry and
the general charge of the production of this publication to my hands. I trust that
their recognition of the value of pure science in the economic development of the fishery
resources of their State may find ample recompense within the near future, both in the
betterment of the condition of certain sections of the coastal population and in the
continued increase of revenue returns derived from the State monopolies constituted
by the pearl and chank fisheries of Okhamandal.

JAMES HORNELL.

GOVERNMENT FISHERIES STATION, TUTICORIN,
July 10th, 1916.

ERRATUM

NOTE UPON THE IDENTIFICATION OF THE EDIBLE OYSTERS OF

OKHAMANDAL.

THE rock oysters which densely crowd the surfaces of the rocks about half tide
level at Poshetra Point were wrongly identified in Part I, pp. 22 and 23. This species
appears really to be the Ostrea cucullata of Born and of Lamarck.

Its main distinguishing characters are as follows : Outline roughly oval ; the left
valve extensively attached, the cavity deep and cup-shaped, with a sacciform extension
into the hollow beak region of the hinge, which is moderately elongated in freely-grown
individuals ; the edges of this valve have a distinct tendency to grow upwards. Ex-
ternally the left valve is folded into deep ridges passing radially outwards from the
hinge and ending in a sharply dentate edge which interlocks closely with the edge of
the upper or right valve. The latter is. flattened and opercular in form. The muscle
scars are normally purplish-black in tint, sometimes brown, rarely white. Very
characteristic is a row of closely-set elongated denticulations seen a short distance
inwards from the margin on the inner surface of the upper valve ; these fit into a corre-
sponding series of furrows in the lower valve. Externally the shell is tinted an opaque
pinkish-purple. Internally it is white, margined with purple or black. The size is
generally small, seldom exceeding three inches in length.

The true identity of the fine edible oyster found on the muddy bottom of Aramra
Creek and named tentatively 0. cucullata in Part I, p. 22, is somewhat doubtful. It
is certainly not 0. cucullata, as that is the name of the small rock-oyster above described.
In general form and in habit it closely approximates to the common estuarine oyster
of South India, a species which appears to be identical with or extremely closely related
to Ostrea virginiana, Gmelin, the most abundant oyster of the Atlantic Coast of North
America. The Kattiawar oyster differs from both these forms by the possession of
white muscle scars, which in both 0. virginiana and the South Indian form are con-
sistently purplish black. It is the same species as is found in local abundance in the
muddy creeks of the Sind and Kutch coasts. Its shell sometimes attains a remarkable
size ; I have one valve 14 inches long. The thickness is also characteristically great.
It wili probably be found to be Ostrea gryphoides, Schlotheim, or a variety thereof.

Very little is known at present concerning the number and relationship of the
different Ostreids of Indian waters, and their nomenclature is in a state of considerable

xu ERRATUM

confusion. A review of the species is urgently needed. To this end I am accumulating
material, and I shall be very grateful for any representative specimens from different
parts of India, Ceylon, Burma, and the Persian Gulf that anyone in a position to help
may be kind enough to send. Not less than half a dozen shells should be sent from
each locality and these should comprise as great a variety of form as possible. The
typical or most common form assumed in each locality should be distinctively indicated ,
wherever possible.

J. HORNELL.

MARINE ZOOLOGY OF OKHAMANDAL, Pt. II

PLATE I (FRONTISPIECE)

[Photographed Ini firidlia Kala Mandir, Barmla.

Figs. 1, 2 and 3.— Sinistral chanks respectively in the Satya Bhamaji, Shank Narayan

and Lakshmi Temples, Bet, Kathiawar.

THE INDIAN CONCH

(TURBINELLA PYRUM, LINN.)

AND ITS RELATION TO HINDU LIFE

AND RELIGION

JAMES HORNELL, F.L S.,

Marine Assistant, Department of Fisheries, Madras, and formerly Marine
Biologist to the Government of Ceylon

[WITH SEVEN PLATKS AND Two TEXT-FIGURES]

PARE

INTRODUCTORY 2

I. — LIFE-HISTORY AND LOCAL RACES OF THE CHANK . . . . ' . . 3

II.— THE SOURCES OF INDUSTRIAL SUPPLY 8

III.— THE ROLE PLAYED BY THE CHANK IN INDIAN RELIGION AND LIFE 11

(1) LEGENDARY AND HISTORICAL . . . . . . . . . 11

(2) PRESENT DAY USES : —

PAOK

(a) IN RELIGIOUS CEREMONIAL AND (g) TOTEMS 33

VULGAR SUPERSTITION (IN- (A) EVIL-EYE SUPERSTITIONS .... 34

CLUDING MARRIAGE AND ((') PERSONAL ADORNMENT 33

DEATH RITES) ... 18 (j) FEEDING SPOUTS . . . . . . 43

(6) BRANDING AND TATTOOING . 25 (k) As CURRENCY AND IN ARMORIAL BEARINGS 43

(c) THE MENDICANT'S CONCH . 28 (/) LIME-MAKING 45

(d) DEDICATION OF TEMPLES AND (m) MEDICINE 45

HOUSES 29 (n) FOOD 47

(e) MARRIAGE CEREMONIES 30 (o) INCENSE STICKS 47

(/) DEATH CEREMONIES ... 32 (p) ASSEMBLY CALLS, ETC 43

IV. — THE USE OF CHANK BANGLES -

(1) IN NORTHERN INDIA 49

(2) IN THE MADRAS PRESIDENCY 55

(3) THE ANTIQUITY OF THE CHANK-BANGLE INDUSTRY : —

(a) IN THE TlNNEVELLY DISTRICT IN THE BEGINNING OF THE CHRISTIAN ERA . . .68

(li) ITS FORMER EXISTENCE IN THE DECCAN, GUJARAT AND KATHIAWAR . . . . . 64

V. — THE SCOPE FOR THE DEVELOPMENT OF A SHELL-BANGLE INDUSTRY IN OKHAMANDAL 74

INTRODUCTORY

To all Hindus the important position occupied by the Sacred Conch (Turbinella
pyrum, Linn.), the Sankha of Sanscrit literature and the Sanku or Chanku of Tamil
speech, in their religion is a commonplace of everyday knowledge ; few, however, are
aware of the intimate relationship it bears to a hundred common incidents in the ordinary
life of the people in many localities and among widely sundered races, tribes, and castes.
In the following pages an attempt will be made to survey this little known by-way in
the life and history of the Indian world, to show how superstition looks upon the Sankha
as an amulet against the powers of evil, how this belief is among the oldest and most
tenaciously held by Animist and Hindu, by Muhammadan and by Buddhist ; to indicate
the way this belief has brought the shell into prominence in the Hindu religion, and to
detail how it subserves as well, a hundred different uses in the daily lives of millions of
Indians — how it is associated with infancy, marriage and death, and finally how its
employment in the form of ornamental bangles was once the subject of an important
industry in several widely separated parts of India, prominent being Kathiawar and
Gujarat where to-day all memory of it has vanished.

In connection with this latter point I trust that the attention now drawn to it,
may result in the revival in Kathiawar of an industry that has more good things than
usual in its favour if we consider other aims than that of mere money-making ; no objects
manufactured in India are more artistic and pleasing than the handsome milk-white
bangles made in Dacca workshops for the ladies of Bengal.

In the course of collecting the materials for this essay, difficult problems have
taken form. Some we can solve, but others remain obscure ; among the latter may be
mentioned the unknown cause for the cessation and disappearance of the chank bangle
industry in Kathiawar, Gujarat and the Deccan, and the question whether the use of
chank bangles among a few sections of several castes in South India is in the nature
of a survival of a once universal custom.

Another point to which I wish to draw attention is the bearing upon the antiquity
of trade relations between India and the Persian Gulf in the recognition I made last
year of several exhibits in the Louvre as consisting of objects carved from the shell of
the Indian conch. True it is that they go back only some 500 years B.C. but it is some-

HORNELL— THE INDIAN CONCH 3

thing definitely accomplished when we can say with certainty that at this date we have
positive evidence of the employment of the Indian chank, probably in religious service,
in one of the great capitals of the Achsemenid dynasty of Persia. An examination
of remains from Babylonian and Assyrian sites with a view to discover articles cut from
this shell might well have profitable and important historical results.

I am well aware of the many lacunae in the story I offer, particularly in regard to
ancient references to the Sankha preserved in the Sanscrit classics ; there may also
be many valuable passages in ancient Gujarati works ; these were sealed books to me
and I can but hope that others with the necessary qualifications may make good what
is wanting in the present account.

LIFE-HISTOKY AND LOCAL RACES

THE Indian Conch, Turbinella pyrum, Linn., is a handsome gastropod mollusc found
in Indian waters in large numbers in comparatively shallow water. Its geographical
distribution is peculiar ; on the west coast of India large numbers are fished off the Kathia-
war coast, but southward of this, we find no trace of the chank till we reach the southern
coastline of Travancore where this shell again appears and forms the object of a
small fishery. On the East Coast of India its distribution is more extensive, being found
and fished everywhere from Cape Comoriu to Madras City. The northern limit on this
coast may be placed at the mouths of the Godaveri, where I have found a few shells,
all marked by stunted growth — individuals living in an unfavourable environment.
The northern shores of Ceylon, from Puttalam in the north-west to Trincomalee on the
north-east, yield large numbers of this shell ; it is also to be found at the Andaman
Islands.

The bottom most favoured of Turbinella pyrum is a sandy one containing a moderate
proportion of mud ; this character of bottom is admirably suited to the luxuriant growth
of tube-building polychset worms which constitute the main food supply of the chank.
These polycheets are of several genera, the most abundant being Terebellids. In some
places in the Gulf of Mannar, square miles of sea-bottom are monopolised by these
Terebellids and the chanks and echinoderms which prey upon them ; a veritable Tere-
bella — Turbinella — Echinoderm formation. The edges of rocky reefs and the sandy
patches interspersed among the rocks are other favourite haunts of chanks as the worms
on which they prey are usually abundant there.

A -2

4 OKHAMANDAL MARINE ZOOLOGY— PART II

The shell of the adult chank is characteristically thick-walled and massive ; in the
live condition the exterior is covered with a dense brown and velvety horny layer, the
periostracum ; after death this dries, becomes brittle and eventually peels off, so that
shells exposed for any length of time on the sea-shore or buried for any considerable
period, become naked and reveal the characteristic snowy-white porcellaneous nature
of the shell clearly. In fully grown shells the rows of chestnut brown spots which are
normally very distinct on the outer surface in the immature, tend to disappear and often
become entirely obliterated. Around the mouth, especially along the inner edge of
the lip, a faint pink tint is often seen in fully mature shells, while occasionally the whole
interior surface of the mouth may assume a brick-red colouration particularly in shells
from certain localities. Characteristic of this genus are the three strong columellar
plications or ridges upon the columella ; the trace of a fourth is also frequently present.

Local races. — So considerable are the divergences in external form and colour shown
by shells of normal individuals of Turbinella pyrum that conchologists have made several
pseudo-species out of the principal variations. Intimate acquaintance, extending
over a period of upwards of ten years, with all the chank-producing localities on the
coasts of India and Ceylon, together with the advantage of having scrutinised several
hundred thousand shells, convince me that all the so-called species of Turbinella found
in Indian seas are referable to a single species — T. pyrum — and constitute, at the most,
varieties of the nature of emphasized local races.

Three types exist which may be conveniently described as the obtuse, the central
and the elongate. All three forms have their own particular geographical distribution
and characteristic physical and biological environment. Normally they respectively
occupy well-defined localities. Intermediate forms linking the different main varieties
frequently occur, especially where the habitats of two varieties march, forming a debat-
able land where it becomes difficult or even impossible to assign the shells there found
to one or other variety, so evenly are the respective characteristics balanced. But the
vast bulk of the shells fall readily into one or other of the three types.

(a). The central variety preponderates largely over the other two ; both numeri-
cally and industrially it is the most important and most valuable, being the form specially
valued by the shell-bangle manufacturers of Bengal for the purposes of their trade.
This type forms the bulk of the immense number fished off the Indian coast of the Gulf
of Mannar and around the north of Ceylon, a number averaging some 1| million
annually. Its range on the Indian 'coast is straitly limited on the north, where
Pamban Pass and the north coast of Rameswaram Island at the head of the Gulf of
Mannar mark the boundary, thence southward it extends along the shores of the
Ramnad and Tinnevelly Districts. North-westward of Pamban Pass this variety
marches with that which I shall term obtusa. Off Ceylon this central form is
restricted to the northern coasts from the Pearl Banks and Mannar round to
Trincomalee on the north-east coast.

MARINE ZOOLOGY OF OKHAMANDAL, Pt. II

INDIAN CONCH, PLATE II

Figs. 1 and 2.— Immature chanks from Okhamandal ; ^both
show persistence of the protoconch.

Fig. 3. — The Indian Chank, Tur-
binella pyrum, Linn., central
type of form.

Fig. 4.-Elongated variety from the
Andaman Islands. Adult and two
very young specimens. The latter
show the protoconch persisting.

[Fiffl. 1, 2, and 3 plintoffra/ilied by Viv'ull'ii Kula MamH,-, Biii-mln : f'iri, 4 '»/ iMlinn Musev.tn, Calcutta.]

HORNELL— THE INDIAN CONCH 5

The central type has a heavy massive shell, with a well-proportioned and well-
defined apical spiral. Apart from the chestnut coloured surface-markings which are
extremely variable and tend to disappear with age, these shells are of an opalescent
lily-white colour, porcellaneous in texture and slightly translucent when cut into sections.
Unlike var. obtusa the colour of the body at the mouth is not deeply tinted ; usually it
is of a very pale flesh tint or even quite white, whence comes the Tamil name Vel-vayan
or " white-mouth " used to designate this variety, which again is split into several
local races whereof the principal are known locally as (1) big-bellied chanks and (2)
small-bellied chanks. The former come from the beds lying off the mouth of the Tam-
braparni River, the other chiefly from the northern beds off Tuticorin.

A second location of the central type is found on the coast of Kathiawar, separated
from the Gulf of Mannar and Ceylon area by hundreds of miles of sea bottom from which
all forms of Turbinella pyrum are absent.

(b). Turbinella pyrum, variety obtusa, is very distinctive in shape and sometimes in
colouration. The apical spire is emphatically abbreviated, the whorls appearing as
though telescoped while the colouring of the body at the mouth opening is frequently of
a dark brick-red in shells from the Tan j ore coast, most distinctive when compared with
the delicate pallor of this region in shells of the central type.

The distribution of the obtusa variety includes the littoral waters of the eastern
coast line of the Madras Presidency from Pamban Pass in Palk Bay to the delta of the
Godaveri River in the north ; the chief fishing grounds are situated in Palk Bay and off
the Tanjore coast, where this variety furnishes annually about 200,000 shells to the
Calcutta market.

The shortening of the apical whorls often becomes extreme and so abbreviated
as to be subtruncate and to approach in general form the aboral appearance of a
typical Conus. This extreme form appears to arise when the environment is highly
unfavourable, such as life in a muddy sea co-ordinated with scarcity of food ; it re-
presents an extreme form of stunting, analogous to that seen on those pearl banks where
dense over-population reduces the available food supply to a degree insufficient to meet
the normal individual requirements.

(c). Variety elongata is a much elongated spindle-shaped form found in the Andaman
Islands. As var. obtusa is the extreme in variation in the coniform direction so elongata
approaches to the long-drawn-out spiral of Fusus. (PI. II., fig. 4). Shells of this variety
are very rare on the coast of the Indian mainland but I cannot accept it as anything
more than a very well-marked variety of Turbinella pyrum.

The anatomy of the body of Turbinella. pyrum follows closely the lines of such a
typical gastropod as Buccinum (Whelk), and it is unnecessary to go into any details
here, save to mention that the sexes are separate, the proboscis much elongated, and the
columellar muscles extremely powerful. It is well nigh an impossibility to extract
the animal intact from its shell even after death ; the ordinary methods which are read'ly

fi OKHAMANDAL MARINE ZOOLOGY— PART II

successful in the case of Buccinum are ineffective here, and the aid of a sledge hammer
has to be evoked if we require to dissect the animal.

The ova are deposited in a many-chambered chitinous egg-capsule of striking and
peculiar appearance. In general form it is an elongated loosely spiral annulated
cylinder, divided transversely by partitions into a number of compartments ; it reminds
one of a corrugated and loosely twisted ram's horn. When newly formed it is pale
opaque yellow in colour ; with age it darkens and becomes covered with low growths
of algse. It stands upright on the sea-bottom, the lower and first formed end rooted in
the sand by means of a broad flange-shaped basal disc. The lower end is narrow, the
chambers there small ; these gradually increase in size, till at a point about one-third
of the length from the base they attain a maximum size, which is maintained thence
to the abruptly truncate summit.

Considerable complexity is introduced by the fact that the partitions between the
chambers are double and separated by a slight space except along the hinder edge.
The partition which forms the floor of each compartment has a crescentic slit parallel
with the front edge, hence when the capsule sways to the current the partitions
between the chambers gape slightly and so allow a circulation of sea-water within
each compartment, thus providing the aeration needed by the larvae. The transverse
partitions are much thinner than the outer wall of the capsule.

In each chamber a considerable number of fertilised ova are deposited, embedded
in a transparent colourless gelatinous nutritive matrix which entirely fills the chamber.
In this nutritive jelly, the stronger of the embryos develop rapidly ; in each
chamber an average of six reach the larval stage characterised by the possession of
a larval external shell or protoconch ; those that do, grow rapidly and soon pass
beyond the protoconch stage, and assume the semblance of adult form provided with
a brown-flecked shell measuring eventually slightly over half an inch in length,
inclusive of the two and a half whorls of the protoconch which persists both here and into
adult life. By the time this size is attained all the nutritive contents of the incubatory
chamber, including the remains of the weaker of the brethren, have been eaten up
by the ravenous young who now find it necessary to leave home in search of food. Their
first step is to eat through the partitions dividing the chambers, a proceeding that results
in the bringing together of the whole surviving family numbering usually from 200 to
250 in all. The stronger next eat a way through the outer wall and the whole brood
follow to scatter over the adjacent sea-bottom to lead independent lives.

The breeding season, when the capsules are fashioned and rooted in the sand, extends
throughout January, February and the first half of March. Some divers assert that new
capsules are also to be found in June and July, but I have had no opportunity to test
this statement. The sexes being separate the divers have several stories to tell anent
the breeding habits. Among others they assert that the females are the larger and are
attended each by a number of smaller males, who assist in the making of the capsule

HORNELL— THE INDIAN CONCH 7

a manifest impossibility. The post-larval rate of growth is quite unknown. On grounds
where the food supply is abundant, it appears to be very rapid, for beds which are ex-
hausted of adult shells measuring over 2|- inches diameter in one year are found the
succeeding year with plenty of shells, 3 inches and over in diameter.

The chank is an excellent instance of the acquisition by an animal of characters which
appear for all practical purposes absolutely perfect to enable it to hold its own with
ease in its struggle for existence. Against every one of its known enemies it has evolved
suitable means of defence. The massive strength of the shell protects it from the attacks
of all ordinary fishes ; the density and thickness of its periostracum give during youth
and maturity adequate protection against the insidious attack of the boring sponge,
Cliona, and its shell-burrowing congeners ; the strong capsule it constructs for its young
gives them adequate protection till they reach a self-supporting stage endowed even at
this early period with a fairly strong and resistant shell — one cannot crack it between
one's finger and thumb. Its semi-burrowing habits give it great protection against
those fishes which have the habit of snapping off the protruded feet of gastropods. Finally
the pale yellowish-brown periostracum assimilates closely in colouring to the sand and
should be a further protection against its discovery by foes ; to this form of protection
I am, however, not inclined to assign great value, for chank divers can distinguish
the presence of a chank even when half buried in the sand, and if they can, I feel
assured that predatory fish are equally clever.

On rare occasions chank shells are found with the larger whorls crushed in ; this is
believed to be the work of the great goggle-eyed Ray, Rhinoptera adsp&rsa, which has
the most powerful milling teeth of any Indian ray or shark.

As chanks grow old, their resisting powers diminish, the protecting periostracum
receives damage and the burrowing sponge Cliona obtains a lodgment in the shell. Once
there, it runs its branching tunnels everywhere in the substance of the shell, converting
it into a honey-comb mass. I greatly doubt if this contributes except very occasionally
to the death of the chank. This probably occurs usually from senile decay on beds
that are not fished commercially. It is noteworthy that beds which have not been
fished for some years, contain great numbers of Cliona-burrowed shells, whereas on beds
fished regularly the proportion of " wormed " shells is so low as to be practically non-
existent.

II

SOURCES OF INDUSTRIAL SUPPLY

EVERY year many lakhs of chank shells are required by the shell-bangle manufac-
turers of Dacca. To supply this demand hundreds of fishermen and divers, from Kathia-
war in the north to Cape Comorin in the south, search the reefs and scour the sand-
flats within the 10-fathom line during the fine weather season ; in some places they wade
about in the shallows, in others they bring their catch to the surface in nets, or they may
descend themselves to the bottom and hunt it by sight as they swim.

Six distinct chank fisheries are carried on at the present day in India seas ; ranked
in their geographical order they are situated in the following localities : —

(a). Kathiawar ;

(6). Travancore ;

(c). The Gulf of Mannar (usually called the Tuticorin Fishery) ;

(d). Palk Strait (known generally as the Rameswaram and Ceylon Fisheries) ;

(e). The Coromandel Coast, from Point Calimere to Madras City.

Without exception the chank fishery in each of these localities is considered as a
royal prerogative, the monopoly of Government. In practice this prerogative is vari-
ously exercised. In the Gulf of Mannar and on the Indian side of Palk Strait, the
Madras Government work the fishery departmentally through an officer of the Fish-
eries Department styled the Superintendent of Pearl and Chank Fisheries. On the
Coromandel coast the exclusive right to collect is farmed out to a renter for a term of years.
The latter administration of the prerogative is also in force in Okhamandal (Kathiawar),
where His Highness the Gaekwar of Baroda exercises sovereign rights in the local fishery.
In Ceylon the renting system was in force till 1890, when it was abandoned in favour of
an export duty, a method of securing Government revenue from this source which has
continued ever since. In Travancore the dues of Government are collected in the same
manner as now prevails in Ceylon.

The shells fished off the Kathiawar coast are of good quality, well esteemed in the
Bengal trade where they are known as Surti shells — an echo of the day when Surat
was the great emporium of the Kathiawar and Konkan coasts. To-day the shells
are sent to Bombay, whence they are shipped to Calcutta. The quantity yielded is
approximately 200 to 250 bags per annum.

HOR NELL— THE INDIAN CONCH

9

Okhamandal, the north-western extremity of Kathiawar, which forms an outlying
portion of the Gaekwar of Baroda's dominions, furnishes a considerable proportion of
this export. The right to collect the shells is leased out at intervals for a term of years.
Unlike other Indian chank fisheries the shells on this coast are all collected at spring
tides when great areas of the littoral are uncovered at the time of low water. A certain
proportion of the shells are sold to pilgrims who resort to the holy shrines at Bet and
Dwarka, the district of Okhamandal from its association with Krishna forming one of
the holy lands of the Hindus, who delight to take home as a sacred souvenir one of the
shells loved of this god. Full details of this fishery and of the enactments made to safe-
guard it, are to be found in Part I of the present Report.

Were diving for these shells to be introduced on the Kathiawar coast a great increase
in the yield should be possible. As it is, the warm weather when alone diving can be
systematically carried on is unfortunately that when the monsoon is most violent.

The chief centres of supply at the present day are the Gulf of Mannar and Palk
Bay where physical and climatic conditions permit of diving being ultilised to an extent
impossible in any of the other fisheries. Indeed chank-diving in these waters has been
the regular calling of hundreds of the fishing population from time immemorial and
references in Tamil classics make it clear that this fishery was prosecuted with vigour
under Pandyan rule as long ago as the beginning of the Christian era ; in those days the
headquarters of the industry was at Korkai, an important emporium with a population
composed, we are told, in the main of traders, jewellers, pearl-fishers and chank divers.
This city, the Kolkhoi of Greek geographers, was situated at the mouth of the Tambra-
parni which then entered the sea some 12 miles south of the present-day port of Tuticorin.

The relative importance of the various centres is shown clearly in the following
table compiled from the Bengal customs returns for the eight years from 1905-1913.

TABLE SHOWING THE VALUE OF IMPORTS OF CHANK SHELLS INTO CALCUTTA FROM 1905 to

1913.

1905-1906.

1906-1907.

1907-1908. 1908-1909.

1909-1910.

1910-1911.

1911-1912.

1912-1913.

Rs.

Rs.

Rs. Rs.

Rs. Rs. Rs.

Rs.

From Ceylon
From Madras — •

144,772

189,280 86,515 181,223

166,060 87,716 133,495 : 127,940

Chief Port
Other Ports . . .
Travancore

1,583
32,172
114

14,435
21,622

324 2,842
5,265 52,399
592

1,648
66,371
500

1,945
44,504

2,025
65,600
96

3,086
74,153

Bombay*

1

?

1 1

?

?

J

1

Total ... Rs.

178,641

225,337 92,696 : 236,464

234,579

134,165

201,216

205,179

* Not ascertainable as chanks and cowries have been lumped together in the Customs returns

of the Bombay exports,

10 OKHAMANDAL MARINE ZOOLOGY— PART II

The export from Bombay consists solely of shells fished on the Kathiawar coast,
so far as I am able to learn. These " Surti " shells, as they are known in trade circles,
share with the Tuticorin and Rameswaram shells the honour of being esteemed of first
quality. To-day these shells command an average price of not less than Rs. 180 per 1,000
in the wholesale market at Calcutta.

From the foregoing facts it is abundantly clear that, barring the pearl fisheries
of the Gulf of Mannar, the chank fishery is beyond comparison the most important
of Indian shell-fish industries. Year in and year out the chank fishery produces
regularly from 2 to 2| lakhs of rupees per annum (£13,300 to £16,600) whereas the pearl
fishery may prove wholly unproductive for many years together, though when it does yield
a return, it does so with no niggard hand. The pearl fishery owes much to the chank-
fishery ; the latter is the school where boys and men of the fisher population pass their
apprenticeship and qualify for the feverish days of a pearl fishery — a time when a gambler's
lust for gain enters the diver's breast and makes him forget or ignore danger, and despise
sleep and comfort in the race to gather the pearly treasures put at his disposal for an
all-too-brief season, counted by days, and limited by the tempest.

Ill

THE ROLE PLAYED BY THE CHANK IN INDIAN RELIGION AND LIFE

(1). LEGENDARY AND HISTORICAL.

WHEN and how the cult of the chank as a religious symbol originated in India are
questions going back so far beyond any traditions now existing that the utmost diffi-
culty confronts us when we seek to find their solution. One main fact alone seems certain
and that is the non-Aryan origin of this symbol. The Aryan-speaking hordes which
descended upon the Punjab through the N.W. passes perhaps 2,000 years or more B.C.,
certainly did not bring the custom with them. They, the warrior ploughmen and herds-
men of the plains of Eastern Europe and Western Asia, had never seen the sea; they knew
not as yet the deep sonorous boom of the snow-white chank — a note on a curved cattle-
horn was with them the signal between scattered bands, while their hymns tell us that
in music they used the drum, the flute and the lute. Vishnu, the God whose emblems
include the chank, is barely mentioned in the Rig- Veda and the few Vedic hymns to
him were probably composed after long intercourse had been established with the Dravi-
dians, the chief race whom the invaders found in possession of the new land. He is
almost certainly one of the gods borrowed from the indigenous people as his complexion
is characteristically represented as dark-hued whenever his image is shown in colour.

I am strongly of opinion that the key to the problem is to be sought in the custom
prevalent among animists — the worshippers of evil spirits — of employing noise to scare
the demons they fear. At the present day Bengali Hindus make a practice of sounding
the conch which each household keeps for religious rites whenever an eclipse or an earth-
quake occurs. Clearly this is a survival of the use of loud blasts on a shell to scare away
evil spirits — the demon intent on devouring the moon or the sun, or shaking in fury the
foundations of the world. This custom at once conferred religious significance upon the
shell whose noise frightens the evil spirit and restores peace ; thenceforward the shell
would be honoured and held sacred. Probably it was associated with one particular god,
the prototype of Vishnu, and, with his adoption into the Aryan Pantheon, his emblem
and weapon against the powers of evil would accompany him, and become endowed
with still greater and deeper religious significance.

When the hungry swarms of Aryan tribesmen descended upon north-west India,
the whole land, with the exception of the north-east corner, was occupied by a long-

12 OKHAMANDAL MARINE ZOOLOGY— PART II

settled aboriginal population, split into many states and tribes vastly differing in civili-
sation. Many tribes, particularly those living in the mountains and dense forests and
less accessible districts, were in the lowest possible stage, naked savages living on fruits
and small game and maintaining a precarious defence against wild beasts by means
of rude stone weapons and cudgels. In the south, especially in the maritime districts,
a high civilisation developed at a comparatively early date and when the Aryan invaders
were fighting their way into the Punjab against wild and semi-savage tribes, in appearance
and customs probably much like the Santals 50 years ago, the men of the south were
then or shortly later engaged in commercial relations with Babylon and the coastal
districts of the Persian Gulf and Red Sea ; partly through the stimulation received
from this intercourse with these seats of ancient civilisation and partly from indigenous
effort these southern Dravidians were evolving a language unsurpassed for its richness
and flexibility and its power to express with perfect felicity the highest flights of imagina-
tion which poets and philosophers can reach, together with a material civilisation of no
mean order. It is to these coastal Dravidians settled in the prosperous sea -ports situated
on the western shore of the Gulf of Mannar or to men of the same race living on the
Kathiawar coast that the first use of the chank must be traced. Both localities were
the seats of pearl fisheries and the centres whence much oversea traffic flowed coastwise
to Persia, Egypt and the adjacent Semitic lands. The chank and the pearl-oyster are
usually associated in Indian waters, the chank on the sandy patches interspersed with
the rocky patches which form the habitat of the pearl-oyster ; pearl fishers often
bring chanks ashore and thus the beauty of their snowy white porcelain-like massive
shells would early become familiar to the merchants gathered from many lands to
purchase pearls.

The word itself, in its chief forms of chanku or sanJcu in Tamil and sankha in Sanscrit,
appears to be of Aryan origin ; the Indian names are all obviously variants of one root
and this is identifiable with the Greek kongche (KOJ^) and the Latin concha, both
meaning a shell. It is probable that when the Aryans swarmed into India they applied
their generic name for shells to the great white conch so conspicuous an object in the
hands of their enemies, and so it is that, just as Christians in calling their scriptures by
the Greek word for book (Biblos) imply thereby that it is " The Book " pre-eminent
and supreme, so the Aryans appear to have similarly lauded the chank as " The Shell "
and considered no qualifying term necessary. To them it was the shell of shells — the
one shell above all others worthy of honour and even of worship. In the oldest Tamil
literature the word is found in its present-day form but almost as often the shell is
termed valai ; the latter is probably the original Tamil or Dravidian name, a term
which has now come to be displaced by one derived from the Sanscrit sankha.

We may infer that when these ancient poems were written the Brahmans had
already acquired great influence and were engaged in forcing Aryan forms and Sanscrit
terms into the language of the Tamil country by virtue of their religious and literary

HORNELL— THE INDIAN CONCH 13

ascendancy. As the religious leaders of the country, Sanscrit terms in constant use in
their prayers and services would soon become dominant and be absorbed into the language
of the country, in the way that Norman culinary terms readily found acceptance in the
English language because of the superiority of the Normans to the Saxons in this
particular art. Thus it has happened that the Sanscrit term for Turbinella pyrum
gradually but surely ousted the indigenous Tamil term. Finally it is worthy of note that
the English word " conch " is again another variant of the words chanku and sankha.

The earliest notices of the use of the chank are entirely of a secular nature, and occur
in the two great Indian epics, the Ramayana and the Mahabharata. In these we get
frequent reference to the employment of the chank as a martial trumpet by the great
warriors whose more or less mythical exploits are recounted. Particularly is this the
case in the Mahabharata, where in the Bhagavat-Gita we find the heroes endeavouring
to terrify their foes with loud blasts on their battle-conchs, even as their forefathers
scared away the evil- working spirits of the village. Each hero has his famous conch
distinguished by some high-sounding name, just as the famous swords of European
legendary heroes were frequently given names that have become immortal in song and
story. The beautiful Excalibur wielded by Arthur in many glorious fights, Charlemagne's
famous Joyeuse, and the magic Tyrfing so oft the theme of Viking sagas, have their
parallels in the names of the conchs of the Mahabharata heroes.

When the opposing hosts of Kauravas and Pandavas confronted each other on the
field of Kurukshetra, we read in the Bhagavat-Gita (verses 11 to 19) how the prelude
to battle was the deafening clamour sounded by the leaders on their great conchs.

" The Ancient of the Kurus, the Grandsire (Bhisma), the glorious, sounded on high
his conch, ' The Lion's Roar.'

" Then conchs and kettledrums, tabors and drums and cowhorns suddenly blared
forth with tumultuous clamour.

" Stationed in their great war-chariot yoked to white horses, Madhava (Krishna)
and the son of Pandu (Arjuna) blew their divine conchs.

" Panchajanya was blown by Hrishikisha (Krishna) and Devadatta by Dhananjaya
(Arjuna). Vrikodara (Bhima) of terrible deeds blew his mighty conch, Paundra.

" The king Yudhishthira, the son of Kunti, blew Anantavijaya ; Nakula and Saha-
deva blew their conchs Sughosha and Manipushpaka.

" And Kashya of the great bow, and Shikhandi, the mighty carwarrior, Dristadyumna
and Virata and Satykai, the unconquered.

" Drupada and the Draupadeyas, 0 Lord of Earth, and Saubhadra, the mighty-
armed, on all sides their several conchs blew.

" That tumultuous uproar rent the hearts of the sons of Dhritarashtra, filling the
earth and sky with sound."

Here the names of the conchs possessed by all the five Pandava brothers are given,

14 OKHAMANDAL MARINE ZOOLOGY— PART II

— Paundra, Devadatta, Anantavijaya, Sughosha and Manipushpaka. Paundra appears
to have been named from Pundra, a demon killed by Bhima ; the others signify re-
spectively God-given, Eternal Victory, Sweet Voice or Honey-tone, and Jewel-blossom.
Here too we hear for the first time the name Panchajanya given to the conch of
Krishna, King of the Yadavas, who had espoused the Pandavan cause. Around this
famous shell many legends have gathered and now we see it held on high in most figures
of Vishnu, who is considered by Hindus to have been re-incarnated in Krishna, the wise
and good king of the Yadavas. According to one legend Panchajanya was originally
the shell home of a terrible marine demon, Panchajana, so named as he was a foe to
the five kinds of beings (jana), to wit, gods, men, gandharvas, serpents and ghosts or
non-incorporated spirits. Panchajana lived on the sea bottom and at last filled the
measure of his misdeeds by seizing the son of Sandipani, who had taught Krishna the
use of arms. The God, fearless of consequences, rushed to the help of the child,
assuming the form of a fish ; after a terrible struggle he vanquished the demon and
brought away his shell as a trophy, since accounted one of the emblems of Vishnu and
Krishna.

Tod, the author of the famous Annals of Rajasthan, in his " Travels in Western
India," published 1839, in describing his visit to Dwarka and its neighbourhood gives
a variant of this story and as the passage is most interesting no apology is needed for its
reproduction in full. Under date January 1, 1823, he writes, " Crossed over to the
Pirates' isle, emphatically called Bate, or ' the island/ but in the classic traditions of
the Hindu, Sankhodwara, or ' the door of the shell/ one of the most sacred spots of his
faith. It was here that Crishna 1 or Kanya acted the part of the Pythian Apollo, and
redeemed the sacred books, slaying his hydra foe, Takshac, who had purloined and con-
cealed them in one of those gigantic shells whence the island has its name. The whole
history of Kanya, or Crishna, who assumed the form of Vishnu, is allegorical, but neither
devoid of interest nor incapable of solution. There is no part of their mythology more
easy of illustration than this, which is allusive to the sectarian warfare carried on at
this period between the new sect of Vishnu vites and the more ancient one of Budha.
The races who supported the religion of Crishna are typified under his emblem, Garuda,
or the eagle ; while their wily adversary, the Budhist, is figured by the Tacshac,
Naga, or serpent, a denomination given to the races of northern origin, which at
various periods overran India, and of which were Taksiles (the friend of Alexander,
the site of whose capital is still preserved in the Memoirs of Baber) and the still more
famed Tacshac Salivahan, the foe of Vikrama. In the legend of the Yadu (Yadava)
prince, Crishna (himself a seceder from the faith of Budha Trivicrama to that of
Vishnu, if not its founder), receiving the sacred volumes from his hydra foe at this

1 Tod notes " Kanya, or Vishnu, resembles the sun-god of the Egyptians in name as well as
symbols. Kan was one name of the sun in Egypt, and his eagle-head is a well-known type." With
regard to the extract given in the text, it has to be remembered that Tod's mythological explanations
are not always reliable.

HORNELL— THE INDIAN CONCH 15

remote point of Hinduism, as well as his first combat with him in the Jumna, we
have but the continuance of the same sectarian warfare, in which Crishna was in this
instance successful, driving them before him both in the north of India and here : thus,
his title of Rinchar was given on his defeat by Jarasindha, the King of Magadha, of
the heretical faith, and at length these religious and civil conflicts led to his death,
and the dispersion of the Yadu race of which he was the chief support. These Yadus,
I surmise to have been all originally Budhists, and of Indo-Getic origin, as their
habits of polyandrism alone would almost demonstrate ; and when we find the best-
informed of the Jains assuring us that Nemnat'h, the twenty-second Budha, was not
only Yadu, but the near kinsman of Crishna, all doubt is at an end ; and I am
strongly inclined to pronounce decidedly, what I have before only suggested, that the
Yadus are the Yute, or ancient Getes of the Jaxartes, amongst whom, according to
Professor Newmann from Chinese authorities, one of the Shamanean sages sprung,
eight hundred years before Christ. Both Nemnat'h and Sham-nat'h have the same
personal epithets, derived from their dark complexions, the first being familiarly called
Arishta-Nemi, ' the black Nemi/ the other Sham and Crishna, both also meaning
' dark-coloured ' ; and when this is not only confirmed by tradition, but the shrine of
Budha itself is yet preserved within that of Crishna at Dwarica we have no reason to
question that his faith, prior to his own deification, was that of Budha."

Always is Krishna's chank represented as a sinistral abnormality, and legend pic-
tures to the mind of the devout Hindu every shell of this rare form when alive, as a
marvellous production receiving the homage of thousands of chanks of ordinary form,
which crowd around it on all sides. Another myth is related by Baldseus, the chaplain
to the Dutch forces which wrested Ceylon from the Portuguese, to the effect that Garuda,
the eagle, vehicle or attendant (almost certainly the hawk-headed deity of Assyria)
of Vishnu flew in all haste to Brahma and brought to Krishna " the chianko or kink-
horn twisted to the right." 1 Vishnu derives several of his alternative names from his
chank symbol, as Chankapani, the " chank-armed," and Chankamenthi, the " chank-
bearer."

Krishna, when represented as a herdsman under the form of Govinda or Gopala,
usually bears a conch in his hand and possibly the origin of this may be sought in the
use a herdsman may make of it to call together his scattered charge just as the shepherds
of Corsica and Sardinia at the present day use a great Triton shell (Tritonium nodu-
liferum) for a like purpose.

A curious and most significant fact is that the twenty -second Tirthankar of the Jains,
Nemi or Nemnath, who, legend says, was the son of King Samudravijaya of the race

1 With the contradiction which exists between East and West in so many matters, the abnormal
twist in these shells is termed the left-handed or sinistral by Europeans, whereas Indians term it right-
handed. They view it from the mouth end, we from the apex, and accordingly confusion is frequent
in conversation on this subject with Indians.

16 OKHAMANDAL MARINE ZOOLOGY— PART II

of Harivansa and a cousin to Krishna, has a conch as his emblem and is represented
in Jain statues as being of a black colour. The black image of Nemi in the Nemnath
temple on Mount Girnar in Kathiawar is a well-known example. The dark hue under
which Vishnu and Krishna are always represented by Hindus and the black colour
of his cousin Nemi, the Jain Tirthankar, go far to show that these gods and teachers
belonged not to the Aryan race but to nations of Dravidian origin in the forefront of the
earliest indigenous civilization in North India or Hindustan. With them in particular,
is the conch most definitely associated ; there is strong presumption on this and other
grounds already referred to, to believe that it was the Dravidians who first employed
the chank as a battle-conch and that this custom was adopted by the Aryan invaders
as blood connections began to be formed in increasing numbers with the Dravidian
nobility of the land and when certain of the Dravidian gods were admitted to the Aryan
pantheon. The Aryans would be particularly eager to acquire fine conchs both for use
and ornament ; their deep-voiced boom would prove their utility as battle-trumpets to
enspirit and to give signals, while their rare white beauty would appeal to the religious
sense as making them fit vessels wherewith to offer libations to their gods. To an inland
people the beautiful products of the sea assume a double value from their strangeness
and rarity and mysterious origin. To-day the people of Tibet, cut off from all know-
ledge of the sea, esteem pearls and red coral, tortoise-shell and amber, among the greatest
treasures within their knowledge. The wild Nagas of the Assam hills equally prized the
snow-white chank shell itself till some 50 years ago, using it as part of their accepted
currency at the rates enumerated on page 43. As the Aryan hosts advanced into
India they must have captured numbers of battle conchs from time to time and there can
be no doubt they early adopted them in place of their own less sonorous cow-horns.
Indeed the boom of the conch has been the battle signal throughout the ages in
India, and this custom has lasted almost to the present day. Ancient Tamil and
Rajput poems descriptive of battles and raids continually refer to the clamour of the
conchs blown as the opposing parties approached each other ; the etiquette of old Indian
chivalry required a prelude of challenging conch-blowing before the serious fight was
begun ; the long-drawn hollow sonorous note of the chank often greeted early British
commanders as they led their forces to the assault ; until the beginning of last century
Marathi and Pindari chiefs called their followers together and heartened them for the
fray by loud blasts on conch-horns. Even in very recent days the chank 's voice has called
our enemies to the attack, and this too by other foes than Hindus. The graphic pen of
Percival Landon in his " Lhasa "—an account of the British Mission to Thibet in 1903-4,
in describing a night cannonade of the British Commissioner's post outside of Gyantse
by the Thibetans, paints a word picture worthy of quotation:— " As one peers out into
the warm night a long monotone is faintly droned out from the darkness ahead.
It is one of the huge conch shells in the Jong and it may only mean a call to prayer — the
' hours ' of Lamaism are unending — but as the moaning note persists softly and steadily

HORNELL— THE INDIAN CONCH 17

a vivid speck of flame stabs the darkness across the river. A second later the report
of the gun accompanies a prolonged ' the-e-es' overhead."

From the earliest times the conch has also been used in India to call the people to
their sacrifices and other religious rites, as an instrument of invocation to call the attention
of the gods to the ceremonies to be performed, and as a means to keep evil spirits at a
distance and to prevent their entry into houses and temples.

With this intimate association with the chief religious rites, the people gradually
came to reverence the instrument itself, and to adore and invoke it (see p. 22 for details),
as is also done with regard to many other instruments or articles of sacrifice in Hindu
rites at the present day ; these latter do not, however, appeal in equal measure to the
religious feelings of the multitude, for around them have not been woven the myths and
legends pertaining to the chank.

In the ceremonies attending the coronation of great kings the chank naturally played
an important part. At the time when the Mahabharata was put into its present form,
this custom was fully developed to judge from the description of the coronation of
King Yudhistira given in the Shantiparwa of that epic. To quote from an interesting
summary by Rao Sahib P. B. Joshi1 " Kings of different countries, learned Brahmans and
sages were invited for the ceremony. A Vedi or sacrificial altar was prepared. There
were brought Samidhas or pieces of sacred wood, five kinds of sacred leaves, waters of the
holy rivers and of the four seas, seven kinds of holy earth, the sacred conch shell, a
white umbrella, and white Chamars. The horses and elephants used in connection with
the coronation ceremony were also white. Yudhistir was then made to sit on a throne
of gold, and other members of the royal family sat on seats made of ivory, and close by
sat the king's spiritual guide and other sages. The king was now requested to touch
such auspicious articles as corn, white flowers, swastika, gold, silver and jewels. The
king's ministers and other high functionaries were now brought betore him by the royal
priest and they all paid their homage to their sovereign. The sacred fire was then
kindled, the king and queen sat in front of the holy fire on seats covered over with tiger's
skin, and made offerings to the fire. Shri Krishna then got up, took in his hand the
sacred conch-shell, which was filled with holy water, sprinkled the water over the heads
of the king and queen and said : ' I sprinkle this holy water over you to indicate that
from this day you have become the paramount sovereign of Bharat Varsha.' At this
time dhundubi and other musical instruments were sounded, drums were beaten, the
king's bards sang the praise of the king and wished him victory and long life."

In Bengal every marriage conducted according to Hindu ceremonial includes the
placing of chank-bangles, lacquered red, upon the bride's wrists. An iron bangle placed
on the left wrist is also essential to the ceremony. Elsewhere this paiticular marriage
custom is scarcely ever practised, but sufficient instances are recorded among a few
widely scattered castes and caste-sections of the existence of a similar marriage custom

1 The Times of India Illustrated Weekly, 20th September, 1911.

B

18 OKHAMANDAL MARINE ZOOLOGY— PART IT

elsewhere to tempt us to believe that at one time it was the general custom of all fully
Hinduised castes throughout India. Finds of fragments of chank-bangles in places
where the use of these ornaments is non-existent at the present day strengthen the theory.
Legendary lore can also be quoted in support. For instance, among the Balijas of
Telugu districts, who there constitute the chief trading caste, a legend is current (Thurston,
I, p. 137) that " on one occasion Siva wanted his consort Parvati to appear before him
in all her glory. But, when she stood before him, fully decorated, he laughed and said
that she was not as charming as she might be. On this, she prayed that Siva would
help her to become so. From his braid of hair Siva created a being who descended on
the earth, bearing a number of bangles and turmeric paste, with which Parvati adorned
herself. Siva, being greatly pleased with her appearance, told her to look at herself
in a looking-glass. The being who brought the bangles is believed to have been the
ancestor of the (lazula Balijas."

The latter sub-division of the Balijas peddle glass bangles only at the present day,
but it is reasonable to suppose that before the discovery of glass, their stock in trade
consisted instead of chank bangles. It is indeed probable that the introduction of glass
dealt a heavy blow to the employment of the chank shell in feminine adornment in certain
districts, particularly, for instance, in those where, as in Vizagapatam, glass factories
being established, glass bangles were put on sale at a fraction of the cost of the compara-
tively expensive chank ones, which require the expenditure of much time and labour
to render them attractive.

Another legend, prevalent among the Sangukatti Idaiyans, the great pastoral or
shepherd caste of Tamil India, narrates that when Krishna desired to marry Rukmani,
her family insisted on marrying her to Sishupalen. When the wedding was about to
take place, Krishna carried off Rukmani and placed a bangle made of chank-shell on her
wrist (Thurston, II, p. 354). These particular Idaiyans belong to one of the sections
of this caste which to-day require their married women to wear these bangles — now a
very rare custom in South India.

Indian sources give the barest indications of the traffic in chank shells that must
have been brisk for 3,000 years or more between the fisheries in the Gulf of Mannar
and on the Kathiawar coast and the inland nations of the Deccan and Hindustan.

(2) PRESENT DAY USES.

(a) IN RELIGIOUS CEREMONIAL (INCLUDING MARRIAGE AND DEATH RITES), AND VULGAR

SUPERSTITION.

We have already seen that the chank is one of the two most important symbols—
the other being the wheel or chakram — associated by Hindus with Vishnu, the second
person in the Brahmanic trinity or Trimurthi. The majority of the avatars or incar-

HORNELL— THE INDIAN CONCH 19

nations of Vishnu are also occasionally represented as holding a chank in the hand :
Matsya,in the form of a fish, Kurma the tortoise, Varaha the boar, and Narsingha the man-
lion, are avatars sometimes sculptured holding Vishnu's chank : more frequently Krishna
is thus depicted. Narayana, the god dwelling in the sun, another form of Vishnu,
is similarly represented in human form with a chank in one hand and a discus (chaki-am)
in the other. On rare occasions Siva is also depicted as holding a chank in one hand.
In all these instances the chank represented is of the sinistral or left-handed form, a
rarity so choice and valuable as to be worthy to form an adornment of a god. No
more fitting gift to a deity can be imagined ; as the symbol of the god who divides with
Siva the worship of the Hindu world, as a production of nature so scarce as to appear
once only in several millions of normally shaped shells and as an emblem of purity,
could Hindu find more fitting offering at the shrine of his god ? Thus it is that the pious
wealthy have from time to time dedicated these shells to favourite temples — particularly
to those that are in high esteem at centres of Hindu pilgrimage.

Chanks are held in special veneration at those shrines where the cult of Krishna is
the predominant feature of worship, and in those localities rendered sacred to Hindus
because of their reputed association with this deity while sojourning on earth.

Chief among the former is the great religious centre of the Madhva sect at Udipi
in South Canara. There around a temple dedicated to Krishna are disposed eight mutts,
or religious colleges, whereof the priesthood is a peregrinating fraternity, who tour the
country to confirm in their faith the scattered members of their community — a procedure
which consists in branding with hot irons the deity's symbols of the chank and the wheel
upon the arms of the faithful.

Usually the priests of seven mutts are on tour ; the head of one mutt takes charge
in rotation of the Udipi temple and for a period of two years carries on the services
and provides the funds necessary for the lavish charity dispensed daily by the temple.

The members of this sect have a great reputation for active piety and the priesthood
of each mutt vie with those of the others in collecting precious vessels for use in the service
of the deity. The sinistral chank is the principal of these, and so it comes about that
in the eight mutts at Udipi is gathered an unique collection of these rarities. During
a recent visit to this place the Swamiyar or high-priest was good enough to show me the
four which belong to his mutt — the Puttige mutt. Of these, two were handsomely
mounted in richly chased gold, the others in silver. All were rather small in size and one
appeared to be of the sub-fossil description frequently found buried in muddy deposits
in shallow lagoons in the north of Ceylon. A census of the sinistral shells possessed by
the whole eight mutts was not possible, but from the information received from the high
priest there cannot be less than thirty of these shells in the town, allowing the low average
of three to each mutt and six for those possessed by wealthy Hindus who are not of the
priesthood.

