Marine Biological Laboratory

Received_l£ A'/t 1941

Accession No. 55159

Given By_ 3. Hanson

Rockef ell or Foundation"
Place,_ York City

i

. .

; • '

Jl

'

•
-

:
•i

,

•

I

-

TEXT-BOOK

OF

ru

JOMPABATIVE ANATOMY

of

m
a

0/ /

TEXT-BOOK

OF

COMPARATIVE ANATOMY

BY

DR. ARNOLD LANG

PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF ZURICH
FORMERLY RITTER PROFESSOR OF PHYLOGENY IN THE UNIVERSITY OF JENA

TRANSLATED INTO ENC4LISH BY

HENRY M. BERNARD, M.A. CANTAB.

AND

MATILDA BERNARD

PART II.

ILontion

MACMILLAN AND CO., LTD.

NEW YORK : MACMILLAN & CO.

1896

.TRANSLATORS' PREFACE

THE fact that this second volume of the translation appears four years
after the first is due partly to the delay in the issue of the third and
fourth German parts of which it is composed, and partly to the
increased difficulty in the work of translation. A comparison of the
two volumes will show at a glance that the work has developed
under the hands of the author: the treatment has become more
elaborate. The two " chapters " which practically fill this volume are
in reality more like comprehensive treatises on the groups with
which they deal, and as such could only be adequately translated from
the German by some one with a very special knowledge of both
groups. There are probably few zoologists who have attempted to
make a special study of two such heterogeneous phyla as the Mollusca
and the Echinodermata. In addition, therefore, to frequent references
to the original literature and to constant applications to kind friends,
the whole of the text relating to the two chief groups was submitted
to specialists for revision. The translators beg to tender their
warmest thanks to their friends who kindly undertook this laborious
task. Mr. B. B. Woodward read the text of the chapter dealing
with the Mollusca, revising the terminology, and suggesting slight
alterations, which have been either adopted without comment in the
text or else placed in short footnotes. Mr. W. Percy Sladen and Mr.
F. A. Bather revised the text dealing with the Echinodermata, each
with special reference to the group with which his name is most asso-
ciated. Thanks are also due to Professor Jeffery Bell for his kind
assistance in the solution of difficulties. We have no hesitation in
saying that it is to the generous help of these gentlemen that

vi COMPARATIVE ANATOMY

our translation owes much of the value it may possess for the English
student.

In the use of certain technical terms we have given the English
or the Latin form indifferently, e.g. pinnule or pinnula, auricle or
auricula, with deliberate inconsistency. On the other hand, we have
throughout used the terms madreporite, madreporitic, and Echinoder-
mata, although some authorities are more in favour of madrepore,
madreporic, and Echinoderma. AVe feel it our duty to call the atten-
tion of students to these points.

The following author's preface is a free translation of the
Nachwort which appeared at the end of the fourth German part.
In it the author answers the only serious charge against the work
as a text-book which has been brought to our notice. It finds its
most appropriate place as a preface to the second volume of the

translation.

H. & M. BERNARD.

AUTHOR'S PREFACE TO THE SECOND VOLUME

WITH the publication of the last two chapters, dealing with the
Echinodermata and the Enteropneusta — that is of the fourth German
portion — I bring this text-book to a close for the time being, as a
comparative anatomy of the Invertebrata.

I feel that some excuse is necessary for the tardy appearance of
the separate parts, especially of the third (Mollusca). This was
mainly due to my call to the University of Zurich, where official
duties left only the holidays and vacations for my own work. When
I add that the greater number of the illustrations were drawn by
my own hand, the reader will, I trust, pardon the lapse of time.
Indeed, if he be a trained zoologist, he will be specially sympathetic
and indulgent, and will be able to realise my feelings as I watched
the fresh relays of books piling up before me at the commencement
of each new chapter. Original sources alone have been relied upon
for the subject matter of the work.

In spite of the imperfections and deficiencies of which I am
only too conscious, the book appears to have been found useful,
judging from the favourable reception almost universally given to
it, and from the circumstances that, even during its appearance,
it was translated into foreign languages.

I am fully aware that the matter is unequally worked up. The
divisions treated in the first volume are too briefly dealt with, a
defect which must be remedied in a new edition. Any criticisms
or advice with which my colleagues may favour me will be gladly
accepted in the spirit in which they are intended.

I have been blamed by many for not mentioning the names of

via COMPARATIVE ANATOMY

authors in the text. From the very first this question caused me
much perplexity, and I made repeated attempts to indite single
chapters so as to bring in the historical development of the branch
dealt with, together with the names of the most important authors.
I then found that if this course were pursued the book would
attain twice its present dimensions, that is, if strict impartiality
were to be invariably observed. This latter I Avas resolved on no
account to renounce, and I therefore determined to exclude from
the text the names of all authors without distinction. Any one who
is interested in knowing how a special question stands, can easily
find his bearings by careful comparison of the text with the illustra-
tions (the origin of which is everywhere given), and by consulting
the literature. I have convinced myself of this among my own
students.

I must here express my thanks to my honoured and dear friend,
Mr. Gustav Fischer, for the care and patience he has exercised in

connection with this work.

ARNOLD LANG.

ZURICH, July 1894.

CONTENTS

CHAPTER VII

MOLLUSCA

PAGE

Systematic Review ........ 2

Class I. AMPHINEURA .... 2

II. GASTROPODA (CEPHALOPHORA) ..... 3

III. SCAPHOPODA ....... 13

IV. LAMELLIBRANCHIA (PELECYPODA, BIVALVA, ACEPHALA,

AGLOSSA) ....... 14

V. CEPHALOPODA . . . .21

I. ORGANISATION OF THE PRIMITIVE MOLLUSC ... 26

II. REVIEW OF THE OUTER ORGANISATION CHARACTERISING THE

CHIEF GROUPS OF THE MOLLUSCA .... 28

A. PLACOPHORA OR POLYPLACOPHORA (CHITONHXE) . . 29

B. APLACOPHORA, SOLENOGASTRES . . 29

C. GASTROPODA (CEPHALOPHORA) . . 30

D. SCAPHOPODA ....... 34

E. LAMELLIBRANCHIA . . . 34

F. CEPHALOPODA ....... 36

III. THE INTEGUMENT, THE MANTLE, AND THE VISCERAL DOME . 39

A. PLACOPHORA ...... 39

B. SOLENOGASTRES . .41

C. GASTROPODA . . 42

D. SCAPHOPODA . . 49

E. LAMELLIBRANCHIA . 49

F. CEPHALOPODA . . 53

IV. THE SHELL . 55

A. AMPHINEURA . . .58

B. GASTROPODA 58

COMPARATIVE ANATOMY

PAGE

C. LAMELLIBRANCHIA ....... 61

D. CEPHALOPODA ....... 67

^T. ARRANGEMENT OF THE ORGANS IN THE MANTLE CAVITY, AND

OF THE OUTLETS OF INNER ORGANS IN THAT CAVITY . 71

A. GASTROPODA ... .71

B. SCAPHOPODA ....... 80

C. LAMELLIBRANCHIA ....... 81

D. CEPHALOPODA ... ... 81

VI. THE RESPIRATORY ORGANS . . . 84

THE TRUE GILLS OR CTENIDIA ... .84

A. AMPHINEURA .... .86

B. GASTROPODA .... .88

C. LAMELLIBRANCHIA ... . .91

D. CEPHALOPODA ... .96
ADAPTIVE GILLS . . . 97
LTTNGS ..... 99

VII. THE HYPOBRANCHIAL GLAND . 101

VIII. THE HEAD . 101

A. GASTROPODA . . . . . . .102

B. SCAPHOPODA . ... 104

C. CEPHALOPODA ..... . 105

IX. THE ORAL LOBES OF THE LAMELLIBRANCHIA . . 105

X. THE FOOT AND THE PEDAL GLANDS . 106

A. AMPHINEURA ... . . 106

B. GASTROPODA . . . 107

C. SCAPHOPODA ... . .112

D. LAMELLIBRANCHIA .... . 112

E. CEPHALOPODA .... . 115

XI. SWELLING OF THE FOOT (Turcjescencc] .... 118

XII. MUSCULATURE AND ENDOSKELETON . 119

A. AMPHINEURA ..... .120

B. GASTROPODA . . . . . . .120

C. SCAPHOPODA . . . . . . .123

D. LAMELLIBRANCHIA . . . . . .124

E. CEPHALOPODA ... . 126

XIII. THE NERVOUS SYSTEM . . 128

A. AMPHINEURA ..... 128

B. GASTROPODA . • 132

CONTENTS xi

PAGE

1. THE AREAS OF INNERVATION OF THE VARIOUS GANGLIA . 133

2. ORIGIN OF THE CROSSING OF THE PLEUROVISCERAL CON-

NECTIVE (CHIASTONEURY) . . . . .135

3. SPECIAL REMARKS ON THE NERVOUS SYSTEM OF THE GAS-

TROPODA ....... 137

C. SCAPHOPODA . . . . . .142

D. LAMELLIBRANCHIA ....... 143

E. CEPHALOPODA ....... 145

XIV. AN ATTEMPT TO EXPLAIN THE ASYMMETRY OF THE GASTROPODA 149

XV. THE SENSORY ORGANS .... .162

A. INTEGUMENTAL SENSORY ORGANS .... 162

1. TACTILE ORGANS .... .162

2. OLFACTORY ORGANS . . . . . .162

3. THE "LATERAL ORGANS" OF THE DIOTOCARDIA . . 165

4. GUSTATORY ORGANS .... 166

5. SUBRADULAR SENSORY ORGAN OF CHITON . . 166

6. THE SENSORY ORGANS ON THE SHELL OF CHITON . 166

B. AUDITORY ORGANS ...... 167

C. VISUAL ORGANS ....... 169

1. OPTIC PITS ..... .169

2. OPTIC VESICLES OR VESICULAR EYES . . .170

3. THE EYE OF THE DIBRANCHIATE CEPHALOPODA . . 170

4. THE DORSAL EYES OF ONCIDIUM AND THE EYES AT THE

EDGE OF THE MANTLE IN PECTEN . .173

5. THE EYES ON THE SHELL OF CHITON . . . 175

6. THE COMPOUND EYES OF ARCA AND PECTUNCULUS . 175

7. DEGENERATION OF THE CEPHALIC EYES . . 176

XVI. THE ALIMENTARY CANAL . . . . . .176

A. BUCCAL CAVITY, SNOUT, PROBOSCIS . . . .178

B. THE PHARYNX AND JAWS, THE TONGUE AND SALIVARY

GLANDS ...... 180

FORMATION OF THE RADULA . ... 183

C. THE (ESOPHAGUS ...... 187

D. THE MID-GUT WITH THE STOMACH AND DIGESTIVE GLAND

(LlVER) ........ 190

1. AMPHINEURA ... ... 191

2. GASTROPODA ... . .192

3. SCAPHOPODA ... ... 193

4. LAMELLIBRANCHIA ... 194

5. CEPHALOPODA . . ... 194

E. HIND-GUT (RECTUM) ..... 195

XVII. THE CIRCULATORY SYSTEM . . 198

A. GENERAL 198

xii COMPARATIVE ANATOMY

PAGE

B. SPECIAL ... . 201

1. AMPHINEURA . .... 201

2. GASTROPODA .... . . 201

3. SCAPHOPODA . . ... 206

4. LAMELLIBRANCHIA .... 206

5. CEPHALOPODA . ..... 208

XVIII. THE BODY CAVITY . .211

XIX. THE NEPHRIDIA . .... 215

A. AMPHINEURA . .... 216

B. GASTROPODA . . . 217

C. SCAPHOPODA .... 221

D. LAMELLIBRANCHIA . ..... 221

E. CEPHALOPODA ... ... 222

XX. GENITAL ORGANS .... 225

A. GENERAL ...... . 225

B. SPECIAL . . . .227

XXI. PARASITIC GASTROPODA ...... 244

XXII. ATTACHED GASTROPODA .... . 248

XXIII. ONTOGENY . . 248

A. AMPHINEURA ....... 248

B. GASTROPODA . .... 252

XXIV. PHYLOGENY ... . . 268
Rcviciv of the most Important Literature .... 269
APPENDAGE,— RHODOPE VERANII . . . . .281

CHAPTER VIII

ECHINODERMATA

Systematic Review . . . . . . .285

CLASS I. HOLOTHURIOIDEA . ... 285

II. ECHINOIDEA 288

III. ASTEROIDEA . .... 295

IV. OPHIUROIDEA . . . .299
V. PELMATOZOA . .... 302

1. CRINOIDEA . 302

2. CYSTIDEA . . 313

3. BLASTOIDEA ... ' . 314

I. GENERAL MORPHOLOGY OF THE ECHINODERM BODY . . 315

II. MORPHOLOGY OF THE SKELETAL SYSTEM . 317

CONTENTS xiii

PACE

INTRODUCTION . . . 317

A. THE APICAL SYSTEM (Calyx) . . 319

1. ECHINOIDEA .... . 319

2. ASTEROIDEA ....... 326

3. OPHIUROIDEA . ..... 327

4. PELMATOZOA ....... 328

(a) CRINOIDEA . . .... 328

(b) BLASTOIDEA . ..... 330

(c) CYSTIDEA . . . . . . .332

B. THE ORAL SYSTEM OF PLATES ... .333

C. THE PERISOMATIC SKELETON ..... 337

1. HOLOTHURIOIDEA ...... 337

2. ECHINOIDEA ....... 338

(«) THE NUMBER OF THE VERTICAL Eows OF PLATES . 339

(b) THE PORES OF THE AMBULACRAL SYSTEM . . 340

(c) THE SYMMETRY OF THE ECHINOID SHELL . . 340

(d) THE RELATION OF THE AMBULACRAL AND INTERAMBU-

LACRAL PLATES TO THE PERISTOME . . . 344
(c) MANNER IN WHICH THE SKELETAL PLATES ARE CON-
NECTED ....... 345

(/) SPECIAL MODIFICATIONS OF THE AMBULACRA . . 346

(g) SPECIAL MODIFICATIONS OF THE INTERRADII . . 348

(h) FORM OF THE PERISTOME ..... 349

(i) ORNAMENTATION ... . . 349

(&) MARGINAL INCISIONS OR PERFORATIONS . . . 349

(1) THE PERIGNATHIC APOPHYSIAL GIRDLE . . . 350

3. ASTEROIDEA ....... 351

(a) THE AMBULACRAL SKELETON .... 351

(b) THE INTERAMBULACRAL SKELETON . . .353

(c) THE ACCESSORY SKELETAL SYSTEM . . . 354
(d} COMPARISON OF THE PERISOMATIC SKELETON OF THE

ASTEROIDEA WITH THAT OF THE ECHINOIDEA . . 355

4. OPHIUROIDEA ....... 355

(a) SKELETON OF THE ARMS ..... 355

(b) THE ORAL SKELETON ..... 358

5. CRINOIDEA ....... 362

(a) THE PERISOMATIC SKELETON OF THE CALYX . . 362

a. THE APICAL CAPSULE OR DORSAL CUP . 367

b. THE TEGMEN CALYCIS ..... 369
(6) THE BRACHIAL SKELETON ..... 370

(c) THE STEM (COLUMNA) ..... 373

(d) THE MANNER OF CONNECTION BETWEEN THE SKELETAL

PIECES ....... 376

(e) THE NERVE CANALS OF THE ARMS AND OF THE APICAL

CAPSULE 377

xiv COMPARATIVE ANATOMY

PAGE

(/) THE WATER PORES . . 377

6. BLASTOIDEA ... ... 379

(a) THE AMBULACRAL SKELETON .... 379

(6) THE STEM . 384

7. CYSTIDEA ....... 384

D. THE SPINES AND THEIR DERIVATIVES — THE SPH^RIDIA AND

THE PEDICELLARI.E . . ... 387

E. THE MASTICATORY APPARATUS OF THE ECHINOIDEA. (Aris-

totle's Lantern) ....... 400

F. THE CALCAREOUS RING OF THE HOLOTHURIOIDEA . . 403

G. FURTHER DEPOSITS OF CALCAREOUS MATTER . . . 405
H. CONCLUDING REMARKS ON THE SKELETON . . . 405

III. THE OUTER MORPHOLOGY OF THE HOLOTHURIOIDEA . . 406

IV. THE POSITION AND ARRANGEMENT OF THE MOST IMPORTANT

ORGANS IN THE RADII ...... 409

V. THE INTEGUMENT . ..... 414

VI. THE WATER VASCULAR SYSTEM . . 416

A. THE MADREPORITE AND STONE CANAL . . . .417

B. THE WATER VASCULAR RING ..... 423

C. THE RADIAL CANALS, THE CANALS OF THE TENTACLES AND

TUBE-FEET, ETC. ....... 426

D. THE AMBULACRAL APPENDAGES . ... 431

VII. THE CCELOM . 436

A. THE BODY CAVITY . . . . 437

B. THE BRACHIAL CAVITIES ...... 440

C. THE PERKESOPHAGEAL SINUS ..... 441

D. THE PERIANAL SINUS ...... 444

E. THE AXIAL SINUS . .... 444

F. THE AXIAL ORGAN . . .445

G. THE CHAMBERED SINUS . . 446

VIII. THE PSEUDOH^MAL SYSTEM . . 447

IX. THE EPINEURAL SYSTEM . . 448

X. THE BLOOD VASCULAR OR LACUNAR SYSTEM . . 449

XI. THE NERVOUS SYSTEM . . . . 453

A. THE SUPERFICIAL ORAL SYSTEM . . . 454

B. THE DEEPER ORAL NERVOUS SYSTEM . . . 458

C. THE APICAL OR ABORAL NERVOUS SYSTEM 459

D. THE THIRD NERVOUS SYSTEM OF THE CRINOIDEA . . 461

CONTENTS xv

PAGE

XII. THE SENSORY ORGANS ..... 462

A. THE AMBULACRAL APPENDAGES AS SENSORY ORGANS . 462

B. NERVE ENDINGS IN THE INTEGUMENT . . . 466

C. AUDITORY ORGANS, ORGANS OF ORIENTATION . . . 468

D. EYES ..... .468

XIII. THE BODY MUSCULATURE ... . 470

A. HOLOTHURIOIDEA . . 471

B. ECHINOIDEA . . 471

C. ASTEROIDEA . ... 472

D. OPHIUROIDEA ... . . 474

E. CRINOIDEA .... . . 474

XIV. THE ALIMENTARY CANAL . . 474

A. GENERAL REVIEW . . . . . 474

B. HOLOTHURIOIDEA ... . . 476

C. ECHINOIDEA . 479

D. CRINOIDEA . . . . . 481

E. ASTEROIDEA ....... 483

F. OPHIUROIDEA ....... 485

XV. RESPIRATORY ORGANS ...... 485

A. THE (INNEE) RESPIRATORY TREES OF THE HOLOTHURIOIDEA . 487

B. REVIEW OF THE RESPIRATORY ORGANS OF THE ECHINODER-

MATA ........ 487

XVI. THE CUVIERIAN ORGANS OF THE HOLOTHURIOIDEA . . 488

XVII. EXCRETION . . 489

XVIII. THE SACCULI OF THE CRINOIDEA . . . 489

XIX. GENITAL ORGANS . . 490

A. GENERAL MORPHOLOGY ... . 490

B. HOLOTHURIOIDEA . . . 491

C. ASTEROIDEA . . . 492

D. OPHIUROIDEA . 494

1. THE BURSTS ... 494

2. THE GENITAL APPARATUS . . . . 495

E. ECHINOIDEA . . 498

F. CRINOIDEA .... 500

G. ORIGIN OF THE SEXUAL PRODUCTS . . . 501
H. HERMAPHRODITISM IN ECHINODERMS . . . 501

1. CARE OF THE BROOD AND SEXUAL DIMORPHISM . 502

XX. CAPACITY FOR REGENERATION AND ASEXUAL REPRODUCTION . 504

XXI. ONTOGENY ....... 5.06

A. THE VARIOUS LARVAL FORMS OF THE ECHINODERMATA 506

xvi COMPARATIVE ANATOMY

PAGE

B. ONTOGENY OF THE HOLOTHURIOIDEA . . . 510

C. ONTOGENY OF THE ECHINOIDEA .... 519

D. ONTOGENY OF THE ASTEROIDEA . ... 524

E. ONTOGENY OF THE OPHIUROIDEA ..... 532

F. ONTOGENY OF THE CRINOIDEA . . 533

XXII. PHYLOGENY ... 545

Review of the most Important Literature ... 551

CHAPTER IX

ENTEROPNEUSTA

I. OUTER ORGANISATION . . . 562

II. THE BODY EPITHELIUM ..... 563

III. THE NERVOUS SYSTEM . . 564

IV. THE SENSORY ORGANS . . . 565
V. THE ALIMENTARY CANAL .... . 565

VI. THE CCELOMIC SACS AND THE BODY MUSCULATURE. . .' 571

VII. THE "HEART VESICLE" ... 578

ATIII. THE LIMITING MEMBRANES, THE PROBOSCIDAL SKELETON, AND THE

BRANCHIAL SKELETON ...... 579

IX. THE BLOOD VASCULAR SYSTEM . . . 581

X. THE GONADS ........ 585

XI. ONTOGENY ..... .586

XII. PHYLOGENY . ...... 591

Literature ...... . 595

APPENDAGE TO THE ENTEROPNEUSTA

I. CEPHALODISCUS . ..... 596

II. RHABDOPLEURA . .... 600
Literature . . . . . . . .602

INDEX . 603

CHAPTEE VII
SIXTH RACE OR PHYLUM OF THE ANIMAL KINGDOM

MOLLUSCA.

THE Mollusca are essentially bilaterally symmetrical animals with
unsegmented bodies. The ventral Avail is thick and muscular, and
forms a foot which is used for locomotion, and assumes the most
varied shapes. A fold of the body wall forms a circular mantle, which
hangs down round the body, enclosing a space which is called the
mantle or pallial cavity. This cavity is originally deepest and
most spacious posteriorly, and contains, at the sides of the median
anus, symmetrically grouped, the two gills and the renal and
genital apertures. The dorsal portion of the animal is generally
developed into a visceral dome or sac, and is protected down
to the edge of the mantle by a shell. The mouth lies at the
anterior end of the body and leads into a pharynx, which is usually
provided with jaws and a rasp- like organ called the raclula. The
mesenteron or mid-gut is supplied with a large digestive gland (liver).
The secondary coelom (enclosed lay its own walls) is reduced, but
always persists as a pericardium. The blood vascular system is open,
and generally to a great extent lacunar. The heart is dorsal and
arterial, and was primitively provided with two symmetrical auricles.
The nephridia were originally paired, and in open communication
with the pericardium. The central nervous system consists of paired
cerebral, pleural, pedal, and visceral ganglia. The Mollusca are either
sexually separate or hermaphrodite. The gonads are usually single,
with paired or unpaired ducts. In the course of development a
modified Trochophora arises from the gastrula ; this is the Veliger
larva, typical of the Mollusca.

These general characteristics of the Molluscan body have to be modified for each

class. In each class there are series of forms which deviate from the typical

organisation in some one important point, or in several. The shell may disappear,

and so may the mantle. Either one or both of the gills or ctenidia may be lost,

VOL. II B

COMPARATIVE ANATOMY

CHAP.

and new, morphologically different respiratory organs may be substituted. The
visceral dome may be flattened down, and the foot become rudimentary or disappear.
Teeth of all kinds may be wanting. The complex of the sub-pallial organs may be
so displaced as to lie anteriorly, thereby causing a very pronounced asymmetry of
the whole organism. But the typical Molluscan characteristics are never so entirely
obscured that the members of the race cannot be recognised, on the one hand by
meaiis of transition forms leading to well-known Molluscan types, and on the other
by their developmental history.

The Molluscs are divided into the five following classes : —

I. Amphineura.
III. Scaphopoda.

II. Gastropoda.
IV. Lamellibranchia.
V '. Cephalopoda.

Systematic Review.

CLASS I. Amphineura.

Bilaterally-symmetrical Molluscs. The nervous system consists of two lateral
and two ventral nerve trunks, bound together by numerous commissures, and

m fipTJlTp^-'jp^:'

'"mg^^

FIG. 1.— Chiton, from life (after Pretre, in the Voyay: <!•• I'

provided with ganglion cells throughout their whole length ; these pass anteriorly
into the cerebral ganglion. Special sensory organs are reduced. Marine.

ORDER 1. Placophora (Polyplacophora) sive Chitonidse.

On the dorsal side there are eight consecutive shelly plates overlapping like the
tiles on a roof. There is a distinct snout. The branchiae are numerous, and are
arranged in two longitudinal rows, one on each side in the groove between the foot
and mantle. The foot (except in Chitonellus) is strongly developed, with a large flat

vii MOLLUSC A— SYSTEMATIC REVIEW 3

sole for creeping or for attachment. The sexual ducts and the nephridia are paired.
The sexes are separate. The heart is provided with two auricles. Radula (3 + 1),

(2 + 1), (1 + 1 + 1), (1 + 2), (1+3). Chiton (Fig. 1), ChitoncUus.

ORDER 2. Aplacophora sive Solenogastres.1

The body is almost cylindrical, and generally vermiform. There is no shell.
The much thickened cuticle contains calcareous spicules. The foot is rudimentary,
a mere ridge being left, and the mantle cavity is reduced to a groove at the sides of
this ridge, and a cavity (cloaca) at the posterior part of the body, into which the
intestinal canal and nephridia open, and in which are found, when present, the
rudimentary gills. The nephridia serve as ducts for the genital products.

Family 1. Neomeniidse.

The foot is a longitudinal ridge, which rises from the base of a medio-ventral

FIG. 2.— Proneomenia Sluiteri, two-thirds natural size. A, From the right sick- ; B, from
beneath ; o, mouth ; c?, cloaca.

longitudinal furrow. This family is hermaphrodite. Proneomenia (Fig. 2), Neo-
menia, Lcpidomenia, Danders in.

Family 2. Chsetodermidse.

The foot and the pedal farrow are quite degenerated. The sexes are separate.
Cficctoderma.

•
CLASS II. Gastropoda (Cephalophora). Snails.

The body is asymmetrical. The head, which carries tentacles and eyes, is
generally distinct from the body. The foot is well developed — usually with a flat sole
for creeping. The large protruding visceral dome may be flattened down secondarily
in all the groups. It is covered by a shell, consisting of a single piece, into which the
animal can withdraw. In all divisions, however, though rarely among the Proso-

1 Simroth, in the new edition of Bronn's Klassen und Ordnunyen dcs Thierreiches,
vol. iii., 1893, divides the Solenogastres as follows :—

Sub-Order. Fain.

Cnsetodermatina . . Chaetodermatidse.

Neomeniidae.

,T Proneomeniidse.

Neomemma . . T

Dondersiidse.

Parameniidse.

COMPARATIVE AX ATOMY

CHAP.

6

branc/iia, this shell may become more or less rudimentary (generally in connection
with the reduction of the visceral dome).

The pallial complex becomes shifted forward on to the right (seldom the left)

side, or along this side so as to lie quite
anteriorly. The visceral dome and
shell (with some exceptions) are spirally
coiled.

In all except the lowest Proso-
b/'iii/chia, the asymmetry is evidenced
by the disappearance of one gill, of one
kidney, and of one auricle.

The radula is rarely wanting.

ORDER 1. Prosobranchia.

The pleuro- visceral connectives are
crossed. The mantle complex is twisted

round to the front side of the visceral
Fir, 3.-Margarita Groenlandica (TrodM, after dome_ Ju mogt forms there ig Qnl

Pelseneer). 1, Head ; 2, anterior epipodial lobes ; ... , .

3, foot ; 4, pigmented prominence at the base of the one glU> l'laced anteriorly to the heart,
epipodial tentacles '(5); 6, visceral dome. and in the heart the auricle lies anter-

iorly to the ventricle. The Proso-
branchia are chiefly marine, and are sexually separate. The foot is generally pro-
vided with an operculum for closing

the aperture of the shell. A shell is c 1

wanting only in Titiscania, a genus
of the Xeritacea. \

Sub-Order 1. Diotocardia.

The heart has two auricles (except-
ing in Docoglossa). There are two
kidneys. Instead of the pedal ganglion
of other Gastropoda, there are two
longitudinal nerves in the foot, sup-
plied with ganglia and connected with
one another by numerous commissures.
The gills are feathered on two sides,
their points projecting freely. The
epipoclium is well developed, and
there is a circle of more or less
numerous tentacles around the base
of the foot. Proboscis, penis, and
siphon arc all w7aiiting.

a. Zeugobranchia (Rhipidoglossa,
Aspidobranchia). — Two gills ; both
auricles well developed. Heart tra-
versed by the rectum. Shell with

\

FIG. 4. — Patella vulgata (from beneath, after

marginal cleft, or with apical perfora- Lankester). «, Tentacle ; d, efferent branchial vessel
tion or with a row of perforations.
Generally without operculum.
Marine. Fam. Haliotidce, radula
col. (5.1. 5)1 °° , Fissurellidce (Fissu-
rella, rail. eel. (4.1. 4)1. oo, with secondarily symmetrical shell.

c, free edge of the shell ; e, free edge of the mantle
x-y, median line; g, afferent branchial vessels
/, branchial lamellfe ; h, one of the afferent vessels
i, spaces between the shell muscles ; ft, foot.

Emurginula, Scutum

VII

MOLL USG A— SYSTEM A TIC RE VIE W

= Parmopkorus), Plcurotomaridce (Pleurotomaria, Sdssurella, Polytrcmaria), Bellero-
phontidoe (exclusively fossil).

b. Azygobranchia. — One gill, homologous with the left gill of the Zeugo-
branchia. Right auricle ending blindly. Heart perforated by the rectum. Fain.

Turbonidce, rad. c»0. (5.1.5. )0.oo , Trochidce (Fig. 3) Stomatiidce, Neritopsidcc, rad.

ool. (2.0.2.)!. oo, marine, Ncritidce, rad. ool.(3.1.3.)l.oo (marine, but along the shore
able to live out of water), Neritince (marine and fresh-water). The Hydrocoenidcc, rad.

ool.(l.l.l.)l.oo, and Helicinidce, rad. oo 1. (4.1.4.)!. oo , have no gills but a lung

resembling that of the Pulmonata. The Helicinidce are terrestrial.

c. Docoglossa. — Heart with one auricle, and not perforated by the rectum. Left
kidney shifted to the right side of the pericardium. Visceral dome and shell
secondarily symmetrical, the latter usually cup-like. Operculum wanting. Marine.

FIG. 5.— Phorus exutus (after Lankester). n, Proboseidal snout or rostrum; I, tentacle;
c, eye; d, foot; e, metapodmm with operculuni/.

1. Left true ctenidium present. Acmaeidce, rad. 1.2.(1.0.1.)2.1.; with numerous
accessory gills in the mantle furrow : Scurria ; — without such gills : Aemaea
(Tectura).

2. True ctenidia altogether wanting, accessory gills very numerous in the mantle
furrow.— Fam. Patdlldcc (Fig. 4), rad. 3.1. (2.0.2.)!. 3.

3. Neither ctenidia nor accessory gills found (Lcpetida:), rad. 2.0.1.0.2.

Sub-Order 2. Monotocardia (Pectinibranchia).

Heart with one auricle. A single true ctenidium feathered on one side, the point
not projecting freely (except in Valvata). Pedal nerve trunks a rare exception, pedal
ganglia the rule. Only one kidney. Siphon and penis generally present. Epi-
podium weakly developed or wanting. The Monotocardia are very numerous and
are chiefly marine.

a. Architaenioglossa. — Pedal nerve trunks. In Cypraea (and in other forms ?) a
rudiment of the right auricle persists. Fam. Cypraeidai, rad. 3.1.1.1.3, Paludinidce
(fresh-water), Cyclophoridce (terrestrial, pulmonate).

6

COM PA RA TIVE A NA TOM Y

CHAP. VII

b. Taenioglossa. — Typical radula, 2.1.1.1.2. Semiproboscidifera. Fam.
JYaticidcc (Fig. 98, p. 107), Lamellar idee. Rostrifera. Fam. Falvatidcc (fresh-water),
Ampullaridcc (fresh - water), Littorinidie, Cyclostomidce (terrestrial), Planaxidcc,
Hydrobiidcc (fresh-water), Aciculidcc (terrestrial), Truncatellidce (partly terrestrial),
Hipponycidcc, Capulidce, Calyptraeidcc, Pseudomelanidce, Melanidce, Cerithiidce,

ll

FIG. 6.— Rostellaria rectirostris (after Owen), a, Snout ; l>, tentacle ; c, stalked eye ; d, foot ;
e, metapodiuin with operculuin/; h, beak (for the siphon).

I ',, •,!!>•/ id(c, Turritdlidcc., Xcnophoridcc (Fig. 5), Struthiolaridce, Clicnopidcv,
Strombidcc (Fig. 6). Proboscidifera holostomata. Fam. Scalar idee, rad. cr.0*> ;
Solaruhc, rad. ccO» ; PymmideUidcc, rad. 0; Eulimidcc, rad. 0. Proboscidifera
siphonostomata. Fam. Colombellinidce, Tritoniidce, Cassidiidce (Fig. 7), DoUidcr.

KII;. 7.— Cassis suclosa (after Poll), a, Shell; /), beak; c, siphon; d, head; g, proboscis; c,
eye ; /, tentacle; h, foot; i, operculuin.

Janthinidfe. rad. ccO » . Heteropoda (marine Taenioglossa, with foot transformed
into a perpendicular rowing fin). Fam. Atlantidrc (Fig. 8), Pterotrachaeida
(Fig. 9).

c. Stenoglossa. — Xormal rad. 1.1.1. Rachiglossa. Fam. Tnrl!nclli<ln\
Fusidce, Mitridcc, Suecinidce, J//' ;•/<•/'/«, Purpurida:, Haliadca, Canccllariidcc,
Volutidce, Olivida', MurgincUiihv, Harpidir. Toxiglossa. Fam. Pleurotomidae,
Tcrebridw, Conida:.

n

FIG. 8.— Atlanta Peronii (after Gegenbaur). «, Pharynx ; 6, buccal ganglion ; c, tentacle ; d,
eye ; e, cerebral ganglion ; /, aorta cephalica ; g, pleuro-visceral connective ; h, coluinellar muscle ;
i, k, osphradium ; I, vagina ; m, ctenidium ; n, anus ; ", uterus ; ji, nephridium ; q, aorta cephalica ;
r, auricle ; s, ventricle ; t, aorta visceralis ; u, digestive gland (liver) ; v, ovary ; w, stomach ; x,
pedal ganglion ; y, operculum ; :, metapodiuin ; 1, sucker of the fin-like foot (rudimentary sole); 2,
foot ; 3, auditory organ ; 4, oesophagus ; "i, snout ; (*>, salivary gland.

Fie. 9.— Pterotrachea (Firola) coronata (after Leuckart). a, Pharynx; 6, proboscidal snout ;
c, rye; d, cerebral ganglion; c, pedal ganglion; /, pedal artery; <j, intestinal canal; li, pleuru-
visceral connective; i, parieto-visceral ganglion; 7,-, osphradium; 1, ventricle; MI, auricle; «,
anus ; o, ctenidium ; p, metapodiuin ; ij, appendage ; /•, aorta cephalica ; s, nerve running to the
metapodiuin ; f, artery ; «, foot ; c, common pedal artery ; w, cephalic artery ; .r, auditory organ ;
y, buccal ganglion.

8

COMPABA TIVE ANA TOMY

CHAP.

OIIDER 2. Pulmonata.

The plenro-visceral connectives are not crossed. The cteuidium has disappeared
from the mantle complex and is replaced by a lung, or respiratory vascular network,
on the inner surface of the mantle. The pallial organs lie primitively to the right,
anteriorly on the visceral dome. The edge of the mantle, with the exception of a
branchial aperture on the right, unites with the integument of the neck. In terres-
trial Pulmonata the visceral dome is often
flattened down and the shell becomes rudi-
mentary (Slugs). The operculum is always
wanting. The heart has one auricle, which
almost always lies anteriorly to the ventricle.
The Pulmonata are hermaphrodites with herma-
phrodite glands or ovotestes, and complicated
efferent ducts. They are either terrestrial or
fresh -water.

FIG. 10. — Amphipeplea leuconensis
(after Adams). «, Lobe of the mantle
bent back over the shell ; I, portion of
the shell uncovered ; c. foot.

Sub-Order 1. Basommatophora (fresh-water).

Eyes at the bases of the uon-iuvaginable
optic tentacles. Genital apertures separate, to

the right anteriorly, the male in front of the female. Fam. Limna-icta; (Limmea,
Amphipeplea [Fig. 10], Physa. [Fig. 11], Phmorbis, Ancylus], Anriculida:

FIG. 11.— Physa fontinalis (after L. Reeve), a, Mantle lobes folded back over the shell; I,
evaginated penis.

Eyes at the tips of the optic tentacles ; tentacles iuvaginable.

Sub-Order 2. Stylommatopkora.

a. Monogonopora.— With a single genital aperture to the right. Fam. Helicidse
(Helix [Fig. 12, A], Arion [Fig. 12, D], Bidlmus). Testacellidae (l><nnl, 7,,-,rt7/,-,
[Fig. 12, B], Tcstacella [Fig. 12, C]. Limacidse (Arioplifintii, Li,n>is. J'itrinci.
Zonitcs, Helicarioii). Bulimulidse (Fig. 131, Pupidse (Bulimiiius, Pupa, Clausilid),
Succineidae.

b. Digonopora. — Shell-less snails with separate male and female genital apertures,
the male anterior, the female at the posterior end of the body, both to the right.
Pallial complex at the posterior end of the body, lung cavity reduced. Fam. Vagi-
nulidse (terrestrial), Oncidiidse (marine or amphibious) ; respiration'partly by means
of dorsal branchial appendages.

VII

MOLL USC A —S YSTEMA TIC RE I 'IE W

D

FIG. 12.— .4, Helix pomatia ; /;, Daudebardia (Helicophanta) brevipes ; r, Testacella halio-
tidea; />, Arion ater ; x, shell, in D sliielil (from Lankester).

FIG. 13.— Peltella palliolum (BuU m/ilkl, aftei- Ferussac).

10

COMPARATIVE ANATOMY

CHAP.

ORDER 3. Opisthobranchia.

The pleuro- visceral connectives do not cross.1 There is one auricle placed behind
the ventricle. A shell is sometimes present, more frequently wanting. An
operculum is rarely found. Respiration by means of true ctenidia, or of adaptive
•(ills, or through the skin. The visceral dome is very often levelled down. Herma-
phrodites with ovotestes. Marine.

Sub-Order 1. Tectibranchia.

The pallial complex is to the right of the body, and is more or less covered by
the mantle fold belonging to that side. One true ctenidium (viz. that which was
originally the right) is always retained in the mantle cavity, but is often very
incompletely covered by the mantle. The visceral dome tends to disappear. A
shell is always present, but tends to become rudimentary. Generally with para-
podia, and mantle lobes covering the shell.

A. Reptantia.

frontal or cephalic

disc. Fam. Actseonidse
Accra], Gastropteridse

(Fig.

(with

14),

a. Cephalaspidse. — "With Irontal or
operculunO, Scaphandridae, Bullidse (Bulla,
Philinidae, Doridiidse.

b. Anaspidae. — Head without frontal disc ; four triangular or ear-like tentacles.
Fam. Aplysiidse (Aplysia, Dolabclla, Notarchus).

„

FIG. 14. — Gastropteron Meckelii,
with internal shell (after Vayssiere).
1, Cephalic shield (frontal disc) ; 2, para-
podium ; 3, ctenidium, left almost un-
covered by the rudimentary mantle fold ;
4, flagellum= appendage of the mantle
fold.

FIG. 15.— Pleurobranchus aurantiacus, with internal
shell (after Leuckart's Wamltcifdn), seen from the right
side, a, Rhinophores ; &, labial sail; c, genital aperture ;
«/. nrphridial aperture (?) ; c, ctenidium ; /, anus.

c. Notaspidae. — Head short, with or without tentacles. Large dorsal disc
(notfeum) in or on which a shell may lie. Fam. Pleurobranchidse (Pleurobranchus
[Fig. 15], Plcurobrancluca, Oscainius), Umbrellidae ( Umbrella, Tylodina), Peltidse.

B. Natantia sive Pteropoda.-

These formerly constituted a separate class of the Molluscs, but are now recog-
nised to be Tectibranchia adapted to a free-swimming pelagic life. The parapodia of
the Tectibranchia develop as fins or wing-like swimming organs.

1 Except in Actun/i, which is streptoueurous, and thus forms a connecting link
between the Opisthobranchia and Pulmonata on the one hand, and the remaining
Gastropods on the other [Bouvier and Pelseneer], r. Ifat. Sci., July 1893.

J The classification of the Opisthobranchs, which places the Pteropoda thecosomata
with the Cephalaspidre, and the Pteropoda gymnosomata with the Anaspidse, is accepted
on p. 110 and elsewhere.

VII

MOLL USCA—S YSTEMA TIC RE VIE W

11

a. Pteropoda thecosomata. — These are nearly related to the Cephalaspidea,
and possess a mantle, mantle cavity, and shell. The head is not distinct, and has
only one pair of tentacles. The fins, at their anterior edges, are fused over the mouth ;
the anus lies to the left. Fam. Limacinidse. An external calcareous shell, with
left-handed or sinistral twist, and a spiral operculum. Anus to the right (Lima-
fin n [Fig. 16], Pcradis). Fam. Cavoliniidae. External symmetrical shell (Clio,
Cavolinia}. Fam. Cymbuliidse. Internal cartilaginous shell (Cymbulia, Cymbuli-
opsis, Gli'bn}. The Thecosomata feed chiefly on small Protozoa and Alga?.

O.

FIG. Hi.— Limacina Lesueuri (dorsal aspect,
after Pelseneer). 1, Penis ; 2, tin (parapo-
diuin) ; 3, seminal furrow ; 4, mantle process
("balancer"); 5, visceral dome; 6, head with
two tentacles and the seminal furrow 3.

FIG. 17.— Pneumoderma (diagram from tin-
right, after Pelseneer). 1, right evaginated
process bearing hooks (hook sac) ; 2, proboscis ;
'•>, right buccal tentacle ; 4, position of the right
nuchal tentacle ; 5, right tin (parapodium) ; 0,
seminal furrow ; 7, genital aperture ; 8, position
of the .jaw ; 9, ventral proboscidal papilla ; 10,
right buccal appendage provided with suckers ;
. 11, head ; 12, aperture for penis ; 13, right an-
terior pedal lobe ; 14, anus ; 15, posterior pedal
lobe ; 16, ctenidiuin ; 17, posterior adaptive
gill; d, v, «, fi denote dorsal, ventral, anterior,
and posterior.

17

b. Pteropoda gymnosomata. — These are nearly related to the Anaspidnc.
They have no mantle, mantle cavity, nor shell. The head is distinct, and carries
two pairs of tentacles. The fins are separate ; the anus lies to the right. Fam.
Pneumodermatidse. One ctenidium to the right (Dexiobranckcca, Spongiobranchcea,
Pneumoderma [Fig. 17]). In the last two genera there is an adaptive posterior gill
as well. Fam. Clionopsidse and Notobranchseidae. X"o ctenidium, but a posterior
adaptive gill. Fam. Clionidse. Neither ctenidium nor adaptive gill. All Gymnoso-
mata are carnivorous, feeding principally on Tlt<-'-n»nnata>

Sub-Order '1. Ascoglossa.

This sub-order is characterised by the fact that the worn-out teeth of the long
narrow radula, which consists of a single row of dental plates, are preserved in a sac

COMPARATIVE ANATOMY

CHAP.

at its anterior end. No jaws. The anus almost always dorsal. Except in the
Steganobranchia, the disappearance of the mantle and its cavity is accompanied by
the disappearance of the single ctenidinm of the Tectibranchia.

Section 1. Steganobranchia. — With mantle, cavity, and ctenidium to the right ;
with a shell and parapodia. Fam. Oxynoidea (Oxynoe, Lobiger).

Section 2. Cirrobranchia. — Leaf- or club-shaped processes found laterally on the
back. Fam. Hermseidae, Phyllobranchidae.

Section 3. Pterobranchia. — The sides of the body produced into lobes, in
which the branches of the glands of the mid -gut spread out. Fam. Elysiadse,
Placobranchidse.

Section 4. Abranchia. — Neither ctenidium, nor dorsal appendages, nor leaf-like
lateral expansions of the body. Respiration through the skin. The body is almost
like that of a Planarian. Fam. Limapontiidse.

Sub-Order 3. Nudibranchia.
Without mantle fold, shell, or ctenidium. Jaws almost always found. Radula

Fie. IS.— Aeolis ruflbranchialis (right aspect, after Alder ami Hancock). «., Eye ; 6, oral
tentacle ; c, cephalic tentacle ; </, anus ; e, genital aperture ; /, dorsal respiratory appendages
(cerata).

generally well developed, with teeth which fall off and are lost. Adaptive gills
very variously developed, but occasionally wanting.

JJ 12,

FIG. 19. — Phyllirhoe bucephalum (lateral aspect, alter Souleyet, modified). 1, Tentacle ;
•2, cerebral ganglion ; 3, stomach ; 4 and (12, intestinal c;tca (funning the* digestive gland) ;
5 ventricle ; 6, auricle ; 7, pericardial aperture of the kidney; S (kidney ; 9, external aperture of
the same (on the right side) ; 10, anus (on the right side) ;«11, hermaphrodite glands, the ducts not
drawn ; 13, genital apertures ; 14, buccal ganglion ; 15, salivary glands.

Section 1. Holohepatica. — One large unbranched hepatic gland (liver). Fain.

VII

MOLL USC A —S Y STEM A TIC BE VIE W

13

k-T

Phyllidiidae. Numerous branchial lamella; lie in a groove which encircles the body.
No jaws and no raclula. Pharynx transformed for sucking. Fam. Doridopsidse.
Without jaws or radula ; pharynx adapted for sucking. Branchial rosette round
the dorsal anus. Dorididse crypto-
branchiatse. The branchial rosette
round the dorsal anus can be with-
drawn into a cavity. (Batltydoris,
ArcMdoris, Discodoris, Diaulula, K< 'nf-
rodoris, Platydoris, Chromodoris, etc.)
Dorididae phanerobranchiatse. Bran-
chial rosette not retractile. (Gonio-
doris, Polyccra, Acaidhodoris, Idalia,
Ancula, Euplocamus, Triopa, etc.)

Section 2. Cladohepatica. — Diges-
tive glands more or less broken up into
separate branched canals spreading
widely in the body. Variously formed
dorsal appendages chiefly connected
with respiration. Anus usually to the
right. Fam. Aeolidiadse (Aeolidia
[Fig. 18], Ecrghia, Tcrgipcs, Galcim/,
Coryptella, Eizzolia, Faccllina, Flabel-
lina, Fiona, Glaucus, Janus, Hero).
Fam. Tethymelibidse, without radula
(Tcthys, Melibe). Fams. Lomanotidse,
Dotonidse, Dendronotidse, Bornellidae,
Scyllaeidse, Phyllirhoidse (Fig. 19 ;
marine free - swimming animals with
narrow laterally - compressed body,

without foot or respiratory append-
ages). Fam. Pleurophyllidiidae. Nu-

1 —

2..—

FIG. 20.— Pleurophyllidia lineata (from below,
after Souleyet). 1, Genital apertures; 2, branchial
merous branchial lamella; arranged in leaflets ; 3, anus ; 4, pedal gland ; 5, mouth ; t>,
a single row on each side along a tentacle shield ; 7, foot.

furrow between the dorsal shield and the foot (Fig. 20). Fam. Pleuroleuridse,
Tritoniadse ( Tritonia, Marionia).

CLASS III. Scaphopoda.

The body is symmetrical, and elongated dorso-ventrally. The mantle is a tubular
sac with a narrow dorsal and a wider ventral aperture. Posteriorly, the mantle
cavity reaches to the apical (dorsal) aperture. The shell forms a high tubular cone,
and, like the mantle, has a small apical and a larger ventral aperture. Ctenidia are
wanting ; the kidneys are paired. The vascular part of the circulatory system is
reduced to a ventricle ; without auricles. The sexes are separate. There are no
special ducts for the sexual products, which are ejected through the right
kidney. The mouth lies at the end of a barrel-shaped snout, and is surrounded by
a circle of leaf-like appendages. At the base of this snout there are numerous
filamentous appendages, which can be protruded through the lower aperture of the
shell and mantle. The foot is ventrally elongated. A radula is found. Limicolous.
Marine. Fam. Dentalium (Fig. 101, p. 113). The foot is relatively short ; it is shaped
somewhat like an acorn, with a conical central portion and two lateral lobes.
Siphonodentalium. The foot is long and worm-like, but broadens out at the end
into a disc edged with papilla;.

14

GOMPABA TIVE ANA TOM Y

CHAP.

CLASS IV. Lamellibranchia (Pelecypoda, Bivalva, Acephala, Aglossa). Mussels.

The body is symmetrical and more or less transversely flattened ; it has two
large lateral leaf-like mantle lobes, enclosing a spacious mantle cavity large enough
to contain the foot, which is usually hatchet- or wedge-shaped. The shell consists
of two lateral valves connected together only at the dorsal hinge. It is closed by means
of two adductor muscles passing transversely from one valve to the other (Dimyaria) ;
occasionally the anterior adductor degenerates and only one remains (Monomyaria).
On each side in the mantle cavity there is a ctenidium. There are no jaws, no
pharynx, no radula, no tentacles, and no distinct head. The kidneys and genital
organs are paired, and the latter either have separate ducts or eject their products
through the nephridia. The heart has two auricles. At each side of the mouth
there arc two oral lobes. Either sexually separate or hermaphrodite. They live
in salt or fresh water, and are either limicolous or attached.

ORDER 1. Protobranchia.

The gills with two rows of leaflets, in the posterior part of the mantle cavity ;
they correspond in all respects with the ctenidia of the Zeugobranchia, their ends

Fio. 21.— Nucula nucleus, left aspect after removal of the left valve and mantle (after
Pelseneer). a, Anterior adductor ; b, anterior retractor of the foot ; c, elevator of the foot;//, genital
mass ; e, hypobranchial gland ; /, posterior retractor of the foot ; y, posterior adductor ; 7i, cteni-
dium ; i, mantle cavity ; k, creeping sole of the foot (?); in, oral lobes (labial palps) with posterior
appendages ti and o.

project freely backward into the cavity. The foot has a sole for creeping. The
pleural ganglion can be distinguished from the cerebral. Fam. Nuculidse (Xt/cula
[Fig. 21], Leda, Yoldia, Solenomyidce).

ORDER 2. Filibranchia.

The branchial leaflets of the ctenidium have become lengthened out into long
filaments hanging far down into the mantle cavity. Each is in two parts, the proxi-
mal descending and the distal ascending (cf. Fig. 88 B). Fam. Anomiidse: mantle open

VII

MOLL USGA—SYSTEMA TIC REVIEW

15

without siphons ; Monomyarian. Foot small ; body and shell asymmetrical. Attached.
Branchial filaments entirely free (Anomia, Placuna). Fam. Arcidse : the branchial
filaments of each row connected by ciliated discs ; Dimyarian. No siphons. Foot
large (Area, Pcctunculus). Fam. Trigoniidae : ctenidia like those of the Arci</" .

FIG. 22.— Mytilus edulis (after Meyer ami Mo'bius), left aspect, with extended foot attaching a
byssus thread ; d, byssus threads ; a, exhalent aperture (anal siphon); l>, fringed edge of the in-
halent mantle aperture ; c, object to which the animal is attached.

Dimyarian. No siphons (Triyonia). Fam. Mytilidse (excluding Arii-i<lii/«_:} : cten-
idia connected by means of non - vasctilarised trabecula?. The anterior adductor
is smaller than the posterior (Heteromyarian). With siphons. Foot long. (Mytilus
[Fig. 22], Modinla, Lithodomus [boring mussel], Modiolaria).

ORDER 3. Pseudolamellibranchia.

The consecutive ctenidial filanipnts of each row are connected by means either
of ciliated discs or of vascularised trabecuhu ; and the ascending and descending

16 COMPARATIVE AX ATOMY CHAP.

portions of each filament are similarly united (cf. Fig. 88. p. 92). Fam. Pectinidse:

"

' -

FIG. 23. — Pecten Jacobaeus, ventral aspect, shell opened. The mantle cleft is seen between
the fringes of the mantle, which are beset with numerous tentacles and eyes (after Leuckart and
Nitsche, Zoi.'l. n'o

o

FIG. 24.— Anatomy of the Oyster (Ostrea edulis), right aspect (after Mobius, Leuckart, and
Nitsche, '/.iv>l. U'liiiiltujUii). br, Gills ; Pn, posterior mantle nerve ; :e. ti, apertures of the cavities
between the fused plates of the two left gills ; .U, large adductor muscle ; a, anus ; Mm, posterior
portion of the adductor muscle ; Pa, mantle ; P, pericardium ; I", heart ; 30, gonad (hermaphrodite
gland) ; d, intestinal canal ; I, digestive gland (liver) ; o, mouth ; o$, o$i, oral lobes (labial palpsXof
the left side ; Cg, cerebral ganglion ; it, kidney ; bn, branchial nerve ; Vg, visceral ganglion ; PI, abdo-
minal process ; Pni, nerve of the pallial edge ; »i, stomach, with the apertures of the digestive gland.

Monomyarian with mantle entirely open, and eyes at its edge. Without siphons.
Foot small and linguiform. Valves of the shell equal or unequal. Capable of

VII

MOLLUSC A— SYSTEMATIC REVIEW

17

(Pcctai [Fig. 23], L'hlannjs). Fam. Aviculidse : Monomyarian or Hetero-
myarian without siphons. Valves equal or unequal (Acicula [Meleagrina], Malleus,
Vulsclla, Pcrna, Inoccramus, Pinna, Meleagrina man/aritifera, pearl mussel).
Fam. Ostreidse : Monomyarian without foot, with completely open mantle, without
siphons. Valves unequal, the left valve attached to the substratum. (Oatrn/ :
oyster [Fig. 24]).

ORDER 4. Eulamellibranchia.

The gills no longer consist of distinct filaments. On the contrary, the filaments
in each row and the two parts of each filament are so connected by means of
vascularised trabeculte or sutures as to form together a lamella or trellis -work.
There are, on either side, two such branchial lamellrc (hence the name of Lmnclli-
branchia), which in fact correspond with the two rows of leaflets of the typical
ctenidium. This order includes the majority of the Lamellibranchia.

Sub-Order 1. Submytilacea.

Branchial lamellae smooth. The mantle edges usually grown together only
between the inhalent and the exhalent apertures. Dimyarian. Fam. Carditidse :
with open mantle and large foot (Cardita, Vcncricardia). Fam. Lucinidse :
with simple, and as a rule single, siphonal aperture. Foot often vermiform. Fam.

FIG. L'5.— Anatomy of Unio (Margaritana) margaritiferus, left aspect (after Leuckart ami
Nitsche). o, Mouth ; &j, cerebral ganglion ; J/i, anterior adductor muscle ; a?, oesophagus ; J,
digestive gland (liver); no, nephridial aperture ; lo, aperture of the digestive gland in the stomach
•m ; Aa, anterior aorta ; «., nephridiuui, the. outlines given in dotted lines ; V, heart ; r, proctodseum ;
Ap, posterior aorta ; M2, posterior adductor ; a, anus ; V<j, visceral ganglion ; Br, gill ; £/,-, mantle
cavity ; go, gonad and ducts j/oj ; Pg, pedal ganglion ; p, foot. The arrows mark the direction
of the inhalent and exhalent streams of water.

Erycinidse : mantle closed except at the two siphonal and the pedal apertures.
Foot long. (Erydna, K<Utja, Lascca, Li'pton, Galeomma.) Fam. Crassatellidse :
mantle open without siphons. Foot moderately developed. Fam. Cyrenidae : mantle
open, two siphons. • Foot large. In fresh or brackish water. (Cyrena, Corbicula,
Sphcerium, Pisidium, Galatea.) Fam. Dreissensiidse (fluvial). Fam. Unionidse: fresh-
water ; foot large, hatchet- or wedge-shaped, two simple siphonal apertures or clefts,
mantle open (Unio [Fig. 25], Painter's Mussel ; Anodonta, pond Mussel ; Mutcla).
VOL. II C

18

COMPARATIVE ANATOMY

CHAP.

Sub-Order 2. Tellinacese.

Diniyarian with completely separate siphons. Foot large. Gills smooth. Fam.
Tellinidse (Tcllina). Fam. Donacidse (Donax), Mactridse (Martm).

Sub-Order 3. Veneracea.

Dimyariau with somewhat folded branchial lamella?.. Siphons separate, and foot
large. Fain. Veneridae (/"•/*!«, Aft'trtrix [C>jthe>vf<~\, Tapes). Fam. Petricolidse :
boring muscles.

Sub-Order 4. Cardiacea.

Dimyariau or Monomyarian. Branchial lamella? much folded. Mantle closed
except at the two siphonal and one pedal apertures. Fam. Cardiidse : Dimyarian.

710,

Fa;. -JO.— Anatomy of Cardium tuberculatum, h-ft aspect (after Grobben, Leuckart, and
Nitsche, Zonl. n'miJtufeln). ji, Foot ; ijo, gonad ; S, shell ; Pa, mantle ; os, labial palps ; o, month;
-l/i, anterior adductor muscle; a1, oesophagus; m, stomach; I, digestive gland; </, intestinal
canal ; 300, genital aperture ; /"<i, pericardia! aperture of the kidney ; V, ventricle. ; At, auricle ; /',
pericardium ; no, aperture of the kidney in the mantle cavity ; n, kidney ; jl/o, posterior adductor ;
Bl, point of concrescence of the right and left ctenidia behind the foot ; a, anus ; At;, anal chamber
of the mantle cavity, with anal siphon ^ls ; l!k, branchial chamber of the same cavity with
branchial siphon Bs ', Br, ctenidium.

(Cardinal [Fig. 26].) Fam. Cbamidse : Dimyarian. Valves of shell unequal. (

Jt'i'ijii.ii'ittit.) To these the fossil forms MmmpJ, //,•/'/", C<tprini<la. Hip-
Huil inl it iiii<_\ Fam. Tridacnidse : ilonomyarian. (Tridacna, ffippopus.)

Sub-Order 5. Myacea.

Diniyarian with folded branchial lamella. Tendency towards concrescence of the
edges of the mantle folds. Siphons very long and foot large. Fam. Psammobiidae :

pedal cleft of the mantle still

large (Pxuiinii<A>i<i'\ Fam. Mesodesmatidse.

VII

MOLL USC A —S YSTEMA TIC RE VIE W

19

Lutrariidae, Myidse (Mya, Coi'luta). Fam. Glycymeridse ('////<•/////. ris, Saxicava
[boring mussels]). Solenidse : shell with anterior and posterior cleft ; foot very
large (Solenociirtus, Cultcllus, Eiisis, Solo/).

Sub-Order 6. Pholadacea.
Dimyarian with closed mantle and well-developed siphons. Foot varies, and is

M,

Cf

Npcr

FIG. 27.— Anatomy of Pholadidea, left aspect (after Egger). Lettering as before. In addition,
Npa, Npp, anterior and posterior nerves of the mantle ed^'e ; mo, anterior aperture of mantle; A's,
sac of the crystalline stylet ; A'r, branchial vein ; nl, anterior upper mantle lobe ; Rpp, posterior
retractor of the foot ; Ss, partition between the two siphons ; JlTy, accessory adductor ; ml>, intestinal
caecum ; r, pericardial section of the kidney, which opens into the pericardium by means of the
renal funnel at •».

M,

HFC

Fir,. -28.— Anatomy of Jouannetia Cumingii, left aspect (after Egger).

last figure.

as in

sometimes rudimentary. Shell open, often having accessory pieces added to it.
Fam. Pholadidse: boring mussels ( Pholas, Pholadidm [Fig. 27], Jouannetia [Fig. 281,

20

COMPARATIVE ANATOMY

CHAP.

FIG. 29.— Teredo Navalis in its boring,
ventral aspect (after Meyer and Mobius).
The centre is omitted, the calcareous tube
is for the most part uninjured.

FIG. 30.— Shell of Aspergillum (Bre-
chites) vaginiferum, dorsal view. «, An-
terior; p, posterior; d, right; s, left; 1,
siphonal aperture of the pseudoconch ; 2,
pseudocouch (calcareous tube) ; 3, true shell
embedded in the pseudoconch ; 4, anterior
aperture of the pseudoconch.

VII

MOLL USC A —S YSTEMA TIC RE VIE W

21

Xylophaga). Fam. Teredinidae : boring mussels (Tercitu [Fig. 29]). Fam. Clava
gellidse (Clavagelht, B whites \_Aspcrgillum, Fig. 30]).

Sub-Order 7. Anatinacea.

Mantle to a great extent closed. With siphons and foot. Hermaphrodite.
Fam. Pandoridse, Lyonsiidse, Anatinidse (Am/fiiin, Thracia).

ORDER 5. Septibranchia.
The ctenidium on each side is transformed into a muscular septum pierced by

a

10

FIG. 31.— Soft body of Silenia Sarsii (Cuspidaria), after Pelseneer. A, Left aspect after
removal of the mantle ; B, ventral aspect after removal of most of the mantle ; a, p, anterior ami
posterior ; d, v, dorsal and ventral ; d, s, right and left ; 1, anterior adductor ; 2, mouth ; 3, anterior
group of branchial slits; 4, hepatic mass ; 5, branchial septum ; 6, posterior group of branchial slits ;
7, posterior adductor; 8, anal siphon ; 9, siphonal tentacles ; 10, valve of the branchial or inhalent
aperture ; 11, point where the free mantle edges limiting the pedal aperture fuse ; 12, median group
of branchial slits ; 13, free edges of mantle ; 14, foot ; 15, posterior labial palps ; 16, anterior labial
palp.

slits, which divides the mantle cavity into two chambers, one lying above the other.
Hermaphrodite. Fam. Poromyidae, Cuspidaridas (Fig. 31 A and B).

CLASS V.— Cephalopoda (Cuttlefish).

Body symmetrical with high visceral dome. The mouth is surrounded by
tentacles or prehensile arms, which may be considered as portions of the foot
developed round the mouth. Another portion of the foot forms the siphon. In

22

COMPARATIVE ANATOMY

CHAP.

the posterior mantle cavity there are two or four ctenidia. The heart has two
or four auricles, and there are two or four kidneys. Gonad unpaired, with single
or paired ducts. The sensory organs are highly developed, especially the eyes,
which lie anteriorly and laterally on the "head" (Kopffuss). The jaws and
radula are powerful. There is sometimes a shell, either external or internal. An
ink-bag is generally present. The Cephalopoda are large, highly-developed marine
carnivora. Diojcious.

OKBKK 1. Tetrabranchia.

An external chambered shell, the animal inhabiting the last (and largest)
chamber. It is symmetrical, and exogastrically coiled. The mouth is surrounded
bv numerous tentacles without suckers, which rise from large lobes and can be

FIG. 3'2.— Nautilus Pompilius, after Owen. Median section of shell. K, Cephalic hoi id;
6, tentacles; «-, infumliliuluin (siphon); <', eye : e, projection causeil by nklainental gland; /, point
of attachment of the adductor muscle; ;/, upper portion of the visceral dome ; h, last (inhabited)
chamber of the shell; i, anterior lol t the mantle ; /.•, last chamber but one; I, siphuncle.

retracted into special sheaths. There are four ctenidia, four auricles, and four
kidneys. The siphon consists of two lateral lobes distinct from one another, which
by the apposition of their free edges form a tube. Without ink-bag. The eyes are
simple pits. The only living form is the Nautilus, radula 2.2.1.2.2 (Fig. 32).
The two large divisions of this order, Nautiloidea and Ammonitidea,1 occur as
fossils.

1 Tlie Ammonitidea, owing to the uncertainty concerning their anatomy, are by
many authorities arranged in a separate order, "Ammonea," and placed between tlie
other two.

VII

MOLLUSGA—SYSTEMA TIC REVIEW

23

Km. :«.— Spirula prototypes, riicht aspect (after Chun and Owen),
from Leuckart ami Nitsche, /ool. ll'iniiltiifrhi. Both portions of the
shell are visible, the inner portion seen through the mantle. The eye
should be placed more anteriorly on the " head" (Kopllus-o.

Fin. 34.— Loligo vulgaris (alter D'Orbigny). .1, Dorsal (physiologically ventral) view; 7;,
anterior (physiologically dorsal) view, of the five pairs of arms, the fourth are seen to be
developed as lon^ prehensile tentacles ; the eyes, the cd.ue of the mantle, the lins, and the
ohromatophores in the skin are depicted.

24

COMPARATIVE ANATOMY

CHAP.

ORDER li. Dibranchia.

The shell is either internal, rudimentary, or altogether wanting. When
present it is endogastrically coiled. There are two ctenidia, two auricles, and two
kidneys. The mouth is surrounded by eight or ten sucker-bearing prehensile arms.
The free edges of the two lobes which form the siphon have grown together. The
eyes are vesicular. An ink-bag is present.

Sub-Order 1. Decapoda.

Shell internal and often rudimentary. There are ten arms, the fourth pair being
developed into long prehensile tentacles, which can be withdrawn into special cephalic
cavities. The Decapoda are good swimmers ; their bodies are elongated dorso-
ventrally, and provided with lateral fins. The oviduct is unpaired. Fam. Spirulidse :
internal shell spirally (endogastrically) coiled. Spirilla (Fig. 33). Fam. Belem-
nitidse : fossil forms with internal chambered shell, usually long and straight
(Behmnites, Spirulirostra, Belemnoteuthis). Fam. Oigopsidse ( OmmastrcpJies, radula
3.1.3, Loligopsis, Cranchia, Chiroteuthis, Owe/tin, Thysanoteufivis, Onychotevihis,
Ommatost replies). Fam. Myopsidse (Rossia, Sepiola, Scpiadarrinn, Idiosi'pion, Loliyo
[Fig. 34], Sepiotcuthis, Belosepia [fossil], Sejria. radula 3.1.3).

Sub-Order 2. Octopoda.

Without shell or "guard " (rostrum) ; eight arms ; without specialised prehensile
tentacles. Body thick, generally without fins, and little adapted for swimming.

FIG. 35.— Female Argonauta, in the swimming position, right aspect (after Lacaze-Duthiers).
1, Uncovered part of the shell ; -2, the right arm of the first (anterior) pair, with its lobe-like expan-
sion'(sail) 3, covering a large part of the shell; 4, fourth arm; 5, third arm; 6, siphon; V, eye;
S, jaw; 9, second arm. The second, third, and fourth arms are stretched backwards inside the
shell.

Oviducts paired. Fam. Cirrhoteuthidse : with fins. Fam. Philonexidse : Ai'<j«,i<nit<i
(Figs. 35, 36, and 200, p. 243). Female with external unchambered shell. Philonexis,
'/';•• i/ioctopus. Fam. Octopodidse (Ck-lnpua, radula 1.3.1. [Fig. 37], Eledonc).

VII

MOLL USO A —S YSTEMA TIC EEVIE V '

25

FIG. 36.— Female of Argonauta Argo (after Verany). Second, third, and fourth pairs of arms
stretched downwards, a, Siphon; ?>, eye; c, first pair of arms, covering with its sail d nearly the
whole shell e.

_s-2--.<

-.-

FIG. 37.— Octopus vulgaris, after Merculiano (in "Aquarium Neapolitanum"). Above, in
swimming position ; below, quiescent, watching for prey.

26

COMPARATIVE ANATOMY

CHAP.

I. Organisation of the Primitive Mollusc.

The hypothetical primitive Mollusc, reconstructed from the results
of morphological research, may be described as follows :—

The body is bilaterally symmetrical and dorsally arched ; its

anterior end carries the mouth, eyes, and
tentacles, forming a distinct head. The
ventral side forms a powerful muscular
foot, distinct from the rest of the body,
with a flat sole for creeping.

The soft integument of the arched
dorsal side forms a fold, which hangs
down all round the body, and is called
PIG. 3.s. -Hypothetical Primitive the mantle or pallium. The mantle

Mollusc, diagrammatic, left aspect, o, -, •, ., .,

Mouth; fc, head; sm, shell muscle ;osoi encloses a Circular Cavity, the mantle-
upper aperture of the shell; a, anus; m, or pallial cavity, which surrounds the

renal aperture; mil, mantle cavity; c.t, ^ an(j communjcates freely With the
cteniilinin ; /, foot. it

surrounding medium between the tree

edge of the mantle and the foot. The dorsal integument of the body
and of the mantle secretes a closely-applied shell, which consists of a
chitinous matrix (conchyolin) in-
ter-stratified with deposits of car-
bonate of lime. This shell repeats
the form of the dorsal surface, and
is thus bilaterally symmetrical and
arched. Such a shell detached and
turned over would resemble a cup
or plate. Since the dorsal shell
covers the whole, or at any rate
the greater part of the body, it
forms a protection for it and at
the same time plays the part of a
skeleton, to which the muscles run-
ning more or less dorso-ventrall}'
into the foot and head, can be
firmly attached.

The mantle is of special im-
portance as a protective Structure.
Apart from the fact that its edge

pic. 3H.— Hypothetical Primitive Mollusc.
°m "bove. o, Mouth : «?..\ ult,l, »//-. primitive
,1 F .-, i 11 left cerebral pleural and pedal ganglia; ulpa,

secretes the greater part of the shell urpa> prim'iti/e left and right parietal ,a]ldi:, .

Substance, and in this Way adds to nl«, primitive U-ft auricle ; uos, wros, primitive

the Shell as the animal grOWS, it left

, , ,. ... *-

covers the delicate glllS,

tllUS also share the protection mantle cavity ; c. visceral ganglion; re, ventricle;

afforded by the shell. Analogous "' anus-

arrangements are to be found in other divisions of the animal kingdom,

riSht ™l'l>ra.lia (Spengel's organ) ; Oat,
,. , urct, primitive left and right ctenidia (gills) ; ml),
which base Of the mantle ; mr, edge of the mantle ; TO,

vii THE HYPOTHETICAL PRIMITIVE MOLLUSC 27

r.<'/. the dorsal fold or carapace which, in the higher Crustacea, covers
the branchial cavity, and the operculum of Fishes. The relations
existing between the branchiae, the mantle, and the shell in the
Mollusca are of the highest importance ; these organs should always
be regarded as essentially interdependent structures.

The branchiae lying in the mantle cavity are paired and symme-
trical. It may be left an open question whether the primitive Mollusc
possessed more than one pair of gills. If only one, we must suppose
that one gill lay on each side of the mantle cavity posteriorly ; if more
than one, that there was a row of branchiae on each side.

Each gill is feather-like, with a shaft and two rows of very
numerous leaflets. The shaft stands out freely from the body in the
mantle cavity. Close to the base of each gill, a sensory organ, con-
sidered to be olfactory, and called the osphradium, is found. Such
a gill with an osphradium at its base has a very definite morphological
value ; in order to distinguish it from analogous though not homologous
respiratory organs found in certain Mollusca, it has been named a
etenidium.

The head is provided with one pair of tentacles and one pair of
eyes. The mouth lies anteriorly and ventrally. The remaining open-
ings of the inner organs lie posteriorly above the foot ; the anus in the
middle line, and on each side, between it and the etenidium (supposing
that there is only one pair of cteniclia), an aperture for the sexual
organs, and another for the kidney (nephridium). These five apertures
are covered by the mantle, and thus lie in the mantle cavity. We
have thus, to recapitulate, in the posterior part of the mantle cavity
two ctenidia, two osphradia, and five apertures, the median anus, and
the paired symmetrical sexual and renal apertures. These, taken
together, form what is known as the pallial complex.

The inner organisation may thus be briefly described.

The intestinal canal. The mouth leads to a muscular pharynx, with
horny jaws. At its base lies a chitinous rasp-like ribbon called the
tongue or radula, which carries numerous consecutive transverse rows
of sharp chitinous teeth. Paired salivary glands enter the pharynx,
which passes into an oesophagus, which latter leads into the mid-
gut. This, which we will suppose to be more or less coiled, runs
right through the body, passing posteriorly into a very short hind-
gut, which opens outward through the median anus. The mid-gut
has large paired glandular diverticula (mesenteric gland, diges-
tive gland, hepatopancreas, liver).

Musculature. --The muscles of the foot are powerful, and are
adapted for the creeping movement. There are, in addition, muscles
running from the inner surface of the shell into the foot and head
(eolumellar or shell muscles), and special muscles for the different
organs.

Nervous system. --Two well - developed cerebral ganglia lie
dorsally in the head, and are connected by means of a short cerebral

28 COMPARATIVE ANATOMY CHAP.

commissure, which runs over the oesophagus. Each cerebral ganglion
gives rise to two powerful nerve trunks which are provided along
their whole length with ganglion cells ; there are thus two pairs of
nerve trunks running right through the body longitudinally. One pair,
the pedal cords, run right and left in the foot ; the other pair, the
visceral cords, which lie more dorsally and are more deeply embedded
in the body, run through the body cavity. The two visceral nerves
are connected posteriorly.

If we leave the Amphineura and Diotocardia out of the question,
the following modified sketch of the Molluscan nervous system holds
good. Two cerebral ganglia, two pedal ganglia, two pleural
ganglia lying at the sides of the pharynx, two visceral ganglia
lying posteriorly in the body cavity. Giving the name connectives
to such nerves as unite the ganglia of one side of the body, i.e. dis-
similar ganglia, and that of commissures to the nerves that unite the
similar ganglia of the two sides of the body, we have the following
system: Commissures are found --(1) between the two cerebral
ganglia (over the fore-gut) ; (2) between the two pedal ganglia
(under the fore-gut) ; (3) between the two visceral ganglia (under
the hind-gut). The connectives on each side are: (1) the cerebro-
pedal connective ; (2) the cerebropleural connective ; (3) the pleuro-
pedal connective ; (4) the pleurovisceral connective.

There is a secondary ecelom or body cavity lined with endo-
thelium, which has at least two divisions. In the anterior division,
the genital chamber, the sexual products arise from the endothelium ;
this chamber is connected by means of two canals, the genital ducts,
with the mantle cavity. In the posterior chamber, or pericardium,
lies at least one organ, the heart ; this chamber is connected with the
mantle cavity by means of two nephridial duets or vesicles.

The circulatory system is partly vascular and partly lacunar.
The arterial heart lies in the pericardium above the hind-gut. It
consists of one ventricle and two lateral auricles.

II. Review of the Outer Organisation characterising the Chief

Groups of the Mollusca.

Having given above a general plan of the morphology of the Mollusca, let us
now see how far the various groups of Molluscs agree with this description in their
outer organisation. We shall at first only mention in connection with each group
those special features which are now considered to be typical or characteristic of
that group. In other words, we shall again give a general scheme of the outer*
organisation of each class of the Mollusca, in order that these more specialised
schemes may be compared with that of the hypothetical primitive Mollusc above
described.

Later sections will deal with the changes which the separate organs undergo,
not only in the different classes, but within one and the same class, so far, that is,
as these modifications bear on external morphology.

vii MOLLUSC A— OUTER ORGANISATION 29

A. Plaeophora or Polyplaeophora (Chitonidse).

The body of the Plaeophora is bilaterally symmetrical, and dorso-
ventrally flattened ; viewed from the dorsal or ventral surface its
shape is that of a long oval. On the ventral side there is a large
muscular foot with a flat sole, the outline of which runs very nearly
parallel with that of the body. In front of the foot, and also on the
ventral side, there is a distinct snout which carries the mouth in the
middle of its ventral surface. There are no eyes or tentacles on the
head. Between the mantle, which forms the outer edge of the body,
and the body and head it covers, there is a deep groove, in the base of
which lie numerous lancet-shaped gills, arranged in a single row on each
side. These two rows of gills sometimes approach each other so nearly
both anteriorly and posteriorly that there is an almost complete circle
of gills around the foot, or else they are more or less shortened, and
are in some forms so reduced as only to occupy the posterior third
of the branchial furrow. The anus lies posteriorly in the median
line, ventrally, immediately behind the foot. The two apertures
of the nephridial ducts lie in the branchial furrow on each side, and
slightly in front of the anus. The two genital apertures lie imme-
diately in front of the nephridial apertures, also in the branchial
furrow.

The median dorsal region is covered by eight consecutive imbri-
cating calcareous plates. The peripheral dorsal region, between the
edge of the body and these shell plates, carries calcareous spicules,
granules, etc. The corresponding peripheral region on the ventral
side forms one of the boundaries of the branchial groove, and may be
considered as the mantle.

B. Aplaeophora, Solenogastres.

The body is here bilaterally symmetrical and vermiform ; in section
it is round, and is sometimes long and thin, at others short and thick.
The large oral aperture lies in the form of a longitudinal slit on the
ventral surface of the anterior end of the body. The cloacal aperture
—or common opening for the intestinal canal and the urogenital
organs — lies ventrally at the posterior end of the bodjr. A narrow
median ventral groove runs forward from the cloacal aperture
and terminates anteriorly near the mouth. In the base of this pedal
groove rises a ciliated ridge or fold which runs along its Avhole
length ; this ridge, in cross section, is triangular, and represents
the reduced foot. In the Chcefodcrma both foot and pedal groove
are wanting. The Solenogastres have no distinct compact shell ;
its place is taken by calcareous spicules embedded in the integu-
ment.

30

COMPARATIVE ANATOMY

CHAP.

C. Gastropoda (Cephalophora).

Although there can be no doubt as to the relationship to one
another of the Mollusca grouped together in this class, it is almost
impossible to give a general scheme of the outer form of the whole
class. The greatest variation occurs, the body being sometimes out-
wardly bilaterally symmetrical, sometimes in a high degree asym-
metrical. Further, forms such as Fissuretta, Oliva, Turritdlu, Cleodom.
Pterotrachea, Phi/llirhoc, Lima?, Pleurobranchus, TJtcti/s, differ so greatly
in outward appearance that, at the first glance, it is almost impossible
to believe that they are related. A shell may be present, and may show
the most marvellous variation in form ; or it may be rudimentary or
even (in adult forms) altogether wanting. The foot also may assume
the most varied forms, or may be entirely wanting. The same may
be said of the mantle fold, the gills, etc.