After Udipi, probably the largest number of sinistral chanks are to be found in the

B 2

20 OKHAMANDAL MARINE ZOOLOGY— PART II

temples of the sacred land of Kathiawar, rich in its association with the later life of
Krishna ; while in Bet in 1906, I found richly ornamented sinistral chanks in the Shank
Narayan, Lakhsmi and Satya Bhamaji temples. (Plate I.) That in the last named is
a particularly lai-ge and handsome shell, probably the finest sinistral chank in existence
and consequently an almost priceless treasure. The shell possessed by the Shank
Narayan temple is a small elongated specimen offered at the shrine some twenty years
ago by a Bathia from Zanzibar ; that of the Laksmi temple is a short broad one of small
size with handsome arabesque ornamentation on the mounting — it has been in the
possession of the temple since Samvat 1890 (A.D. 1835). At Benares, temple treasures
include similar examples, while in the south of India, where opportunities to obtain
these shells are great, a considerable number of the temples of that devout land possess
one or more. Those at Rameswaram, Trichendur and Madura may be instanced.

The temple at Rameswaram, one of the four most holy places of pilgrimage among
Hindus, possesses four shells ; of these three are very ancient and date from the days
when the munificence of the Madura and Ramnad rulers endowed this temple with a
share of the produce of the local pearl and chank fisheries. The shells in the Trichendur
Temple, a towering pile overlooking a sea where chank fishing still flourishes, are
derived likewise from a privilege share possessed formerly by the temple in this fishery.

It is noteworthy that although Rama, an incarnation of Vishnu, is especially
associated with the Rameswaram temple, the actual deity represented in the inmost
shrine is the phallic lingam of Siva. On important religious festivals water is poured
into the sinistral chanks, and after Vedic chants have been sung and prayers offered,
the water thus consecrated is poured over the lingam.

It is remarkable also and indicative that this custom has not originated with modern
Hinduism, that sinistral chanks are objects of adoration among northern Buddhists.
Sarat Chandra Das, the intrepid survey officer who spent some perilous years in Thibet,
mentions (Journey to Lhasa, London, 1902) that in the Sakya Monastery lying to the
south-west of Shigatze there is preserved in the temple a chank of this rare form. Its
history is invested with more than ordinary interest, for the monastic records state
that it was a present from Kublai Khan, the great Tartar conqueror of China and patron
of the Polos, to Phagpa, a hierarch of Sakya whom Kublai made ruler of Thibet in the
second half of the thirteenth century. Sarat Chandra Das mentions that this famous
shell is blown by the lamas only when the request is accompanied by a present of seven
ounces of silver, but to have it blown " is held to be an act of great merit."

In Thibet these left-handed chanks are called Ya chvil dung-kar and in Chinese
Yu Jisuan pai-lei. The people of both countries consider such shells as treasures of
inestimable value. In 1867, one was known to be kept at Fuchu by the Ti-tuh (Pekiny
(rdzette, February 23rd, 1867) and one at Lhasa (fide an editorial note in Das, " Journey
to Lhasa," above quoted).

At one time the value of these shells is said to have been assessed at their weight in

HORNELL— THE INDIAN COXCH 21

gold and this statement is probably correct. To-day they are less valuable and small
and imperfect ones occasionally change hands in the north of Ceylon at Rs. 60 to Rs.
90 each (say £4 to £6) ; such shells are usually sub-fossil ones found buried in deep mud.
Large good-conditioned sinistral shells obtained alive, and therefore of good colour, still
command high rates — anything from Rs. 400 to Rs. 5,000 or more — so infrequently
are they found.

The method of estimating the approximate value of these latter shells is as follows —
if the shell be of 100 rupees' weight or over (one standard coin weighs exactly one tola ;
2| tolas = one ounce) the value is calculated at the rate of Rs. 50 per tola or rupee's
weight ; thus a shell of 110 rupees' weight would be estimated accordingly to this method
at Rs. 5,500. When the weight is between 30 and 100 tolas then the rate per tola may
be taken at Rs. 30/- to Rs. 40/- according to size and quality ; a 50 tola weight shell
would be worth from Rs. 1,500 to Rs. 2,000. If of 25 tolas' weight the rate may vary from
Rs. 5/- to Rs. 20/- per tola.

Wealthy Hindu Rajahs and Zamindars also aspire to possess these sacrosanct shells,
the orthodox in order that they may use them to perform abeyshekam ' in their family
shrine and others for the ostentatious pride of possession and because of the superstition
generally current which accounts the ownership of a Valampuri chank as conferring
prosperity upon the household owning it.

I know also a wealthy Muhammadan Marakkayar who has refused offers of several
hundreds of rupees for one of these shells ; to him the shell is a mascot, the bringer of
good luck.

The result of my inquiries indicates that there exist at present in India not less
than 120 sinistral chank shells. I know the actual whereabouts of 60 and I am well
satisfied that it is improbable that I have overestimated the remainder at an equal
number.

The chank shell is so massive and strong as to be practically indestructible under
ordinary circumstances. I have found chank artifacts in the ruins of Korkai, the Kolkhoi
of the old Greek and Egyptian geographers, which though over 1,000 years old are in-
distinguishable except in colour from fragments cut in Dacca workshops to-day. Hence
the sinistral shells now found in Hindu temples should represent the aggregate catch
of this abnormality for several hundreds of years. If we place the period at the conser-
vative limit of 300 years we shall find the percentage borne by sinistral to normal
dextral shells an extremely small one. Elsewhere (Madras Fisheries Bulletin, No. 7)
I have shown that the average catch of chank shells off the Indian and Ceylon coasts
aggregates 2,500,000 per annum. There is reason to believe that a hundred years ago
the average was considerably higher but some allowance must be made for interruptions
in the fishery due to political unrest in the troublous times prior to the coming of the
Pax Britannica. 2| millions of shells are therefore probably a fairly accurate approxi-

1 The worship of the gods with libations.

22 OKHAMANDAL MARINE ZOOLOGY— PART II

mation. At this rate the produce of the past 300 years would amount to a total of
750,000,000 shells ; the 120 sinistral shells believed to exist now thus give the infinitesimal
ratio of one shell to 65 millions of normal ones, and this proportion undoubtedly gives a
substantially accurate idea of the great rarity of the abnormal form.

Among the ignorant who usually have never even seen a Valampuri chank, the belief
is prevalent in Tamil South India that it blows of its own accord during the night ;
even the Roman Catholic chank divers of Tuticorin entertain this quaint superstition,
and say that the shell is specially clamorous on Tuesday and Friday nights ! A yogi
when controlling or retaining his breath is also credited with hearing the sound of a
Valampuri chank blowing within his abdomen though why the rumbling should be that
of a Valampuri and not of an ordinary chank is hard to answer.

Worship of the chank as one of the three essential articles used in domestic worship
among Brahmans should occupy an important part in the daily liturgy of this priestly
caste and in the Brahma Karma, the work which sets forth in minute detail the order and
phrasing of the sacred rites of the Brahmans, the prayer to the sacred chank may be
translated as follows :—

Taking the chank in his hand the Brahman recites : " At the mouth of this shell
is the God of the Moon, on its sides is Varuna, on its back Prajapati, and on its apex
the Ganges, the Sarasvati, and all the other sacred rivers of the three worlds in which
they make ablutions according to the command of Vasudeva.1 In this chank is the
chief of the Brahmans (Brahmendra or Brahmanaspati). This is why we must worship
the sacred chank. Glory to thee, sacred shell, blessed by all the gods, born in the sea,
and formerly held by Vishnu in his hand. We adore the sacred chank and meditate
upon it. May we be filled with joy !

' I offer (to the chank) everything needful for worship — perfumes, rice and flowers."

Here they make the sign of the chank, but Bourquin (" Annales du Musee
Guimet," Vol. VII, p. 45), from whom I quote, says he was never able to discover the
manner of making this sign and I have had a like ill-success.

While the above is a portion of the liturgy which the head of each Brahman family
is theoretically bound to recite daily, under present-day conditions this is impossible,
and in fact it is only on specially important puja or holy days that even a greatly curtailed
version of this and the accompanying prayers is recited by the head of the family, and
this, too, only in thoroughly orthodox families. The curtailed prayer usually runs. " Oh,
chank shell, thou wast produced in the sea and art held by Vishnu in his hand. Thou
art worshipped by all the gods. Receive my homage."

In this connection an interesting chank legend centres round the temple tank in
Tirukalikundram, a holy Saivite village in the Chingleput District, some 37 miles south-
ward of Madras. The village munsiff, Mr. T. A. Vedachala Gurukkal. to whom I am in-
debted for the following particulars, states that once every twelve years a chank rises

1 One of the names of Krishna.

HORNELL— THE INDIAN CONCH 23

to the surface of the large sacred bathing tank called Sanku Theertham. Three days
before this marvellous occurrence, the water in the tank is greatly agitated, foaming and
boiling to the accompaniment of considerable noise. On the appearance of the chank
the foam subsides and the sound ceases. Then the chank comes to the side where it is
picked up, purified and holy water sprinkled upon it. Thus consecrated it is carried
in solemn procession through the village to the Vethagiriswarar temple, where it is placed
in the treasury with the rest of the temple treasures.

Locally this chank is considered superior to all other chanks in the world and the
origin of the periodic miracle is narrated as follows :—

When the sage Markkandeya was visiting the sacred places with his disciple, th3
latter forgot to bring the Siva puja-box containing the chank required for the libation
necessary in the ritual of worship. The sage took his bath in the tank and as puja time
was fast approaching he prayed to be helped in this misfortune. As a result of his prayer
and by the special grace of the deity, a chank appeared in the tank. Then he placed a
lingam before him and, with the help of the chank miraculously provided, performed his
worship in an acceptable and proper manner. He also prayed that a chank might appear
each day he should worship at this tank. The Puranas say that this request was granted
and has been since continued until the present day. In this connection my informant
remarks that it is to be noted that twelve ordinary years constitute one divine day.

Thousands of pilgrims resort to the Tirukalikundram shrines each year and the
marvel of this story is one of the great assets of the place just as is the periodic lique-
faction of blood to the shrine of San Gennario in Naples.

In temple worship, the chank fulfils important service. The ordinary and sinistral
forms are both employed whenever the temple possesses them. The former is used in
the menial duty of summoning the god's attention, announcing the commencement of
the principal rites as well as in calling the devout to worship ; such are among the general
explanations given for its employment, but some ethnologists hold that the innate and
primitive significance of the use of the blowing chank in temple worship is to scare away
hostile and evil-working spirits. This is a reasonable belief as there is little or no doubt
that the chank was used originally as a horn or trumpet by tribes holding animistic
beliefs prior to the development of the Brahman religion which appears to have adopted
the use of the chank in religious ceremonies together with many other rites from the
devil-fearing tribes who gradually came into the fold of the new and higher religious
belief.

In this connection should be mentioned the custom which prevails largely in Bengal of
keeping blowing chanks in the houses of the better class people for use in family worship.
Mr. A. R. Banerji, I.C.S., the late Dewan of Cochin State, informs me that it is a
general custom in Bengal to turn ovit with these shells during eclipses and earthquakes
and to keep up a continuous blowing till the eclipse or earthquake be over.

A rather striking effect is produced when the chank is used in temple ritual as a

24 OKHAMANDAL MARINE ZOOLOGY— PART II

sort of rhythmical accompaniment when it plays the part of kannagolu or
talavinyasa.1

In Hindu temples the four daily services take place before dawn, at noon, at sunset,
and at 9 or 10 P.M. At these times the Ochchans, the caste entrusted solely in the Tamil
country with the duty of chank-blowing in temples, announce the commencement of
each service and punctuate the various rites with the drone of their shell according to
the customary ritual. Among the Uriyas the corresponding caste of temple servants
is that of the Ravulos, whose caste duty is to sound the chank during services in Saivite
temples when the god is being taken in procession as also to prepare garlands of flowers
for the adornment of the god. Like the Ochchans, they are not usually whole-time
servants of the temple, but while the former earn money as musicians at weddings,
performing upon a long silver trumpet, the Ravulos make and sell garlands to the laity.
The Ochchans never use the chank except in the temples, whereas the Ravulos are em-
ployed to sound it at Brahrnans' weddings. It is a rule among the latter that they
must possess at least two blowing chanks, lest, losing one, the temple service should
suffer in consequence.

In Bengal, the Ramavat sect of Vaishnavas pay particular attention to the call
of the chank. By them all forms of worship except the unceasing repetition of the name
Rama or Hari are deemed useless, but in every akhara or monastery of the sect an idol
is tended at regular hours to the sound of chank shells and gongs, while offerings of
fruit and flowers are presented by the laity (Risley, II, p. 340).

Chanks to be used as wind instruments are chosen of as large size as procurable,
often 8 inches long by 4 inches in diameter. The only preparation they require is to
have the extreme apex removed, usually by hammering. No tune properly so called
can be played, but the tone is capable of much modulation by the lips, and the long drawn
notes as they drone clear and mellow on the evening breeze have a haunting charm that
clings sweet and seductive in the memory ; it has a mystic wail perfect in appropriateness
to its religious use before the shrines of the gods of a profoundly philosophical creed.

Sinistral shells whenever possessed by a temple, are usually mounted in handsomely
decorated golden settings and used as libation vessels in the service of the god. Whether
the god be Siva in the form of a lingam, or Vishnu or other deity represented in
anthropomorphic shape, the officiating priests must lave it with water rendered sacred
by being poured from the mouth of a chank.2 On certain auspicious days cow's milk
is used for libations in lieu of water. And if the doubly sacred sinistral chank is not
possessed by the temple, then a choice example of the ordinary form must be used.

In family devotions the chank is also employed as a libation vessel by strictly
devout Brahmans, both Saivites and Vaishnavites. Daily before the mid-day meal

1 Day, "Music and Musical Instruments of S. India and the Deccan," 1891.

- A Tamil proverb says : " If you pour water into a chank it becomes holy water ; if you pour it
into a pot, it remains merely water."

HORNELL— THE INDIAN CONCH 25

the Brahman head of the family, after taking his bath, prostrates himself before the family
shrine and then, chanting some hymns from the Vedas, pours water over the image
of the deity from the mouth of a chank shell. Then he dresses the god and commences
his prayers.

In Thibet the call of the chank is amongst the most familiar sounds to be heard in
the monasteries and temples of the Lamaistic faith. The writings of travellers in that
most priest-ridden of countries, contain frequent mention of the custom. Sven Hedin
for example, when describing the opening ceremonies of the Losar or new year festival
which he saw in the great monastery of Tashilunpo in Shigatze — the seat of the Tashi
Lama, says : — " Suddenly from the uppermost platforms on the roof ring out deep, long
drawn-out blasts of horns over the country ; a couple of monks show themselves against
the sky ; they blow on singular sea-shells, producing a penetrating sound, which is
echoed back in shrill and yet heavy tones from the fissured rocks behind the convent :
they summon the Gelugpa, the brotherhood of yellow monks, to the festival."

Tea-drinking among the Lamas must never be missed ; the monks partake of it even
in the midst of the most important ceremonies, and to prevent the terrible misfortune
of a brother being too late for any distribution of tea, the departure of the novices from
the kitchen bearing their loads of hot tea in large copper vessels on their shoulders,
is signalled to all in the various halls and cells by a loud call upon a chank-horn from
the temple roof.

Sven Hedin also describes (" Trans-Himalaya," Vol. II, p. 19) a cave inhabited by
a hermit reputed to be one hundred years old, who passed his days crouching in a niche
in the wall continually saying his prayers and occasionally blowing a faint blast on a
chank.

And when a monk, no longer able to answer the shell's call to gather with his brethren
round the teapots and the bowls of tsamba, passes quietly away, the same sound summons
those who remain to attend his funeral mass.

In the purer Buddhism of Ceylon the chank cult also finds place, and figures pro-
minently among the imisical instruments employed to lend eclat to the periodic procession
(perahera) of the tooth relic at Kandy.

(b) — BRANDING.

All Sri Vaishnavite Brahmans, irrespective of profession, are expected to undergo
a ceremony of initiation into Vaishnavism after the Upanayanam ceremony of invest-
ment with the sacred thread, in the belief that it is the duty of all of their creed to carry
throughout life a memorial of their god upon their person. To effect this, resort is
had to branding with heated copper seals made in the conventional form of the various
symbols of Vishnu.

Members of this sect are not compelled to undergo this ordeal more than once

26 OKHAMANDAL MARINE ZOOLOGY— PART II

during their lifetime, but the Madhva sect which comprises chiefly Canarese-speaking
Brahmans, have to submit to it as often as they visit their Guru. Men of other castes
who become followers of a Vaishnava or a Madhva Achilrya (Guru) are expected to present
themselves before the Guru for the purpose of being branded. But the ceremony is optional
and not compulsory as in the case of a Brahman. Even the women in Vaishnavite
families must submit to this branding ; in their case it takes place after marriage in
the case of Sri Vaishnavites, while among the Madhvas one form of branding should
be performed at any age before marriage should the Guru visit the neighbourhood, and
a more formal one again after marriage. Regarding Sri Vaishnavites, Thurston (I,
370) states that " the disciples after a purificatory bath and worship of their gods,
proceed to the residence of the Achfirya, or to the mutt where they are initiated into
their religion, and branded with the chakra on the right shoulder and with the chank on
the left. The initiation consists in imparting to the disciple, in a very low tone, the
Mula Mantram, the word Namonara-yanaya, the sacred syllable Om, and a few man-
trams from the Brahma Kahasyam (Secrets about God). A person who has not been
initiated thus is regarded as unfit to take part in the ceremonies which have to be per-
formed by Brahmans. Even close relations, if orthodox, will refuse to take food prepared
or touched by the uninitiated."

As Vaishnavite Gurus are few in number, it is necessary for them to peregrinate
the country, halting at suitable centres to brand those of their followers living in the
neighbourhood just as a Bishop in certain Christian churches tours his diocese to afford
confirmation (i.e., initiation) services at periodical intervals. In populous districts
where Vaishnavites are in numerical strength the scene at each of the Guru's halting
places is intense with interest. Thousands of his disciples gather round eager to be
branded. Brahmans are there in force, but men of many other castes and even Parai-
yans are there. The ceremonies begin by the making of a fire in a mud pot (homa
kunda), accompanied by the chant of hymns and the offering of prayer to Vishnu. As
Brahmans present themselves for the rite the Guru lifts the copper brands which have
been heating meanwhile in the fire and applies them to the shoulders of the people,
the chakra on the right and the chank on the left. As each stamp is made the Guru's
assistant, usually a Dasari or Vaishnavite mendicant, smears the burnt spots with a
paste of namakJcatti, the same white clay used by Vaishnavites when painting the namam
or sect mark (improperly called caste mark) on their forehead.

Paraiyans and low caste disciples may not be branded directly by the Guru ; in
their case he heats the instruments and hands them to the Dasari, his assistant, who
performs the actual operation.

With regard to the branding customs of Madhva Vaishnavites, who believe that to
carry a lasting outward and visible sign of their deity on their body helps them to obtain
salvation through him, Thurston (I. pp. 371-373) supplies an interesting account :—
" Madhvas have four mutts to which they repair for the branding ceremony, viz., Vaya-

HORNELL— THE INDIAN CONCH 27

saraya, Sumathendra, and Mulabagal in Mysore, and Uttaraja in South Canara. The
followers of the Uttaraja mutt are branded in five places in the case of adult males,
and boys after the thread investiture. The situations and emblems selected are the
chakra on the right upper arm, right side of the chest, and above the navel ; the chank
on the left shoulder and left side of the chest. Women and girls after marriage, are
branded with the chakra on the right forearm, and the chank on the left. In the case
of widows, the marks are impressed on the shoulders as in the case of males. The disciples
of the three other mutts are generally branded with the chakra on the right upper arm,
and chank on the left. As the branding is supposed to remove sins committed during the
interval, they get it done every time they see their Guru. There is with Madhvas no
restriction as to the age at which the ceremony should be performed. Even a new-
born babe, after the pollution period of ten days, must receive the mark of the chakra,
if the Guru should turn up. Boys before the upanayanam, and girls before marriage
are branded with the chakra on the abdomen just above the navel. The copper or
brass branding instruments (mudras) are not heated to a very high temperature, but
sufficient to singe the skin, and leave a deep black mark in the case of adults, and a light
mark in that of young people and babies. In some cases, disciples who are afraid of
being hurt, bribe the person who heats the instruments ; but as a rule, the Guru regulates
the temperature so as to suit the individual. If, for example, the disciple is a strong
well-built man, the instruments are well heated, and, if he is a weakling, they are allowed
to cool somewhat before their application. If the operator has to deal with babies,
he presses the instrument against a wet rag before applying it to the infant's skin. Some
Matathipathis (head priests of the mutt) are, it is said, inclined to be vindictive, and to
make a very hot application of the instruments if the disciple has not paid the fee (guru-
kanika) to his satisfaction. The fee is not fixed in the case of Sri Vaishnavas. whereas
Madhvas are expected to pay from one to three months' income for being branded.
Failure to pay is punished with excommunication on some pretext or other. The area
of skin branded generally peels off within a week, leaving a pale mark of the mudra,
which either disappears in a few months, or persists throughout life. Madhvas should
stamp mudras with gopi paste1 daily on various parts of the body. The names of these
mudras are chakra, chank or sankha, gfitha (the weapon of war used by Bhima, one
of the Pandavas), padma (lotus), and Narayana. The chakra is stamped thrice on the
abdomen above the navel, twice on the right flank, twice on the right side of the chest
above the nipple, twice on the right arm, once on the right temple, once on the left
side of the chest, and once on the left arm. The chank is stamped twice on the right
side of the chest, in two places on the left arm, and once on the left temple. The gatha
is stamped in two places on the right arm, twice on the chest, and in one spot on the
forehead. The padma is stamped twice on the left arm, and twice on the left side of
the chest. Narayana is stamped on all places where other mudra marks have been made.

1 Properly yopi chandiram, a paste made of white kaolin mixed with sandalwood.

28 OKHAMANDAL MARINE ZOOLOGY— PART II

Sometimes it is difficult to put on all the marks after the daily morning bath. In such cases,
a single mudra mark, containing all the five mudras, is made to suffice. Some regard
the chakra mudra as sufficient on occasions of emergency."

So far as I can learn the branding instruments which are employed to sear the two
chief symbols, chank and chakra, by means of heat are usually made of copper. In
other localities brands of different metals appear to be used as Risley (II, p. 339) states
that the Ramanuja, a Vaishnavite sect in Bengal, when undergoing the initiatory rite
(tapta-mudra) are branded with the chakra on the right shoulder and the chank on the
left, by means of a brand made of eight metals (ashta-dhatu) — gold, silver, copper,
brass, tin, lead, iron, and zinc.

Various deviations from the standard ceremonial exist in certain districts ; among
these is that followed by the Bedar or Boya caste of the Southern Deccan, a caste which
largely constituted the old fighting stock of this district. Among them the men are
branded on the shoulders by the priest of a Hanuman shrine with the sign of the chank
and of the chakra, in the belief that this will enable them to go to swarga (heaven).
Female Bedars who are branded become Basavis (temple women) and are dedicated to
a male deity and called Gandu Basavis or male Basavis (Thurston, I; p. 194).

This branding of temple girls, or Deva-dasi as they are termed in the Tamil country,
with symbols of the chank and chakra is always an essential feature in the ceremonies
which mark their dedication to the god of their temple, whom thenceforward they serve
with dance and song.

Allied to branding is tattooing. The Tandans of Malabar, a caste about the level
of the Tiyyans, adopt this method to show devotion to the deity, and among the religious
symbols worked into the skin of their arms is that of the chank (Thurston, VII, 10).

(c) THE MENDICANT'S CONCH.

Beggars throughout India occasionally use the chank shell as a musical instru-
ment, and with certain castes of religious mendicants it is an essential part of their pro-
fessional paraphernalia, so much so that a Tamil proverb likens things in continual
association to " the breech of the chank and the mouth of the mendicant."

The Dasari, who belongs to a caste of Vaishnavite mendicants well represented in
the Madras Presidency, is often seen in North Arcot and the Southern Deccan, announcing
his arrival in a village by blasts on the chank-shell which in that part of the country is
one of his five insignia. In Telugu districts the Dasaris are more secular and less reli-
gious, and the caste is known as Sanku Dasari or vulgarly Sanku jadi, the chank-blowing
caste.

A mendicant's conch sometimes has the apical orifice mounted in brass ; temple
conchs are usually without any ornamentation, but the Udipi temple owns one very
handsomely mounted in brass and this is sounded whenever the god (Krishna) is carried
in procession in the temple car.

HORNELL— THE INDIAN CONCH 29

((I) DEDICATION OF TEMPLES AND HOUSES.

Wherever a new temple is built, or when a new shrine or god is established and
added to the number already there, its dedicatory ceremonies include as one of the most
important a special libation to the god from the mouths of 108 chanks, or better still,
from 1,008 chanks if so many can be afforded, filled with water and flowers.

The building of an ordinary house in the Tamil country must also have its ceremonial
dedication at the time the foundation trenches are dug, though among low caste people
the rite consists merely of a superstitious act to ensure good luck or to baulk the evil eye.
It is carried out with the help of the sacred chank which thus is seen to touch the lives
of the people at every point from the cradle to the grave. Before a single stone in the
foundation is laid, the ceremony is carried out on a day carefully chosen as being highly
auspicious. A chank is then buried beneath the first stone laid. An old reference to this
occurs in a petition quoted by Wheeler (fide Thurston, III, p. 147) from two natives of
Madras, in connection with the founding of a village called Chintadrepettah, now a
populous division of Madras City. The entry runs : — " Expended towards digging a
foundation where chanks were buried with accustomary ceremonies." Roman
Catholic converts from low castes follow this custom, as well as Hindus ; in Tuticorin,
if a Roman Catholic Parayan desires to build a house, the carpenter employed by him
chooses an auspicious day by reference to a native calendar, a chank is bought in the
bazaar and on the day chosen, having dug a foundation trench and prepared at the
bottom a bed of coral stone and mortar, the chank is laid thereon. In the cavity of the
shell small fragments of five metals (panjalokam), gold, silver, copper, iron and lead,
are placed, turmeric and sandalwood water is sprinkled over, and then the whole is hidden
under a mass of sweet-smelling flowers. The ceremony is ended ; the first stone may now
be lowered into place upon the chank and its contents, and good luck is believed to
be assured to him who will inhabit the house.

It may, however, happen that in spite of every precaution, an inauspicious site
appears to have been chosen as shown by a sequence of misfortunes happening to the
householder. In such cases Hindus may perform a special ceremony called Chank -
usthabanam to remedy the evil. A chank-shell is filled with water and incantations
made for forty-five days. At the end of this period of propitiation, the chank is buried
under the house wall. (Winslow, " Tamil and English Dictionary," p. 390, Madras 1862.)

Among the Parawa caste living in Tuticorin and other coast towns and villages
in Tinnevelly and Ramnad, misfortune is often sought to be averted from the individual
by almost completely burying a chank shell in the floor of the hall (kudam), about two
feet on the inner side of the doorway to the street. A small portion of the back of the
shell shows as a patch of white on the surface of the floor. The explanation of the
custom current among Parawas is that the shell is so placed that when an inmate leaves

30 OKHAMANDAL MARINE ZOOLOGY— PART II

the house, lie must pass over it, usually touching it, and this will prevent misfortune
happening to him.

(i>) MARRIAGE CEREMONIES.

The chank has important but variable fxmctions to perform at weddings among
all Hindu non-Brahman castes in the districts of the south of India, where this shell
is blown by the barber (ambattan) particularly at or immediately after the tying of
the tali or marriage badge around the bride's neck ; the bridal couple usually occupy a
raised platform, and round and round this the barber walks while blowing his chank. In
Bengal this custom of chank-blowirig during weddings is even more general ; a common
formula which runs

" Gauga ka pani samundra ki sank
Bar kanya jag jag anand."

(" May Ganges water and sea-chank betide
Enduring bliss to bridegroom and bride.")

is usually recited during the marriage (Risley, II, p. 116) with reference to the blasts
on the conch which accompany the ceremonies — the equivalent of the marriage bells
in Christian ritual.

In Telugu districts the chank is not used by any caste, non-Brahman as well as
Brahman, during weddings, as this is considered inauspicious because chank-blowing
is specially associated with funeral ceremonies.

Usually a man of a special caste is engaged to blow the chank at the customary times ;
in the Tamil country the caste barbers (ambattans) perform this duty ; among the
Telugus the chank blower is usually a Dasari, among the Uriyas, a Ravulo.

Sometimes, however, women of the family or of the caste perform the chank-blowing
duty. Among Bengal Brahmans, for instance, one section of the elaborate and lengthy
marriage ceremonies consists of a procession of seven married ladies headed by the bride's
mother going round the bridegroom seven times, some sprinkling libations of water
and vociferating the hymeneal cry of ulu-ulu. One of the seven carries a conch and blows
it as she goes (Risley, I, p. 150). A custom somewhat akin to the above prevails among
the Kalian caste of Tanjore, Madura, Trichinopoly and Tinnevelly. On the wedding
day the sister of the bridegroom goes to the house of the bride, accompanied by
women, some of whom carry flowers, coconuts, paddy, turmeric, milk, ghee, etc. On
the way two of these women blow chank shells. (Thurston, III, p. 80.)

In passing it is interesting to note that a section of this caste, the Puramalai nildu
Kalians, practise the rite of circumcision probably as a survival of a forgotten forcible
conversion to Muhammadanism. The rite is carried out in a grove or plain outside the
village and en route to the place, and throughout the ceremony a chank is blown at fre-
quent intervals. (Thurston, III, p. 74.)

It is noteworthy that Brahmaus in the Tamil and Telugu districts do not employ the

HORNELLr— THE INDIAN CONCH 31

chank during marriage ceremonies though their brethren in Bengal do. Among Telugu
Brahmans living in Uriya districts the custom of Bengal used to be followed at marriages,
but this is gradually dying out ; as one Brahman in Berhampur (Ganjfun) remarked
" The present day Brahmans here have more regard for the magic flute than for the
divine voice issuing from the chank."

In Bengal the association of the chank with marriage is more intimate and deep than
elsewhere ; no Bengali lady is properly or legally married unless chank bangles, which
should be lacquered red, be placed upon her wrists. In the Madras Presidency, marriage
bangles are used only by a few sections of the agricultural and pastoral castes (Vellalans
and Idaiyans).

One of the most interesting facts brought to light during the present research is
the weighty evidence we have that in former days the tali, the essential marriage symbol
among Tamils, was directly connected with the chank, either composed of a piece
of the shell or of a metal ornament in the form of a miniature chank shell. We find
this marriage badge named specifically sankhu tali among four castes widely separated
both geographically and in status and civilisation. First are the Chanku tali Vellalans,
a section of the great Vellalar caste, who wear, according to Winslow (" Tamil and English
Dictionary," Madras, 1862), a representation of the chank on either side of a central
symbol. Unfortunately, apart from this reference, I have been unable to trace the
location of these Vellalans at the present day, or to obtain any details of the custom.

Two other castes with the same marriage badge occur on the West Coast, and it
is significant that one is undoubtedly of Tamil origin. This is an immigrant branch
of the Idaiyans known locally as Puvandans, settled in Travancore. On ceremonial
occasions the women wear the Tamil Idaiyan dress while in ordinary life they attire
themselves after the fashion of Nayar women. Their tali is known as sankhu tali and a
small ornament in the form of a chank is its most conspicuous feature. (Thurston, II,
366.)

The other West Coast caste using a sankhu tali is that of the Thandan Pulayans,
a small division of the Pulayans who dwell in South Malabar and Cochin. The women
dress in a leaf skirt made from the stems of a sedge called thanda, which are cut into
equal lengths, woven at one end and tied round the waist so that they hang down below
the knees.

According to Ananthakrishna Aiyar (Thurston, VII, p. 23.) : "At the marriage
ceremony, the tali (marriage badge) is made of a piece of a conch shell (Turbindla rapa)
which is tied on the bride's neck at an auspicious hour. She is taken before her landlord,
who gives her some paddy, and all the coconuts on the tree beneath which she happens

to kneel To ascertain whether a marriage will be a happy one, a

conch shell is spun round. If it falls to the north, it predicts good fortune ; if to the east
or west, the omens are favourable ; if to the south, very unfavourable."

Lastly, and most interesting of all, we find a caste calling their marriage badge

32 OKHAMANDAL MARINE ZOOLOGY— PART II

sanklm tali, which on examination shows no likeness to a chank shell. These are the
Parawas of the coast towns on the Indian side of the Gulf of Mannar. When the Portu-
guese arrived there early in the sixteenth century, these people, who were principally
pearl fishers, chank divers, and fishermen, were orthodox Hindus, but the stress of
Muhammadan competition drove them into alliance with the Portuguese and they went
over in a body to the Roman Catholic church. To-day the badge tied around the bride's
neck on marriage consists of three ornaments, a central cross flanked on either side by
the symbol of the Holy Ghost ; nevertheless it is called sankhu tali as among the castes
first mentioned. There is no doubt that when the caste was a Hindu one the tali was
true to name, indeed Parawa tradition is definite, for it asserts that originally the central
ornament was a small figure of some Hindu God (probably Krishna) flanked by one of
a chank shell on each side. The use of the original name is a strange persistence in view
of nearly 400 years' sojourn within the Christian fold; it is one of the many signs of toler-
ance shown by the Roman Catholic priesthood towards their converts' prejudices on
immaterial points — a tolerance in petty matters that has done much to help that church
in its propaganda.

Among some castes, including the Bauris and Dandasis of Ganjam, turmeric water
from a chank shell is poured seven times over the hands of bride and bridegroom, which
are tied together with seven turns of a turmeric-dyed thread. (Thurston, Vols. I. and II.)

(f) DEATH CEREMONIES.

Throughout the Tamil country all non-Brahman castes which observe Hindu
rites have the chank sounded as the body is being taken to the funeral pyre or to the
burial ground. It is usual also to employ the conch-blower on the last day of the sraddh
ceremonies in those castes which follow the orthodox ritual. Among the Telugus
these same rites are largely followed, but it is said that Vaishnavites do not observe
them. Among both races, the Brahmans do not have the conch blown at any period
of the obsequies — a sign that lends weight to the theory that the chank has been borrowed
by Brahmanism from another religion. •

In the Madura and Tinnevelly districts the conch-blowers at a funeral are Ambattans
or barbers, the same caste as performs likewise at weddings. Among the Idaiyans
of these and the neighbouring districts one part of the funeral rites consists in the son
perambulating the pyre thrice with a pot of water on his shoulder ; at each turn the barber
makes a hole in it with a shell when the head of the corpse is reached. Finally the pot
is broken near the head. (Thurston, II, p. 362.)

Further north, in the East Coast districts from Tanjore and Salem to the Kistna
River, the Panisavans are by caste custom the funeral conch-blowers ; they may indeed
be accounted the undertaker caste, as it is their duty to carry news of the death to the
relations of the deceased. It is they who generally keep all the materials necessary

HORNELL— THE INDIAN CONCH 33

for the funeral including the palanquin required for the conveyance of the corpse to the
cremation ground. At the funeral, the Panisavan follows the corpse, blowing his conch.
When the son goes round the corpse with a pot of water, the Panisavan accompanies
him sounding his conch the while. On the last day of the death ceremonies (Karmand-
hiram) the Pansivan should also be present and blow his conch especially when the tali
is removed from the widow's neck. (Thurston, VI, p. 56.)

The insignia of the Panisavans are the chank and the tharai, a long straight trumpet.

In Coimbatore district, the duty of sounding the death conch belongs to the
members of an important sub-division of Paraiyans, called on this account, Sankhu
Paraiyan. (Thurston, VI, p. 81.) In Travancore when a headman or kaikkaran of
the Paraiyans settled there happens to die, a chank-shell is buried with the corpse.
(Thurston, VI, p. 134.)

The chank sometimes has a place in the death ceremonies of castes which are not
Hinduised. Thus among the Cherumans of Malabar and Cochin, a caste of agricultural
serfs, according to Mr. Ananthakrishna Aiyar (Thurston, II, p. 81.), " The son or nephew
is the chief mourner, who erects a mound of earth on the south side of the hut, and
uses it as a place of worship. For seven days, both morning and evening, he
prostrates himself before it, and sprinkles the water of a tender coconut on it. On the
eighth day, his relatives, friends, the Vallon, and the devil-driver assemble together.
The devil-driver turns round and blows his conch, and finds out the position of the ghost,
whether it has taken up its abode in the mound, or is kept under restraint by some deity.
Should the latter be the case, the ceremony of deliverance has to be performed, after
which the spirit is set up as a household deity."

How far the conch is used in funeral rites outside the Madras Presidency, I am not
in a position to say, except in regard to Thibet, where as already incidentally mentioned,
it is a custom to sound it as the body of a monk or a nun is being conveyed from the
place where death occurred.

(g) TOTEMS.

Totems as the distinctive signs of exogamous septs must have been at one time
universal among tribes of Dravidian origin. To-day a well developed totemistic system
characterises the tribal organisation of the Santals and Oraons who retain languages
distinct from those of the surrounding peoples, know nothing of the caste system, and
who continue to worship non-Aryan gods. Among the Santals 91 septs are known,
and one of these is known as Sankh. The members of this sept may not cut, burn,
nor use the shell, nor may the women of this sept wear it in personal adornment.

Above these still primitive tribes and between them and the fully Hinduised peoples
who are split up into castes, are a large number of partially Hinduised tribes which
in many cases show distinct traces of a totemistic organisation. Among these the

c

34 OKHAMANDAL MARINE ZOOLOGY— PART IT

Kurmis of Bengal, and the nomad Koravas who wander throughout the peninsular
part of India, both have an exogamous sept or gotra of which the totem is the chank-
shell. Among the Kurmis this sept is called Sankhawar ; its members are prohibited
from wearing ornaments made from chank-shells. With the Koravas it is termed
Samudrala, signifying the sea, and people of this sept may not use the chank in any way.
Higher than these are the Kalinjis, an Uriya agricultural caste, and the Kurubas, a caste
of shepherds and weavers widely spread throughout the Madras Presidency. Both
castes comprise septs named after the chank, in the case of the Kalinjis, Sankho, in that
of the Kurubas, Sankhu. I am not aware whether the septs among the former caste
have now totemistic value, or if it has become merely a name, a gotra name ; in any
case it may be taken as certain that in the pre-Hinduised condition, the name of the gotra
was of real totemistic value. Bhago (tiger) and nago (cobra) are names of two other
gotras of obviously totemistic origin. With the Kurubas, the sept is undoubtedly
exogamous and its totemistic character certain.

Another caste or sub-caste showing by the names of its sections a probable totemistic
origin is that of the Koppala or Toththala, a sub-division of the Velamas, a caste of
agriculturists in the Vizagapatam district. Among their sections are some named
Naga (cobra), Sankha (chank), Tulasi (basil or tulsi) and Tableu (tortoise). At the present
day these divisions although apparently of totemistic origin, have no significance so
far as marriage is concerned. (Thurston, VII, p. 340.)

(h) EVIL-EYE SUPERSTITIONS.

Belief in the reality of the malign results which ensue from being overlooked by the
evil-eye is frequently present in an acute form in the Madras Presidency. It is specially
dreaded in the case of houses under construction and in respect to valued cattle. Every-
where in Tamil districts the custom prevails more or less extensively of seeking protection
for draft bullocks by tying a small chank-shell upon the forehead of such as are in
good condition or in any way specially valuable or beautiful in their owner's eyes. Of
late years the custom has tended to fall into abeyance in certain districts. As is to be
expected the people of country villages cling to it with greater tenacity than those in
towns. Many, however, decorate their bulls in this way without thought of it as an
amulet against evil — to them it is merely an old custom to be followed, or else they put
it on their favourite animals as an ornament to mark the pride they have in them.
Again, some, from a peculiar shyness often met when discussing such matters with the
peasantry, deny that the chank-shell is used as an amulet although in reality it may be
so used by them.

In the southern Deccan the custom appears to be falling more quickly into abeyance
than in Tamil districts. The Collector of Kurnul informs me that though the practice
survives in parts of Dhone, Kumbum, Koilkuntla and Sirvel taluks of tying a chank

HORNELL— THE INDIAN CONCH 35

on the forehead or round the necks of bullocks and ponies, it is gradually dying out ;
in Bellary the Collector states that it still prevails in Bellary, Hospet and Hadagalli
taluks in respect to bullocks, but adds that the people do not now attach any religious
or secular significance to it. In Anantapur the Collector states that the custom is not
now followed in that district.

In Madras City it is quite common to see it and there also I have seen a shell hung
by a chain or a cord round the neck of a cow-buffalo when in milk to prevent her being
" overlooked " and her milk thereby dried up prematurely. In country villages this
latter custom is not infrequent both in regard to ordinary cows and to cow-buffaloes.
In the Madura, Trichinopoly, Salem and adjacent districts a shell is often hung round
the necks of jutka and pack ponies, not only by Hindu owners, but also by Muhammadans.
In Madura specially valued sheep occasionally are similarly guarded from evil, and in
the same district I have seen milch-goats protected in a like manner. In all these cases
the shells used are of small size, the great majority being dead or sub-fossil shells from the
mud-beds of Ceylon. They are the same as are sold as feeding spouts for infants in
every Tamil bazaar. A hole is bored or broken in the back and the rope passed through
this and out at the mouth of the shell. The surface is generally roughly engraved in
a coarse spiral or scroll pattern.

Used probably for a similar purpose may have been the handsomely engraved large
chank shell obtained from an oblong sarcophagus of red pottery found in the prehistoric
burial site at Perambair in the south of Chingleput district. The size is much greater
than any I have ever seen used to decorate ordinary cattle ; no ordinary person owned
it we may be certain — probably it decorated the forehead of a bull or possibly of an
elephant belonging to a man of great local importance. From the same tombs came
three other handsomely decorated chank ornaments, two of which were probably
ornaments for the hair (see p. 41 for further particulars).

Once only have I seen ' more than a single shell hung round the neck of any animal,
but that in ancient times a different habit at least occasionally prevailed is possible,
to judge from the string of 16 small chank shells from a barrow near Guntakal junction
in the Anantapur district, now to be seen in Madras Museum. All these shells have
had the apices broken in and partly rubbed down and each has the thickest part of the
body perforated from side to side so permitting them to be strung together. They
appear to have formed a necklace, but whether they were suspended round the neck of
a bull or may be hung like a chain of office round the neck of some person of importance,
we have no means of determining.

In Malabar the chank is little in evidence, but Logan (" Malabar Manual," 1887.
Vol. I, p. 175) records a belief there prevalent that a cow will stop giving milk unless a
shell (not necessarily a chank) is tied conspicuously about her horns and at Tanur,
Malabar, I have seen valuable sheep with shells other than chanks hung round the neck.

1 At Tirupalakufli, on Palk Strait.

C -2

3G OKHAMANDAL MARINE ZOOLOGY— PART II

Further north, in the coast villages of South Canara, south of Mangalore, rings
made from Strombus shells but known locally as chank rings are employed by parents
to avert the evil eye from their young children. At Kasargod, Bekal and the adjacent
villages I have found the custom especially common among the Mukuvans, a caste of
immigrant Malayali fishermen. Children from 3 to 4 years old in these villages are
frequently given necklaces made of Strombus rings alternating with elongated glass
beads. Some other castes in the same neighbourhood, Mokayans and Tiyyans together
with some Mappillas, are said to follow the same custom. Usually the rings do not
exceed twelve in each necklace. Adults do not wear these amulets as is the habit of the
woman of certain sections of the Cheruman caste in Malabar (cf. p. 39) and of the
Hill Vedans of Travancore. How far these and other facts connote former wide or
even universal prevalence of this habit among the indigenous population of Malabar,
is a line of inquiry likely to repay careful investigation.

Finger rings purporting to be made from chank shells, but usually cut from a small
species of Strombus common on the western coast of the Gulf of Mannar, are also used
very freely throughout the Tamil country and also in Malabar and Cochin, chiefly by
non-Brahmans among Hindus, as amulets against evil spirits, the evil eye and certain
sicknesses. In Tinnevelly, Madura and Ramnad the custom is very prevalent among
both sexes of non-Brahmans. Labbai and Marakkayar Muhammadans in whose veins
much Hindu blood is present, also affect the custom. The Vellalans, although like
Brahman adults, they wear, except in one section, neither chank-bangles nor rings,
often provide their children with chank-rings or else with pieces of chank-shell tied on
the wrist of the right hand by means of black thread, as an amulet against the disease
called chedi which is, I believe, rickets. In some cases the ring or piece of chank is
placed on the wrist only when the disease has laid hold of the child, in others it is tied
on when the child attains its second month and kept there till it is three years old,
when it is believed that all danger of contracting the disease has passed. Among the
castes ranking next — Chettis, Kollans, Thachchans, Thattans, Naidus, Idaiyans, and
Chaluppans, a chank-ring is often worn as an amulet against pimples on the
face ; occasionally their young children are provided with small and roughly
ornamented chank-bangles to safeguard them against chedi. The low castes or
Panchamas such as Pallans, Vallayans, Paraiyans, etc., are the most regular devotees
of these amulets.

The Roman Catholic Parawas of Tuticorin and the other Parawa strongholds
on the Pescaria Coast have also been great believers in the virtue of chank amulets,
and till recently all babies were given chank-bangles to protect from convulsions and
from chedi. Even now the poorer and more ignorant continue to employ these amulets,
keeping them on the wrists for about three years. The richer and better educated
have either abandoned the practice or keep the bangles on for a much abbreviated
period. The Parawas formerly also employed pieces of the curious egg-capsule of the

HORNELL— THE INDIAN CONCH 37

chank for the same purpose as the bangle, a fragment of the capsule (chanku-pu, literally
" chank-flower ") being tied by means of thread upon babies' wrists.

In Madura chank amulets are used even more freely than in Tinnevelly. In addition
to bangles and rings used as amulets against the evil eye, or ailments such as chedi
and pimples, very roughly fashioned and imperfectly rounded fragments of chank
shells are used in the manner of beads to make necklaces which are used as amulets.
Maravans, Paraiyans, and Chakkiliyans are among the castes chiefly addicted to the
wearing of these and other chank amulets ; these people often give their children both
chank-bangles and necklaces of chank-beads with a view to multiplying the counter-
vailing influences against the evil eye and against disease.

Chank-bead necklaces (chanku malai) are also worn largely by children of the
poorer Chettis and of the Vanniyans (oilmongers) who, though they do not generally
wear chank-bangles, will wear these chank-bead necklaces. The people of the lower
castes also use the same rough beads to make bracelets (chanku pasi) worn on the wrists
for the same object as the bead necklaces. Similar customs in regard to rings and
bead necklaces prevail in Tan j ore and in South Arcot, where the low castes, especially
Vanniyans, Koravans, Paraiyans, Chakkiliyans and wandering Lambadis, generally
wear them as amulets against evil spirits, the evil eye and sickness. Koratti women
appear to be the only ones in South Arcot who wear bead bracelets in addition to other
bangles according to information kindly supplied by the Collector, Mr. Azizuddin.

The trade in these amulets is of considerable dimensions. Thus the Tahsildar
of Chidambaram reports that rings worth about Rs. 500 are sold yearly in his taluk
to the residents and to the thousands of devotees who flock to the great temple of
Nataraja. The price runs from two pies to one anna per ring — some are very rough
untrimmed Strombus rings, while the higher priced may be of real chank with a few
oblique lines of ornament sawed or filed on the outer surface.

The rough beads used in making necklaces and bracelets look very much like tiny
carpal bones ; they sell about eight beads for one pie — a whole necklace may be
bought for one anna ; the small bangles worn by babies in Madura and Tinnevelly cost
3 to 5 annas per pair. The latter are made largely in Kilakarai, a large Muhammadan
settlement on the Ramnad coast, near the head of the Gulf of Mannar. The majority
are cut chiefly from under-sized shells too small to have any value in the Bengal market.
By rights these shells should not be fished ; they should be put back alive by the divers
in order to grow to adult dimensions. This precaution is followed as far as circumstances
permit in the Tuticorin chank fishery which is conducted departmentally by the
Government of Madras ; in the Ramnad fishery the short-sighted greed of the renter
and his employees takes no account of any such precaution for the future prosperity
of the fishery.

38 OKHAMANDAL MARINE ZOOLOGY— PART II

(i) FOR PERSONAL ADORNMENT.

In personal adornment, and apart from any uses it is put to in the form of amulets,
the chank-shell is employed principally as the crude material wherefrom beautiful
bracelets of many patterns are made for use in Bengal and the adjoining provinces ;
subsidiary uses to which it is put are to fashion from it finger rings, necklaces, disc
ornaments for headdresses and caps, and as a recent addition, coat and dress buttons.

The bangle industry in all phases is treated separately and at full length in Section
II, to which the reader is referred for details.

Rings actually made from chank-shell are not manufactured in any quantity ;
their place is taken largely by those made from a much smaller shell, a species of
Strombus, found on the Riimnad coast of the Gulf of Mannar. So far as I can learn,
the industry is localised at Kilakarai, a seaport of Ramnad inhabited largely by
Muhammadan (Labbai) fishermen, pearl-fishers and chank-divers. The rings are usually
exported with a minimum of finish ; only the roughness of the edge is rubbed off and
nearly always the chestnut stippling is clearly recognisable. Many are sold in the
bazaars or by peddlers throughout the Tamil and Malayalam districts, usually as amulets
against the evil eye and against such minor ailments as pimples on the face and various
skin troubles. So far as I can ascertain, the only people who use these rings in personal
adornment are two tribes of low civilisation living in the Malayalam country — the Hill
Vedans of Travancore and certain sections of the Cheruman tribe in Cochin and
Malabar. The former I have not seen. They are described by Thurston (VI, p. 333)
as living in wretched huts and employed chiefly as rice-field watchmen. He states
that both the men and women of this tribe wear numerous bead necklaces interstrung
or otherwise associated with a few of these rings. In a photograph given by the same
authority (Vol. Ill, p. 177), a man is shown wearing numerous strings of glass beads
passed through eight Strombus rings. In the case of the women these necklaces hang
down as far as the abdomen.

The Cherumans were formerly the agrestic serfs of Malabar, Cochin and Travancore
—the Malayalam country or Kerala. To-day they still remain largely in a servile
condition, carrying on for their masters the heavy labour of the fields ; they receive
their pay almost always in kind. They are divided into a considerable number of
endogamous sections differing in appellation in different districts — a sign of long-
continued residence in a country of difficult intercommunication.

All Cheruman women are greatly addicted to the use of necklaces, particularly
of the showy strings of beads now put within the reach of the poorest by the enterprise
of Austrian and Italian manufacturers. Of other clothing they wear the scantiest —
a very dirty, once white cloth, pendant from the waist, being their usual garb. Certain
sections wear as a distinctive badge, in addition to numerous bead necklaces, a long cord
whereon are strung large numbers of Strombus rings (chanku modiram) — they believe

MARINE ZOOLOGY OF OKHAMANDAL, Pt. II

INDIAN CONCH, PLATE III

[Photo by J. Horndl.