Setting aside those forms which are quite one-sidedly differenti-
ated, it may be said in general- — (1) that, in the Gastropods, the
protective shell consists of one piece, and follows in a remarkable way
the forms assumed by the bod}' ; (2) that the dorsal portion of the
body, which contains the viscera, becomes constricted almost hernia-
like from the head and foot, making a sac-like protuberance (visceral
dome) ; (3) that, for the diminution of its surface, this dome or hump
becomes coiled spirally, the shell repeating its shape ; (4) that the head
and foot, which project through the aperture of this shell for purposes

of locomotion, can be withdrawn
into it. The large, long foot
generally has a flat sole for creep-
ing. The head is distinct, and
provided with tentacles and eyes.
At some part of the body, the in-
tegument of the visceral dome
forms a mantle fold which hangs
downwards, covering and protect-
ing the respiratory organs. The
outer surface of this mantle takes
part with the rest of the integu-
ment of the visceral dome in the
formation of the shell. The follow-
ing are more special descriptions of
the outer organisation of the chief
Gastropodan groups.

1. Prosobranchia.

The large visceral dome is
coiled spirally, generally to the
right (dextrally), the shell naturally assuming the same form. The well-

FIG. 40.— Diagram of the Organisation of a
Zeugobranchiate Diotocardian. </, Aims ; w .
ventricle' ; nln, ri^Ht aurk-le ; tint, left ctenirlium ;
tiros, It't't osphradium.

VII

MOLL USCA— OUTER ORGAN IS A TION

31

CVIA

I

Avvt

T

ue-

developed foot has a flat creeping sole. On the dorsal side of the

posterior portion of the foot, the metapodium, there is a calcareous

plate, the opereulum, which, when the animal withdraws its head and

foot, closes the aperture

of the shell. The mantle

fold hangs down from

the anterior side of the

visceral dome, and covers

the spacious branchial or

mantle cavity, in which

lie certain organs of

special morphological

importance. These,

which may be called the

mantle or pallial organs,

are, in such forms as \

' . , . . 'Utd

may be considered primi-
tive, (1) the anus, which
lies, not posteriorly, but
on the anterior side
of the visceral dome,
shifted forwards to-
wards the mouth ; (2)
the two apertures of the
paired nephridia, one on
each side of the anus; (3)
the tAvo gills, one to the
left and one to the right :
(4) the two osphradia
near thebases of thegills.
In most Prosobranchia,
however, the organs just
mentioned as paired are
unpaired; only the gill Fio.ii.-DiagramofaProsobranchiateMonotooardian. The

nephndial aperture, and outer form, sliell, mantle, pallial complex, heart and perirardi ,

Osphradilim to the left aervous system and operculum, are depicted. Lettering mostly

n , i -I as in Fig. 3'.i. In addition : /, foot ; si, siphon ; tup, xtiti. supr.-i-

and sub -intestinal connectives ; op, operculum ; ^.auditory

tailied, while the hind- organ: y, penis; sr, seminal groove; wh, mantle cavity; hy,

°"llt with the anUS moves nyi"-'w'am'llial gland ; ,5, male genital aperture ; /. rectum ; au,

f . • i, • -I "" r , 1 «-ye : ', tentacle.

to the right side ot the

mantle cavity. The single genital aperture lies on the right side, in
the head, or on the floor of the mantle cavity. (In the Prosobranchia
the sexes are separate.) The abortion of one of each of these original!}'
paired organs, gills, nephridia, and osphradia, produces a very striking
asymmetry of the whole body. The name Prosobranchia indicates the
fact that the gills lie in front of the heart.

32

COMPARA TIVE ANA TOM Y

CHAP.

Mitt,

2. Pulmonata.

Type : Helix pomatia. — The visceral dome is well developed, and
protrudes hernia-like from the rest of the body; it is dextrally coiled,

and has a corresponding shell. The
foot is large and long, and has a
flat creeping sole. The head has
two pairs of feelers, one of which
carries the eyes. The mantle fold
hangs down from the anterior side
of the visceral dome, and covers a
spacious mantle cavity (respiratory
or pulmonary cavity). The free
edge of the mantle fold unites with
the integument of the neck near it,
only leaving an aperture to the
right, the respiratory aperture.
This aperture serves for the inhala-
tion and exhalation of the air.
The anus and the unpaired nephri-
dial aperture lie close to the re-
spiratory aperture, and are thus on
the right side. There are no gills
in the mantle cavity, which con-
tains air. Respiration takes place

Fir;. 4-->. —Diagram of a Basommatophoran ,, . ,. , ,

Pulmonate. «l, Respiratory aperture; r,jn, vas- afc the inner Surface ot the mantle

eular network on the inner surface of the mantle, fold, ill which 1'lins a fine network

The kidney is incorrectly drawn. Further letter- f vessels lying in front of fcj e

ing as in Figs. 39 and 41. J »

heart. Ihe toot, unlike that of

the Prosobranchia, has no operculum. There is a common genital
aperture on the neck, to the right, in front of the respiratory cavity
(the Pulmonata being hermaphrodite). Many Pulmonata, however,
differ greatly in their outer organisation from the Helix type.

3. Opisthobranehia.

The respiratory organs lie behind the heart.

(a) Tectibranehia. — The visceral dome is usually not large. It
may be either spirally coiled or symmetrical, and is covered by a
variously shaped shell. The foot is large, and usually has a flat
sole for creeping. The head is variously shaped, and often carries
tentacles or rhinophores, and unstalked eyes. The small mantle
fold hangs down from the right side of the visceral dome, and
often does not quite cover the single gill tying beneath it. The
anus lies behind the gill, more or less removed from it. The Tecti-
branehia are, like all Opisthobranehia, hermaphrodite ; the genital

VII

MOLL UBGA— OUTER ORGANISA TION

33

and nephridial apertures lie on the right side of the body in front of
the anus.

(b) Nudibranehia. - - The body is outwardly symmetrical, the
visceral dome does not protrude from it, but is closely applied to
the whole length of the foot, from which it is often not distinctly

FIG. 43. — Diagram of a Tectibranchiate Opistho-
branchiate. Lettering as before. In addition : gg, genital
ganglion ; s, shell ; 9 , female genital aperture ; Ipp, rpp,
left and right parapodial lobes, that on the right laid back.

FIG. 44. — Dentalium, diagram-
matic, left aspect, g, Sexual glands ;
kt, cephalic tentacles ; other letter-
ing as before.

differentiated. The foot has a Hat creeping sole. There is no distinct
mantle fold, no gill corresponding with that of the Tectibranchia,
and no shell. The head carries tentacles or rhinophores, and sessile
eyes. The anus lies either dorsally in the median line, or laterally to
the right. The genital and renal apertures lie to the right in front
of the anus. The gills, which vary much in form, number, and
arrangement, are found dorsally or laterally, and are not homologous
with the typical Molluscan ctenidia.

VOL. II

D

34

COM PA RA TIVE A NA TOMY

CHAP.

r,

kb

D. Seaphopoda.

The body is symmetrical and long, i.e. the visceral sac is elongated
dorso-ventrally, and is completely enveloped in a tubular mantle. The
mantle cavity lies posteriorly, and is prolonged ventrally far enough to
allow the snout and retracted foot to be completely concealed in it.
Besides the large ventral aperture, there is a smaller dorsal aperture
further placing the mantle cavity in communication with the exterior.
The shell, like the mantle, is tubular, or like a tapering cone, slightly
curved anteriorly. It has two apertures corresponding with those in
the mantle. The head is developed into a barrel-shaped snout, and
has no eyes. The mouth, which lies at its ventral end, is surrounded
by a circle of leaf-like tentacles. At the base of the snout there arise
two tassels of long filamentous contractile tentacles, which hang down
into the mantle cavity and can be projected far beyond the ventral aper-
ture. Behind the snout, the cylindrical muscular foot rises from the
body, and can be protruded downwards. There are no gills. The

median anus lies posteriorly above
the foot. The two njephridial
apertures are at the sides of the
anus. There are no special genital
apertures (Figs. 44 and 101,
p. 113).

E. Lamellibranehia.

The body is bilaterally sym-
metrical ; somewhat elongated
(from before backward). The
integument forms leaf-like mantle
folds to the right and to the left,
which at their bases are attached
to the trunk along its whole
length, and grow down ventrally.
If the body of a Lamellibranch,
from which the shell has been
removed (the foot being re-
tracted), be viewed from the side,
the outline will be found to be

(Keber's organ); re, kidney (glandular portion) : forme(J clorsallv, by the dorsal

..1,^ nl.ni-nV.nno of -Hio. Vincac nf tlio rrill< • ftrJ crPliitnl ** * "

median line of the body ; an-
teriorly, posteriorly, and ventrally
by the free edge of the mantle
fold. The two mantle folds en-
close a space whose transverse
axis is always markedly shorter than either its dorso-ventral or its
longitudinal axis, i.e. the animal with its mantle is laterally compressed.

ibc

Kn.i. 45. — Transverse section of Anodonta
cygnsea (ordinary freshwater mwsseZ) (after Howes).

lij, Ligament; tij, typhlosolis ; kb, pericardialglam

sic, chambers at the bases of the gills ; yd, genital
ducts ; /;/•/!, brl;, outer and inner branchial lamella- :
ibc, mantle cavity ; s, shell ; .?i, edge of the shell ;
fl, foot ; pm, pallia! muscle ; i, intestine ; pli, right
mantle fnld ; ggl, gonad ; r, rectum: 171, etTt'bro-
pedal connective ; rt'i, non-glandular vestibule of
kidney; ;r._,. renal aperture ; )"', pericardium.

VII

MOLL UXCA — OUTER ORGANISATION

35

Projecting into the mantle cavity, there is a large muscular process of
the body, the foot, which is directed downward and somewhat forward,
and can be protruded between the free edges of the mantle. This foot
is also laterally compressed. In certain cases which, though excep-
tional, deserve special mention, its free end is flattened, and it thus has
a fiat sole. The outer surface of the trunk and mantle folds secretes
a bivalve shell which covers the whole body. One valve lies to the
right, the other to the left of the median plane, and the two are exactly
alike. Each valve repeats the outline of its own side of the trunk
with its mantle fold.

The two valves articulate dorsally, and are open anteriorly, ventrally,
and posteriorly. Two strong muscles (adductors) run transversely

FIG. 40.— Anatomy of Unio (Margaritana) margaritiferus, left side (after Leuckart and Nitsche).
o, Mouth ; Cg, cerebral ganglion ; MI, anterior adductor muscle ; <r, oesophagus ; I, digestive gland
(liver) ; no, nephridial aperture : lo, apertures of the digestive gland in the stomach m ; Act, anterior
aorta; n, nephridium, the outline given in dotted lines; I", heart; r, hind-gut; A]>, posterior
aorta ; Mn, posterior adductor ; a, anus ; I";/, visceral ganglion ; Kr, gill ; /;/,-, mantle cavity ; </<>,
gonads with genital duct </f>i ; 1'fi, pedal ganglion; p, foot. The arrows indicate the direction of
the inhalent and exhalent streams of water.

from one valve to the other. Their contraction serves to shut the
shell completely. One of these muscles lies anteriorly, the other
posteriorly. Their points of attachment produce impressions on the
inner surface of the shell, which are always distinctly visible when
the shell is removed.

The mouth lies below the anterior adductor, between it and the
anterior base of the foot. The anus lies behind the posterior adductor.
There is no distinct head. Near each side of the mouth, the body
carries two leaf-like processes, the oral lobes or labial palps. At the
line of insertion of the foot in the mantle cavity, a longitudinal
ridge rises on each side in the middle and posterior regions of the
body ; this carries two rows of long branchial leaflets. There is thus,

36

COM PA RA TI VE ANA TOMY

CHAP.

on each side of the mantle cavity, one plumose gill, the shaft of which
is attached lengthwise to the body (Figs. 45, 46, etc.).

In various divisions of the Lamellibranchia, the outer organisation
deviates very greatly from the above.

F. Cephalopoda.

The body is bilaterally symmetrical. The visceral dome is large

and often much elongated dorso-ventrally. The head is more or less

distinct, and is surrounded by the foot,
«*- which is transformed in a peculiar man-

ner. The foot has, in fact, grown round
the head, and has developed numerous
differently -shaped processes (arms and
tentacles) arranged in a circle round the
mouth ; these serve principally for seizing
and holding prey. In viewing the body
of a Cephalopod, it must be remembered
that the apex of the visceral dome (which
a casual observer might take to be the
posterior end of the body) is really the
highest dorsal point, while the head and
its arms lie lowest. We may thus dis-
tinguish, both in the visceral dome and
in the transformed foot which has been
combined with the head, and drawn out
into tentacles, an anterior and a posterior
part (which to a casual observer would
seem upper and lower parts), and a right
and a left side. This at first sight seems
a paradox to those not acquainted Avith
the comparative anatomy of the Mollusca,
since the normal position in the water of
certain well-known Cephalopods does not
agree with it. A Sepia, for example,
swims or lies at rest in such a way that
the strongly pigmented anterior side of
the visceral dome and of the "head"
(Kopffuss) is uppermost, and the posterior
side lowermost. The accompanying dia-
gram illustrates the strict morphological

position of the body, which alone concerns the comparative anatomist

(Fig. 47).

On the right and left of the " head " there is a highly-developed

eye, and near it an olfactory pit.

The mantle fold hangs down posteriorly from the visceral dome,

covering a spacious mantle- or respiratory cavity, which communicates

FIG. 47. — Diagram of Sepia, median
section from the left side, v, Ventral
(physiologically anterior) ; d, dorsal
(physiologically posterior) ; an, anterior
(physiologically upper) ; po, posterior
(physiologically lower) ; 1, '2, 3, 4, 5, the
five arms of the left side ; au, eye ; co,
internal shell; 30, gonad; d, pigment gland
= ink-bag; m, stomach; n, kidney; ct,
ctenidium ; a, anus ; mh, mantle cavity ;
in, siphon. The arrows indicate the
direction of the respiratory current.

VIE

MOLL USG A —0 UTER ORGANISA TION

3K
i

with the exterior at the free edge of the mantle fold, above the
"head." Within the mantle cavity there are two or four gills,
arranged symmetrically, the median anus, and the apertures of the
sexual and excretory organs. Two symmetrical lobes are found on
the posterior lower side of the visceral dome ; the edges of these
are apposed in such a way as to form a tube, the funnel or siphon,
one aperture of which lies in the mantle cavity, while the other
protrudes from the mantle cleft. The respiratory water enters the
mantle cavity through the mantle cleft, and escapes through the
siphon. The faecal masses, waste and sexual products, and the
secretion of the ink-bag also leave the body through the siphon.
Originally, no doubt, all Cephalopoda possessed a shell which covered
the whole visceral dome as well as the mantle fold. In recent
Cephalopods the shell is rarely developed in this way ; it is often
rudimentary, and may, indeed, be altogether wanting. Recent
Cephalopods fall into two entirely distinct divisions, the Tetra-
branchia and the Dibranchia.

01

The Tetrabranehia (Nautilus, Fig. 48).

These have a shell coiled anteriorly (exogastrically) in the plane
of symmetry, and divided by septa into consecutive chambers.
The animal occupies
the last and largest
chamber ; the others
contain gas.1 The septa
separating the consec-
utive chambers are
pierced in the middle
to allow of the passage
of a siphunele, which
runs through all the
compartments, and is
attached to the visceral
dome. That portion of
the foot which sur-
rounds the mouth is
produced into numerous Fic; 48._Diagram of NautiluSj left view. ve> Ventral . do>

tentacles, which can be dorsal ; vo, anterior ; hi, posterior ; /, foot (tentacles and siphon) ;
retracted into Special sm> shell muscle ; ct, cteniclia ; mil, mantle cavity ; a, anus ; s,
•• , , shell ; si, siphunele ; cm, eye ; o, mouth.

The anterior portion of the foot, which lies in front of and over
the head, is widened out into a concave lobe, the hood ; this is applied
to the outer, surface of the occupied chamber of the shell anteriorly,
and, when the tentacles are withdrawn, can close its aperture. The
hood carries two tentacles, and on each side of the head there is an eye.

1 Or water ; r. Ford's Introduction to Brit. Mus. Cat., Fossil Cephalopoda, 1889.

vs.

COMPARATIVE ANATOMY CHAP.

Above the head, the mantle fold encircles the whole body. It is short
at the sides, but anteriorly it forms a large lobe which is folded back
over the shell in the way shown in Fig. 32, p. 22. Posteriorly, the
mantle covers a very deep cavity which contains the whole posterior side
of the visceral dome. The siphon consists of two entirely distinct
lateral lobes (epipodial lobes), whose free edges overlap in such a
manner as to form a tube, open above and below. As we shall see
later, this siphon is a part of the foot. Deep down in the mantle
cavity, two pairs of pinnate gills — a lower and an upper pair — spring
from the visceral dome. Nine apertures of inner organs are also
found in this cavity ; a single median anal aperture, and four paired
apertures, viz. one pair of genital, two pairs of nephridial, and one
pair of viscero - pericardial apertures. The position of these is
depicted in Figs. 78 and 79, p. 82.

The Dibranehia.

With one exception, viz. the female .7/y/"/""//", which has an
external unchambered shell, the Dibranehia either have an internal
shell lying on the anterior side of the visceral dome, covered by an
integumental fold, or no shell at all. The visceral dome is sometimes
compact and pouch-like (in reptant animals, Fig. 37), sometimes, in
the good swimmers, much elongated dorso-ventrally, produced dorsal! y
to a point, and flattened antero-posteriorly (Fig. 34). In the latter
case, the body is further generally encircled by a fin-like integumental
fold, which marks the limit between the anterior and posterior sides of
the visceral dome.

The " head " is usually distinct from the visceral dome, and carries
to the right and left the well -developed eyes. The mouth is sur-
rounded by eight or ten arms for seizing prey ; these are provided
with suckers on their lower adoral sides.

The mantle fold covers nearly the whole posterior surface of the
visceral dome, and thus encloses a very deep and spacious cavity.
Laterally and anteriorly to the visceral dome, the mantle fold is
continued as a narrow border which, immediately above the " head, '
covers a shallow groove or furrow.

The two lateral lobes which form the siphon of the Tetrabranchia
have in the Dibranehia grown together at their free edges, and form
a tube open at each end. There are only two gills in the mantle
cavity, one right, and one left. Near the upper siphonal aperture in
the mantle cavity lie the anus, and the genital and nephridial apertures
as well as that of the ink-bag. Details as to the arrangement and
number of these apertures will be given further on.

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 39

III. The Integument, the Mantle, and the Visceral Dome.

The whole body is covered by a single layer of epithelium, which,
in parts not protected by the fl

shell, may be more or less
ciliated. This layer is very
rich in glands Avhich are
almost exclusively unicellular :
some of these lie in the epi-
thelium itself, while some
have sunk into the subjacent
tissue, their ducts, however,
passing between the epithelial
cells.

The layer immediately
beneath this body epithelium
is called the corium, and con-
sists of connective tissue and
muscle fibres. It is, how-
ever, not distinctly marked
off from the tissues beneath
it.

The pigment is almost
always found in the cells of
the subepithelial connective
tissue.

FKJ. 40.— Section of the integument of Daudebardia
rufa (after Plate). 1, Epithelium ; 2, 3, i>, various forms
of unicellular glands ; 4, globular pigment cells ; 5, 7,
unpigmented cells of the connective tissue ; 6, muscle
fibres ; 8, branched and anastomosing cells of the con-
nective tissue containing pigment.

A. Placophora. (Of. the sketch of the Outer Organisation, p. 29.)

The CKitini is provided dorsally with eight consecutive shell-plates (Fig. 1. p. 2),
which overlap in such a manner that the posterior edge of each plate covers the anterior
edgeof the next. These plates are bilaminar. The outer and upper layer which forms the
dorsal surface is called the tegmentum, the lower hidden layer the articulamentum.
As a rule, the tegmentum of the anterior plate only is as large as the articulamentum
beneath it ; in the other plates, the latter is the larger and projects beyond the
former laterally and anteriorly. These projecting parts of the articulamentum,
called apophyses, slide under the plate next in order anteriorly. Between these
two layers, tissue is found, which is a continuation of the dorsal integument.
The tegmentum is penetrated by canals of various sizes, which open at its
surface through characteristically arranged pores.1 The tegmentum consists of a
horny or chitinous substance, which may be considered as a cuticular formation,
impregnated with calcareous salts. The articulamentum is compact and free from
canals ; it contains little organic substance, and much calcareous salt. It alone
answers to the shell of other Molluscs, while the tegmentum must be considered as
a calcified cuticle covering the true shells (the articulamenta) as a continuation of
the cuticle of the zone which encircles the eight shell-plates. This zone carries

1 On the relation of these canals and pores to peculiar sensory organs and eyes on
the shell of the Chiton, cf. section on Sensory Organs, p. 166.

40

GOMPARA TIVE ANA TOMY

CHAP.

t

chitinous or calcareous spines, setae, scales, granules, etc., varying in number and

arrangement according to the genus and species.

Each spine, as a rule, arises as a globular vesicle within an epithelial papilla and

above a very large formative cell (Fig.
50). As it grows, it is pushed upwards by
the newly - forming cuticular layers. The
formative cell at its base persists, but remains
connected with the epithelial papilla only by
a protoplasmic process which continually
lengthens, and may surround itself with a
nucleated sheath. In fully-developed spines,
the remains of this cell are still found as a
small terminal swelling (Endkiilbchen).

There are, however, spines and specially
flat scale- or plate-like calcareous formations
in the integument which do not arise from
single" large formative cells, but are probably
produced by several cells in the base of an
epithelial papilla.

Just as we have recognised the tegmentum
covering the articulamenta to be merely a
special portion of the general cuticle, so we
may further recognise in the articulamenta
the homologues of the calcareous spines,
scales, etc., which are developed in the integu-
ment of the mantle. The articulamenta

would thus be nothing more than very large and expanded calcareous scales.

FIG. 50.— A, B, C, Three stages in de-
velopment of a spine in the Chiton (after
Blumrich), diagrammatic, st, Spine; bz, its
formative cell ; e, epithelium ; c, thick
cuticle secreted by the epithelium ; el;,
terminal swelling (Endkolbchen) = remains
of the formative cell.

FIG. 51.— Transverse section through a Chiton near the nephridial apertures, highly
diagrammatic (after Sedgwick), somewhat modified. 1, Pericardium ; 2, ventricle ; 3, auricle ;
4, branchial "vein"; 5, branchial groove (mantle cavity); (>, gill (ctenidium) ; 7, foot; S, pleuro-
visceral connective; 9, branchial "artery"; 10, secondary cceloni ; 11, intestine; 12, posterior
portion of the gonacl lying below the pericardium ; 13, 14, the two posterior branches of the
nephridium, one of which (13) opens into the branchial groove (at 16), the other (14) being connected
in a way not here depicted with the pericardium ; 15, pedal nerves.

This view, finally, leads to the conclusion that the shell (if it may here be so

vn MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 41

called) of the Molluscs originally consisted of isolated calcareous spicules or spines,
which were enclosed in a thick cuticle, and projected from the same as in the
Proiieomenia, Neomenia, etc. (v. below).

In Cryptochiton the shell is internal, i.e. it is entirely covered by a fold of the
integument, which grows over it from all sides. It consists exclusively of the
articulamentum, since the whole dorsal integument is covered by an even cuticle,
which therefore forms no tegmentum.

The only part of a Chiton which can be called the mantle fold is the marginal
zone of the body, the ventral side of which encircles the head and foot and forms
the lateral boundary of the branchial groove or furrow. Just as the dorsal side of
this mantle, which is called
the zone, carries large
spines, setae, or scales, so
may the under surface be
covered with small closely-
crowded spines. The rest
of the integument is bare,
being merely covered with
a simple epithelium.

The genus Chitonellus
is of great importance in
comparing the outer or-
ganisation of the Placo-
phora with that of the
Solenogastres. The body
is not dorso-venti'ally flat-
tened, as in the Chiton,
but nearly cylindrical ; the
ventral surface, however,
is flattened (Fig. 52), and

-00

-C

DC--

has a median longitudinal

FIG. 5i>. —Transverse section of Chitonellus, diagrammatic,
adapted from figures by Pelseneer and Blumrich. g, Shell (articu-
lamentum) ; 30, gonacl ; i, intestine ; ab, vb, branchial arteries and
veins ; pv, pleuro-visceral nerves ; x, latero-ventral thickening of
the cuticle ; p, foot ; ct, ctenidium ; pn, pedal nerve ; h, digestive
gland (liver) ; c, secondary ccelom ; ao, aorta.

groove. The foot is not
externally visible, but can
be discovered, much reduced, in the base of the median groove, itself possessing
a ventral median groove representing a narrow contracted sole. The flat ventral
surface is therefore the mantle. In the narrow cleft on each side, between mantle
and foot, in the posterior half of the body, lie the gills. The lateral margin of
the body in Chiton is represented in Chitviti'1/nx by a mere blunted ridge, which
is almost exclusively caused, as may be seen in transverse sections, by a great
thickening of the cuticle.

B. Solenogastres.

In the Solenogastres (Aplacophora), whose outer organisation has already been
sufficiently described (p. 29), the shell is altogether wanting, but the cuticle secreted
by the epithelium over the whole body is usually exceedingly thick (Fig. 53). It
contains calcareous spicules, which sometimes project above the surface. These, like
the spines of the Polyplacophora, rise from cellular cups, which are connected with
the basal epithelium of the cuticle by nucleated stalks. There can be no doubt
that the spicules are formed by these cups and nourished by them during growth.
The foot, as we have seen, is reduced to a narrow ciliated longitudinal ridge, which
rises from the base of the medio-ventral groove. The term mantle is here inappli-
cable, except perhaps to the integument which forms the lateral boundary of this

42

COMPARATIVE ANATOMY

CHAP.

Iii Chaetoderma the foot finally atrophies, and the medio-ventral groove also

lisappears.

The long series of undoubtedly primitive characteristics in these two groups — the

Placophora and Solenogastres — obliges
us to place them, as we shall have
repeatedly to point out, near the root
of the Molluscan phylum. In some
points the Solenogastres are perhaps
more primitive than the Pul>/j>/itc<i-
/iltum, and the vermiform body, the
slight development of the mantle, the
foot and the gills have been thought
to be primitive characteristics. More
recently, however, it has been main-
tained, as the present writer thinks,
with justice, that these conditions are
rather the result of secondary adapta-
tion to a limicolous habit of life (most
Solenogastres inhabiting mud). The
shell, mantle, gills, and foot are such
essential characteristics of the Mol-
lusca that we must assume their
existence in the racial form.

The series Chiton, Chitoncllus,
Neomenia,, Chaetoderma does not, there-
fore, illustrate for us the rise and
development of typical Molluscan

characteristics, but rather their progressive obscuration and disappearance.

Fi<;. •>'-'i. — Transverse section of Proneomenia
Sluiteri in the region of the mid-gut. 1, Mid-^ut ;
•2, rudimentary Toot ; 3, sepia projecting into the mid-
^ut ; 4, testicnlar portion of the gonad ; 5, ovarial
portion of the same ; 6, thick cuticle secreted by the
epithelium.

C. Gastropoda. (Cf. Sketch of Outer Organisation, pp. 30-33.)

Integument.

The free edge of the mantle, which takes the chief part in the
formation and growth of the shell, is particularly rich in mucous, pig-
ment, and calcareous glands.

The epithelium is ciliated over areas of varying extent, especially
in aquatic Gastropods. In many of the shell-less Opisthobranchia the
whole surface of the body is ciliated.

The remarkable marking and colouring of the integument especi-
ally seen in the Nudibranchia are caused by pigment cells, which are
more often found in the cutis than in the epithelium.

Where there is no firm shell, calcareous granules or spicules may
be found scattered throughout the cutis.

In several Nudibranchia stinging cells have been discovered in
the integument.

Mantle, Visceral Dome.

The mantle fold is, as a rule, well developed in Gastropods, and
covers a spacious pallia! cavity. Whenever the fold is small or alto-
gether Avanting, the condition is secondary rather than primitive.

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 43

1. Prosobranehia.

In the Prosobranehia, the mantle fold develops on the anterior
side of the visceral dome, and there covers a spacious cavity. It
further usually extends like a narrow collar right round the base of
the visceral dome.

In the symmetrical Fissurellidce, the mantle cavity is short, and opens out-
wardly by means of a dorsal aperture through the mantle fold, which corresponds
with the perforation at the apex of the shell. A circular fold, provided with a
highly sensitive fringe, is formed by the mantle around the aperture, and projects for
a short distance beyond the perforation in the shell. The water needed for respira-
tion passes into the pallial cavity through the slit-like aperture at the free edge of
the mantle fold, over the nuchal region, and flows out through the apical aperture
just described. This aperture also serves for the ejection of excretory matter from
the rectum, which lies immediately behind it. In Simula, the apical apertures in
shell and mantle have moved somewhat forward, and lie anteriorly between the
apex and edge of the shell. In Emarginula, the mantle has an anterior cleft, the
edges of which, in the living animal, are folded in such a way as to form a tubular
siphon, which can lie protruded through the marginal cleft of the shell. In Par-
mophorus there is no second opening into the mantle cavity, but the lateral edges
of the mantle are very much widened, and bent back dorsally over the outer surface
of the shell in such a way as to cover the greater part of it.

In Haliotis, the enormous development of the columellar muscle on the right
side confines the mantle cavity to the left. The mantle fold has a long slit reach-
ing from its edge to the base of the pallial cavity. This slit lies under a row of
perforations in the shell which are characteristic of Hull: >tis, and through these the
respiratory water is expelled. In the spaces between the consecutive perforations, the
edges of the mantle cleft are apposed, merely separating beneath each aperture to
allow of free communication between the cavity and the exterior. The edges carry
three tentacular processes, which can be thrust outward through the perforations.
The anus is always found under the posterior perforation. The edge of the mantle
surrounding the body splits into two narrow lamell;e, which bend round to form
a groove for the reception of the edge of the shell.

The Trochidu; Turbiniifa:, Xcritiibr, and nearly all MonotocanUu have no second
aperture and no mantle cleft.

In Docoglossa (Patella, etc.) the mantle forms a circular fold round the visceral
dome, which is in the form of a blunt cone. It covers the edge of the almost
circular broad-soled foot. The mantle is broadest anteriorly, where it covers the
head and neck, i.e. the pallial cavity or groove is here deepest.

The visceral dome, in the Monotocardia, is almost always distinctly constricted
at the base, and spirally coiled. The pallial cavity occupies its typical position.
In many Monotocardia, the free edge of the mantle fold is prolonged on the left side,
projecting forward, sometimes to a great extent ; the lower edges of this projecting
fold bend round towards each other to form a tube or semi-cylindrical channel,
which is called the siphon. Through this siphon, the water needed for respiration
Mows into the mantle cavity. It can generally be told, by the shape of the shell, if
there is a siphon or not, since most Momitncitnlia which possess one have either a
notch in the edge of the shell at the columella, or a process called the canal or beak,
at this same point, which encloses the siphon. The length of this latter canal need
not, however, correspond with that of the siphon.

44

COMPARATIVE ANATOMY

CHAP.

A

The Monotocardia have even been grouped, according to the presence or absence
of a siphon, into the Siphoniata or SipTumostomata, and the Asiphoniata or Holo-

stomata ; but this classification is artificial,
since siphons are sometimes present and
sometimes absent in forms which are un-
doubtedly nearly related.

In most Monotocardia, the shell is not
outwardly covered by the mantle, but in
some groups, the edges of the mantle bend
back over the shell, and finally grow over
it .to "such an extent as to unite above it.
The external shell in such cases becomes
an internal shell.

In the Harpidce among the Jihachiglossa,
the mantle bends back over the columellar
lip of the shell. In the Marginellirfce, it
covers a large part of the outer surface, and
the same is the case in Pyrula among the
Taenioglossa, in most Cyprccidcc and in the
Lamcllaridcc. In Lamcllaria, the shell is
completely grown over by the mantle. In
Stilifer among the EuHnn<l<i- also, the shell
is more or less covered by the mantle.

The edge of the mantle may be fringed
or notched, or (Cyprneidce) provided with
wart -like, tentacular, or branched ap-
pendages.

2. Pulmonata.

In the Pulmonata, the arrange-
ments of the mantle fold and visceral
dome and of the shell, which is in-
timately connected with them, are of

ff

great interest.

We have, on the one

pharynx evaginated through the buccal cavity,
carrying on its surface the radula (a) ; c, open-
ing of the pharynx into the oesophagus ; d,
position of the genital aperture ; c, latero-dorsal
groove along the body ; /, latero-ventral groove ;
g, mantle, rudiment of the visceral dome. B,
dorsal view : a, b, the two pairs of tentacles ;
o, the latero-ventral groove ; d, the latero-
dorsal groove ; e, shell.

hand, forms such as Helix, with large
FIG. 54. — Testaceiia naiiotidea (after protruding spirally -coiled visceral

Lacaze-Duthiers). A, right view ; 6, enormou>

a spacious cavity ; on the other,
forms such as Oncidium, without
distinct visceral dome or mantle fold
and without shell. Between these
two extremes there are numerous
transition forms ; indeed, complete
series of such forms may be found

even within some of the natural divisions of the Pulmonata. The

following are a few characteristic types.

Helix (Figs. 12 A, p. 9 ; 72, p. 75). — The visceral dome is large and spirally
coiled, and is covered by a spiral shell sufficiently large to shelter with ease the whole
body. The mantle fold covers a cavity lying anteriorly to the visceral dome (pul-
monary cavity). Its free thickened glandular edge unites with the nuchal integument

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 45

near it in a way characteristic of the Pulnionata, leaving only one aperture, the
respiratory aperture — on the right. (In Pulnionata whose shells have the sinistral
twist, the respiratory aperture lies to the left. ) The apertures of the hind-gut and
excretory organ are close to the respiratory aperture, through which their excreta
have to pass out.

In many species of the genus Vitriiia, the shell cannot contain the whole animal.
The mantle fold projects in front of the shell, and has a process which is bent back
over the shell, and is used for cleansing it.

In Daudebardia (Hdicophanta) (Fig. 12 B, p. 9) the visceral dome and shell are,
in comparison with the rest of the body, much smaller than in Vitrina. The animal
cannot be sheltered by the shell. The visceral dome begins to be levelled down to a
certain extent, disappearing into the dorsal surface of the foot. It lies far back
on the body, the respiratory aperture being on its right side.

A somewhat similar arrangement is found in the genus Homalonyj', in which the
low visceral dome lies on
the centre of the back.
The respiratory aperture
lies to the right at the edge
of the mantle. The edge
of the flat ear-shaped shell
is fixed into the mantle
fold. Daudcbardia and
Hbmalonyx begin to look
like slugs.

In Testacella (Figs. 54
and 55) a visceral dome
hardly exists. The only re-
mains of it is a small mantle
at the dorso-posterior end
of the body, which is
covered by an ear -shaped
shell. Beneath the mantle
lies a reduced respiratory
cavity. The respiratory aperture lies to the right posteriorly, beneath the edge of
the shell. The viscera lie dorsally on the foot.

The common terrestrial snails Limax and Arion (Fig. 12 D, p. 9) resemble
Testacella in the reduction of the visceral dome, but in them the mantle or so-called
shield which takes its place lies anteriorly behind the head. At its right edge lies
the respiratory aperture. In Limax there is a small round rudimentary shell which
is internal, i.e. it is entirely enveloped in or overgrown by the mantle fold. In
Arion this shell is represented by isolated calcareous granules. In Onchidium and
Varjinulus there is no trace of a visceral dome, nor, in the adult, of a shell. The
visceral dome has to a certain extent spread out over the whole dorsal surface of
the foot, and has disappeared. There is, further, no outwardly recognisable mantle
fold distinct from the rest of the dorsal integument. A longitudinal furrow still
divides the dorsal part of the body from the foot. The respiratory aperture with
the anus lie posteriorly in the median line.

In the genus Physa (Fig. 11, p. 8), the edge of the mantle takes the form of
lobe-like or finger-shaped processes, which bend back over the shell, and can be
applied to its outer surface. In Amphipcplca (Fig. 10, p. 8) the mantle is much
widened and, when bent back over the shell, covers all but an oval spot on the
dorsal side of the last coil.

The dorsal integument of the Oiichidia has wart-like protuberances or (in Peronia)

h b

FIG. 55.— Testacella haliotidea, posterior portion of the body
from the right (after Lacaze-Duthiers). The shell is removed to
show the rudimentary visceral dome, a, latero-dorsal groove ; 6,
latero-ventral groove ; c, end of the muscle attached to the shell ;
c, mantle edge of the visceral dome ; <j, respiratory aperture.

46 COMPARATIVE ANATOMY CHAP.

branched appendages. These are richly supplied with blood-vessels, and serve for
respiration. In Peronfa there are besides these also dorsal prominences which carry
eyes.

The dorsal integument projects all round the body above the foot, and thus
forms, as in Chiton, a peripheral zone, which is ventrally separated from the foot
by a groove. In Oncii'ic/la the edge of this zone, i.e. the lateral edge of the body,
is dentate or fringed.

3. Opisthobranehia.

The typical outer organisation of the Gastropoda here suffers
even more varied and thorough modification than in the Pulinonata.
We have, on the one hand, forms with head, foot, visceral dome, shell,
mantle and gill ; on the other, forms which possess none of these
organs and nevertheless are both Gastropods and Opisthobranehia. In
one principal division of this order, the PtilUaia or Tc<:til>rancliia, the
mantle fold is retained on the right side of the body, and partially
covers a typical Molluscan ctenidium ; in other divisions both mantle
and ctenidia are wanting. We do not here apply the term mantle to
the fold or edge of the dorsal integument which surrounds the body
at the part where the head and foot take their rise ; such an edge is
more or less developed in most Opisthobranehia and distinctly marks
off the foot and head from the rest of the body or back. The mantle
here means only the broader fold which covers the mantle cavity, in
which lies a typical molluscan gill. The edge of the mantle never
forms a distinct siphon in the Opisthobranehia, though there is an
approach to such a structure in the Ringwulida'.

(a) Tectibranchia.

(a) Reptantia. — In this division we have, on the one hand, forms which still
have a distinctly projecting visceral dome, whose integument secretes a coiled shell,
into which the whole body can be withdrawn. On the other hand, forms occur in
which the flattened visceral dome has spread out over the whole dorsal surface of
the foot, the shell being rudimentary and internal. Examples of the former are
found in the Cephalaspidce, e.g. the Actaeonida; Tonnitiindce, and some Scnp/innili-i'f"-
(Atys, Cyliclincf, Amphisphyra), a few Bullitlcc (Bulla), and the fiiiigicitlida:

In Scaphander among the Scaphandridce, and Accra among the Bullidce, the body
cannot be completely withdrawn into the shell.

In the Cephalaspidce, to which so far reference has been made, the shell is
external.

In (.Tdstropterun the mantle is rudimentary, and is provided posteriorly with a
filiform appendage. It covers a delicate membranous internal shell, into which the
body cannot be withdrawn. The same is the case in Pliiliiic and !'<>,•/'</ in,jt, where
there is also a delicate internal shell covering only a small portion of the viscera ; this
shell, in Doridium, is produced in the form of two lobes, the one to the left endino-
in a filiform process.

The visceral dome in the Amtr<jii</<r is small as compared with the size of the
animal, but rises distinctly above the rest of the body, and is covered by a thin
inconspicuous shell. The mantle and shell often only partially cover the gill. In
Aplysia, the shell is internal, i.e. it is entirely overgrown by the mantle ; in Dola-
Mla. this enveloping overgrowth is not quite complete, as a circular median dorsal

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME

The mantle

aperture is left, through which the dorsal surface of the shell is visible.
in Dolalclln forms a small anal siphon posteriorly.

Notarchus has a microscopically minute shell. In certain species of this genus,
the integument forms protuberances or delicately
branched appendages. A

In the Oxynoidcit, the shell is only partially covered
by the mantle, and is, further, much too small to
shelter the body.

Among the Notaspiihc, the UinbreUidec have a
small flattened cap-like visceral dome lying upon the
massive foot. The visceral dome is surrounded by a
mantle fold which, on the right side, covers the gill.
The integument of the dome and mantle is covered
by a flattened disc-shaped shell.

In Pleurobraiii'/iii/, the visceral dome is relatively
large. The right and left margins project as short
mantle folds, but there are no such folds to the front
and back, so that at these latter parts the flattened
visceral dome is not distinct from the rest of the body.
In Plfurolj-ranehus, the integument of the flatten*-. 1
visceral dome broadens out into a large fleshy disc
which projects on all sides beyond the large, broad -
soled foot ; its margin (mantle fold) is separated from
the foot by a deep continuous groove running right
round the body ; in this groove, to the right, lies the
large gill, while in Pleurobranchus a small flat internal
shell, thin and membranous, is still found ; in related
forms this may be wanting. The dorsal integu-

layer of calcareous

ment is often strengthened by a
granules.

FIG. 56.— Diagrammatic trans-
verse sections of Gastropods, to
illustrate the arrangement of the
shell (black, 1), visceral dome and
mantle (dotted, 2), and foot (streak-
ed, 3). A, Prosobranchiate with

in -.-I • • , i , • , outer shell and epipodium (4). B.

shell, with simstral twist ; the shell can be closed by Abranchiate with lobes (6) of

means of a typical operculum. The mantle fold covers the mantle turned back over the
a cavity which lies anteriorly to the visceral dome, outer surface of the shell. Dorsally

Q3) Natantia.

Pteropoda Thecosomata. — The Limaciiiittcc have
a well - developed visceral dome and corresponding

The anus is to the
draw into its shell.

right. The animal can with-
in the Cavoliniidcc the dome

the shell is still uncovered ; 5, para-
podia ; 7, ctenidium. C, Tecti-
branchiate with internal shell, i.e.
and shell are bilaterally symmetrical, not twisted, and completely overgrown by the lobes
the body can be entirely hidden within the shell, of the mantle.
The mantle cavity here lies posteriorly to the

visceral dome, on what is usually called its lower side. The symmetrical shell of
the CymbuliidK does not correspond with the shell of other Thecosomata ; it is a
cartilaginous " pseudoconch " covered with body epithelium. In the Cymbal ii<l,i
the mantle cavity also lies posteriorly. We shall return later to the varying
position of this cavity among the Thecosomata.

The mantle, in the genus Cavulinin, shows peculiarities which can best be described
in connection with the shell. In the latter, two surfaces are distinguished, a slightly
arched anterior surface (usually described as the upper), and an arched posterior
surface. The anterior surface projects forwards and downwards beyond the
posterior for a third of its length. The shell has three slit-like apertures, one

48 COMPARATIVE ANATOMY CHAP.

anterior and ventral, through which the fin-like processes of the foot can be protruded,
and two lateral apertures stretching far up, so that the shell appears almost bivalve.

At these lateral slits, which admit water to the mantle cavity, the mantle bends
round on to the outer surface of the shell, covering the greater part of it ; and, at
the upper angles of the slits, has two freely projecting processes.

Pteropoda Gymnosomata. — In these, the long outwardly symmetrical body is
naked and without a mantle, and the foot, which is much rediiced, is found on
the ventral side of the most anterior part of the body.

(b) Ascoglossa and Nudibranchia.

In mature Ascoglossa and Xudibranchia, with the single exception of the Stegano-
lirauchia, a shell is always wanting, as also a distinctly demarcated visceral dome.
The latter, indeed, spreads out over the whole dorsal surface. The dorsal integu-
ment, nevertheless, forms a circular fold (mantle fold) separated from the foot by a
groove sometimes deep, sometimes shallow ; but, except in the Phyllidiidce, no gills
lie in this groove. Where this groove has nearly disappeared, the animals strongly
resemble Planaria.

Phyllidiidae. — In these, the mantle fold is distinct, and carries on its lower
surface, to the right and left, a row of branchial leaves, herein recalling PntclJa and
Chiton.

The genus Dcrmatobranchus, which, judged by its organisation, belongs here,
has, however, no gills.

Dorididae. — The dorsal integument (notieum), which here covers the body like
a shield, being generally distinctly demarcated from the foot and the head, contains
numerous calcareous particles, which give it a firmer consistency. Anteriorly, there
are two feeler-like processes, the rhinophores, which can generally be withdrawn
into special sheaths or pits ; these are not to be confounded with the tentacles.
The anus lies in the median line, generally behind the middle of the body, and
is surrounded by an ornamental circlet of pinnate gills. The notseum is often
covered with prominences, and in some genera the margin carries variously shaped
processes.

Cladohepatica. — Here there are no anal gills. The dorsal integument has
variously formed and variously arranged appendages ; these may be conical, club-
or finger-shaped, lobate or branched ; they are, for the most part, very striking in
colour and appearance. Sacs of nematocysts are generally found at their tips, and
cseca of the intestinal canal (branches of the digestive gland) penetrate them. These
dorsal appendages, which, like the rest of the body, are ciliated, have, at least
partly, a respiratory function. In many forms they easily fall off, and are later
regenerated (Fig. 18, p. 12).

Many Cladohepatica have a certain external likeness to Planaria Avith dorsal
papilla; (Thysanozoon), but this likeness is still more marked in the following
family :—

Ascoglossa. — Anal gills and also, as a rule, dorsal appendages are here wanting.
The whole body is naked and ciliated. The back is indistinctly demarcated from
the head.

Phyllirhoe. — This Nudibranchiate genus, of all Opisthobranchia, shows least of
the typical external organisation of the Mollusca. The body here is naked and
laterally compressed, with sharp dorsal and ventral edges. It has neither foot
nor gills (Fig. 19, p. 12).

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 49

D. Scaphopoda. (Cf. Eeview of Outer Organisation, p. 34.)

E. Lamellibranehia.

From each side of the body there typically hangs a large leaf-like
mantle fold of the same shape as the shell-valve formed by it. These
mantle folds project beyond the body in front, below, and behind, and
enclose a mantle cavity which everywhere, except dorsally, opens
outward by means of the slit left between the edges of the folds.
This large single cleft serves for the admission of nourishment and
water into the mantle cavity, and for the expulsion of the excreta,
genital products, and respired water ; through it also the foot is
protruded. Such a primitive mantle is thus completely open, its
simple edges (i.e. without folds, papillae, tentacles, or eyes) are quite
free, coalescing nowhere.

The above serves for a description of the mantle of Nucula—
one of the Protobranchia — and must be considered as the primitive
arrangement.

In most Lamellibranehia, however, special differentiations of the
margin of the mantle occur ; these take the form of folds, thickenings,
protuberances, papilla?, tentacles, glaneft, eyes, etc., and this is the
case both in forms which have an open mantle and in those in which
the mantle is partially closed.

The partial closing of the mantle is brought about by the con-
crescence at one or more points of the free edges of the mantle
folds.

A. A completely open mantle, i.e. one single large cleft entirely separating the
edges of the mantle, is found :

(a) Among the Protobranchia in Nucula.

(b) Among the Filibranehici in the Anomiidcc, Arcidce, Trigoniidce, and a few
Mytilidcc (Pinna).

(c) In all Pscudolamellibranchia except Meleagrina.

(d) Among the Eulamellibranchia, only in a few species of Cmssatclla.

B. The mantle folds of the two sides grow together at one point.— In this
case the point of concrescence almost always lies high up posteriorly ; and marks
off a small aperture from the originally simple cleft. This aperture, occurring on
a level with the anus, forms the exhalent or anal aperture of the mantle. Its edge
may be more or less prolonged posteriorly to form au anal siphon, which can be
protruded beyond the valves of the shell.

At a point a little below this exhalent aperture, the mantle edges usually become
applied to one another, although no concrescence takes place. Above this point,
between it and the anal siphon, they separate to form an inhalent or branchial
aperture. The edges of this aperture also may be produced posteriorly into a
branchial siphon, which, however, in this case, has a cleft extending along the
whole of its lower side, which is a continuation of the large cleft of the mantle. A
branchial siphon formed in this way, by mere apposition of the mantle edges, is found
in the genus Malletia among the Protobranchia.

VOL. II E

50

COMPARATIVE ANATOMY

CHAP.

An anal aperture, separated by a point of concrescence from the large mantle cleft,
is found in the following Lamellibranchia :

(a) Among the Protobranchia in Mallet in.

(b) Among the Filibranchia in most Mytilidce.

(c) Among the Pseudolamellibranchia in the Aviculidce (genus Mcleagrinci).

(d) Among the Eulamellibranchia, in the C'ardituhe (I'cncrn'n ,•</;</, Cardita
Miliicria}, the Astartidce, and most Grassate llidcc ; among the Cyrcnidcc, in the genus
Pisidium, ; among the Unioiiidce in the Uni<niina: (U/tio, Anodonta) ; and among
the Lucinacca, in Cryptodon *}[u»leyi.

In Sole iwni i/a, among the Protobi-micliin. the two mantle edges grow together
only at one point, but to such an extent as to close the whole posterior half of the

FIG. 57.— Diagrams to illustrate the various ways in which concrescence of the mantle
and formation of siphons take place in the Lamellibranchia. The foot (7) protruded forward
through the mantle cleft ; A, mantle completely open ; B, mantle open, but with its edges applied
to one another at two points, thus giving rise to incompletely separated anal and respiratory
cavities ; C, edges of the mantle grown together at one point (1), the anal or exhalent aperture of
tin- mantle (4) is separated ; D, edges grown together at two points (1, 2), the branchial or iulialent
aperture (5) is also separated, the mantle has three apertures ; E, mantle closed by the extension
of the place of concrescence (2), three limited apertures remain, viz. the anal, branchial, and pedal
apertures — the first two are produced into siphons ; F, a third concrescence (3) takes place.
Mantle with four apertures (4, 5, Ga, Gb), the most anterior (6b) for the protrusion of the foot. The
siphons have united.

ventral mantle opening. In this way the mantle cleft is divided into two ; the anterior
aperture serves for the protrusion of the foot, while the posterior serves at the same
time as inhalent (branchial) and exhalent (anal) aperture. Solaioiiiya is the only
1 livalve in which this arrangement is found.

C. The mantle folds grow together at two points, thus forming three
apertures. — This condition arises in consequence of the complete separation (through
concrescence) of the branchial aperture from the rest of the large anterior mantle
cleft. The anal and branchial apertures may remain as slits, or may be produced
into longer or shorter anal and branchial siphons. The large anterior and ventral
mantle cleft serves for the protrusion of the foot, and is called the pedal cleft.
These two points of concrescence are found :

(a) Among the Protolraitchin, in Yoldia and Lain.

vii MOLLUSC A— INTEGUMENT, MANTLE, VISCERAL DOME 51

(6) In most Eula/nellibranchia, viz. in most Lucinidce, most Cyrenidoc ; among
the UnionidcK, in the Mutclince, in the Donacidce, Psammobiidce, Tcllinidcc, Scrobi-
culariidce ; among the Veneracca, in the Vencridce, in the Cardiidce, the Mactridce,
Mesodcsmatidce, and the Solcnidce (excepting Solen and Lutraria,).

(c) In all Septibranchia (Poromyia, Cuspidaria).

In tlie above forms the mantle is still wide open, i.e. the points of concrescence
are small and local. But these points may become lines of concrescence of con-
siderable length. In the Cha/nacea, for example, and especially in the Tridacnidce
among the Eulamcllibranchia, the three apertures of the mantle are found at con-
siderable distances from one another, being divided by long intervals where the
edges have grown together.

In some groups of Lamellibranchia, the concrescence between the anal and
branchial apertures or siphons remains short, i.e. the one aperture lies directly
below the other, but in such cases the edges anterior to the branchial aperture
grow together to a greater extent, so that the pedal aperture becomes reduced to
a small anterior fissure. In this condition the mantle is closed. Such a mantle is
found :

Among the ISulamettibranchia in the Mudiolarca, Drcisscnsia, Pctricola, all
Phohididic (Pholns, Pholadidca, Jouannctia, in which the pedal aperture is said
to close entirely in old animals, Xylophaga, Martcsia) ; in the Tcredinidae, and
among the Pandoridcc, Pandora, the Vcrticordiidce and Lyonsiidcc (Anatinacca).

D. There are some Lamellibranchia with closed mantle, in which a fourth
aperture is added to the three found in the above groups, the mantle thus having
three points of concrescence. The fourth aperture is always small, and is found
between the pedal and branchial apertures ; it probably corresponds with a rudi-
mentary fissure for the byssus.

This arrangement is found in the Eulamcllibranchia ; among the Solenidce, in
Solcn and Lutraria ; among the Pandoridce, in Myochama ; in Glycymeris ; among
the Anatinacea, in the genus Thracia, ; in the Pholadomyidce and the Clavagellidce
(Clavagclla and Brechitcs [Aspcrgillum]) ; and, finally, in Lyonsia nonvgica.

The anal aperture is often and the branchial aperture nearly always
fringed, or in various ways edged with protuberances, papillae, or ten-
tacles, and this is the case whether these apertures are found on the
edge of the mantle or at the ends of (longer or shorter) siphons.

The siphons can be contracted and extended, and either wholly
or partly withdrawn into the shell, by means of special muscles.
These muscles are attached on the inner surface of the shell-valves to
the right and left posteriorly, and their line of attachment forms the
pallial sinus, which will be described later on.

The length of the siphons varies greatly. Specially long siphons are found in
the Mactridce, Donacidce, Psammobiidce, Tcllinidcc, Scrobiculariidce, many Veneracea
and Cardiidce, the Mcsodesmatidcc, Lutraria, the Pholadidce, Teredinidce, Anatinidce,
and Clavi/ijcUliln .

The siphons may be separated throughout their whole length, and often diverge
one from the other (e.g. Galatea among the Cyrenidce, the Donacidce, Psammobiidce,
Tellinidcc, Scrobicularidcc (Fig. 58), Mcsodesmatidce, Pharus, etc.).

In other forms they coalesce along their entire length ; they may even look like
a single tube, which is, however, always internally divided into an upper (anal) and
a lower (branchial) channel. This common siphon is sometimes protected by
a special sheath of epidermis, particularly in those forms in which it cannot be

52 COMPARATIVE ANATOMY CHAP.

withdrawn into the shell. Siphons nnited throughout their whole length are found
in the Mactridos, a few Veneracc-a, Lutraria, Solenocurtus, Solen, the PJioladidce,
many Anatinidce, and the Clavagellidae.

In some cases, siphons which are united for some distance at the base, separate
near their ends and even diverge, e.g. in Petricola among the Veneracea, Teredo, etc.

The two siphons are often of unequal length. In the Modiolaria (MytiUdcc) only
one, the anal, is developed, while the branchial aperture remains unseparated from
the large mantle cleft. The reverse is the case in Dreissensia and Scrobicularia,
where the branchial siphon is much longer than the anal.

The siphons are sometimes provided with valves ; these occur more often in the
anal than in the branchial siphon.

Significance of the development of the Anal and Branchial Apertures

and Siphons.

Most Lamellibranchia inhabit mud or sand, into which they sink the anterior part
of the body, burrowing by means of the protrusible foot. The water necessary for
bathing the gills and for respiration can only be received into and expelled from the
mantle cavity through the cleft at the posterior end of the body which projects above
the mud. The foecal masses from the anus near this point must also here be ejected
from the cavity. The development of localised inhalent and exhalent apertures is
explained by the fact that a constant regulated stream of water into and out of the
mantle cavity is necessary both for respiration and for conducting particles of food

Fio. 58.— Scrobicularia piperata buried in mud. The inhalent siphon takes in mud as
nourishment ; the anal siphon stands erect (after Meyer and Mobius).

to the mouth. The most advantageous point for the exhalent aperture is obviously
directly behind the anus.

Siphons attain development in consequence of the habit of life of many bivalves,
which bury themselves deep in mud, sand, wood, and even rock. By means of their
siphons they can still remain connected with the water which bathes the surface of
their place of concealment, and, as long as the animals remain undisturbed, a con-
stant current enters the mantle cavity through the branchial and leaves it through
the anal siphon.

Where the mantle folds have grown together to a large extent (closed mantle) the
siphons are always well developed. Such closing of the mantle is found principally
among bivalves which bore into wood, clay, rock, etc., and in which the foot of the
adult is weakly developed, or altogether rudimentary. The degeneration of the foot
leads to the shortening of the pedal aperture which originally served for its pro-
trusion.

The mantle is found completely open with only slightly developed anal and
branchial apertures or none at all, in bivalves which do not burrow, but live
surrounded by water, either attached to the bottom or lying freely on it.

&

vii MOLLUSCA— INTEGUMENT, MANTLE, VISCERAL DOME 53

In such animals the surrounding water can circulate through the usually open
mantle cleft and the mantle cavity. We here find protuberances, papillae, tentacles,
etc., carrying sensory organs, all along the free edges of the mantle, whereas, in
bivalves which inhabit mud or bore into wood, rock, etc., such organs are mostly
found massed together round the edges of the branchial and anal apertures.

The Edge of the Mantle.

The edge of the mantle often forms a number of diverging folds, which in trans-
verse section look like finger-shaped processes. The outermost fold is always applied
to the. shell. The edge of the mantle may also be beset with one or more rows of pro-
tuberances, papillfe, or tentacles, and often contains unicellular or multicellular
glands, mucous glands, and others which have been considered to be poison glands for
protective purposes. Tactile sensory cells are very common. Eyes are rarely developed
on the edge (cf. section on the Sensory Organs).

In the Pectinidce, Spondylidcc, and Limidce, the inner fold of the mantle has a
somewhat broad border, which, when the shell is open, projects from the mantle
towards the median line of the body (Fig. 23, p. 16). The free opposite edges of
these folds (flaps, or curtains), springing from right and left, may meet in such a
way as to shut off the central part of the mantle cavity even while the shell is open,
apertures only remaining anteriorly and posteriorly.

F. Cephalopoda.

Integument.

The integument of the Cephalopoda consists of an outer cylindrical
epithelium, and a subjacent cutis in the form of thick connective tissue.
In this cutis, not far removed from the epidermis, and above a layer
of connective tissue plates (which are refractive and often shimmer like
silver), there are large pigment cells or ehromatophores which, by their
alternate contraction and expansion, bring about the well-known changes
of colour in these animals.

These ehromatophores are single cells containing yellow, brown, black, violet or
carmine pigment, either as fluid or in small grannies. The layers containing them
are either single or double ; in the latter case, the colour of the pigment in the one
layer of ehromatophores differs from that of the ehromatophores in the other.
Radial fibres, arising from the surrounding connective tissue, are attached to each
chroniatophore, round that equator which lies parallel to the integument. The
ehromatophores are enveloped in a special, possibly elastic, membrane, and when
contracted are almost globular ; the pigment corpuscles are then crowded together.
The ehromatophores expand equatorially, diminishing the distance between their
poles, i.e. they become much flattened. In this condition, they may, further, throw
out fine branches, the pigment granules being thus spread out over a large surface.
It was formerly believed that the expansion of the ehromatophores was caused by the
contraction of the radial fibres, which were thought to be muscular, but more recent
investigations have shown the fibres to be of the nature of connective tissue. The

o

changes of colour, which are of great physiological and biological interest, and which
are partially under the control of the animal, are brought about by the alternate
contraction and expansion of these variously coloured ehromatophores

54 COMPARATIVE ANATOMY CHAP.

Mantle, Visceral Dome.

Some of the most important points connected with the mantle and visceral dome
have already been mentioned (pp. 36-38).

In Xii.ittiliix, the body is attached right and left to the inner surface of the
shell of the last or inhabited chamber by powerful muscles, which may make
a slight impression on the shell. Between the points of attachment of these
lateral muscles, the integument of the visceral dome coalesces with the inner surface
of the shell of the inhabited chamber in a narrow circular zone, so that the gas/
enclosed in the upper chambers of the shell cannot escape. While the integu-
ment and mantle beneath this zone of concrescence (i.e. towards the free aperture
of the last chamber) are rough, fleshy, and muscular, the integument of that
portion of the visceral dome which lies above the zone and is applied to the last
septum is delicate and soft. The siphuncle, which arises at the dorsal end of the
visceral dome and passes through all the septa, is membraneous and hollow and filled
with blood. It is said to communicate with the pericardium. In the female j\7/i/-
tilus, the nidamental gland (see Genital Organs, p. 241) lies in the free mantle fold,
near the point at which it separates from the visceral dome. We thus have parts
which usually lie in the visceral dome wandering into the mantle fold.

Among the DibrancJiin, which are good swimmers, fins are found. In the
Octopodfi, which are distinguished by the round, compact form of the visceral dome,
these are wanting, except in the remarkable genus Cirrhotewthis. Fins are universal
among the Dccapoda, and vary much in form, size, and arrangement.

In Sepia (Fig. 80, p. 83) and Sepioteuthis, the fins are inserted on the lateral
edges of the body, along the whole height (length) of the visceral dome, forming the
boundary between the anterior and posterior (physiologically the dorsal and ventral)
surfaces of the latter. In Rossiu, Sepivhi, and SfinoJindfu. they are almost
semicircular, and are like distinct appendages situated on the anterior surface of
the dome, about half- way up it. This is also the case in Cirrhnti'uthis, where the
more or less circular fin-lobes rise from the body on stalk-like bases.

The triangular or semicircular fins of Cmiiclin, Histiotnithis, Onychoteuthis, Loligo
(Fig. 34, p. 23), Loligopsis, Ommastrephes, etc., are found at the dorsal end of the
visceral dome, on its anterior side.

In many Dibranchia, there is a concrescence of the free edge of the mantle fold
with the integument of the "head" (Kopffuss), which lies below it. This connec-
tion is effected by means of a muscular band, which passes over the neck (nuchal
band). In most Dempwlii, this connection is wanting, and the edge of the mantle
is free all round the body ; the exceptions are the genera Sepiolit, Cn'in-liiu, and
Loligopsis, which have a narrow connection of this sort. All Odopoda have this
concrescence, commencing with the A /•i/m/n uta ; it lengthens in Philonexis and
Octopus, till in Cirrlioteuthis it spreads to the posterior surface (physiologically the
ventral surface), so that the edge of the mantle remains free only at the aperture
through which the funnel or siphon is protruded.

Arrangements for fastening the mantle fold to the adjacent
body wall are very common. Such attachment is either temporary
or permanent. In the former case, there are prominences with cor-
responding depressions for locking the mantle (appareil de resistance) ;
in the latter case, dermal or muscular fusions take place between the
mantle and body wall.

1 Cf. note, p. 37.

vn MOLLUSCA—THE SHELL 55

1 . Apparatus for locking the mantle. — These are paired or unpaired. The former
are to be found on the posterior side of the body, in the mantle cavity, near its
lower end ; they lie to the right and left at the base of the funnel, and on the
corresponding points of the inner surface of -the mantle fold. The unpaired, on
the contrary, are found on the anterior surface of the neck. Since all the arrange-
ments serve the purpose of cutting off the mantle cavity from the external medium,
it is easy to see that their development is in inverse ratio to the extent of the con-
crescence of the edge of the mantle round the neck before mentioned. Where no
concrescence is found, as in Sepia, the arrangements for locking the mantle are
highly developed ; while, where the line of concrescence is very long, as in Octopus,
the locking apparatus may be altogether wanting. The locking apparatus consists,
in general, of cartilaginous prominences (often accompanied by depressions) on the
inner surface of the mantle fold, i.e. the surface turned towards the mantle cavity,
which exactly fit corresponding cartilaginous depressions accompanied, as the case
may be, by prominences, on the opposite body wall (of. Fig. 80). The special forms
of the mantle and nuchal locking cartilages are of importance in classification.

The cartilaginous arrangements for locking, which are almost always found in the
Dccapoda (they are wanting only in Owen-ia and Cranchia), are still retained in a
few Octopoda in the form of more or less modified fleshy processes (Argonauta,
Trcmoctopus). The nuchal locking apparatus is the first to disappear on the rise of
the pallio-nuchal concrescence. It has disappeared among the Dccapoda in the
genus Scpiola, where the mantle is firmly attached to the neck.

2. Permanent connections between the mantle fold and the adjacent body wall
traversing the mantle cavity are found only in those Cephalopods which have no
locking apparatus. Thus in Octopus and Eledone the mantle is attached to the body
wall by means of a median muscle above the funnel. This muscle consists of two
closely -applied lamella?, having the anus between them. In Cranchia the free
dorsal edge of the funnel (at its so-called base) has become united by an integu-
mental band on the right and left with the mantle fold, and a similar arrangement
is found in Loligopsis.

Water pores. — Near the mouth, or at the bases of the arms, or laterally on the
head, in many Cephalopods, there are apertures leading to integumental pouches of
varying size. The function of these organs is unknown.

IV. The Shell.

General.

The Shape of the Shell, and its Relation to the Soft Body.

All the various forms of shell found in the Mollusca are deducible
from a cup- or plate -like shell covering the dorsal region. Such a
shell affords sufficient protection for animals such as Fisswella, Patrllti,
etc., which can firmly and almost immovably attach themselves to a
hard surface by the sucker-like foot. The soft body is in this case
protected on one side by the shell, and on the other by the surface
of attachment. Free -moving Mollusca, however, show a tendency
to protect the whole body exclusively by means of their shells, and
this object is attained in various ways.

In the Chitonida', for instance, the shell is made up of consecutive

56 COMPARATIVE ANATOMY CHAP.

joints, overlapping in such a way as to be movable one upon the other.
This segmented shell can protect the whole body, since it alloAvs the
Chiton to roll up like an Armadillo or a Wood-louse.

In the Lamellibranchia, the protection of the whole of the soft body
is provided for by the development of a bivalve shell, from which the
foot can be protruded, and which, by the closing of its two valves,
completely envelops the soft body as well as the retracted foot.

In the Gastropoda, Seaphopoda, and Cephalopoda, the most complete
protection on all sides of the body by means of the shell is attained
on another plan. The shell becomes much elongated and turret-
like, and is thus so capacious that not only the visceral dome but the
head and foot also can find place in it. Even the only remaining
unprotected aperture, the one weak point of this fortification, can very
often be completely closed by a hard operculum.

A long, turret-like shell is an inconvenient burden for a freely
moving animal, being, in consequence of its large surface, a hindrance
to locomotion. A reduction of the surface is brought about in the
Gastropoda and Cephalopoda by the coiling of the shell, either in one
plane or in a conical spiral.

In the latter case the spiral twist is almost always right-handed or
dextral.

In order to decide the direction of the twist, the shell should be held in such a
manner that the point of the spiral is uppermost, while the aperture is directed
downwards and towards the observer (Fig. 60, p. 60). If, in this position, the
aperture lies on the right of the axis, the shell has a dextral twist if to the left its
twist is left-handed or sinistral.

We have a striking and in most cases unexplained phenomenon in
the reduction and even complete disappearance of the shell, which takes
place not only in nearly all the classes, but even within some of the
smaller groups of Molluscs, e.g. the Solenogastres among the Amphineura,
a few Heteropoda and Titisctnria among the Prosobranchia, many Pul-
monata, very many Opisthobranchia, and most extant Cephalopoda.

In almost all cases the forms in which the shell is rudimentary or
wanting can be shown to be derived from forms in which it is well
developed. All shell-less snails (slugs) have shells in the early stages
of their development.

The process of the gradual reduction of the shell to a rudiment,
which will be more fully described later on, is often as follows : (1)
the shell becomes internal ; (2) it decreases in size, so that it no
longer can cover the body ; (3) the visceral dome disappears ; (4) the
shell is only to be found in the form of isolated calcareous particles
in the dorsal integument ; (5) even these vanish, and the shell is only
to be found in the embryo.

Only in a few cases is it possible distinctly to recognise the reason
or the advantage of this reduction of a protective covering so useful
to and exercising so profound an influence on the organisation of the

VII

MOLLUSGA—THE SHELL 57

whole race. The following are a few cases in which the utility of
the reduction of the shell in adaptation to special conditions is to
some extent evident: (1) In free-swimming marine forms, where the
shell is too heavy and increases friction ; (2) in Testacella and allied
forms, which prey upon earthworms, where a large shell would prevent
them from following their prey into narrow holes and passages ;
(3) in Gastropods, which browse among thick tangles of Corals,
Bryozoa, Hydroida, or Algse (e.y. many Nudibranchid).

The loss of the shell is generally followed by compensatory
adaptations for protection, such as great capacity for regeneration,
especially of the easily detachable appendages, voluntary amputation of
portions of the body, stinging cells, and colouring which may be
protective in various ways.

The carnivorous Cephalopods are protected (1) by their extraordinary
swimming powers, which are in keeping with their highly developed
organisation ; (2) by their well-developed sight ; (3) by great muscular
strength ; (4) by strong jaws ; (5) by the discharge of the secretion
of the ink-bag; (6) by their partly mimetic changes of colour, etc.

Certain peculiarities of organisation, which can only be under-
stood as remains of a shelled condition (e.g. the lateral position of the
genital and renal apertures and also to some extent of the anus in the
Nudibranchia), always persist after the shell has disappeared.

Chemical Composition of the Shell.

The shell of the Mollusca consists principally of carbonate of lime, with traces
of phosphate of lime and of an organic substance related to chit-in, — conehyolin.
Besides these, various colouring matters may occur.

Structure of the Shell.

The shell of the LamelHbranchm consists of three layers, the innermost layer
being applied to the surface of the mantle. The shell is to be regarded as a
cuticular structure.

The outer layer (shell-integument, epidermis, cuticle, periostracum) is, so far a.s
its physical constitution is concerned, horny and wanting in lime. It generally
disappears off the older portions of the shell.

The middle layer (columnar, prismatic, or. porcelanous layer) consists of slender
prisms of carbonate of lime, usually perpendicular to the surface of the shell and
closely crowded together.

The inner (nacreous) layer has a finely lamellated structure. The very delicate
transparent lamime of which it is composed are thrown into slight waves ; these
cause the wavy lines on that surface of the shell which lies on the mantle, which,
by interference, produce the characteristic nacreous lustre. The pearls of the pearl
oyster are formed of the same substance as this layer.

The constitution of these three layers varies greatly in details both in the Lamelli-
branchia and in other Mollusca. The outer and middle layers are formed at the free
margin of the mantle, the inner layer is yielded by the epithelium of its whole outer
surface.

The shell in the Gastropoda and Cephalopoda consists principally of the middle

58 COMPARATIVE ANATOMY CHAP.

or porcelain layer, which, however, has a structure very different from that of the
same layer in the Lamcllibranchia. This layer is generally, if not always (at least
in the young), covered by a periostracum. The inner (nacreous) layer is very often
wanting.

Growth of the Shell.

In the Arthropoda, the chitinous exoskeleton, which we may
compare with the Molluscan shell, develops at the surface of the
whole body and its appendages. This skeleton, when once formed
and hardened, encases the body on all sides within fixed boundaries,
and is incapable of growth. Hence the moults of the Arthropoda, by
which alone growth becomes possible.

The Molluscan shell, on the contrary, is open. In the Gastropoda
and Cephalopoda, it assumes the shape of a conical mantle, coiled round
a single axis and open at the base of the cone. By continual additions
at the edge of its aperture, it grows with the growth of the animal,
without materially altering its form. The lines on the surface of the
shell of the adult snail register its phases of growth. During growth,
the oldest, uppermost coils or whorls of the shell either continue to be
filled by the apex of the visceral dome (in many Ga4roi><><lx\ or are
deserted by the animal which, as the shell grows, withdraws farther
and farther from its tip. These whorls may remain empty, or may be
partially or completely filled with shell substance. In the latter case,
they may be successively thrown off. The Nautilus and allied forms,
during growth, periodically form transverse septa, so that the forsaken
parts of the shell become chambered and filled with gas,1 the animal
occupying the largest and last-formed chamber, which opens externally.
In the LameUibranchia, the growth of the shell keeps pace with the
growth of the body in exactly the same manner, the free edge of the
shell valve continually receiving additions of shell substance from the
edge of the mantle to form the periostracum and the prismatic layers,
while the whole external surface of the mantle yields an additional
nacreous layer. The consecutive phases of growth are here also
registered by the concentric markings on the surface of the shell.

Special.
A. Amphineura. (Cf. pp. 39-42.)

B. Gastropoda.

A few details concerning the shell of the Gastropods must here he added. As a
rule, the shell is coiled spirally round an axis. This spiral is, in rare instances, so
flattened that the coils come to lie almost in one plane, giving rise to a nearly
symmetrical shell (e.g. Planorbis).

There are, however, among the Gastropoda, uncoiled shells which are symmetrical,
and these require special attention. The most important are the cup-shaped or more
or less bluntly conical shells of the Pati'lli<l«' and Fissurellidce. Since (1) we derive

1 Cf. note, p. 37.

VII

MOLLUSGA—THE SHELL

59

the Gastropoda from bilaterally-symmetrical ancestors with symmetrical shells ; and
since (2) the Fissurellidic undoubtedly possess the most primitive organisation of all
the Gastropoda, and thus stand nearest to the racial form, and are moreover (3)
strikingly symmetrical in their organisation, it seems, at first sight, natural to
consider this symmetry a primitive feature. Certain peculiarities of the nervous
system, however, especially the crossing of the pleuro-visceral connectives, taken in
connection with other conditions explained more fully elsewhere, make it certain
that the cup-shaped shell of Fissurella is only secondarily symmetrical, i.e. that
Fissurella is descended from forms which possessed a spirally coiled shell. The same
is the case with the PateUidce.

The following important facts are in harmony with this conclusion : (1) the
young shell of Fissurella. is asymmetrical and coiled, and it only gradually assumes

PIG. 59. — Shells of— A, Pleurotomaria ; B, Polytremaria ; C ami E, Emarginula ; D, Haliotis ;
F, Fissurella ; G- and H, stages in the development of the shell of Fissurella ; I, shell of the
twisted racial form of the Gastropoda with marginal cleft ; K, the same with apical aperture ;
L, shell of Lamellibranch ; M, shell of Dentalium, seen from the apical aperture. The holes and
clefts of the shells are black ; o, mouth ; a, anus ; ct, ctenulium.

the symmetrical form (Fig. 59, G, H) ; (2) the apparently symmetrical shape of some
forms nearly related to FissurcUa and Patella prove on closer inspection to lie
somewhat asymmetrical, the apex especially being more or less excentric ; (3)
other forms nearly allied to Fissurella, such as Haliotis, Scissurella, and Plcuro-
tomaria, have spirally coiled shells (Fig. 59, A, B, C, D).

In the Fissurellid.ee, many Pleurotomariidce, and the Haliotida', i.e. in the most
primitive Gastropods, peculiar and noteworthy perforations of the shell occur, such
as are occasionally found in other divisions. These perforations lie above the slit
in the mantle which is characteristic of this order (ef. p. 43), and they everywhere
establish communication between the mantle cavity and the exterior, especially
needed when the mouth or edge of the shell is closely applied to the object on which
the creature crawls.

60

COM PA EA TI VE ANATOMY

CHAP.

In Scissurclla, Plcurotomaria, and Emarginula, tliere is a median indentation in
the anterior edge of the shell, which corresponds with an incision in the mantle
edge. This is the case in the young Fissurella, but, during further development,
the edge of the shell grows across the incision, so that in the adult animal the
aperture lies near the apex. Beneath it is the anus, placed high up in the mantle
cavity. If such a cleft were to arise at both the anterior and posterior edges, and
to become very deep, a double shell would result comparable with the bivalve shell
of the Lamellibranchia. It is in fact probable that this notching of the shell edge
is of great phylogenetic significance.

In Haliotis we have a row of perforations of the shell, the process of formation
of the perforation in Fissurella being often repeated ; the older apertures are always,
however, closed by shell substance, and the younger only remain open as long as
they lie immediately over the respiratory cavity.

In very many ProsobrancMa (the Siphon inta of earlier writers), there is, at the
columnar edge or lip of the shell, a notch which gives passage to a channel -like fold
of the mantle margin. This channel keeps up communication between the mantle
cavity and exterior, even when the shell is closed by the operculum. Instead of a

A B

FIG. 60.— A, Dextrally twisted ; B, sinistrally twisted shell of Helix pomatia.

notch, a more or less long process or beak may enclose a corresponding process of the
mantle, the siphon. The latter may become a tube by the apposition of its edges.

It has already been mentioned that the shells of most Gastropods are
dextrally twisted. There are, however, a few families, genera, or species in which
the shell has a sinistral twist ; and in some species where the twist is dextral, a few
individuals with sinistral twist may occur, and vice versa. It is a curious fact that
some species, in which the shell has a sinistral twist, show the asymmetry of the
dextral twist in the soft body, whereas, in others, the asymmetry of the soft body
corresponds with the twist of the shell. We shall return to this point.

For details as to the growth of the shell, and the capacity of the animal to
dissolve the shell already formed, both of which are points full of interest, we must
refer to special works on Conchology, as also for detailed descriptions of forms of the
shell and opercula, and differences due to age.

Progressive reduction of the shell occurs in each of the three divisions of
the Gastropoda. In the ProsobrancMa, this has only been observed in marine,
free-swimming Hetcropods and in Titian n in ,• in the Pulmonata, it is much more
common ; and in the Opisthobranchia, so frequent that nearly all the members of
this division have more or less rudimentary shells. Many adult Opisthobranchia
have even lost every trace of a shell (Pteropoda gymnosomata, Nudibranchia, and
most Ascoglossa), although, in their earliest stages at least, they possessed a coiled

vii MOLLUSGA—THE SHELL 61

shell, for the closing of which there may even be an operculum, secreted by the foot,
as in the ProsobrancMa.

The following are some of the principal stages and concomitant phenomena of
the redaction of the shell : (a) The well-developed shell ceases to be large enough to
shelter the whole body, (b) The shell, which becomes thinner and smaller, is dorsally
overgrown, partially or altogether, by extensions of the mantle, (c) As the shell
(which is then either cup- shield- or ear-shaped) becomes continually smaller, the
visceral dome begins to be levelled down, till it no longer rises above the rest of the
body, its contents spreading out to a certain extent over the dorsal surface of the foot.
(d) The external asymmetry of the body passes by degrees, into symmetry, whereas
the internal asymmetry never entirely disappears, (e) The shell is reduced to a
number of isolated calcareous particles in the integument of the flattened visceral
dome. (/) There is at last no trace of a distinct visceral dome ; calcareous particles
are to be found in the dorsal integument of the long and now naked Gastropod.
(g) Even these particles finally disappear.

In connection with the reduction of the shell in Opisthobranchia and Pulmonata,
compare the section on the mantle, pp. 43-48.

The Heteropoda present the following interesting series :—

Atlanta. The shell is very light and thin, but large and spirally coiled (with
au incision at its aperture) ; the animal can entirely withdraw into it, and close
it by means of an operculum developed on the distinct metapodium.

Carinaria. The shell is thin, light, and delicate ; it is cup-shaped, and covers the
large stalked visceral dome, but is incapable of sheltering the long and thick
cylindrical body and foot. There is no operculum.

Pterotrachea. The visceral dome is small, and there is no shell and no operculum.

C. Lamellibranchia.

The two lateral valves of the Lamellibranch shell are connected, at their
dorsal edges, by means of a hinge and a ligament. The ligament counteracts the
muscles of the shell, which will be described later on, and which, by their contraction,
close the shell. It is usually composed of two layers, the inner layer being elastic,
while the outer is not. The outer non-elastic layer passes into the epidermis or
periostracum of the shell. The inner layer of the ligament is elastic and calcareous,
and is often called cartilage, but this is histologically incorrect.