Fig. 1. — Group of Cheruman women wearing necklaces of
so-called chank-rings.

[rhoto l>y Vividha Ktdn Mandir.

Fig. 2.— Selling chank shells to pilgrims returning from Bet.

HORNELL -THE INDIAN CONCH 39

them to be of chank-shell. The bead necklaces are usually wound many times round
the neck itself, roughly forming a collar often reaching as high as the chin. The chank
necklace is worn at a lower level, and lies on the shoulders and on the upper part of
the breast ; it looks much like a chain of office and is indeed the badge of the tribal
sept. At Tanur (Malabar) where after much trouble three Cheruman women were
got together for my inspection, one of the husbands had to be paid a day's wages to keep
guard over them to prevent their flight. They were all exceedingly shy, and it was
with much reluctance that they stood up in front of my camera. As will be seen by
reference to PI. Ill, Fig. 1, the chank ring necklaces (chanku modira mala) are made
up of a very large number of rings not strung but tied by the upper edge to a strong
cord in such a way that each ring overlaps its neighbour on one side and is similarly,
overlapped on the other side by the succeeding ring, much as the rings in chain armour
are arranged. From 50 to 100 rings are required to form a full necklace of this pattern ;
as each ring costs from 3 to 6 pies in the local bazaar, the total cost may amount to
Re. 1-8-0 or Rs. 2, a large sum to these exceedingly poor people. The Cherumans who
wear these chank chains in the Tanur neighbourhood say they belong to a sept named
Kalladi Cherumans and that they wear them to distinguish themselves from the Paliya
and other septs with which they may not intermarry. In Tanur bazaar I saw a single
example of another pattern of this strange necklace worn by a woman also said to be
a Kalladi Cheruman. In this case the number of rings used were comparatively few,
20 in all, and between each pair were strung a couple of glass beads of different colours.
Each ring was separated by an interval of about an inch from its neighbour on either
side, and instead of being fastened to the common cord by a single loop, it was fastened
by two separate loops which enabled it to lie flat upon the skin. The woman shrank
against the wall, averting her face and trying to sidle away, and it was with great
difficulty she could be persuaded to answer a few particulars. Among other information
she gave, was the statement that this necklet is believed to protect from evil spirits.

So far as I have been able to ascertain, these chank necklaces are assumed soon
after a girl attains puberty if her parents can afford it. If they be very poor and cannot
afford it, then, when her marriage is arranged, it is generally settled that the bride-
groom shall provide the needful ornament. There is no special ceremony followed
at the time a girl puts on her chank necklet for the first time. As a rule the men of the
family attach the rings to the cord.

This custom seems to be losing ground quickly, for while many people knew of it
further south in Malabar, I never saw this ornament in use in North Malabar. Many
Cherumans were seen between Cannanore and Mount Dilly, but all said few use it now,
preferring glass or imitation coral beads for their necklets. There is no doubt that
formerly the custom was widely spread among the servile population of Kerala, and as
these people's religious beliefs consist almost solely of the dread of malignant spirits,
it is extremely probable that originally the necklet was used as an amulet against demons

40 OKHAMANDAL MARINE ZOOLOGY— PART II

and the evil eye, though now it is more generally considered as a sept badge. The
custom of long-settled Malayali immigrants (Mukuvans, etc.) on the South Canara
coast, of putting similar necklaces round their children's necks already referred to on
p. 36, appears to furnish strong corroboration of this conclusion.

In Bengal a few ornamental finger rings are now made, carved in simple patterns
and highly polished. These are not in great demand, and I am uncertain as to whether
they are worn as ornaments or as amulets. At Kilakarai a few roughly decorated
thin finger-rings to be used as amulets are also produced, in addition to the roughly
made, thick and clumsy sections cut from Strombus shells.

The first mention of the use of discs cut from chank-shells to ornament caps and
headdresses occurs in Ta vernier's " Indian Travels." In 1666 he was in Dacca and
records the fact that Bhutan merchants took home quantities of " round and square
pieces (of shell) of the size of our 15 sol coins." He also states that " all the people
of the north, men, women, girls and boys, suspend small pieces of shell both round
and square from their hair and ears."

Whether the trade is as large as in former days, I cannot say. It is now of small
monetary value.

The Nagas of Assam, lately brought to prominent notice through the good work
they did as carriers during the Abor punitive expedition (1912), employ these discs
both to form necklaces and to decorate the handsome plaited cane helmets worn by the
men. These latter are conical in shape, about a foot high, and covered with a layer
of fur and hair, black or red in colour. When decorated with chank-shell discs, these
are arranged as coronals, adding most effectively to the general design (W. Crooke,
" Natives of Northern India," p. 47, London, 1907). As the Nagas are known to have
set much greater store by the chank in former times, say prior to the middle of the
nineteenth century, it is probable that then the use of charik discs as items of ornament
was much more general among this race than it is now. Still the custom is quite common,
for Mr. Stanley Kemp, who accompanied the Abor expedition as naturalist, informs
me that the Naga coolies employed as carriers frequently wore necklaces formed of
square concave portions of chank-shell with a large cornelian set en cabochon in the centre.
Sometimes long cylindrical beads made from chank shell, tapered slightly at either end,
were used instead and cornelian beads were often seen in conjunction.

In the middle of last century Major John Butler mentions (" Travels and Adven-
tures in the Province of Assam," p. 148, London, 1855) that at sixteen years of age a
Naga youth " puts on ivory armlets or else wooden or red-coloured cane ones round
his neck. He suspends conch shells with a black thread " (round his neck) " puts
brass ornaments into his ears and wears the black kilt ; and if a man has killed another
in war he wears three or four rows of cowries round the kilt." From a specimen of
chank-shell necklace from the Naga hills contained in the ethnological collection of the
Indian Museum, Calcutta, it appears that the shells before being used were bisected

HORNELL— THE INDIAN CONCH 41

longitudinally, each half being hung as a pendant by one end from the cord encircling
the neck, the whole forming a most uncomfortable-looking decoration, particularly
as the custom is to wear them slung at the back of the neck.1

Sixty years ago chanks constituted the currency of the Naga tribes, but with the
advent of the rupee, the consideration in which these shells was held largely disappeared,
and now these quaint chank necklaces are seldom worn. Mr. Kemp saw them worn
on only one or two occasions during the Abor expedition (1912).

At death these ornaments and all the other items of the deceased's dress, together
with all his treasured weapons, are laid upon the grave.

Among the Abors the custom of wearing chank ornaments must be very rare,
for Mr. Kemp, who most kindly gave attention to this subject, saw only a single
instance — a Gam or headman of Komsing village, who was found wearing a necklace
composed of round concave discs of shell.

The furthest point east to which I have been able to trace the use of chank discs
is the banks of the Upper Mekong to the northward of Tali-fu in the Chinese Province
of Yunnan. Here Prince Henri d'Orleans (" From Tonkin to India," p. 174, London,
1898) found the women of the wild Lissu tribe, a branch of the Lolo race, " often naked
to the waist ; they had a little hempen skirt and a Chinese cap decked with cowries
and round white discs which are said to be brought from Thibet and looked to me as
if cut out of large shells." In some villages they wore a heavy turban in place of the
little white-disc'd cap.

The finest discs I have seen are prehistoric in age, having been taken from the very
peculiar oblong sarcophagi, made of red pottery and raised on 6 or 8 stumpy legs, from
the ancient graves at Perambair in the south of Chingleput District, near Madras. These
discs, two in number, now on exhibit in the Madras Museum, are respectively about
2| and 3 inches in diameter ; in both there is a small central perforation. They appear
to have been cut from the belly of large shells as the convexity is not great. The convex
surface in each case is ornamented with geometrical patterns (different in each case)
of much delicacy. One is illustrated on PI. XXXIII. of the report for 1908-09 of the
Director-General of Archaeology.

The shape and size of these ornaments and the character of the incised patterns
suggest that they have been used as boss ornaments for the back hair in the manner
affected by native women of certain castes in South India. Until I had seen these
ancient chank ornaments I had never heard of the chank shell being used for this purpose,
but subsequently I have been told that the custom still survives in Travancore and that

1 Similarly bisected chariks hung by a cord round the neck are also seen among the Chins of the
Central and Northern. sections of the Chin hills in Burma. My informant, Mr. W. Street, of the Burma
Commission, states that the women alone wear this neck ornament ; usually a single shell is used and
apparently fresh supplies no longer come into the country as those now worn are heirlooms in the
families of the wearers. It is probable that cessation of the supply synchronised with the discontinuance
of chank shell currency among the Naga tribes living to the north of the Chin country.

42 OKHAMANDAL MARINE ZOOLOGY— PART II

when the wearers cannot afford gold these boss ornaments for the back hair may also
be made from ivory, bone, horn and even coconut-shell. These are usually richly
carved and frequently mounted in gold. The central hole in the chank disc would in
such cases be used to secure the ornament to the hair.

Beads made from chank -shells do not seem to be used except to form bracelet
and necklace amulets. I have seen no carefully worked and polished beads suitable
for purely ornamental use. It is possible, however, that necklaces have been made
from the pearls which are occasionally, but very rarely, found in the flesh (mantle)
of the chank. Such pearls are not uncommon in the West Indian conch which produces
them in sufficient frequency to constitute them regular items in the jewellery trade.
These " pink pearls " as they are called, are usually made up into necklaces. The Indian
chank is a much smaller shell, and although fished in far greater numbers than the
West Indian shell, it is exceedingly rare for a pearl to be found. The colour of the few
found varies from porcelain white to pale pink, and while it would be a matter of the
greatest difficulty to obtain enough during many years' search to make a necklace,
matching the colour and grading the size of the pearls to make the ornament a thing
of beauty, is well-nigh an impossibility. I have three of these chank pearls in my
possession ; they are the only ones I have ever seen. The largest is a perfect sphere,
If inch in diameter, porcelain or opal white in colour, of lovely skin grained with a most
peculiar mottling something after the fashion of the " watering " of watered silk.
Another is slightly elongated in one axis (y\ inch x £ inch), oval or elliptical in outline,
of a very pale pink tint and possessing also the peculiar watered grain shown by its
fellow. The third is salmon coloured, almost spherical, with a diameter of \^ inch.

A few coat buttons are now made from chank-shells at Dacca — a recent departure
on the part of one or two cutters who have made a feeble and ill-sustained attempt
to open up new sources of demand. The main obstacle to the success of this new
departure lies primarily in the lack of power-tools to cut up, drill and polish the material
more cheaply than is possible so long as dependence is placed upon hand labour,
however low be the wages paid. Granted even this change, great difficulties in the way
of success exist in the lack of artistic versatility characterising the chank cutters' trade
and the inability of the ordinary Indian manufacturer to appreciate the value of a
judicious advertisement of his wares. He grudges to pay out money in advertisement
and when he does so he usually brings about the loss he fears by lack of foresight in
keeping up his stock of the advertised article or by the foolish as well as dishonest trick
of sending an inferior article to that ordered and paid for by the customer who answers
his advertisement.

With power machinery utilised by firms trading on sound and honest principles,
there should be a very great field for the sale of chank buttons. There is nearly always
a good demand for handsome buttons suitable for the decoration of ladies' jackets and
coats and owing to the beautiful porcellaneous appearance of chank-shell when cut

HORNELL— THE INDIAN CONCH 43

and engraved with some attractive or distinctive pattern, suitably designed buttons
should meet with appreciation in the European and American dress trade. Rough
cut buttons priced at what seems to the European ridiculously high rates are worse
than useless, and beyond this the imagination of the Dacca manufacturer cannot soar
— at present.

(j) FEEDING SPOUTS.

In the ordinary everyday life of the people of Southern India, the chank subserves
several useful functions. Some of these have already been touched upon, but the
most useful remains to be mentioned — that of small shells used as feeding spouts when
weaning infants. The bazaars in every big Tamil town furnish these primitive
utensils, made from undersized shells usually of the sub-fossil description obtained from
the muddy lagoons near Jaffna in Ceylon.

The shells are prepared for market by breaking down parts of the inner portion
of the terminal whorls just inside the mouth and by removing the central part of the
columella. The canal-shaped canaliculum of the mouth is deepened and straightened
to form a rough spout ; the exterior surface of the shell is rubbed down and upon it
is engraved a rude pattern, usually in the form of a spiral scroll with a few star-shaped
emblems ; last of all it receives a thin coating of fine lime or whitewash to hide imper-
fections and improve the colour (PL VI, Fig. 4). For the purpose intended it is quite
effective, but how far the crevices of the interior, by offering obstacles to efficient
cleansing, harbour and promote the rapid growth of bacteria and so lead directly to
infantile diarrhoea, it is difficult to say. If the shell be boiled daily, a very simple pre-
caution and easier to do in the case of a chank than in that of a glass bottle, there would
be no danger, but I fear this is seldom thought of. In feeding baby monkeys just taken
from their mother I have found this feeding shell most useful ; the sight of the little
creature hanging on with both fore-paws to the shell, half choking in its eagerness to
swallow the milk and all the time trying to locate every noise and movement in the
room with its great nervous eyes is one of the quaintest pictures imaginable.

(k) CURRENCY.

That the chank once served a savage people as a form of currency is little known,
but so it was in the Naga country of Assam until less than 50 years ago. Major-General
John Butler, who commanded an early expedition into the Naga hills, tells (loc. cit.,
p. 157) that he found the Nagas of many villages using charik-shells as currency with
a fixed and thoroughly well-determined exchange value relative to the price of all
articles of trade. Slaves arid cattle in particular were always valued in chank-shells.
Thus while a male slave was worth one cow and three chank-shells, a female slave—

41 OKHAMANDAL MARINE ZOOLOGY— PART II

much more valuable, the suffragettes will learn with pleasure, than a mere man — was
worth as much as three cows and four or five charik-shells. Now a cow was valued
at ten chank-shells, a pig at two shells, a goat was the same rate, and a fowl at one
packet of salt. As a chank-shell was considered worth one rupee, a short calculation
will show that a male slave was worth Es. 13, and a female slave Rs. 34, or 34 shells.
The ransoms of villages captured during raids in these good old days were largely paid
in chank-shells, beads, cows, pigs and other portable wealth. Charik-shells and beads
were the chief items of currency, but even in Butler's time the inevitable invasion of
the rupee was already successful in the valleys most accessible to low-country traders.
At the village of Hosang-hajoo the chief remarked to Major Butler, with a show of con-
siderable pride, " since we became British subjects, we have paid revenue in coin and
with it we can procure anything we require ; we therefore no longer want shells and
beads."

I see no reason to believe that chank currency ever extended beyond the hill peoples
of Assam and possibly some of the adjoining hill tracts. On some coins issued by the
ancient Pandiyan and Chalukyan dynasties of southern India a chank-shell appears
as the principal symbol (Thurston, I, 328) ; this might be held as evidence of a pre-
ceding currency consisting of the actual object so represented, whereof the memory
was perpetuated in pictorial form upon one face of the coins and tokens which came to
take its place as more convenient units of exchange. But there is much more reason to
believe that the chank was represented on such coins for a similar reason to that which
actuates the present-day States of Travancore and Cochin to adopt a similar symbol
on their current coins. In these two States, the homes of southern Hindu orthodoxy,
the chank-shell symbolises the religious belief of the ruling race and is their emblem
as the rose stands for England and the thistle for Scotland. Both these States utilised
it as a distinctive symbol on their earliest issues of local postage stamps in place of and
to the exclusion of the sovereign's head — the customary pivot of design in European
stamps.

Both the States of Travancore and Cochin also employ the chank in their recently
designed armorial bearings. In the case of Travancore, the arms described in heraldic
terms consist of : — Argent, on a fesse azure, three reversed (sinistral) chank-shells or ;
Crest — a sea-horse proper. Motto — Dharmosmal Kaladevatam. In Cochin the shield
bears more numerous devices ; in addition to a left-handed chank, a palanquin, a brass
lamp and an umbrella are depicted, with elephant supporters as in the case of Travancore.
In all cases where these States use the chank symbol, it is necessary to note that it
should occur in the abnormal sinistral or reversed form, this being the Royal and Sacred
Chank — the Chank of Vishnu.

When the Maharaja of Travancore performs tulabharam, a coronation ceremony
wherein he weighs himself in scales against gold, special gold coins are struck called
tulabhara kasu (cf. our Maundy money). On one side a figure of a chank-shell appears,

HORNELL— THE INDIAN CONCH 45

on the other the legend " Sri Padmanabha " in Malayalam characters. After the
ceremony these coins are distributed among the Brahmans who have assembled from
all parts of the country.

(I) CHANK LIME.

A minor use to which chank-shells are put in the coastal districts where they occur,
and also in those localities in Bengal where bangle factories exist, is to calcine these
in kilns. The lime so produced is esteemed the best quality obtainable in India, fully
equal to, if not better than, that obtained by burning pearl oyster shells. The
auspicious nature of the shell adds further value to the product, and when a temple
or shrine or specially fine newly -built house has to be whitewashed, chank lime is greatly
sought after for this purpose in the Tamil districts. I have even received petitions
praying that permission be granted for the collection of chanks for this purpose.

At the present day the fact that almost all the produce of the South Indian chank
fisheries is exported to Bengal, makes it very difficult to obtain chank lime — the shells
are too valuable to calcine. That it was not so in former times, in some cases at least,
is to be seen if we inspect the walls of the old temples at Korkai, the seat of the
Tinnevelly chank fishery 800 to 2,000 years ago. The mortar still contains many
recognisable fragments of chank-shells.

(m) IN MEDICINE.

Apart from the uses to which chank rings and bracelets are put as amulets against
certain ailments, the shell itself in several ways is used medicinally. Except in cases
which have come under my personal notice, it is somewhat difficult to ascertain the
exact nature of the diseases for which native practitioners employ this specific ; custom
appears to vary with different districts and even with different " doctors " living in
the same town.

Of some there is no doubt. The belief is general throughout Tamil districts and
Malabar that water which has been in contact with an article formed from a chank-
shell is a charm against and a remedy for blotches, pimples and other skin troubles on
the face and body. A chank ring worn on a finger is an easy way of applying the remedy,
as water applied to the face or body by the hand must necessarily have been in contact
with chank substance and so able to transmit the virtue thereof. This remedy is believed
to act still more beneficially if the ring be rubbed upon the affected parts. In South
Arcot, Tanjore, Coimbatore, Salem and Trichinopoly, certain skin diseases, eruptions,
warts and even haemorrhoids are believed to yield to this treatment. In Coimbatore
native doctors prescribe a paste made by mixing chank powder in water or by rubbing
it up with human milk for use as a salve in the case of eruptions (sties) on the eyelids.
Chank ointments (basmams) are also employed in the same district to cure inflammation

46 OKHAMANDAL MARINE ZOOLOGY— PART IT

of the eye, the growth of bad flesh (granulation) on the interior surface of the eyelids
and also for piles and leprosy.

Chank-shell in the form of powder is also stated to be taken internally in South
Arcot, Salem, Madura and Tinnevelly, either in water or mixed with ghee, as a specific
for skin eruptions, asthma, coughs, and also to cool the system. In Salem and also in
Ceylon it is used as a remedy for consumption. Both in Tanjore and Salem mixed with
milk or water it is also employed as a salve or lotion applied to pimples and boils. In
Malabar and South Canara, I am told, it is used in the case of rickets (grahani), chank
ring powder ground in water being rubbed on the breast. At Tanur a street quack
told me he used chank-shell powder internally as a remedy in cases of varchcha (gonorrho3a,
I believe).

Among the Tuticorin Para was a mixture of camphor and chank powder is commonly
used to relieve soreness of the eyes. A small piece of camphor is partially burned and
then ground down in a small quantity of human milk upon a flat stone by means of a
small well-cleaned chank-shell ; a small amount of powder from the shell is thus incor-
porated with this peculiar ointment ; sometimes the white of an egg is substituted
for human milk. The ointment thus made is applied round the eyelids ; it is reputed
to effect a sure and speedy cure.

Pounded chank-shell is also given internally by native practitioners in Trichinopoly,
Salem and Coimbatore to those who suffer from an acute form of dyspepsia called
kunman. It is administered after each meal — a treatment perfectly rational as the
carbonate of lime of which the shell is composed is well adapted to counteract
hyperacidity of the gastric fluids.

In Gujarat and Kathiawar chank powder is prescribed as a specific in the following
diseases : — Jaundice, phthisis, coughs, shooting pain in the side, general debility and,
very commonly, in affections of the eyes.

With regard to the practice of prescribing it in the case of asthma, cough and con-
sumption, a medical friend points out that while of no value in asthma, this treatment
has reason for its employment in phthisical cases — the introduction of quantities of lime
into the system facilitating the deposit of lime salts around tubercular centres, encap-
sulating them and rendering them innocuous.

In rickets the use of lime taken internally is also indicated emphatically, the
disease being characterised by an insufficient deposit of lime in the bones. He also
points out that in the case of haemorrhoids, the use of lime administered internally may
assist a cure by increasing the coagulative property of the patient's blood.

It appears therefore that the employment of chank-shell powder by native prac-
titioners is not without reason in regard to certain diseases, and while it may be objected
that a non-organic form of carbonate of lime should prove equally beneficial, it has to
be remembered that the carbonate of lime of shells is laid down within a delicate frame-
work of animal membrane, and this minutely divided form may possibly render it more

HORNELL— THE INDIAN CONCH 47

easy of assimilation in the body and therefore more efficacious. The religious associa-
tions surrounding the chank have also their value in inspiring the confidence of patients
in the value of this medicine, faith that may help largely towards a cure.

The wearing of chank rings, the rubbing of the affected parts with them and the
laving of them with water which has been in contact with these rings, are forms of
treatment on a different footing. They are to be considered purely as charms, without
direct therapeutic value. They bear the same relation to the internal employment
of powdered shell as does the quack exploitation of electricity by means of belts and
bands containing discs of metal to the legitimate use of current electricity in the hands
of qualified medical practitioners. If the former have any value it is by reason of
faith alone.

The egg-capsule of the chank is employed by the chank and pearl divers of Tuticorin
to relieve headache. They grind up a portion of the egg-capsule (sanku-pu or " chank-
flower ") in gingelly-oil, together with pepper and coriander seed, and apply the paste
to the forehead and temples.

Finally, according to Risley (II, p. 223) the shell-workers of Dacca are accustomed
to extract the dried remnant of the visceral coil (called pitta) from the shells they receive
and to sell this to native physicians as a medicine for spleen enlargement. He also
states that the dust produced in sawing the shells is employed to prevent the pitting
of small-pox and as an ingredient of a valuable white paint.

(n) FOOD.

During the run home from the chank beds, the divers are accustomed to extract
the foot and anterior part of the body of the chank from the shells they have collected.
The work is roughly performed by means of a pointed iron rod and all the apical mass,
comprising the hepatic and reproductive glands, remains within the shell. What is
extracted consists a1 most entirely of tough muscular tissue carrying the adherent horny
operculum at one end. These fragments are collected in the little palmyra-leaf baskets
used for bailing water out of the canoe. The flesh, called chanku-chathai, is carried
home and there prepared for family use. The preparation consists of separating the
operculum, boiling the flesh for a short time and then cutting the foot and head region
transversely into thin slices. These are dried in the sun ; when required for use they
are fried in oil and eaten with rice and curry stuffs. On one occasion I essayed to try
this much-esteemed food, but my taste was not sufficiently cultivated ; the fried slices
tasted or rather smelled like frizzled shoe-leather and were altogether too tough for
my teeth.

(o) INCENSE STICKS.

The horny operculum is also put to use. It is dried, reduced to powder, and then
employed, after soaking in water, as an adhesive matrix to bind together the powdered

48 OKHAMANDAL MARINE ZOOLOGY— PART II

sandal-wood and other sweet-smelling incense material used in coating the incense sticks
burned before shrines, or used in native shops. In Tamil these sticks go by the names
of uthupaththi or sambiranik-kuchchi ; the best quality sells at the rate of 1 or 2 for
3 pies, while inferior sorts retail at from J to 1 pie each. The operculum itself is called
naganam or navanam ; the usual rate is 2 annas per palam ( = 8 tolas or just over 3 oz.).

(p) ASSEMBLY CALLS, ETC.

A call on a chank-shell is frequently employed upon native-owned plantations in
South India and Ceylon to summon the workpeople to their duties : there can be no
doubt that these long drawn out and penetrating booming calls are particularly well
adapted to this purpose.

In the Laccadive Islands all the inhabitants are required under penalty to attend
the call of the chank, sounded in cases of emergency and public requirement. Among
these are counted the beaching of boats and the inauguration of rat hunts.

To conclude this account of the miscellaneous uses to which the chank is put and of
which the foregoing summary has by no means exhausted the list, the following
instance of the ingenuity of the Indian countryman may not be amiss. For it I am
indebted to Mr. C. A. Innes, I.C.S. Apropos of a flight of winged termites, he told me
that once when travelling in the Madura district, he chanced upon a low caste man
engaged upon some mysterious work on a large termite anthill : the man had a chank-
shell in his hand. When asked what he was doing, he replied, " I am catching white
ants to eat," and gave a blast upon the chank at one of the major openings into the hill.
Hardly had he finished ere crowds of ants sallied forth from other openings, and these
the man scooped up in handfuls and ate without any preparation.

IV.
THE USE OF CHANK BANGLES.

(1) IN NORTHERN INDIA.

(2) IN THE MADRAS PRESIDENCY.

(1) In Northern India. — Although evidence is strong in favour of the belief that
the custom of wearing chank bracelets was in old times prevalent throughout the length
and breadth of India, more especially in the Tamil country, in the Deccan, in Kathiawar,
Gujarat, and Bengal, at the present day only in Bengal, the adjacent hill regions to the
west, north, and east and in a few Tamil-speaking districts in the extreme south of
India, does the custom continue to be widely observed and of notable social importance.

In Bengal and wherever in the adjoining Provinces of Assam, Behar and Orissa
there are colonies of the Bengali race, every married woman of all castes which are
thoroughly Hinduised is bound to possess a pair of chank bangles lacquered in vermilion
as one of the visible tokens of her married state ; the red sankha, or shakha as it is called
in Dacca, is indeed as necessary of assumption during the marriage ceremonies as is the
performance of that other Hindu custom of smearing a streak of vermilion on the fore-
head or down the parting of the bride's hair or as the wedding ring of English women.
Garcia da Orta's curious statement quoted on page 73 is to be explained in the light of
this custom ; his informants doubtless meant to convey no more than that among
the better classes an essential part of the marriage ceremony consisted in placing chank
bracelets on the arms of the bride. The women of- castes holding good social status
appear, however, to have no great liking for the custom, particularly if their husbands
be well-to-do, and I was informed that they frequently lay them aside temporarily
in favour either of more handsomely carved ones or replace them when means permit
by gold and jewelled ones. Chank bangles are occasionally ornamented with gold and
set with jewels ; the price of these may reach several hundreds of rupees. The great
majority of married women, however, wear them permanently, never removing them
so long as their husbands are alive. Occasionally some of the modern sankha (marriage)
bangles are made in two sections secured together after the bangle is placed on the wrist
by means of tiny bamboo pins as it is otherwise impossible to pass one of the right size
over the hand without great difficulty and the infliction of acute pain.

49 D

50

OKHAMANDAL MARINE ZOOLOGY— PART II

In spite of the rapid spread of a desire for bracelets of more showy appearance,
there are very large numbers of prosperous Hindu households, especially in the country
districts, where the womenfolk remain attached to the old and less ostentatious custom
of wearing chank bangles solely as ornaments. Among these conservative folk a large
demand exists for the handsome products of the sankhari workshops. The ornamental
bangles made to meet these requirements are of two kinds called respectively bdla and
clinri. The former are broad bangles worn one on each wrist. The chfiri on the con-

TEXT-FIOURE 1.

im iniiiiiiiiimnmiiu

Edge and side views of a Bengal marriage bangle to show ornamentation
in yellow upon a red-lacquered ground. From Pabna, Bengal. X 2.

trary is always quite narrow, generally | to ^ inch in width, and usually of conventional
scroll design, worn in a set of three on each wrist.

The use of these ornamental bangles (bdla and churi) and also of the red marriage
bangle is limited almost entirely to the thoroughly Hinduised sections of the Bengali
people, together with the Hindu communities settled in Assam, Behar and Orissa.
Baishnab women, however, do not wear these bangles according to the Collector of
Birbhum. Information received from a Muhammadan source indicates that women of
the lower classes of this community in Dacca, Darjeeling and Assam occasionally wear
chank bangles as wrist ornaments.

As elsewhere in India, it is the invariable custom in Bengal in orthodox Hindu
households for widows to discard all their jewellery on the death of their husbands.
In the case of chank and glass bangles, it is usual for the widow to break and throw them
away on the first occasion when she bathes after her husband's death. They never

HORNELL— THE INDIAN CONCH 51

resume the use of similar bangles except in the very rare cases where re-marriage is
permitted to widows. Tavernier says l " when a man dies, all his relatives and friends
should come to the interment and when they place the body in the ground they take
off all the bracelets which are on their arms and legs and bury them with the defunct."
This burial of the widow's bangles with the dead may still be continued by some castes,
but as earth-burial is now rapidly being displaced by cremation as orthodox Hinduism
secures a firmer hold on the people, this custom must tend to die out. Generally in Bengal
the Hindu women wear sankhas as visible tokens of the possession of living husbands.
The Hindu Shastras are said to enjoin their use as it is believed that this contributes
to the prosperity and longevity of their husbands.

Surti, Tuticorin and Rameswaram chanks are necessary in the manufacture of
both bdla and cMri bangles as these require to be made from the finest quality of shells
—those possessing a pure white porcellaneous appearance and a dense well-conditioned
substance susceptible of high polish.

Among Bengal castes of inferior social status, particularly those whose physical
characteristics bespeak Dravidian descent and whose customs are not yet thoroughly
Hinduised the use of chank bangles made up into massive gauntlets composed of
numerous separate bangles is very prevalent. Prominent among these are the widely
spread Kochh tribe in their two principal sub-divisions of Rajbansi and Paliya. It is
largely to supply the women of this tribe with their characteristic ornaments that the
chank bangle workshops in Dinajpur and Rangpur exist, as in these districts the tribe
has its chief settlement with an approximate total of one million individuals. Kuch
Behar and Jalpaiguri account for another half million, while considerable numbers are
found also in Purnia, Maldah, the Darjeeling Terai, Bogra, Murshidabad, Nadiya and
Dacca. The Rajbansi and Paliya gauntlets are composed usually of ten separate
bangles. As the wearers belong largely to the labouring and agricultural classes the
bangles forming these gauntlets are broad and thick, frequently without any ornament
whatever ; where decoration is attempted, it consists of simple line patterns made of
shallow groovings which impair very little the strength of the bangle and yet are very
effective and elegant (Text-figure 2). Neither are they usually polished, hence dead
shells from Jaffna are largely employed in this manufacture, although inferior shells
of the better qualities from the Indian side are also extensively made use of.

The Muchi is another important Bengal caste where the wearing of numerous
chank bangles is a distinctive custom among the women. This is a leather-dressing and
cobbler caste, socially a shade higher than the allied Chamars from whom the Muchis
appear to be an offshoot. One of the obvious distinctions between the women of these
castes lies in the character of the bracelets worn. Thus while the female Chamar prides
herself on huge bracelets of bell metal adorning her arms, the Muchi woman always
substitutes chank bangles. The Muchis, like the Paraiyar of the South, are largely

1 Loc. dt., Vol. II, p. 285.

D 2

52 OKHAMANDAL MARINE ZOOLOGY— PART II

the caste drummers of the province and as they are fond of the violin and the pipe are
usually employed as musicians at Hindu weddings.

In Western Bengal and in Behar the Santals take the place occupied by the
Rajbansis and Paliyas in North-Eastern and Eastern Bengal as the chief chank bangle
wearing tribe. Many of their women follow the same habit of disposing of a number of
chank bangles, three to five usually, as a massive cuff-like gauntlet or compound bracelet.
These people being generally poor, the quality employed for these compound bracelets
is inferior and red and yellow lac are freely used upon them to enhance their appearance
and to disguise imperfections. Many indeed are too poor to afford these ornaments
and others belong to families which do not observe the custom ; in Birbhum which

TKXT-FIGURE 2.

Gauntlet pattern of compound bracelet worn by Paliya women, Bengal.

may be taken as a characteristic Santal district, it is estimated that about half the female
Santal population follow this custom. Sometimes Santal girls wear them from an early
age but generally they are assumed at marriage. It has no religious significance and
marriage may be performed without the putting on of these bangles which are worn
more for ornament and because of custom than for any more serious reason. Like the
Hindus, Santal women break and throw away their bangles on the occasion of widowhood,
re-assuming others, if they wish, if they remarry. Dead shells are often employed by
the cutters for Santal bangles.

Eisley states (II, p. 225) that the Santals in point of physical characteristics may be
regarded as typical examples of the pure Dravidian stock and in view of the similar

HORNELL— THE INDIAN CONCH 53

origin attributed to the Kochh tribe which includes both the Rajbansi and the Paliya,
this becomes a matter of great significance as well as of much difficulty, for whereas
the Kochh people are professed Hindus, the Santfils hold the animistic beliefs
characteristic of Non-Hinduised Dravidians. However Oldham, as quoted by
Risley (I, p. 492), states that " the adhesion of the Kochh tribe to Hinduism is
comparatively recent as shown by their own customs as regards burial, food and
marriage."

The section of the Kurmi caste found in Chota Nagpore and Orissa also wear chank
bangles. In view of what has been said above in regard to the Dravidian origin of the
Kochhs and Santals, it is of importance to find that Risley (I, p. 530) considers this
territorial section of the caste as undoubtedly Dravidian, as shown by their physical
characteristics, religious beliefs and social customs. In appearance, he says that in
Munbhum and the north of Orissa, it is difficult to distinguish a Kurmi from a Bhumij
or a Santal. In their religion the animistic beliefs characteristic of the Dravidian
races are overlaid by the thinnest veneer of conventional Hinduism, and the
vague shapes of ghosts and demons who haunt the jungles and the rocks are the
real powers to whom the Kurmi looks for the ordering of his moral and physical
welfare.

Alike with the Santals the internal structure of that branch of the Kurmi caste
living in Chota Nagpur and Orissa is founded upon a distinct and well-defined totemism
in which a large proportion of the totems are still capable of being identified. Risley
(II, appendix, p. 88) enumerates 60 totemistic sections or septs in this caste, among
which is one termed Sankhawfir whose members are prohibited from wearing chank
shell ornaments. Among the Santals, the place of this sept is taken by one called
Sankh, wherein all individuals are forbidden, under pain of caste punishment, the use
of the chank shell in any form ; they may neither cut, burn, nor use the shell, nor may
the women of this sept use it in personal adornment (I, p. xliii).

The prevalence of the use of chank bangles among these Dravidian races, the
present animistic beliefs of the Santills and Chota Nagpur Kurmis, and the comparatively
recent renunciation of the same cult by the great Kochh tribe, taken in conjunction
with other facts and especially with the widely spread archaeological finds detailed
elsewhere in these pages, point to the use of chank bangles as having had a purely
Dravidian origin and as having been a custom prevalent and solidly established among
at least certain sections of the race throughout India anterior to the advent of the Aryan
invaders and the rise of the Brahmanic faith. The cult of the chank would therefore
appear to be one adopted (and modified) by the Brahmans from the religious belief
which they found indigenous to India.

Finally, in the hill tracts of Chittagong, we find the women of the Maghs, a race
of Indo-Mongolian extraction and Buddhists by religion, using very broad unornamented
sections of chank shells as bracelets in similar manner as we shall next see is the habit

54 OKHAMANDAL MARINE ZOOLOGY— PART II

in Thibet and Bhutan, inhabited by other Mongolian races. To supply the needs of the
Maghs, bangle cutters are established in Chittagong ; these work-people are chiefly
Muhammadans and the work they do is of the roughest and crudest description in
conformity with the undeveloped artistic taste of their customers who appear to wear
these bracelets rather as amulets than as ornaments. Broad arm ornaments of similar
simple form are used by the Papuans and by the wild inhabitants of several groups
of the Melanesian islands ; sometimes round the wrist, sometimes' on the upper arm
above the elbow. I do not know, however, whether the shell employed in these instances
be Turbinella or not. Among these island tribes it is the men who wear these
ornaments.

Outside of Bengal and Assam the only considerable demand for chank bracelets
comes from Thibet and Bhutan. The trade is one of long standing, for Tavernier, in
1666, found Bhutanese merchants taking home from Pabna and Dacca bracelets sawn
from " sea-shells as large as an egg." He also states that 2,000 men were occupied
in these two places in making tortoise-shell and sea-shell bracelets and " all that is
produced by them is exported a to the kingdoms of Bhutan, Assam, Siam and other
countries to the north and east of the territories of the great Moghul " (loc. cit.,
p. 267).

Chank bangles appear to be worn very generally throughout Thibet, from Ladakh
in the west to the Kham country in the east. Neve records 2 seeing the poorer women
in Kashmiri Thibet wearing broad shell-bangles in shape like a cuff on both wrists, while
on the march of the British expedition to Lhassa in 1904 they were noted as in frequent
use by Thibetan women. This ornament is assumed early in life while the hand is
still small and pliable ; after a few years it becomes impossible to remove it without
breakage which these women will suffer only in the last resort, as it cannot be replaced
except by one of large diameter which will fit more loosely on the arm than they like.
A medical officer with the Thibet mission has informed me that in one instance a
Thibetan woman was brought to him for the treatment of a festering wound on the
wrist. On examination the cause of the trouble was found to be the presence of a
chank bangle so small that the wrist had been wounded and circulation impeded ;
gangrene was imminent and although the woman was loth to part with her bangle
it had to be filed off to save the hand.

The export of round and square discs of chank shell to the Buddhist countries of
the north appears to be much less than in Tavernier's time, as it is now comparatively
insignificant. Among the Njlgas, the discs are employed to ornament the men's hair-
bedecked helmets, and Prince Henri d'Orleans 3 found the women of the wild Lissus,

1 Evidently a lapsus calami as the custom of wearing chank bangles was even more prevalent in
Tavernier's day among Bengali women than it is to-day, ride Orta, loc. cit.

2 "Beyond the Pir Panjal," London, 1912.

:; " From Tonkin to India," English Translation, London, 1898.

HORNELL— THE INDIAN CONCH 55

a section of the Lolo tribe, mountaineers living in the upper valley of the Mekong in
Yunnan, employing chank shell discs to ornament their Chinese caps. It may be
that these Lissus and cognate tribes represent those chank jewel wearers whom
Tavernier refers to as belonging to the kingdom of Siam. In this latter country at the
present day I know of no utilisation of chanks in personal adornment.

The chank is one of the eight lucky signs recognised by Buddhists of the Northern
cult and as such is constantly reproduced in Buddhist ornamentation in Thibet and
Bhutan.1 It may therefore be inferred that the use of it in personal adornment has
a like reason ; whether in the form of a bangle, a cap or a hair ornament, a necklace
or a breast disc, it is employed as a talisman to ensure good fortune, and possibly even
as an amulet against the evil eye, as is the chank shell placed on the forehead of draft
bulls in Southern India.

(2) The Tribes and Castes which wear Chank Bangles in the South of India.

In the Madras Presidency and the associated native States, the castes whose
women systematically wear chank bangles are few, and if we except the wandering
tribes of the Lambadis (or Brinjaris), Koravars and Kurivikkarans, the custom appears
confined to a sub-division of each caste or tribe. Whether it had a totemistic origin
and significance as it still has among non-Hinduised tribes in Bengal, Behar and Chota
Nagpur is not at present clear. If it had, the original tribal sept, usually exogamous,
has become changed to a caste sub-division, invariably endogamous. And whereas
among the septs of those animistic tribes in Northern India which are named after the
chank this shell is taboo with them, it forms the characteristic ornament of the women
of the caste sub -divisions named after it, in Southern India.

Only in the Kongu country, which coincides roughly with the present inland districts
of Coimbatore and Salem, does the custom continue to flourish at all strongly.
Coimbatore is the great centre of the custom, for there the numbers of chank-bangle
wearers greatly exceed those found in any other district. The Collector reports eleven
castes and sub-divisions as following this custom, viz. : —

(1) Pala Vellalas.

(2) Puluvans (so-called Puluva Vellalas).

(3) Konga and Golla Idaiyans.

(4) Konga Shanans.

(5) Konga Vannans.

(6) Thotti Chukkiliyans.

(7) Sangu, Konga, Sangudu or Sanguvalai Paraiyans.

(8) Thottiya Naiks.

(9) Okkilians (not universal).

(10) Kurumbars.

(11) Lambadis in parts of Kollegal and Gobichettipalaiyam divisions.

1 J. Claude White, " Sikhim and Bhutan," p. 46. London, 1909.

56 OKHAMANDAL MARINE ZOOLOGY— PART II

The custom is associated particularly with those caste sub-divisions whose
territorial cognomen indicates a long settled residence in the Kongu country (Coimbatore
and Salem). The Konga sub-divisions of the Idaiyar, Paraiyar, Vannar and Shanar
castes have this custom in common and as several other castes in Coimbatore also
adhere to it, we may infer that at one time the custom was general in the Kongu country
among at least the generality of the lower castes.

Another caste sub-division where the women wear chank bangles is that of the
Sangukatti Idaiyans. Among them the marriage ceremony requires, as in Bengal,
the placing of a chank bangle on one of the bride's wrists.

Among Coimbatore castes the chank bangle is worn always upon the left wrist,
usually singly but occasionally a pair in certain cases, e.g., among the Konga Paraiyans.

The wearing of these bangles is considered as a symbol of the permanence of the
marriage tie, a belief probably derived from the custom of breaking the bangle after the
death of the husband. The Collector of Coimbatore states that a widow discards her
bangle one month after her husband's death. He adds that if a woman accidentally
breaks her bangle, she thinks it unlucky and regards it as an omen that her husband
will chance on some evil ; when the husband is sick the wife prays that it may be her
good fortune to wear the bangle during her whole life. A woman considers it improper
to appear before her husband or in public without the bangle. The wearing of this
ornament appears to be followed in Coimbatore largely because it is an ancient
custom, with no further significance beyond what is implied above. It is not
now connected with belief in the evil-eye though it is said to have had this
significance in former times. The general belief in the efficacy of the chank as a
specific against skin diseases may however be counted as one of the obscure reasons for
its continued usage.

The practice is more general among low castes. The Collectors of Madura and
Trichinopoly both inform me that among Paraiyans, Chukkiliyans (leather workers, etc.),
Oddans, Koravas and the Naick sections known as Kavaraja and Thottiya Kambalathans,
together with the wandering tribe of Lambadis, the custom of wearing chank bangles
is found to prevail here and there in both districts. There appears to be no general
observance of the custom — in some villages and taluks none among the women of the
castes named wears chank bangles ; elsewhere, as in the Namakkal Taluk (Trichinopoly
District), a definite section of the Paraiyan caste called Sengudimi Paraiyans adopts this
ornament as a distinguishing sept distinction, while in other parts of the country, the
women of these various low castes wear it chiefly if not entirely for its ornamental value.
The custom appears to be dying out, as witness the vagueness of the people who still
adhere to its observance as to the reason for so doing, its partial and sporadic geographical
distribution in the districts where it lingers, and the comparatively small numbers who
adhere to it. As a typical instance of the irregular distribution of the custom, the
report of the Collector of Trichinopoly states that in the Musiri and Karur Taluks, no

HORNELL— THE INDIAN CONCH 57

people wear chank-bangles ; in Namakkal Taluk only the Paraiyan sept call
Sengudimi Paraiyans wear them ; in Udaiyarpalaiyam Taluk the habit is confined
to Koravans and Lambadis ; in Trichinopoly and Perambalur Taluks to Chukkiliyans
and Paraiyans, while in Kulittalai Taluk besides the two castes just named, the Oddans,
Koravans and Thottia Naickers are given as castes following the custom — it is to be noted
that all are amongst the lowest of castes, constituting what may be termed the inferior
labouring population. The bangles employed are manufactured in Kilakarai and pass
to Trichinopoly by the intermediary of traders in Madura. The price in Trichinopoly
town is from Rs. 3 to 4 per 100 bangles, all very roughly made, with little or no
ornamentation.

In Tanjore and Salem the Paraiyans and Chukkiliyans use chank-bangles in certain
villages and these Salem Pariayans are said to belong to the Konga division as in
Coimbatore. In both districts the Lambadis and Koravas have the same custom,
and the Collector of Tanjore adds that the women of the Uppu Koravar, Panni Koravar,
and Vari Koravar sub-divisions, together with the Oddar and Domba castes, all agree
in following the habit. He states that the Koravas put on the bangles during the
marriage ceremonies.

In Salem the Malayali women of the Chitteri Hills, Uttangarai Taluk, also use
chank-bangles .

Hawkers called Dasam Chettis, who bring their wares from Rameswaram and
Kilakarai, attend village fairs and temple festivals with these bangles in Salem and
Coimbatore, charging from 4 to 12 annas a pair. In Coimbatore well-to-do Puluva
Vellalans not infrequently wear bangles of superior quality costing even Rs. 3 to 5
per pair ; these probably are brought from Calcutta, as no expensive patterns are made
at Kilakarai. In Tanjore about the same prices prevail, but here the most expensive,
said to come from Calcutta, are said to cost one rupee per pair ; common qualities as
usual come from the Ramnad coast.

In the Nilgiri hills, especially in Ootacamund, Konga Paraiyan women who have
come from Coimbatore are often to be met with wearing chank-bangles. Their sub-
division is considered one of superior standing and the people belonging to it will not
act as sweepers. The bangle is always worn on the left wrist ; frequently two are
worn, always plain and massive, and about f inch in width, exactly similar indeed to
those worn by Chanku (Puluva ?) Vellalas.

The women of the Kota hill-tribe in the Nilgiri hills have an allied custom, but
instead of a massive bangle cut from the entire shell, they wear around the left wrist a
bracelet of roughly-made chank beads strung on a thread. When chank beads are
not obtainable they wear a string of white glass beads as near the colour and shape of
the usual chank beads as they can obtain. In answer to my questions, the elders of a
Kota village situated close to Kotagiri informed me that all married women must possess
and wear one of these bracelets on the left wrist together with two massive copper

58 OKHAMANDAL MARINE ZOOLOGY— PART II

bangles on the right wrist during their husband's life. They are assumed, however, before
marriage, and the putting on of the chank bracelet is not a part of the marriage
ceremonies. On the death of the husband the widow discards this with her other orna-
ments, but is permitted to resume them after a decent interval of mourning — three
months according to my informant. It is to be noted that the wife's ornaments are not
broken or destroyed at the husband's death as is the custom among the chank-wearing
Hinduised plains-people. At the woman's own death they are put to burn with her body
on the funeral pyre. The Kotas can adduce no special reason for the wearing of these
bracelets, except that their god Bhagawani long ago ordered that their women should
do so.

From the foregoing it is seen that chank-bangle wearing is confined to the Tamil
districts in the south of the Madras Presidency. It is unknown in the central and
northern sections — the Madras Deccan and the coastal Telugu districts — except in respect
of the wandering Lambadis ; the Collector of Kurnul informs me that chank-bangles
for sale to the local representatives of this tribe are occasionally brought from the
Raichur side (Hyderabad State), a significant fact as the Raichur Doab is one of the
localities where Mr. Bruce-Foote found numerous fragments of chank factory waste,
indicative of the former existence of a bangle-making industry in the vicinity of his
discoveries. Where these modern Raichur bangles come from I do not know, but I
should expect them to be of Bengal manufacture.

(1) THE ANTIQUITY OF THE CHANK-BANGLE INDUSTRY.
(«) IN THE TINNEVELLY DISTRICT.

Reference to ancient Tamil classics furnishes evidence scanty but conclusive of the
existence of an important chank-cutting industry in the ancient Pandyan kingdom
in the early centuries of the Christian era. Similar evidence is also extant of a wide-
spread use of carved and ornamented chank bangles in former days by the women
of the Pandyan country which may be considered as roughly co-extensive with the
modern districts of Tinnevelly, Madura, and Ramnad, forming the eastern section of the
extreme south of the Madras Presidency.

Among the more important references which prove the ancient importance of this
industry on the Indian shore of the Gulf of Mannar, is one contained in the
" Maduraikkanchi," a Tamil poem which incidentally describes the ancient city of
Korkai, once the sub-capital of the Pandyan kingdom and the great emporium familiar to
Greek and Egyptian sailors and traders and described by the geographers of the 1st and
2nd centuries A.D. under the name of Kolkhoi. From the purity of the Tamil employed

HORNELL— THE INDIAN CONCH 59

in this poem and the similarity of the names of the towns, ports and goods mentioned
incidentally with those employed by Ptolemy and the author of the " Periplus of the
Erythraean Sea," we may date it as approximately contemporaneous with the writings
of these authors and certainly not later than the 2nd or 3rd century A.D.

In one passage (LL. 140-144) the Parawas are described as men who dived for pearl
oysters and for chank shells and knew charms to keep sharks away from that part of
the sea where diving was being carried on. Another passage depicts the city of Kprkai,
then a seaport at the mouth of the Tambraparni, as the chief town in the country of
the Parawas and the seat of the pearl fishery, with a population consisting chiefly of
pearl-divers and chank-cutters. The great epic, the Silapjxithikkarram, or " Lay of the
Anklet," written about the same period by a Jain poet, gives further information about
Korkai, from which we gather that on account of the great value of the revenue derived
from the pearl fishery, this city was a sub-capital of the Pandyan realm and the usual
residence of the heir-apparent, boasting great magnificence and adorned with temples
and palaces befitting its wealth and importance.

Another valuable reference to the chank trade is contained in two Tamil stanzas
which chronicle a passage at arms between a Brahman and Nakkirar, the celebrated
poet-president of the Madura Sangam in the reign of the Pandyan king Nedunj Cheliyan
II, who flourished probably about the beginning of the 2nd century A.D.

The Brahman, named Dharmi, presented to the Sangam a poem purporting to
be composed with the aid of Siva. Nakkirar, the President, in spite of its alleged
divine origin criticised the poem mercilessly, and rejected it as unworthy of literary
recognition. The Brahman took revenge by presenting another poem also purporting
to be inspired by Siva ; in it he held the President up to ridicule on account of his
caste trade in pungent lines which may be translated literally as follows :—

IS

Is Kiran fit to criticise my poem ? Spreading his knees wide, his joints
loosened (by the labour) does he not saw chanks into sections, his ghee-smeared saw
murmuring the while kir — kir ? "

Besides the insult intended to be given, the verse contains a play on the President's
name and the sound given out during the sawing of chank shells.

The reply of Nakkirar was " Chank-cutting is indeed the calling of my caste ;
of that I am not ashamed. But of what caste is Sankara ? (one of the many names
of Siva). We earn our livelihood by cutting chanks ; we do not live by begging as he
did " —an allusion to the fable popularised by the Brahmans wherein Siva is represented
as a mendicant seeking alms with a skull in his hand as begging bowl.