The ligament lies either externally, distinctly seen dorsally between the
prominences of the hinge edges of the valves, or internally, stretched between the
apposed edges of the hinge, which are furnished with depressions for its reception.
These depressions can easily be distinguished from those belonging to the hinge
itself, since the former are alike on the two valves, whereas the furrows and other
depressions belonging to the one face of the hinge correspond with teeth, ridges,
etc., on the opposite face.

When the elastic "cartilage" of the ligament is at rest, as in a dead bivalve, or
when the adductor muscles of the living animal are relaxed, the valves open. When
the adductors contract, the "cartilage" is — apparently in all cases — compressed.
On the other hand, when the adductors are relaxed, the elasticity of the "cartilage "
forces the shell open again (Fig. 61).

The continuity established between the two valves, by means of this dorsal
ligament, causes the Lamellibranch shell to appear to consist, strictly speaking, of
one dorsal piece, developed to the right and left ventrally into two valves. The
constitution of the ligament and hinge are of importance in classification.

We must refer the reader to systematic zoological works for the special forms
taken by the shell, and content ourselves with the following remarks :—

62

COMPARATIVE ANATOMY

CHAP.

The Lamellibranch shell is originally symmetrical, that is to say, the two valves,
apart from the almost invariable asymmetry of the hinge, are exactly alike
(equivalve). This is the case in most of the Lamellibranchia. The two valves
may, however, become unlike, i.e. the shell (and to a much lesser extent, and only
in unimportant details, the soft body also) may become asymmetrical. As far as
we can at present judge, this asymmetry is caused by adaptation to an attached
manner of life.

The left valve of the Oyster is firmly cemented to the surface on which it
rests. This valve is thicker, more convex and spacious, and forms a sort of
basin in which the soft body lies, while the right valve acts rather as a lid, and
is thinner and natter. We have thus an 'upper" (the right) and a "lower" (the
left) valve, but it is hardly necessary to point out that this use of the terms upper
and lower has as little morphological significance as in the Pleuronectidte among the
fishes. The attached valve is sometimes the right, sometimes the left, and this

FIG. in.— Diagram in illustration of the mechanism for opening and closing the Lamelli-
branch shell. 1, 2. 3, The three layers of the shell— 1, prismatic layer ; 2, cuticle or periostraeiini ;
3, nacreous layer. A, Shell closed by the contraction of the adductor muscle (6), by means of
which the elastic inner portion of the ligament (5) is compressed. B, Shell opened by the elastic
pressure of the inner portion of the ligament during the relaxation of the adductor muscle. 4,
Non-elastic outer portion of the ligament, which passes into the periostracum.

variation may occur within one and the same genus (Chama), or even species
(Aether ia).

Besides the above-named, the following bivalves are also attached, and have
dissimilar valves: Spo/idi/lus, Gnjphcca, p. p., E.^'<j,i,-n p. p., and especially the
fossil Nijjpt'rifrs (Budistes), in which the right valve assumes the form of a high cone
attached by its point, while the left looks like a lid. The conical valye has, however,
no corresponding internal cavity, but is almost entirely filled up with shell substance,
so that, in spite of the form of the shell, the space occupied by the animal between
the two valves is very limited.

This same condition is found in certain fossil C/tioncicrn. In JKcquienia, the left
valve is produced spirally and is attached by its point, while the spirally-coiled
flattened right valve covers it like a lid, so that the whole shell closely resembles
a Gastropod shell closed by its operculuni.

There are also free, unattached bivalves with unequal valves, c.y. many
In these animals, however, many peculiarities of organisation, such as

VII

MOLLUSGA—THE XHELL

63

the rudimentary foot, the constitution of the mantle edge, and the absence of siphons,
indicate descent from sedentary forms. In the case of other forms with unequal
valves, however, no such descent can be established.

In Anomia we have an example of an inequivalve bivalve, in which the valve
turned to the surface it rests on is flat and the
upper arched. The lower valve is here the right
one, and takes the exact imprint of the surface
on which it rests, so that, for example, the mark-
ings of the shell of the Pecten or the Oyster, to
which Aiwmia frequently attaches itself, are
exactly reproduced. In this right attached valve
there is a perforation into which a shelly plug,
the calcined byssus, fits ; by means of this, the
animal tixes itself to its substratum. The ex-
planation of this perforation is seen in the course
of development. It commences as a simple notch
at the edge of the shell, as found also in other
bivalves, for the passage of the byssus. By the

further growth of the shell, this notch to a the byssils notch surroimded by the shell
certain extent is grown round, and thus ap- and persisting as a hole (after Morse).
parently travels away from the edge of the shell,

with which, however, it is still really connected (Fig. 62). In related forms (Carotin]
this aperture becomes quite filled up by a homogeneous calcareous mass.

Impressions on the inner surfaces of the shell. — Various organs of the Mollusc,
attached to or adjacent to the inner surface of the shell, leave more or less distinct
impressions on this surface, which are visible when it is empty. These impressions
are of great importance, especially to the palaeontologist, for by their means fairly

FIG. t>2.— Three stages in the develop-
ment of the shell valves of Anomia.
A, Very young shell ; B, older shell with
notch for the byssus ; C, still older shell,

FIG. i)3.— Dimyaria, inner surface of the left shell valve. A, Cytherea chione
B, Lucina Pennsylvanica (IntegripcdliMta) ; 1, impression of the anterior; 2, impression of the
posterior, adductor ; 3, sinus of the pallial line (4) ; 5, ligament.

safe conclusions may be arrived at as to certain points in the organisation of the
soft body which has disappeared.

1. The most distinct impressions are those caused by the adductor muscles.
Where there are two powerful adductor muscles, one anterior and the other posterior
(Dimyaria), there are two impressions in the corresponding parts of the inner surface
of the shell (Fig. 63). In cases where the anterior muscle is rudimentary, while
the posterior is unusually powerful, and has moved anteriorly towards the middle
of the shell (Monomyaria), there is only one large impression (Fig. 64). The anus

64

COMPARATIVE ANATOMY

CHAP.

is always to be found close to the posterior (which in the Monomyaria is the only)

adductor.

2. Parallel to the edge of the shell, and more or less removed from it, we find

on the inner surface of the shell the so-called pallial
line, caused by the muscle fibres which attach the
edge of the mantle to the valves.

The course taken by this line undergoes charac-
teristic modification in such Larnellibranchs as have
siphons ; at the posterior part of the shell it suddenly
bends forward and upward, and then again passes
backward and upward towards the lower edge of the
posterior adductor. The pallial line, in this case,
forms an indentation, leaving a sinus or bay opening
posteriorly, the pallial sinus, which has been utilised
for systematic purposes (Simtpallfata,

FIG. 64.— Monomyarian, internal
surface of a shell valve of Perna
Ephippium. 1, Hinge edge; 2, im-
pression of adductor.

liata, Fig. 63). The sinus marks the line of attach-

ment of the sipho-retractor muscles ; the stronger
these retractors and the better developed the siphons
the larger and clearer is the sinus.

3. The foregoing impressions are the most dis-
tinct and constant, but others may occur as well, caused by the protractors and
retractors of the foot, by the muscles or ligaments which attach the visceral dome
to the shell, etc. ; but these cannot be further described.

In most Lamellibranchia, when the shell is closed, the edges of the two valves
meet exactly, so that the soft body can be entirely enclosed and cut off from the
exterior (closed) shell. There are, however, shells in which, in the closed condition,
the valves gape posteriorly, or, more frequently, both posteriorly and anteriorly
(e.g., M/iidce, Glycymrridcc, Sotenidcc). This is accounted for by the great develop-
ment of the siphons and of the foot, which can only partially (Myidce, Solcnocurtus)
or with difficulty be withdrawn into the shell. Such gaping shells are found in most
boring bivalves, whose shell formation is specially interesting owing to the develop-
ment of accessory valves or calcareous tubes. In this respect Pholas, Pholadidea,
and Jouannetia represent the most important stages in a remarkable series.

The shell of Pholas is elongated longitudinally, and gapes anteriorly and ventrally
for the passage of the short club-shaped foot, and posteriorly for that of the strongly
developed siphons. As many as three accessory valves are developed dorsally
(prosoplax, mesoplax, metaplax).

The shell of Pholadidea somewhat resembles that of Pholas. In the young
animal it gapes anteriorly, as in Pholas, for the passage of the foot. Posteriorly,
each valve is produced into a horny process, which is succeeded by an accessory
piece (siphonoplax), hollowed out like a trough. The siphonoplax of the one valve
often fuses with that of the other to form a single tube for the reception of the
siphons. There are two pieces of prosoplax, while the rneso- and metaplax are
rudimentary. In the adult the boring activity is suspended, and the anterior
opening becomes entirely closed by the secretion of an accessory piece, the callum
(hypoplax). The functionless foot atrophies, and the animal can move no farther
in the substance into which it has bored.

The shell of the adult Jouannctia is much shortened longitudinally, and is
globular, and the animal cannot move in the round hole it has bored for itself in a
block of coral. Any alteration in its position in the hole, which might be fatal to
the animal, is avoided by means of a posterior tongue-like process of the shell,
which, however, only belongs to the right valve. The shell is completely closed
anteriorly, and a foot is wanting (cf. also Figs. 27, 28, p. 19, and 66, p. 67).

vii MOLLUSCA—THE SHELL 65

The adult condition of Juiiannctia is explained by its developmental history.
The shell of the young animal is like the segment of a sphere, whose greatest height
is hardly half of the radius. It covers the dorsal upper portion of the body, its free
edges thus bounding a very wide aperture, which corresponds with the anterior pedal
gape of Pholas.

In this Pholas-st&ge, in fact, Jouannetia really possesses a foot. Twisting the
body about and rasping the stone with the anterior edge of the shell, the animal
excavates a hole, which is spherical in consequence of the shape of its shell. When
this hole is made, new accessory shell material is secreted at the free edge of the
shell ; this forms the "callum," and as the edge of the mantle follows the lines of
excavation, the form of the accessory shell is here (as in Teredo) determined by the
form of the hole, and the sphere of which the original shell was but a segment is
completed.

Setting aside a few related forms (Martesia, Tcredina, Xylojjhaya, Gastrochaena,
and Fistula no), in which the conditions are somewhat similar, we come to the ship-
worm Teredo (Fig. 29, p. 20). This animal has a long tubular mantle which is
produced posteriorly in two long siphons. The body lies at the anterior end of the
mantle. Teredo bores cylindrical passages in wood. The valves of the shell are
very small in comparison with the body ; they take the form of tri-lobate pieces,
which encircle the anterior end of the mantle. This rudimentary shell gapes
anteriorly for the passage of the pestle-shaped foot, and very widely posteriorly.
The mantle further secretes over its whole surface a calcareous tube which lines its
burrow, but which does not fuse with the shell valves. Two small accessory shell-
pieces, the so-called "palettes," lie at the place where the siphons separate. If the
anterior portion of the animal reaches (i.e. if it bores through to) the water, the
calcareous tube is rounded off and closed.

Aspergillum (Brcchitcs, Fig. 30, p. 20, and Fig. 65) and ClavagcUa show similar con-
ditions. In the club-shaped shell, which inserts its anterior thicker end into rock,
shell, coral, or sand, we can distinguish a true and a false shell. The false shell
forms by far the larger portion of the tube, and corresponds with the secreted tube
of Teredo, and with a callum like that of Pholas. The true shell is very small
and lies anteriorly. The two valves of this true but rudimentary shell are, in
Aspergillum, placed saddle-like over the anterior end of the tube, with which they
are firmly fused > (Fig. 30, p. 20). Were they isolated, their gape would be
unusually wide, not only anteriorly and posteriorly, but ventrally. The shell-tube is
open posteriorly, over the apertures of the siphons ; anteriorly, however, it is closed
(in the adult) by means of a disc perforated like the rose of a watering-can, which
corresponds in position with the callum of the Pholttdidcc. The perforations at the
edge of the disc, or even over its whole surface, are sometimes produced into cal-
careous, and at times dichotomously branched tubules. In the middle of the disc
there is sometimes found a narrow slit-like aperture corresponding with the pedal
aperture in the mantle beneath, but this is often wanting. Less frequently, we find
another aperture in the ventral middle line, corresponding with the fourth mantle
aperture above described (p. 51).

Aspergillum buries its anterior end in mud or sand, but its whole organisation,
and especially its shell arrangement, point to a former boring mode of life.

Clavagella, which is nearly related to Aspergillum, bores into rock or the cal-
careous shells of various other animals. The arrangement of its shell differs from
that of Aspcrgilliun chiefly in the somewhat greater size of its true valves, and in
the fusion of only the left valve with the calcareous tube, the right lying free
within that tube.

In the Pkolndidcc, the ligament, which is still found at the hinge, no longer acts
for opening the shell. In consequence of a peculiar arrangement of the anterior
VOL. II F

66

COMPARATIVE ANATOMY

CHAP.

§1 I? s

D « c' %

55 ^
^ ^ 5T "g t 2

"S £ ?•§ M£
^-" a3 d '-'•••««

^3 -S a> „ cs c

» eT^ 1 o P

-"^

01 _ +-> Q; SO -*J

-4^ C3 « ^ — ^

^s '- rt o> "S o

.S5 o .'ft- -c

i:li§

— rt c; O c^r. -

,2 § S s >,^

I -!-!! =
s « I- i 1 1

S 2 § s -5 5

3 ^v " -^ " -—

a -~ ir - °

•g ~ ^" •- o p
S .= -^' f 'i |

I = ~ I M ••

s §
*

!

-U _ -3 S> a)

• - S .. =

>0 A . . P< _, ^,

° •- = - I -^

s 5 i ~ -^ f

•J. SU £ - r—

VII

MOLLUSGA—THE SHELL

adductor, the opening of the shell, such as it is, is brought about by the muscles.
The anterior and upper edges of the valves are bent outward, and to these edges the
anterior muscle is attached. We thus have external instead of internal points

FIG. 66. — Pholas dactylus, right valve, internal aspect (after Egger). 1-2, Axis round which the
valves move upon one another ; 3-4, longitudinal axis of the shell ; 5-8, line connecting the shell
muscles ; 6, anterior muscle ; T, posterior muscle ; 0, rotating point of the valves ; 10, anterior and
upper edge of the shell, which is bent outwards, and to which the muscle 6 is attached ; 6-9, shorter
anterior ; P-7, longer posterior arm of the lever.

of attachment, and the whole shell may be compared to a double -armed lever
acting along the longitudinal axis of the body, its fulcrum being at the point where,
in other bivalves, the hinge is found. When the anterior muscle contracts, the shell
opens posteriorly and ventrally ; when the posterior adductor contracts, the shell
closes (Fig. 66).

D. Cephalopoda.

The Cephalopoda are all to be derived from an ancient fossil form which possessed
a chambered shell, in the last and largest portion of which the animal lived, leaving
the rest of the shell empty, or rather filled with gas (or water) and traversed by
the siphon or siphuncle. Such a shell is now found only in the sole living repre-
sentative of the Tetrabranchia, the Nautilus, an animal of great importance to the
comparative anatomist. Many fossil forms allied to the Nautilus, and grouped
in the order Nautiloidea, possessed such a shell, as did also the Ammonoidea, with
their enormous wealth of forms which, rightly or not, are generally considered to
be nearly related to the Nautiloidea, i.e. to belong to the Tetrabranchia. In nearly
all these animals the shell, when coiled at all, is, unlike the Gastropod shell, coiled
anteriorly or exogastrically.

One group of the Navtiloidea, the Endoceratidce, which includes only very old
forms (Cambrian and Lower Silurian), is distinguished by the fact that the chambers
of its straight shell, which were filled with gas (or water), lay at the side of and not
behind the inhabited chamber. There was no real siphuncle, but the upper end of
the visceral dome, much narrowed by the air chambers, stretched as far as to the
apex of the shell.

In other Nautiloid<.'«, ^the air chambers always lie, as in Nautilus, above the
occupied chamber, and are traversed by a thin membraneous siphuncle, which, how-
ever, in old forms, is much thicker, and represented the narrow prolonged portion of
the visceral dome (Fig. 32, p. 22).

Some forms of Nautiluidi'n have shells coiled endogastrically ; this is never the
case, however, when the shell forms a complete spiral. The sutures, which corre-
spond with the lines of insertion of the septa, are simple in the Nautiloidea, as

68 COMPARATIVE ANATOMY CHAP.

compared with those in the Ammonoidea, in which they are folded in a complicated
manner.

Nautiloidea. — In the following table we have the chief forms of the shell among
the Nautiloidea : J—

(a) Orthoceras group. — Shell straight or slightly bent. Silurian — Trias.

(b) Cyrtoccras group. — Shell curved like a horn, but not regularly spirally

coiled. Cambrian — Permian.

(c) fhjroceras group. — Shell regularly spirally coiled, the coils, however, not

touching each other. Silurian — Permian.

(d) Nautilus group. — Shell regularly spirally coiled, the coils touching, or the

outer clasping the inner. Silurian — recent.

(e) Lituites. — Shell at first regularly spirally coiled, straightening later.

Silurian.

The siphuncle runs either through the centre of the septa, or through their
anterior or posterior sides.

Ammonoidea. — The shells of the (fossil) Ammonoidea are distinguished by very
complicated sutures, their zigzag lines are like the outlines of sharply-indented leaves
or richly-branched mosses, they are diie to the extraordinary folding of the edges of
the septa, which are attached to the inner surface of the shell. The siphuncle is
always very thin in the Ammonoidea, and almost always pierces the septa on the
posterior side.

The following quotation summarises the chief peculiarities in the form of the
Ammonite shell : —

"The shell, as a rule, forms a closed symmetrical spiral, the coils touching or
clasping one another. Some of the oldest forms are straight, or in youth incom-
pletely coiled. In certain groups of the Ammonoidea we find a tendency repeated at
different times (Trias, Jurassic, Chalk) to depart from the close symmetrical spiral,
and to adopt what are called accessory forms. The first step in this process of change
is in most cases the detachment of the occupied chamber from the next inner whorl ;
then, little by little, the inner whorls also separate, though they still remain in the
same plane — the Crioccras stage. Sometimes the shell grows straight for a time,
then becomes hooked — the Ancyloceras and Hamites stages, and, if only the occupied
chamber separates from the coiled part — the Scaphitcs stage. Finally, entirely straight
shells arise in the Baculites stage. Rarely, the coils leave the symmetrical plane and
assume the shape of a snail's shell ; in this case, the shells may be either closely or
loosely coiled, — the Turrilitcs stage."

Dibranchia. — The shells of all known DibmncMa, extinct or recent, are more or
less rudimentary, since they are never capable of sheltering more than a small portion
of the animal. Further, they are always internal, on the anterior side of the visceral
dome, and are overgrown by a fold of the integument. In Spiruht (Fig. 33, p. 23)
alone, the shell is not completely overgrown, a portion at the apex of the visceral
dome remaining uncovered.

The shell of the (fossil) Belemiiites (Fig. 67 C) is straight, conical, and chambered;
the septa are near one another, and are pierced on the posterior or ventral side by
the thread-like siphuncle, which is enclosed in short, calcareous sheaths. The apex
of the shell (phragmocone) is protected by a conical calcareous sheath (rostrum or
guard), the only part usually preserved. The anterior wall of the last chamber
is produced downwards into a broad thin process, the pro-ostracum.

In Spirulirostra (Fig. 67 D), the phragmocone begins to bend posteriorly (endo-
gastrically). The rostrum is triangular and pointed at the top.

1 Steinmann-Doderlein, Elemente der Palaontoloyie, 1890. - Ibid.

VII

MOLLUSCA—THE SHELL

69

In Spirula (E), the shell is coiled spirally and endogastrically. The siphuncle
is thick, and is surrounded along its whole length by septal envelopes. The rostrum
is rudimentary, and there is no pro-ostracum.

Starting again from the Belemnitcs, the modification of the shell may take another
direction. The phragmocone may become smaller and shorter in comparison with
the continually lengthening pro-ostracum (e.g. Ostracoteuthis, F). The rostrum
also may become thinner and smaller. Finally, the shell may be reduced to a very

P>

M

FIG. 67.— A-H., Diagrammatic median sections through the shells of eight extant or fossli
Dibranchia, from the left side. The point of the visceral dome is turned downwards, the posterior
side of the shell is to the left and the anterior to the right (c/. the position of the Cephalopod body,
p. 36). A, Sepia; It, Belosepia (fossil); C, Belemnite (fossil); I), Spirulirostra (fossil); E,
Spirula; F, Ostracoteuthis (fossil); G, Ommastrephes ; II, Loligopsis; ph, chambered shell =
phragmocone ; pr, pro-ostracum ; r, rostrum (guard) ; s, siphuncular canal, or space which con-
tains the siphuncle ; 1, 2, 3, last three septa (the most recent) ; a, anterior wall of the siphuncle ; p,
posterior ; x, anterior edge of the first septal or siphuncular en velope= anterior or posterior edge of
the siphuncular canal.

small hollow cone at the end of a long narrow horny lamella which corresponds
with the pro-ostracum, and is called, in the extant Deeapoda, the gladius or calamus
(or pen) (Loligo, Ommastrephes (G), Onychotcuthis). In Dosidicus, this terminal cone
is almost solid, and in Loligopsis (H) it is nothing more than a thickening at the
upper end of the gladius ; in other Deeapoda, there is no trace of it on the gladius.
In the Octopoda, the shell has completely disappeared.

Again starting from the Belemnite, the shell may develop in a third direction
to form the Se^na shell. The transition form is found in Belosepia (B) (Eocene),

70

COMPARATIVE ANATOMY

CHAP.

-1

that is, if this interpretation is correct. This shell is somewhat bent, the septa
are crowded together and slope downwards anteriorly. They are penetrated
posteriorly by an extremely thick siphon, which is enclosed throughout its whole

length in an envelope with a very thick anterior
wall. The completely enclosed siphuncular
space is thus a wide funnel running through the
chambers of the shell on its posterior side (Fig.
67 13). The phragmocone is enclosed in a thick,
strongly - developed rostrum, and its anterior
and lateral walls are produced downwards into
a broad, posteriorly concave shell (pro-
ostracum ?).

These arrangements seem to have culminated
in the extant Sepia (Figs. 67 A and 68). The
siphuncular space fits over the visceral dome
like a mould. The anterior portions of the
septa slope downward much more obliquely
from behind anteriorly, so that, in a back view
of the shell, the whole area of the last septum is
visit Je at the surface (Fig. 68, 1). The septa
are thin calcareous lamellae, closely superim-
posed one upon the other, with very narrow air
chambers between them ; and these latter are
traversed by perpendicular trabeculre. The shell
is thus very light, its specific gravity is less than
that of water. Behind the siphuncle, on the
posterior very much shortened side of the shell,
the short septa are closely contiguous, without
any intervening air spaces.

The dorsal end of the shell is enclosed in a
small pointed rostrum. The whole anterior sur-
face is covered by a thin lamella of conchyolin,
which projects laterally beyond the edge of the
shell, and is itself covered by a calcareous layer
which is an anterior and ventral extension of
the rostrum.

The female Argonaut is the single exception

Posterior (physiologically ventral) aspect, to the rule stated above, that in the Octopoda
Lettering as in Fig. »37. The last septum the shell has entirely disappeared. This animal
1 is seen in its whole extent ; s, the mouth hag & u ht tMn external shell coiled anteriorly
(if the broad, slipper-shaped siphuncular , . ,, , . , . „ ,

cavity ; /, lateral wall of the cavity ; a-ft or exogastrically, which is not firmly attached
line of the section which in Fig. (57 A is to the body at any point, and serves more for
diagrammatised. The two figures should receiving the eggs (Figs. 35, 36, pp. 24, 25) than
be compared (principal details after for protecting the body. This shell is sur-
rounded and secured by lobate processes of the

anterior pair of arms. It has 110 nacreous layer, but is porcelanous, and is
apparently produced from the integument of the visceral dome and the mantle.
The dorsal pair of arms is said only to deposit the so-called black layer on its
surface.

It is usually considered that this Argonaut shell is not the homologue of the
shell of other Cephalopods, but is a formation peculiar to the Argonaut female. An
opposite view has, however, recently been very ably advanced — that the Argonaut
shell is an Ant inn/life shell which has lost its septa and siphuncle and also its

r

a

FIG. 68. — Shell of

Sepia aouleata.

vii MOLLUSGA—THE PALLIAL COMPLEX 71

nacreous layer.1 Should this view prove correct, the Cephalopods would have to
be differently classified. The division into Tetrdbranchia and DibrancMa would
have to disappear, as we cannot tell whether the fossil Ammonoidea were Tetra-
branchia, and are also ignorant as to when the Dibranchia developed from the Tetra-
branchia. The Cephalopods would then have to be divided into (1) Nautil<ii<l<«
with the extant genus Nautilus ; (2) Ammvimidca with the still living Octopoda ; and
(3) Belcmnoidea with the extant Dccapoda.

Bivalve shelly plates called aptychi have been found sometimes in the last
cha.mber of the Ammonoidea, sometimes isolated. These have been proved to belong
to the bodies of certain species of Ammonoidea, and have been considered by some to
be protectives for the nidamental gland, by others as opercula, and by others again
as the analogues or homologues of the infundibular cartilage of the Deeapoda. No
one of these three views has as yet been generally accepted.

V. Arrangement of the Organs in the Mantle Cavity and of
the Outlets of Inner Organs in that Cavity.

A discussion of this subject at this stage will help to explain the asymmetry of
the ('riia/ni/inda and to simplify the discussion in later chapters.

There are, in the mantle cavity, many important organs crowded together in a
comparatively small space, and into it also open all the apertures of the inner organs
except the oral aperture of the alimentary canal. The term "circum-anal complex."
though especially applicable to the arrangement in the Gastropoda, is not so suitable
as " pallial complex," which applies to nearly all Mollusca, and comprises not only
the pallial organs themselves, but the apertures of inner organs that lie in the
mantle cavity.

The most important constituents of the pallial complex are the ctenidium, the
osphradium (Spengel's organ, olfactory organ, or accessory gill), the hypobranchial
gland, the anus, and frequently the rectum as well, the nephridial apertures and
often the renal organ also, the genital apertures, and frequently the pericardium,
with the enclosed heart.

Starting with the Chiton/idee, which, as has already been described (p. 42),
must be considered as the most primitive of all living Molluscs, we have :—

The median anus, lying at the posterior end of the body in the mantle groove ;
on each side of it anteriorly the nephridial apertures, and again on each side, in
front of these, the genital apertures.

Assuming this to be the primitive arrangement, we have the following important
variations.

A. Gastropoda.
1. Prosobranchia.

a. Diotocardia. — In Fissta'cltit, the pallial complex is still quite symmetrical,
but instead of lying posteriorly, as in Chiton, it, together with the mantle and the
pallial cavity, lies on the front of the visceral dome. We have to imagine that the
whole complex has shifted forward along the right side of the body, so that the gill
originally on the left has come to lie on the right anteriorly, and that originally on
the right now lies anteriorly on the left, and the same applies to the other organs
belonging to the complex.

1 Steimnann, Bericht Freiburg Gesellsch., iv. pp. 113-129.

72

COM PAR A TI \7E A NA TOMY

CHAP.

In order to prevent confusion, the hypothetical original position of each organ
will be denoted by vr (= originally right) and it I ( = originally left) in brackets.

In the upper part of the mantle cavity in Fissurella, beneath the median
aperture in the mantle and shell, lies the anus, and immediately to its right, the
right (ul) nephridial aperture, immediately to its left the left (ur) nephridial
aperture ; the right (ul) and left ()(/•) ctenidia, again, lie symmetrically to the right
and left. There are no distinct osphradia, and the genital apertures are wanting as
the genital gland opens into the right nephridium.

Haliotis. — The mantle cavity has here shifted to the left, and the rectum,
attached to the mantle fold, runs forward some way through it, so that the anus is
at a considerable distance from the posterior apex of the cavity. On the right of the
rectum lies the right (id), and to its left the left (ur) ctenidium, both fastened to
the mantle, and stretching far forward. The right and left nephridial apertures
lie near the bases of the ctenidia, in the upper and posterior part of the mantle,

•m.

IIV

FII

side.

TO.— The same specimen from the left
Lettering as before ; o, mouth.

•n

FIG. 69.— Anterior portion of Patella, from
above, after removal of the mantle fold (after
Ray Lankester). a, Tentacle ; b, foot ; <•,
pedal muscles (shell muscles) ; cl, osphradia ;
e, mantle fold ; /, aperture of the right nephri-
dium ; g, anal papilla and anus ; li, papilla and
aperture of the left nephridium ; /, left nephri-
dium ; /,-, right nephridium ; I, pericardium ;
n, digestive gland (liver); m, cut edge of the
mantle ; p, snout.

cavity. Between the rectum and the left ctenidium, also on the mantle, is found
the long, well-developed hypobranchial gland (mucous gland), which stretches as fat-
forward as the gill. Only a small portion of the gland lies to the right between
the rectum, as far as it runs, and the right ctenidium. There are two osphradia
which run as bands along the axes of the ctenidia facing the mantle cavity.

Turbinidse and Trochidae. — Only the left (ur) ctenidium of Haliotis is here
retained ; it lies far to the left on the roof of the mantle cavity, i.e. on the mantle.
The rectum runs far forward along this roof. Two nephridial apertures lie on
papilke in the base of the cavity, at the sides of the rectum. The hypobranchial
gland is found in various stages of development, the highest being attained in the
Turbinidcc. It is largest between the rectum and ctenidium, i.e. between the
right side of the latter and the left side of the former. In the T-urbinida; however,
a portion of it lies to the right of the rectum. There is a diffuse osphradium on the
axis of the gill.

Neritina.— There is here only one gill (the left (ur) in Hn.liotis) shifted somewhat far
to the right. The rectum lies asymmetrically to the right in the respiratory cavity,

VII

MOLLUSGA—THE PALLIAL COMPLEX

reaching so far forward that the anus is found near the right edge of the mantle cleft.
There is only one nephridial aperture to the left of the base of the ctenidium, far up in
the mantle cavity. The inner surface of the mantle, between the rectum on the right

1Z

PIG. 71.— Pyrula tuba, male, taken out of the shell (after Souleyet). The mantle is cut open
along its base and right side, and laid back to the left ; the position of the pallial organs is thus
reversed. 1, Proboscis; 2, snout ; 3, foot ; 4, penis ; 5, seminal duct, which is continued at 15;
6, floor of the pallial cavity = nuchal integument; 7, columellar muscle; S, intestine; 9, heart in
the opened pericardium ; 10, digestive gland (liver) ; 11, testes ; 12 and 13, renal organs ; 14, renal
aperture ; 15, seminal duct ; Hi, rectum ; 17, hypobranchial gland ; 18, anus ; 10, ctenidium (gill) ;
20, mantle ; 21, osphradium ; 22, respiratory siphon.

and the gill on the left, is glandular, and represents the slightly differentiated hypo-
branchial gland. The genital aperture lies close to the anus.

Docoglossa. — In the Patell-idcc (Figs. 69, 70) a short conical portion of the

74 COMPARATIVE ANATOMY CHAP.

rectum projects into the small mantle cavity. This anal cone is not median, but
is distinctly shifted to the right. To its right and left lie the nephridial apertures,
raised on short conical papillfe. There is no separate genital aperture. In some
forms (Tcctura, Scurrut, Acmtca,} one ctenidium is found attached to the mantle, on
the left side of the pallial cavity. Further details as to the gills in the Patellidrc
will be given later on. We further find, on the floor of the cavity, on each side,
traces of an osphradium in the shape of a small patch of sensory epithelium, which
may be raised on a prominence. It is doubtful if the prominence found in Patella
close to each osphradium, containing a blood sinus divided up by septa, can be
considered as a rudimentary gill. These prominences rise from the floor of the
mantle cavity, whereas in Tectum, for example, in which a true gill still occurs on
the left, it lies far removed from the left osphradium, in the usual position on the
'of the cavity, i.e. on the inner surface of the mantle.

b. Monotocardia. — In this division, the numerous forms of which show little
variety of organisation, the arrangement of the pallial complex is very uniform.
The single genital aperture is always distinct from the single nephridial aperture.
The position of the organs in the spacious pallial cavity (Fig. 71), from right
to left, is as follows :—

1. To the extreme right, lies the afferent duct of the genital organs (ovary or
seminal duct), which runs more or less far forward, in the mantle cavity.

2. In contact with this, but quite on the roof of the cavity, is the rectum.

3. To the left of the rectum, far back in the base of the mantle cavity, lies the
slit-like nephridial aperture, which pierces the wall separating the cavity from the
renal organ behind and above it. Exceptions occur in Palurlinn and Valrat/t, in
which this aperture is shifted forward to the end of a urinary duct which runs on
the mantle.

4. On the roof of the mantle cavity are found the hypobranchial glands (mucous
and purple glands), which are developed in varying degrees.

5. Quite to the left, and also on the roof of the cavity, the ctenidium, feathered
on one side (the left (iir) of ffaliotis and FissurcUn), at whose base, deep back in the
cavity, the pericardium is visible with the ventricle and auricle seen through it.

6. Finally, to the extreme left, lies the osphradium, which is always well
developed and sharply circumscribed, and is either filamentous or feathered on two
sides, and attached to the roof of the pallial cavity.

The position of the organs in the pallial complex of the Heteropoda, certain forms
of which, such as Atlanta, are closely related to the other Monotocardia, requires to
be re-investigated. The osphradium lies at the base of the gill.

2. Pulmonata.

In the Pulmonata, the single or double ($ and 6 ) genital aperture (Fig. 72) no
longer belongs to the pallial complex, but lies outside the mantle cavity laterally on
the head or neck. In Onciilium the male aperture lies anteriorly under the right
tentacle, the female posteriorly, near the anus.

Bearing in mind that the mantle or pulmonary cavity communicates with the
exterior only by means of the respiratory aperture lying on the right, we have the
following arrangement of the pallial complex as typical (excluding such aberrant
forms as DandchtnUa, Ti'Sdict'lla, and Oncii/iiim).

1. On the extreme right of the pulmonary cavity lies the rectum, the anus
opening in the respiratory aperture.

2. On the roof at the back of the cavity lies the nephridium (kidney).

3. To the left, near the kidney, also far up in the cavity, and on its roof, lies

vii MOLLUSGA—THE PALLIAL COMPLEX 75

the pericardium, containing the ventricle and auricle, the latter lying in front of

s

7

Fir.. 72.— Helix aspersa, fully extended from the right (after Howes), a, Anus appearing in the
respiratory aperture, pi? ; s, shell ; p, edge of shell aperture ; </«, genital aperture ; tlt optic tentacle ;
t, anterior tentacle ; ?o, upper lip.

the former. From the ventricle the aortic trunk runs upward and backward, and
from the auricle rises the pulmonary vein,
which runs forward along the roof of the
pulmonary cavity.

4. The respiratory vascular network
spreads over the whole remaining surface
of the roof of the pulmonary cavity, and
is thus in front of the kidney and peri-
cardium.

5. An osphradium has till now only
been found in the Basommatophora (Pl«n-
orbis, Physa, Limnaeus), near the respira-
tory aperture, and among the Stt/lommato-
phora in Tcstticella on the floor of the
pulmonary cavity at its extreme posterior
angle.

The floor of the pulmonary cavity (the
dorsal nuchal integument) is smooth and
devoid of organs.

The arrangement of the efferent ducts
of the renal organ varies and deserves
special description (Fig. 73).

1. The anterior side of the renal sac
opens on a simple papilla in the mantle
cavity (£ulu>iusoblongus,a.nd some species
of Planorbis) (Fig. 73 A).

2. The papilla lengthens and runs for-
ward as a straight ureter (primary ureter).

FK;. 73. — Six diagrams illustrating the
variations in the renal ducts in the Pul-
monata. The organs are supposed to he seen
through the mantle above them. 1, Free edge
nf mantle ; 2, respiratory aperture ; 3, rectum ;

This occurs inmost Basommatophora, and 4- kidney; 5, pericardium; 6, auricle; 7, ven-
tricle ; 8, primary urinary duct ; 0, secondary
urinary duct, which, in D, is a groove. Further

explanations found in the text.

some species of Bulimus, Gionella, Pupa,
Helix (B).

3. The ureter runs backward along the

kidney, and opens at the base of the respiratory cavity. Testacclla, and some forms
of Helix (C).

4. A secondary urinary duct is added, becoming constricted from the wall of

76 COMPARATIVE ANATOMY CHAP.

the pulmonary cavity, and at first forming a more or less closed channel along
which the urinary discharge can be forwarded from the base of the cavity to the
respiratory aperture. Some species of Bulimus and Helix (D).

5. The secondary urinary duct becomes closed, and opens either alone or with
the anus into the pulmonary cavity. Some species of Bulimus, Helix, Daudebardia,
Vitrina, Hyalinia, Zonitcs, Anon, etc. (E).

6. The end of the secondary urinary duct and the end of the rectum together
form a cloaca which is distinct from the pulmonary cavity/ and opens close to
the respiratory aperture. Limax, Amalia, and some species of Daudebardia (F).

When the primary urinary duct runs back along the kidney it is externally in-
distinguishable from the substance of the latter, and it thus often appears as if the
duct rose from the posterior end of the renal organ.

The variations which occur in the position of the organs of the pallial complex
in the carnivorous Fulmonata are .specially interesting. In a series of car-
nivorous forms, commencing probably with Hyalinia among the Stylommatophora,
and proceeding through Daudebardia to the extraordinary genus Testacella, we find
progressive diminution of the visceral dome and its displacement to the posterior
end of the body, simplification and diminution of the shell, and further, a shifting
back of the liver and genital organs from the visceral dome into the nuchal portion
of the ccelom, which now is found along the whole length of the dorsal surface of the
foot. Finally, in Testacella and certain Daudebardia, the visceral dome completely
disappears, and the pulmonary cavity covered by the shell is alone left, the cavity
reaching up to the apex of the shell. The floor of this cavity, and indeed the whole
cavity, with the mantle and the shell, sink down into the body. In this way
Tcstucellii, which follows its prey, the earthworm, into its underground passages, is
admirably adapted to its manner of life ; its body is slender, and the somewhat
flat shell at its posterior end, which does not stand out above the surrounding sur-
face of the body, in no way hinders its movements. These alterations, however,
especially the displacement of the visceral dome to the posterior end of the body,
are accompanied by important alterations of position in the pallial organs, which
finally lead to the condition called opisthopneumonic.

It is important to note that concrescence of the mantle and the subjacent dorsal
integument is complete except at the respiratory aperture on the right, and that the
latter shifts farther and farther back, in its relation to the pulmonary cavity, till, in
Testacella, its position is almost terminal.

The first important step in the displacement of the pallial organs is seen in
Daudebardia rufa. The pericardium, instead of lying far back at the base of the
pulmonary cavity, here lies far forward on its roof, so that by far the greater portion
of the vascularised pulmonary tissue lies on the roof behind the pericardium (Fig.
74 A). Daudebardia rufa is thus actually opisthopneumonic. But in this case the
relative position of the ventricle and auricle is still unaltered. The auricle is, as
before, placed in front of the ventricle ; the pulmonary vein from the auricle is thus
obliged to bend round in order to run backward, while the aorta, which becomes
almost exclusively the anterior or cephalic artery, supplying that portion of the body
which lies in front of the visceral dome (by far the greatest part), must bend forward
from the ventricle.