Dharmi's description of a chank-cutter's trade is wonderfully vivid in the original
Tamil ; in a dozen words he paints a realistic word-picture of a cutter's workshop — the
men seated on the ground with the knees widely spread and depressed outwards almost
to the ground to give free play to the great crescentic two-handled saw monotonously

60 OKHAMANDAL MARINE ZOOLOGY— PART II

droning a single note as it cuts its way laboriously through the hard substance of the
shell.

Tradition has it that Nakkirar, the chank-cutter President of the Sangam, was a
Parawa by caste. It would be most appropriate if this be correct as we have already
seen that at the beginning of the Christian era chank-fishing and chank-cutting were
among the important trades carried on in Korkai, the chief settlement of the Parawas
in early days.

No Parawas to-day are engaged in chank-cutting although they still largely
monopolise the shore industries of Tinnevelly where they continue as from time imme-
morial to provide the contingent of divers required for the exploitation of both the
pearl- and the chank-fisheries of the Gulf of Mannar.

It is noteworthy that though their women do not now wear chank-bangles their
children from four months to about two years old are often given roughly-made chank-
bracelets to wear in the belief that such will protect them against the baleful influence
of the evil eye, from vomiting and from a wasting disease called chedi which appears
to be rickets and reputed to be caused by the touch or near approach of a woman during
her menses ! This custom has now been abandoned or is perfunctorily performed
by some of the better class Parawas, but the great majority, including naturally the
whole of the poorer and more ignorant sections of the community, continue to
adhere strongly to the custom. The bangles are roughly fashioned and with the crudest
of ornamentation ; they are made by Muhammadans at Kilakarai, their chief settlement
on the coast of the Gulf of Mannar.

Apart from this evidence we have nothing of importance till we come to the sixth
century when the travelled monk, Cosmas Indico-Pleustes, after mentioning the island
of Ceylon, proceeds to say " and then again on the continent and further back is Marallo
which exports conch shells (/co^Aiou?)." Sir J. Emerson Tennent in his " Sketches
of the Natural History of Ceylon " (London, 1861) misses the significance of the
expression " on the continent " and identifies Marallo with Mantotte near Mannar on
the north-west coast of Ceylon, where chanks are collected in the neighbourhood in
large quantities even at the present day. Yule l with closer adherence to the old text
would place this ancient chank-fishery on the Indian coast (i.e., on the continent opposite
Ceylon), and he suggests that Marallo is a corrupted form or misrendering of Marawar,
the name of the chief caste living in the coastal district of Ramnad, now the location
of one of the most productive and accessible present day chank-fisheries. The name
of the local people not infrequently was applied by old travellers to the chief town in
their territory and so, very reasonably, we may identify Marallo generally with the
Maravan coast and particularly with either the town of Rameswaram or of Pamban
situated at the western extremity of Adam's Bridge and directly opposite to Mantotte
and Mannar at the western extremity.

1 Cathay and the Way Thither, Vol. I, p. clxxviii, London, 1866.

HORNELL— THE INDIAN CONCH 61

Ma'bar or Maabar, the Arab name for the western coast of the Pandyan country,
lias probably a parallel derivation, Maabar being indeed a very fair rendering by guttural
Arab lips of the Tamil term Maravar.

The next writer to mention the chank is the Arab Abouzeyd, who in 851 A.D.
stated that " they find on the shores of Ceylon the pearl and the shank, which serves
for a trumpet and which is much sought after." l

A long gap occurs in references by travellers to chank-fisheries till the days of the
Portuguese and Dutch when they became fairly frequent. A few years before the
establishment of the former power in the Gulf of Mannar, the traveller Barbosa visited
the old town of Kayal, and from him we learn that it was then still an important seaport
where many ships from Malabar, Coromandel and Bengal resorted every year to trade
with the rich Hindu and Muhammadan merchants living there, a definite statement
which shows that there was even then no difficulty in forwarding supplies of chanks
direct by sea to the Dacca workshops.

Barbosa also tells us that at the time of his visit the fishery off this coast belonged
to the king of Koulam (Quilon in the southern part of Travancore) who generally resided
at Kayal and who farmed the pearl-fishery to a wealthy Muhammadan.2 The chank -
fishery so far as we know has always been an adjunct to the more romantic pearl-fishery
and must almost certainly have been treated in a similar manner, both fisheries being
considered everywhere in India from immemorial times as prerogatives of the sovereign.
About 1524, the Portuguese seized the Tinnevelly pearl-fishery, stationing a factor and
guard boats on this coast — the Pescaria or fishery coast as it soon became termed. In
1563, Garcia da Orta speaks of the trade with Bengal having declined owing to the unrest
caused by Muhammadan invaders in that country, but in 1644, Boccaro in a detailed
report upon the Portuguese ports and settlements in India records that a large quantity
of chanks fished off Tuticorin were exported to Bengal " where the blacks make of them
bracelets for the arm." He adds rather quaintly the name of another Tuticorin
production — " the biggest and best fowls in all these eastern parts." 3 Exactly how the
Portuguese conducted this trade and what profits it yielded them are not known to me ;
the Dutch, so far as they were able, destroyed the Portuguese archives in Tuticorin
as well as in Ceylon, and we must await further research among the records at Lisbon
before we can gain any further information.

The Dutch, keen to distinguish the substance from the shadow, paid great attention
to the development of the chank -fishery as distinguished from the pearl-fishery whereof
one of their most able local Governors, Baron Van Imhoff, once queried (1740) whether
the latter " is not more glitter than gold as so many things are which belong to the
Company, which shine uncommonly but have no real substance." 4

1 Fide Yule's " Hobson-Jobson." 2 Fide Yule's " Hobson-Jobson," article "Chank."

3 Fide Yule, "The Book of Ser Marco Polo," Vol. II, p. 307, London, 1871.

4 Hornell, " Report to the Government of Madras on the Indian Pearl Fisheries in the Gulf of
Mannar," Madras, 1905.

62 OKHAMANDAL MARINE ZOOLOGY— PART II

In 1700, Father Martin, a Jesuit Missionary, wrote (Lettres Edifiantes, II, p. 278,
edition of 1843), " It is scarcely credible how jealous the Dutch are of this commerce.
It is death to a native to sell them to any one but the servants of the Company. The
shells are bought by the Dutch for a trifle, and then despatched to Bengal, where they
are sold at great profit. These shells, which are round and hollow, are sawn and
fashioned into bracelets equalling the most brilliant ivory in lustre. Those fished on
this coast (Tinnevelly) are extraordinarily abundant ; they have their spiral from right
to left, but if one be found twisted in the other direction, it is a treasure valued by
Hindus at an extravagant price, for they believe that it was in a chank of this
description that one of their gods hid himself in order to escape the fury of enemies
pursuing him in the sea."

With the transfer to the British of all Dutch ports on the Coromandel coast and in
Ceylon together with the acquirement of the Tanjore and Carnatic territories about the
same time, the control of all the chank-fisheries in these localities passed to the
British.

The evidence furnished by the Tamil classics of the existence of an extensive chank-
bangle industry in the extreme south of India during the height of ancient Tamil
civilisation 1,200 to 2,000 years ago has received unexpectedly conclusive corroboration
within the present year (1912) through discoveries which I have made on the sites of
the once famous Tamil cities of Korkai and Kayal (now Palayakayal). These cities
are now represented by mounds of rubbish adjacent to villages still bearing the
appellation of their celebrated predecessors. The greatest find was at Korkai, which as
already noted flourished from a date well antecedent to the Christian era down to some
indeterminate date prior to 1000 A.D. when the accretion of silt at the mouth of the
Tambraparni drove the inhabitants to build another city (Kayal) at the new mouth of
the river. Here, on the landward outskirts of the village, I unearthed a fine series of
chank workshop waste — seventeen fragments in all. The whole number were found
lying on the surface of the ground in a place where old Pandyan coins have from time
to time been discovered according to information gathered in the village. The fragments
unearthed all bear distinct evidence of having been sawn by the same form of instrument,
a thin-bladed iron saw, and in the same manner as that employed in Bengal at the
present day. Eight fragments represent the obliquely cut " shoulder-piece," six consist
of the columella and part of the oral extremity of the shell and the remaining three
are fragments of the lips — all show a sawn surface, the positive sign of treatment by
skilled artisans.

At Kayal or Palayakayal (i.e., old Kayal) as it is now termed, the daughter city of
Korkai, which flourished in the days of Marco Polo and appears to have grown rich as
Korkai gradually passed away as a sea-port owing to physical changes in the delta of the
Tambraparni, I found an excellently preserved sawn shoulder-piece, with marks of
the apex having been hammered in after the present-day habit in Dacca workshops.

HORNELL— THE INDIAN CONCH 63

This was found on the surface in an open space within the present village. Time did
not allow me to prosecute a detailed search, but in my own mind the single fragment
found is conclusive evidence of the industry having once been located here. No shell
cutting of any description is carried on anywhere in this neighbourhood.

Again, at Tuticorin, I have found a sawn and hammered shoulder-piece of typical
form, hence as the three discoveries were all made at places which in turn have been the
head-quarters of the chank-fishery, I am fully convinced that at all three, chank-bangle
workshops formerly existed, to treat on the spot this product of the neighbouring sea.
Why the seat of the bangle-cutting trade became transferred or limited to Bengal is
obscure, and may never be satisfactorily elucidated ; I am, however, inclined to suggest
the hypothesis that the decay of the industry in Tinnevelly may have been consequent
upon the Muhammadan invasion. The date of the passing away of the chank-cutting
industry I am inclined to put tentatively at about the fourteenth century, a time which
marks the close of unchallenged Hindu supremacy in the south, the spoliation of the
vast riches of the Pandyan cities by the Moslem and the heyday of Arab sea-power on
this part of the Indian coast. With the depression and decay entailed by the loot and
ruin of their enormously wealthy temples and long prosperous cities by the invaders
under Malik Kafur and his lieutenants it is far from improbable that the particular
trade here referred to became disorganised within the Pandyan realm and forced into a
different channel, the whole of the shells being exported to Bengal to be cut there instead
of being treated locally at the seat of the fishery.

It is also noteworthy that the huge funeral urns found in tumuli of the Tambraparni
valley (at Adichanallur) have yielded a few fragments of working sections cut from
chank shells, associated in the urns with beautifully formed bronze utensils, iron weapons
and implements and gold fillets. So old are these tumuli that they are classed as
prehistoric though it is obvious that the people of those days were skilful artizans in
gold, bronze, and iron and must have been contemporaries of historic periods in the
story of Egypt and Mesopotamia.

In this connection, as bearing upon the antiquity of the trade connections between
India and Persia, some interesting exhibits are to be seen in the Louvre, Paris ; the most
noteworthy is a chank-shell cup, probably used in libations, found by the Mission
Dieulafoy in the ruins of Susa. It measures 6| inches in length and has been longitu-
dinally bisected a little to one side of the median line, the larger portion being retained
to form the cup. The columella has been cut out and the result gives a very serviceable
form of spouted cup. It is classed as of Acheemenid age — say circa 500 B.C. From the
same ruins the Mission de Morgan brought back a fragment of a large chank bangle
nearly one inch wide (No. A7532) roughly ground to a triangular ridge pattern in section
and with a rough < mark engraved at one side.

Again the Fouilles de Tello yielded to the Mission de Sarzec a number of perforated
chank-shell fragments sometimes very finely engraved ; one in particular (No. 221)

64 OKHAMANDAL MARINE ZOOLOGY— PART II

bears a spirited representation of a lion leaping upon the back of a bull. These objects
were apparently ornaments worn suspended from a cord.

As there are no chank-fisheries in the Persian Gulf, these objects must have been
imported from India either from Kathiawar or from the Gulf of Mannar. The libation
cup would seem to show a much more widely spread use of the chank shell in ancient
times than we have hitherto suspected and research directed to this special point in
collections coming from Babylonian, Assyrian and ancient Persian sites should throw
further light on this subject.

(6) THE FORMER EXISTENCE OF BANGLE FACTORIES IN THE DECCAN AND IN

GUJARAT AND KATHIAWAR.

I have been unable to obtain any evidence from ancient Indian writings of the
existence elsewhere than in the extreme south of the country of any ancient custom of
wearing bangles cut from chank shells. Probably such references do exist, and if this
be so, I trust the present notes may elicit their quotation by scholars who are familiar
with the ancient Sanscrit and Gujarati classics, the most probable sources of information.
Fortunately, in this apparent absence of written records, archaeology has important
evidence to offer, and although it is difficult to date the greater portion of this testimony
with any exactitude, it offers irrefutable proof that the industry of chank-cutting and the
custom of wearing chank bangles had once much less restricted geographical range than
at the present day.

Apart from the finds in the Pandyan country, mentioned above, a considerable
amount of archaeological evidence exists proving that extensive chank-bangle factories
were located in ancient times in the Deccan and in Gujarat and Kathiawar. The principal
finds proving this were made by the latfe Mr. Bruce Foote ; they are now deposited in
the Madras Museum and form an exhibit of wonderful and unique interest.

The remains obtained in these localities consist of two distinct series ; one comprising
fragments of finished chank-bangles which appear to have been once actually worn,
and discarded owing to accidental breakage or else purposely broken as a sign of
mourning, the second, either of sawn rings of chank shells — working sections — ready for
the bangle carver and polisher or of waste material from chank-bangle cutters' workshops.
In the Southern Deccan and neighbourhood chank-bangle fragments have been
found in the following districts and States, namely :—

Mysore.

Bellary.

Anantapur.

Kurnul.

Raichur Doab.

Guntur.

Kistna.

HORNELL— THE INDIAN CONCH 65

Careful examination of the museum collection leads me to believe that shell-bangle
factories existed at four centres in the area indicated, the principal being :—

(1) Hampasagra on the Tungabhadra, Bellary District.

(2) Bastipad on the Hindri River, Kurnul District.

(3) Maski in the Raichur Doab (Hyderabad State) ;

and (4) at the great Buddhist ruins of Amaravati in Guntur District.

Of all these, the last named discovered by Mr. Rea, lately Archaeologist to Govern-
ment, is probably by far the most important, as we are better able to date the remains
here than anywhere else owing to their association with buildings of known origin.

The chank fragments discovered at Amaravati are very numerous and
important, giving evidence that the chank-bangle industry was carried on in this locality
even earlier than the construction of the Buddhist buildings. They consist in the main
of large numbers of fragments of working sections of the shell together with a very
few pieces of finished and ornamented bangles. Besides these are numerous waste
pieces — shoulder and oral rejects, showing that the methods of cutting up were identical
with those of the present day. The whole of these fragments were found beneath the
foundations of buildings which the most competent authorities date circa 200 B.C.,
hence these bangle fragments are antecedent thereto and must be over 2,100 years old.
It may be that these remains constituted part of the town's rubbish heap before the
erection of the Buddhist buildings which have survived to the present time and that
this rubbish heap was employed in making or raising the ground prior to the putting in
of the foundations, or it may be that an old village site, including the waste of a
village bangle factory, was selected as a site by the Buddhist architects of Amaravati.

Besides bangle fragments, a few rudely carved chank finger-rings figure among the
remains, together with small discs of J inch diameter sometimes perforated in the centre ;
the latter were used in the fashioning of necklaces — at Peddamudiyam in Cuddapah,
Mr. Rea has found complete necklaces formed of spherical chank beads alternating
with chank-discs of the pattern here referred to. The perfectly circular outline of these
small discs is remarkable.

The two fragments of chank bangles found in the Kistna District are probably
of approximately the same age as those from Amaravati as they also were associated
with Buddhist remains. The finds in all the other districts cannot be placed in time
in spite of their frequent association with neolithic weapons and implements. Mr. Bruce
Foote has indeed suggested that finely worked serrate and bi serrate chert and agate
flakes found in one place in association with bangle fragments were employed by neolithic
man to saw through chank-shells and to fashion bangles therefrom. This I cannot accept.

66 OKHAMANDAL MARINE ZOOLOGY— PART II

GUJARAT AND KATHIAWAR.

Mr. Bruce Foote's labours prove that the custom of using chank bangles was widely
spread and that chank-bangle factories were numerous in these two provinces in
ancient times.

The finds which he records are as follows : —

In Kathiawar : —

(a) Damnagar, Amreli Prant. In the fields (presumably upon the surface) north
of the camping tope at this town a great number of chank bangles in a fragmentary
condition were found and of these 41 pieces are represented in the Museum collection.
Three working fragments were also found at the same place, together with a couple of
cowries, and a Trochus shell ground upon three sides. Associated were such neoliths
as a basalt corncrusher, a bloodstone hammer and chert and agate cores.

(b) Babapur. At this village situated 13 miles westward of Amreli, the alluvium
of the left bank of the Shitranji river yielded a large and important series of neolithic
chert flakes, scrapers, slingstones, and cores in association with 13 fragments of finished
chank bangles, together with two working fragments and part of the columella of a
chank. Several of the flint flakes are worked upon one or both edges, and one of the
bangle fragments exhibits a chaste design executed with considerable delicacy. The
other bangles are of plain and crude design.

(c) Ambavalli. Seventy-one fragments of broken bangles from an old site at this
place are represented in the Museum collection (Nos. 3622-1 to 65 and 81 to 89). Of
these the greater number are ornamented by pattern grooving and many show an
elaboration of design as great as those now manufactured in Bengal. The designs in
many instances are precisely the same as those in vogue to-day.

Associated with these bangle fragments were numerous portions of sawn sections
of chank-shells, constituting the rough working material required by the bangle carver ;
33 fragments are shown (Nos. 3622-63 to 65 and 90 to 119).

With the exception of a few unimportant potsherds the only other object of
importance found at this site was a small iron knife with tang. No stone implements
were discovered, and no information is given as to the precise conditions under which
any of the exhibits were found ; presumably they lay on the surface of the ground
examined.

(d) Sonnaria. Fragments of two chank bangles of simple pattern apparently
found on the ground surface. A chert scraper comes from the same locality.

(e) Kodinar. On the surface of Mr. Foote's camping ground were found several
sawn portions of chank-shell, two being shoulder slices such as are found in the wastage
from a bangle workshop.

(/) Valabhipur (the modern Walah). From the ruins of the ancient city Mr. Foote

HORNELL— THE INDIAN CONCH 67

obtained a large and most interesting series of chank-bangle fragments, 62 in number.
With them were a smaller number (7) of sawn working sections. A few marine shells
(Nerita, Nassa, Ostrea, and Conus) were also found among the ruins.

In Gujarat —

(a) Sigam, on north bank of the Heran River. Five weathered sawn working
sections of chanks are represented in the collection from this site. No finished remains
of bangles were seen. The site yielded a variety of neolithic flakes and cores and two
sandstone hammers or pestles. No indication is given of the precise mode of occurrence,
but I conclude they were all surface finds.

(b) Kamrej, 12 miles north-east of Surat. The summit of a small islet in the
Tapti river at this place yielded three sawn shoulder slices (workshop waste) of chank
shells and a single fragment of finished bangle. The latter is remarkable for the pecu-
liarity and elegance of its pattern, a broad and closely worked zig-zag groove such as
I have never seen either among ancient bangle fragments or on any of the present-day
productions of Bengal. With these chank remains were two fragments of sandstone
hammers.

This site is notable as being on an islet in the Tapti river protected against assault
by steep and almost inaccessible sides — a place very defensible and therefore an ideal
place for the settlement of craftsmen.

(c) Mahuri, in Vijapur Taluk. From " an old site at the head of the gully system
which cuts deep into the alluvium of the Sabarmati " at this place, a series of working
sections and waste pieces of chank shell was found sufficiently numerous and varied
to convince me, after examination of the fragments, that a bangle workshop undoubtedly
existed here at a remote period. The presence of sawn waste associated with sawn
working sections is conclusive.

Of completed bangles the remains found were few (8 are shown in the collection)
but of these, three are of special interest on account of the elaboration of ornament
exhibited. Two of these fragments are of broad bangles richly carved in patterns
very closely approximate if indeed not identical with forms in use at the present day.
The third fine example is a tiny fragment of the narrow form of bangle known as churi
in Bengal, usually worn in sets of three on each wrist. The other fragments found are
of simpler patterns.

An interesting associated find was that of a small " finial " carved out of shell,
probably mother-of-pearl. It is identical in form with a mother-of-pearl nose-pendant
now in use among the poorer castes in some country districts in Bengal. An example
which I purchased in Eastern Bengal is carved from the shell of a river mussel (Unio sp.).
From the alluvium at Mahuri whence the bangle fragments came, a few neolithic imple-
ments, chert flakes and scrapers principally, were unearthed, together with several
noticeable pieces of pottery. Of the latter, one is of special importance as it affords

E 2

68 OKHAMANDAL MARINE ZOOLOGY— PART II

some evidence better than the neoliths touching the age of the bangle factory once situated
at this place. It is a small headless figure of a sacred bull, of polished earthenware,
red externally and black within. Two garlands are indicated around the hump by means
of rows of tiny impressed punctures and there can be little doubt that it is of early
Brahmanical age.

(a) Kheralu. A single fragment of a sawn working section of chank shell was found
on the surface of the loess at this place.

Eight sites can clearly be indicated as probable centres of the chank-bangle industry
in Gujarat and Kathiawar, namely : — (a) Sigam, Hiran Valley, Baroda Prant, (6)
Kamrej, on the Tapti, (c) Mahuri, on the left bank of the Sabarmati, Baroda State,
with (d) Ambavalli, (e) Damnagar, (/) Kodinar, and (g) in and on the alluvium of the
Shitranji river above Babapur, all four in Amreli Prant, Kathiawar, also (h) Valabhipur
in Vala State, Kathiawar. At all these places working fragments of chank shells have
been found. The most important sites appear to have been those at Mahuri in Gujarat
and Ambavalli and Valabhipur in Kathiawar. The unworked sections and waste
pieces of shells found at these three places are so numerous, and so characteristic in their
form of stages in shell-bangle manufacture, that we are perforce compelled to admit
these sites as having been in old times locations of important factories, a conclusion
to which further weight is given by the discovery at each of these places of fragments
of completed bangles, in many instances of highly decorated patterns. At Ambavalli
and Valabhipur fragments of finished bangles are especially plentiful ; the ornamentation
is well executed and exhibits considerable taste, a high degree of skill, and undoubtedly
the employment of effective tools of several sorts — saws, drills, and files. Iron is the only
metal suitable for making tools fit for carving the extremely hard substance of chank
shells and it is of the greatest interest and significance that at the Ambavalli site
associated with the many fragments of worked and unworked chank circlets found there,
an iron knife with a tang was discovered which from personal examination I am satisfied
may well represent such a chank-saw as is to-day in common use in Bengal chank
factories for cutting patterns upon the bangles.

From a consideration of the details given above a certain number of facts and
conclusions of importance emerge, to wit : —

(a) In all cases the fragments of bangles and of chank shells appear to have been
surface finds. In several cases this is definitely stated and in the remainder wherever
no statement of horizon is given, the context points to a like provenance. From this
it follows that association with neolithic artifacts in itself has little value or significance ;
both neoliths and chank fragments are practically indestructible by atmospheric weather-
ing agencies and their association may merely connote the fact that particular surface
areas have suffered little or no denudation or change since neolithic times whereby the
broken implements and discarded ornaments of a later age have mingled with those
of an earlier one. Or it may be the result of the artifacts of different ages having been

HORNELL— THE INDIAN CONCH 69

weathered out of different alluvial strata in such way that they come eventually to lie
together at a lower level of the original ground or else in some newer river deposit into
which floods may have rolled them.

(6) The facts already noted that all sections of chank shells, working pieces as well
as wastage scraps, show cleanly sawn surfaces as verified by examination of the originals
now in the Madras Museum, and that these surfaces show series of striae often at two
or more angles to one another, are sufficient to negative the tentative suggestion made
by Mr. Bruce Foote assigning a neolithic origin to the workmanship. Neither serrate
nor biserrate chert flake saws however delicately made could possibly produce such
cleanly sawn sections as we see represented in the collection. The aid of thin metal
saws must be invoked and it is most significant that in two instances (Ambavalli in
Kathiawar and Muski in the Raichur Doab) fragments of iron knives were found asso-
ciated with the remnants of chank working sections. In several other cases (Srinivaspur
in Mysore, Havaligi Hill in Anantapur, and Bastipad in Kurnul) pieces of iron slag were
found in association.

As the working sections of chank shells retain visible evidence of being sawn by
means of a metal (iron) saw and as iron fragments are frequently associated with them,
the evidence is to me satisfactory that the age of the former cannot possibly be neolithic ;
knowledge of the manufacture of iron into somewhat elaborate tools — saws, files, and
drills — must have been possessed by the bangle makers. This would appear therefore
to rule out the early iron age, when iron weapons and tools were of primitive design.

Incidentally this conclusion is likely to affect the estimate of age accorded to the
potsherds so frequently associated with fragments of chank bangles and to render
doubtful their identification as neolithic or even of early iron age.

(c) Three sites alone give other than negative evidence in regard to age. These
are Gudivada in Kistna district, Valabhipur in Kathiawar and Mahuri in Gujarat.
The remains at the first named are indubitably Buddhistic while the occurrence of a
figurine of a bull with a double garland round the hump points distinctly to an age
when the adherents of Brahmanism were in the land holding in especial reverence Siva's
sacred bull. Most important find of all was that made in the ruins of Valabhipur, for
the history of this old city is fairly well known ; the dates of many of the great events
that happened there are on record and the descriptions of two Chinese Buddhist pilgrims
who visited the city are extant. The story of Valabhipur goes back some centuries
before the Christian era and for long it was the seat of the Valabhis, a Rajput race,
and the centre of their rule, till the middle of the eighth century when the last of the line
was overthrown by Arab invaders from Sind. Valabhi was visited by the Chinese
pilgrim, Hiuen Tsang, in the course of his fifteen years' sojourn in India (A.D. 630-645)
and by I. Tsing in the succeeding century. Both pilgrims describe it as a large and
flourishing city and a great centre of Buddhist learning, its streets and schools crowded
with students. The reigning dynasty, themselves of the Brahman faith, appear to have

70 OKHAMANDAL MARINE ZOOLOGY— PART II

been tolerant of Buddhism like many of their contemporaries. In Hiuen Tsang's days
the latter religion was still followed by great numbers of the populace, especially in
Orissa and Southern India ; elsewhere Hinduism was rapidly becoming the popular
religion and the mass of the people were of this faith when the last Valabhi dynasty
ended.

As the chank is a religious symbol both to Hindus and to Buddhists, we may
reasonably conclude that the remains of chank bangles found in Valabhipur were made
for the use of the women of the town and neighbourhood not later than the eighth
century. The trade must have been long established at that time to judge by the excel-
lence of the work turned out, which fully equals that of average Bengal workmanship
of the present day.

Taking all facts into consideration I am inclined to date the majority of the bangle
fragments found in the Kathiawar and Gujarat sites as roughly contemporary with the
Valabhipur specimens or at most not antedating them by more than 300 to 400 years.

To date the Deccan chank -bangle factories is more difficult ; one outstanding
fact is the simplicity of all the patterns. The great majority are devoid of ornament
save for a boss roughly carved at one side. This plainness of design would seem to
bespeak less skill on the part of the Deccan workman than on that of his fellow craftsman
in Gujarat. If that be the explanation, and if it be not due to lack of taste or of the
means to pay for good work on the part of the buyers, then we may reasonably date
the majority of these fragments back to the first few centuries before or after the
beginning of the Christian era. The presence among the pottery mixed with the bangle
fragments found near Srinivaspur in Mysore of a flat sherd similar in pattern to one
found with the Buddhist remains at Gudivada in Kistna District is noteworthy as
lending further countenance to this conclusion.

(d) The finds made by Mr. Bruce Foote argue two great centres of chank -bangle
manufacture and usage apart from that in the extreme south of the Madras Presidency,
namely, one in the Southern Deccan and the other round the shores of the Gulf of
Cambay. It is most probable that other centres of the industry did exist, but at present
there is no direct evidence to this effect. For instance, it is not likely that an industry
which was firmly established in Eastern Bengal at the time of the arrival of the
Portuguese in India ' and of Tavernier's travels in the seventeenth century, and which
continues to flourish at the present day, should be of modern growth.

With regard to the third known seat of the industry in ancient times, that which
flourished in. the early centuries of the Christian era in the Tinnevelly district, its
geographical location in the coastal section of the Pandyan kingdom made it the natural
centre and home of a great chank-cutting industry. Its Pandyan sovereigns were from
time immemorial overlords of the Pearl and Chank Fisheries of the Gulf of Mannar and

1 Garcia da Orta writing in the sixteenth century states that the chank was then an article of
importance in the Bengal trade, though less valuable than formerly.

HORNELL— THE INDIAN CONCH 71

Palk Bay, the most important source of supply of the raw material then and now, and
it is a curious vagary of trade that the present seat of the industry should be situated
1,500 miles from the scene of the fishery.

From the fact that among a few widely separated castes, sub-castes and tribes of the
extreme south of India, including among others the Kotas of the Nilgiri Hills and certain
sections of the Vellalans and Idaiyans in the inland Coimbatore district, the custom
prevails of wearing chank bangles for ceremonial reasons, we may also reasonably infer
the former wider prevalence of the custom. Indeed it is probable that the custom was
at one time prevalent throughout a large section of Southern India.

Kathiawar and adjacent Gujarat are also both maritime provinces and this
geographical situation is the key to the location of a chank-bangle industry in those
provinces in early times ; the coast of Kathiawar is the only considerable source of chank
shells apart from the Gulf of Mannar and Palk Bay. No chank-cutting is now done
either in Kathiawar or Gujarat ; the women there have abandoned their former habit
of wearing chank bangles and all the shells fished in this locality are exported from
Bombay to Bengal where they are known in trade as " Surti " shells, Surat having been
the port of shipment prior to the rise of Bombay.

Why the Southern Deccan should once have been the home of a shell-cutting industry
is not so easy of explanation, seeing that it is situated in the heart of the country and
distant from 400 to 500 miles from the nearest sources of supply (Eameswaram and the
Tanjore coast). Possibly the location of this trade in the Deccan was due to the superior
skill as craftsmen of the people in this region inherited from stone-using ancestors
who found in the quartzite and trap rocks of the district more suitable material for their
weapons and tools than the men to the southward where intractable gneiss constitutes
all the rocky outcrops. Certainly in prehistoric times, Bellary, Kurnul and Cuddapah
were more thickly populated than the country to the south if we may judge from the
evidence of the number of stone implements found respectively in these two sections
of India. The neolithic remains of these Deccan craftsmen show their makers to have
been comparatively highly-skilled workers and with the discovery of the use of iron,
haematite ore being abundant in Bellary, the men of this district may reasonably be
supposed to have developed special skill in the working of the new material into tools
and in the manufacture of many articles, ornamental as well as useful, with the aid
of these improved tools. Add to this the natural conservatism of tribes isolated from
the coast by hill ranges — the customs and manners of the Deccan tribes have been
less changed by contact and intermixture with surrounding races than the majority
of the tribes or races living in the coastal plains. To these inland people the wonder
of the great shell honoured by their gods would appeal vividly ; the mystery to them
of its origin would confer added importance and, as we find the wild hill tribes of Thibet,
Assam, and Bhutan do at the present day, they would end by endowing ornaments made
from it with mysterious powers of ensuring well-being and good luck, even as the

72 OKUAMANDAL MARINE ZOOLOGY— PART II

Buddhist cartmen of Ceylon and their Hindu brethren throughout the Southern Carnatic
adorn their bulls with a chank shell as an amulet against the evil eye.

Chank shells for the Deccan bangle workshops may probably have come from the
Tan j ore coast, this being the nearest source of supply. The Tan j ore fishery appears
to have been fairly lucrative down to 1826 when economic changes caused a collapse
of the industry. Tirumalavasal at the mouth of one of the northern branches of the
Kaveri is the centre of the chank -fishery on this part of the coast and is not far from
Kaveri-pattanam, once the chief port of the Chola kingdom and in the height of its
prosperity in the early centuries of the Christian era. From Kaveri-pattanam to the
inland districts of Kurnul and Bellary the transit of goods would be comparatively
easy and safe ; coasters would be used to the mouth of the Kistna, 350 miles to the north,
whence river craft would carry the goods direct to their destination, 200 miles inland.
Or it may be that the shells required in the industry were fished further south, for we have
mention by Cosmas Indico-Pleustes in the sixth century (circa 545) of a place called
Marallo on the continent adjoining Ceylon, where a shell called by him *OX\MW?
(Kochlious) T was produced in quantity, and Yule in " Cathay and the Way Thither "
(London, 1866), Vol. I, p. 81, suggests that this Marallo is the same word as Marawa,
the name of the ruling caste in the district of Ramnad ; if this be accepted, the reference
would indicate the chank-fishery carried on off the coast of the Mara war country and now
operated by lessees of the Raja of Ramnad. Again, a chank fishery, the most productive
in the world, exists to-day in the shallow seas in the neighbourhood of Jaffna in Ceylon
and direct communication by means of large native craft having existed from time
immemorial between the north of Ceylon and the port of Masulipatam, for centuries
the eastern sea-gate of the Deccan, this fishery may have been drawn upon also to
supply the needs of the latter locality.

The cause of the cessation of the chank industry in the Deccan, Gujarat, and
Kathiawar is to be looked for in the constant strife which kept India in a welter of blood
through the six centuries of Muhammadan dominance in the land. From the days of
Mahmoud of Ghazni, the northern and central portions of India in particular were
harried by successive waves of fanatic invaders sweeping down through the north-west
passes, and from the thirteenth century onwards to the end of the seventeenth the
story of India is that of an unceasing contest between Muhammadan and Hindu for power
on the part of the former and for existence and religion on that of the latter. Well
may certain old Hindu customs have disappeared ; during the worst periods when the
intolerance of the conquerors was at its height, their influence was often exerted towards
the suppression of Hindu customs and this, combined with the dislocation of trade
consequent upon the general insecurity of the country and the frequent recurrence
of raids and widespread warfare, may be considered the main reason for the decay of

1 In the Norman-French dialect still spoken in Jersey and the other Channel islands, the common
whelk (Buccinum), which is the European representative of the Eastern chank, is known as coqueluche !

HORNELL— THE INDIAN CONCH 73

the chank-bangle industry in the Deccan and Cambay provinces. A striking confirmation
of this conclusion is afforded by Garcia da Orta, a Portuguese writer of the sixteenth
century, whose colloquies on Indian drugs and simples have recently been translated
by Sir Clements K. Markham (London, 1913). In the thirty-fifth colloquy this old
physician remarks :— ' There is another (shell) they call chanquo, of which they make
boxes, tables, and counters, for though it is rough outside, inside it is very smooth
and beautiful. This chank is an article of trade to Bengal, and used to be worth more
than it is now. The large ones, which we call Buzios, go to Bengal and are worked
up very beautifully, remaining very smooth and white. For this only a small quantity
is used, the rest being wanted for bracelets and other ornaments. It was the custom
in Bengal that no person of distinction who was a virgin could be corrupted unless
she had bracelets of the chank shell on her arms.1 After the arrival of the Patans this
custom was neglected and the chank became cheaper in consequence. You see here a
chess table at your service where you may see the chank when you like."

1 Another translation renders this sentence rather differently, to wit : — " There was formerly a
custom in Bengal that no virgin in honour and esteem could be corrupted unless it were by placing
chank bracelets on her arms." Da Orta's statement refers really to the fact that an essential ceremony
in a Bengali marriage consists in placing a chank bangle on each of the bride's wrists ; the marriage
would not be formally valid if this were to be omitted.

THE SCOPE FOR DEVELOPMENT OF A SHELL-BANGLE
INDUSTRY IN OKHAMANDAL

AT the present day the Indian shell-bangle industry is limited to Bengal, where
in Dacca, Calcutta, Nadia, and other centres it gives lucrative employment to large
numbers of skilled artisans. How economically important it is as an art industry will
be realised when we learn that the value of the raw material required ranges between
2 and 2J lakhs of rupees per annum as given by the importers, an amount probably
understated so considerably that it is quite safe to put the average actual value at
2 1 lakhs. This is the wholesale value ; the shells usually change hands twice before
they reach the actual bangle-workers ; in turn the latter sell their products to bangle
merchants who distribute them to shopkeepers and peddlers throughout Bengal, Behar,
Assam, Bhutan and Thibet. The value increases several times in this process of
manufacture and distribution till it is certain that the final retail value of the products
of this industry at a conservative estimate is considerably over 15 lakhs of rupees, or
£100,000. From one shell an average of three wide bangles can be cut ; the wholesale
value of good shells may be taken at from 1 to 2 annas each, whereas the one and a half
pair of bangles that one shell produces will be sold at anything from 9 annas to rupees 18
according to the amount of ornamentation. Certainly the three bangles yielded by
one shell have an average retail value of not less than one rupee. On this basis the
shells increase in value eight times on the import value if we take the latter at the high
rate of 2 annas each.

The tools employed by the shell-cutters and gravers are of the simplest and most
primitive description ; they consist of heavy double-handled saws employed in sawing
the circlets which constitute the working sections for the bangle carver (vide PI. IV,
Figs. 1 and 2) and of small handsaws (PI. V, Figs. 1 and 2), bow-drills and files for smoothing
and carving the patterns. No labour-saving devices are used ; the tools with the
exception of European-made files, are undoubtedly identical with those used 1,500 years
ago, in particular the great saw — the most important of these tools — which is referred
to in an ancient Tamil poem quoted on page 59.

The present concentration of the shell-bangle industry in Bengal is consequent

>

LJ

l

o
z
o
o

Q

Z

u
u
(S
Q

(0

e

(B

U)

c

c
I)

Q.

CO

b

JlC

c

(0

Q

Z

I
X

o

o

o
o

N

QC

Q.
O

a
o
u

(0

Q

C
(0

bfl

E

'c

O

+-•

(J
<u
CO

LU
<

I

o
z
o
o

Q

Z

D
_Q.

C

b

fan

c

hb

c

0)

c

(Q
hD

'E
o

D.

CO

h)

L

<

D

X
X

o

o

o

N

cr
<

s

LJ
<

I
O

z
O
O

D

Z

Q

Z

I
*
O

u.
O

>
C3

3

O
O
N

UJ

Z

cc

s_
3
^0.

C

b

bO

C
C8

bO

OS

bO

HORNELL— THE INDIAN CONCH 75

upon this province being the only one in India where shell bangles are extensively used
by high and low among the female population. Even in Bengal itself shell-cutting
proper — the sawing of the shells into working sections — is restricted to a very few centres ;
this part of the work requires a great amount of skill such as can be acquired only after
a long and toilsome apprenticeship, begun at an early age ; shell-slicing calls for the
possession of a highly trained eye, perfect steadiness of hand and arm and an ironlike
capacity to sit for long hours in a position of great discomfort. As a consequence the
sawing of working sections is limited to a few centres and a good cutter is a valuable
asset to his employer. To retain a hold upon these men, employers willingly give large
advances in cash.

The towns and villages where the working sections are fashioned into the bangle
patterns favoured locally are much more numerous than the sawing centres ; the bangle-
workers in the great majority of such places do not attempt to slice the shell itself ;
they depend upon the wholesale sawyers of Dacca and a few other great centres for their
working sections. Particularly is this the case with Muhammadan bangle-workers,
for these men are comparatively new recruits to the trade and therefore are employed
chiefly in the less skilled sections of the industry and in fashioning bangles of the simple
and crude patterns affected by the poorest and most ignorant of the population.

In spite of the two great advantages possessed by the bangle trade in Bengal — (a)
location in the centre of the only province where chank -bangle wearing continues to be
widely prevalent, and (b) the possession of a caste of highly skilled hereditary chank-
bangle sawyers and carvers, there is ample scope for the establishment of a large chank-
sawing industry in those localities where the shells are fished extensively — in Kathiawar
and on the coast of the Gulf of Mannar. Indeed the proposition is one economically
sound. At present the whole shell has to be transported from the place of fishing to
Calcutta and Dacca, distances in both cases upwards of 1,200 miles, where the mouth
part, the apex, and the columella, constituting over 60 per cent, of the total weight,
are cut or broken away and thrown out as waste in the process of manufacture. Under
present conditions freight and handling charges have thus to be paid on more than double
the weight of the material actually employed. A radical change in the method of
preparing working sections would, however, be necessitated as the present hand-sawing
of the shell is an art too difficult for unskilled labour to acquire and I believe no Bengal
sawyers would leave their own districts to work elsewhere. The one alternative is to
adopt a form of machine saw capable of employment by comparatively unskilled
workpeople. A band saw working with emery, or preferably with carborundum powder,
probably would give satisfactory results, the motive power being either foot power
leaving the hands free to hold and to guide the shell or, preferably, a small oil engine
or else electric current in towns where the latter is available. The introduction of
a chank-cutting industry into Okhamandal is, I believe, eminently feasible and practical
and commercially sound. To begin with there is a considerable supply of shells available

76 OKHAMANDAL MARINE ZOOLOGY— PART II

on the Okhamandal coast itself, and if this were to prove inadequate, it could be supple-
mented by supplies drawn from other parts of the coastline of Kathiawar.

The supply being therefore assured, the chank-sawing industry might be developed
along either of two lines, either the shells might be sawn into working sections and these
exported to Bengal in this partly manufactured state, or the finished bangle might be
produced. With provision of a power machine suitable for cutting sections, this could
readily be adapted to do smoothing, carving, and polishing, so that an intelligent workman
—especially if he be one having some knowledge of lapidary work — if he were given a
series of simple patterns to begin with and direction how to use the machine he has to
handle, should have no insuperable difficulty in doing passable work in the course of
a few weeks. Once he acquires a command over his machine and the ability to carve
simple patterns, the power to attack more complicated patterns — such as those depicted
on Plate VII — would soon come. As to a market, Bengal has now increased so greatly
in population that the chank-bangle workers there begin to find difficulty in obtaining
a sufficient supply of the raw material, so even if no local demand should develop in
Okhamandal, the Bengal market would absorb readily all that an Okhamandal workshop
would be likely to produce for years to come. It seems, extremely probable, however,
that a more remunerative local demand may be counted upon to develop concurrently
with the progress of the industry. Many thousands of pilgrims resort to the holy shrines
of Bet and Dwarka annually and these people, who value the chank shell as one of the
emblems both of Vishnu and Krishna, as a vessel for use in their household religious
ceremonies and, in many cases, as a curious souvenir of a memorable pilgrimage, would,
I believe, esteem still more a handsome ornament made from this shell for their women-
folk, particularly if judicious efforts were made to draw attention to the appropriateness
of such an ornament on the hands of Hindu ladies ; the beauty of its pattern, its snow-
white colour — emblem of purity — and its general superiority to the gaudy glass bangles
of European manufacture that not infrequently inflict serious wounds on the wrist
when they are accidentally broken. Patriotic preference for a home-made ornament,
especially when backed up by the more solid advantages of greater real beauty and
greater strength and lasting property, should also count for much. Lastly, a most
valuable potential market for high-priced chank bangles certainly exists in Bombay,
Karachi, and other great cities where European and American tourists are to be found
in yearly increasing numbers. These people are all eager to take home some " curios-
ities " of the East, preferably such as are not too bulky ; from my own knowledge
I can say that well-carved chank bangles do appeal greatly to these visitors from abroad
and it is only lack of enterprise on the part of Bengal chank bangle manufacturers that
has prevented the exploitation of this promising field. Time after time people who have
seen the chank bangles in my collection have expressed their desire to purchase — a desire
baulked by the fact that none are to be had in any shops visited by tourists in any city
in India.

>

LU

X

o
z
o
o

Q

Z

V

V
V
\

V
k

V
V
\

\

\

\

\

\

\

\

I E

£ 3

a «
- u>

13

(O

2

73

t
O

•*-•

o

_OJ

"o
O

o
o
U.

O

_c

C
<U

I

I I

Q

Z

LL

o

?"

o
o

o
o

N

LU

Z

oc

«
jQ

o

c

V

E
o

c
o

u

3

o
u

o

c
u

0)

•*-•

"to
Q.

HORNELL— THE INDIAN CONCH 77

Here, then, is an artistic industry requiring a minimum of capital to ensure hand-
some returns provided it be conducted with discrimination, enterprise, and honesty ;
well worthy of Government encouragement and of support by those who profess their
desire to revive and develop the indigenous art industries of India.

At present the waste material produced when sawing bangle working-sections is
not properly utilised. The bulk of it is burnt to produce lime, whereas much of it is
capable of utilisation for button-making, seeing how great and increasing is the demand
for buttons of an attractive appearance. In former days chank shell was also sometimes
employed in inlaying ; Garcia da Orta mentions the use of squares of chank shell in making
the white squares of chess-boards.

EXPLANATION OF PLATES.

PLATE I (FRONTISPIECE).

Figs. 1, 2, and 3. — Sinistral chanks respectively in the Satya Bhamaji, Shank Narayan, and Lakshmi
Temples, Bet, Kathiawar.

PLATE II.

Figs. 1 and 2. — Immature chanks (Turbinella pyrum) from Okhamandal ; both sh'ow persistence of the

protoconch.

Fig 3. — Adult chank from Okhamandal.
Fig. 4. — Elongated variety from the Andaman Islands. Also two very young individuals with protoconch

well shown.

PLATE III.

Fig. 1. — Group of Cheruman Women wearing necklaces of so-called chank-rings.
Fig. 2. — Under a banyan tree. Selling chank shells to pilgrims returning from Bet.

PLATE IV.

Fig. 1. — Sectioning chank shells in a Dacca workshop.

Fig. 2. — Sharpening a chank saw, Dacca.

Fig. 3. — Breaking away the remains of the septum from a sawn chank circle (working section).

Fig. 4. — Rubbing down the inner surface of a working section, Dinajpur, Bengal.

PLATE V.

Fig. 1. — Carving an incised pattern, Rangpur.

Fig. 2. — Sharpening an engraving saw, Dinajpur.

Fig. 3. — Forming a simple ridge pattern by rubbing down the sections on a stone, Rangpur, Bengal.

Fig. 4. — Rubbing down Cinnabar (Hingol) to colour lacquer red, Dinajpur.

PLATE VI.

Fig. 1.— Lacquering marriage bangles, Dinajpur.

Fig. 2. — Rest used when finishing off an inlaid lacquer pattern, Dinajpur.

Fig. 3. — Chank-shell waste from ancient bangle factory sites at Korkai, Kayal and Tuticorin (two upper

rows) compared with modern waste pieces from Dacca (bottom row, where a working section

is also shown).
Fig. 4. — Making children's feeding spouts from chank shells, Karimanal, near Pulicat (Madras).

PLATE VII.
Pattern reconstruction of some of the ancient bangles in the Foote Collection, Madras Museum.

REFERENCES.

The two most frequently quoted works in this Report being : —

THUBSTON, EDGAR. — "The Castes and Tribes of South India," 7 Vols., Madras, 1909, and
RISI.EY, H. H. — "The Tribes and Castes of Bengal," 2 Vols., 1891, .

references thereto arc abbreviated to the author's name followed by the number of the volume and the
page therein quoted, eg., Thurston, II. p. 21.

MARINE ZOOLOGY OF OKHAMANDAL.

CALCAREOUS SPONGES, PLATE I.

CALCAREOUS SPONGES.

Cambridge University Press

REPORT

ON THE

CALCAKEOUS SPONGES

COLLECTED BY

MR. JAMES HORNELL

AT

OKHAMANDAL IN KATTIAWAR IN 1905-0.

BY

ARTHUR BENDY, D.Sc., F.R.S.,

Professor of Zoology in the University of London (King's College).

[With Two Plates.]

THE collection of sponges from Okhamandal placed in my hands by Mr. James
Hornell contains six species of Calcarea, unfortunately not in a very good state of
preservation. I have found it necessary to describe as new one species of Sycon (S.
grantioides) and one of Leucandra (L. dwarkaensis), together with a variety of Leucandra
donnani (var. tenuiradiata). More interesting than these new forms, however, are the
specimens of Grantessa hastifem (Row) and Heteropia glomerosa (Bowerbank), which
enable me to add a good deal to our knowledge of these little known species.

The classification adopted is that of Dendy and Row's [1913] "Classification and
Phylogeny of the Calcareous Sponges, with a Reference List of all the described Species,
systematically arranged." I have followed my usual practice of quoting the Register
Number (R.N.) of each specimen in order to facilitate accurate reference.

1. Sycon grantioides, n. sp. — (Plate I, Fig. i.)

The single specimen, unfortunately in a bad state of preservation, has the form
of a slightly compressed cylinder, rather wider above than below, but contracting

79

80 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

suddenly to a moderately wide vent (Fig. 1). There is now no peristome, but it is
impossible to be certain that there were no hair-like oxea surrounding the vent in life.
The dermal surface is minutely reticulate, without conspicuously projecting oxea in
its present condition, but this is because the outer ends of the numerous large oxea
have all been broken off short. Size of specimen about 16 mm. high by 7 mm. in greatest
width. Texture soft and flabby. Colour in spirit white.

The gastral cavity is wide, and the total thickness of the sponge wall only about
TO mm. The gastral surface is smooth and pierced by the numerous small apertures
of the exhalant canals. The dermal surface is formed by the fusion of the conical outer
ends of the radial chambers to form a reticulate pattern. The gastral cortex is moder-
ately thick and pierced by the short exhalant canals. The radial chambers taper towards
their distal extremities, while proximally they open, usually in groups, into the short
exhalant canals that pierce the gastral cortex. They are provided with well developed
diaphragms at the apopyles. They exhibit the " linked " arrangement described
by Jenkin [1908B] in his genus Tenthrenodes. The inhalant canals open on the dermal
surface by wide apertures between the distal conuli.

The skeleton is typically syconoid in its arrangement, except that the usual tufts
of oxea which crown the distal ends of the chambers in typical species of the genus are
replaced by large oxea which cannot really be said to be arranged in tufts at all, but
extend sometimes through the entire thickness of the sponge-wall, between the
chambers, with their distal ends projecting from the surface more or less at
right angles. The articulate tubar skeleton consists of many joints, but there are
no specially differentiated subgastral sagittal radiates. The spicules are so much
broken and eroded that I have not found it practicable to get perfect specimens
suitable for figuring. The following descriptions, however, may be taken as
substantially correct.

(1) Sagittal triradiates of the many- jointed articulate tubar skeleton ; with
very long and slender rays ; the oral angle wider than the paired angles ; the oral
rays curved around the chamber as usual and sometimes slightly and irregularly bent.
Dimensions of a specimen whose oral rays lie in about the middle of the chamber
wall : — Basal ray about 0'2 by 0*005 mm. l ; oral rays about 0'12 by 0'005 mm.