In another Dand/b/irdia, D. snulryi, the case is somewhat similar, but the
kidney and pericardium together form a sort of sac which hangs down into the pul-
monary cavity from its roof. In this sac, the ureter lies dorsally and the peri-
cardium ventrally to the kidney. The floor of the cavity sinks right and left deep
into the subjacent region of the body.

If we imagine that the pulmonary vein which runs back from the anteriorly

VII

MOLLUSCA—THE PALLIAL COMPLEX

77

placed auricle, and the aorta which runs forward from the chamber lying behind the
auricle have pulled these chambers round in such a way that the flow of blood can
have a straight course (cf. diagram, Fig. 74), the ventricle will then come to lie
in front of the auricle. Indeed, the pericardium (with the ventricle and auricle)
has actually twisted round 180°. In this twisting it has been followed by the
kidney, which is connected with it by the reno-pericardial aperture, so that the
latter organ no longer lies to the right but to the left of the pericardium, the aper-
ture of the urinary duct remaining at its former place. The whole reno-pericardial
complex, as compared with its typical position in the Pulmonata, is quite reversed.
This reversal is characteristic of Tcstacrlln.

It is, further, noteworthy that, in Testacella, the floor of the pulmonary cavity
becomes invaginated anteriorly into the body below it to form a large air sac. The
walls of this sac are not supplied with blood vessels, and it seems to serve merely as
a reservoir of air. In many Testaccllidcc the reno-pericardial complex hangs down in
the shape of a sac into this air sac from the roof of the pulmonary cavity.

In the Vaginulidce and the Oncidia the arrangement of the organs, originally
belonging to the pallial complex, deviates still further from the type. A shell is

FIG. 74. -Diagrams to illustrate the changes of position in the pallial organs of Daude-
bardia and Testacella (adapted from figures by Plate). Mantle organs drawn as in Fig. 73. A,
Daudebardia rufa : B, Hypothetical stage, the pallial complex of A twisted round 00° ; C,
Testacella. 1, Respiratory aperture ; 2, kidney ; 3, ureter or urinary duct ; 4, reno-pericardial
aperture (renal funnel) ; 5, ventricle ; 6, auricle ; 7, aorta ; 8, pulmonary vein ; 9, pulmonary
vascular network.

wanting in the adult and a mantle also ; and the mantle- or pulmonary cavity
seems in consequence to have atrophied. The pericardium lies posteriorly to
the right, sunk into the integument, the ventricle lying, as in TcstaccHa, in front
of the auricle. Respiration takes place principally through the skin ; in the amphib-
ious Oncidia it is assisted by dorsal papillae. In Vaginulus, the urinary duct joins
the proctodaeum to form a cloaca which somewhat widens at the point of junction,
and opens externally at the posterior part of the body. The same is the case in
most Oncidia, but in Oncidium cdticum, the urinary duct and the rectum emerge
separately, but one close to the other, at the posterior end of the body. Close to
these apertures lies, in all cases, the female genital aperture ; the male aperture,
however, lies anteriorly to the right below the tentacle.

The cloaca just mentioned, which is tilled with air, has given rise to interesting
discussions. From its wall there rise into the lumen closely packed folds, which
may also be continued along the posterior portion of the urinary duct. The cloaca
has therefore been considered by some to be a rudimentary pulmonary cavity, into
which the urinary duct and the rectum open. The present writer holds the opinion,

78

COMPARATIVE ANATOMY

CHAP.

provisionally, that this cloaca has arisen by the junction of the terminal portions of
the secondary ureter with the rectum, as in other Pulmonata, but that here the pul-
monary cavity having atrophied, it opens outward direct, i.e. no longer through a
respiratory aperture. Others, again, have thought the arrangement in Oncidiinn
and T"aginulu-s to be primitive, the pulmonary cavity appearing here first as an
insignificant widening of the terminal portion of the primary ureter.

If this were the case, then the condition described above (p. 75. 1) for BiiUiimx
oblongus, where the kidney opens on a papilla direct into the base of the pulmonary
cavity, would be thus explained : the pulmonary cavity would have to be considered
as a much widened primary urinary duct. Then, in this primary ureter (pulmonary
cavity) would follow the successive stages of the development of the secondary
ureter, at first an open and later a partially closed channel, and finally a closed
tube, so that at last, as in Helix pomatia, the primary ureter is divided into two
distinct portions, viz. the much widened pulmonary cavity and the secondary
ureter. But in the Li miuvidce, for example, the pulmonary cavity admittedly
corresponds with the mantle cavity of other Gastropods. The Pulmonata would
thus fall into two groups, the Nepltropucusta (Stylommatophora), in which the
pulmonary cavity = the widened primary ureter, and the Branchiopneusta (Basum-
matophora, p. parte), in which the pulmonary cavity = the mantle cavity of other
Gastropods.

We consider this view incorrect because of the uniformity of the whole organisa-
tion in the Pulmonata, and especially because of the occurrence of an osphradium in
the pulmonary cavity of a Stylommatophore (Nephropneusta), viz. in the genus
Tcstacella. For the osphradium invariably belongs to the mantle cavity, being
primitively connected with the ctenidium, it never lies in the urinary duct.

3. Gastropoda Opisthobranchiata.

We can here speak of a pallial complex only in connection with the Tcdibranchia,
since in them alone is a distinct mantle fold developed on the right side of the

FIG. 75. — Aplysia, right aspect, the right parapodium (15) turned downwards; the pallial
complex is seen under the mantle fold 7 (after Lankester). 1, Anterior tentacle ; 2, eyes ; 3,
posterior tentacle (rhinophore) ; 4, left parapodium ; 5, seminal furrow; fi, genital aperture; 7,
mantle fold ; 8, gland ; !>, osphradium ; 10, outline of some inner organ seen through the integument :
11, nephridial aperture; 12, ctenidium; 13, anus; 15, right parapodium; 16, anterior portimi
of the foot. (There should be no connecting line between C. and '.'.)

body. The general order of the organs in the pallial cavity (Fig. 75) is as follows : —

1. Far back, and often hardly or not at all covered by the mantle, sometimes at the

summit of a conical prominence, lies the anus, and near it occasionally an anal gland.

VII

MOLLUSC A— THE PALLIAL COMPLEX

79

8

2. In front of the anus, between it and the ctenidium, is the nephridial aperture.
Following these there may be—

3. A hypobranchial gland.

4. The ctenidium.

5. At the base of the ctenidium or on its axis,
the osphradium.

Were this complex of organs to be shifted along
the edge of the body, we should have the arrange-
ment found in the Monotocardia among the Pr<»><>-
bmnchia. The correspondence is, however, appar-
ently marred by the position of—

6. The genital aperture, which in the Opistho-
branchia lies farthest forward of all the pallial
organs.

In all other Opisthobranchia (after excluding the
Ti'ctibmncliia] the pallial complex is broken up
when the mantle and the true ctenidium disappear.
The only exception to this is found in the Phyl-
Hiliiilcc, where, apart from the gills, a similar
arrangement to that in the Tectibranchia occurs.
The single or paired genital aperture always lies
asymmetrically on the right side in front of the
anus, which is sometimes found asymmetrically on
the right side, and sometimes has a median dorsal
position between the middle and the posterior end
of the body. The renal aperture lies between the
anus and the genital aperture, sometimes close to
the latter.

In the Pti'i'Dpiiilii yipiirtosoiniitn (Fig. 76} the
shell and mantle are wanting. The ctenidium,
when retained, as in the Dexiobranchia and Piifiino-
ili'/'uKt, lies somewhat far back on the right side of
the body, far behind the anus. On the disappear-
ance of the mantle, it evidently shifted back front
its original position between the anus and the genital
aperture, while the osphradium, which is generally

found close to the ctenidium, has, as far as has yet
been observed, retained its original position.

The. anus lies anteriorly behind the right tin ;

the nephridial aperture lies close by, either distinct

or united with the anus at the base of a common

cloacal depression. Immediately in front of this

lies the osphradium, then follows, considerably

farther forward on the neck, to the right behind

the base of the right fin, the genital aperture, from

which, as in many Tectibranchm, a ciliated furrow

17

FIG. 76.— Pneumoderma, from tin-
right side, diagrammatic (after Pel-
seneer). 1, Right process bearing
hooks (Hakensack) evaginated ; '2,
proboscis ; 3, right buccal tentacle ;

4, position of the right nuchal ten-
tacle ; 5, right fin (paru podium) ; 6,
seminal furrow; 7, genital aperture;

5, position of the jaw ; f>, ventral pn>-
boscidal papilla ; 10, right buccal ap-
pendage bearing suckers; 11, head;
12, aperture of penis ; 13, right anterior
pedal lobe; 14, anus; 15, posterior
pedal lobe; 16, ctenidium; 17, pos-
terior adaptive gill ; <', c, », y.
ventral, anterior, posterior.

runs forward along the surface of the body to the

aperture of the penis, which lies to the right in front of the foot.

All ThfcoKiniiiifa have a mantle and a mantle cavity, and often a shell as well ;
in the Cymbuliidce, the latter is replaced by a cartilaginous pseudoconch, a sub-
cutaneous formation of the mantle.

Among the Thccosouu/ta, the Liin<n-iii!iln: indicate the primitive arrangement ;
they possess a dorsal or anterior mantle cavity, a coiled shell, and an operculum.

80

COMPARATIVE ANATOMY

CHAP.

The ctenidium, however, is wanting. In the base of the pallial cavity, to the left,
lies the pericardium, and immediately in front of it the kidney, with a narrow
aperture into the cavity ; then follow the osphradium (where this has been found),
and, at the extreme right of the cavity, the anus with the anal gland. The mantle
gland (hypobranchial gland, shield) is found on the roof of the pallial cavity. The
genital aperture lies to the right anteriorly in the cephalic region ; from it a
ciliated channel or furrow runs dorsally to the aperture of the penis, which lies
anteriorly between the fins.

As compared with the Linmcinidce, i.e. the Tliccosomata with coiled shell, the

— 7

Fie. 77.— A, B, C, Three diagrams to illustrate the relation of the Limacinidae to the
Cavoliniidae (after Boas). A, Limacinidse ; B, hypothetical intermediate stage between
the Limacinidae and the Cavoliniidae. The visceral dome twisted 90°. C, Cavoliniidae. All the
diagrams from the ventral or posterior side. In A the visceral dome is drawn straight, whereas it
is in reality coiled. 1, Right fin (parapodium) ; 2, foot bent forward ; 3, genital aperture ; 4, ten-
tacular appendage of the mantle edge ; 5, anus ; 6, masticatory stomach ; 7, gonad.

CavoliniidcK and Cymlndiidce, or Thccoso/imta with straight shell, show a very
different arrangement of the pallial complex, which can only be explained by the
supposition that the larger posterior portion of the body (the visceral dome) ot
the Limacinidae^ with all the pallial organs belonging to it, lias twisted round the
longitudinal axis of the body 180°, in relation to the cephalic region with the
genital apertures belonging to it. Such a twist gives the organs the position they
actually occupy in the Cavoliniulac and Cymbuliidce ; the posterior (ventral) pallial
cavity containing, on the left the anus, on the right the pericardium and kidney and
the osphradium, the genital aperture occupying its original position to the right.
The cause and significance of this twist are at present unknown.

B. Scaphopoda.

There is no gill in the posteriorly placed mantle cavity. The anus lies in the
middle line above the foot, having a nephridial aperture on each side of it. There
are no distinct genital apertures.

vii MOLLUSGA—THE PALLIAL COMPLEX 81

C. Lamellibranchia.

The general arrangement of the organs in the mantle cavity of the Lamelli-
branchia has already been described. The strict symmetry of the body in this
class must again be pointed out. All originally paired organs remain paired and
symmetrical.

The two nephridial apertures lie on the body above the base of the foot, or
farther back near the posterior adductor muscle ; they usually lie beneath the point
of attachment of the gill-axis, between it and the line of concrescence of the (inner)
ascending lamella of the branchial leaf with the foot, where such concrescence takes
place. In the Scptibranchia, on the contrary, the apertures open into the upper
pallial chamber.

The outer genital apertures may be wanting, and in this case the genital
products are ejected through the nephridial apertures, which is the primitive
arrangement. When present, in diceceous bivalves, they are always found in one
pair, and lie on each side just in front of the nephridial apertures, sometimes in the
base of a common pit or furrow, less frequently at some distance from these aper-
tures. There are no special copulatory organs.

In hermaphrodite Lamellibranchia the arrangements may vary as follows : —

1. Both kinds of sexual products may be ejected on each side through a
common aperture (Ostrcca, Pectcn, Cydas, Pisidium, etc.).

2. There may be, on each side, two distinct apertures, one male and the other
female (Anatinacea).

3. The seminal ducts and the oviducts may unite before opening to form a
short, common, terminal piece (Septibranchia).

The osphradium is paired in the Lamellibranchia, and always lies near the
posterior adductor muscle over the visceral ganglion, at the point of insertion of the
branchial axis on the body. A pair of sensory organs is found in many Lamelli-
branchia, one on each side of the anus (abdominal sensory organs), or to the right
and left on the mantle at the inner aperture of the siphons of the Siphoniata (pallial
sensory organs).

Hypobranchial glands have been found in the Protobranchia (Nuculidcc and
Solenoiiujidce}. They are large and well developed, and belong to the mantle, lying
in the posterior part of the body above the base of the gill on each side, to the right
and left of the pericardium, and in front of the posterior adductor.

The leaf-like oral lobes (labial palps), one occurring on each side of the mouth,
between it and the anterior end of the base of the gill, will be described more in
detail in another place.

D. Cephalopoda.

In the Cephalopoda the primitive symmetry of the pallial complex is on the
whole retained.

If we cut open the mantle of the Nautilus (Figs. 78 and 79), which covers the
posteriorly placed pallial cavity, and lay it back on all sides, the following organs
are revealed : —

1. On each side there are two gills, an upper and a lower.

•2. The anus lies on the visceral dome, between the bases of the four gills.

3. Below the base of each gill is found a nephridial aperture — making four in all.

•i. Close to the two upper nephridial apertures lie the two so-called viscero-
pericardial apertures.

5. Between the bases of the lower gills there are in each sex, two genital
VOL. II G

82

COMPARATIVE ANATOMY

CHAP.

11

ill

FIG. 78.— Pallial complex and siphon of Nautilus pompilius ? (after Bourne and Lankester).
v, Valve of the siphon ; ro, right genital aperture ; m, the mantle fold, with the nidamental gland,
folded back ; an, anus ; cp, left aperture of the secondary ccelom ; Ihn, left upper nephridial aper-
ture ; lo, aperture of the left rudimentary oviduct ; Ivn, left lower nephridial aperture. The four
ctenidia are not lettered.

FIG. 79.— Pallial complex of Nautilus pompilius <J (after Bourne and Lankester). pe, Penis ;
a, muscle band of the siphon ; Isp, aperture of the left rudimentary seminal duct ; nrplm, m.'^Jifi.
lower and upper nephridial aperture of the left side ; olf, left osphradium ; viscper, left aperture of
the secondary ccelom ; an, anus ; x, supra-anal papilla of unknown significance ; c, mantle cut off.

VII

MOLLUSCA—THE PALLIAL COMPLEX

83

FIG. 80.— Sepia Savignyana, from behind (after Savigny). The greater part of the mantle cut
open and laid back on the right side (left in the figure), a, Prehensile tentacle ; &, oral arm ; c,
mouth with jaws ; d, lower aperture of siphon ; e, eye ; /, locking apparatus of the mantle g ; h, right
ctenidium ; i, siphon ; k, locking apparatus of the mantle on the visceral dome ; I, upper aperture
of siphon ; TO, anus ; n, depressor infundibuli ; o, penis ; p, right nephridial aperture ; q, posterior
integument of the visceral dome ; r, tin.

84 COMPARATIVE ANATOMY CHAP.

apertures, but only that on the right side is functional. In the male, the aperture
is produced into a tubular penis.

6. Above the bases of the lower gills there is an osphradium on each side placed
on a papilla.

7. Above the anus there is a large median papilla of unknown significance.

8. The nidamental gland lies dorsally in the mantle.

If we compare with the above the pallial complex of a dibranchiate Cephalopod,
such as Sepia (Fig. 80), we find the following arrangements :—

1. There is one gill on each side.

2. Along the median line of the visceral dome, the rectum and the duct of the
ink-bag descend together, to open through a common aperture at the tip of a papilla
at the base of the siphon.

3. On each side near the rectum, above the anus, a nephriclial aperture occurs
on the point of a papilla.

4. Of the two paired genital apertures only the left has been retained in Scjna
and many other Cephalopoda ; this lies near the left nephridial aperture at the
summit of a large papilla (penis). In the female Octopus, the genital apertures are
paired and symmetrical, and lie to the right and left of the rectum.

5. The two nidamental glands (in Dccapoda) lie in the visceral dome, sym-
metrically with regard to the median line ; they open above the nephridial apertures
into the mantle cavity.

VI. The Respiratory Organs.

The True Gills or Ctenidia.

The most important of the pallial organs in the Mollusca is the gill,
for it is in order to protect it that the mantle, and with it the pallial
cavity, develop. The gill found in the mantle cavity is throughout
all the divisions of the Mollusca a homologous organ, to be derived
from the gill of a common racial form. But since this gill is wanting
in certain Mollusca (e.g. many Opisthobranchia), and is functionally
replaced by new organs which are morphologically altogether uncon-
nected with it, it has been found useful to distinguish the primitive
Molluscan gill by the name of etenidium. This word, therefore, has
a special morphological significance.

The etenidia of the Mollusca are originally paired and symmetri-
cally arranged ciliated processes of the body wall, carrying two
rows of branchial leaflets, and projecting into the mantle cavity.

Venous blood flows into the gills through afferent vessels
(branchial arteries), and after becoming arterial by means of the
respiration, flows through efferent vessels (branchial veins) back to
the body, passing first through the heart. At or near the base of
each etenidium there always lies a sensory organ, which is considered
as olfactory, the so-called osphradium or Spengel's organ.

Such primitive etenidia are met with first in that group of the
Mollusca which has undoubtedly retained more primitive characteristics
than any other, viz. the Chitonidce among the Amphineura. They are,
further, found in all other Mollusca which have retained the original

VII

MOLLUSC A— -RESPIRATORY ORGANS

85

bilateral symmetry of the body, such as the Lamellibranchia, the
Cephalopoda, and — a point of great importance — also in the primitive

FIG. SI. — Ctenidia of various Molluscs (after Ray Lankester). A, Chiton; B, Sepia ; C,
Fissurella; D, Nucula; E, Paludina. ft, Longitudinal branchial muscle; abv, afferent branchial
vessel ; ebv, efferent branchial vessel (branchial vein) ; gl, paired lamellss (leaflets) of the feathered
gill ; in D : d, position of the axis ; a, inner ; 5 and c, outer rows of branchial lamellse ; in E : i,
rectum ; 6r, branchial filaments ; «, anus.

Gastropoda, the Zenyobranchia. In the latter, however, the left ctenidium
was originally the right and vice versa, but this will be dealt with more
in detail later.

86 COMPARATIVE ANATOMY CHAP.

With regard to the number of gills originally present on each side
of the body, opinions are divided. Those who hold that there were
several seem justified by the arrangement in Chiton, where numerous
consecutive ctenidia lie in a longitudinal row in the branchial furrow
(mantle cavity) on each side, and also by that in the Nautilus, which
is rightly considered the most primitive of extant Cephalopods, where
four gills are found (Tetrabranchia). We shall, however, see later that
the other view, viz. that the Mollusca originally possessed only one
pair of ctenidia, has, to say the least, equal claim to be accepted.

In all other Mollusca with paired ctenidia, including the Lamelli-
branchia, there is only one pair at the posterior part of the body.
Further, in the racial form of the Prosobranchia, a single pair of gills
must be assumed to have occupied a posterior position in a mantle
cavity which, with them, shifted forward later to the anterior position.
The Zengobranchia still retain this single pair of gills.

In most Prosobranchia, the asymmetry of the body is also seen in
the gills, only the left gill of the two in the Fissurellidce and Haliotidce
being retained, the right completely disappearing. In the forms which
most resemble the Fissurellidce and Haliotidce, the single-gilled Dioto-
cardia (Turbinidce, Trochidce, etc.), the gill is still feathered on both
sides, but in all Monotocarclia it has only a single row of leaflets.

In one division of the Opisthobranchia, the Tedibranchia, one
ctenidium is still retained, that on the right side. Other Opistho-
branchia have lost the true ctenidium together with the mantle cavity ;
it may be replaced by analogous (but not homologous) respiratory
organs, such as adaptive gills.

The Pulmonata, in consequence of their adaptation to aerial
respiration, have lost the ctenidia.

The blood, which has become arterial in the ctenidia, reaches the
heart through the auricle, and passes into the body through the
arteries. It is therefore evident that a close relation must exist
between the gills and auricles. This relation is briefly as follows :
where the gills are paired, the auricles are paired, and unpaired gills
are accompanied by a single auricle on that side of the body on which
the gill is retained. Where gills are paired, there is almost always
only one pair, and then there is one right and one left auricle.

The Nautilus has four gills, and, to correspond, two right and two
left auricles. The Chitonidce, on the other hand, in spite of their
numerous pairs of gills, have only one right and one left auricle.

The Scaphopoda possess neither true ctenidia nor any other localised
gills. Kespiration may take place at the various soft-skinned surfaces
which come in contact with the water, such as the inner surface of the
mantle, the tentacles, etc.

A. Amphineura.

Chitonidse. — A single ctenidium of a Chiton (Fig. 82) may serve as a type of the
Molluscan gill with its two rows of leaflets. The plumose ctenidium rises freely from

VII

MOLLUSCA— RESPIRATORY ORGANS

87

the base of the branchial groove (mantle cavity). The axis here takes the shape of
a thin septum. At each side, on the broader surface of the septum, extending from

B

FIG. 82.— Structure of the ctenidium of a Chiton (after B. Haller). A, Single ctenidium with
its double row of branchial leaflets. B, Transverse section of the gill along the line a-li in Fig. A.
1, Narrow blood sinus in the branchial leaflet; 2, septum in its axis; 3, longitudinal muscle; 4,
afferent branchial vessel ; 5, efferent branchial vessel ; 6, nerves ; 7, long cilia on the branchial
axis. C, 2 pairs of branchial leaflets cut through at right angles to their surfaces, along the line c-f
in Fig. B. 1, Same as in Fig. B ; S, space between the consecutive branchial leaflets. D, Longi-
tudinal section of the ctenidium somewhat laterally to the axis, and parallel to its septum, along
the. line c-<? in Fig. A. This section is part of a transverse section of the body. Lettering as in
Figs. B and C. In addition : 9, olfactory ridge of the branchial epithelium ; 10, general afferent
branchial vessel ; 11, general efferent branchial vessel ; 12, pleuro-visceral strand of the nervous
system. The branchial epithelium is everywhere indicated by a thick black line.

base to tip, there is one row of smooth, delicate branchial leaflets. In outline they
are more or less semicircular, and stand crowded together in great numbers almost
like the leaves of a book. The entire surface of the branchial epithelium is
ciliated ; on the axial epithe-

Hum, the cilia are remarkably A ' »

long. On that side of the axis
which is turned towards the
foot, a blood-vessel runs from
base to tip, conducting venous
blood to the gill (afferent
branchial vessel). On the op-
posite side, which faces the
mantle, another vessel, the
branchial vein, runs from the
tip to the base of the gill, and
carries the blood, which has
become arterial by respiration,
to the general branchial vein,
and through it to the auricle.
These vessels have no special

endothelial walls, but are surrounded by circular muscle fibres,
is accompanied by a powerful longitudinal muscle.

Fio. SM.— Diagrams illustrating the arrangement of the
gills in the Chitonidae. in, Mantle ; o, mouth ; 7^, snout ; /,
foot; ct, ctenidia ; «, anus.

The branchial vein
At the base of each branchial

leaflet, the blood flows out of the branchial artery through an aperture into the
narrow cavity of the leaflet, and passes through a similar aperture on the opposite
side of the axis to enter the branchial vein. Nerves are supplied to the ctenidium
from the pleuro-visceral nerve which runs close to its base.

88

COMPARATIVE ANATOMY

CHAP.

The number of ctenidia in each row varies very much in the different species of
Chitonidce ; it ranges from 14 to 75. The row extends along the
whole length of the branchial furrow (Fig. 83 A), or else (in
Chiton Icevis, 0. Pallasii, and Chitonellus) is confined to its
posterior half (B, C).

Solenogastres. — (Proneomenia, Neomenia, Chcctodcrma}.
The mantle cavity, in these forms, is much reduced, consisting
only of the groove on each side of the rudimentary foot ; it
opens into the cloacal cavity, or rather widens to form that
cavity. The cloaca is thus the posterior portion of the mantle
cavity. In Cluctoderma (Fig. 84) the foot has disappeared, and
the mantle cavity is reduced to the cloaca, in which one typical
gill lies on feach side of the anus. These gills are regarded as

the last ctenidia of the rows found in the Chitonidce. which in
FIG. 84.— Posterior c m -t i j « i ^

end of the body of C'LitoneHus an(i some species of Chiton are already confined to

Chaetoderma (ilia- the posterior half of the body. In Neomenia, there is no longer

gram after Hubrecht). a pair of ctenidia, but a mere tuft of filaments rising from the

l, Gonad ; 2, pencar- waj} of tne- c}oacal cavity, and in Proneomenia, there are only

dium ; 3, rectum ; 4, ir kr fo]ds of the doacal WftU-

nepnridium ; 5, anus ;

6, ctenidium • 7 ^n "ie relation of the gills in the Chitonidce to certain

cloaca. patches of epithelium, which may perhaps be considered as

osphradia, see the section on Olfactory Organs, p. 165.

B. Gastropoda.

The Fissurellidce (Fig. 85, A and B) among the Prosobranchia stand nearest to
the racial form of the Gastropoda. The mantle cavity is anteriorly placed ; into it
from behind and above project two long gills feathered on each side ; these lie
symmetrically to the middle line, and to the right and left of the anus. The
posterior portion of their axes is connected by a band with the floor of the respiratory
cavity, while the anterior pointed portion projects freely.

The fact that in the Fissurellidce (and related forms) the gills are paired and
symmetrical is very significant. It points to the primitive character of these forms,
and enables us to compare, their gills with those of the lower Lamellibranchia., i.e.
the Protobranchia, and of the Cephalopoda. We must, however, again emphasise
the generally-assumed fact that the left gill of Fissurella answers to the right gill of
the Lamellibranchia and Cephalopoda, and the right gill of the former to the left of
the latter, these latter having retained their primitive symmetry in this respect.
This assumption becomes the more plausible when we consider that the mantle
cavity with its organs originally lay posteriorly on the body, and shifted forward
secondarily along its right side.

The Haliotidcc are closely connected with the Fissurellidce. Their spacious mantle
cavity is, however, forced to the left side by the great development of the columellar
muscle. There are two gills, feathered on both sides, of which the right is the
smaller. The axis of each gill has united, for nearly its whole length, with the
inner wall of the mantle, and only its anterior end is free ; its tip even projects a
short distance beyond the respiratory cavity.

Although the Fissurellidce and Haliotidcc still possess two gills, other Diotoc«r<li<i
have retained only the left (ur) and larger gill of Haliotis. This gill is, however,
still feathered on both sides, although this characteristic is obscured in a peculiar
manner. The septum or axis of the gill, to the broader surfaces of which the branchial
leaflets are attached, and one edge of which had, in Haliotis, already fused with the

VII

MOLL USCA—RESPIRA TORY ORGANS

89

inner wall of the mantle, becomes attached to the mantle by its other edge
also (viz. that along which the branchial artery runs), somewhat to the right
of the first line of concrescence. In this manner, which is illustrated by the
accompanying diagrammatic sections (Fig. 86), the mantle cavity is divided
by the branchial septum into two unequal parts, which open into one another
anteriorly.

Into the much smaller upper division the one row of smaller branchial leaflets
projects, while the opposite row of larger leaflets hangs clown into the lower and
larger chamber. The anterior end of the
gill, however, is still free, its point pro-
jecting anteriorly (Trochidcc, Turbinidcc,
Ncritidcc}.

In the Docoglossa (Patcllidcc) the ar-
rangement of the gills is very varied.
While the LepctidcR have no gills whatever,
we find in Patella a single row of numerous
small branchial leaflets right round the
body, on the inner or under side of the
short encircling mantle fold, between it and
the foot. This row is broken only in one
place anteriorly on the left. It is, how-
ever, evident that these gills, which some-
what resemble those of the Chitonidcc, are
no true ctenidia, from the fact that there
are Docoglossa (e.g. some forms of Tcctura
and Scurria) which possess, in addition
to this marginal row of leaflets, a typical
ctenidium corresponding in every way with
that of the Turbinidce, Trochidce, etc.
Other forms, such as Acmcea, have only the
true ctenidium and no marginal branchial

FIG. 85.— Subemarginula after removal of the

In the large second division of the shell (arter Fischer). A, from above ; B, from
Prosobranchia — the, Monotocarclia — the right side. The mantle cavity is exposed by
arrangement of the gills is, on the whole, bending back the mantle fold 4. l, Snout ; 2,
remarkably uniform. There is only a tentacle with the eye on its short stalk behind
.,/. . -, ,„. »n it; 3, right cteuidmm; 4, mantle fold; 5, shell

single gill feathered on one side (Fig. i\, miujc.le; 6j edge of the mantle encircling the
p. 73), united to the mantle along almost body . 7] epipodium; 8, foot,
its whole length ; this gill corresponds

with the left ' gill in Fissurclla and Halwtis, and the single gill in Turbo and
Trochus. It generally lies quite to the left in the mantle cavity.

The rise of this gill can best be explained by recalling the arrangements already
described in Turbo and Trochus. We have only to assume that the row of small
leaflets turned towards the mantle in Turbo disappears, and that the branchial
septum unites with the mantle across its whole width (Fig. 86, C, D).
A few anomalous forms alone require special mention.

1. In a series of terrestrial Monotocardia, aerial respiration has taken the place of
aquatic respiration, and the ctenidium has disappeared (Acicula, Cydostoma, Cydo-
phorus, etc.).

2. The Ampullaria are amphibian Prosobranchia. A doubling of the mantle
gives rise to a very spacious pulmonary sac, 011 the inner surface of which the respira-
tory vascular network spreads out. The lower wall of this pulmonary sac, which
forms at the same time the roof of the mantle cavity, is perforated by an aperture

-V-- 1

90

COMPARATIVE ANATOMY

CHAP.

for the inhalation and exhalation of air. l The ctenidium is placed to the extreme
right of the mantle cavity, a position which is in some' way connected with the great
development of the pulmonary sac. It nevertheless answers to the left gill in other
Mfinotocctrdia, as can be seen from its innervation.

3. The genus Valvata is unlike all other Monotocardict, in that its gill is
feathered on both sides and projects freely. It can, further, be protruded from
the pallial cavity.

4. In Atlanta, among the Hcteropoda, the gill is well hidden in the spacious
mantle cavity. In Carinaria, it is only slightly protected in consequence of the
small development of the mantle fold. In Ptcrotrachca there is no mantle fold, and
the filamentous branchial leaflets project free and uncovered. Firoloides has no
gills.

Opisthobranchia. — A true ctenidium is here found only in the Tectibranchia and

FIG. 86.— General Morphology of the gills of the Prosobranchia. Diagrammatic sections
in the region of the mantle cavity, from behind, t' A, Haliotis ; B, Trochus, anterior portion of the
pallial cavity. C, Trochus, middle or posterior portion of the cavity. D, Monotocardia. 1,
Mantle cavity ; 2, rectum or anus, r right, I left gill of Haliotis (A), which latter is the only gill
present in the Azygabranchia (B, C) and MonotomnHu (D). i, Branchial leaflet of the inner row ;
e, ditto of the outer row, between them the branchial axis or septum with the afferent and efferent
branchial vessels (3 and 4) ; 5, position of the mantle slit in Haliotis (<•/. p. 43). Further explana-
tions in the text.

in the Steganoliranchia among the Ascoglossa. It lies, often incompletely covered,
in the mantle cavity which is developed on the right, and is, in some cases at least
(e.g. Pleurobrauchus), distinctly feathered on both sides.

In the Pteropoda, which must be derived from the tectibranchiate Opistho-
branchia, the ctenidium, when present, is little developed, and lies on the right side
of the body. It answers to the tectibranchiate ctenidium.

In the Gymnosomata, this true gill is retained only in the Pneumodermidce as a
simple, or less frequently (Pneumoderma) fringed, process on the right side of the
body (Fig. 76, p. 79). New gills, on the other hand, may develop at the posterior
end of the body, occurring either together with the true ctenidium (Spongiobranchcea,
P iieumodcrma), or alone (Clwnopsis, Notobmncluea}, until they in their turn dis-
appear (Clionc, Halopsyche).

Among the Thecosomata, the Cavoliniidcc alone (Fig. 87) possess a gill which

1 In this and in the closely-allied Lanistes there is in addition a protrusible siphon
on the left side (v. Fischer and Bouvier, C. R. cxi. p. 200).

vii MOLLUSCA— RESPIRATORY ORGANS 91

rises in the form of a series of fold-like elevations of the body wall in the pallial

FIG. ST.— Anatomy of Cavolinia
tridentata (after Souleyet). Shell
and mantle removed, and visceral
dome partly opened, seen from behind
and below. (/, Right ; s, left ; 1 , aper-
ture of the penis ; 2, mouth ; 3, left tin
(parapodium) ; 4, foot ; 5, oesophagus ;
6, part of the efferent genital appara-
tus ; 7, ventricle ; 8, auricle ; 9, herma-
phrodite gland ; 10, lateral processes
of the mantle ; 11, columellar muscle ;
12, intestine ; 13, digestive gland
(liver); 14, stomach; 15, ctenidium ;
16, genital aperture ; IT, anus.

cavity, and which, running in a wavy line, forms a semicircle, open anteriorly, the
greater portion of it, however, lying on the right side.

C. Lamellibranchia.

The Lamellibranchia also possess typically two symmetrically placed gills, each
provided with two rows of branchial leaflets. The opinion which until lately was
common, that the Lamellibranchia possessed two gills on each side of the mantle
cavity, has been shown to be incorrect — these two gills in reality answering to the
two rows of branchial leaflets of one typical gill.

It is worth while to follow, step by step, the interesting series of modifications
undergone by the original gill in the Lamellibranchia.

(«) The primitive arrangement is found in the Protobranchia. Taking Nucula
(Fig. 21, p. 14) as an example, we find a gill like that of Fissurella, consisting of
an axis along which the branchial artery and the branchial vein run, and which is
attached by a short membraneous band to the posterior and upper portion of the
body or visceral dome, and to the posterior adductor muscle. On this axis are
attached two rows of short flat branchial leaflets. These two plumose gills converge
posteriorly, and project with their free tips into the mantle cavity. The leaflets of
both rows are directed somewhat downwards, so that they are at right angles to one
another. In Malletia and Solenomya, on the contrary, they lie in the same plane,
the two rows standing out on opposite sides of the axis. In Malletia, this plane is
horizontal, but in Solenomya it trends downwards and inwards. The number of leaflets
on the very slender gill of Malletia is much smaller than on that of Nucula ; they
are consequently neither so crowded nor so flattened. Each leaflet contains a blood

92

COMPARATIVE ANATOMY

CHAP.

sinus, which is a continuation of the branchial artery. Two rods of connective
tissue run along the lower edge of each leaflet from the axis to its tip, and serve for
its support. Similar supports are found in almost all Lamellibranchia and in
many Gastropods.

The epithelium of the branchial leaflets is beset with long cilia — (1) at the
ventral edge ; (2) on both (anterior and posterior) surfaces, near the ventral edge.

The first-named cilia form, with regard to the whole gill, a longitudinal row along
the free ventral edge of each row of leaflets, and bring about a current in the water
along this edge from behind forward. The other cilia mentioned above, mingling
together like the bristles of two brushes which arc pressed together, form a loose
connection between the successive leaflets of the row.

(b) In the Filibranchia (Fig. 88 B) the leaflets in each of the two rows are very
long and filamentous, and hang down far into the mantle cavity. The branchial
filaments of the two rows are recurved and bent back upon themselves, so that in
each filament a descending ami an ascending portion can be distinguished. The
prolongation of the filaments corresponds with a necessary increase of the respiratory

FIG. 88.— Morphology of the gills of the Lamellibranchia, diagrammatic transverse sections.
A, Protobranchia B, Filibranchia. C, Eulamellibranchia. D, Septibranchia. 1, Mantle ;
2, body (visceral dome) ; 3, foot ; e, in A, branchial leaflets of the enter row in the feathered gill,
in B, branchial filaments of the outer row, in C, outer branchial leaf ; i, branchial leaflets or
filaments of the inner row or inner branchial leaf; e\, ascending branch of the outer filament,
or lamella of the outer leaf; ij, ascending branch of the inner filament, or lamella of the inner leaf;
in D, s, signifies the gill which has become transformed into a muscular septum which divides the
mantle cavity into an upper (4) and a lower (5) chamber, the two communicating by means of slits
(o) in the septum. Further explanations in the text.

surface. By this bending back of the filaments, the gills make the most of the
limited space afforded by the mantle cavity. Each filament of the outer row is bent
outwards, and of the, inner row inwards.

The filaments of each row may be so crowded together that the whole row looks
like a leaf or fringe. This branchial leaf consists of two closely contiguous lamellf.
one the descending and the other the ascending, the two passing into one another
at the lower edge of the leaf. The descending lamella is formed by the descending
portions of the filaments, and the ascending by the ascending portions. On the
outer leaf, the ascending lamella is the outer one, on the inner leaf the inner.

In the Filibranchia, the separate branchial filaments retain their independence—
they are free, i.e. the separate filaments of a series are unconnected with one
another, and the descending and ascending portions of one and the same filament
are in no way united. There are, however, on both the anterior and posterior sides
of the filaments places covered with long cilia closely crowded together. These
ciliated tufts on adjoining filaments mingle, and so give rise to a sort of connection
between the filaments of each leaf.

In the Mytilidcc, so-called interfoliar junctions or trabeculte occur at certain

vii MOLLUSGA— RESPIRATORY ORGANS 93

points between the ascending and descending portions of the branchial filaments,
but no blood-vessels run into them.

In Anemia, the dorsal ends of the ascending portions of the outer lamella are
free, but in the Arcidce united, although their internal cavities are not in communi-
cation. In such cases, the interior of each filament is divided by a longitudinal
septum into two canals. In one of these the blood flows from base to tip, and in the
other back from tip to base, i.e. to the axis. In the Myiilidce, the dorsal ends of the
recurved portions of the filaments of eacli branch have grown together, and their
blood-vessels communicate at the points of junction, i.e. along the upper edge of the
ascending lamella.

(c) Pseudolamellibranchia. —Each leaf of the gill is here folded, to secure increase
of surface. The plications run longitudinally with regard to the filaments, and are
thus almost dorso-ventral. There are, therefore, distinct alternate ridges and furrows
on each leaf, the ridges on the one surface corresponding with those on the other, and
the furrows corresponding with furrows. Each ridge or furrow is formed by one
filament ; the filament forming the furrow is in some way, such as greater breadth,
distinguished from the others. The two lamella of each leaf of the gill are united
here and there by trabeculae, which may or may not contain blood-vessels. They occur
either between the opposite furrows or between the opposite ridges, i.e. between the
ascending and descending portions of the filaments which lie either in the furrows
or ridges. The upper edge of the ascending lamella of the outer leaf may unite with
the mantle. The consecutive filaments of the same leaf are only connected by means
of tufts of cilia.