(2) Radiates of the gastral skeleton ; usually triradiate but sometimes with a
feebly developed apical ray. With wide oral angle and very long and slender
facial rays. These spicules are not definitely oriented but form a confused felt-
work in the gastral cortex, in which it is very difficult to follow the individual
rays for their entire length. I have measured the basal ray up to 0*4 mm. in
length, with an oral ray of about 0*2 mm., each having a diameter at the base
of about O'Ol mm., but they are usually rather more slender.

1 The first measurement is the length and the second the greatest thickness in all cases except
where otherwise stated.

BENDY— CALCAREOUS SPONGES 81

• (3) Oxea ; straight or nearly so, and of nearly uniform diameter, but gradually
sharp-pointed at their inner ends ; slender in proportion to their great length.
The outer ends are all broken off so that I cannot say what they are like. I
have measured the remaining portion of the spicule up to about 1'7 by 0'025 mm.

This species is not a typical Sycon. In the arrangement of its oxea it much
more closely approaches the genus Grantia, from which, however, it must be
excluded on account of the absence of a dermal cortex. In the " linked " arrange-
ment of the radial chambers it resembles those species included by Jenkin [1908B]
in his genus Tenthrenodes, but that genus cannot be maintained [Dendy and
Row 1913].

Register Number and Locality. III. 4. Off Dwarka.

2. Grantessa hastifera (Row).— (Plate I, Figs. 2, 2a ; Plate II, Figs, la-if".)
Grantilla hastifera Row [1909].
Grantessa hastifera Dendy [1913].

It is a curious coincidence that this species, first described by my colleague
Mr. Row in 1909, and re-described by myself in 1913, in both cases from very
inadequate material, should again occur in the present collection. Mr. HornelFs
material, however, enables me to add some valuable particulars, especially with regard
to the extremely variable external form and the structure of the very remarkable
hastate oxea.

The sponge (Figs. 2, 2a) may be described as consisting of thin lamella?, folded
into irregularly tubular or cup-shaped forms ; sometimes forming irregularly proli-
ferating masses of larger and smaller tubes (Fig. 2a). The lamella or sponge-wall
is about 1*5 mm. in thickness, but the diameter of the tubes or cups varies from
about 2 mm. to at least 22 mm. Unfortunately the sponge is very fragile and
the material has been much broken up, so that it is difficult to say anything
about the oscula, but these appear to be naked and terminal. The larger fragments
(Fig. 2) look like thin, concave, irregular shells, but they probably formed parts of cups
in life. All the pieces are possibly parts of the same colony, and it is certain that the
wide cups give off narrow cylindrical tubes.

The outer surface is for the most part smooth and subglabrous, but here and there
with a few conspicuously projecting spicules. The inner surface (Fig. 2) is pitted by the
numerous openings of the short exhalant canals, arranged in groups. The colour in
spirit is dirty white.

The canal system is syconoid, but the material is so badly preserved that it is
impossible to make out any details.

The gastral and dermal cortex are each about 0-14 mm. in thickness. The gastral
cortical skeleton is made up of the slender rays of tangential gastral triradiates (Fig. 76)
and to a slight extent of the much stouter oral rays of subgastral sagittal triradiates

82 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

(Fig. 7c). The dermal cortical skeleton is made up of the dermal triradiates
(Fig. la) arranged tangentially, and of the outer rays of subdermal pseudosagittal
triradiates (Fig. le). The tubar skeleton approaches but does not completely
realise the inarticulate type, the stout subgastral sagittal triradiates being
succeeded centrifugally by two or three other sagittal triradiates very similar
to themselves (Fig.- Id). The subdermal pseudosagittal triradiates (Fig. 7e) are
much slenderer than the subgastral spicules (Fig. 7c), and their centripetally directed
rays are commonly grouped in bundles which meet and overlap the opposing basal
rays of subgastral or tubar sagittals. Around the oscular margin there is a narrow
band of strongly alate triradiates which occasionally develop a short apical ray. There
appears to be no oscular fringe of oxea, but a few stout oxea (Figs. If, If, If") are to be
found penetrating the sponge wall more or less at right angles, and often projecting
from the dermal surface.

The spicules may be grouped under the following heads :—

1. Dermal triradiates (Fig. la). Nearly equiangular but commonly inequiradiate,
with indications of a sagittal character. There is no definite orientation, but what
appear to be the oral rays are sometimes more or less crooked while the basal ray is
straight. All rays gradually and sharply pointed. Size very variable. In a typical
example the basal ray measured about 0-42 by 0-025 mm., and the orals 0-36 by
0-025 mm. and 0-3 by 0-025 mm. respectively.

2. Gastral triradiates (Fig. 76). Very similar to the dermal triradiates but con-
siderably smaller. In a typical example the apparently basal ray, which was straight,
measured about 0-16 by 0-02 mm., and the orals, which were very slightly crooked,
each about 0-24 by 0-015 mm. Towards the osculum these spicules become very strongly
sagittal and exhibit a definite orientation, the oral rays being extended in line with
one another and parallel to the oscular margin, while a short apical ray is occasionally
developed.

3. Subgastral sagittal triradiates (Fig. Ic). Stout, with oral rays slightly recurved
towards the basal, which latter is straight. Oral rays bent towards one another in a
plane parallel to the gastral surface, so as to give rise to a deceptive appearance of
inequality according to the point of view.1 All rays gradually sharp-pointed. In a
typical example the basal ray measured 0-51 by 0-04 mm., the orals apparently 0-37 by
0-035 and 0-25 by 0-035 mm. respectively.

4. Tubar triradiates (Fig. Id). These are not really distinguishable from the
subgastral sagittal triradiates except by their more distal position. The most
distally situated show less curvature of the oral rays, and are of smaller size, but they
are connected by intermediate forms with the typical subgastral sagittals.

5. Subdermal pseudosagittal - triradiates (Fig. le). Conspicuously smaller and

1 Actual inequality may, however, occur to a considerable extent.
- For a discussion as to the nature of these spicules vide infra, p. 86.

BENDY— CALCAREOUS SPONGES 83

especially more slender than the subgastral sagittals, with the centripetal ray
straight and the outer rays more or less bent and asymmetrical. All rays gradually
sharp-pointed. In a typical example the centripetal ray measured 0-36 by 0-025
mm., and the outer rays apparently 0-24 by 0-025 and 0-17 by 0-025 mm.
respectively.

6. Large oxea (Figs. If, If, If"). These spicules, though not very numerous, are
highly characteristic. Their outer ends usually project more or less at right angles
from the surface, and often for a great proportion of the length of the spicule, but
except over protected areas of the surface they are almost invariably broken off.
The form of these spicules affords by far the most characteristic feature of the
species, and after a careful study of isolated specimens I am able to add certain
particulars under this heading. The entire spicule is slightly curved (Fig. If). The
inner end is simply sharp -pointed. The outer end is sharp-pointed and flattened
like a spear-head, with a sharp knife-edge on either side. On one side (the
concave side of the spicule) this knife-edge is simply rounded, on the other it is
produced backwards, where it meets the cylindrical shaft of the spicule, into two
short, conical teeth. The presence of two teeth can only be clearly seen when the
spicule is examined edge on (Figs. If, If"), and hence only a single tooth, as seen in
side view (Fig. If), has hitherto been described. These spicules measure about 0-83 mm.
in total length by 0-035 mm. in diameter in the middle, while the spear-head measures
about 0-1 mm. in length.

7. Trichoxea ; long and very slender, typically arranged more or less at right
angles to the dermal surface, but so scarce that they can hardly be regarded as an
essential constituent of the spiculation.

Previously knoivn Distribution. Red Sea (Row) ; Providence I., Indian Ocean
(Dendy).

Register Number and Locality. III. 2. Off Dwarka.

3. Heteropia glomerosa (Bowerbank). — (Plate I., Figs. 3, 3a, 36 ; Plate II.,

Figs. 8a-%)

Leuconia glomerosa Bowerbank [1873].

This well-characterised and very beautiful species was first described by Bower-
bank in 1873, from dry material collected at Port Elizabeth, South Africa. The type
specimen, which is now in the Natural History Department of the British Museum,
is a good deal worn, and possibly beach-rolled, and Bowerbank's figure cannot be taken
to represent its natural appearance very accurately. The species has not been recorded
since its original publication, but it occurs in considerable quantity in Mr. Hornell's
collection. Although the locality is so widely separated from that where it was originally
found, there can be no doubt as to the identification, which is based upon a careful
re-examination of Bowerbank's material. His description and figures are quite in-

F 2

84 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

sufficient, and in several respects misleading. Thus he says that " the terminal orifices
are rarely ciliated ; but when they are so furnished the ciliary fringe is composed of a
prolongation of the layer of large acerate spicula." This really applies only to the
openings accidentally produced by breaking across of branches, the true oscula are
provided with a special fringe of slender, hair-like oxea. Again, he says that there
are no " defensive spicula projected from the oscular surface," by which he evidently
means no gastral quadriradiates ; such spicules, however, occur in his specimen. The
statement that the apices of the subdermal and subgastral triradiates are cemented
together where they meet by " keratode " hardly deserves contradiction, the " kera-
tode " being, of course, simply the dried remains of the soft tissues.

Under the circumstances it seems desirable to give a completely new description,
based upon Mr. Hornell's spirit-preserved material. I may say, however, that I cannot
find any character in which his specimens differ from the type.

The sponge colony (Figs. 3, 3a, 36) consists of very numerous, rather slender,
cylindrical branches, for the most part ascending vertically and lying close together.
The branching is very irregular and takes place by the formation of lateral buds
at varying levels. Each branch terminates, when fully developed, in a distinct
osculum, but the younger buds are blind. The oscula appear to be naked, but
are in reality provided with an inconspicuous fringe of slender, hair-like oxea.
The projecting portions of these spicules appear always to be broken off. Just
within the osculum is a transverse membranous sphincter. All the colonies have
evidently been attached below in life, and in one case a portion of the substratum
is still present in the form of a barnacle shell.

The largest colony, or piece of a colony, in the collection measures about
32 mm. in height by 38 mm. in greatest breadth, and is composed of about 50
branches measuring up to about 19 mm. in length and 2-5 mm. in diameter.
There is considerable variation, both as regards the length of the branches and
the compactness of the colony, in different specimens. The surface of the branches
is longitudinally striated (Ute-like) owing to the presence of the large dermal
oxea, but otherwise smooth. The colour in life was white, in spirit it is dirty
white.

The canal system is typically syconoid, the thimble-shaped radial chambers
extending at right angles through the wall of the sponge from gastral to dermal
cortex.

The preservation of the material is not good enough to enable me to make
any detailed histological observations, but the nuclei of the collared cells are apical.

The gastral cortex is fairly thick, and its skeleton is made up of the facial
rays of gastral sagittal triradiates and quadriradiates, and the oral rays of sub-
gastral sagittal triradiates. The dermal cortical skeleton is very strongly developed
and made up of dermal triradiates, the outer rays of subdermal pseudosagittal

DENDY— CALCAREOUS SPONGES 85

triradiates, and huge longitudinally placed oxea, the latter lying for the most
part on the inner side of the layer of triradiates. The tubar skeleton is of the
inarticulate type (Figs. 8d, 8e), composed of the centrifugal rays of the subgastral
triradiates and the centripetal rays of the subdermal triradiates (Fig. 8e, o1).
The former usually extend outwards right through the chamber layer, while the
latter may extend inwards as far as the gastral cortex or may fall considerably
short of this. Just within the osculum the stout oral rays of the sagittal gastral
triradiates, here distinctly alate, are extended parallel with the oscular margin
and packed close together in a dense feltwork. Immediately on the -inner side
of this feltwork lie the hair-like oxea of the oscular fringe. The oscular fringe
thus arises from the gastral cortex and the colossal dermal oxea take no part
in its formation.

The spicules may be grouped under the following heads :—

1. Dermal triradiates (Fig. 8a). These are nearly equiangular but sagittal owing
to the greater length of the basal ray. The rays are usually straight and rather
slender, tapering gradually to sharp points, the basal ray being a good deal
longer and rather more slender than the orals. In a typical example the oral
rays measured about 0-11 by 0-01 mm. and the basal about 0-24 by 0-008 mm.
There is a good deal of irregularity in the arrangement of these spicules, but
typically the basal ray points away from the osculum as usual.

2. Gastral triradiates (Fig. 86). These are a good deal larger than the dermal
triradiates and more markedly sagittal, the basal ray being usually very long
and slender, while the oral rays are often more or less curved backwards. The
oral rays are sometimes very unequally developed, one beingxmuch longer than the
other. The rays are slender and more or less gradually sharp-pointed. In a
typical example the oral rays measured 0-2 by 0-012 mm. and 0-17 by 0-012 mm.
respectively, and the basal ray 0-44 by 0-012 mm. The basal ray is typically
directed away from the osculum as usual. Near the osculum the oral rays are
curved backwards so much as to lie almost in a line with one another, and are
much more strongly developed than the basal.

3. Gastral quadriradiates (Fig. 8c). These resemble the gastral triradiates, but
they are decidedly stouter. The apical ray may be very strongly developed
(Fig. 8c, a.r.), straight and sharply pointed, and directed obliquely upwards in the
gastral cavity. These spicules are not very numerous.

4. Subgastral sagittal triradiates (Fig. 8d). These are typical alate spicules,
with slender, gradually sharp-pointed rays. The recurved oral rays are extended
almost in line with one another in the deeper part of the gastral cortex ; the
straight basal ray extends through the chamber layer to the dermal cortex. In
a typical example the oral rays measured about 0-145 by 0-01 mm. ; the basal
about 0-245 by 0-01 mm.

86 OKHAMANDAL MARINE ZOOLOGY REPORT— PART IT

5. Subdermal pseudosagittal triradiates (Figs. 8e, 8/). The form and arrange-
ment of these spicules strongly support the view1 that they are really distal
tubar triradiates which have undergone rotation so that the original basal ray (&)
has come to lie in or below the dermal cortex, while one of the original oral
rays (o1) has come to be directed inwards and has become more or less elongated.
I had previously looked upon this inwardly pointing ray as the basal ray of a
sagittal spicule, and upon the true basal ray as one of the orals. It will be well
in future to speak of the former (Figs. 8e, 8/, o1) simply as the centripetal ray,
and the others as the outer or dermal rays. In the present case the dermal rays are
asymmetrical and it is quite easy to see which is the original basal ray (6), for
it is straight, while the other (o2) is often more or less curved or crooked, and
really forms a pair with the centripetal ray, which is also frequently bent. The
centripetal rays usually lie in close juxtaposition with the centrifugal rays of the
subgastral sagittal triradiates. In a typical example the centripetal ray measured
about 0-13 by 0-01 mm., and the dermal rays about 0-11 by 0-01 mm., but the
centripetal ray may be more elongated. All the rays are more or less gradually
sharp-pointed.

6. Colossal oxea of the dermal cortex (Fig. 8g). These spicules are fusiform
but commonly thicker at one end than at the other. They vary much in size,
up to about 2-85 by 0-075 mm.

7. Hair-like oxea of the oscular fringe. Straight and very slender, only about
0-004 mm. in maximum diameter. Their length is probably about 0-6 mm., but
they are almost invariably broken short in preparations.

Previously known Distribution. Port Elizabeth, South Africa (Bowerbank).

Register Numbers and Localities. I., a number of good colonies from Okhamandal
Point (off Buoy ; 5. 1. 06) ; IV. 6, two small fragments from the S.W. Coast of Beyt
Island.

4. Leucandra donnani Dendy, var. tenuiradiata nov. (Plate L, Figs. 4, 4a,
4b ; Plate II., Figs. 9o-9cZ).

Leucandra donnani Dendy [1905].

There are five specimens in the collection which I think must certainly be referred
to this species, although the curious differences in the proportions of the spicules
make it desirable to give them a special varietal name.

The external form in the best example (R.N. IV. 8, Fig. 4) agrees closely with
that of the type of the species from Ceylon, even in the curiously curved character
of the whole sponge. The colour, however, is white, owing to the absence of
the pigment granules found in the type. The brittle texture is pronounced and
has resulted in the breaking of the specimen in two.

1 Cf. Dendy and Row [1913], p. 750.

MARINE ZOOLOGY OF OKHAMANDAL.

CALCAREOUS SPONGES, PLATE

Arthur Dendy.del

CALCAREOUS SPONGES.

Cambridge University Press.

DENDY— CALCAREOUS SPONGES 87

The skeleton is composed, as in the type, of a thin layer of slender dermal
triradiates (Fig. 9o), a thin layer of slender gastral quadriradiates (Fig. 96), with
some triradiates, and a thick layer of much larger triradiates (Fig. 9c) in the
chamber layer between the two. The latter are quite irregularly arranged. The
dermal and chamber-layer triradiates are perhaps less strongly sagittal than in
the type. There are a few large oxea (Fig. 9d) arranged more or less at right
angles to the surface, but with their outer ends broken off short.

The dermal triradiates (Fig. 9a) are considerably smaller and more slender than
in the type. The gastral quadriradiates (Fig. 96) have more slender rays than
those of the type, but they may be longer. The chamber-layer triradiates (Fig. 9c)
are altogether smaller and of more slender build. The oxea (Fig. 9d), however, are
considerably larger than in the type.

This specimen (R.N. IV. 8) may be taken as the type of the variety.

Another specimen (R.N. IV. 17a, Fig. 4a) is of very irregular, sac-like form,
and has a delicate oscular fringe of hair-like oxea. There was no such fringe in
the type of the species, nor have I been able to find one in R.N. IV. 8, but I
cannot say whether or not its absence is due to abrasion. It may possibly be a
characteristic feature of the variety. R.N. IV. 176 is a damaged specimen closely
resembling R.N. IV. 17«. I have been unable to determine whether or not it has
an oscular fringe. There are indications in R.N. IV. 17a that the outer ends of
the large oxea may be lance-headed.

R.N. III. 5 (Fig. 46) is a small specimen, also of curiously curved form,
attached to a stony Polyzoon colony (Retepora ?) whereby its surface has been
protected from abrasion. The large oxea in this case project far beyond the
dermal surface, and many, if not all, of them are distinctly lance-headed. There
is also a feebly developed oscular fringe of trichoxea.

The chief distinguishing features of the variety appear to be the comparatively
large size of the oxea and the comparatively small size of the triradiates of the
chamber layer. The lance-headed form of the oxea is perhaps a specific character,
which I have reason to believe occurs in the type of the species (in which it is
shown by the broken off outer end of one spicule very clearly).

Register Numbers and Locality. III. 5, off Dwarka ; IV. 8, IV. 9c, IV. 17a and 6,
off S.W. Coast of Beyt Island.

5. Leucandra wasinensis (Jenkin). (Plate I. Fig. 5.)

Leucilla wasinensis Jenkin [1908].
Leucandra wasinensis Dendy [1913].

I refer to this species a single small specimen (Fig. 5), about 4 mm. in height
(exclusive of the oscular fringe of slender oxea, which is very prominent). The

88 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

sponge is a typical leucon person of ovoid form. The canal system is typical.
The spiculation consists of the following elements :—

(1) Rather large dermal radiates, mostly with a well-developed centripetal
apical ray.

(2) Large, irregularly scattered radiates of the chamber layer, mostly, at any
rate, without apical rays.

(3) Slender subgastral sagittal radiates, mostly, if not all, with short apical
rays ; of typical form and arrangement. These are not mentioned by Jenkin in
the case of the type, but probably occur there. I also overlooked them in the
specimen recorded by me [1913] from Saya de Malha, but I find on re-examination
that they undoubtedly occur therein.

(4) Gastral quadriradiates ; a good deal smaller than the derma! radiates and
with well developed, sharp-pointed, curved apical rays projecting into the gastral
cavity.

(5) Quadriradiates of the larger exhalant canals ; perhaps merely radiates of
the chamber layer with short, thorn-like apical rays.

(6) Stout fusiform oxea projecting radially from the dermal surface ; their
outer ends too much broken and corroded for description.

(7) Slender hair-like oxea (trichoxea) of the peristomial fringe.

(8) ? A few rather large microxea at the surface in contact with foreign
objects.

A direct comparison with the type has convinced me, as in the case of the
Saya de Malha specimen, that there is not sufficient difference to justify a specific
separation, especially as there are only single specimens known from the three
localities.

Previously known Distribution. Wasin, East Africa (Jenkin) ; Saya de Malha,
Indian Ocean (Dendy).

Register Number and Locality. XXXV. 9. Off Dwarka.

6. Leucandra dwarkaensis, n.sp. (Plate I. Fig. 6 ; Plate II. Figs. 10a-10e).

The single specimen (Fig. 6) is massive and irregular below, compressed above,
and terminating in a wide, slit-like vent. The entire body is sac-shaped, with
a very wide gastral cavity. Height about 28 mm. ; width in the middle 17 mm.
The outer surface is now nearly smooth, but this is doubtless due to the more or
less complete erosion of the outer ends of the large oxea. The oscular margin is
very thin and shows the remains of a fringe of close-packed, hair-like oxea. A
good deal of foreign material is collected on both dermal and gastral surfaces,
including numerous small siliceous oxeote spicules. Colour in spirit dirty white.

The canal system is typical. There is a rather thick gelatinous ectosome

BENDY— CALCAREOUS SPONGES 89

containing irregular subdermal cavities which lead into the wide inhalant lacunae.
The flagellate chambers are spherical, about 0-087 mm. in diameter and thickly
scattered throughout the choanosome. The gastral cortex is thin ; its inner surface
nearly smooth and pierced by numerous exhalant openings.

The arrangement of the skeleton offers no peculiarities. The dermal skeleton
is composed of slender sagittal triradiates, tangentially arranged, with a few
scattered microxea. The gastral skeleton consists of slender sagittal quadriradiates
and triradiates. The skeleton of the chamber layer consists of much larger and
stouter triradiates, irregularly scattered, and of the inner portions of large oxea
whose outer ends are broken off. There are a few slender subgastral sagittal
triradiates and quadriradiates in the younger parts of the sponge towards the
vent. There is an oscular fringe of trichoxea.

1. Dermal triradiates (Fig. 10a). Sagittal, with long, slender, gradually sharp-
pointed rays. The oral rays usually curved, with wide oral angle. Size varying
a good deal, rays commonly somewhat unequal in length, averaging say about
0-28 by 0-017 mm. The arrangement of these spicules, except for the fact that
they all lie tangentially, is quite irregular.

2. Gastral quadriradiates (Fig. 106). Sagittal, facial rays straighter and rather
more slender than those of the dermal triradiates, but of about the same length
and gradually sharp -pointed. Apical ray moderately long, slender, finely pointed,
nearly straight. As usual these spicules become more regularly arranged and
strongly alate towards the oscular margin, where also the paired rays become
much stouter than the basal ray.

3. Gastral triradiates. Similar to the gastral quadriradiates but without the
apical ray.

4. Subgastral sagittal radiates (Fig. lOc). Rays long and slender, gradually
sharp-pointed. Basal ray much longer than orals, say about 0-32 by 0-015 mm.,
while the orals are only about 0-19 mm. long. Occasionally with a short apical
ray. Arranged as usual with the basal ray directly centrifugally.

5. Triradiates of the chamber larger (Fig. lOd). Approximately regular, with
moderately stout, nearly straight, gradually sharp-pointed rays, varying a good
deal in actual size and in proportions ; measuring say about 0-35 by 0-03 mm.

6. Large oxea (Fig. lOe). With their inner ends deeply implanted in the
chamber layer, or perhaps even projecting into the gastral cavity, and their outer
ends projecting obliquely upwards and outwards from the dermal surface. The
inner ends, where perfect, are gradually and finely pointed ; the outer ends are all
broken off short close to the dermal surface. The perfect spicule must be nearly
straight and very long. The portions remaining in the sponge may measure
nearly 2 mm. in length, with a thickness of about 0-05 mm. These spicules are
very numerous.

90 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

7. Hair-like oxea (trichoxea) of the peristome. These are well developed but
all broken off short, their remaining portions being densely crowded together.

8. Microxea. These seem to have been fairly numerous in the dermal layer,
but, owing perhaps to the fact that the specimen was first preserved in formalin,
they have been extensively eroded and perhaps some of them entirely dissolved
away. They seem to have been hastately pointed at one end.

The most characteristic feature of this species appears to be the external
form, and especially the wide, slit-like vent, but, in the presence of only a single
specimen, it is impossible to say how far this may be constant.

The arrangement of the skeleton and the form and size of the principal
spicules agree closely with the corresponding features in Leucandra donnani var .
tenuiradiata, but in the latter I have found no microxea and no subgastral
sagittal radiates.

Register Number and Locality. XXIII. 6. Off Dwarka, 15-17 fathoms.

LIST OF LITERATURE REFERRED TO.

1873. Bowerbank, J. S. " Contributions to a General History of the Spongiadse," Part IV. (Proc.

Zool. Soc. Land., 1873, p. 17.)
1905. Dendy, A. "Report on the Sponges Collected by Professor Herdman at Ceylon in 1902."

(Reports on the Pearl Oyster Fisheries of the Gulf of Manaar, Vol. III., Royal Society,

London.}
1913. Dendy, A. "Report on the Calcareous Sponges collected by the Sealark Expedition in the

Indian Ocean." (Trans. Linn. Soc. Land. Zoology. Vol. XVI.)
1913. Dendy, A., and Row, R. W. H. "The Classification and Phylogeny of the Calcareous

Sponges, with a Reference List of all the described Species, systematically arranged."

(Proc. Zool, Soc. Land., Sept., 1913.)
1908 A. Jenkin, C. F. "The Calcareous Sponges." (The Marine Fauna of Zanzibar and British

East Africa, from Collections made by Cyril Crossland, M.A., in the Years 1901 and

1902. Proc. Zool. Soc. Land., 1908).
1908 B. Jenkin, C. F. " Porifera Calcarea." (National Antarctic Expedition. Natural History.

Vol. IV.)
1909. Row, R, W. H. "Report on the Sponges Collected by Mr. Cyril Crossland in 1904-1905.

Part I. Calcarea." (Reports on the Marine Biology of the Sudanese Red Sea. Journ.

Linn. Soc. Lond. Zoology. Vol. XXXI.)

DENDY— CALCAREOUS SPONGES 91

DESCRIPTIONS OF PLATES.

PLATE I.

Fig. 1. Sycon yrantioides n.sp. (R.N. III. 4). X 4.

Figs. 2, 2a. Grantessa hastifera (Row) (Parts of R.N. III. 2). X 2.

„ 3, 3a, 36. ffeteropia glomerosa (Bowerbank) (Parts of R.N. I.). X 2.
Fig. 4. Leucandra doniiani Dendy, var. tenuiradiata nov. (R.N. IV. 8). X 3£.

„ 4«. Leucandra donnani Dendy, var. tenuiradiata nov. (R.N. IV. 17o). o, vent; X 3§.

,, 46. Leucandra donnani Dendy, var. tenuiradiata nov. (R.N. III. 5). X 4.

„ 5. Leucandra wasinensis (Jenkin) (R.N. XXXV. 9). X 5^.

„ 6. Leucandra dtvarkaensis n. sp. (R.N. XXIII. 6). X If.

PLATE II.

Figs. 7a-7/". Grantessa hastifera (Row) (R.N. III. 2).

Fig. la. Dermal triradiates. X 90.

,, 76. Gastral triradiates. X 90.

.„ 7c. Subgastral sagittal triradiates. X 90.

„ Id. Tubar triradiates. X 90.

„ If. Subdermal pseudosagittal triradiates. X 90.

„ If. Large oxeote, front view. x 90.

„ 7,/". do. do. side view. x 90.

„ If". do. do. outer end, front view. x 360.

Figs. 8a-8ff. ffeteropia glomerosa (Bowerbank) (R.N. I.).

Fig. 8a. Dermal triradiates. X 166.
,, 86. Gastral triradiates. X 166.
,, 8c. Gastral quadriradiate, a. r., apical ray. x 166.
„ 8d. Subgastral sagittal, and 8e, subdermal pseudosagittal triradiate in situ ; the outer paired (oral)

ray of the latter being broken off. b, basal ray ; o1, centripetal paired ray ; o2, outer

paired ray. x 166.

,, 8/. Complete subdermal pseudosagittal triradiate ; lettering as before. X 166.
„ 8g. Dermal oxea. X 50.

Figs. 9a— 9c£ Leucandra donnani (Dendy) var. tenuiradiata nov. (R. N. IV. 8).

Fig. 9a. Dermal triradiate. X 90.

„ 96. Gastral quadriradiate. X 90.

„ 9c. Triradiates of the chamber layer. X 90.

„ M. Oxeote. x 90.

Figs. lOa-lOe. Leucandra dwarkaensis, n. sp. (R.N. XXIII. 6).

„ lOa. Dermal triradiate. X 100.

„ 106. Gastral quadriradiate. X 100.

„ 10c. Subgastral sagittal triradiate (broken). X 100.

,, lOrf. Triradiates of the chamber layer. X 100.

,, I Of. Inner end of large oxeote. x 100.

REPORT

ON THE

NON- CALCAREOUS SPONGES

COLLECTED BY

MR. JAMES HORNELL

AT

OKHAMANDAL IN KATTIAWAR IN 1905-6.

BY

ARTHUR BENDY, D.Sc., F.R.S.

Professor of Zoology in the University of London (King's College).

[With Four Plates.]

MR. HORNELL'S collection of Non-Calcareous Sponges contains about fifty-eight
species, of which I have been able to identify no fewer than forty-two with previously
described forms and to describe fifteen as new. Leaving out of account numerous
fragments which were not sufficiently well-preserved for identification, the following
is a complete list of the species represented. It will be observed that one species
(Higginsia sp.) has only been generically identified, but the genus is of sufficient interest
to make it worth recording.

ORDER TETRAXONIDA.
SUB-ORDER ASTROTETRAXONIDA.

FAMILY STELLETTID^E.

1. Myriastra (Pilochrota) haeckeli Sollas.

2. Jaspis reptans (Dendy).

3. Asteropus simplex (Carter).

94 OKHAMANDAL MARINE ZOOLOGY REPORT— PART I

FAMILY GEODIIDJE.

4. Geodia variospiculosa Thiele.

FAMILY DONATIID^.

5. Donatia seychellensis (Wright).

6. Tuberella aaptos (Schmidt).

FAMILY CHONDKOSIID.E.

7. Chondrilla australiensis Carter.

8. Chondrilla agglutinam n. sp.

SUB-ORDER SIGMATOTETRAXONIDA.

FAMILY TETILLID.E.

9. Tetilla dactyloidea (Carter).

10. Tetilla hirswta Dendy.

11. Tetilla pilula n. sp.

12. Tetilla barodensis n. sp.

*

FAMILY HAPLOSCLERID^E.

13. Gellius fibulatm (Schmidt) var. microsigma nov.

14. Gellius ridleyi (Hentschel).

15. Gelliodes fibrosa Dendy.

16. Reniera permollis (Bowerbank).

17. Reniera topsenti Thiele.

18. Reniera hornelli n. sp.

19. Reniera fibroreticulata n. sp.

20. Reniera semifibrosa n. sp.

21. Halichondria panicea Johnston var.

22. Halichondria reticulata Baer.

23. Siphonochalin-a crassifibra Dendy.

24. SiphonocJialina minor n. sp.

25. Desmacella tubulata Dendy.

26. Thrinacophora cervicornis Ridley and Dendy.

27. Axinella virgultosa Carter.

28. Phakellia donnani (Bowerbank).

29. Auletta lyrata var. glotnerata Dendy.

30. Auletta elongata var. fruticosa nov.

DENDY— NON-CALCAREOUS SPONGES 95

31. Ciocalypta dichotoma n. sp.

32. Higginsia sp.

FAMILY DESMACIDONID^E.

33. Esperella plutnosa (Carter).

34. Desmacidon minor n. sp.

35. lotrochota baculifera Ridley.

36. Guitarra indica n. sp.

37. Psammochela elegans n. gen. et sp.

38. Chondropsis kirkii (Carter).

39. Myxilla arenaria Dendy.

40. Clathria comllitincta Dendy.

41. Clathria spiculosa Dendy.

42. Echinodictyum gorgonioides n. sp. .

43. Raspailia fruticosa var. tenuiramosa Dendy.

44. Acarnus tortilis Topsent.

45. Bubaris radiata n. sp.

FAMILY SPIRASTRELLID.E.

46. Spirastrella vagabunda var. tubulodigitata Dendy.

47. Placospongia carinata (Bowerbank).

FAMILY CLIONID^;.

48. Cliona coronaria (Carter).

FAMILY SUBERITID/K.

49. Suberites carnosus (Johnston) var.

50. Suberites flabellatus Carter.

51. Suberites cruciatus Dendy.

52. Polymastia gemmipara n. sp.

ORDER EUCERATOSA.
FAMILY APLYSILLIDJE.

53. Megalopastas retiaria n. sp.

54. Danvinella australiensis Carter.

FAMILY SPONGELIID^;.

55. Spongelia fragilis var. ramosa (Schulze).

56. Spongelia cinerea (Keller).

57. Spongelia elegans (Nardo) var.

FAMILY

58. Hippospongia dathrata (Carter).

o 2

96 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

The fact that it has been possible to identify seventy-four per cent, of these species
with previously described forms clearly indicates the progress that has been made
in recent years in our knowledge of the Sponges of the Indian Ocean, especially when
we consider that Mr. HornelPs collection was made in a locality from which, so far
as I am aware, no sponges have hitherto been recorded.

As might naturally be expected, a large proportion of the previously known
species are identical with more or less well-known Ceylon forms [cf. Dendy
1905]. The commonest species in the collection is Esperella plumosa (Carter),
which attains a large size and has a very fully developed and very beautiful
spiculation. Other common and characteristic Indian Ocean species are Myriastra
haeckeli, Donatia seychellensis, Chondrilla australiensis, Tetilla dactyloidea, Tetilla
hirsuta, Phakellia donnani, Aulelta lyrata, lotrochota baculifera, Clathria comllitincta,
Clathria spiculosa, Raspailia fruticosa, Spirastrella vagabunda, Placospongia carinata
and Hippospongia clathrata.

Of the fifteen new species, Tetilla pilula, T. barodensis, Guitarra indica, Psam-
mochela elegans (for which a new genus is proposed), Polymastia gemmipara and
Megalopastas retiaria may be mentioned as exceptionally interesting forms.

The scarcity of true Horny Sponges (Euceratosa) in the collection is remark-
able. There is no true bath sponge and, indeed, only one representative of the
family SpongiidaB, viz., the common but useless Hippospongia clathrata. On the
whole, however, the Sponge-Fauna of Okhamandal is undoubtedly a rich one, and
I am very glad to have had the opportunity of investigating it.

All the specimens in the collection seem to have come from shallow water, the
greatest depth recorded being seventeen fathoms. A considerable number of them
were growing upon the large, branching, parchment-like tubes of a polychsete
worm,* which appears to be extremely common. It is unfortunate that so many
of the specimens were originally preserved in formalin, a medium which is
entirely unsuited for sponges and in which they undergo extensive maceration.

As regards the classification employed in this Eeport it will be observed that
certain innovations have been introduced. The Axinellidse are included in the
Haploscleridse, and the Spirastrellidse, Clionidse and Suberitidaa are placed in the
Sigmatotetraxonida. I cannot attempt to justify these changes in this place,
but must refer the reader to my Reports on the Sponges of the Sealark Expedition,
now in course of publication, and to future publications by myself and my
colleague, Mr. R. W. H. Row, in which we hope to discuss the question of the

* I am indebted to Dr. J. H. Ashworth and Professor Fauvel for the information that this worm
is a species of the genus Eunice, possibly Eunice tubifex Crossland. [A photograph of a forest of the
tubes of this worm draped with masses of zoophytes and polyzoa, as seen at extreme low water on
the Kiu littoral, Beyt harbour, is reproduced as Plate VI. in "Marine Resources of Okhamandal,"
in Part I. of this Report. — J. H.]

BENDY— NON-CALCAREOUS SPONGES 97

classification of the Non-Calcarea from the phylogenetic point of view in some
detail.

My account of the six species of Calcareous Sponges collected by Mr. Hornell
has already been published [Dendy 1915].

1. Myriastra (Piloehrota) haeckeli Sollas.

Pilochrota haeckeli Sollas [1888].
Stelletta haeckeli Lendenfeld [1903].
Pilochrota haeckeli Dendy [1905].

There are in the collection nineteen specimens which I refer to this species.
The smallest are approximately spherical and no larger than a pea. The largest
is irregular, like a potato ; measures about 45 mm. in length by 33 mm. in
transverse diameter, and has three vents. They closely resemble the specimens
collected by Professor Herdman at Ceylon, but the thin, membranous lip of the
vent appears to be devoid of oxeote spicules.

The largest specimen (R.N. IV. 3) shows to a very marked degree the curious
abnormality of some of the trisenes which I described and figured in the case
of the Ceylon material. The reduction of the rays, however, is carried to such an
extent in this specimen that in extreme cases the entire spicule is reduced to a
perfectly spherical ball of concentrically laminated silica (opal). A precisely
similar modification of tetract megascleres occurs in the two known species of the
genus Yodomia [Lebwohl 1914, and Dendy 1916].

Previously known Distribution. Philippine Islands (Sollas) ; Ceylon (Dendy).

Register Numbers, Localities, &c. II. 5 (four small specimens), off Poshetra, Jan. 7,
1906 ; IV. 3 (two large specimens), IV. 9 a (seven small specimens), off S.W.
coast of Beyt Island; XIV. (small fragment), off S.W. of Beyt, 6.1.06; XV. 2
(four specimens, varying greatly in size), three miles W.N.W. of Samiani Light-
house, 17 fms., 22.12.05 ; XX. 9 (one specimen), Adatra.

2. Jaspis reptans (Dendy).

Coppatias reptans Dendy [1905].

This species is represented in the collection by three specimens, of which it is
possible that R.N. XX. 4 and R.N. XX. 7 may be parts of the same. The shape is
extremely irregular. R.N. XX. 4 is a flattened, cake-like fragment, measuring about
35 by 25 mm., with a maximum thickness of 12 mm., and a rounded margin except
where broken. R.N. XX. 7 is a very irregular fragment of about the same size, throwing
off irregular digitiform processes. Both have a coarse, firm, harsh consistency and are
of a greyish-fawn colour in spirit (having been first preserved in formalin). R.N.
XXXIII. 2 b is a massive but flattened specimen measuring about 50 by 50 mm., and

9K OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

completely overgrown by a specimen of Reniera semifibrosa. It has now, in alcohol,
a slightly pinkish tint, having also been originally preserved in formalin.

These specimens differ from the Ceylon type in the less abundant development of
pigment. The oxea are a good deal more robust and longer, especially in R.N. XX. 4
and 7 ; E.N. XXXIII. 2 b being intermediate in this respect. The asters also appear
to be somewhat larger, though still very minute.

Previously known Distribution. Ceylon (Dendy).

Register Numbers and Localities. XX. 4, 7, Adatra ; XXXIII. 2 b, Dhed Mora and
adjacent rocky ground between Beyt and Aramra, 1 fm., 21.12.05.

3. Asteropus simplex (Carter).

Stellettinopsis simplex Carter [1879, 1886].
Asteropus simplex Sollas [1888].
Asteropus haeckeli Dendy [1905].
Asteropus simplex Hentschel [1909].
Asteropus simplex Dendy, [1916].

This species is evidently an " epipolasid " form from which the trisenes have
completely disappeared. It was first described by Carter from Fremantle, Australia,
and in the same paper that author also recorded it from Hayti. He subsequently
recorded it again from Victoria, Australia, where it was collected by Mr. J. B. Wilson.

The Okhamandal specimen forms a mass of sponge cementing together and filling
the interstices in an agglomeration of shells (chiefly Siliquaria) and other debris. It
is now a very light pinkish-grey in colour, having been first preserved in formalin, and
in this respect differs from, at any rate, the Victorian and Haytian specimens, which
contain large, dark brown pigment-cells in the more superficial part of the sponge. It
also differs from Carter's Australian specimens in the more robust character of the oxea,
while the Haytian specimen, of which the type slide is in my possession, appears to be
to some extent intermediate in this respect. The spiculation of the Okhamandal
sponge is as follows : —

(1) Oxea ; stout, curved, fusiform, sharply pointed ; measuring about 1-7 by
0-07 mm. (or even more) when full-sized, but often smaller. The main skeleton is a
confused reticulation of these spicules.

(2) Oxyasters ; with small centrum (if any), and rather few (up to about ten) slender
rays, which sometimes seem to be very slightly roughened ; total diameter about
0-03 mm. These spicules have only about half the diameter of those of the type as
given by Carter ; they also seem to be very local in their distribution, so that they may
easily be overlooked if only one sample is examined.

(3) Sanidasters (the " sceptrelliform " spicule of Mr. Carter's description).
Extremely numerous, especially in the dermal membrane ; about 0'02 mm. long, with
slender axis and rather few, moderately long, irregularly arranged, slender spines.

BENDY— NON-CALCAREOUS SPONGES 99

It is interesting to speculate as to whether or not this widely distributed form has
arisen polyphyletically by reduction of species of Ancorina in the different localities
where it occurs, or whether such reduction has taken place only once in one locality
and been followed by extensive migration. In the present state of our knowledge this
question cannot, of course, be settled.

Previously known Distribution. West Australia (Carter) ; Victoria (Carter) ;
Hayti (Carter) ; South West Australia (Hentschel) ; Cargados Carajos (Dendy).

Register Number and Locality. V. 2, S. of Chindi Keef, 6-10 fms., 18.12.05.

4. Geodia variospiculosa Thiele.

Geodia variospiculosa Thiele [1898].

Geodia variospiculosa var. clavigera Thiele [1898].

Geodia variospiculosa Lendenfeld [1903].

Geodia variospiculosa var. typica Lendenfeld [1910].

Geodia variospiculosa var. intermedia Lendenfeld [1910].

Geodia variospiculosa var. micraster Lendenfeld [1910].

Geodia variospiculosa var. aapta Lebwohl [1914].

Although this species has hitherto been recorded only from Japanese waters, I
have no hesitation in referring to it the only specimen of a geodiid sponge in
Mr. HornelPs collection, and in support of my identification I furnish the following
details.

The sponge is almost spherical, about 14 mm. in diameter, and has evidently
been attached to the substratum by one side. The surface is almost smooth, but
slightly uneven. No pores or vents are visible under a pocket lens. The colour
in alcohol (after formalin) is very light grey.

There is a thin external fur of small, radially disposed oxea, containing also
the cladi of protria3nes (and possibly anatrieenes). This rests upon the cortical
layer of sterrasters, which is about 0-26 mm. thick. The main choanosomal skeleton
consists of dense radial bundles composed of large oxea and of the shafts of
trisenes, whose cladi are for the most part extended just beneath the cortical
layer of sterrasters.

Spiculation. (1) Long, slender oxea of the choanosome ; straight or nearly so,
gradually and finely pointed at each end, measuring about 2-0 by 0-028 mm.

(2) Short oxea of the surface fur ; fusiform, almost straight, almost stylote, with
.the narrow inner end slightly rounded off; measuring about 0-2 by 0-0082 mm.

(3) Orthotriaenes ; with long, straight shaft tapering gradually to a slender
point, and simple, conical cladi ; shaft about 1-8 by 0-05 mm., with cladi about
0-24 by 0-034 mm.

(4) Dichotrisenes ; resembling (3) but with cladi once bifurcate.

(5) Mesoprotrisenes ; shaft very long and slender, say about 3-7 by 0-017 mm. ;

100 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

cladi sharp-pointed, approximately equal in length, about 0-07 by 0-0086 mm. ;
prolongation of shaft sharp-pointed and almost equal in length to cladi ; cladi
sometimes irregular.

(6) Anatrisenes ; cladi sharply recurved, sharply pointed ; occasionally split
at the apex into two almost parallel branches. Dimensions much the same as
for protrieenes but shaft rather more slender.

(7) Sterrasters ; of normal form, with well-marked hilum ; elliptical, measuring
about 0-08 by 0-065 mm.

(8) Subcortical spherasters, with numerous fairly long, sharp-pointed rays,
total diameter about 0-012 mm. Not sharply distinguishable from (9).

(9) Oxyasters ; with few or fairly numerous, slender, perhaps faintly roughened
rays ; total diameter varying up to about 0-05 mm.

(10) Minute chiasters or strongylospherasters of the dermal layer, about 0-006 mm.
in diameter.

Previously known Distribution. Japan (Thiele, Lendenfeld, Lebwohl).
Register Number and Locality. XV. 1, three miles W.N.W. of Samiani Light-
house, 17 fms., 22.12.05.

5. Donatia seychellensis (Wright).

Alemo seychellensis Wright [1881].
Tethya seychellensis Sollas [1888].
Teihya seychellensis Keller [1891].
Tethya seychellensis Topsent [1893].
Tethya ingalli (pars) Lindgren [1898].
Tethya seychellensis Kirkpatrick [1900].
Tethya lyncurium var. c. Dendy [1905].
Donatia Ingalli Topsent [1906].
Donatia Ingalli (pars) Hentschel [1909].
Tethya seychellensis Row [1911].
Donatia seychellensis Dendy [1916].

The three specimens in the collection agree so closely with those obtained by
the Sealark Expedition in the Indian Ocean and dealt with by me in my Report
[1916] that it is unnecessary to describe them in this place. Two of them
(R.N. XXVI. 8 a, b) are well provided with buds and are also remarkable for the
presence of immense numbers of oscillatorian algae in the cortex, which I have
rarely seen before in a Donatia. In all the specimens the sex-radiate condition of
the tylasters and large oxyasters is very strongly pronounced though by no
means constant, and the latter frequently have branching rays.

Previously known Distribution. Seychelles (Wright) ; Samboangan (Sollas) ;
Flinders Passage, Torres Straits (Sollas) ; Red Sea (Keller, Row, Topsent) ; ? South

DENDY— NON-CALCAREOUS SPONGES 101

West Australia (Hentschel) ; Gulf of Mannar, Praslin Reef, Egmont Reef and
Salomon, Indian Ocean (Dendy) ; Xmas Island (Kirkpatrick).

Register Numbers, Localities, &c. II. 9, off Poshetra, 7 January, '06. ; XXVI.
8 a, b, Adatra Reefs, 25 December, '05.

6. Tuberella aaptos (Schmidt).

(For Literature and Synonymy, vide Topsent [1900]).

There is one very typical specimen of this curious sponge in the collection.
It is irregularly tuberous, elongated, with an uneven surface beset here and there
with small papillae, some of which have each a small vent. The texture is
compact but fairly compressible, the colour in alcohol (after formalin) brown.
Length of specimen about 60 mm., greatest breadth 32 mm., greatest thickness
21 mm.

The main skeleton consists of loose bundles of large strongyloxea radiating
towards the surface, with scattered spicules between. The dermal skeleton consists
of dense brushes of small styli with outwardly directed apices. The large strongyl-
oxea measure about 1-1 by 0-034 mm. ; the small styli about 0-26 by 0-0086 mm.

This specimen agrees very closely in all respects with the description and
figures given by Topsent, whose views as to the correct generic name and
synonymy I accept provisionally. It seems possible, however, that Gray's generic
name Aaptos [1867] may have to be revived, and also that Keller's Tuberella
tethyoides [1880] may, after all, be a distinct species.

Previously known Distribution. Mediterranean (Schmidt, Lendenfeld, Topsent) ;
Gulf of Mexico (Topsent) ; S.W. Australia (Hentschel) ; Aru Islands (Hentschel).

Register Number, Locality, &c. VI., S. of Chindi Reef, 6-10 fins., 18.12.05.

7. Chondrilla australiensis Carter.

Chondrilla australiensis Carter [1873].
Chondrilla australiensis Lendenfeld [1886].
Chondrilla australiensis Lindgren [1898].
Chondrilla australiensis Dendy [1905].
Chondrilla australiensis Hentschel [1909 and 1912].
Chondrilla australiensis Dendy [1916].

Several pieces of considerable size, representing one or more flat, spreading crusts,
were obtained from Adatra Reef. They appear to have been originally preserved
in formalin, but are now in alcohol, and exhibit the usual light brown colour
of the species.

Previously known Distribution. Port Jackson, E. Coast of Australia (Carter, Lenden-
feld) ; Sharks Bay, S.W. Australia (Hentschel) ; Coast of Cochin China (Lindgren) ;
Ceylon, Cargados Carajos, Amirante, Seychelles (Dendy) ; Aru Islands (Hentschel).

102 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

Register Number, Locality, &c. XXVI. 13, Adatra Reef, 25 December '05.

8. Chondrilla agglutinans, n. sp.— (Plate I., Figs, la, 16.)

The sponge occurs as a sort of matrix, holding together and partially enveloping
a mass of Siliquaria and other shell-fragments, together with sand and pebbles. The
surface is glabrous and approximately smooth except where wrinkled by contraction
in spirit. The colour in spirit is chocolate-brown of varying shades. Oscula small,
in irregular groups, their margins flush with the general surface.

The cortex is barely 0-1 mm. thick, and is covered by a very distinct cuticle. Even
in stained paraffin sections I have been unable to detect the inhalant cortical canals,
so conspicuous in the closely related C. mixta [Schulze 1877]. It is distinctly fibrous and
also contains numerous pigment-cells filled with minute brown granules. Its outer
half is densely packed with spherasters arranged in several layers.

The colour of the sponge appears to be due chiefly to the presence in the choanosome
of numerous spherical cells of a brown colour, scattered singly and in dense groups.
Each of these brown cells is about 0-01 mm. in diameter. Their colour may, however,
possibly be due to staining by colouring matter extracted by alcohol from the pigment-
cells. They remind one of the fat-like bodies described by Schulze [1877] as possible
reserve-material in Chondrosia reniformis, and of the similar bodies described by Carter
[1887 bis] in his Chondrosia spurca, and of the bodies regarded as possible symbiotic algae
by myself [1905] in Hexadella indica, &c.

There are two kinds of aster present, closely resembling those of Chondrilla mixta,
viz. spherasters and oxyasters. The former (Fig. la) occur chiefly in the outer part of
the cortex, but also sparingly in the inner part of the cortex, and still more sparingly
in the choanosome. They have a very large centrum and numerous smooth, sharp,
conical rays touching each other at their bases ; the total diameter is about 0-028 mm.
The oxyasters (Fig. 16) appear to be confined to the choanosome, where they are very
sparingly scattered. They have a small or indistinguishable centrum, and comparatively
few, rather slender, smooth, sharp-pointed, conical rays, say about eight or ten in
number. The total diameter of this spicule is about 0-02 mm.

Register Number, Locality, &c. V. 1, 8. of Chindi Reef, 6-10 fms., 18.12.05.

9. Tetilla dactyloidea (Carter).— (Plate II., Figs. lOa-lOc.)

Tethya dactyloidea Carter [1869].

Tethya dactyloidea Carter [1872].

Tethya dactyloidea Carter [1887].

Tetilla dactyloidea Sollas [1888].

Teiilla dactyloidea Keller [1891].