(rf) Eulamellibranchia (Figs. 89-91). — The branchial leaves are either smooth or
folded, but there is always organic connection, by means of numerous vascularised
junctions, not only between the ascending and descending lamellae, but between the
successive filaments. The junctions are therefore both interfoliar and interfilamentar.
This leads to the entire disappearance of the original filamentous structure of each
leaf, which now becomes an actual leaf or lamella with perforations or slits, the
remains of the spaces between the original filaments, leading into an internal system
of sinuses or canals, which in their turn are the remains of the spaces between the
ascending and descending lamellae. This peculiar arrangement was formerly con-
sidered typical of the Lamellibranchia, and was the origin of their name. It was
supposed that the animals of this class had two leaf-like gills on each side of the
mantle cavity, i.e. four altogether, but we now know how the two branchial leaves on
each side arose, that they are in fact the two, modified, rows of leaflets of the original
plumose gill of the Protobranchia. The Lamellibranchia in reality possess only one
gill on each side in the mantle cavity.

The blood now no longer flows through the primitive filaments of the lamellae
of the gills and back again, but the afferent and efferent channels lie in the trabecular
network between the two lamellfe of a branchial leaf.

Instead of the two leaves of a gill hanging down into the mantle cavity parallel
to one another, the outer leaf may stand up dorsally in the cavity, so that the two
come to lie in the same plane (Tdlinidce and Anatinacca).

The ascending lamella of the outer leaf may be wanting (Anatinacea, Lascea),
and in fact the entire outer leaf may be absent (Lucina, Corbis, Montacuta,
Cryptodon).

In all Lamellibranchia, with the exception of the Protobranchia, and further, of
the Arcidce, Trigonidce, and Pcctinidce, the gill and mantle unite, the dorsal edge of
the ascending (outer) lamella or, where this is wanting, the free edge of the single
lamella of the outer leaf becoming fused with the mantle. In the same way, the
dorsal edge of the ascending (inner) lamella of the inner leaf may become fused with
the upper part of the foot (Fig. 88 C). If the two gills, which have fused with the foot,

94

COMPARATIVE ANATOMY

CHAP.

fuse with each other behind the foot in the middle line of the mantle cavity, they form
a septum which, uniting with the septum formed by the mantle between the inhalent
and exhalent siphons, divides the cavity into an upper and a lower chamber. The
water flows through the lower (inhalent) siphon into the large lower chamber, bathes
the gills, and, streaming forward, conveys the particles of food it contains to the
mouth. It then flows back along each side of the foot in the upper chamber of the
mantle cavity (which is itself divided into two canals by the line of insertion of the

FIG. SO.— Part of a transverse section of the outer branchial leaf of Dreissensia polymorpha
(after Peck). /, The separate filaments ; /, sub-epithelial fibres ; c/i, supporting substance of the
filaments ; lac, lacunar or alveolar tissue ; pig, pigment cells ; be, blood corpuscles ; fe, epithelium
of the free edge of the branchial filaments ; Ifei, lfe%, two rows of lateral epithelial cells of the
branchial filaments, carrying long cilia (ciliated tufts) ; Irf, tissue of the interfilamentar junctions.
Two interfoliar junctions are shown in the figure.

gill) into the single posterior and upper chamber behind the foot, and escapes through
the upper (exhalent) siphon (Fig. 26, p. 18).

(c) Septibranchia (Fig. 31 A and B, p. 21 ; and Fig. 88 D, p. 92).— These Mussels
were formerly erroneously considered to be gill-less. As a matter of fact, the
branchial septum just described has in them been much modified in structure, and
has become a muscular septum, running across the mantle cavity in a horizontal
direction and joining the siphonal septum posteriorly, while anteriorly it passes
round the foot. This septum is broken through by various perforations and slits,
which allow of communication between the upper and lower chambers of the mantle
cavity, and vary in the different genera.

VII

MOLL USCA—RESPIRA TOR Y ORGA XS

95

FIG. 90.— Portions of transverse sections of the branchial lamella of Anodonta (after Peck).
A, Outer ; B, inner lamella. In each leaf the cross sections of both lamellce are seen, and also the
interfoliar as well as the interfilainentar junctions. C, A part of B much magnified, ol, Outer; II,
inner lamella of the same leaf; v, blood-vessels ; /, the separate filaments of which the lamella?
consist; lac, lacunar tissue; ch, supporting tissue of the filaments, with firmer supporting
rods, chr.

ttt

FIG. 91.— Portion of the ascending lamella of the outer branchial leaf of Anodonta.

diagrammatic (after Peck). /, The separate filaments, connected by interfilamentar junctions ;
trf, connective tissue of the latter ; v, blood-vessels ; ilj, interlamellar junctions ; the perforations
in the lamella (of a darker shade) are the spaces remaining between the filaments and their
junctions, through which the water needed for respiration can flow.

96

COMPA EA TIVE ANA TOMY

CHAP.

D. Cephalopoda.

The gills of the Cephalopoda are always feathered on both sides. Those of the
Dibrciiicliia have been the most thoroughly investigated. In Sepia, each gill has
the shape of a slender cone, its whole length being applied to the visceral dome in
the mantle cavity, in such a way that the base is directed dorsalty towards the apex

of the visceral dome, and the point ventrally
towards the free edge of the mantle fold or
the mantle cleft (Fig. 80, p. 83). The
points of the two gills diverge.

The two rows of flat triangular branchial
leaflets (Fig. 92) are carried by the two
branchial vessels, each leaflet being attached
by one end of its base to the branchial artery
and by the other to the branchial vein. In
the axis of the gill between the two vessels,
and also between the bases of the two rows
of leaflets, a channel is formed which com-
municates by a slit between each successive
pair of leaflets with the mantle cavity ;
through this canal the respiratory water
freely flows. The slits in this axial channel
are arranged alternately on each side, like
the leaflets between whose bases they lie.
The branchial vein forms the posterior sup-
port of the gill turned towards the mantle,
and the branchial artery the anterior support
turned towards the visceral dome. The
artery is united along its entire length with

10

FIG. 92.— Diagram to illustrate the struc
ture of the gill of Sepia (after Joubin). 1,
Branchial vein (containing arterial blood) ; 2,
branchial canal ; 3, branchial; artery (contain-
ing venous blood) ; 4, special branchial vein

(vasefferens)of each leaflet; 5, special branchial the integument of the visceral dome by a

artery (vas afferens) of each leaflet ; (i, suspensor
of the gill, which attaches the branchial artery
(3) to the posterior integument of the visceral
dome (12) ; 7, suspensor of each leaflet to the

membrane of connective tissue. The an-
terior edge of each leaflet (that facing the
visceral dome) is connected with this mem-

general suspensor (6) ; 8, one of the connecting braiie, which may be called the gill-suspen-

vesst-ls between the branchial artery and the b meanjj of anotlier triangular mem-

"blood-making gland" (9), through which . , , ,

venous blood flows; 10, 11, vessels carrying bralle' A SPeCial Vem runs alollS tlie

the venous blood which has passed through posterior free edge of each leaflet, and enters

the " blood-making " gland back to the venous the general branchial vein at its base ; and

sinus at the base of the gill. The arrows in- a Special artery runs along the anterior edge,

dicate the direction of the blood-stream. . , , •> c .-, n a , i • i •

i.e. along that edge or the leaflet which is

fastened to the suspensor. Each leaflet is wrinkled in such a way that the folds
on the two surfaces alternate, each fold being creased in its turn. These two systems
of folds cross each other at right angles, and serve to increase the respiratory surface.
At the point where the suspensor of the gill passes into the integument of the
visceral dome, it contains a cellular body, which is traversed by a system of inter-
cellular blood-channels. This may perhaps be a blood-making gland. It receives
venous blood from branches of the principal branchial artery and of the special
arteries of the leaflets, and returns the same along two veins which run back to the
base of the gill, there, with others, to open into the venous sinus of the renal organ ;
from this organ the blood passes for the second time along the branchial artery into
the gill. We thus find that not all the venous blood which' is conducted by the
branchial artery towards the gills enters the leaflets for purposes of respiration ; part

vii MOLLUSC A— RESPIRATORY ORGANS 97

of it streams through the " blood-making " gland, and returns to the venous branchial
heart still unpurified. There are, further, certain fine branchings of the branchial
artery which serve for nourishing the gill and its suspending membranes. The blood
in these returns to the venous sinus through a special vessel which runs parallel to
the branchial artery on its anterior side.

A powerful nerve enters the gill at its base and ramifies through it. A muscle
spreads over the surface of the " blood-making " gland, and a special musculature
brings about the contractions of the principal branchial vein.

The gills of the Octopoda differ considerably, though not essentially, in structure
from those of the Decapoda. The branchial channel is much larger, and the leaflets
are not only folded, but have on each side alternating lamella-, which in their turn
may carry similar lamellae of the second order, and so on till in some cases the
seventh order of subsidiary lamellae is reached. The leaflet is thus an extremely
complicated, folded, or feathered structure with its surface increased to an
extraordinary degree.

Adaptive Gills.

The Scaphopoda and many Gastropoda possess no true ctenidia.
In the Fulmonata and the few air-breathing Prosobranchia, the ctenidia,
as organs adapted for aquatic respiration, have disappeared. It is,
however, at present difficult to determine the cause of their dis-
appearance in Opisthobranchia which inhabit water, and in the gill-less
forms of the Pteropoda, all the more so, as in most Opisthobranchia
they are replaced by adaptive gills, which are new structures in no
way comparable morphologically with ctenidia. These adaptive gills
may even appear (Pnewmoderma) before the true ctenidia have dis-
appeared. The Scaphopoda and many Opisthobranchia have no gills
whatever, and in these respiration evidently takes place at various
suitable parts of the surface of the body. In many cases, also, where
epipodial or parapodial processes are developed as well as gills, or
the mantle possesses exteiisions, these may help the gills in the
function of respiration.

Adaptive gills are found in most Ascoglossa and in the Nudi-
bmnchia ; also, as mentioned above, in the gymnosomatous Pteropoda.
In the latter, they consist of small fringed or plain ridges at the
posterior end of the body ; these may be of various shapes ; a
description of them would be of no special interest to the comparative
anatomist.

The principal forms of adaptive gills of the Nudibranchia are :
(1) the anal gills of the Dorididce ; (2) the longitudinal rows of
branchial leaflets to the right and left under the mantle fold of the
so-called Pk>/Uidiida> : (3) the dorsal appendages or eerata of the
Nudibranchia and most Ascoglossa.

1. The Anal Gills (Fig. 93). — These take the form of delicate leaflets, generally
feathered on both sides, which, in the Dorididcc, form a rosette round the anus,
which has a median dorsal position towards the posterior half of the body. Cerata
may occur with the anal gills (Polyceridce). The view that these gills are ctenidia
has as yet no sufficient foundation.

VOL. II H

98

COMPARATIVE ANATOMY

CHAP.

Of

2. The Longitudinal Rows of Branchial Leaflets (Fig. 20, p. 13). — These organs,
which lie to the right and left of the body in the Phyll idiidcc and Plcurophyl-
lidiidce. bear the same relation to the (lost) true ctenidium as do the respiratory struc-
tures of the Patellidce above described to
the same organ, which in them is some-
times present, sometimes wanting. The
longitudinal rows consist of numerous
small lamellae which project from the lower
side of the enveloping mantle fold into the
shallow pallial cavity. There is either one
long row of these lamellfe running along
the whole length of the mantle fold and
only interrupted anteriorly (Phyllidia), or
a row interrupted posteriorly as well
(Pleurophyllidia) ; or again, the rows of
lamella? are confined to the posterior end
of the mantle fold (Hypobranchicea). The
genus Dcrmatobranchus has no gills.

3. Dorsal Appendages (Cerata) (Fig.
18, p. 12). — These processes vary very
much in form, being sometimes simple,
and sometimes branched ; they differ also
greatly in number and arrangement. At

their tips there are often cnidophore sacs ;
these are invaginations of the ectoderm in
which stinging cells with stinging capsules
are developed. Diverticula of the intestine
(digestive gland) enter the cerata, and may
open outward at their tips. The cerata are
generally striking and beautiful both in

FK:. 93. - Respiratory and circulatory eolour alld markings. In some cases they

system of Doris, after Leuckart ("\Vuml- may serve for protection and concealment,

tafeln"). a, Rhinophure ; i,, posterior edge of in others, where the brilliant colouring is

the visceral dome ; c, end of the foot ;d, plumose combined with stinging properties, they
gills ; tli, two gills cut oft"; e, anus ; /, auricle ;
rj, ventricle; 7i, aorta; i, circular vein around

the gill, and sends it through the branchial vein
into the auriide ; k, circular artery, which receives
the venous blood coining from the body ; x, two
vascular trunks, which conduct venous blood
direct to the heart.

may serve as a warning. They often break

the anus, which receives the arterial blood from off easily at the base (as a protective ar-
rangement), and are always quickly regen-
erated. They no doubt assist, like the
rest of the body surface, in respiration,
especially where they are much branched
and richly supplied with blood-vessels.

Certain Opistkobranckia are altogether gill-less, e.g. the Elysiida:, Limapontidce,
and Phyllirrhoida;.

Among the Pulmotmta, the shell-less genus Onchidium has developed adaptive
gills. The species of this genus are amphibious, living on the sea-coast, within reach of
the tide. Their pulmonary cavity is very small ; respiration therefore takes place by
means of the richly vascularised dorsal integument, and especially of the simple or
branched dorsal papillae, in which there is a rich vascular network, which receives
the blood from an afferent vessel and gives it off to an efferent vessel.

VII

MOLL USGA—RESPIEA TOR Y ORGANS

99

f

Lungs.

The total disappearance of the typical molluscan ctenidium is
characteristic of the Pulmonata, and is connected with their terrestrial
life and aerial respiration. Instead of water, air enters and escapes
from the mantle cavity which lies either anteriorly or laterally on
the visceral dome, and thus the mantle cavity becomes a pulmonary
cavity. The free edge
of the mantle fold, which
forms the roof of this
cavity, unites with the
nuchal integument be-
neath it, except at one
point on the right, where
the respiratory aper-
ture, which can be closed
at will, allows of the
entrance and egress of
air. Along the line of
its concrescence with the
integument, the edge of
the mantle is much
thickened, forming the
mantle border, and is FIG. 94.— Slightly oblique transverse section through the

verv rich in lime-secret- body and sheu of Helix taken J'ust in front of the coiumeiia

" . . „, . (after Howes), pgl, Pedal gland ; fs, lateral pedal blood sinus ;

ing glands. Ihe inner ((0j cephalic aorta; gd, genital duct (uterus); rp, retractor

delicate Surface of the muscle of penis ; phu, pallial muscle, the pallia! edge having

mtlp whiVh forms flip united with the lm''hal llltegument ; si, salivary gland ; cr, crop,

e' VVUK .e or widening of the oesophagus ; s, shell; ms, floor of the pulmon-

1'Oof of the Cavity, IS ary cavity = dorsal integument of the posterior nuchal region

Overspread by a close which is covered by the mantle; sp, spermatheca = stalk of the

Vas •

receptaculuni seminis ; pli, pulmonary cavity ; pv, afferent pul-

monary vessels ; ,n, ,-,.,,al duct ; •»-, rectum ; Jtgl, hermaphrodite
WOrk. A Circular Vein gland or ovotestis ; /, digestive gland (liver) ; hd, hermaphrodite
1'UnS aloil0" the mantle ^uct; m> columellar muscle; aijl, albumen gland ; !, intestine;
® ' st, stomach.

collar, r rom it spring

numerous fine anastomosing vessels which ramify on the mantle.
These vessels are again collected into larger trunks, winch enter the
large pulmonary vein. This vein runs upwards and backwards,
along the right side of the pulmonary cavity, to the left of and almost
parallel with the rectum, and enters the auricle. The circular vein
contains venous blood, but the pulmonary vein conducts blood which
has become arterial through respiration in the vascular network, to
the heart.

Since, in most Pulmonata, as in the Prosobmnchia, the respiratory
organ and the pallial cavity in which it is found lie in front of the
heart, this order is prosopneumonic. An account of the opistho-
pneumonic condition of certain Ptdmonata, which results from the

100

COMPARATIVE ANATOMY

CHAP.

displacement of the visceral dome and mantle to the posterior end of
the bod}', will be found in Section V., p. 76.

Certain Pulmonata (Limna'ida?) have become readapted to aquatic life, but their
respiration is the same as that of the terrestrial forms, they rise periodically to the
surface of the water to take in air. The respiratory cavity, is, however, filled with
water when the animal is young, and it is then a water breather. In Limncea
abyssicola, a deep-water form found in the lake of Geneva, this form of aquatic

FIG. 95. — Helix. The roof of the pulmonary sac cut along the rectum, and along the edgi-
uniting with the nuchal integument, and turned back to show the arrangement of the blood
vascular system, after Howes. The pulmonary veins are of a lighter shade than the afferent
pulmonary vessels and the venous sinuses ; aa, W>, show the cut edges which belong to each other ;
1, afferent pulmonary vessels which draw their venous blood from the large circular venous
sinus (9); this latter receives its blood from the large sinuses of the body, two of which,
that of the visceral dome (0) and that on the right side of the foot (7) are shown. The efferent
pulmonary vessels collect the blood which has become arterial on the roof of the pulmonary
chamber, and conduct it through the pulmonary vein (2) to the auricle (3) ; 4, ventricle ; 5, renal
circulatory system.

respiration continues throughout life, and the pulmonary chamber, in uo way
modified, is constantly filled with water.

Iii certain terrestrial Prosobranchia (Cydostoma, Cyclophorus, etc.) the
respiratory cavity becomes transformed, as in the Pulmonata, into a
pulmonary chamber, and its- roof is covered with a respiratory
vascular network. But there is here no concrescence of the edge of
the mantle with the nuchal integument. Cyclostoma still retains a
rudiment of a prosobranchiate gill, but this is lost in Cyclophorus.
The amphibian Ampullaria possess both a gill and a pulmonary
sac,1 and can breathe either water or air.

1 See note ante, p. 90.

vii MOLLUSGA—HYPOBRANCHIAL GLAND, HEAD 101

VII. The Hypobranehial Gland.

(Slime gland of the Prosobranchia, epithelial shield of the Pteropoda,

etc., anal gland, etc.)

This is an organ very commonly found on the molluscan mantle,
always occurring near the ctenidium, at its base or between it and
the rectum. Of. on its position and occurrence Section V.

The hypobranchial gland varies consider'ably in shape, but is
never a multicellular, acinose, or tubular gland wifeh efferent ducts. It
is originally a more or less extended area of the epithelium of the
mantle cavity (generally of the inner surface of the mantle) in which
epithelial glandular cells are particularly numerous. In this condition
it is not very distinct from the parts around it, but it may become
more definitely localised, and may assume a definite shape ; and in
this latter case, the glandular epithelium of which it consists may also
become folded in order to obtain a larger secretory surface, the folds
being more or less closely crowded together and projecting into the
mantle cavity. This gland often secretes a large quantity of mucus.
The purple gland of certain Prosobranchia (Purpura, Murex, Mitra) is a
hypobranchial gland, the slimy secretion of which is, immediately
after ejection, colourless or only slightly coloured, but under the
influence of light becomes violet or red. In Purpura, the gland consists
of two parts which differ slightly in structure.

VIII. The Head.

If by the word head is meant an anterior portion of the body
more or less distinct from the rest, possessing a mouth and specific
sensory organs, the Lamellibmnehia must be considered headless, and
as such have been distinguished as Accphala, from other Mollusca.
This absence of a head in the Lamellibranchia cannot be regarded as
a primitive condition,1 but is to be accounted for by their general
habit of living in mud, and by the strong and peculiar development
of the mantle and shell, which, by cutting off the anterior portion of
the body (with the mouth) from direct contact with the outer world,
renders specific sensory organs useless. In those Molluscs which
have to seek, seize, and crush their food, a projecting head carrying
sensory organs and furnished with buccal armature is of great use.
Bivalves, however, feed on particles brought to the mouth by the
water which by the motion of cilia is driven through the mantle
cavity ; buccal armatures are thus unnecessary.

In the Cephalopoda, the head is strengthened by the incorporation
with it of the foot, here transformed into a circle of arm-like prolonga-

1 Hence the term " Lipouepliala," suggested by Lankester.

102 COMPARATIVE ANATOMY CHAP.

tions for seizing the prey. We thus have a combined head and foot
(Kopffuss), on each side of which, anteriorly, lies a large highly-
developed eye. This head is more or less separated from the rest
of the body (the visceral dome) by a neck.

The Gastropoda, with very few exceptions, possess a head which on
its anterior lower side is provided with an oral aperture, on its upper
side with eyes and tentacles, and often asymmetrically (generally on
the right side) with a genital aperture or a copulatory organ. This head
is distinctly separated ventrally by means of a groove or furrow from
the foot behind it ; dorsally it passes gradually into the neck. Further
details of this Gastropod head are given below.

A. Gastropoda.

1. Prosobranchia.

The head in this order always carries tentacles, which are solid, simply contractile
(not invagmable) processes of the cephalic wall. It may be assumed that there
were originally two pairs of tentacles, an anterior and a posterior pair. The
posterior are called ommatophores and carry eyes at their tips. Most Diotocardia
possess anterior tactile tentacles, and posterior and slightly lateral optic tentacles.

The cephalic tentacles are always innervated from the cerebral ganglion, and are
thus distinguishable from the tentacular processes which may occur near them on
the head or neck, but belong to the epipodium, and are innervated from the pedal
or pleural ganglia.

In the Docoglossa and most Monotoeardin the optic tentacles do not rise separately
from the head, but are to a greater or lesser extent fused with the tactile tentacles.
Starting with the tentacular arrangements existing in Dolium, S trombus, Rostellaria,

we find the tactile and optic tentacles
fused for a certain distance from tin-
base, but separating later, the tips
projecting independently (Fig. 96, B).
If the two tentacles were of the
same length, and were fused for their
whole extent, there would only be one
F,r, 90. -Relations of the tactile and optic tentacie on each side of the head, which
tentacles in the Prosobranchia. Description in
the tpxt would carry the eye at its tip (Tereora

(7). But if the optic tentacle is shorter

than the tactile, the eye might be met with at any point between the base and tip
of the latter, on a projection which answers to the tip of the fused optic tentacle (D
and E). Finally, the eye may be altogether sessile, i.e. it may lie near the base of
the sensory tentacle in the integument of the head (F).

The snout, which carries the mouth and is anterior to the tentacles, is very
variously developed in the Prosobranchia.

1. It is short and truncated in the Diotocardia, and especially in the herbivorous
Tcvnioglossa,

2. It is prolonged like a proboscis (rostrum), but is only contractile, not invagm-
able (Capulidcc, Strombidce, Ctcnopidcc, Calyptrceidce), or else can be invaginated,
commencing at the tip (C '///"" ''"V'-1, Lamellaridix, Natiddce, Scalarida', Solaridce).

3. It is transformed into a long proboscis with the mouth at its anterior end.
This proboscis can be invaginated in such a way that the invaginated base forms a
proboscidal sheath for the non-invaginated anterior portion or tip. Gastropods

vir MOLLUSGA—HEAD 103

with such proboscides are nearly all carnivorous (the Tritonidw, Doliidoc, and Cassi-
didce, among the Stenoglossa the RacMglossa, and a number of Toxirjlossa).

Most male Monotocardia have a non-invaginable penis, which varies in shape, on
the right (rarely on the left) side of the head or neck, near the tentacle ; this organ
in most cases belongs morphologically to the foot, being innervated from the pedal
ganglion ; less frequently it is a cephalic appendage, and is then innervated from the
cerebral ganglion (Fig. 71, p. 73).

The head of the Hctcropoda carries two tentacles (occasionally rudimentary : Ptrn>-
ti'in-Iii'a, Firoloidea). The eyes are sessile or placed on small prominences near the
bases of the tentacles on their outer posterior sides. That part of the head which
lies in front of the tentacles is prolonged to form a large proboscis-like non-invagin-
able snout.

2. Opisthobranchia.

The shape of the head in this order varies to an extraordinary degree, and can
here be only generally described. It usually carries two pairs of tentacles ; the
posterior pair, which are called rhinophores, are perhaps olfactory. Their surface is
often increased by the formation of circular folds. They frequently rise from the base
of pits into which they can be withdrawn. The head is rarely prolonged into a
proboscidial snout. The eyes are sessile.

Among the Tectibranchia, the Cephalaspidce are distinguished by peculiarities of
the head. It carries dorsally a flat fleshy disc, the cephalic or tentacular disc
(Fig. 14, p. 10), which is regarded as the result of the fusion of the tentacles, and
which, by its shape, recalls the propodium of the Naticidcc or Olimdac among the
Prosdbramchia. This cephalic disc carries the sessile eyes on its dorsal side, and its
posterior lobe, which is sometimes produced in the shape of two lateral tentacular
processes, shifts about over the anterior portion of the shell. The shape of this disc
varies considerably in details.

Of the very numerous Nudibranchia we shall only notice two extreme forms :
Tcthys and Phyllirrhoe.

In Tcthys, the head takes the form of a large flat disc, almost semicircular in
shape and fringed at the edge ; this carries on its upper surface two conical rhino-
phores, which can be retracted into large sheaths.

In Phyllirhoe (Fig. 19, p. 12), the head is produced into a short proboscidial
snout, which carries only two very long curved tentacles ; the bases of these are
encircled by integumental folds, and they may be considered as rhinophores.

Pteropoda gymnosomata. — The head is distinct, and carries two pairs of
tentacles, one labial and the other nuchal. The former answers to the anterior, and
the latter to the posterior tentacles or rhinophores of the Tectibranchia, especially
those of the Aplysiidce. The nuchal tentacles are generally small or rudimentary,
the rudiments of the eyes lying at their bases.

Nearlyall the Gymnosomata, as highly-developed carnivorous animals, are provided
with a proboscidial snout which, commencing at its tip, can be completely in vagina ted,
and carries at its base, when evaginated, buccal appendages innervated from the
cerebral ganglion.

Definite compensatory relations exist between the proboscidial snout and the
buccal appendages : —

1. When the proboscis is specially long, the buccal appendages are wanting
(Clionopsis).

2. When the proboscis is of median length, it carries suckers at its base, or a
pair of long appendages provided with suckers (Pneumodermidce, Fig. 76, p. 79).

3. When the proboscis is short, there are long anterior tentacles, and at the base

104

COMPARATIVE ANATOMY

CHAP.

of the evaginated proboscis three pairs of conical processes (cephalic cones), with
special nerve endings and glands whose sticky secretion helps in the capture of prey
(Clionidce).

4. The proboscis may be wanting. There is then on each side of the mouth a
long extensible buccal appendage carrying at its base the labial tentacle.

Pteropoda thecosomata. — The head is, as a rule, not sharply separated from the
body, and has no invaginable snout, but one pair of tentacles which answer to
rhinophores, and sometimes lie in sheaths at their bases. The left tentacle may
become rudimentary. In the Thecosomata the male copulatory organ lies on the
upper side of the head, near the tentacle.

3. Pulmonata.

The head is here distinct from the foot ventrally, but passes dorsally into the

neck. It carries two or four tentacles. The Stylom-
liialophom, which are terrestrial, have four tentacles
(Fig. 97), an anterior and a posterior pair. The
posterior, which are usually the longer, carry the eyes
on their tips. The tentacles are hollow tubes filled
with blood and connected with the blood spaces of
the head. They can be invaginated from the very
tip into the head, special muscles acting as retractors
which, when the tentacle is evaginated, run from the
head to the tip of the tentacular cavity.

The Basommatophora, which are aquatic, have
only one pair of tentacles which are usually triangular
FIG. or.— Helix, front view, creep- and flat. They are solid, and not invaginable, but

in- with extended tentacles (after merely contractile. The eyes lie on the inner side of

Howes), s, Shell ; tlt optic tentacle ; t]lejr ljases>

t, anterior tentacle; m, mouth ; ?i, , . „ 7 ,~, ,. „ .,

^ jips In certain Pulinonata (G-landinu, Zonites, Onei-

diiiin) the upper lip may be drawn out into a lobe or

labial palp on each side. ThisUabial palp in Glundina can move very freely, and is
the seat of a fine sense of touch.

On the right, behind the right tentacle, lies the common genital aperture, or, in
cases where the male and female apertures are distinct, the male aperture.

B. Seaphopoda (Fig. 101, p. 113).

In this order the non-invaginable snout is ovoid or barrel-shaped,
and projects from the body, over and in front of the foot, downwards
into the mantle cavity. At its extremity lies the mouth, surrounded by
a circle of dentate oral lobes shaped like oak-leaves, — four on each side.

At the boundary between the bases of the foot and of the snout,
to the right and left of the cerebral ganglion, a shield-shaped lobe
rises from the body on each side ; this is attached, at the centre of its
inner side, by a short slender stalk to the body wall, concrescence also
taking place at its lower edge. This shield carries numerous filamentous
or vermiform glandular tentacles, which move very freely and can be
protruded far beyond the mantle aperture.

The ends of the tentacles are swollen into the shape of a spoon, and can become
attached to foreign objects like suckers. Each swelling has long ciliary hairs on

vii MOLLUSCA—ORAL LOBES OF THE LAMELLIBRANCHIA 105

its concave surface, the cilia being continued in a band all along the, tentacle to its
base. Tentacles of this sort are found in all stages of development ; they rise
chiefly from the inner surface of the shield, and easily become detached or broken
off, and are then regenerated. They are no doubt chiefly useful as organs of touch,
and serve for seizing particles of food (Foraminifera, etc.). They may further assist
respiration in the absence of localised gills, by causing increase of surface. The
tentacles are innervated from the cerebral ganglion through the stalk of the shield
on which they stand.

C. Cephalopoda.

In Nautilus, there are on each side one tentacle above and one
below the eye. It is not improbable that these two tentacles cor-
respond with the two pairs of tentacles in the Gastropoda.

IX. The Oral Lobes of the Lamellibranehia.

The oral aperture of the Lamellibranehia is produced right and
left in the form of a groove, which runs backward along the surface
of the body to the anterior end of the base of the gill, or to some
point near it. This groove is bordered by two projecting ridges
above and below it. The two upper ridges, at the point where they
meet, form a sort of upper lip over the mouth, the lower ridges, in
the same way, forming a lower lip. The groove between the ridges
serves for conducting to the mouth the particles of food which are
swept past the gills by the cilia.

The length of the groove is naturally determined by the distance
between the anterior ends of the gills and the mouth.

The two ridges just described are continued posteriorly in the
shape of thin lamella?, which hang down into the mantle cavity. These
lamella?, between which the groove becomes a deep, narrow cleft, are
the oral lobes or labial palps of the Lamellibranehia. They are more
or less triangular, one side of the triangle forming the base by which
the lobe is attached to the body.

In cases in which the gills lie far behind the oral aperture, the bases of these
lobes are long, but in others, where they begin near the mouth, the bases are short,
and each lobe then usually forms a long, free, pointed process. The surfaces of these
two oral lobes are ciliated, and, further, the surfaces which face each other, i.e.
which have the groove between them, are striated at right angles to their bases.
This striation is caused by parallel ridges, and gives the lobes a superficial resem-
blance to gills. The lobes contain .blood lacuna?, and it is probable that, besides then-
chief function of conducting food to the mouth, they may assist in respiration.

In certain forms, the free edge of the upper lip folds over that of the lower
(Ostrea, Tridacna) ; in others, the two edges are closely apposed and interlocked by
means of processes and folds (Pccfcn, S-pmidylus), so that a closed cavity rises in
front of the mouth, into which the groove brings particles of food from each side.
The edges of the upper and lower labial palps may even grow together (Lima).

Nucula (Fig. 21, p. 14), in which the ctenidium lies far back, and has a very
small respiratory surface, may serve as an example of very highly developed oral lobes,

106 COMPARATIVE ANATOMY CHAP.

which were formerly considered to be gills. The base of the lobe here stretches
along the whole length of the base of the foot, and is further prolonged posteriorly
in the shape of a free appendage with a groove running along it. This process can
be protruded beyond the shell, and probably assists in conducting food to the
mouth.

X. The Foot and the Pedal Glands.

The ventral side of the body in the Mollusca is characterised by the
pronounced development of its musculature, which enables the animal
to creep, a fleshy foot, provided with a flat sole suited for creeping,
distinct from the rest of the body and especially from the head, being
developed. This strong ventral musculature must be considered as the
remains of the dermo-muscular tube of the racial form, which attained
greater development on the ventral side in adaptation to a creeping
manner of life, while it degenerated on the dorsal side, being rendered
functionless and useless by the hard shell.

The flat form of the foot with a sole for creeping must be con-
sidered the primitive form. Such a foot is found in the Chitonida.'
among the Amphineura, in most Gastropoda, and in certain Lamelli-
branchia, especially in the Protobranchia, which for other reasons also
must be considered the most primitive form of Lamellibrancliia.

The musculature of the foot and of all parts which become differ-
entiated from it are innervated from the pedal ganglia or pedal nerve
cords.

The foot may become much modified in adaptation to various
methods of life and of locomotion, — in fact, it may entirely lose all
resemblance to the primitive organ. It may, by constriction or by
the formation of lobes or folds, fall into several parts, of which the
following are the most important :—

1. Proceeding from before backward we have the propodium, an
anterior portion distinct from the rest, and the metapodium behind
the former and seldom very distinct, which carries the operculum
when this is present.

2. From below upward there are the parapodia, lobe-like exten-
sions of the edge of the ventral sole, and the epipodium, a projecting
ridge or fold round the base, i.e. round the upper portion of the foot.
Tentacular processes are often developed on this ridge.

Taking the different groups in order, the following variations of
the foot and the pedal glands (mucous glands and byssus gland) are to
be noted.

A. Amphineura.

(Of. Section II., p. 29). The foot is here not divided into separate consecutive
portions, and there are no parapodia or epipodia.

VII

MOLLUSCA—THE FOOT AND ITS GLANDS

107

B. Gastropoda.

1. Prosobranchia.

With rare exceptions, which will be described later, the foot, which is well
developed in this order, has a simple (undivided) flat sole for creeping.

Propodium. — In a few cases, however, the anterior portion of the foot forms a
propodium well marked off from the rest of the organ. This is especially the case
in the Monotocardia (Olividcc, Harpidcc, certain species of Pyrulida:, Strombidce,
Strombus, Pteroccra, Terebellum, Rostellaria [Fig. 6, p. 6], Xenophoridw [Fig. 5, p. 5],
Nariddce, Naticidcc [Fig. 98]).

Among the above, the propodium is particularly well developed in Oliva, sepa-
rated from the rest of the foot by a transverse furrow and forming a semicircular
disc.

In the large foot of Natica (Fig. 98), the propodium is also very distinct. It
has an anterior lobe which bends back over the shell, and so covers the head.

FIG. 98.— Natica Josephina, with protruded proboscis, from the right side (utter Schiemenz).
1, Propodium ; '2, sucker-like boring appendage of the proboscis (3) with boring gland ; 4, siphon
(here formed by the foot) ; 5, tentacle ; 6, lobe of the metapodium, which usually covers a larg'-
part of the shell from behind, and carries the operculum on its inner side ; 7, metapodium.

Sometimes the propodium forms a sort of siphon on the left side, and in other cases
the lobe which bends back over the shell shows a bulging. Both these arrangements
serve to conduct water to the respiratory cavity. The metapodium also, which,
when swollen and expanded, spreads out widely, carries on its dorsal side a lobe
which bends forward over the shell, and carries the operculum on the side nearest
the shell.

In most Prosobranchia the metapodium carries, on its dorsal side, a horny or
calcareous operculum which serves to close the shell.

Epipodium. — The epipodium is very commonly present in the Diotocardia. It
is most strongly developed in Haliotis (Fig. 105, p. 121), where it surrounds the base
of the foot in the form of a large integumental fold. This fold, which may aptly
be called the ruff, has fringed or digitate appendages as well as long contractile
tentacular processes. The tentacles here, as in other Prosobranchia, are organs of
touch, and may be provided at their bases with so-called lateral organs. In the
Fissurellidcv this epipodial ruff is replaced by a row of numerous tentacles or
papilla, rising on each side from the base of the groove between the base of the foot
and the visceral dome. Among the other Diotocardia also, the epipodium is well

108 COMPARATIVE ANATOMY CHAP.

developed as a simple or fringed border, which carries a few tentacles (usually four
on each side) of varying length (Fig. 3, p. 4). At the base of each tentacle there
is a lateral organ. Eyes are said to occur at the bases of the epipodial tentacles in
Euiiiargerita and Scissurclla.

The epipodium is, as a rule, wanting in Docoglossa, but one is found beset with
papillae in the genus Hclcion, and in Patinclla and Nacella it is fringed ; these
epipodia correspond in position with those of other Diotocardia.

A well-developed epipodium rarely occurs among the Monotocardia , but lanthina
lias a typical epipodial border, and the Litiopidce, and many Eissoidce have an
epipodium with several (1-5) tentacles on each side. Many other Monotocardia
have retained either the anterior or posterior portions of the epipodium.

(a) Anterior vestiges of the epipodium are found in Vcrmetus in two anterior
pedal tentacles, and in Paludinu and Ampullaria in two nuchal lobes, which
must not be confounded with true cephalic tentacles. In Paludi/ia, the right
nuchal lobe, and in Ampullaria the left, forming a longitudinal groove, becomes a
sort of siphon. Cahjptrcea possesses on each side under the neck a semicircular
epipodial fold.

(b) Posterior vestiges of the epipodium are found in Lacuna in the form of an
epipodial fold with a process on each side above the foot. Narica has, above the
metapodium on each side, a wing-like epipodial lobe.

(c) Median and posterior vestiges of the epipodium are found in Choristes, where
there is a median papilla on each side, and posteriorly a pair of tentacles below the
operculum.

The epipodium is always innervated from the pedal nerve cords or the homolo-
gous pedal ganglia, or from the pleural ganglia which separate off from the latter.

The foot of Hippomjx undergoes a curious transformation. Hippwiyx is a
Monotocardian genus, with a conical shell ; the animal attaches itself firmly to
rocks or the shells of other Molluscs, which it excavates, either directly or by
means of a shell plate, which probably answers to the operculum. The median part
of the sole of the foot has lost its muscle layer, and its edge has united with the
edge of the mantle, leaving only an anterior aperture through which the head can
be protruded. On the lower side of the foot, the columellar muscle which descends
from the shell gives rise to a horseshoe -shaped muscular area surrounding the
central non-muscular part.

Without going into details as to the method of locomotion of the Prosobranchia,
it may be stated that most of them creep or attach themselves by means of the flat
sole of the foot.

Heteropoda. — The Heterppoda are pelagic Prosobranchia (Mmiotocardia), which
have exchanged the creeping for the swimming manner of life. The foot has in
them become peculiarly adapted to this new method of locomotion. The propodium
has become changed into a narrow vertical rowing fin (carinate foot), which when
the animal is in its swimming position is turned upward.

The development of this vertical fin can be traced almost step by step within
this division, starting with 0,''y</i/rus, and proceeding through Atlanta and Carinarin
to Ptcrutrnchca. In this series, the typical outer appearance of the Prosobranchia te
(its shell, visceral dome, mantle, and gills, which are still retained in Oxygyrus and
Atlanta], gradually disappears owing to development in another direction.