Tetilla dactyloidea var. lingua Annandale [1915].

DENDY— NON-CALCAREOUS SPONGES 103

There are five specimens of this sponge in the collection. They are all subcylin-
drical, provided with a single vent at the upper extremity and a root-tuft of long,
silky spicules at the lower (cf. Figs. lOa-lOc). The general surface is smooth and porous,
the consistence soft and compressible, the colour in alcohol light grey. The smallest
specimen measures about 19 mm. in height by 5-5 mm. in diameter ; the largest 38 mm.
by 13 mm., in both cases excluding the root-tuft. In one specimen the root-tuft extends
downwards for 18 mm. before meeting the mass of sand-grains with which it is still
in connection. In the other specimens a similar mass of sand -grains is attached by the
root-tuft close to the lower extremity of the sponge.

In one of the larger specimens, which I cut open, the vent forms the terminal
aperture of a cylindrical cloacal chamber about 10 mm. in length and 2-5 mm. in diameter,
into which numerous larger and smaller exhalant canals open at various levels, the
larger ones being continuations of the cloacal cavity deep down into the body of the
sponge (Fig. lOa). Just below the point where the larger canals, coming from below,
join to form the cloaca! cavity, lies the so-called " nucleus," from which the principal
fibres of the skeleton radiate, mostly in a downward direction. Curving gently outwards,
these fibres break up, over the general surface, into dense surface-brushes, which,
however, do not project sufficiently to render the surface hispid to the naked eye. At
the lower extremity of the sponge they are continued downwards, outside the sponge-
body, to form the root-tuft.

Throughout the greater part of the sponge-body the fibres appear to be composed
exclusively of very long and very slender oxea, and they are crossed at various angles
by irregularly scattered oxea of similar form, but perhaps shorter.

The dense surface-brushes are composed mainly of long, slender oxea, but mingled
with these occur many protrieenes with very slender shaft and almost hair-like cladi
of unequal length.

The anatria3nes appear to be confined to the lower parts of the sponge, where they
seem to form the principal constituents of the descending fibres both inside the sponge
and in the root-tuft. Their shafts are very long and slender, hair-like, and their cladomes
unusually well developed, with sharp, strongly recurved cladi. Protrisenes also occur
in the root-tuft, but less abundantly than the anatria3nes. The microscleres are minute,
slender, contort sigmata, of the ordinary Tetilla type, not very abundant.

This interesting and easily recognisable sponge appears to be characteristic of
sandy and muddy flats along the shores of the Indian Ocean. Sollas has expressed
a doubt whether the specimen from the Mergui Archipelago identified by Carter as
belonging to this species is really specifically identical with the type from the S.E.
coast of Arabia, but I do not think it at all likely that there is a specific difference.
On the other hand, I have myself [1905] described a distinct, but closely-related species
(T. limicola) from Ceylon, differing from T. dactyloidea in external form and in the
arrangement of the exhalant canal-system. Annandale has recently [1915] described

104 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

a variety of T. dactyloidea (var. lingua) from the Chilka Lake, growing in
fresh water.

Previously knoivn Distribution. S.E. coast of Arabia, on shallow sandy bottom
near shore (Carter) ; ? on the sandy bottom of the Mahim Estuary, off the Island of
Bombay (Carter)1 ; King Island, Mergui Archipelago (Carter) ; Chilka Lake, Bay of
Bengal (Annandale).

Register Number and Locality. II. 2 a-e. From muddy shore, Balapur, Jan.
1906.

10. Tetilla hirsuta Dendy.
Tetilla hirsuta Dendy [1889].
Cinachyra hirsuta Lendenfeld [1903].
Tetilla hirsuta Dendy [1905].

I identify with this species two specimens (R.N. XIX., 5, 6), neither of which is
in a very good state of preservation and neither of which shows the arrangement of
the inhalant and exhalant apertures. They agree closely with the type as regards
spiculation, but the triaenes are very scarce.

R.N. XXI. 5 may also possibly belong to this species. It contains, however,
numerous small, scattered oxea, and may possibly be a specimen of T. poculifera Dendy
[1905]. It is, however, a mere fragment and cannot be safely identified.

Previously known Distribution. Gulf of Mannar and Ceylon coast (Dendy).

Register Number and Locality. R.N. XIX., 5, 6, Vamiani Point, January 5, '06 ;
? R.N. XXL, 5, off Rupan Bander and Kutchegudh, 4-7 fms., 8.12.05.

11. Tetilla pilula n. sp.— (Plate I., Figs. 2o-2c.)

There are three specimens of this fascinating little sponge in the collection,
all closely resembling one another in external form and microscopic details of
structure. The form is spherical, with a single small vent surrounded by a
distinct spicular, membranous collar. The largest specimen (R.N. IV. 9&) measures
only about 5 mm. in diameter, the other two only about 3 mm. The surface
is minutely conulose, not visibly hispid, and there is no root-tuft. The texture
is rather soft and compressible, and the colour in spirit pale greyish-yellow.

The arrangement of the skeleton is very strongly radial. Dense and closely-
placed bundles of slender oxea and trisenes radiate outwards from a central
" nucleus," while the hair-like shafts of the tria3nes are often collected together
in wavy fibres which also diverge from the centre of the sponge. As they
approach the surface the spicule-bundles spread out very gradually into surface-
brushes (Fig. 2a) composed of oxea, protrisenes and anamonames, the apices of

1 The evidence of the specific identity in this case is insufficient.

DENDY— NON-CALCAREOUS SPONGES 105

the more distally placed oxea and the cladomes of the more distally placed trisenes
and monsenes projecting slightly beyond the surface. The spicule-bundles are
crossed at various angles by loosely scattered oxea.

There is no cortex and (in unstained preparations) no visible distinction
between ectosome and choanosome.

Spicules. (1) Oxea ; straight, slender, fusiform ; gradually and finely pointed
at each end ; measuring about 0-85 by 0-012 mm.

(2) Protriaenes (sometimes disenes ?) ; with equal or unequal, slender, straight,
sharp-pointed cladi, and very long, slender shaft tapering off into a fine hair.
Shaft measured up to 1-1 mm. in length with a thickness of 0-004 mm. (near the
cladal end), and cladi about 0-033 by 0-002 mm. Both shaft and cladi are
sometimes of hair-like fineness.

(3) Anamonsenes (Fig. 2fo) ; shaft measured up to 2-5 mm. in length (and then
probably broken off) ; of hair-like fineness throughout the greater part of its
length, attaining a thickness of 0-008 mm. immediately below the cladome ;
the single cladus sharply recurved, gradually and sharply pointed, measuring
about 0-025 by 0-008 mm. (at the base) in a well -developed example.

(4) Slender, spirally twisted (contort) sigmata (Fig. 2c), of the usual Tetilla
type ; measuring about 0-009 mm. in a straight between extreme points. Very
numerous throughout the sponge.

The most characteristic spicules of this species are undoubtedly the ana-
monsenes, and it appears to me a very noteworthy fact that, although these
occur in immense numbers, I have not met with a single anatrieene or anadisene.
The only other known species in which the anatrisenes appear to be represented
exclusively by anamonsenes is, so far as I am aware, Tetilla pedifera Sollas
[1888].

In Tetilla pedifera, however, there are, according to Sollas, no sigmata, nor
any other form of microsclere, an unusual feature which serves at once to distin-
guish it from T. pilula. On account of this character Lendenfeld [1903] has
included T. pedifera in his genus Tethyopsilla, but it is evidently very closely
related to T. pilula, and it may well be doubted whether Tethyopsilla is a
monophyletic genus.

In Tetilla coronida Sollas [1888] anamonsenes occur together with anatrisenes.

In Cinachyra hamata Lendenfeld [1906] anamonocnes sometimes occur alone
and sometimes associated with a few anatrisenes.

Register Numbers and Localities. IV. 9 b, dredged off S.W. coast of Beyt
Island ; XXXV. 8 a, b, dredged off Dwarka, January '06.

12. Tetilla barodensis n. sp.— (Plate I., Figs. 3«-3d).

The single specimen in the collection is approximately spherical and about

106 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

13 mm. in diameter. The surface is not strongly hispid, but is covered with an
encrustation of sand-grains between the projecting ends of megascleres. Three
small mammiform projections, terminally fringed with projecting spicules, pre-
sumably indicate vents ; another smaller one, without a terminal fringe of
spicules, may bear inhalant pores. Colour internally pale greyish-yellow.

There is a well-developed, dense cortex, about 0-26 mm. thick, full of granular
cells and perhaps to some extent fibrous.

The skeleton radiates from a very dense central " nucleus " and consists of
well-defined bundles of oxea and triaenes, the bundles being separated from one
another by fairly wide intervals free from megascleres.

The cladomes of the orthotrieenes and anatrisenes for the most part he in the
cortex, but some of them project, along with the ends of some of the oxea,
into the encrusting layer of sand and foreign spicules. I have seen no protriaenes,
neither in sections nor in boiled-out preparations, but it is quite possible that their
ends lie outside the cortex and have all been broken off.

Spicules. (1) Oxea ; very long, straight, slender, fusiform, tapering very
gradually to a fine point at each end, but one end may taper less gradually
than the other ; size up to about 2-9 by 0-05 mm.

(2) Orthotriaenes (Fig. 3 a) ; shaft tapering very gradually to a long-drawn-out,
hair-like extremity ; cladi simple but often irregularly bent, normally conical,
sharp-pointed, gently recurved. Dimensions of a typical example : shaft 1-6 by
0-03 mm. (thickness just below cladome), cladi about 0-17 by 0-025 mm. In
another example, of different proportions, the shaft measured about 1-0 by
0-037 mm. and the cladi 0-26 by 0-033 mm. Sometimes the hair-like portion of the
shaft is abbreviated and abruptly truncated.

(3) Anatriaenes (Fig. 3 b) ; shaft very long and slender, hair-like, measured
up to about 2-75 by 0-007 mm. (thickness just below cladome) ; cladi slender,
sharp-pointed, about equal, recurved not very abruptly, measuring about 0-06
by 0-006 mm.

(4) Sigmata (Fig. 3 c) ; very slender, spirally curved (contort), measuring about
0-012 mm. in a straight line between extreme points ; of the typical tetillid form ;
extremely numerous in the choanosome.

(5) Trichodragmata (Fig. 3 d) ; bundles of long, slender, hair-like raphides ; the
entire bundle may measure up to about 0-13 by 0-028 mm. Abundantly scattered
through the choanosome. Usually they contain fewer raphides and are therefore
more slender than the specimen measured.

The most characteristic feature of this species is the presence of the tricho-
dragmata, which, so far as I am aware, are known to occur in only one other
species of Tetillidge, viz. Cinachyra eurystoma Keller [1891].

Register Number and Locality. XXIII. 8, off Dwarka, 15-17 fins., 12.12.05.

DENDY— NON-CALCAREOUS SPONGES 107

13. Gellius fibulatus (Schmidt) var. microsigma nov.

Reniera fibulata Schmidt [1862].
Reniera fibulifera Carter [1880].
Gellius fibulatus Topsent [1892].
Gellius fibulatus Dendy [1905].
Gellius ridleyi (pars) Hentschel [1912].

Very little is known about the external form of the European type. Schmidt,
however, says that it is not easily distinguishable from small specimens of Reniera alba.
The latter is described (loc. cit.) as a shapeless crust between the branches of Clathria
and in other situations, with single vents on the summit of short, projecting tubes. This
description applies very well to a small specimen in Mr. Hornell's collection. It consists
of a thin crust growing over a mass of calcareous debris and giving off three short processes
of unequal length, the longest measuring about 10 by 3-5 mm. The processes are
hollow, and each bears a terminal vent.

The specimen seems to differ from the European form only in the small size of
the sigmata. The oxea measure about 0-25 by 0-01 mm., which seems to agree very
well with the type, but the sigmata measure only about 0-0164 mm. from bend to
bend.

Previously known Distribution of Species. Adriatic (Schmidt) ; North Atlantic
(Topsent) ; Gulf of Mannar, Ceylon Seas (Carter, Dendy) : ? other localities in Indian
Ocean (see under Gellius ridleyi).

Register Number, Locality, &c. XXIII. 7, off Dwarka, 15-17 fms., 12.12.05.

14. Gellius ridleyi Hentschel.

(For possible Synonymy, see Hentschel [1912]).

I refer to this species a number of irregularly massive specimens bearing large vents
at the ends of deep, cylindrical oscular tubes. The largest specimen (R.N. XVII. 1)
is a clathrous mass of thick, anastomosing branches, for the most part vertical and more
or less fused laterally ; each pierced by a wide, cylindrical oscular tube. Colour in spirit
(after formalin), light brown. Texture loose and friable, slightly fibrous. The whole
mass measures about 70 mm. in greatest breadth and 65 mm. in height, but it has been
a good deal damaged. The oscular tubes measure up to 10 mm. in diameter.

The skeleton is a sub-isodictyal reticulation of oxea, with only a slight tendency
to collect in fibres. The oxea are gently curved, gradually and sharply pointed at each
end, and measure about 0-2 by 0-0096 mm. ; numerous more slender forms occur,
presumably young. The sigmata are simply C-shaped, or only slightly contort, slender,
and measure about 0-02 mm. from bend to bend.

Hentschel distinguishes his species from the European Gellius fibulatus (Schmidt)
by the spicular measurements. He states that in the Atlantic-Mediterranean form the
oxea measure more than 0-22 mm. and the sigmata more than 0-028 mm. in length,

108 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

while in the Indian Ocean form the oxea measure less than 0-2 mm. and the sigmata
usually less than 0-025 mm.

It appears to me that these differences are too slight to be of any specific value in
themselves, especially as I have in my possession (in Mr. Carter's cabinet) a preparation
from a specimen from the Devonshire coast (Budleigh Salterton), in which the oxea
measure only 0-14 by 0-006 mm., while the sigmata measure about 0-03 mm. from
bend to bend.

On the other hand, I think that it may be possible to distinguish an Indian Ocean
species characterised by its robust growth and large oscular tubes. For this form I
adopt Hentschel's name ridleyi, though I must regard it as very doubtful whether all
the specimens from the Indian Ocean which have been referred by various authors to
Gellius fibulatm can now be referred to Gellius ridleyi. Mr. Hornell's collection contains,
as we have already seen, another Gellius which seems to be quite distinct from G. ridleyi,
and which I regard as a mere variety of G. fibulatus, and I consider it highly probable
that the specimens which I described from Ceylon in 1905 belong to G. fibulatus rather
than to G. ridleyi.

Previously known Distribution. Indian Ocean (?) (auctorum) ; Aru Islands
(Hentschel).

Register Number, Localities, &c. XVII. 1, Kiu, littoral, 24.12.05 ; XXVI. 5,
Adatra Reef, 25 Dec. '05 ; XXXIII. 7, Dhed Mora and adjacent rocky ground between
Beyt and Aramra, 1 fm., 21.12.05.

15. Gelliodes fibrosa Dendy.

Gelliodes petrosioides, var fibrosa Dendy [1905].

In my report on Professor Herdman's Ceylon sponges I suggested that my Gelliodes
petrosioides var. fibrosa might, when better material was forthcoming, have to be con-
sidered as a distinct species. The occurrence of two fairly good specimens in Mr.
Hornell's collection leads me to carry out this suggestion. The two specimens come
from the same locality and are possibly parts of the same. One (R.N. XXXIV. 6)
is an irregular, flattened sponge which has been attached to the substratum at a few
points only, and with a tendency to throw off digitiform processes. The upper surface
is almost flat and minutely conulose. It bears several fair-sized but shallow vents,
whose margins are very slightly raised above the general surface. The specimen
measures about 48 mm. in length by 28 mm. in greatest breadth and 10 mm. in average
thickness. Texture rather soft and compressible. Colour in spirit (after formalin)
very pale brown. The second specimen is an irregularly subcylindrical fragment (?),
about 56 mm. in length by 7 mm. in diameter in the middle, broadening out suddenly
to about 18 mm. at one end.

The skeleton arrangement is identical with that of a typical Pachychalina. The
main skeleton consists of a subrectangularly meshed network of stout spicular fibre

DENDY— NON-CALCAREOUS SPONGES 109

with very little spongin, with numerous scattered oxea in the meshes between the
fibres. The dermal skeleton is an irregular network of similar fibre.

Spicules. (1) Slightly curved oxea, gradually and sharply pointed at each end,
measuring about 0-19 by 0-008 mm. (2) Slender sigmata, usually simply C-shaped,
measuring about 0-0164 mm. from hend to bend.

Previously known Distribution. Ceylon Seas (Dendy).

Register Numbers, Locality, &c. XXXIV. 6, 12, Channel, W. side of S. end of Beyt
Island, 3-4 fms., 3.1.06.

16. Reniera permollis (Bowerbank).

Isodictya permollis Bowerbank [1866].

I identify with this species a number of encrusting specimens growing over the
parchment-like worm-tubes that are so abundant in the collection. The colour in spirit
(after formalin) is rather dark brown ; the texture very soft and friable ; the surface
very minutely hispid owing to the projection of the ends of the primary skeletal lines.
The oscula are small and scattered. The crusts attain a considerable size, but are so
irregular that it is useless to give dimensions.

The skeleton is an irregular isodictyal reticulation, for the most part of single
spicules, but there is a strong tendency to form primary lines several spicules thick and
separated from one another by intervals of about one spicule's length.

The oxea are gently curved, gradually and sharply pointed, and measure up to
about 0-12 by 0-006 mm., agreeing very closely with Bowerbank's figure and
measurement.

This seems to be quite a good identification, especially as Bowerbank mentions
the " nut-brown colour," but whether R. permollis is more than varietally distinct
from R. cinerea may be regarded as an open question.

Previously known Distribution. British Seas (Bowerbank).

Register Numbers, Localities, &c. XX. 1, Adatra ; XXX. 3, 3|-4 fms., N. of
Poshetra, 20.12.05 ; XXXII. 4, off Beyt.

17. Reniera topsenti Thiele.

Reniera cinerea (Grant) var. porosa Topsent [1901].
Reniera topsenti Thiele [1905].

Three well-preserved pieces, possibly parts of the same specimen, agree
remarkably closely in external form with Topsent's figure and fairly closely with
Thiele's. The sponge is depressed and ramo-lobose in form, with numerous large,
slightly prominent vents. The surface has a porous appearance, but is really
covered by a thin, almost aspiculous dermal membrane. The texture is rather
soft and friable ; the colour in alcohol (after at any rate some formalin) light
brown. The skeleton is a rather irregular network of mostly single spicules,

H

110 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

but with a tendency to form slender, multispicular primary lines running towards
the surface.

The oxea are slightly curved, sharply and fairly gradually pointed, and
measure about 0-2 by 0-009 mm., being thus a little larger than those of Topsent's
specimens, from which they also differ in the absence of abnormal forms.

Previously known Distribution. Magellan Straits (Topsent) ; Punta Arenas
(Thiele).

Register Number, Locality, &c. XXVI. 6, Adatra Reef, 25.12.05.

is. Reniera hornelli n. sp.— (Plate II., Fig. 11).

The sponge (Fig. 11) consists of an irregularly subglobose body contracted
almost (or quite) to a short stalk at the point of attachment. There are numerous
large vents, scattered, usually on more or less strongly developed prominences, over
the upper parts of the sponge, and varying in diameter up to about 5 mm.
Each vent is the terminal opening of a very deep, cylindrical oscular tube. The
largest specimen measures about 45 mm. in height by 53 mm. in greatest breadth.
The surface of the sponge has a characteristic woolly appearance, due to the fact
that the sponge-tissue is broken up into a sort of network of villi by the in-
numerable narrow, but deep and close-set, inhalant canals. In life the surface
was doubtless covered all over by a very thin, translucent, pore-bearing dermal
membrane, but this is now nearly all rubbed off. The texture is very soft,
spongy and resilient ; the colour in spirit, pale yellowish-grey.

The main skeleton consists of numerous slender, multispicular fibres running
at right-angles to the surface at distances of about one spicule's length from one
another. These lines are united with one another cross-wise by numerous single
spicules and the whole skeleton forms a rather irregular, almost isodictyal network.
The dermal membrane appears to be almost aspiculous, and one cannot speak of a
definite dermal skeleton. There is very little spongin present in any part of the
skeleton.

Spicules. Rather slender, slightly curved oxea, gradually and sharply pointed
at each end ; measuring about 0-14 by 0-008 mm., but somewhat variable and often
more slender.

I have much pleasure in naming this beautiful and well-characterised species
after its discoverer, Mr. James Hornell. As there are four specimens in the
collection it is probably not uncommon on the west coast of India.

Register Numbers, Localities, &c. II. 12 (locality uncertain) ; IV. 4, three
specimens, dredged off S.W. coast of Beyt Island.

19. Reniera fibroreticulata n. sp.— (Plate II., Fig. 12.)

The sponge (Fig. 12) has the appearance of being made up of short, ana-
stomosing branches, sometimes united laterally so as to give a plate-like form,

DENDY— NON-CALCAREOUS SPONGES 111

and sometimes, at any rate, ending blindly. Vents relatively large, about 2 mm.
in diameter, situated on the sides of the branches or the upper margin of the
plate, each the terminal opening of a deep, cylindrical oscular tube. Surface smooth,
covered by a closely adherent, translucent dermal membrane, through which the
subdermal reticulation shows faintly. Colour in spirit pale yellow, texture fairly firm
but friable.

There are several pieces of this sponge in the collection, which, as they all
come from the same jar, probably belong to the same specimen. The largest,
represented in Fig. 12, is, as a whole, lamellar, measuring about 31 mm. in length,
20 mm. in height and 5 mm. in thickness (this being also about the usual
diameter of freely projecting branches).

The main skeleton is a close, irregular network of single spicules, penetrated
by long; multispicular fibres about 0-04 mm. thick. These fibres are numerous
and run for the most part lengthwise in the branches ; they also form a well
developed subdermal reticulation, with very irregular, unequal meshes.

The dermal skeleton consists of single spicules, thickly and evenly scattered
through the dermal membrane, crossing one another at all angles, but not united
in a regular network.

There is little, if any, spongin present.

Spicules. Short and fairly stout oxea, slightly curved and gradually and
sharply pointed at each end, measuring about 0-1 by 0-006 mm., but often rather
more slender.

This sponge reminds one rather of the European Reniera simulans, but differs
in the smaller size of the spicules and in the strongly developed reticulation of
spicular fibre. In this respect, like Reniera semifibrosa, it approaches the genus
Pachychalina.

It also resembles, both in external form and spiculation, the new species
Siphonochalina minor, described below (cf. Fig. 15). The growth of the sponge,
however, is on a much smaller scale, and the arrangement of the main skeleton
is different, for in S. minor it is entirely composed of multispicular fibres.

Register Number, Locality, &c. II. 8 (exact locality uncertain).

20. Reniera semifibrosa n. sp.— (Plate II., Fig. 13).

The sponge forms massive, convex crusts, closely adherent to other objects or
hollow underneath. The finest specimen (R.N. XXXIII. 1, Fig. 13) has the form of
a deep, inverted cup, the margin of which has been broken away all round from the
substratum, while the interior is quite hollow and empty. The total height of the
specimen is 85 mm. ; the maximum breadth of the base about the same ; the thickness
of the wall of the cup about 15 mm. The wall of the cup is perforated by two large,
irregular apertures (natural) with rounded margins, which lead right through into the

H 2

112 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

spacious interior. The inner surface of the cup is smooth but uneven ; it bears no vents.
The outer surface is similar, but bears numerous large, prominent, circular vents, up
to about 9 mm. in diameter, leading out of wide exhalant canals which come from deep
down in the interior of the sponge. The texture is fairly firm but rather cavernous and
friable ; the colour in spirit (after formalin) pale yellowish-grey, with a transparent
look which is probably due largely to the imperfect preservation.

Another specimen (R.N. XXXIII. 2 a) forms a much flatter crust growing over a
massive specimen of Jaspis reptans, and there are also a number of broken fragments
evidently of the same species.

The main skeleton is a typical unispicular, isodictyal reticulation of short oxea,
and there is a similar unispicular dermal reticulation, except that here the spicules all
lie in one plane, parallel to the surface. Just beneath the surface, however, there is a
well developed subdermal reticulation of short, multispicular fibres. This reticulation
lies parallel to the surface. Its meshes are very irregular in shape and vary greatly
in size, and the component fibres vary very much in thickness, up to at least 0-17 mm.
The dermal and subdermal skeleton are similar on the inner and outer surfaces of the
sponge. Here and there in the interior of the sponge a reticulation of coarse spicular
fibre similar to the subdermal reticulation is to be found, probably representing lines
of growth (earlier surface levels).

There is little, if any, spongin present in the skeleton.

Spicules. Oxea ; fairly stout, slightly curved, sharply and rather abruptly pointed,
but not tornote. Size about 0-16 by 0-0095 mm.

This handsome species, with its well developed subdermal reticulation of stout
multispicular fibre, seems to be intermediate between Reniera and Pachychalina.

Register Numbers, Localities, &c. XXXIII. 1, 2 a, 3, Dhed Mora and adjacent rocky
ground between Beyt and Aramra, 1 fm., 21.12.05 ; XXXIV. 7, 9, Channel, W. side of
S. end of Beyt Island, 3-4 fms., 3.1.06.

Reniera spp.

There are also in the collection a number of more or less fragmentary specimens
probably representing other species of this difficult genus.

21. Halichondria panicea Johnston vars.

(For Literature, Synonymy, &c., vide Ridley and Dendy [1887] and Dendy
[1905].)

There are several specimens in the collection which may be considered as varieties
of this ubiquitous species. R.N. XX. 3 a is the best preserved and characterised. It
consists of a compressed Jobose fragment (? erect), 51 mm. in length and 30 mm. in greatest
breadth (near the top, where it expands somewhat). A number of good-sized venjts
occur around the margin. The surface, under a pocket lens, appears very distinctly

DEN DY— NON-CALCAREOUS SPONGES 113

reticulate. The texture is firm and rather coarse, the colour in spirit (after formalin)
light brown.

The skeleton is a confused reticulation of large oxea which show a strong tendency
to arrange themselves in coarse fibres. There is a very well developed dermal reticu-
lation of spicular fibre varying in diameter.

The oxea are slightly curved, gradually and sharply pointed at each end, and
commonly measure about 0-77 by 0-023 mm., although variable.

R.N. XXXII. 5 is more massive, with a more compact and less distinctly fibrous
skeleton, and comes very near to Ceylon specimens of Halichondria panicea var.
megalorhaphis collected by Professor Herdman [Dendy 1905]. Most of the specimens
are in a very poor state of preservation, owing, doubtless, to the use of formalin as a
preservative.

Previously known Distribution of the Species. Almost cosmopolitan.

Register Numbers, Localities, &c. XVII. 3, Kiu littoral ; XX. 3 a, b, Adatra ;
XXIX. 4, 3f-4 fms. N. of Poshetra, 20.12.05 ; XXXII. 5, off Beyt ; XXXIV. 8,
Channel, W. side of S. end of Beyt Island, 3-4 fms., 3.1.06.

22. Halichondria reticulata Baer [1905].— (Plate II., Figs. Ua, 14&.)

The sponge, of which there is a good deal in the collection, consists of slender,
irregular, often tortuous, creeping branches, which frequently anastomose and unite
with one another in irregular, massive lumps (Figs. 14 a, 14 b). Individual branches
are usually about 2 or 3 mm. in diameter. They are generally subcylindrical, but
may be flattened ; they may be bluntly rounded at the extremity or drawn out into a
long point. The surface is finely granular and marked by ramified, meandering, sub-
dermal canals. Vents minute, few, scattered. The colour in spirit is very pale yellow ;
the texture rather compact and fleshy, but soft and compressible.

The main skeleton consists of slender oxea, partly scattered quite irregularly
and partly in loose wisps or fibres which run towards the surface. The dermal
skeleton consists of irregularly scattered spicules of the same kind, lying tan-
gentially and crossing one another in all directions.

Spicules. Slender oxea ; very slightly curved, gradually sharp-pointed at each
end, measuring about 0-21 by 0-006 mm. ; very uniform in shape and size.

This species seems to be well characterised by its peculiar mode of growth
and by the small size of its spicules. Although Baer's description is very brief,
and although the spicules in the Okhamandal sponge seem to be somewhat
stouter, I think the identification is a fairly safe one. Baer observes that the
flagellate chambers are round and 0-03 mm. in diameter, and that the ground
substance is homogeneous and filled with round granule-cells.

The histological features of the Okhamandal sponge, which I have studied by
means of paraffin sections of material stained with borax-carmine, confirm the identi-

114 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

fication with Baer's species. The flagellate chambers are approximately spherical
and about 0-03 mm. (or a little more) in diameter, scattered rather sparsely in
a compact ground-substance densely charged with minute, spherical, granule-
bearing cells, measuring up to about 0-008 mm. in diameter. There is a fairly
thick ectosome, more or less interrupted by the spacious subdermal cavities and
also containing small granule-cells.

Throughout the sponge, but especially in parts of the ectosome, where they
are densely crowded together, occur numerous large, spherical cells, filled with
granules of various sizes and each with a compact, deeply staining nucleus of
moderate size. The diameter of the entire cell is about 0-025 mm. ; of the
nucleus about 0-008 mm. These cells resemble immature ova, but they may be
merely large amoebocytes charged with food-material or excretory products.

Although Baer has figured both external form and spicules in the case of the
Zanzibar sponge, I have thought it desirable to add illustrations of the Indian
form.

Previously known Distribution. Zanzibar (Baer).

Register Numbers, Locality, &c. II. 6, 11 (altogether a considerable number of
pieces), probably off Poshetra, January 7, 1906.

23. Siphonochalina crassifibra Dendy.

Siphonochalina crassifibra Dendy [1889].
Siphonochalina communis (pars) Dendy [1905].

The best of the two specimens in the collection closely resembles in external
form Carter's Siphonochalina (Patuloscula) procumbens from the West Indies, a
figure of which will be found in my memoir on the West Indian Chalinina3 [1890].
It consists of a spreading base from which a dozen or more tubes rise obliquely
upwards, branching and anastomosing with one another to a slight extent. The
tubes are subcylindrical, about 50 mm. in height and 12 or 14 mm. in diameter,
and each terminates in a wide, circular vent about 6 mm. in diameter. They are
thus considerably smaller than in the type. The surface is smooth but finely
granulated.

In skeletal peculiarities the specimens exaggerate the distinguishing character
of the type. The main skeleton is a rectangularly or polygonally meshed network
of fairly stout fibre, almost completely filled with the very numerous, close-packed
spicules, so that there is only a thin investment of spongin. The primary fibres
are about 0-05 mm. in diameter and the secondaries only a little less. The
dermal skeleton in the best specimen, from which this description is taken, is
chiefly a unispicular reticulation, in which the spicules are held together by very
pale-coloured spongin, with a much coarser subdermal reticulation formed by the

DENDY— NON-CALCAREOUS SPONGES 115

outer part of the main skeleton. The spicules are very slightly curved, fairly
sharply pointed oxea, measuring about 0-078 by 0-004 mm.

In my Report on Professor Herdman's Ceylon Sponges [1905] I suggested
that my Siphonochalina crassifibm should be regarded as a variety of Carter's
Siphonochalina (Tubulodigitus) communis, also an Indian Ocean species. Probably
we shall ultimately have to unite in one species a considerable number of varieties
which exhibit the same characteristic external form, e.g., Tubulodigitus communis
Carter [1881], Siphonochalina crassifibra Dendy [1889], Siphonochalina communis
var. tenuispiculata Dendy [1905], from the coasts of India and Ceylon ; Patuloscula
procumbens Carter [1882, 1885 bis], from Australia and the West Indies ; Siphono-
chalina intermedia Ridley and Dendy [1887], from Australia ; Siphonochalina
spiculosa Dendy and S. ceratosa Dendy [1890], from the West Indies. In all
these localities there seems to be much variation in the relative amounts of
spicules and spongin in the skeleton fibres. There is also a good deal of variation
in the length and diameter of the tubes, but I do not see how it is feasible to
make really satisfactory specific distinctions between the forms mentioned.

Until, however, it is possible to make a more thorough comparative study of
this interesting group of varieties I propose to revert to the specific name " crassi-
fibra "for the Indian form in question.

Previously known Distribution. Gulf of Mannar (Dendy).

Register Numbers, Localities, <&c. VII., Adatra Reef, 5.12.05 ; XXXIV.
5, Channel W. side of S. end of Beyt Island, 3^1 fms., 3.1.06.

24. Siphonochalina minor n. sp. — (Plate II., Fig. 15.)

The sponge (Fig. 15) consists of a horizontal, subcylindrical or vertically
somewhat flattened, branching and anastomosing stolon, from which arise short,
slightly branching and sometimes anastomosing, ascending tubes, terminating each
in a wide vent. In the best specimen (R.N. XXVI. 4) there is a continuous
stolon 53 mm. in length by about 6 mm. in diameter ; the largest tube arising
from it is about 35 mm. in height and 10 mm. in average diameter, with a
terminal vent about 3-5 mm. in diameter. The stolon itself, though penetrated by
smaller canals, does not contain a single, wide, central cavity, as do the tubes
that arise from it. The surface is subglabrous, slightly uneven, very minutely
granular. The texture is very compressible and resilient, easily torn. The colour
in spirit (after formalin) is light brown.

The main skeleton is a rectangularly meshed network of stout multispicular
fibre, the meshes becoming irregularly polygonal in the deeper parts. Primary and
secondary fibres are alike and of about the same diameter, 0-034 mm. They are
composed of a great number of oxea and a small quantity of spongin, which does
not seem to form a continuous investment but is sometimes visible in the angles

116 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

of the reticulation. The dermal skeleton is an irregular, unispicular reticulation
of oxea, in which the individual spicules overlap one another extensively ; without
obvious spongin.

The spicules are very slightly curved oxea, gradually and sharply pointed at
each end, measuring about 0-13 by 0-006 mm.

This species shows very clearly the difficulty of differentiating between the
genera Pachychalina and Siphonochalina. The whole structure of the stolon is
typically that of a Pachychalina, and only the presence of the tubular ascending
branches justifies its inclusion in Siphonochalina. The entire sponge resembles
Siphonochalina crassifibra on a smaller scale, but with decidedly larger spicules.
It makes a near approach to my Pachychalina subcylindrica from Ceylon [1905]. I
have already pointed out the resemblance which it bears to Reniera fibroreticulata
(vide supra and cf. Fig. 12).

Register Numbers, Localities, &c. XXVI. 4, Adatra Reef, 25 December, '05 ;
XXXIII. 5, Dhed Mora and adjacent rocky ground between Beyt and Aramra,
1 fm., 21.12.05.

Chalininae spp.

There are several other small chalinine sponges in the collection which are
not in a sufficiently good state of preservation or sufficiently well characterised to
make identification or description desirable.

25. Desmacella tubulata Bendy [1905].

It is very interesting to meet with this curious and well characterised species
again in Mr. Hornell's collection. So far as I am aware it has not been recorded
since I first described it from Ceylon in 1905. The material now before me
consists of a number of fragments of thin-walled tubes, some of which seem to
have been 10 mm. or more in diameter. At one point some of these tubes have
grawn into close union with a specimen of Gellius sp.

The agreement in spiculation with the type is very close, but I must add
that I have observed a few slender toxa, about 0-032 mm. long, on one occasion
arranged in a sheaf or toxodragma, differing only in the curvature of the toxa
from the smaller trichodragmata. I did not observe any toxa in the type speci-
mens, but probably they occur there amongst the vast number of trichodragmata.
I must also add to my original description that the individual trichites of which
the trichodragmata are composed, although slender, are not nearly so slender as
in some cases. The sheaves or dragmata themselves may be very thick.

Previously known Distribution. Gulf of Mannar (Dendy).

Register Number, Locality, <&c. XXVI. 7, Adatra Eeef, 25 December, '05.

DENDY— NON -CALCAREOUS SPONGES 117

26. Thrinacophora eervicornis Ridley and Dendy.

Thrinacophora eervicornis Ridley and Dendy [1887].
Thrinacophora eervicornis Hentschel [1912].

This species is represented in the collection by two small specimens. The
branching, so far as the specimens show, is dichotomous and in one plane, the
branches being very short and bluntly rounded off. Surface sparsely hispid owing
to the long, slender styli, which project for nearly two millimetres, with shorter
projecting styli between. The two specimens are very similar to one another in
appearance and closely resemble the young antlers of a stag " in velvet." Total
height of each about 20 mm. ; diameter of main stem and branches about 3-5 mm.,
but variable. Colour in spirit light brown.

There is a very stout skeletal axis composed of a dense reticulation of stout
and rather short oxea. This is surrounded by a comparatively thin layer of soft
tissue in which loose bundles of very long styli run lengthwise. Similar styli,
with their bases implanted in the central axis, extend through the outer layer and
for a long distance beyond the surface. These are arranged singly and each is
surrounded, where it leaves the surface, by a radiate tuft of comparatively short
and very slender styli. There is no conspicuous spongin in any part of the
skeleton (in balsam preparations).

Spicules. 1 . Oxea of the central axis ; rather short, stout, distinctly
but not very strongly curved or angulated in the middle; usually fairly
gradually and sharply pointed at each end. Size fairly uniform, about 0-26 by
0-013 mm.

2. Large styli, running lengthwise in the outer part of the sponge and also
projecting more or less at right angles from the surface. These spicules are stoutest
at the base and from there taper very gradually to very fine points. They
measure about 2-24 mm. in length by 0-02 mm. in thickness at the base. They
are slightly curved.

3. Small styli of the radiate surf ace- tufts ; very slender and finely pointed ;
measuring about 0-55 by 0-0043 mm.

4. Trichodragmata ; occurring in immense numbers in the outer part of the
sponge ; each bundle rather long and narrow, measuring about 0-1 by 0-0082 mm.,
often curved, easily separating out into trichites.

In spite of some slight apparent differences in spicular proportions, especially
as regards the smaller size of the large styli, I think that there can be no reason-
able doubt that we have here two small and probably young examples of the
Challenger species described by Mr. Ridley and myself under the name Thrina-
cophora eervicornis.

In the Challenger Report the length of the trichodragmata is given as 0-0126
mm. It should be 0-1 mm., as determined by re-measurements of the type.

.

118 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

I believe that this remarkable and well-characterised species has only been
recorded once since it was first described in the Challenger Report.

Previously known Distribution. Philippine Islands (Ridley and Dendy) ; Aru
Islands (Hentschel).

Register Number, Locality, &c. III. 6 a, b, dredged off Dwarka.

27. Axinella virgultosa Carter [1887].

There is in the collection one remarkably beautiful specimen, which agrees
so closely in its very characteristic external form with the description and figure
given by Carter of his Axinella virgultosa from Mergui that I have little doubt
in making a specific identification. Carter's description of the spiculation, however,
without either figures or measurements, is so inadequate that it is impossible,
without referring to the type, presumably in Calcutta, to be quite certain. Un-
fortunately, there is no microscopic preparation of the sponge in Mr. Carter's
cabinet.

Mr. Hornell's specimen consists of a large number of slender, vertical, stiff
processes, rising up side by side from a thin, encrusting base which measures
about 35 by 22 mm. The larger processes are in the middle and measure about
20 mm. in height. They are about 2 mm. in diameter at the base and taper
gradually to sharply pointed apices. The processes may give off branches, chiefly
from their outer sides near the base. They are all abundantly but shortly hispid
with projecting spicules. The colour in spirit (after formalin) is almost white.

Little more than the skeleton of the sponge remains. Bach process consists
of a plumose column of short, stout spicules arranged in typically axinellid
fashion.

The spicules are typically stout styli, sometimes becoming oxeote by more
or less pronounced sharpening of the inner end. They are commonly a little
bent and the outer end is gradually sharp-pointed. I have seen nothing of the
subterminal inflation which Mr. Carter says is often present in his specimens.
These spicules measure about 0-77 by 0-04 mm.

Previously known Distribution. Mergui Archipelago (Carter).

Register Number, Locality, &c. IX., off Dwarka, 15-17 fms., 12.12.05.

28. Phakellia donnani (Bowerbank).

Isodictya donnani Bowerbank [1873].
Axinella donnani Dendy [1887].
Phakellia donnani Dendy [1905].
Phakellia donnani Row [1911].

This well-known Ceylon species is represented in the collection by a single
specimen of the ordinary, pedunculate, cup-like form. The colour in life was, as

BENDY— NON-CALCAREOUS SPONGES 119

usual, orange. It differs from my Ceylon specimens only in the somewhat larger,
and especially stouter, spicules, which measure, when full-grown, about 0-38 by
0-025 mm.

Previously known Distribution. Gulf of Mannar, Ceylon Seas (Bowerbank,
Dendy) ; Red Sea (Row).

Register Number, Locality, <&c. II. 7, Poshetra Head, 7.1.06.

29. Auletta lyrata var. glomerata Dendy. — (Plate II., Fig. 16).

Spongia lyrata Esper [1794-1806].
Raspaigella lyrata Ehlers [1870].
Auletta aurantiaca Dendy [1889].
Auletta lyrata var. glomerata Dendy [1905].

The single specimen (Fig. 16) in the collection agrees very closely in general
form with the type of the variety. The strongly sphinctrate vents are situated
each in a cup-shaped depression at the extremity of a short branch. The average
thickness of the spicules is, however, much greater than in the type of the
variety.

The species of the genus Aiiletta are evidently extremely variable both in
external form and spiculation, and it will be extremely difficult to differentiate
them from one another.

Previously known Distribution of Species. Ceylon, Gulf of Mannar (Esper,
Dendy).

Register Number, Locality, &c. III. 3, off Dwarka.

30. Auletta elongata Dendy var. fruticosa nov.— (Plate II., Fig. 17).

Auletta elongata Dendy [1905].

The single specimen (Fig. 17) differs from the type in its much more spreading
mode of branching and in the considerably smaller average size of the spicules.
Previously known Distribution of Species. Ceylon, Gulf of Mannar (Dendy).
Register Number, Locality, &c. XXIII. 2, off Dwarka, 15-17 fms., 12.12.05.

31. Ciocalypta dichotoma n. sp.— (Plate III., Fig. 18).

The single specimen (Fig. 18) consists of a cylindrical stem dividing at half
the total height of the specimen into two approximately equal branches diverging
from one another at an acute angle. Each branch terminates in a bluntly pointed
apex. The base of the stem is slightly enlarged and attached to it are a few
grains of coarse sand and a comparatively large shell-fragment, which seem to
indicate that the stem was directly attached to the substratum and did not
spring from a massive body. The surface is stellately reticulate as in Ciocalypta
hyaloderma Ridley and Dendy [1887], though hardly so distinctly. There is a

120 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

translucent dermal membrane, supported by a reticulation of spicular fibre and
overlying extensive subdermal cavities. There are a number of small, incon-
spicuous vents, chiefly in single series along the sides of stem and branches.
Colour in spirit very pale yellow ; texture stiff, resilient. Total height of specimen
46 mm. ; diameter of stem and branches about 4-5 mm.

The sponge consists, as usual in the genus, of a central axis, surrounded by
wide subdermal cavities which are traversed by spicular columns supporting the
dermal membrane. The axis is very thick and the radiating spicule columns very
short. Numerous loose fascicles of large oxea run lengthwise through the axis,
separated from one another by a fair amount of soft tissue and crossed here and
there by scattered oxea. The radiating columns which support the dermal
membrane are loose fascicles of similar spicules, ending in surface brushes of
short styli. The dermal skeleton is a very irregular reticulation of loose spicular
fibre composed of the large oxea. In the dermal membrane also occur numerous
small styli, mostly arranged in the above-mentioned brushes at the ends of the
radial columns.

The spicules are very sharply differentiated into two kinds : — (1) Large oxea ;
slightly curved, fusiform, symmetrical, gradually and sharply pointed at each end,
size about 0-8 by 0-02 mm. (smaller ones also occur). (2) Small styli ; short,
usually slightly bent ; well rounded off, but somewhat narrowed, at one end, and
gradually sharp-pointed at the other, size about 0-2 by 0-008 mm.

The spiculation of this sponge seems to be identical with that of my Hymeni-
acidon (?) foetida [1889], originally from the Gulf of Manaar, which has been
included by Lindgren [1898], Thiele [1900] and Hentschel [1912] in the genus
Ciocalypta, probably quite rightly. The external form of Ciocalypta dichotoma,
however, is so definite, and so different from that of C. foetida, that I think they
may, for the present, be regarded as distinct, though closely related species. It
should be borne in mind, on the other hand, that Hentschel [1912] describes
specimens of C. foetida with finger-shaped processes.

Register Number, Locality, &c. IV. 21, dredged off S.W. Coast of Beyt Island.

32. Higginsia sp.

I have no hesitation in referring to this genus a subcylindrical fragment
measuring about 30 by 8 mm. Unfortunately, the specimen was preserved in
formalin, and is in a very badly macerated condition, practically nothing but the
skeleton remaining, while it contains a large number of evidently foreign spicules.

The main skeleton is a very irregular, subfibrous reticulation of large, stout
oxea, measuring about 0-9 by 0-03 mm., only very slightly curved and gradually
and sharply pointed. These are accompanied by a number of very much longer
and very much more slender oxea and styli, which seem to belong to the sponge.

DENDY— NON-CALCAREOUS SPONGES 121

There are two kinds of microxea ; (1) covered with small sharp spines; more
or less sharply angulated in the middle; size about 0-15 by 0-006 mm.; (2) smooth,
but with a distinct swelling at one side of the central angulation (subcentrotylote) ;
of about the same length as the spined ones but much more slender, even when
allowing for the absence of spines. An intermediate form may occasionally be
found, but on the whole the two kinds seem to be fairly distinct. It is possible,
however, that the smooth ones may be merely young forms of the other.

I refrain from giving a name to this species until better material is forth-
coming.

Register Number, Locality, &c. XVIII. 4 a, Channel W. of Beyt Island, 3-4
fms., January 7, '06.

33. Esperella plumosa (Carter).— (Plate I., Figs. 4a-4g ;. Plate III., Fig. 19).
Esperia plumosa Carter [1882, 1887].
Esperella plumosa Dendy [1905].
Not Esperella plumosa Arnesen [1903].

This appears to be a very common and characteristic Indian Ocean species
and is by far the most abundant sponge in Mr. Hornell's collection. It was also
abundant in Professor Herdman's Ceylon collection. As yet, however, no figures,
either of the external form or of the very well developed spiculation, have been
published, and it seems desirable to make good this omission on the present
occasion.

The external form appears to be very characteristic, though varying much
according to the stage of growth. The sponge seems to begin life as an irregular
crust, which becomes 'massive and then grows out into long, flattened, tongue-
shaped processes, finally breaking up into slender, pointed, digitiform branches ; or
possibly such branches may be formed first as outgrowths of the massive crust,
and subsequently fuse to form the flattened tongue-shaped portions. One of the
best pieces is represented of the natural size in Fig. 19, but this is not the largest
specimen.

The irregularly conulose or cactiform surface also seems to be characteristic.
The surface is subglabrous between the conuli and the dermal reticulation of
spicular fibre appears to be very unequally developed ; in one specimen which I
have dried it is quite conspicuous in some places, under a pocket lens, while
apparently absent in others. The oscula are represented by larger and smaller
circular apertures scattered here and there between the conuli.

Most of the specimens are now, in spirit and dry, of a dull reddish colour,
but I suspect that this may be due to their having been preserved in the first
instance in formalin.

I have nothing further to add to my previous description of the skeleton arrange-

122 OKHAMANDAL MARINE ZOOLOGY REPORT— PART IT

ment, but a more complete account of the spiculation seems desirable to accompany
the illustrations, especially as I omitted before to mention two types of spicule which
occur both in Carter's type and in the Ceylon and Okhamandal specimens. These
are the small, palmate anisochelse and the small, slender sigmata. I think I
must have regarded these formerly as merely young forms, but I do not think
that that view can be accepted. There are, then, no fewer than seven different
kinds of spicule in the sponge, and the constancy in form and size in all the
specimens I have examined, including the type, is very remarkable.

1. Tylostyli (Fig. 4«), of the usual Esperella type, generally slightly crooked ;
size about 0-3 by 0-009 mm. (much more slender forms also occur).

2. Large, broad, palmate anisochelse (Figs. 46-46"), about 0-049 mm. in
length by 0-022 mm. in greatest width from one lateral palm to the other.

3. Small palmate anisochelee (Figs. 4c, 4c'), about 0-02 mm. long by 0-008 mm.
in greatest width.

4. Minute palmate isochelse (Figs. <id, 4^'), resembling those of Clathria,
about 0-012 mm. long.

5. Large, stout sigmata (Figs. 4e, 4e'), probably all really more or less
contort, with abruptly recurved and very sharply pointed ends ; length in a
straight line from bend to bend about 0-094 mm., thickness about 0-0054 mm.

6. Small, slender, contort sigmata (Fig. 4/) ; length in a straight line from
bend to bend about 0-033 mm., thickness about 0-0013 mm. ; but variable in
dimensions.

7. Slender toxa (Fig. 4</), gently curved like a parenthesis mark ; length about
0-065 mm., thickness about 0-0013 mm. ; sometimes arranged in toxodragmata.

Pending a much-needed revision of the esperelline sponges I adhere to the
genus Esperella for this species, but, apart altogether from the question whether or
not that name should be replaced by Mycale, it seems probable that the genus
will have to be split up in the near future.

I have already [1905] pointed out the close resemblance of the characteristic
large anisochelse of this species to the corresponding spicules of Esperella simonis
Ridley and Dendy [1887], from Simon's Bay, Cape of Good Hope. It is obvious
that the two species are closely related, but there are certain well-marked
differences in the spiculation. Chief amongst these are the absence from the
spiculation of E. simonis of the minute, palmate isochela? and of the small, slender
sigmata, both of which are very abundant in E. plumosa, which has an extra-
ordinarily full complement of spicules. Other differences concern the size of the
spicules and the shape of the toxa, which are much more strongly arcuate in
E. simonis.

Previously known Distribution. Mauritius and Mergui Archipelago (Carter) ;
Ceylon (Dendy).

BENDY— NON-CALCAREOUS SPONGES 123

Register Numbers, Localities, &c. XX. 2, 8, Adatra ; XXII. 1,3,^ mile N.
of Poshetra, 20.12.05; XXIV. 1, 2, Kiu, low water, 24.12.05 ; XXVI. 3, 12,
Adatra Reef, 25 Dec. '05 \\ XXXII. 1, 2, 7, off Beyt ; XXXIV. 3 a, channel,
W. side of S. end of Beyt Island, 3-4 fms., 3.1.06.

34. Desmacidon minor n. sp.— (Plate III., Figs. 20a, 206.)

There are two specimens of this sponge in the collection, both from the same
locality. One of them (R.N. XX. 6 a, Fig. 20a) is compressed, flabellate, with a much
contracted base, almost forming a short peduncle. The upper margin of the sponge
is widely extended and produced into short conical processes, apparently the ends of
laterally fused branches. The vents are of fair size, shallow and scattered on one surface
of the sponge only, not marginal ; owing to the maceration of the specimen they are no
longer very distinct. The surface is finely granular. The specimen measures about
26 mm. in greatest height, 41 mm. in width, and 3 mm. in average thickness. The
texture is soft and resilient ; the colour in spirit (after formalin), light brown.

The second specimen (R.N. XX. 6 b, Fig. 206) is very similar in most respects, but
is divided into digitiform branches from the contracted base upwards, most of the
branches lying in approximately the same plane. The branches are subcylindrical or
slightly compressed, and about 3-6 mm. in diameter. The shallow vents tend to arrange
themselves in longitudinal series on the branches. The maximum height of the specimen
is about 33 mm.

The skeleton is a sub-isodictyal reticulation of short oxea, but with the meshes
composed of plurispicular fibres in which the spicules are held together by a considerable
amount of very pale-coloured spongin.

The spiculation consists of the following :—

(1) Fairly stout, slightly curved oxea, gradually and sharply pointed at each end,
measuring about 0-13 by 0-008 mm. Much more slender oxea also occur, which are
probably young forms.

(2) Slender, palmate isochelae(" naviculiform ''), about 0-01 64 mm. long ; abundant.
This species might almost be regarded as a dwarf variety of Desmacidon compressa

(Carter's Chalina compressa), which appears to be a South African species, but the
spicules of the latter are very much larger, the oxea measuring, according to Carter
[1882 bis], about 0-37 by 0-023 mm, and the isochelse (" naviculiform equianchorates ")
0-025 mm. in length.

Register Numbers, Locality, &c. XX. 6 a, b, Adatra.

35. lotrochota baculifera Ridley.

lotrochota baculifera Ridley [1884].

lotrochota baculifera var. flabellata Dendy [1887].

lotrochota baculifera Topsent [1893].

124 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

lotrochota baculifera Topsent [1897].

lotrochota baculifera Lindgren [1898].

lotrochota baculifera Thiele [1899].'

lotrochota baculifera var. tumescens Kirkpatrick [1900].

lotrochota baculifera Thiele [1903].

lotrochota baculifera Dendy [1905]. •

lotrochota baculifera var. minor Hentschel [1911].

lotrochota baculifera Hentschel [1912].

There is in the collection only a single specimen of this widely distributed and well
known Indian Ocean species. It forms a rather thick, irregular crust of a dark brownish-
purple colour, attached to a calcareous nodule, the greatest diameter of the specimen
being about 30 mm. The specimen was preserved in alcohol and is in good condition.

The skeleton arrangement and spiculation are typical. The styli measure about
0-13 mm. in length by from 0-004 to 0-008 mm. in thickness. The diactinal megascleres
are strongylote or have only very feebly developed heads ; they measure about 0-2 by
0-004 mm. The " birotulates " are very abundant but very minute, only about
0-0125 mm. long. The spicules therefore are decidedly smaller than in Professor
Herdman's Ceylon specimens.

Previously known Distribution. North Australia and Mascarene Islands (Ridley) ;
Gulf of Mannar and Ceylon Seas (Dendy) ; Seychelles and Amboina (Topsent) ; Coast
of Cochin China (Lindgren) ; Celebes and Ternate (Thiele) ; Christmas Island (Kirk-
patrick) ; S.W. Australia and Aru Islands (Hentschel).

Register Number, Locality, &c. XIII., Adatra Reefs, 25 December, '05.

36. Guitarra indiea n. sp.— (Plate I., Figs. 5a-5b'"" ; Plate III., Fig. 21).

This very interesting sponge is represented in the collection by eight good specimens.
Five of these (Fig. 21) are attached to a branching, parchment-like tube, belonging
to some polychsete worm*, along with other sponges, including an Esperella and an
encrusting Aplysillid. Two other loose specimens, in the same jar, have probably
been broken away from the same association. The eighth specimen (R.N. II. 4) is also
loose and comes from a different locality.

The specimens are irregularly cushion-shaped and tend to surround by overgrowth
the object to which they are attached ; they bear a general resemblance, in form and
colour, to some species of Chondrilla.

The largest measures about 28 mm. in maximum diameter, with a true thickness
(from the outer surface to the surface of attachment) of about 7 mm. The colour,
in spirit, ranges from slate grey (on the surface which was evidently exposed to the
light) to pale yellow (on what was evidently the shaded surface). R.N. II. 4 is pale

yellow all over.

* Probably Eunice tubifcx, see p. 96.

BENDY— NON-CALCAREOUS SPONGES 125

The vents are numerous and irregularly scattered on the more exposed parts of
the surface. They are mostly minute and each at the summit of a small conical pro-
jection, formed by a contracted spicular, membranous margin, but I have seen one
expanded up to nearly 2 mm. in diameter. The surface is smooth and appears porous
under a pocket lens, but I have not been able to detect the actual inhalant pores. These,
however, are no doubt scattered in the thin dermal membrane between the surface-
brushes of megascleres.

The main skeleton is a very irregular, rather close-meshed reticulation of somewhat
loose spicular fibre, with a certain number of isolated megascleres scattered between.
At the surface this gives place to a velvety pile composed of well denned brushes of
megascleres with outwardly directed apices projecting for a short distance beyond the
surface. I have detected no spongin.

Spicules. (1) Styli (tornostrongyla) (Fig. 5a) ; nearly straight, rather abruptly
sharp-pointed at one end and rounded off at the other ; often a little crooked ;
measuring about 0-266 by 0-007 mm., but often more slender.

(2) Placochelse (Figs. 5b-5b'"") ; of the typical Guitarra form but with the shaft
very abruptly constricted in the middle ; length about 0-041 mm., with greatest breadth
of expanded shaft about 0-0143 mm. Numerous smaller forms occur with less sharp
constriction in the middle of the shaft ; also numerous very slender forms of various
sizes, without fimbrisB or with very feebly developed fimbriee, which I take to be early
developmental stages. The placochelaa are abundantly scattered throughout the
choanosome.

So far as I am aware, only three species of Guitarra have hitherto been described ;
viz. Guitarra fimbriata Carter [1874], from deep water in the North Atlantic, Guitarra
valuta Topsent [1904], from deep water off the Azores, and Guitarra Antarctica Hentschel
[1914 ?], from deep water in the Antarctic. Unfortunately, no depth was recorded for
our new species, but it was associated with typical shallow water sponges and the
depth was probably not more than a few fathoms.

Guitarra indica differs from all its congeners in the comparatively small size of the
placochelao. It seems to come nearest to G. valuta, but differs from that species in the
smaller size of the spicules, the presence of the surface pile of megasclere-brushes (in
which it agrees with G. fimbriata), and in the abrupt constriction of the shaft of the
placochelse. It differs from G. fimbriata and agrees with G. valuta in having torno-
strongylote megascleres (styli), and it differs from G. antarctica in having no sigmata.
The youngest stages (Fig. 56'") in the development of the placochela), however, resemble
sigmata, but they have a peculiar, indefinite, rough outline which seems to indicate
immaturity.

Locality, Register Numbers, &c. II. 4, off Poshetra, 7 January '06 ; IV. 5 a-g,
dredged off S.W. coast of Beyt Island.

126 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

Genus Psammochela n. g.

Desmacidonidee with a reticulate skeleton composed of sandy and sometimes
partly spicular fibres. Megascleres styli or strongyla or both. Microscleres isochelee,
which may be very minute and with vestigial teeth ; to which sigmata may be added.

I propose this genus for some interesting sand-sponges from the neighbourhood of
Beyt Island, which seem to represent a stage in the regressive evolution of such genera
as Phoriospongia and Chondropsis. These two genera have no chelae, and I have hitherto
excluded them from the Desmacidonida?, but the occurrence of the vestigial isochelse
in Psammochela seems to indicate a probable desmacidonid origin.

I regard Psammochela elegans n. sp. as the type of the genus. Hentschel's
Desmacidon psammodes from S.W. Australia [1911] evidently belongs to the same
genus. My own Desmacidon (?) arenifibrosa [1896] and Carter's Dysidea chaliniformis
( = Desmacidon (?) chaliniformis Dendy [1896]), may belong to a closely related genus
in which the megascleres have been completely suppressed.

37. Psammochela elegans n. sp. (Plate I., Figs. 6a-6e ; Plate III., Figs. 22<z,
226).

Sponge (Figs. 22a, 226) irregular, lamellar or digitate, often running out into long,
slender, sometimes bifurcating processes, from 3 to 10 mm. in diameter and up to about
90 mm. in length in the specimens before me. Surface irregularly rugose or conulose,
but with the intervals between the rugae or conuli spanned over in life by a delicate,
translucent, minutely reticulate and finely porous dermal membrane. In formalin
specimens (Fig. 226) the dermal membrane has completely disappeared, and the
sponge has a curiously eroded appearance, the surface being deeply and irregularly
grooved and pitted,the grooves and pits being obviously uncovered subdermal cavities.
Vents probably rather small and scattered.

The main skeleton is a fairly close-meshed but very irregular reticulation of rather
slender fibre, composed of sand-grains and proper megascleres in varying proportions,
with no visible spongin ; with numerous megascleres and sand-grains scattered in the
soft tissues between the fibres. The dermal skeleton is composed chiefly of very fine,
scattered sand-grains, with a more or less pronounced tendency to arrange themselves
in a fine-meshed reticulation.

Spicules. (1) Slender styli (Fig. Qa) ; evenly rounded off at the base, rather
abruptly sharp-pointed at the apex, often a little crooked ; size about 0-16 by 0-005 mm.,
but very variable in thickness (? sometimes becoming strongylote). (2) Tridentate
isochelse (Figs. 66, 66') ; fairly robust, with stout, curved shaft ; length about
0-024 mm. These are usually scarce. (3) Very minute, C-shaped isochelse (Figs.
6c, 6c') ; resembling in side view slender, strongly curved, C-shaped sigmata with
slightly enlarged ends ; in front view like an lotrochota "birotulate," with indications
of three vestigial teeth at each end ; length from bend to bend about 0-012 mm.

DENDY— NON-CALCAREOUS SPONGES 127

Very numerous. (4) Slender or fairly stout, more or less contort sigmata (Figs.
6d, Qd') ; with short, strongly recurved, finely "pointed ends ; measuring commonly
about 0-033 mm. from bend to bend. Much smaller sigmata also occur (Fig. Qe), which
may belong to a different category.

Register Numbers, Localities, &c. IV. 12, 16, 20, dredged off S.W. coast of Beyt
Island; X., XXXIV. 11, channel, W. side of S. end of Beyt Island, 3-4 fathoms,
3.1.06; XVIII. 3, channel, W. of Beyt Island, 3-4 fathoms, Jan. 1906; XXXIII. 6,
Dhed Mora and adjacent rocky ground between Beyt and Aramra, 1 fathom, 21.12.05.

38. Chondropsis kirkii (Carter).

Dysidea kirkii Carter [1885].
? Sigmatella australis Lendenfeld [1889].
Sigmatella corticata Lendenfeld [1889].
Chondropsis kirkii Dendy [1895].

I identify with this species a single well preserved specimen which differs in no
important respect from the common Australian form. The specimen is subdigitate,
consisting of two or three tubular processes partially fused together side by side and each
terminating in a conspicuous but constricted vent, surrounded by a membranous collar.
The surface is slightly conulose, minutely reticulate where rubbed, sub-glabrous where
uninjured. The texture is rather soft and compressible, the colour in spirit pale grey.
Height of specimen 31 mm., breadth 36 mm., diameter of digitiform processes about
12 mm.

The main skeleton consists of the usual irregular reticulation of fine-grained sand-
fibre. The dermal sand- reticulation is less strongly developed than is usual in Australian
specimens. The spicules are the usual slender strongyla and small sigmata, the
strongyla perhaps rather better developed than in most Australian specimens.

Previously known Distribution. Australian Seas (Carter, Lendenfeld, Dendy).

Register Number, Locality, &c. IV. 11, dredged off S.W. coast of Beyt Island.

39. Myxilla arenaria Dendy [1905].

I identify with this species three specimens in the collection, all from about the same
locality. The one (R.N. IV. 2 a) upon which the following notes are based is irregularly
and massively lobose, slightly clathrous, about 66 mm. in height and 50 mm. in greatest
breadth. It looks (in spirit) like a mass of sand held together by pale grey, gelatinous
" sarcode," which forms a thicker or thinner surface layer. The surface is uneven and
slightly conulose or rugose. The vents are rather small and arranged in a row on the
prominent ridges which form the top of the sponge.

The skeleton is composed chiefly of a dense agglomeration of coarse sand-grains
which, to the naked eye, show, through the translucent dermal membrane, a distinct
tendency to be arranged in ascending columns. These sand-grains are very sparsely

I 2

128 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

echinated by small spined styli, while slender diactinal megascleres occur scattered
and in loose wisps in the soft tissues between. Towards the surface the diactinal
megascleres become -much more abundant and form radiating dermal brushes. The
megascleres are rather more robust than in the type "and the diactinal megascleres
are distinctly tornote, i.e. very abruptly pointed at each end, rather than rounded off.

Previously known Distribution. Gulf of Mannar, Ceylon Seas (Dendy).

Register Numbers, Locality, &c. IV. 2 a, b, dredged off S.W. coast of Beyt Island ;
XXXIV. 4, channel, W. side of S. end of Beyt Island, 3-4 fathoms, 3.1.06.

40. Clathria corallitincta Dendy.

Clathria corallitincta Dendy [1889].

Clathria frondifera Dendy [1905].

? Clathria frondifera Ridley [1884].

? Halichondria frondifera Bowerbank [1875].

There are three good-sized specimens and one small specimen of this sponge in the
collection. They have, each as a whole, a massive form with a slight tendency to
become lamellar, but each is in reality made up of an immense number of slender,
anastomosing branches, forming a close network, with the ends of the branches
projecting on the surface in the form of conuli. All three specimens exhibit large
pseudoscula on prominent parts of the sponge.

The spiculation calls for no special comment except as regards the presence of
numerous large, stout, strongly bent toxa.

It is quite possible that Bowerbank's Halichondria frondifera may be a very
variable and widely distributed species, with which the Indian specimens might be
identified, as I did with Ceylon specimens in 1905 ; but I am now more inclined, in view
of Bowerbank's original description and the new material to hand, to regard the Indian
and Ceylon specimens as belonging to a distinct species, as I did in 1889, under the
name Clathria corallitincta.

Previously known Distribution. Gulf of Mannar, Ceylon (Dendy).

Register Numbers, Localities, &c. IV. 14, off S.W. coast of Beyt Island ; XVIII. 2,
channel, W. of Beyt Island, 3-4 fathoms, January, 1906; XIX. 1, Vamiani Point,
January 5, 1906 ; XXXIV. 2, channel, W. side of S. end of Beyt Island, 3-4 fms.,
3.1.06.

41. Clathria spiculosa Dendy.

Rhaphidophlus spiculosus Dendy [1889].
Clathria spiculosa Dendy [1905].
Clathria spiculosa var. ramosa Dendy [1905].
Clathria spiculosa vars. ramosa and macilenta [Hentschel 1912].
There are several specimens in the collection which I identify with this common

BENDY— NON-CALCAREOUS SPONGES 129

Ceylon species. The branching is extremely irregular, and the branches vary in form
from short, broad and flattened to slender, long and cylindrical. The long, slender-
branched specimens evidently belong to my var. ramosa, but I do not think that this
can be at all sharply distinguished from the type.

One of the specimens contains a number of grapnel spicules, which seem to be
foreign, as they were seen only in one small part.

Previously known Distribution. Gulf of Mannar, Ceylon Seas (Dendy) ; Aru Islands,
Arafura Sea (Hentscliel).

Register Numbers, Locality, &c. XX. 5, 10, Adatra.

42. Echinodictyum gorgonioides n. sp. — (Plate I., Figs. ia-iV ; Plate IV.,

Fig. 23.)

So far as external appearance goes this is perhaps the prettiest sponge in the
collection, with an extremely characteristic mode of growth. There are three separate
pieces of it, but they all come from the same locality and may possibly be parts of one
and the same specimen. The sponge (Fig. 23) consists of a number of slender stems
springing close together from a somewhat spreading, encrusting base. The stems rise
almost vertically from the points of attachment and bifurcate repeatedly and very
frequently. All the branches — and no distinction can be drawn between stems and
branches — are of about the same diameter, say 3 mm. They mostly lie in approximately
the same plane and occasionally anastomose with one another. The surfaces of the
branches are subrugose and minutely conulose. The branches terminate in rounded
extremities and there are no conspicuous vents.. The whole growth reminds one
forcibly of some species of Gorgonia.

The total height of the largest piece is about 110 mm., and the greatest width
nearly as much.

The colour in spirit (after formalin) is very pale brown. The texture is fairly tough,
very compressible and resilient.

The skeleton is composed of stout primary fibres, which branch repeatedly as they
approach the surface and are more or less interconnected by secondary fibres. The
main fibres consist chiefly of plumose columns of spined styli, some of which are com-
pletely embedded in the pale-coloured spongin, while others echinate the surface of the
fibre at various angles. The secondary fibres consist mainly or entirely of spongin,
echinated more or less by the spined styli.

Fairly numerous oxea (tornotoxea) accompany the fibres or are scattered between
them, but they rarely, if ever, form a spicular core to the fibre. They are not nearly
so numerous as the spined styli, which form the chief part of the spiculation.

Spicules. (1) Oxea (tornotoxea) (Fig. la) ; straight, slender, thicker at one end
than at the other, abruptly pointed at the thicker end, gradually sharp-pointed at the
other ; occasionally with a bulbous inflation at from | to \ the length from the thicker

130 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

end ; size about 0-14 by 0-0041 mm. (2) Spined styli (acanthostyli) (Figs. Ib, 76') ;
tapering gradually from the base to the sharply pointed apex ; straight ; rather sparsely
covered with small spines, except towards the apex ; size variable, say about 0-11 by
0-01 mm. ; a number of very slender forms also occur (Fig. 7&').

The sponge also contains numerous other types of spicule, most of which, at any
rate, have certainly been derived from other sponges, including Spirastrella, Esperella,
Donatia and Reniera. There are, however, a fair number of stoutish, strongly angulated
toxa, a very few minute Clathria-like, palmate isochelae, and a fair number of long,
straight, slender tylostyles with well-developed heads, which may possibly belong to
the species.

The plumose character of the main skeleton fibres suggests a close affinity with
the genus Plumohalichondria.

Register Number, Locality, &c. XXI. 1, 3, 4-7 fms. off Rupan Bandar and
Kutchegudh, 8.12.05.

43. Raspailia fruticosa var. tenuiramosa Dendy.

Raspailia fruticosa Dendy [1887].

Raspailia, fruticosa var. tenuiramosa Dendy [1905].

Raspailia fruticosa var. aruensis Hentschel [1912].

The two specimens in the collection differ from the Ceylonese types of the variety
in their much more sparingly branched character, being, in fact, less instead of more
bushy than the type of the species. Both specimens, however, have much more
slender branches than the type of the species. As regards skeleton arrangement and
spiculation they agree very closely with the types of the variety.

Previously known Distribution of the Species. Gulf of Mannar (Dendy) ; Aru
Inlands, Arafura Sea (var. aruensis Hentschel).

Register Numbers, Localities, &c. II. 13, off Poshetra, January 7, 1906 ; XXIX. 2,
N. of Poshetra, 3f-4 fms., 20.12.05.

44. AearnilS tortilis Topsent[18926is, 1897, 1904].

I identify with this widely distributed but apparently rare species a small, pre-
sumably encrusting sponge of a brown colour and irregular shape. The spiculation
agrees closely with that described and figured by Topsent for his species, even down
to the minute spination of the bases of the stout styli and the ends of the diactinal mega-
scleres. The very characteristic grapnel-spicules, with usually four strongly recurved,
sharp hooks at the apex, strongly spined shaft, and base with spines curved in the
opposite direction, are identical. The toxa, however, appear to be all of the strongly
arcuate form, though varying much in dimensions. The chief difference that I have
been able to detect, however, lies in the presence of a considerable amount of spongin,
partially uniting and enveloping some of the spicules, but this difference can hardly

BENDY— NON-CALCAREOUS SPONGES 131

be regarded as of specific value and may be associated with temperature conditions.
A great many foreign spicules are present in the sponge, possibly owing to its association
with other sponges in the dredge.

Previously known Distribution. Mediterranean, Azores, Amboina (Topsent).

Register Number, Locality, &c. XXIX. 3, N. of Poshetra, 3f-4 fms., 20.12.05.

45. Bubaris radiata n. sp. (Plate I., Figs. 8a-8b ; Plate IV., Figs. 24a, 246).

Sponge (Figs. 24a, 24&) encrusting, cushion-shaped, composed of close-set, stout,
radiating skeletal columns, united by a small quantity of gelatinous soft tissue.
Surface conulose owing to the projecting ends of the spicular columns, with a thin,
translucent dermal membrane in the grooves between the conuli. No oscula
or pores seen. The largest piece (R.N. II. 10) measures about 18 mm. in maximum
diameter and 6 mm. in thickness, the thickness being equal to the length of the
spicular columns, which extend from the base to the upper surface almost without
branching and approximately parallel to one another. The individual columns are
about 1 mm. in diameter. The coloiir in spirit is light, dull yellow.

A second specimen (R.N. III. 7 a), rather smaller, looks like a fragment of an almost
spherical sponge. The spicular columns are rather stouter, a little more branched,
and radiate almost from a common centre. The colour in spirit is rather darker.

The skeleton consists of very strong columns of spicules arranged in a plumose
fashion. Each column contains a very dense axis composed of an interlacement of
short strongyla, in which are implanted the bases of some of the stout styli which radiate
obliquely outwards and upwards from the axis. There is sometimes a tendency to
the formation of secondary plumose spicular columns coming off from the axis of the
primary. At the ends of the columns the apices of the styli project freely beyond
the surface of the sponge, but lower down between the columns they are completely
enveloped in gelatinous tissue.

Spicules. (1) Stout styli (Fig. 8a), more or less bent towards the base,
tapering gradually to a sharply pointed apex ; commonly measuring about 0-55 by
0-026 mm., often smaller and occasionally a trifle larger. (2) A few very much longer
and more slender styli (Fig. 8a'") occur in boiled-out preparations, measuring up to
about 1-3 by 0-016 mm. (3) Strongyla (Fig. 8b) ; comparatively short and irregularly
curved or bent to a varying but not very high degree, approximately equal-ended ;
size commonly about 0-26 by 0-01 mm., but variable. Intermediate forms are repre-
sented by Figs. 8a' and 80".

In general appearance and skeleton arrangement this sponge reminds one very
strongly of the common Ceylon species Aulospongus tubulatus. The latter, however,
has not got any strongyla in the axes of the plumose spicular columns, while, on the
other hand, it has got minutely spined, echinating styli, which are absent in the present
species.

132 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

From Bubaris vermiculata the species differs in its mode of growth and in the
replacement of the " vermicular " spicules by much less bent strongyla.

The descriptions, measurements and figures of the spicules, as well as the figures
of the external form, are taken from R.N. III. 7 a, which may be regarded as the type
of the species.

Register Numbers, Localities, &c. II. 10, off Poshetra, 7 January, '06 ; III. la, b,
(two fragments), off Dwarka.

46. Spirastrella vagabunda var. tubulodigitata Dendy.— (Plate IV., Fig. 25).

(For possible Synonymy, vide Vosmaer [1911]).

There is in the collection a single specimen (Fig. 25) which agrees very closely in
external form and spiculation with the Ceylon types. It consists of a single tubular
process with terminal vent, and a few much smaller, irregularly ramified, blind pro-
cesses, all arising from a common base containing much coarse sand.

Previously known Distribution of the Variety. Gulf of Mannar and Ceylon Seas
(Dendy).

Register Number, Locality, &c. IV. 10, off S.W. coast of Beyt Island.

47. Placospongia carinata (Bowerbank).

(For Literature and Synonymy vide Vosmaer and Vernhout [1902] and Dendy
[1905]).

This remarkable sponge appears to be quite common in the Indian Ocean.

Previously knoivn Distribution. Tropical Seas between 30° N. and 20° S. of the
equator (Vosmaer and Vernhout, &c.).

Register Numbers, Localities, &c. XXVI. 9, Adatra Reefs, 25 December, '05 ;
XXVJI. 2, 3 and XXXIII. 4, Dhed Mora and adjacent rocky ground between Beyt and
Aramra, 1 fm., 21.12.05.

48. Cliona coronaria (Carter).

Suberites coronarius Carter [1882, 1887].

This interesting species was first described by Carter from specimens in the Bower-
bank collection in the British Museum, coming from Honduras, Jamaica and the
Bahamas. These specimens were described as " massive, lobate, verrucose on the
surface." The characteristic microsclere was described as a " Spinispirula consisting of
one bend, semicircular, with the spines on the outside and over the ends only ; spines
capitate and in single file."

Mr. Carter's original preparations of these three sponges are in my possession,
and I am able to verify the general accuracy of his brief description and figures of the
spiculation. The capitate character of the spines of the microscleres is, however, not
always recognisable, and may be in part due to optical illusion.

DENDY— NON-CALCAKEOUS SPONGES 133

In 1887 the same species was recorded by Mr. Carter from the Mergui Archipelago
and some interesting particulars added as to its mode of growth, as follows : — " Its
growth is more remarkable than in that [the Honduras] example, for it is laminar,
and extends in a horizontal direction for several square inches ; the superficial stratum,
which is comparatively thin and buff-yellow in colour, changes to black or dark brown
in the cancellated cavities to be presently mentioned for half an inch downwards, where
it rests on granite. The explanation of this abrupt termination is that the lower
portion is mingled with a layer of coral which has been cancellated by the excavating
habit of these sponges, which exhibit an apparent fondness for calcareous material,
whether in a mineral or organic form."

Mr. Carter thus clearly recognises that his Suberites coronarius may, at any rate
under some circumstances, be an excavating sponge. A specimen in Mr. Hornell's
collection (R.N. XXVI. 10), also excavating and encrusting a piece of coral, agrees
very closely with the Mergui specimen, and its examination, I think, fully justifies the
transference of the species to the genus Cliona. In all probability the massive West
Indian specimens stand in exactly the same relation to the excavating Indian Ocean
specimens as does the massive " Raphyrus griffithsii " to the excavating Cliona celata
of European seas, or to the encrusting form of the latter described by Topsent [1900].

Mr. Hornell's specimen consists, in the first place, of a thin crust (about 1 mm.
thick), with smooth outer surface. The underlying coral has been eroded and largely
disintegrated by the sponge, giving rise to the " cancellated " structure described by
Mr. Carter, which, when teased up and examined microscopically, is seen to consist
of a mixture of coral fragments and sponge.

In the outer portion of the sponge the skeleton consists of dense, irregular wisps
of tylostyles running towards the surface, where they form a thick dermal pile with
outwardly directed apices. In the cancellated portion the tylostyles seem to be quite
irregularly scattered. The spiculation consists of tylostyles and spirasters only ; I
have seen no oxea such as sometimes occur in Cliona (fide Topsent). The tylostyles
are straight or very slightly curved, usually very sharply and gradually pointed. They
have well developed, almost spherical heads. When fully grown they measure about
0-35 by 0-012 mm., with head 0-014 mm. in diameter. The greatest diameter of the
shaft is at about one-third of the distance from head to apex. The spirasters are very
slender and measure about 0-02 mm. in a straight line from end to end. Many of them
have the typical " semicircular " form described by Carter, but many show a very
obvious spiral twist. They resemble bent fragments of broken fretsaws. It is not
easy to assure oneself that the short spines are really capitate, though they may have
the appearance of being so.

As might be expected, the spiculation, as well as the mode of growth, of the Okha-
mandal sponge seems to agree more closely with that of the Mergui than with that of
the West Indian specimens. Thus the heads of the tylostyles are more nearly spherical

134 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

in both the Indian Ocean specimens (compare Carter's figures of the Mergui and
West Indian sponges), but I do not think there is anything to justify a specific
distinction.

It should be noted that the Okhamandal specimen was first preserved in formalin,
which may possibly have assisted in bringing about the eroded condition of the coral
on which it is growing.

Some larger fragments from the same locality (R.N. XXVI. 2), possibly parts of the
same specimen, have been completely stripped off from the substratum in the form
of a thin sheet about 48 square centimetres in extent and about 1 mm. thick. They
have an almost smooth surface, minutely reticulate in places when viewed under a
lens, and no visible vents.

Previously known Distribution. Honduras, Jamaica, Bahamas, Mergui (Carter).

Register Number, Locality, t&c. XXVI. 2, 10, Adatra Reefs, 25 December, '05.

49. Suberites carnosus (Johnston) var.

(For Literature and Synonymy, vide Topsent [1900]).

This common and widely distributed species is represented in the collection by
two specimens of massive, subovoid form, without any indication of stalks, and with
no visible vents. One of the two specimens (R.N. XIX. 3) measures about 44 by
29 mm. and is compact and solid. The other (R.N. XIX. 2), of about the same size,
has been damaged in collecting, and a superficial cortical layer, made coherent by the
extremely dense surface-skeleton, has been separated in large measure from the
remainder, which seems to have shrunk away or been partially removed, so that
the specimen looks like a broken egg with a dried up yolk adherent to the shell at one
side only. In both the surface is smooth but very minutely velvety, and the colour
throughout (in spirit) is pale yellow.

The skeleton consists exclusively of tylostyles. In the interior of the sponge
these spicules are thickly but quite irregularly scattered, while at the surface they
arrange themselves with their apices projecting outwards, not in distinct brushes but
in a dense, continuous layer.

The tylostyles are approximately straight, gradually and sharply pointed, and
with well rounded heads only occasionally of the " enormispinulate " type. They vary
in size up to about 0-77 by 0-018 mm.

A direct comparison with a preparation of Johnston's type-specimen shows that
the latter has decidedly smaller spicules with a strong tendency to develop a secondary
inflation just below the head. The measurements of the spicules in our variety also
exceed the range of variation for the species given by Topsent.

Previously known Distribution. Atlantic, Mediterranean, Red Sea, Indian Ocean
and Australian Seas (vide Ridley and Dendy [1887] and Topsent [1900]).

Register Numbers, Locality, &c. XIX. 2, 3, Vamiani Point, January 5, '06.

DENDY— NON-CALCAREOUS SPONGES 135

so. Suberites flabellatus Carter.

Suberites flabellatus Carter [1886].

? Suberites globosa (elongated form) Carter [1886].

1 Suberites carnosus Keller [1891].

Suberites flabellatus Dendy [1897].

A single good specimen in the collection closely resembles in external form Keller's
figure of " Suberites carnosus " from the Red Sea. It agrees very closely, however,
with the Victorian species described by Carter and myself, and I have no hesitation
in identifying it with that, which may be, after all, merely a variety of S. carnosus.
The tylostyles in the Okhamandal sponge are long, straight and slender, and they all
seem to have well-developed heads, usually of the " enormispinulate " type. In the
interior of the sponge they attain a length of about 0-7 mm., with a diameter of about
0-01 mm. ; but they are much smaller in the surface-brushes.

Previously known Distribution. Near Port Phillip Heads, Australia (Carter, Dendy) ;
? Red Sea (Keller).

Register Number, Locality, &c. II. 1, off Poshetra, January 7, '06.

51. Suberites cruciatus Dendy [1905].

I identify with this species, originally described from Ceylon, two small, irregularly
massive specimens, each of which shows a tendency to give off digitiform processes.
The larger of the two measures about 18 mm. in greatest length by 9 mm. in greatest
.breadth. The colour in spirit is pale yellow. The surface is uneven but nearly smooth,
and shows small, rounded, translucent (pore ? — ) areas, as described for the type.

The arrangement of the skeleton agrees on the whole with that found in the type.
The tylostyles exhibit the same peculiar form of the heads but are considerably larger
than in the type ; in fact, they may be nearly twice as large, at any rate in the deeper
parts of the sponge, while they diminish in size in the surface-brushes. A large number
of tylostyles occur scattered tangentially in the dermal membrane between the ends
of the surface-brushes, a feature which is not conspicuous in the type.

Previously known Distribution. Ceylon (Dendy).

Register Numbers, Locality, &c. III. 10 a and b (possibly parts of same specimen) ;
dredged off Dwarka.

52. Polymastia gemmipara n. sp.— (Plate I., Figs. 9o, 96; Plate IV., Figs.

26a, 266).

The single specimen (Figs. 26a, 266) has evidently been torn off from the sub-
stratum, part of which, in the form of a small pebble, remains attached at one side.
It now has the form of a hollow, thin-walled sac, widely open below, where it has been
damaged, and produced above into five slender, hollow processes or fistulse, one of

136 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

which is branched. In life, no doubt, the interior of the sponge was filled with soft,
pulpy, choanosomal tissue, part of which remains in the form of irregular masses
adherent to the inner surface of the cortex, as shown at ch. in Fig. 266.

The thin cortex is supported internally by a few very stout spicular columns, one
of which is shown at sp. c. in the illustrations. No doubt in life these columns, of which
I have only been able to find two in the specimen before me, were attached to the
substratum, and formed pillars supporting the dome-like cortex.

The fistulee are very remarkable structures. They have no visible openings, but
most of them taper off distally into fine threads, which in three cases are swollen out to
form either one or two small buds, as shown at b. in the figures.

The colour of the cortex (in alcohol) is white, that of the soft internal pulp very
pale yellow.

The greatest diameter of the specimen, at the base, is 17 mm.

The skeleton may be subdivided as follows : (1) that of the soft internal choano-
some or pulp consists of loosely scattered tylostyles or subtylostyles, varying in size
but mostly small and often very slender ; (2) that of the internal spicular columns
consists of a dense mass of relatively large subtylostyles closely packed together length-
wise ; (3) that of the cortex consists of an inner and an outer portion ; the inner portion
is a fairly close interlacement of the larger subtylostyles lying tangentially, not much
more than one layer thick, with a tendency to the formation of broad bands, which can
be seen under a pocket-lens converging towards the bases of the fistular processes ;
the outer portion is a thin but rather close pile or fur of short tylostyles or subtylostyles
arranged more or less vertically to the surface, with outwardly projecting apices ; '
(4) that of the fistular processes is merely a continuation of the cortical skeleton. The
principal bundles of the larger subtylostyles run lengthwise in the wall of the fistula ;
where the fistula is drawn out into a solid thread at the end they unite to form a single
stout spicular fibre, and in this region the surface-fur of short spicules is almost absent,
but becomes strongly developed again when the thread swells out to form a bud.

Spicules. (1) Subtylostyles of the internal columns and cortex (Fig. 9a) ; straight,
slender, tapering gradually to a finely pointed apex, also tapering gradually to the base,
where there is usually a very slightly developed head ; the base sometimes appears
constricted somewhat suddenly, the " head " being a good deal narrower than the
adjacent part of the shaft, as in the specimen figured ; commonly measuring about
0-72 by 0-0164 mm. ; (2) small tylostyles (Fig. 9fe) ; chiefly in the surface pile ; usually
curved ; apex gradually sharp-pointed ; with rather feebly developed head of smaller
diameter than the middle of the spicule ; size about 0-15 by 0-0068 mm. ; (3) spicules
of the soft internal pulp ; a mixture of the two kinds already • described, together
with intermediate forms and very slender forms which are probably young ; perhaps
never quite so large as in the spicular columns and cortex, and usually much smaller.

This species seems to be intermediate in character between the genera Polymastia

BENDY— NON-CALCAREOUS SPONGES 137

and Quasillina as at present understood, the soft, pulpy internal structure and thin
cortex being suggestive of the latter, and the hollow fistular processes of the former.
It seems quite possible that, after all, the distinction between these two genera may
have to be abandoned.1

The process of bud-formation by gemmation from the filiform ends of tapering
fistula? appears to be identical with that described, with admirable illustrations, by
Merejkowsky [1878] in Polymastia (Rinalda) arctica, a species from the White Sea.
A somewhat similar process is, of course, familiar in certain other genera, such as
Donatia.

In Polymastia arctica there are one or more osculum-bearing fistula) between the
gemmiparous ones. This may also be the case with the present species, but the
specimen is not in a fit condition to enable me to determine the point with certainty.

Register Number, Locality, &c. IV. 7, dredged off S.W. coast of Beyt Island.

53. Megalopastas retiaria n. sp.— (Plate IV., Fig. 27.)

This is certainly one of the most interesting, and at the same time one of the most
beautiful species in the collection, in which it is well represented by three good spirit
specimens and a couple of washed-out skeletons. The general appearance of the sponge
is well shown in Fig. 27, which represents the best preserved specimen twice the natural
size. The sponge is irregularly lobose, or simply massive. Apparently it is sessile,
with several points of attachment to relatively small foreign objects, as though it had
grown on a gravelly or shelly bottom. The surface is produced into moderate-sized,
acute conuli, scattered at rather wide and irregular intervals, the height of the conuli
being about 1-2 mm. Internally the sponge is cavernous and the wide vestibules
open on the surface by large pseudoscula. These pseudoscula, of various shapes and sizes,
are frequently (perhaps always in life) covered over by a very beautiful network, which
is a continuation of the delicately reticulate dermal membrane, but with much coarser
meshes. The dermal membrane in general covers the whole sponge like a gauzy veil.
It is strengthened by a close reticulation of delicate, deeply staining, fibrillar bands (not
horny). Under a pocket-lens the principal lines of this network appear (in the spirit
specimens) as fine white lines radiating from the apices of the conuli into the hollows
between, where they lose themselves in a network of finer lines. The meshes of this
network are occupied by the pore-bearing dermal membrane, which is itself reduced
to a secondary, quite microscopic network by the very numerous inhalant pores, each
only about 0-03 mm. in diameter. The dermal membrane is interrupted here and
there, pretty frequently, by small circular apertures about 1 mm. in diameter. These
apertures have well-defined margins, being bounded each by an unusually large, circular

1 Cf. my remarks OH Quasil/ina brevis in Journ. Linn. Soc. Zool., Vol. XXXII. (1914), p. 271.

138 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

mesh of the network of fibrillar bands described above. They are the openings of deep
cylindrical canals of similar diameter, and probably represent true oscula. Between
them numerous smaller apertures, the openings of the inhalant canals leading from
the subdermal cavities, can be seen through the transparent dermal membrane. In
some places the soft tissues seem to have shrunk away from the skeleton, showing
the coarsely reticulate, dark-coloured, horny fibre projecting beyond the surface ; while
the pale-coloured growing points of the primary fibres appear in the conuli, supporting
the dermal membrane.

The largest specimen (R.N. VIII.) is an irregular lobose mass about 100 mm. in
diameter. The texture (in spirit) is very soft and compressible, but resilient owing to
the coarse, stiff, horny skeleton. The colour in life (R.N. VIII.) was recorded as pink,
in spirit all three specimens are very pale grey.

Of two specimens preserved in formalin (R.N. XXIX. 1) practically nothing re-
mained but the skeleton.

The skeleton is a very coarse reticulation of dark-brown, almost black, horny
fibre. It is only occasionally possible to recognise main stems springing from points
of attachment. Indeed the abundant development of secondary connecting fibres
almost completely conceals any original tree-like growth there may have been, except
towards the surface, where branching main fibres terminate in the conuli. The con-
necting fibres are developed chiefly in the angles of the main branching system, with
the result that we can distinguish meshes of two quite distinct orders, large and small,
the small ones occurring in groups at the nodes of the reticulation formed by the large
ones. The large meshes average perhaps 5 mm. in diameter, the small ones, say, about
0-5 mm., but very variable. The thickest primary fibres measure up to about 0-34 mm.
in diameter and the thinnest secondary ones about 0-034 mm., between which extremes
all gradations occur. The fibres are entirely free from foreign inclusions and consist
of more or less numerous concentric layers of spongin. It is not possible to distinguish
sharply between the so-called pith and the surrounding spongin-lamellae, and, in spite
of the numerous descriptions in which such distinction is insisted upon, I believe that
this is frequently the case in the Aplysillidse.

There is no dermal or subdermal reticulation of horny fibres.

The canal-system is typically aplysillid. The flagellate chambers are very large
and thimble-shaped, measuring about 0-17 by 0-085 mm. Favourable sections show them
arranged in a single, much folded layer between the inhalant and exhalarit canals.
Each chamber has numerous prosopyles.

Register Numbers, Localities, &c. II. 3, off Poshetra, January 7, '06 ; IV. 13,
off S.W. coast of Beyt Island ; VIII., Mangunda Reef ; XXIX. 1 (two skeletons), N. of
Poshetra, 3f-4 fms., 20.12.05.

DENDY— NON-CALCAREOUS SPONGES 139

54. Darwinella australiensis Carter.

Darwinella australiensis Carter [1885].

Darwinella australiensis Lendenfeld [1889].

Darwinella australiensis Topsent [1905].

Darwinella australiensis Hentschel [1912].

There are in the collection the remains of two or three specimens of this sponge,
which has thickly encrusted some of the large polychsete worm-tubes before mentioned.
Unfortunately the specimens were preserved in formalin and are very badly macerated.
There can, however, be little doubt about the identification.

The surface, where preserved, is strongly conulose, the colour (in alcohol after for-
malin) is dull red. The largest specimen is about 25 mm. in maximum thickness
(the height of the largest skeletal trees).

The skeleton is composed of elegant little trees of pale amber-coloured horny fibre,
growing up vertically and side by side from a thin basal layer of spongin. The trees
are well branched, but anastomosis between the branches takes place only rarely.

The horny spicules seem to be much less numerous than in Australian specimens.
They are triradiate, with long, slender, gradually sharp-pointed rays. They all seem
to lie freely in the soft tissues. I have measured the rays up to about 0-85 mm. in
length. I have observed in some of the older fibres apparently the same parasitic
fungus (?) as I described [1905] in the case of Darwinella simplex, Topsent, from Ceylon.
It now appears to me very doubtful whether that species is distinguishable from
D. australiensis.

Previously known Distribution. Victoria, Australia (Carter, Lendenfeld) ; Arafura
Sea (Hentschel).

Register Number, Locality, &c. XXV. 2, Kiu, littoral at low water, 24.12.05.

55. Spongelia fragilis var. ramosa (Schulze).

Spongelia pallescens subspecies fragilis var. ramosa Schulze [1879].
Spongelia fragilis var. irregularis (pars) Lendenfeld [1889].
Spongelia fragilis var. ramosa Dendy [1905].
? Spongelia fragilis var. claihrata Hentschel [1912].

There are several specimens of this sponge in the collection. I pointed out in
my Ceylon Report that the species, at any rate, is probably cosmopolitan, if not the
variety.

Previously known Distribution of Variety. Adriatic (Schulze) ; Ceylon (Lendenfeld,
Dendy) ; ? Aru Islands (Hentschel).

Register Numbers, Localities, &c. XII., XXXII. 3, off Beyt ; XXV. 5, Kiu, littoral.

140 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

56. Spongelia cinerea (Keller).

Dysidea cinerea Keller [1889].

Spongelia elastica var. crassa Dendy [1905].

There are three specimens of this sponge in the collection, one large one (R.N. IV. 1)
very similar to that figured in my Report on Professor Herdman's Ceylon Sponges,
and two small ones more nearly resembling the figure given by Keller. I think there
can be little doubt that my var. crassa is identical with Keller's Dysidea cinerea from the
Red Sea, which is possibly, merely a variety of Spongelia elastica, which, in turn,
according to Schulze [1879]. is merely a subspecies of Spongelia pallescens. It is very
doubtful how far it is desirable to separate these different forms.

Previously known Distribution. Red Sea (Keller) ; Gulf of Mannar (Dendy).

Register Numbers, Localities, &c. IV. 1, 15, 23, dredged off S.W. coast of Beyt
Island.

57. Spongelia elegans Nardo var.

(For Synonymy, vide Schulze [1879] and Lendenfeld [1889].)

A number of branched and anastomosing, digitiform pieces, all from the same
locality and probably all parts of one specimen, closely resemble Schulze's figure of
Spongelia elegans. The specimen seems, however, to have been less robust than that
figured by Schulze and the conuli less strongly developed. The colour in spirit (after
formalin) is light brown ; texture very soft and compressible. The skeleton is a loose,
irregular network of rather slender, strongly arenaceous fibre, in which the fibres are all
equally charged with sand, there being no secondary horny fibres more or less free
from sand and connecting the fibres which run into the conuli, as described by Schulze.
Stained preparations, however, show darkly staining fibrillar bands of nucleated cells
running lengthwise through the sponge. These commonly contain a few sand-grains.
Their relation to the ordinary skeletal fibres is doubtful, though they appear to be
connected with them in places.

Schulze says nothing about the occurrence and distribution of spongin, and one
merely assumes that the sand-free secondary fibres in his material are composed of
that substance. In the Okhamandal specimen spongin is nowhere strongly developed,
and I have recognised it only in the sandy fibre.

Previously known Distribution. Mediterranean (Nardo, Schmidt, Schulze, &c.) ;
East Coast of Australia (Lendenfeld).

Register Number, Locality, &c. XXIII. 4, off Dwarka, 15-17 fms., 12.12.05.

DENDY— NON-CALCAREOUS SPONGES 141

58. Hippospongia clathrata (Carter).
Hircinia clathrata Carter [1881].
Hircinia clathrata Dendy [1887, 1889].
Hyatella clathrata Lendenfeld [1889].
Hippospongia clathrata Dendy [1905].

There are several large pieces of this sponge in the collection.

Previously known Distribution. Gulf of Mannar and Red Sea (Carter) ; Gulf of
Mannar (Dendy) ; Australia (Dendy, Lendenfeld) ; American coast of N. Atlantic
(Lendenfeld).

Register Numbers, Localities, &c. XVIII. 1, channel, W. of Beyt Island, 3-4 fms.,
January, '06 ; XXXII. 6, off Beyt.

LIST OF LITERATURE REFERRED TO.

1915. Annandale, N. " Fauna of the Chilka Lake. Sponges." (Mem. Indian Museum.. Vol. V.).
1903. Arnesen, Emily. " Spongien von der norwegischen Kiiste. II. Monaxonida : Halichon-

drina." (Be.rge.ns Mus. Aarbog, 1903, No. 1.)
1905. Baer, L. " Silicispongien von Sansibar, Kapstadt und Papeete." (Archiv fur Naturge-

schichte, Jahrgang 72, Bd. 1).

1866. Bowerbank, J. S. " A Monograph of the British Spongiadte." Vol. II.
1873. Bowerbank, J. S. " Report on a Collection of Sponges found at Ceylon by E. W. H.

Holdsworth, Esq." (Proc. Zool. Soc. Land,., 1873.)
1875. Bowerbank, J. S. " Contributions to a General History of the Spongiadee." Part VII.

(Proc. Zool. Soc. Lond., 1875.)
1869. Carter, H. J. " Description of a Siliceous Sand Sponge found on the South-East Coast

of Arabia (Tethya dactyloidea) ." (Ann. and Mag. Nat. Hist., Series 4, Vol. III.)

1872. Carter, H. J. " Additional Information on the Structure of Tethya dactyloidea, Carter."

(Ann. and Mag. Nat. Hist., Series 4, Vol. IX.)

1873. Carter, H. J. " On Two New Species of Gumminese (Corticium abyssi, Chondrilla austral-

iensis) with special and general Observations." (Ann. and Mag. Nat. Hist., Series 4,
Vol. XII.)

1874. Carter, H. J. " Descriptions and Figures of Deep-sea Sponges and their Spicules from the

Atlantic Ocean," &c. (Ann. and Mag. Nat. Hist., Series 4, Vol. XIV., pp. 207 and
245.)

1879. Carter, H. J. " Contributions to Our Knowledge of the Spongida." (Ann. and Mag.

Nat. Hist., Series 5, Vol. Ill, pp. 284 and 343.)

1880. Carter, H. J. " Report on Specimens Dredged up from the Gulf of Manaar," &c. (Ann.

and Mag. Nat. Hist., Series 5, Vol. VI, pp. 35 and 129).

1881. Carter, H. J. " Supplementary Report on Specimens Dredged up from the Gulf of Manaar,"

&c. (Ann. and Mag. Nat. Hist., Series 5, Vol. VII.)

K

142 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

1882. Carter, H. J. " Some Sponges from the West Indies and Acapulco," &c. (Ann. and Mag.

Nat. Hist., Series 5, Vol. IX., pp. 266 and 346.)
1882 bis. Carter, H. J. " New Sponges, Observations on Old Ones, and a Proposed New Group

(Phceodictyina)." (Ann. and Mag. Nat. Hist., Series 5, Vol. X.)
1885. Carter, H. J. " Descriptions of Sponges from the Neighbourhood of Port Phillip Heads,

South Australia." (Ann. and Mag. Nat. Hist., Series 5, Vol. XV., p. 196.)

1885. bis. Carter, H. J. "Descriptions of Sponges from the Neighbourhood of Port Phillip Heads,

South Australia, continued." (Ann. and Mag. Nat. Hist., Series 5, Vol. XVI., p. 277.)

1886. Carter, H. J. " Descriptions of Sponges from the Neighbourhood of Port Phillip Heads,

South Australia, continued." (Ann. and Mag. Nat. Hist.: Series 5, Vol. XVII.,
p. 112.)

1887. Carter, H. J. " Report on the Marine Sponges, chiefly from King Island, in the Mergui

Archipelago," &c. (Journ. Linn. Soc. Zool. Vol. XXI.).
1887. bis. Carter, H. J. " Description of Chondrosiaspurca, n. sp., from the South Coast of Australia."

(Ann. and Mag. Nat. Hist., Series 5, Vol. XIX.)
1887. Dendy, A. " The Sponge Fauna of Madras," &c. (Ann. and Mag. Nat. Hist., Series 5,

Vol. XX.)

1889. Dendy, A. " Report on a Second Collection of Sponges from the Gulf of Manaar." (Ann.

and Mag. Nat. Hist., Series 6, Vol. III.)

1890. Dendy, A. " Observations on the West Indian Chalinine Sponges, with Descriptions of

New Species." (Trans. Zool. Soc. Lond. Vol. XII.)

1895. Dendy, A. " Catalogue of Non-Calcareous Sponges collected by J. Bracebridge Wilson,"

&c. Part 1. (Proc. Royal, Soc. Victoria. Vol. VII. n. s.)

1896. Dendy, A. " Catalogue of Non-Calcareous Sponges collected by J. Bracebridge Wilson,"

&c. Part 2. (Proc. Royal Soc. Victoria. Vol. VIII. n. s.)

1897. Dendy, A. " Catalogue of Non-Calcareous Sponges collected by J. Bracebridge Wilson,"

&c. Part 3. (Proc. Royal Soc. Victoria. Vol. IX. n. s.)

1905. Dendy, A. " Report on the Sponges collected by Professor Herdman at Ceylon in 1902."
(Report on Pearl Oyster Fisheries. Part 3. Royal Society.)

1915. Dendy, A. " Report on the Calcareous Sponges," &c. (Report to the Government of Baroda

on the Marine Zoology of Okhamandal in Kattiawar, Part II.)

1916. Dendy, A. " Report on the Homosclerophora and Astrotetraxonida collected by H.M.S.

Sealark in the Indian Ocean." (Trans. Linn. Soc. Lond. Zool. Series 2, Vol.

XVII. In the Press.)
1870. Ehlers, E. " Die Esper'schen Spongien in der Zoologischen Sammlung der K. Universitiit

Erlangen." (Programm zum Eintritt in den Senat, Erlangen, 1870.)
1794-1806. Esper, E. J. C. " Fortsetzung der Pflanzenthiere." (Niirnberg.)
1867. Gray, J. E. " Notes on the Arrangement of Sponges, with the Descriptions of some New

Genera." (Proc. Zool. Soc. Lond., 1867.)
1909. Hentschel, E. " Tetraxonida " 1 Teil. (Die Fauna Siidwest-Australiens. Bd. II. Jena.)

1911. Hentschel, E. " Tetraxonida " 2 Teil. (Die Fauna Siidwest-Australiens. Bd. III. Jena.)

1912. Hentschel, E. " Kiesel- und Hornschwamme der Aru-und Kei-Inseln." (Abh. Senckenberg.

Nat. Ges. Frankfurt a.M. Bd. XXXIV.)
1914 (?) Hentschel, E. " Monaxone Kieselschwamme und Hornschwamme." (Deutsche Siidpolar-

Expedition 1901-1903.)
1880. Keller, C. " Neue Coelenteraten aus dem Golf von Neapel." (Archiv Mikrosk. Anat.

Vol. XVIII.)

BENDY— NON-CALCAREOUS SPONGES 143

1889. Keller, C. " Die Spongienfauna des rothen Meeres." I. Halfte. (Zeit. wiss. Zool. Bd.
XLVIII.)

1891. Keller, C. "Die Spongienfauna des rothen Meeres." II. Halfte. (Zeit. wiss. Zool.

Bd. III.)
1900. Kirkpatrick, R. " On the Sponges of Christmas Island." (Proc. Zool. Soc. Lond.,

1900.)
1914. Lebwohl, F. " Japanische Tetraxonida." I., II., III., IV. (Journ. Coll. Sci. Imp. Univ.

Tokyo. Vol. XXXV.)

1886. Lendenfeld, R. von. " A Monograph of the Australian Sponges. Part IV. The

Myxospongise." (Proc. Linn. Soc. N.S.W. Vol. X.)
1889. Lendenfeld, R. von. " A Monograph of the Horny Sponges." (Royal Society.)

1897. Lendenfeld, R. von. " Spongien von Sansibar." (Abh. Senckenberg. nat. Ges. Frankfurt

a.M. Bd. XXI.)

1903. Lendenfeld, R. von. " Tetraxonia." (Das Thierreich, Berlin.)

1906. Lendenfeld, R. von. "Die Tetraxonia." (Wiss. Ergeb. d. deutsch. TieJ 'see-Expedition auf
d. Dampfer " Valdivia," 1898-1899. Jena.)

1910. Lendenfeld, R. von. " The Geodidae." (" Albatross " Expedition. Mem. Mus. Comp. Zool.

Harvard College. Vol. XLI.)

1898. Lindgren, N. G. " Beitrag zur Kenntniss der Spongienfauna des Malayischen Archipels

und der chinesischen Meere." (Zool. Jalirb. Ab. Syst. Bd. XI.)

1878. Merejkowsky, C. " fitudes sur les Eponges de la Mer Blanche." (Mem. Acad. St. Peters-

burg. Vol. XXVI.)
1884. Ridley, S. 0. " Spongiida." (Zool. Coll. H.M.S. " Alert." Brit. Mus. Nat. Hist.)

1887. Ridley, S. 0. and Dendy, A. " Report on the Monaxonida," &c. (" Challenger " Reports

Zoology. Vol. XX.)

1911. Row, R. W. H. "Report on the Sponges collected by Mr. Cyril Grassland in 1904-5. Part II.

Non-Calcarea." (Reports on the Marine Biology of the Sudanese Red Sea. Journ.

Linn. Soc. Lond. Zool. Vol. XXXI.)

1862. Schmidt, 0. " Die Spongien des adriatischen Meeres." (Leipsig.)
1877. Schulze, F. E. " Untersuchungen iiber den Ban und die Entwicklung der Spongien. III.

Die Familie der Chondrosidse." (Zeit. wiss. Zool. Bd. XXIX.)

1879. Schulze, F. E. " Untersuchungen iiber den Bau und die Entwicklung der Spongien. VI.

Die Gattung Spongelia." (Zeit. wiss. Zool. Bd. XXXII.)

1888. Sollas, W. " Report on the Tetractinellida," &c. (" Challenger " Reports. Zoology.

Vol. XXV.)

1898. Thiele, J. " Studien iiber pazifische Spongien." I. Heft. (Zoologica. Heft 24.

Stuttgart.)

1899. Thiele, J. "Studien iiber pazifische Spongien." II. Heft. (Zoologica. Heft 24".

Stuttgart.)

1900. Thiele, J. " Kieselschwiimme von Ternate." I. (Abh. Senckenberg. nat. Ges. Frankfurt

a.M. Bd. XXV.)
1903. Thiele, J. " Kieselschwamme von Ternate." II. (Abh. Senckenberg. nat. Ges. Frankfurt

a.M. Bd. XXV.)
1905. Thiele, J. " Die Kiesel- und Hornschwamme der Sammlung Plate." (Zool. Jahrb. Supp.

VI. Bd. III.)

1892. Topsent, E. " Contribution a 1'etude des Spongiaires de 1'Atlantique Nord." (Camp.

Scient. du Prince de Monaco. Fasc. II.)

K 2

144 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

1892 bis. Topsent, E. " Diagnoses d'Fyponges nouvelles de la Mediterranee et plus particulierement

do Banyuls." (Arch, de Zool. exp. et gen. T. X., Serie 2.)

1893. Topsent, E. " Note sur quelques Eponges du Golfe de Tadjoura," &c. (Bull. Soc. Zool.
de France. T. XVIII.)

1893 bis. Topsent, E. " Mission scientifiqne de M. Ch. Alhiaud aux lies Sechelles (Mars-Mai, 1892),

Spongiaires." (Bull. Soc. Zool. de France. T. XVIII.)
1897. Topsent, E. " Spongiaires de la Bale d'Amboine." (Revue Siiisse Zool. &c. T. IV.)

1900. Topsent, E. " fitude monographiqne des Spongiaires de France. III. Monaxonida

(Hadromerina)." (Arch, de Zool. exp. et gen. T. VIII., Serie 3.)

1901. Topsent, E. "Spongiaires." (Result. Voyage Belgica. Zool.)

1904. Topsent, E. " Spongiaires des Agores." (Camp. Scient. du Prince de Monaco. Ease.

XXV.)

1905. Topsent, E. " iStude sur les Dendroceratida." (Arch, de Zool. exp. et gen. T. III.,

Serie 4).

1906. Topsent, E. " Sponges recueillies par M. Ch. Gravier dans la Mer Rouge." (Bull. Mus.

Hist. Nat. Paris, 1906.)

1911. Vosmaer, G. C. J. " The Porifera of the Siboga Expedition. II. The Genus Spirastrella."
(Siboga Expedition. Monographic Via').

1902. Vosmaer, G. C. J., and Vernhout, J. H. " The Porifera of the Siboga Expedition. I. The

Genus Placospongia." (Siboga Expedition. Monographic VI.)

1881. Wright, E. P. " On a New Genus and Species of Sponge with supposed Heteromorphic
Zooids." (Trans. Royal Irish Acad. Science, Vol. XXVIII.)

DESCRIPTIONS OF PLATES.
PLATE I.

Figs, la, 16. Cliondrilla agglutinans n. sp. (R.N. V. 1).

Fig. la. Spherasters from cortex. X 650.
,, Ib. Oxyasters from choanosome. X 650.

Figs. 2a-2c. Tetilla jnlula n. sp. (R.N. XXXV. 8 b)

Fig. 2a. Outer part of surface-brush of megascleres, partially separated by teasing. X 280.
,, 26. Cladome of anamonsene. X 860.
„ 2c. Sigmata. X 860.

Figs. 3a-3d. Tetilla barodensis n. sp. (R.N. XXIII. 8).

Fig. 3a. Orthotriaenes. x 70.

,, 36. Cladome of anatrisene. x 290.

„ 3c. Sigmata. X 870.

,, 3d. Trichodragma. x 385.

DEN DY— NON-CALCAREOUS SPONGES 145

Figs. 4a~&g. Esperella plumosa (Carter). (R.N. XXXII. 1).

Fig. 4«. Tylostyles. X 290.

„ 46. Large anisochela, front view. X 870.

„ 46'. „ „ side view, x 870.

„ 46". „ ,, end view (small end). X 870.

„ 4c. Small anisochela, front view. X 870.

„ 4c'. „ „ side view. X 870.

,, 4eZ. Small isochela, front view. X 870.

„ 4d'. „ „ side view. X 870.

„ 4e. Large sigma, side view. X 870.

' „ 4:6'. „ „ front view. X 870.

„ 4/. Small sigmata. X 870.

„ 4gr. Toxa. x 870.

Figs. 5a-56'"". Guitarra indica n. sp. (R.N. IV. 5).

Fig. 5a. Styli (tornostrongyla) . X 290.
„ 56. Placochela, front view. X 870.
„ 56'. „ | front view. X 870.

„ 56". ,, side view, x 870.

„ 5 6"'-56'"". ,, developmental stages, side views. X 870.

Figs. 6 0-6 e. Psammochela elegans n. gen. et sp. (R.N. XVIII. 3).

Fig. 6a. Styli. X 290.

,, 66. Larger isochela, front view. X 870.

,, 66'. „ „ side view. X 870.

„ 6c. Small isochela, front view. X 870.

„ 6c'. „ „ side view. X 870.

„ &d. Larger sigma, side view. X 870.

„ 6d'. „ „ back view. X870.

„ Qe. Smaller sigmata. X 870.

Figs, la-lb'. Echinodictyum gorgonioides 11. sp. (R.N. XXI.).

Fig. la. Tornotoxeote. x 380.
„ 76, 76'. Acanthostyli. x 380.

Figs. 8a-86. Bubaris radiaia n. sp. (R.N. III. la).

Figs. 8o-8a'". Styli of various forms. X 90.
Fig. 86. Strongyla of various forms. X 90.

Figs. 9a, 96. Pdlymastia gemmipara n sp. (R.N. IV. 7).

Fig. 9a. Subtylostyle from skeleton column. X 110.
„ 96. Small tylostyle. X 380.

146 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

PLATE II.

Figs. 10rt-10c. Tetilla dactyloidea (Carter) (R.N. II. 2). Three specimens ; one cut in half length-
wise to show exhalant canal-system, x 1J.
Fig. 11. Seniera hornelli n. sp. (R.N. IV. 4). xl£.

,, 12. Reniera fibmreticulata n. sp. (R.N. II. 8). X 2.

„ 13. JReniera semifibrosa n. sp. (R.N. XXXIII. 1.). Nat. size.

,, 14a. Halichondria reticulata Baer (R.N. II. 6). Nat. size.

,, 146. „ „ „ (R.N. II. 11).

,, 15. Siphonochalina minor n. sp. (R.N. XXVI. 4). Nat. size.

,, 16. Auletta lyrata var. glomerata Dendy (R.N. III. 3). X 2.

„ 17. Auletta elongata var. fruticosa nov. (R.N. XXIII. 2). Nat. size.

PLATE III.

Fig. 18. Ciocalypta dichotoma n. sp. (R.N. IV. 21). X 2.
„ 19. Esperella plumosa (Carter) (R.N. XXXII. 2). Nat. size.
,, 20a. Desmacidon minor n. sp. (R.N. XX. 6 a). Flabellate specimen ; surface without oscula.

Nat. size.
„ 206. Desmacidon minor n. sp. (R.N. XX. 6 b). Digitate specimen ; osculum-bearing surface.

Nat. size.
,, 21. Guitarra indica n. sp. (R.N. IV. 5). Several specimens attached to the branching tube

of a polychaete worm (Eunice, sp.) ; the uppermost (a) showing oscula. Nat. size.
„ 22a. Psammochela elegans n. gen. et sp. (R.N. IV. 20). Specimen preserved in alcohol, with

dermal membrane remaining, x 1|.
,, 226. Psammochela elegans n. gen. et sp. (R.N. XVIII. 3). Specimen originally preserved in

formalin ; the dermal membrane has been removed by maceration, x 1|.

PLATE IV.

Fig. 23. Echinodictyum gorgonioides n. sp. (R.N. XXI. 1.). Nat. size.
,, 24a. Bubaris radiata n. sp. (R.N. III. 7 a). Upper surface, x 2.
,, 246. ,, ,, „ Lower surface, x 2.

,, 25. Spirastrella vagabunda var. tubnlodigitata Dendy (R.N. IV. 10). Nat. size.
,, 26a. Polymastia gemmipara n. sp. (R.N. IV. 7). Viewed from above, b. buds ; p. pebble ;

sp. c. internal spicular column. X 3.
,, 206. Polymastia gemmipara n. sp. (R.N. IV. 7). Viewed obliquely from below, showing interior.

ch. coagulated masses of choanosomal tissue attached to inner surface of cortex. Other

lettering as before. X 3.
,, 27. Megalopastas retiaria n. sp. (R.N. IV. 13). X 2.

(All the figures on Plates II-IV are reproduced from drawings by Mr. T. P. Ceilings, with the
exception of Fig. 23. which is based on a photograph.)

MARINE ZOOLOGY OF OKHAMANDAL. PART II.

NON-CALCAREOUS SPONGES, PLATE I.

Arthur Dendy del

NON - CALCAREOUS SPONGES.

Cambridge University Press

MARINE ZOOLOGY OF OKHAMANDAL, PART II.

NON-CALCAREOUS SPONGES, PLATE II.

14b.

T. P. Collints.

NON-CALCAREOUS SPONGES.

MARINE ZOOLOGY OF OKHAMANDAL, PART II.

. «

i V . . -w -^

»*'•'• \w>\* * ' .<*• . .

A*' i.' "-'f' ' • ' . ft '

" ****^^t*J* ''-*^a>*V < *<*Oi

>-: **f«t .y •? i?
'**•'£*& ^'/J

'-.'-. i- vt, ^.^ f. - jP>

NON-CALCAREOUS SPONGES, PLATE

T. P. Ceilings.

NON-CALCAREOUS SPONGES.

MARINE ZOOLOGY OF OKHAMANDAL, PART

NON-CALCAREOUS SPONGES, PLATE IV.

23

26b.

T. P. Ceilings.

NON-CALCAREOUS SPONGES.

REPORT

ON THE

HYDROIDA

COLLECTED BY

MR. JAMES HORNELL

AT

OKHAMANDAL IN KATTIAWAR IN 1905-6.

BY

MISS LAURA ROSCOE THORNELY

[PKEFATORY NOTE. — Of the 17 species described below, it is remarkable that
whereas six only of these are found among the 43 species in the Ceylon collec-
tion described in Herdman's "Ceylon Pearl Oyster Fisheries Report," Pt. II., p. 108,
as many as 10 are among the species recorded by Bale from Australian waters.
When we add to these the two " Siboga " forms (Plumularia concava and Mono-
staechas fischeri var. simplex), the preponderance of Australian and Malayan species
is extremely significant, constituting indeed 12 out of the 17 species represented.

Four of the six species common to Kattiawar and Ceylon are also common
to Australia, and it is reasonable to expect that when the Gulf of Mannar hydroid
fauna is completely investigated the remaining six species common to Kattiawar
and Australia will also be found there. The present list as it stands comprises
the following 10 species recorded for the first time from Indian and Ceylon
coasts, viz. :—

Hebella scandens (Bale).
Synthecium maldivense Borradaile.
Pasythea quadridentata Ell. & Sol.
Thuiaria fenestrata Bale.
Monostaechas fischeri var. simplex

Billard.

Plumularia badia Kirch.

,, caliculata Bale.

,, concava Billard.

Lytocarpus philippinus Kirch.
Halicornaria gracilicaulis (Jaderholm),

147

148 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

Special attention could not be given to the collection of Hydroids during the
short period of my stay in Okhamandal, and many small and inconspicuous
species undoubtedly were missed. To obtain such forms I know no better way
than to suspend, from channel buoys or from a ship's side, wire net cages con-
taining oyster shells at a depth of 'a few feet below the surface. — J. HORNELL.]

HYDROIDA— SYSTEMATIC LIST.

1. Hebella scandens (Bale).

Bale. — •" On Some New and Rare Hydroida in the Australian Museum
Collection," Proc. Linn. Soc. N.S.W., Ser. 2, Vol. III., 1888, p. 758 ; Proc. Roy.
Soc. Viet., 26 (N.S.), Ft. L, 1913, p. 117.

Locality : — Dwarka, 15-17 fathoms.

2. Clytia geniculata Thomely.

Thornely. — " Report on the Hydroida," in " Ceylon Pearl Oyster Fisheries,"
Pt. II., Suppl. Rept. No. VIII., 1904, p. 112.
A very small part of a colony.
Locality : — Off Poshetra.

3. Synthecium orthogonia (Busk).

Bale.— Op. cit., Proc. Linn. Soc. N.S.W., Ser. 2, Vol. III., 1888, p. 767.
The tip of a colony most probably belongs to this species. The hydrothecee
on the lower portion of the stem are semi-alternate.
Locality : — Dwarka.

4. Synthecium maldivense Borradaile.

Borradaile. — " Fauna and Geography of the Maldive and Laccadive Archi-
pelagoes," Vol. II., Pt. IV., p. 841.

The fragments belonging to the top portions of colonies of this species do
not exceed 1| inch in length. The branches do not show so acute an angle with
the stem as in Borradaile's figure, Plate LXIX. Fig. 5a. The description of the
hydrothecse corresponds with this specimen ; the membranous hood I at first
overlooked, thinking the orifice had two lateral teeth. There are several gonothecee
present on stems and branches, mostly protruding from hydrothecse, but some are
situated on the stem just below these. The grouping of hydrothecae is remarkable,
the upper four being united as in Pasythea. There are usually a semi-alternate
pair and one in the axil of a branch to each internode. There are some nodulous
internodes as described for Thuiaria fabricii (Levinsen).1 There is a general

1 Nutting, "American Hydroids," 1904, p. 71.

THORNELY— HYDROIDA 149

resemblance between this species and Thuiaria tubulifera, Marktanner-Turneretscher,
as described by Clarke.1

Locality : — Kiu, low water.

5. Sertularia rugosissima Thornely.

Thornely.— " Hydroida " in "Ceylon Pearl Oyster Fisheries," Pt. II., Suppl.
Kept. No. VIII., 1904, p. 118.

Growing in quantities over a coralline. This species appears to be identical
with S. hupferi discovered by Broch. 2

Locality : — Dwarka.

6. Sertularia tenuis Bale.

Bale. — Cat. Aust. Hydroid Zoophytes, 1884, p. 82.
Localities : — Kiu ; off Beyt.

7. Pasythea quadridentata Ellis & Sol.

Bale. — Cat. Aust. Hydroid Zoophytes, 1884, p. 112.

There are usually four pairs of zooecia to an internode, but sometimes only
one or three pairs.

Locality : — Dwarka, 15-17 fathoms.

8. Idia pristis Lamouroux.

Bale. — Cat. Aust. Hydroid Zoophytes, 1884, p. 113.

Localities : — West of Beyt Island, 3-4 fathoms ; off Dwarka, 15-17 fathoms ;
off Poshetra Head, 7 fathoms ; Adatra, 7 fathoms.

A widely-distributed species recorded from Torres Strait, Brazil, and the Persian
Gulf.

9. Thuiaria fenestrata Bale.

Bale. — Cat. Aust. Hydroid Zoophytes, 1884, p. 116.

Good colonies growing on barnacles in seven fathoms of water.

Localities : — North of Poshetra, 3|-4 fathoms ; off Poshetra Head, 7 fathoms.

10. Monostaechas fischeri Nutting (1905), variety simplex Billard.

Billard.— " Siboga Expedition," 1913, p. 16.
Localities : — Rupan Bandar ; Kutchegad.

1 1 . Plumularia badia Kirchenpauer.

Kirchenpauer. — Abh. ver. Hamb. VI. ; Bale. — Cat. Aust. Hydroid Zoophytes,
1884, p. 128.

Locality : — Off Beyt.

1 Mem. Mus. Comp, Zool. Harvard, Vol. XXXV. No. 1 ; VIII. "The Hydroids," 1907, p. 14.

2 " Kenntnis der Meeresfauna West Africas," 1914, p. 35.

150 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

12. Plumularia buskii Bale.

Bale. — Cat. Aust. Hydroid Zoophytes, 1884, p. 125.
Locality : — Kiu.

13. Plumularia caliculata Bale.

Bale. — " On some New and Rare Hydroida in the Australian Museum Collection,"
Proc. Linn. Soc. N.S.W., Ser. 2, Vol. III., 1888, p. 780.
Locality : — Chindi Reef, 6-10 fathoms.

14. Plumularia concava Billard.
Billard.— " Siboga Expedition," 1913, p. 19.

The specimen is only half an inch in height and has no gonothecse but is otherwise
in correspondence with this species.

Locality : — Mangunda Reef, off Poshetra.

15. Lytocarpus philippinus Kirch.

Bale.- — " On some New and Rare Hydroida in the Australian Museum Collection,"
Proc. Linn. Soc. N.S.W., Ser. 2, Vol. III., 1888, p. 786.

A broken colony showing however the fasciculated stem, one sided growth of the
branches and hydrothecse as described by Bale.

Locality : — Adatra.

16. Lytocarpus plumosus Thornely.

Thornely.— " Hydroida " in " Ceylon Pearl Oyster Fisheries," Pt. II., Suppl. Rept.
No. VIII., p. 124.

Locality : — Channel west of Beyt Island, 3-4 fathoms.

17. Halicornaria gracilicaulis (Jaderholm).

Jaderholm. — " Aussereuropaische Hydroiden in Schwedischen Reichsmuseum,"
Ark. Zool, Bd. I. (1903), p. 299, Taf. 14, XIV., Figs. 3-4.

Billard.— "Hydroides de Madagascar et du sud-est del'Afrique," Archives de Zoologie
Expe'rimentale et Generate, IV., Tome VII., p. 364, 1907.

There are some broken pieces, three of which belong to one colony and put
together reach to a height of 6 inches and correspond with Billard's description,
but there are no gonangia to help in the identification. The hydrothecee are very
delicate, having a plain rim with a point on either side. The mesial sarcotheca
is always short, not protruding further than the rim.

Localities : — West of Beyt Island, 3-4 fathoms ; 3 miles W.N.W. of Samiani
lighthouse ; Rupan Bandar ; Kutchegad : Kiu, low water.

NOTES ON

SOME JELLY-FISHES FEOM
OKHAMANDAL IN KATTIAAVAR

COLLECTED BY

MR. JAMES HORNELL IN 1904-5,

BY

EDWARD T. BROWNE, M.A., F.L.S.

SOME years ago Mr. James Hornell very kindly sent to me four jars of unsorted
plankton and a jar containing specimens of Cassiopea. The plankton was. collected
off the coast of Okhamandal, at the entrance to the Gulf of Kutch, on the north-west
coast of India, during December, 1905, and January, 1906.

The plankton was thoroughly searched for jelly-fishes, and as there were no records
of their occurrence along that part of the Indian coast for several hundred mites any
results would be of interest on the geographical distribution of species.

It soon became evident that the collection was more or less a failure ; at least, from
my point of view. There was a decided scarcity of jelly-fishes, and an unreasonable
amount of very fine mud in the jars, so fine that it took a whole day to settle. The
mud not only clung tight to, but clogged up and spoilt everything in the plankton.
The scarcity of jelly-fishes may be due to the collecting having been done at the wrong
season of the year, or at stations too close inshore.1

1 Possibly this paucity was due to the fact that the time of collection was coincident with
the period when the temperature of the surface water off the Kattiawar coast is at its lowest,
as this factor would be likely to influence reproduction adversely, especially in the case of the
Ilydromedusse. In the tropics I have found that numerous groups of animals have two well-
marked reproductive maxima in the year, and in the majority of cases these are roughly coincident
with the equinoxes. Such seasonal periodicity is the most probable reason for the scarcity of
Hydromedusse at the time I visited Kattiawar. In contrast with this, the abundance of the
Hydroid fauna characteristic of the littoral waters is notable. Judging from my experience
elsewhere in India, I am inclined to think that March and April together with September and
October are the months when plankton is most likely to be specially abundant both in species
and in bulk in the Gulf of Kutch. Faunistic investigations should give the richest results at
these seasons. — J. H.

151

152 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

The Anthomedusse were not represented at all, and the Leptomedusse had one
species : — an Irene, but the specimens were too dilapidated for the determination of
its species. The Trachomedusee were represented by two species : — Amphogona apstelni
and Liriope sp., which latter was fairly common, but the specimens were spoilt by the
mud. The Narcomedusse consisted solely of the well-known and widely-distributed
Solmundella bitentaculata. The Siphonophora belonged to two species, namely Diphy-
opsis chamissonis and Bassia bassensis (Quoy et Gaimard). The identification of Bassia
bassensis rests on the finding of four bracts belonging to its Eudoxid.

TRACHOMEDUSyE.

Amphogona apsteini (Vanhoffen, 1902.)

Amphogona apsteini, Browne, 1904. " The Fauna and Geography of the

Maldive and Laccadive Archipelagoes," Vol. 2, p. 740, PI. 54, fig. 5 ;

PL 56, fig. 1 ; PL 57, figs. 10-15.
Amphogona apsteini, Bigelow, 1909. Mem. Mus. Comp. Zool. Harvard

College, p. 126, PL 2, figs. 1-2 ; PL 34, figs. 12-15 ; PL 45, fig. 10.
Amphogona apsteini, Mayer, 1910, " Medusae of the World," p. .405, text

fig. 257.

A single specimen was taken in December, 1905, off the coast between Rupan Bundar
and Kutchegad.

The umbrella is conical, about 3 mm. in height and 4 mm. in width. The stomach
is on a short peduncle, and its mouth has four lips. There are eight radial canals, and
upon each one there are traces or remains of gonads which are situated about three-
quarters the distance down the canals. The tentacles are all broken off, but their stumps
indicate that there were twelve to fifteen in each octant. No sense-organs could be
found.

The umbrella of this specimen is rather more highly arched, being higher in pro-
portion to its width, than usually recorded, and its tentacles are more numerous, but
I have no reasonable grounds for assuming it to be a new species.

Amphogona apsteini has been previously recorded from the Maldives, from Sumatra,
and from Acapulco Harbour in Mexico.

NARCOMEDUSyE.
Solmundella bitentaculata (Quoy et Gaimard, 1833.)

Solmundella bitentaculata, Browne, 1904. " The Fauna and Geography of
the Maldive and Laccadive Archipelagoes," Vol. 2, p. 741, PL 56, fig. 3.

BROWNE— SOME JELLY-FISHES FROM OKHAMANDAL IN KATTIAWAR 153

Solmundella bitentaculata, Browne, 1905, Medusae, in " Report on the Pearl
Oyster Fisheries of the Gulf of Manaar (Ceylon)," Part IV., p. 153,
PI. 4, figs. 1-6.
Solmundella Utentaculata, Mayer, 1910. " Medusae of the World," p. 455.

There are twenty-two specimens from off the coast of Okhamandal. They were
collected during December, 1905, and January, 1906. They are all in rather bad con-
dition and very dirty. In general appearance they resemble the specimens described
by me from Ceylon. The umbrella is highly arched, with rather a keel-like summit.
The largest specimen measured 7 mm. in width and 6 mm. in height, and its tentacles
about 30 mm. in length. There are no traces of peronial grooves in the perradii without
tentacles. Owing to the condition of the specimens it was impossible to recognise
sense-organs.

SIPHONOPHORA.

Diphyopsis chamissonis (Huxley).

Diphyes chamissonis, Huxley, 1859. " The Oceanic Hydrozoa," p. 36,

PI. 1, fig. 3.
Diphyopsis weberi, Lens and van Riemsdijk, 1908. " Siphonophora of

the Siboga Expedition," p. 53, PI. 8, figs. 67-68.
Diphyes chamissonis, Browne, 1904. " Fauna and Geography of the

Maldive and Laccadive Archipelagoes," Vol. 2, p. 742, PI. 54, fig. 6.
Diphyes chamissonis, Browne, 1905. " Pearl Oyster Fisheries of the Gulf

of Manaar," p. 155.
Diphyopsis chamissonis, Bigelow, 1911, Mem. Mus. Comp. Zool. Harvard

College, Vol. 38, p. 347.

This species was common along the coast of Okhamandal during December, 1905.
The plankton was sent to me without any sorting out of the specimens, and I failed to
find the posterior nectophore of any Diphyid in it, so that the posterior nectophore of
this species still remains unknown.

Among the plankton were a number of Eudoxids which looked very much like
the eudoxid of Diphyopsis dispar. The complete absence of any trace of the poly-
gastric generation of Diphyopsis dispar in the plankton led me to suspect that these
eudoxids had a connection with Diphyopsis chamissonis, whose eudoxid was unknown.
On comparing the specimens with the eudoxids of Diphyopsis dispar from the Indian
Ocean I was able to detect minute differences, sufficient, however, to isolate them.
The description of the eudoxid of Diphyopsis chamissonis will be given in my forth-
coming Report on the Siphonophora of the " Sealark" collection.

Diphyopsis chamissonis is widely distributed in the Indian and Pacific Oceans.

154 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

SCYPHOMEDUS^:.

Cassiopea andromeda var. maldivensis, Browne.

Cassiopea andromeda var. maldivensis, Browne, 1905. " The Fauna and
Geography of the Maldive and Laccadive Archipelagoes," Vol. 2, p. 962.

Eleven specimens were taken at low water on the 9th of January, 1906, at Aramra,
Beyt Bay, in the entrance to the Gulf of Kutch. These specimens are from 20 mm. to
60 mm. in diameter, and in very good condition, even after having been in the original
formalin solution for nine years. They were sent from India in a glass-stoppered jar,
one litre capacity, with the stopper sealed over with hard paraffin wax, which kept the
jar perfectly air-tight and prevented any loss of formalin by evaporation. I may
here add that I always use a mixture of vaseline and beeswax for covering the rim of
the stopper and it makes even common glass-stoppers proof against evaporation.

The specimens have lost all traces of colour and are now whitish. They were
originally of a dark greenish colour owing to the presence of " Green Cells " or Zoo-
xanthellse in the jelly. The colour of the cells is due to chlorophyll, and it very slowly
disappears, but at a faster rate if the specimens are kept in strong daylight.

The normal number of sense-organs for this species is sixteen, but a variation in
number is very frequent. Out of ten of these specimens only three have the normal
number ; the others all show an increase, two have 24 sense-organs, three have 19 sense-
organs, and two have 21 and 22 sense-organs respectively.

The number of velar lobes between the ocular lobes is also very variable, as the
sense-organs are not usually at equal distances apart. In these specimens three velar
lobes appear to be the normal number between the ocular lobes, but any number up to
five occurs, and occasionally two sense-organs are so close together that their ocular
lobes are adjacent to one another, without an intervening velar lobe.

The outer margins of the ocular and velar lobes form a continuous and even margin,
without any indentations, so that the lobation of the margin is practically absent
between the deep ocular clefts in which the sense-organs lie.

The oral arms extend to the margin of the umbrella and are normally arranged in
four pairs, but the number is not always constant. In this series two specimens have
nine arms, and one has seven arms.

The appendages on the oral arms are by no means constant in shape and size ;
one specimen has long, tapering appendages predominating, another shows broad, band-
shaped appendages. In two specimens there is a band-shaped appendage in the centre
where the arms meet, surrounded by numerous small appendages, varying in shape and
size, forming a kind of rosette, but generally there are only a few large appendages near
the central one.

BROWNE— SOME JELLY-FISHES FROM OKHAMANDAL IN KATTIAWAR 155

The specimens from Beyt Bay agree with the variety from the Maldives in having
no trace of colour-markings upon the umbrella, in having a circular band upon the
ex-umbrella, used as a pad for holding the medusae fast by suction when lying on its
back, and in the length and shape of the oral arms. There are no conspicuous, extra
large, band-shaped appendages on the oral arms as found in the typical specimens from
the Maldives. Some of the specimens have band-shaped appendages, but they are
not more than 12 mm. in length and 3 mm. in width.

Cassiopea has a very wide geographical range throughout the tropics, and lives in
shallow bays and lagoons. It is rather tied to definite localities by having to pass
through a fixed scyphistoma stage in the course of its life-history. It is also a variable
medusa, and the finding of a Cassiopea in a new locality usually means the finding of
a new race, called either a variety or species, or a doubtful association with a variety
already named. The only method for satisfactory work on the species and varieties
of this genus is the examination of a very large number of specimens from each locality.

REPORT

ON THE

POLYZOA

COLLECTED BY

MR. JAMES HORNELL

AT

OKHAMANDAL IN KATTIAWAR IN 1905-6.

BY

MISS LAURA ROSCOE THORNELY

[With Six Text-Figures]

[PREFATORY NOTE.— The following report deals with a collection made in the
main by means of dredging off the western coast of Okhamandal between Dwarka
and Samiani Point, and in the faunistically rich channel which constitutes Beyt
Harbour, supplemented by shore-collecting on the reefs and islets at the entrance
to the Gulf of Kutch. Forty-two species, and varieties have been identified by
Miss Thornely, including one new species, Beania regularis, together with a new
variety of Bugula neritina, which has been named var. • fastigiata.

Considering that this is entirely a littoral and shallow water collection — the
greatest depth dredged from being 17 fathoms — and that it was made by precisely
the same methods as were employed during Prof. Herdman's pearl fishery investi-
gations in Ceylon, it is remarkable that out of the 42 species now recorded, 17
only are common to the Ceylon and Kattiawar lists. The great majority of the
remaining 25 species are new records for Indian waters.

The following table shows the species common to two or more of the four

158

OKHAMANDAI, MARINE ZOOLOGY REPORT— PART II

principal collections of Polyzoa described from India and the Indian Ocean, viz.,
from (a) Ceylon, reported upon in Vol. IV. of Herdman's " Ceylon Pearl Oyster
Fisheries " ; (6) various Indian localities, in the possession of the Indian Museum,
described in Vol. I. Pt. III. of the Records of the Indian Museum ; (c) the Indian
Ocean, obtained during the " Sealark " Expedition of the Percy Sladen Trust
Expedition in 1905 (Trans. Linn. Soc., Vol. XV. Pt. I.) ; and, lastly, (d) the
present one from the north-west extremity of Kattiawar. It is fortunate for the
sake of uniformity of treatment that these collections have all been described
by the same reporter, Miss Laura Roscoe Thornely, whom I take this opportunity
of thanking most sincerely for having acceded to my request to identify the
present collection.

DISTRIBUTION OF INDIAN POLYZOA

Species.

Kattiawar.

Ceylon.

, ,, . , „ Indian
Sealark ,. ,
„ ... Museum
Expedition. ~ ., . .
i Collection.

Aetna anyuina, ........

X

x

Serupocellaria cervicoriiia ..... x

x

—

X

Canda reti/brmis ....... X

—

X

X

Nellia oculata ........ X

X

—

X

Jiugvla nerilina var. rubra ...

X

—

X

—

Buyula neritina .......

—

X

X

X

Synnotum aviculare .......

X

—

—

X

ThalainoporeUa rozieri ......

X

X

X

—

Steganopwella magnilabrit . . . . .

X

—

X

—

Jletepora monilifera . . ...

X

• —

—

X

Microporella ciiiata .......

X

X

X

X

Lf.pralia yiyas . . . . . . . ' .

X

X

—

—

Mticronella thenardii . . ...

X

X

. —

—

,, coccinea .....

X

X

X

— •

tfmittina triapinosa var. bimucronata ...

X

—

—

—

,, Irispinosa .......

—

X

X

X

Cellepora meyasoma . . . . .

X

X

X

X

„ albirostris . . ...

X

X

X

—

„ tridmticitlata ....

X

— .

X

—

Idtnonea radians .......

X

—

X

—

„ milneana .......

X

X

X

—

Pherusa tubulosa . . . • .

X

X

—

—

Amathia dintans ....... X

X

—

X

Duskia set iy era ....... X

X

—

—

Alcyonidium my till .......

X

X

—

—

CylindToecium yiganteum . . . ...

x

X1

X

24

17 + 1 H

10

Recorded elsewhere from Ceylon but not represented in the Herdman collection.

THORNELY— POLYZOA 159

It is notable that only seven species in the Kattiawar list are recorded in
Waters' list of Red Sea species (" Repts. Soudanese Red Sea," Journ. Linn. Soc.
Vol. XXXI., Nos. 205 and 207, 1909-10), viz. : — Scrupocellaria cervicornis, Nellia
oculata, Bugula neritina, Synnotum aviculare, Tkalamoporella rozieri, Microporella
ciliata, and Cylindroecium giganteum, and that all these are cosmopolitan species ;
the six first extend from the Atlantic Ocean (or Mediterranean) eastward to
Australia, while the seventh, Cylindroecium giganteum, is recorded from British
Columbia, Britain, and the Mediterranean and Red Seas.

The chief conclusion to be drawn from this comparison of collections appears
to be that the polyzoan fauna of Indian Seas is still most imperfectly known and
requires much more study before any useful deductions upon geographical dis-
tribution can be drawn. — J. HORNELL.]

SYSTEMATIC LIST— SUB-ORDER : CHEILOSTOMATA

1. Aetea anguina (Linn.).

Large colonies creeping over sea-weeds.
Locality : — Off Dwarka.

2. Brettia tropica Waters.

Waters. — " Bryozoa from Zanzibar," Proc. Roy. Soc. London, Sept. 1913, p. 465-
Growing in quantities on stems of the Hydroid Idia prislis.
Locality : — Off Poshetra Head, 7 fathoms.

3. Catenicella buskii Wyv. Thomson.

MacGillivray. — In McCoy's " Prodromus of the Zool. of Victoria," Decade
III., 1879, p. 24.

A small fragment without ovicells.
Locality : — North coast of Beyt Island.

4. Scrupocellaria pilosa Busk.

Busk.— " Polyzoa," Challenger Reports, Zool, Vol. X. Pt. XXX., 1884, p. 24.
A small fragment corresponding with Busk's description of the species but
without a fornix.

Locality : — S.W. of Beyt Island (dredged).

5. Scrupocellaria cervicornis Busk.

Busk.— Brit. Mus. Cat. Mar. Polyzoa, Pt. I., 1852, p. 24.

The outer spine on the present specimens is forked as described by Waters
(" Repts. Soudanese Red Sea," Journ. Linn. Soc. Zool., Vol. XXXI., March 1909,
p. 166).

Localities : — Rupan Bandar ; Kutchegad, 4-7 fathoms ; Dwarka.

L*

160 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

6. Scrupocellaria macandrei Busk.

Busk.— Brit. Mm. Cat. Mar. Polyzoa, Pt. I., 1852, p. 24.
Localities : — Dwarka ; Kiu, low water.

7. Canda retiformis Pourtales.

Pourtales. — Bull. Mm. Comp. Zool. Harvard Coll., Vol. I., No. 6, 1867, p. 110.
Waters. — Proc. Roy. Soc. Lond., Sept. 1913, p. 479.
Locality : — Dwarka.

FIGS. 1 — 5. Bugula neritina var. fastigiata, n. var.

1. Tip of a brand) ; 2. back view of the lower portion of a branch ;
3. front view of same ; 4. back view of portion of a branch showing
origins of fascicles ; 5. lower portion of a branch.

8. Nellia oculata Busk.

Busk.— Brit. Mm. Cat. Mar. Polyzoa, Pt. I., p. 18.

Localities : — Kiu, low water ; off Poshetra Head, 7 fathoms ; off Beyt.

9. Bicellaria glabra (Hincks).

Hincks.— Ann. Mag. Nat. Hist., Ser. 5, Vol. XL. 1883, p. 195.
Locality : — Off Beyt Island.

THORNELY— POLYZOA 161

10. Bugula neritina var. rubra Thomely.

Thornely. — Trans. Linn. Soc. Lond., March 1912, p. 141.
Localities : — Off Beyt ; Dwarka, 15-17 fathoms ; Kiu, low water.

11. Bugula neritina var. fastigiata nov. (Figs. 1-5).

Zoarium with a stout, dark brown, branched, fasciculated stem rising from a
fibrous root to a height of about three inches, the branches also fasciculated and
dark coloured below, becoming gradually less stout and rigid until near the top
where they curl gracefully over in delicate transparent tufts like those of Bicellaria
ciliata. Zooecia biserial, stout and showing through the fascicles of the branch
below, delicate and smaller in the curled upper parts ; a small avicularium below
each oral aperture. Ooecia opening laterally in the usual B. neritina manner.

FIG. 6. Beania regularis, n. sp.

There are several colonies of this species which must be very beautiful when
perfect, the contrast of the thick dark stems and branches and their delicate
white tips being remarkable.

Localities : — Dwarka, 15-17 fathoms ; Chindi Reef, 6-10 fathoms ; Rupan
Bandar ; Kutchegad.

12. Beania re'gularis, n. sp. (Fig. 6).

Zoarium loosely attached ; zooecia erect, distant, oval, open in front, widest
at the base, five points on the margin above the orifice.

No lateral spines, no avicularia, connecting tubes proceeding from the lower
portion of the zooecium.
Locality : — Dwarka.

13. Synnotum aviculare (Pieper).

Hincks.— Ann. Mag. Nat. Hist., Ser. 5, Vol. XVII., p. 257.
Localities : — Dwarka, deep water ; off Beyt Island.

L* 2

162 OKHAMANDAL MARINE ZOOLOGY REPORT— PART II

14. Membranipora serrata MacGillivray.

MacGillivray. — " Descr. Aust. Polyzoa," Trans. Roy. Soc. Viet., Vol. IX.,
Nov. 1868, p. 6; McCoy.— Prod. Zool. Victoria, Decade XIII., p. 105.
A small broken colony looking like a Flustra.
Locality : — Off Dwarka, 15-17 fathoms.

15. Membranipora curvirostris Hincks.

Hincks. — Ann. Mag. Nat. Hist., Ser. 3, Vol. IX., p. 29; Brit. Mar. Polyzoa, 1880,
p. 153.

Growing on a Retepora.
Locality : — Dwarka.

16. Thalamoporella rozieri (Andouin).

Hincks. — Ann. Mag. Nat. Hist., Ser. 5, Vol. XIX., p. 164.

Some large bilaminate foliaceous pieces and also some unilaminate pieces
clinging to an Amathia.

Localities : — S.W. of Beyt Island ; channel west side of Beyt Island, 3-4 fathoms ;
Kiu, low water ; south of Chindi Keef, 6-10 fathoms ; off Dwarka, 15-17 fathoms.

17. Steganoporella magnilabris (Busk).

Busk.—" Polyzoa," Challenger Repts. Zool, Vol. X. Pt. XXX., p. 75.
Localities : — Channel south-west of Beyt Island ; 3 miles W.N.W. of Samiani light-
house, 17 fathoms.

18. Cellaria gracilis Busk.

Busk.— Brit. Mus. Cat. Mar. Polyzoa, Pt. I., 1852, p. 17.

There are some good colonies growing on Pherusa.

Localities : — Dwarka, 15-17 fathoms ; off Poshetra, 7 fathoms.

19. Retepora monilifera MacGillivray.

MacGillivray.— Trans. Phil. Inst. Viet., Vol. IV. Pt. 2, p. 168.
McCoy.— Prod. Zool. Victoria, Decade X., Vol. I. p. 20.

Some good colonies, several having a pair of long curved jointed spines one
on either side of the orifice.

Localities : — Poshetra, 3f-4 fathoms ; south of Chindi Reef, 6-10 fathoms ; Rupan
Bandar and Kutchegad ; off Dwarka.

20. Microporella ciliata (Pallas).

Hincks. — Brit. Mar. Polyzoa, 1880, p. 206.

Growing on Pherusa tubulosa.

Locality : — Off Poshetra Head, 7 fathoms.

THORNELY— POLYZOA 163

21. Lepralia gigas Hincks.

Hincks. — Ann. Mag. Nat. Hist., Ser. 5, Vol. XV., p. 255.
Locality : — Half-mile north of Poshetra, 3f-4 fathoms.

22. Lepralia japonica Busk.

Busk.—" Polyzoa," Challenger Repts. ZooL, Vol. X. Pt. XXX. p. 143.

Large broken foliaceous pieces and also old colonies showing no division
between zooecia.

Localities : — Off Dwarka, deep water ; Hanuman Dandi Reef, Beyt ; north of
Poshetra, 3f-4 fathoms.

23. Mucronella thenardii (Aud.).
Savigny.— Zool. Egypte, Plate X. Fig. 3.
One dried specimen only.

Locality : — Off Dwarka.

24. Mucronella coccinea Abild.
Hincks. — Brit. Mar. Polyzoa, 1880, p. 371.
Locality : — Off Dwarka.

25. Porella compressa Sowerby.

Sowerby.— Brit. Miscel I., 1806, p. 83.
Hincks.— Brit. Mar. Polyzoa, 1880, p. 330.
Locality : — Poshetra, 3f-4 fathoms.

26. Porella concinna var. gracilis Hincks.
Hincks.— Brit. Mar. Polyzoa, 1880, p. 324.
Encrusting Lepralia japonica.

Localities : — Off Dwarka and Kutchegad, deep water ; off Poshetra, 7 fathoms.

27. Smittina trispinosa var. bimucronata Hincks.
Hincks.— Ann. Mag. Nat. Hist., Ser. 5, Vol. XIII., p. 362.

The Dwarka specimen differs from the others in its ooecium having an arched
ridge with minute pores in the area it encloses.

Localities : — Rupan Bandar ; Kutchegad ; Poshetra, 7 fathoms ; Dwarka, deep
water.

28. Haswellia australiensis (Haswell).

Haswell.— Proc. Linn. Soc. N.S.W., Vol. V. Pt. I., 1880, p. 33.
Busk.— Challenger Repts. Zool, Vol. X. Pt. XXX., p. 172.
Locality : — Off Dwarka, 15-17 fathoms.

164 OKHAMANDAL . MARINE ZOOLOGY REPORT— PART II

29. Cellepora tridenticulata Busk.

Busk.— Challenger Repts. Zool., Vol. X. Pt. XXX., p. 198.
Localities : — Rupan Bandar ; Kutchegad ; Poshetra, 3|-4 fathoms.

30. Cellepora megasoma MacGillivray.
MacGillivray. — Descr. New Pol., Vol. VIII., p. 10.
McCoy.— Prod. Zool. Victoria, Decade IV., Vol. I., p. 33.
Locality : — Off Dwarka.

31. Cellepora bispinata Busk.

Busk.— Brit. Mm. Cat. Mar. Polyzoa, Pt. II., 1854, p. 87.

There are the bases only "of two spines on the one small colony present.
The rostrum sometimes rises to a thin point and has no avicularium upon it.
Vicarious avicularia differ in size from being very small to larger than the zooecia.

Localities : — Rupan Bandar ; Kutchegad.

32. Cellepora albirostris (Smitt).

Smitt.— " Floridan Bryozoa," Pt. II., p. 70, in Vetensk. Akad. Handl., Vo1. XI.,
1872.

Busk.— Challenger Repts. Zool., Vol. X. Pt. XXX., p. 193.

Localities : — South-west coast of Beyt Island ; Dwarka, deep water, growing on
Lepralia japonica ; north of Poshetra, 3|-4 fathoms.

SUB-ORDER: CYCLOSTOMATA.

33. Idmonea radians (Lamk.).

Busk.— Brit. Mus. Cat. Mar. Polyzoa, Pt. III., p. 11, 1875.
Locality : — Off Poshetra Head, 7 fathoms.

34. Idmonea australis MacGillivray.
McCoy.— Prod. Zool. Victoria, Decade VII., p. 30.
Locality : — Off Dwarka.

35. Idmonea milneana D'Orb.

D'Orb. — Voy. Amer. Merid., " Polypiers," p. 20.
Busk.— Brit. Mus. Cat. Mar. Polyzoa, Pt. III., p. 12, 1875.
Locality : — Dwarka, 15-17 fathoms.

36. Pustulopora deflexa Smitt.

Smitt.— " Floridan Bryozoa," Pt. I., p. 11.

Busk.— Challenger Repts. Zool, Vol. XVII. Pt. L., p. 20.

Locality :• — Off Dwarka.

THORNELY— POLYZOA 165

SUB-ORDEK : CTENOSTOMATA.

37. Pherusa tubulosa Lamx.

Lamx. — Polypi&rs Corall. Flexibles, p. 119.

Waters. — Ann. Mag. Nat. Hist., Ser. 5, Vol. III., p. 279.

There are some good colonies one inch in height growing amongst a Retepora.

Localities : — Off Poshetra Head, 7 fathoms ; north of Poshetra, 3f-4 fathoms.

38. Amathia distans Busk.

Busk.— Challenger Repts. Zool, Vol. XVII. Pt. L., p. 33.

Fragments of colonies only.

Localities : — South-west of Beyt Island ; Kiu, low water ; Poshetra, 3|-4
fathoms ; Rupan Bandar ; Kutchegad ; off Poshetra, 7 fathoms ; channel west of
Beyt Island, 3-4 fathoms ; Chindi Reef, 6-10 fathoms.

39. Amathia connexa Busk.

Busk.— Challenger Repts. Zool, Vol. XVII. Pt. L., p. 35.
Localities : — Kiu ; Dwarka.

40. Buskia setigera Hincks.

Hincks. — Journ. Linn. Soc., Vol. XXI., p. 127.
Growing on stems of Idia pristis.

Locality : — Off Poshetra Head, 7 fathoms.

•

41. Alcyonidium mytili Dalyell.
Hincks. — Brit. Mar. Polyzoa, 1880, p. 498.
Locality : — South-west of Beyt Island.

42. Cylindroecium giganteum (Busk).

Busk. — Quart. Journ. Micro. Soc., Vol. IV., p. 93.

Hincks.— Brit. Mar. Polyzoa, 1880, p. 535

Localities : — Dwarka, 15-17 fathoms ; off Poshetra Head, 7 fathoms.

(MS. received from author 29th July, 1915).

PKINTKI) I.V GRKAT BUITAIN BY
RlTHAKD Cl.AY AND So.NS, LIMITED,
BRUNSWICK STREET, STAMFORD STKF.ET, S.E ,
AND liUXOAY, SUFFOLK.

V