O.cygyrus (Fig. 99, A) still has the characteristics of a Prosobranchiate. The
foot consists of (1) a propodium, the creeping sole of which has been somewhat
hollowed out or deepened ; anteriorly it possesses a fin- like outgrowth, which is used
as a propelling organ in swimming ; and (2) a distinct metapodium directed backwards
like a tail, and bearing an operculum. The derivation of such a foot from that of
certain Prosobranchia, which have distinct proporlia and metapodia, such as the

VII

MOLLUSGA—THE FOOT AND ITS GLANDS

109

saltatory StrombidcE, is clear. The sole of the foot in Oxygyrus, although it can In-
used for creeping, is looked upon as a sucker.

In Atlanta (B), the arrangements of the foot are similar to those in <//•//(///, •//*.
but the fin-like outgrowth of the propodium has become its most important part,
the comparatively reduced sole or sucker appearing merely as an appendage to it.

In Carinaria (C) both the foot and the general external appearance of the whole

FIG. 99. —Comparative Morphology of the Heteropoda. A, Oxygyrus. B, Atlanta. C,
Carinaria. D, Pterotrachea 9 , adapted from figures by Souleyet. 1, Visceral dome and shell ;
2, head with eyes and tentacles and proboscidal snout (3) ; 4, gills ; 5, foot with sole, which latter in
B and C is reduced to a sucker, and in D is wanting ; (i, tin-like appendage of the foot ; 7, meta-
podium with, S, operculum.

animal are much changed. The metapodium, which here has no operculum, appears
as a mere tail-like posterior prolongation of the body. The fin is much broader and
longer, and the sucker seems to have shifted backward along its free edge.

Finally, in the Pterotrachea (D), the sucker (the original sole of the foot) is still
further reduced, and only present in the male.

The Heteropoda are said to attach themselves occasionally by means of the sucker.

'2. Pulmonata.

The foot is here almost always undivided, and provided with a large fiat sole
for creeping. In a few Auriculidce, however (Melampus, Lcuconia, Blauneria,
Pedipes), it is divided into two portions by a temporary or permanent transverse
groove.

3. Opisthobranchia.

In almost all Opisthobranchia the foot has a well-developed sole for creep-

110

COMPARA TIVE ANA TOMY

CHAP.

ing. There is no division into parts, and the adult rarely (Actceon) carries an
operculum.

The epipodium is wanting.

The parapodia, on the contrary, i.e. lateral lobes or fold-like extensions of the edges
of the sole, are highly developed in many Opisthobranchia (e.g. the Elysiadcc among the
Ascoglossa, and very many Tcdibraiichia, such as the Scaphandridce, Bullidcc, Aplus-
Gastroptendce (Fig. 14, p. 10), Philinidce, Doridiidcc, Aplysiidcc (Fig. 75,
]>. 78), Oxynocidce). The parapodia are often bent
" back over the shell, their edges sometimes touching,

so that the shell may be entirely roofed over by them.
In many forms which are provided with parapodia
(Gastropteridce, Philinidce, Doridiidce, Aplysiidc?) the
mantle also bends back over the shell, more or less
completely covering it. In these cases the shell is
to some extent doubly internal, being covered first
by the mantle and then (not in Philiiu' and Doridium)
by the parapodia (Fig. 100).

The parapodia may fuse posteriorly along their
upturned edges (Aplysiidoe, Oxynoe). In Lobiger each
parapoclium is transversely slit, so that two long
wing-like processes are formed on each side. Many
Opisthobranchia (Aplysiidcc, Oxynoe, Gastropteridce)
can propel themselves through the water by means of
the waving motion of their parapodia. Phyllirhoe
is a Jfudibraiich which appears to have become
adapted to a pelagic swimming manner of life by the
compression of its body into the shape of a long
narrow leaf with sharp dorsal and ventral edges ; it
travels through the water with an undulating motion
(Fig. 19. p. 12). The foot has disappeared.

Pteropoda. — The Pteropoda, which are Tecti-
branchiate Opisthobranchs, have, like the Proso-
branchiatc Heteropoda, become pelagic animals adapted
for swimming.

While in the Heteropoda, the propodium becomes
transformed into a medio-ventral vertical rowing fin,

mantle (dotted, 2), and foot (streak- in the Pteropoda the paired Tectibranchiate parapodia
ed, 3). A, Prosobranchiate with which, as we have already seen, can be used for swim-
ming, develop into the paired fins or wings of these
animals (Figs. 16 and 17, p. 11 ; 87, p. 91).

In the Theeosomata (Fig. 87, p. 91), which must
be derived from Cc.plmlaspidcc (Bulloidca), in which
the parapodia lie on each side as direct prolongations
of the reptant surface of the foot, this organ, i.e. the
foot, has become confined to the anterior end of the
body, and consists of three portions — the median un-
paired mesopodium and the two lateral parapodia or fins. The mesopodium is small,
and the ventral side of it (which corresponds with the sole of the Ccphalaspidce, but
can no longer be used for creeping) is strongly ciliated. The ciliary movement is from
behind forward, i.e. towards the oral aperture which lies anteriorly on the foot,
and no doubt serves for conveying to it the minute marine animals on which the
creature feeds. On the dorsal side of the mesopodium, which projects freely back-
wards, the Limacinidce carry a delicate transparent operculum, which often becomes

Fie;. 100.— Diagrammatic trans-
verse sections of Gastropods, to
illustrate the arrangement of the
sh,ll (black, i), viscera] dome and

outer shell and epipodium (4). B,
Tectibranchiate with lobes (ii) of
the mantle turned back over the
outer surface of the shell. Dorsally
thf shell is still uncovered ; 5, para-
podia ; 7, ctenidiurn. C, Tecti-
branchiate with internal shell, i.e.
completely overgrown by the lobes
of the mantle.

vii MOLLUSCA—THE FOOT AND ITS GLANDS 111

detached.1 The parapodia are large, fin- or wing-like, and anteriorly inserted on each
side of the median portion of the foot ; they unite in front of and above the month.

The Gymnosomata (Fig. 16, p. 11) are to be derived from the Aplysiidiv, in
which the parapodia are not exactly lateral extensions of the sole of the foot, but
arise somewhat above the edge of the sole on each side. This may be explained by
-supposing that they are fused for a certain distance from their bases with the lateral
wall of the body. In the Gymnosomata, also, the foot is distinctly separated from
the two lateral fins or parapodia. The mesopodium and the fins lie anteriorly on
the ventral side of the body, behind the head.

The foot itself, which is distinct from the head, consists of three parts — a pair of
anterior lobes, which converge anteriorly till they unite, and a median posterior lobe
drawn out to a point posteriorly. The fins never unite in front of or above the head.

Pedal glands of the Gastropoda. — Many Gastropods, and especially
most Prosobranchia and Pulmonata, possess, besides the various unicellular
glands scattered over the upper and lower sides of the foot, larger multi-
cellular localised pedal glands. These belong to two morphologically
distinct groups.

1. In the Prosobranchia an anterior pedal gland opens at the
anterior edge of the foot. In those forms in which this anterior edge
is divided into an upper and a lower lip, this " labial gland " opens
between the lips. In the Pulmonata it opens externally between the
head and the foot. It consists of an epithelial tube of varying length,
not infrequently as long as the foot itself ; this tube runs backward
in the median line mostly through the base of the foot ; less frequently
it lies upon this base, projecting into the body cavity.

This tube serves both as reservoir and duct for the numerous
unicellular mucous glands which lie in the surrounding tissue of the foot
and open on its walls. It secretes mucus, though it has been incorrectly
described as an olfactory organ. It undergoes considerable modifica-
tions with regard to its size, the form of its lumen, and the number and
arrangement of its glandular cells.

'2. Among the Prosobranchia, opening on the sole of the foot, there
is commonly found an unpaired gland. Its outer slit-like aperture is
median, and lies behind the anterior edge of the foot. It leads into a
cavity in the foot which serves as a reservoir ; the epithelial wall of
this cavity projects in the form of folds into its lumen. As in the
former case, unicellular glands pour their secretions into it through
ducts which pass between the epithelial cells. This sole gland in the
Prosobranchia has rightly been considered homologous with the byssus
gland of the Lamellibranchia. It is developed in varying degrees, and
not infrequently is altogether wanting. Its slimy secretion forms
threads by means of which many Prosobranchia attach themselves to
objects in the water. Some terrestrial Pulmonata also loAver themselves
from a height (from plants) by means of the tough threads which they
secrete.

1 With regard to the derivation of the Thecusomata from the Cephalaspidic, which, like
other Opisthobranchia, have as a rule no operculum. it must be noted that Actceon, which
is in many respects a primitive Gephdlaspid genus, possesses an operculum.

112 COMPARATIVE ANATOMY CHAP.

Besides these two, other pedal glands are occasionally found. Only one need be
mentioned, which is found in some Opistftobranch/ia (Pleurobranckus, Pleurobranchcca,
PleurophyUidia). It lies at the posterior end of the sole, and consists of glandular
creca, each of which opens separately.

C. Seaphopoda.

The foot of Deidalium (Fig. 101) is almost cylindrical ; it projects
downwards into the tubular mantle cavity, and can be protruded
through its lower aperture. The free end of the foot is conical ; the
base of the cone carries on each side a fold or ridge which has been
compared, with questionable propriety, to an epipodium. These two
lateral folds or ridges encircle the base of the conical end without
uniting either anteriorly or posteriorly. A groove runs along the
anterior middle line of the foot.

In Siphonodentalium both this groove and the lateral lobes are
wanting, and the anterior end of the foot is broadened into a round
disc carrying on its edge small conical papillae.

D. Lamellibranchia.

The foot in this class is, as a rule, laterally compressed, and has a
sharp edge directed downwards and forwards, which can be stretched
out beyond the shell. It may be called hatchet-shaped (Pelecypoda) or
linguiform, and is especially suited for forcing its way into mud by
means of alternate contraction and expansion.

This peculiar shape must be considered as acquired. Originally
the foot of the Lamellibranchia also possessed a flat sole for creeping.
The Protobranchia, in fact, have a foot with a ventral disc (Fig. 21,
p. 14), and so has Pecfmirithis. The edge of this pedal disc is notched
or toothed. When the foot is retracted, this disc folds down the
middle line.

The foot in the Lamellibranchia varies much in details, according
to the manner of life or of locomotion of the animal, and according to
the development of the byssus. One of the special characteristics of
the Lamellibranchiate foot is the gland which secretes the byssus, the
latter being a bundle of tough threads varying in thickness, and
resembling horn in their physical properties. The Lamellibranch, with
these threads, anchors itself to foreign objects. The byssus can
generally be thrown off and replaced by a new one, and many forms
can move about on a smooth perpendicular pane of glass by means of
alternate attachment and rejection of portions of the byssus applied
by means of the foot.

Stationary bivalves, i.e. those attached by one of the shell valves,
are in the first instance attached by means of the byssus, for a byssus
is, as a rule, present in the young stages of those bivalves which do
not possess it as adults.

VII

MOLLUSCA—THE FOOT AND ITS GLANDS

113

P'IG. 101.— Anatomy of Dent-
alium entale, after Leuckart
(Wandtafeln) and Lacaze-Du-
thiers. Tlie right half of the
shell and the lower portion of
the mantle are removed. a,
Pallial nerve running up from
the visceral ganglion ; 6, shell ;
c, space between the mantle and
shell ; d, anus ; e, visceral gan-
glion ; /, mantle cavity ; g,
mantle ; h, lower, t, upper buccal
ganglion ; i, auditory organ ; k,
pedal ganglion, m, lateral folds
of the foot ; n, terminal pedal
cone ; o, filamentous tentacles ;
I, lower edge of the mantle ; p,
leaf -like oral appendages; q,
snout; r, ! cerebral ganglion; s,
shell or columellar muscle, cut
through ; it, right nephridial (and
genital) aperture ; v, digestive
gland (liver) ; w, gonad ; x, upper
end of the columellar muscle ; y,
upper open end of the mantle.

VOL. II

114

COMPARATIVE ANATOMY

CHAP.

The complete byssus apparatus (Fig. 102) consists of : (1) a cavity
in the foot, into which the byssus gland opens ; (2) a duct connecting
this cavity with the exterior ; (3) a groove which runs from the
aperture of the duct along the ventral edge of the foot to its anterior
end ; and (4) a crescent-shaped or cup-like widening of the groove at
its anterior end.

(1) The byssus cavity is divided into narrow shelves by numerous folds, which
project from each side into its lumen. A septum, descending from its roof, further
divides it into two lateral pails. The byssus secretion is yielded partly by the cells
of the epithelial walls, and partly by glandular cells which lie in the surrounding
tissue, their ducts passing between the epithelial cells. The secretion takes the
form of the cavity, and is thus held fast as with roots by the numerous lamella5
which occupy the shelves. As the amount of the secretion in the cavity increases,
these lamellae are pressed into the duct (2), where they unite to form the main stem
of the byssus.

The walls of the groove (3) and its terminal expansion (4) are also glandular.

When a bivalve attaches itself it forms a byssus thread
in this groove, which fuses with the end of the main
stem. The tip of the foot presses against some surface,
such as a rock, and attaches the thread by means of a
cement secreted by the widened terminal portion (4) of
the groove. In this way the main stem of the byssus
may be fastened to a rook by means of numerous threads
successively secreted in the groove.

The relation existing between the development of
the foot and that of the byssal apparatus may be
sketched as follows : —

1. The foot in its primitive form, with a flat sole
and no groove, lias a simple imagination without

•byssus (Solenomya).

2. "With the same foot, a small lamella rises from
the base of the simple invagination ; the byssus is
very slightly developed (Nucula, Lcda).

3. The invagination becomes differentiated into a
cavity and a duct, and the byssus and its glands are

Fio. 10:.'. —Byssus of a Lamel-
libranch with its cavity and
duct. 1, Diagraiiiiantir trans-
verse section through the foot ; 2, strongly developed. In consequence of this the foot
main stein ; 3, terminal threads ceases to be a locomotory organ ; its flat sole
attaching the byssus to a foreign
object.

disappears, and it becomes finger- or tongue-shaped,
often more or less reduced in size, and serves for
attaching the byssus. In very many cases the groove is formed from the cud of the
din!, widening at the tip of the foot as above described. This is especially the case in
forms which anchor themselves by the byssus to stones, plants, or the shells of other
Molluscs. This attachment may be more or less firm, and may be temporary or
permanent (Limidrc, Hp<nidiilid<i\ Pcctinidcc,1 J///////</«, Ari-idte,1 Cnrdltiihr,1 Ery-
I'iniil'c, Galeommidce, Tri<l:n-,ii,i.<\ i 'yprinidce,* Veneridce? Glycymeridce, Mi/iiln,1 etc.)
When the byssus is very highly developed, some of the pedal muscles become
attached to the byssus gland and form the retractors of the byssus.

4. Many Lamellibranchs, iu the adult state, have neither byssus nor byssus
glands, but the cavity, the duct, and even the retractors (e.g. Trigonia) may be

I' rn

vii MOLLUSCA—THE FOOT AND ITS GLAXI* 115

retained. The byssal apparatus may be found, in closely-related forms, sometimes
with and sometimes without the byssus itself. In the latter case the foot is
generally more strongly developed, and serves for locomotion, i.e. for forcing a way
forward into sand or mud, which most of these forms inhabit, or for the saltatory
motion of Triyoida. In these cases it is linguiform, or wedge- or hatchet-shaped
(Arcidce,1 Carditidrc,1 Cyprinidce,1 Tellinidce, Scrobiculariidce, Mijidw,1 Cardiidce,1
Lucinidcc (foot vermiform), Donacidce, etc. ).

5. When the linguiform, or hatchet-shaped, and often bent, foot becomes more
strongly developed as a fleshy and extensible organ, every trace of the byssus and its
apparatus disappears, at least in the adult ( Unionidce, many VeneridoB, Cyrenidce,
Psammobiidcc, Mcsodcrmatidce, Solenidcc, Mactridce}. All these live in mud. The fleshy
foot of the Solenidcc, which is directed forwards, is so strongly developed that it can
often no longer be wholly withdrawn into the shell, which therefore gapes anteriorly.
The foot is thick and linguiform in Solenocurtus ; club-shaped and truncated at the
tip in Pharus, CuUellns, Siliqua, and Ens is ; and cylindrical, with an egg-shaped
tip, in Solen.

6. In forms where one of the valves has become firmly attached to some hard
substance, the foot (the byssus being absent) may become rudimentary (Chamacca),
or may altogether disappear (Ostreidcc). In forms which inhabit mud or excavations
made by themselves in stone, etc., and which surround the body with an accessory
calcareous tube (Gastrochcenidce, Clavagcllidcc), the foot is also reduced to a small,
usually finger-shaped rudiment. The series of boring Pholadidtc is specially
interesting. Pholas has a pestle- or sucker-shaped foot, which, projecting through
the shell cleft, serves to attach the animal while boring. In Pholadidea and
Jouanndia only the young while boring their habitations possess such a foot ;
as soon as they have finished this work the. pedal aperture of the mantle closes, the
anterior cleft of the shell is also closed by means of an accessory shell-piece called
the callum, and the foot completely atrophies, so that the animals are no longer
capable of locomotion.

In the attached Anomia, also, the foot is small : it is of great importance, how-
ever, as bearer of the byssal apparatus. The shelly plug (see p. 63), by means of
which the animal is fastened to the ground, and whicfc occupies the deep notch cut
by the byssus into the right or under valve, must be regarded as a calcified byssus.

Many Lamellibranchs (Crcnella,, Lima, Mod tola} weave a byssus web which they
inhabit like a nest, and which they strengthen by the addition of foreign bodies
attached by byssus threads.

E. Cephalopoda.

The question, what part of the body in Cephalopoda corre-
sponds with the foot of other Mollusca, has led to much discussion and
careful investigation. It may now be considered as pretty well estab-
lished that the foot in Cephalopoda forms : (1) the arms, (2) the siphon.

The arms are considered as lateral processes of a Molluscan foot
which have pushed past the head to the right and left, and have
united in front, so that the head is entirely encircled by the foot, and
the mouth has come to lie in the middle of the ventral pedal surface,
i.e. at the centre of the circle of arms or brachial umbrella. That this
circle of arms is a derivative of the foot is supported by important
anatomical and ontogenetic facts: (1) The arms are innervated from

1 Pro parte.

116

COMPARATIVE ANATOMY

CHAP.

the brachial ganglion, which lies under the oesophagus, and is an
anterior division of the pedal ganglion. (2) The arms do not
occupy, in the embryo, their definitive position round the mouth, but

rise on the ventral side behind the
mouth, between it and the anus, in a
row on each side. These two rows shift
secondarily forward to form the circle of
arms round the mouth. (According to
another view, the arms are cephalic ap-
pendages, comparable with the cephalic
tentacles of the Pteropoda.)

The pedal nature of the siphon or
funnel has rarely been doubted. It is
innervated from the pedal ganglion. Its
two lateral lobes, which in the Nautilus
remain separate throughout life, but in
the Dibranckiata overlap, may be con-
sidered as epipodia. The accompany-
ing figure of a Cephalopod embryo con-
FIO io3.-Embryo of a cephaiopod, fi i}{ ini the radimentary

seen obliquely from the left posterior side _ _ i . } . . ,

(after Grenadier). 1, Mantle ; -2, anus ; siphon is 86611 in the typical position of

3, right ctenidium ; 4, rudimentary epipodia in the shape of two lateral folds

siphon ; 5, auditory organ ; 0, arms ; 7, • i_ i j i_ ±1 £ j

yolk-sac • s left eye. running backward above the toot and

under the visceral dome.

In Nautilus and the Decapoda. (excluding the Loligopsidce) a valve
is present within the siphon. For the form of the siphon, see p. 38.

1. The Arms of the Tetrabranchia (Nautilus).

The "head" of the Nautilus (Fig. 104) carries numerous tentacles placed in*a
circle round the mouth ; these do not rise directly from the integument around the
mouth, but stand upon special lobes which are differently developed in the two
sexes. These lobes may be compared with the arms of the DibrancMa, and the
tentacles they carry, perhaps, with the suckers on those arms. Each tentacle can be
retracted into its own basal portion as into a sheath.

If the head be viewed from the ventral side, so that the mouth appears lying in
the centre of the extended lobes and tentacles, we see in the female (lower figure)
three inner lobes close to the mouth, two lateral and one posterior. The posterior
inner lobe consists of two fused lateral lobes, the line of fusion being indicated by a
lamellated (olfactory ?) organ. It carries twenty-eight tentacles, fourteen on each side.
Each lateral inner lobe carries twelve tentacles. Besides these three inner lobes,
the foot develops a muscular circular fold ; this is particularly thick anteriorly, and
here forms a lobe, the so-called hood (Fig. 32, a, p. 22), which, when the head is
retracted, covers the aperture of the shell like an operculum. The outer circular
fold carries nineteen tentacles on each side.

Besides these tentacles which belong to the foot, there are two more on each side
which probably belong to the head, one lying above and the other below the eye.

In the male Nautilus (upper figure) the posterior inner lobe is rudimentary.
Each of the lateral inner lobes is divided into two portions. In the right lobe, the
anterior portion carries eight tentacles and the posterior (antispadix) four, three of

VII

MOLLUSGA—THE ARMS OF THE CEPHALOPODA

117

which have a common sheath. The anterior portion of the left lobe also carries
eight tentacles, and the posterior portion forms the conical spadix, which, instead

e

FIG. 104.— Circumoral ring of tentacles in Nautilus pompilius (after Lankester and Bourne).
From the oral or ventral side. Upper figure male, lower female. «, Shell ; 1), circular fold or hood
with its tentacles, g ; c, the two lateral inner lobes, in the male the left inner lobe forms the spadix
or hectocotylus p, and the right the antispadix q ; d, the posterior inner lobe, reduced in the male ;
», lamellateil organ (olfactory?) ; e, jaws in the buocal cone ; /, the tentacles of the outer muscular
circular fold ; I, eye ; m, paired lamellated organ ; o, siphon or funnel.

of tentacles, carries imbricated lamellae. This spadix is looked upon as the hecto-
cotylised limb of the Nautilus, and probably takes some part in copulation (see
the Copulatory Apparatus, p. 242).

2. The Arms of the Dibranchia.

The Dibranchia have either eight or ten arms, which stand in a circle round the
mouth and carry two longitudinal rows of suckers (acetabula) ; rows of cirri may
accompany the suckers, and the cirri may here and there become transformed into
hooks or claws (e.g. Onychotcuthis}.

118 COMPARATIVE ANATOMY CHAP.

In many Octopoda, the long arms are connected by means of membranes near
their bases, and occasionally as far as their tips. In the latter case the circle of
arms has the appearance of an umbrella, of which the arms are the ribs. The
mouth lies in the centre. The Octopoda can creep by means of their circle of arms,
the visceral dome standing erect. In this position they may best be compared
with snails, the ventral side of the circle of arms functioning like the sole of the
snail's foot.

The Dccnpoda- have ten arms ; eight of these correspond with the eight arms of
the f'ldojioilit, but are shorter and are never connected by membranes. The two
others, the prehensile tentacles, are inserted between the third and fourth Octopodan
arms on each side and differ from the latter in structure, being long and vermiform,
with swollen ends armed with suckers, hooks, etc. The prehensile tentacles are
very contractile, and in many Decapoda (e.g. Sepia) are concealed in special cavities
of the head when the animal is at rest. These cavities probably correspond morpho-
logically with the water pores, which often occur elsewhere at the bases of the arms
or on the head. When pursuing prey the Decapods dart these tentacles out of their
cavities with great force.

One (less frequently two) of the eight or ten arms of the male Dibranchia is
almost always transformed (hectocotylised) to assist in copulation. In some
Octopoda it even becomes detached from the body and is regenerated.

The hectocotylised arm is, in the Octopoda, usually the third arm on the right
side, and in the Decapod a- the fourth on the left. (The arms are counted from
before backward. )

In the female Argonaut, each arm of the first pair is widened into a sail-like
expansion, which stretches back over the outer surface of the shell.

All Cephalopocls, even the more massive Octopoda, are good swimmers. In swim-
ming, the mantle and funnel play the chief parts. Water is alternately taken into
the mantle cavity through the mantle cleft, and expelled from it forcibly through
the funnel, the reaction propelling the animal backwards. When the water is being
ejected, the mantle cleft is closed by the locking apparatus, so that all the water in
the mantle cavity has to pass out through the funnel. Many Decapoda can also
swim with the head directed forward, the lower (distal) end of the funnel being bent
round, so that the water is expelled in the direction of the visceral dome. In
swimming the arms are apposed to one another, so as to diminish the friction as
much as possible. Some Octopoda, especially those which have interbrachial
membranes, assist themselves in swimming by opening and shutting their circle of
arms like an umbrella.

XT. Swelling- of the Foot (Turgescence].

Imbibition of Water.

The foot in many LameUibranchia and Gastropoda may swell when
it has to be protruded from the shell and used for locomotion. Until
recently opinions varied very much as to the way in which this swell-
ing or expansion took place. Many believed that water was taken
up from without into the blood vascular system or into a special
water vascular system, but there was difference of opinion as to the
manner in Avhich it was taken in. On the one hand it was said to
enter through apertures or pores in the foot, which, however, do not
exist, the only pores found being the apertures of the pedal glands

vii MOLLUSC A— MUSCULATURE 119

(byssus and sole glands). On the other hand, the water was sup-
posed to enter the foot through intercellular ducts between the
epithelial cells, but this theory has also been disproved.

Others, again, maintained that the water was conducted by the
nephridia to the pericardium, and conveyed thence through the blood
vascular system ; but the pericardium has been shown to be entirely
separated from the vascular system. Indeed, many theories on this
subject have been put forward and disproved.

It is now the received opinion that, except in the case of one
animal, which will be presently described, the foot is swelled by a
rush of blood which, flowing into the foot, is prevented from returning
to the body by sphincter muscles.

The exceptional case is that of Natica Joscphina. In this animal
there can be no doubt that water is taken in to swell the foot. The
swelling takes place very quickly — in less than five minutes. When
the foot is stimulated it gives out an amount of water which would
fill the empty Natica shell two or three times. The water is taken in
through very small slits, invisible to the naked eye (probably indeed
through a single very narrow slit, lying at the edge of the foot), and
finds its way to a system of water sinuses, quite distinct from all
other cavities of the foot, and also distinct from the blood vascular
system (which in Natica is closed). There can thus be no question of
a direct taking in of water into the circulatory system. The water
slits at the edge of the foot can be closed by muscles, which extend
from their upper to their lower edges.

XII. Musculature and Endoskeleton.

This chapter has for its subject simply the general musculature of the body. It
would be impossible to describe in detail the musculature of special organs, such as
the intestine, the heart, and the copulatory organs, that of the cutis, or even that of
the most muscular of all the organs— the foot ; since, owing to the varied development
and functions of this organ, its musculature is liable to innumerable modifications.

The character of the general body musculature of the Mollusca is
determined by the degree of development of the shell, whose function
is to protect the soft portions of the body. In order to make this
protection complete, the Molluscan body is, as a rule, though differing
greatly in details, so arranged that the soft parts can be entirely con-
cealed in the shell, which can itself in many cases be closed. The shell
thus functions as skeleton and passive locomotory organ, to which are
attached such muscles as draw the body into the shell by their contrac-
tion, and such as partially or wholly close the shell.

It is obvious that the arrangement of the musculature becomes
much modified secondarily in cases where the shell aborts or altogether
disappears.

The musculature of the Mollusca is not transversely striated.

120 COMPARATIVE ANATOMY CHAP.

A. Amphineura.

The musculature of the Chitonidce has neither been sufficiently
investigated nor systematically described. According to the figures
of various writers on the subject there are — (1) a large longitudinal
muscle mass on each side above the foot; (2) numerous muscle fibres
which run down from the latero-dorsal region and radiate into the
sole ; and (3) the special fibres of the foot, which run through it in
various directions. The muscle fibres mentioned under (2) no doubt
correspond with the shell muscles of the Fissurellidce, etc., and the
columellar muscle of other Gastropods. Some of the fibres descending
from one side cross those from the opposite side. These crossings are
very marked in the median plane between the two pedal nerve
cords.

Among the Solenocjastres, the muscular system of Proneomenia has
been the most thoroughly investigated. In connection, no doubt,
with the degeneration of the foot and the vermiform development
of the body, a kind of dermo-muscular tube has been formed ; its
layers, consisting of muscle fibres running in various directions, are
very thin in comparison with the thick epidermis. This muscular
tube lies immediately under the epidermis. Its outer layer consists
of circular muscle fibres, then follows a layer of diagonal fibres, cross-
ing each other at right angles, but crossing the circular and longi-
tudinal fibres at an angle of 45°. The innermost layer consists of
longitudinal fibres, and is most strongly developed on the ventral
surface on each side of the ventral groove. Groups of fibres are
detached from the circular layer on both sides, and converge towards the
base of the rudimentary foot, some of them crossing above it. The
bundles which arise from the lateral and upper walls of the body run
within the septa which separate the consecutive lateral diverticula of
the intestinal canal.

So far as a comparison between these animals and the Chitonidce is possible, the
abortion of the foot and vermiform shape of the body being taken into account, and
Chitonellus taken as the transition form, it may be assumed that the circular muscle
layer, and in particular the groups of fibres converging towards the foot, correspond
with the dorso-ventral muscles of Chiton, and the longitudinal layer with their lateral
longitudinal muscle masses.

B. Gastropoda.

The only important muscle to be considered in this class is the
columellar muscle. This muscle is attached inside the shell to the
columella, along which it runs on the right side of the visceral dome
and along the right edge of the mantle cavity; it then enters the dorsal
side of the foot in which it spreads out. The columellar muscle acts
as a retractor to withdraw the animal into its shell.

VII

MOLL USCA— MUSCULATURE

121

1. Prosobranehia.

The columellar muscle is here always developed in its typical form.
It is attached at one end to the columella in the last coil of the shell,
and at the other to the operculum, which lies on the dorsal side of the
metapodium.

f

A few Prosobranehia, such as most Fissurellidce, Haliotidce, and Docoglossa, use
their foot chiefly as a sucker for
attaching themselves to some firm
surface. These forms have no
operculum. The columellar muscle
descends vertically into the foot,
and by its contraction presses the

shell against the surface to which
it is attached. In Haliotis (Fig.
105), the ear-shaped shell of which
is coiled, this muscle is cylindrical
and is very highly developed ; it
runs somewhat to the right of the
median plane, at right angles to
the pedal disc, thereby pushing the
mantle cavity and the viscera to
the left. In many Fissure! lit /«.'
and the Docoglossa, the shell has
become cup - shaped and sym-
metrical ; the columellar muscle,
which is very much shortened,
descends direct from the inner
surface of the shell to the foot,
and is no longer cylindrical. The
whole muscle has the form of a
short truncated hollow cone, open
anteriorly, which is attached to
the shell by its upper horseshoe-
shaped sectional surface, and, by
its base of the same shape, to the
sucker-like foot. The viscera are
contained in its hollow axis (Fig.
106). The same arrangement
occurs in all cases where the shell

P.

FIG. 105.— Haliotis, from above, after removal of the
shell, the mantle, and the entire dorsal integument (after
Wegmann). t, Snout ; s and p, salivary glands ; pi, lateral
pockets of the oesophagus ; i, mid-gut ; a, oesophagus ; r,
rectum; e, stomach with csecum (c); li, digestive gland
(liver), its right-hand portion which lies next the large
columellar muscle (?») is covered by the genital gland.
A fringed epipodium encircles the body.

is flatly conical, cup- or plate-shaped, as in the Hipponycidce and the Capulidce
among the Jlonotocardia.

Heteropoda. — In this order, in which the atrophy of the shell, the transformation
of the foot, and the gradual obliteration of all resemblance to a Gastropod can be
traced, step by step, the musculature deserves special attention.

In Atlanta, where the head and foot can still be completely withdrawn into the
shell, the columellar muscle retains its typical form. It descends from the shell,
dividing into three strands ; the strongest median strand stretches out into the fin and
the sucker, the posterior into the operculum-bearing metapodium, and the anterior,
which is the smallest, into the head and snout.

122

COMPARATIVE ANATOMY

CHAP.

eer

em/

The cutis in Atlanta is still comparatively thin. The network of muscles lying
immediately beneath it is not more strongly developed than in other Gastropods.
A special system of crossing muscle fibres independent of the other dermal muscula-
ture lies on each side under the cutis of the
tin. This is the case in all Hetcro'poda.
The integument greatly increases in
thickness in the typical Heteropoda
(< 'in-iiiiii-in, Picrotrachca), and the sub-
cutaneous muscular tube also grows
thicker. Over the body the latter con-
sists of two superimposed layers of fibres
crossing one another diagonally. In the
outer layer, the fibres run from above
anteriorly downwards posteriorly, and in
the inner from below anteriorly to above
posteriori}*. On the head and snout, the
visceral dome and the tail-like metapo-
diurn, the diagonal fibres of both layers
nm longitudinally. In addition, an ex-
ternal circular musculature is found in
FIG. lOO.-Patella, from above, after removal C'arinaria nearly all over the body, in
of the shell (after Lankester). c. The separate _. 7 , . , ,

bundles of the shell muscle, the section of which Merotrachea only m the snout,
is horseshoe-shaped ; I, pericardium ; Ix, fibrous Turning now to Carinariu, which

septum behind the same ; n, digestive gland ; hit, still possesses a delicate, easily detachable
intestine ; k, larger right nephridium ; i, smaller sjleu covering the visceral dome, but tm-

left ditto; c, mantle border, widening anteriorly i i i j_i ^j.i i j

J able to protect any other part of the body,
into the mantle fold ; ecr, em, edge of mantle. J J

we find the columellar muscle still per-
sisting in the form of two bands descending from the visceral dome into the fin and
radiating out to its edge.

In Pterotrachca, where the shell is wanting and the visceral dome rudimentary,
the columellar muscle is also reduced. It has now no connection with the visceral
dome, and commences half-way up the body wall as three short bands running down
into the fin and radiating out to its edge.

The columellar muscle, which originally served for drawing the foot back into
the shell, now serves chiefly to bring about the lateral movements of the vertical
rowing fin into which the foot has been transformed.

2. Opisthobranchia.

The columellar muscle is well developed in forms possessing a shell into
which the body can be partly or wholly withdrawn. Where, however, the shell
is rudimentary or wanting, as is the case with most Opisthobranchia, the
columellar muscle atrophies or perhaps forms part of the pedal musculature. The
subcutaneous dermo- muscular tube, on the other hand, develops in proportion
to the activity of the animal. It consists of longitudinal, circular, and diagonal
muscle fibres, which occasionally form a regular network. The pedal musculature
is merely a thickened portion of this dermo-muscular tube in which longitudinal
fibres predominate. The development of the musculature varies much in detail.
Where movable or contractile dorsal appendages, gills, oral lobes, oral discs, para-
podia, etc., are developed, their musculature is detached from the dermo-muscular
layer, and the latter, in combination with the occasionally tough skin, forms a
passive organ of support for the former.

A columellar muscle is further found in the Pti'ropoda thccosomata. It is ventral

VII

MOLL USCA—MUSCULA TURE

123

ITJl

in the Limaciin'ilcr, but dorsal in the Cavoliniidce, in which family the body, as
compared with the head, seems to have been twisted through an angle of 180°
(p. 80). The muscle divides anteriorly into two lateral branches, which radiate
out into the tins.

3. Pulmonata.

In the shell-bearing Pulmonata, the columellar muscle is strongly
developed. It is paired, and attached
at one end by many roots to the foot,
behind the buccal mass, and at the
other to the columella of the first coil
or whorl of the shell. It gives oft'
three subsidiary branches — (1) the
retractor muscles of the optic and
other tentacles ; (2) the retractors of
the buccal mass ; (3) muscles running
to the viscera.

In the Daudebardia and TcstaccUiihc, in

which the dwindling visceral dome with the FlG- lOT.-Shell of Helix, in longitudinal

... M. i i,-f4. i 4-1 section through the columellar axis (after

shell which covers it have shifted to the Howeg)_ ^ Columel]a . ^ colnmel]ai.

posterior end of the body, and in which all muscle . p> edge of oral aperture (peritreme).
possibility of the retraction of the body into

the shell has ceased, only parts of the columellar muscle are retained, and naturally
those parts which are still functional. In the Daudebardia and Tcstaccllidce these
are the retractors of the tentacles, and in Daudebardia the retractors of the pharynx.
The tentacular and pharyngeal retractors are distinct.

The retractors of the tentacles, in Daudebardia rufa, run back separately to the
base of the visceral dome, not entering it, but fusing with the body wall on each
side of it. In D. saulcyi the retractors do not run so far back, but the two on the
right fuse with the two on the left, and pass into the pedal musculature in the
anterior half of the body. The same is the case in the Testacellidce.

The Retractors of the Pharynx. — In Daudebardia rufa there are found, attached
to the pharynx, two retractors, which, passing through the cesophageal nerve ring,
fuse to form one muscle, which runs back along the base of the pharyngeal cavity
somewhat to the left, then ascends in the visceral dome to be attached to the
columella of the last coil of the shell. In D. saulcyi, where there is no projecting
visceral dome, and the shell merely covers a mantle cavity, the cesophageal re-
tractors, which are not in this case fused together, no longer run up into the shell,
but end in the middle of the body, where they enter the pedal musculature.

The numerous cesophageal retractors which, in Testacella, are arranged in two
asymmetrical rows, cannot for several reasons be considered as the remains of a
columellar muscle.

Oncidium when adult has neither shell nor columellar muscle, but its shell-
bearing larva also possesses a columellar muscle.

C. Seaphopoda.

In Dentalium (Fig. 101, p. 113) two closely contiguous muscle
bands run on each side along the anterior side of the body, and are
attached anteriorly to the dorsal end of the tubular shell. At the

124 COMPARATIVE ANATOMY CHAP.

base of the foot these two bands unite to form a single muscle on each
side, which enters the foot and radiates out through it in the form of
numerous longitudinal bundles. This then is a paired eolumellar
muscle which retracts the foot, and draws the whole of the lower
portion of the body back into the upper part of the shell.

D. Lamellibranehia.

The two principal groups of muscles to be considered in this
class are :—

1. The pallial musculature.

2. The pedal musculature.

The former is principally developed near the free edge of the
mantle, and consists of three systems: (1) Fibres which run in the
plane of the mantle fold towards and at right angles to its edge.
These are, in the narrower sense, the muscles of the pallial edge,
and leave on the shell the scar known as the pallial line. (2)
Fibres running parallel with the edge of the mantle. (3) Short trans-
verse fibres running more or less straight between the inner and the
outer surfaces of the mantle. In the siphons, which are formed from
the mantle, these three systems become circular, longitudinal, and
radial layers. The retractors of the siphons are a special differentia-
tion of the pallial musculature ; their development is in direct relation
to the size of the siphons ; their crests of attachment to the shell
valves cause the scar known as the pallial sinus ((/. p. 64).

The important adductor muscles for closing the shell must
also be regarded as differentiations of the pallial musculature. These
are exceedingly thick and powerful and run transversely from the
inner surface of one valve to the corresponding surface of the other
valve. They counteract the ligament at the hinge, their contraction
causing the two valves to approach one another, till the shell is closed.
These adductors leave scars on the inner surfaces of the valves.
Typically, there are two adductors, an anterior and a posterior
(Dimyaria), situated nearer the dorsal than the ventral edge of the
valves. In the Mytilacm, the posterior adductor is larger than the
anterior (Heteromyaria as opposed to Isomyaria). In one large
series of forms the anterior adductor completely atrophies, and the
posterior adductor, which is all the more strongly developed, shifts
forwards towards the middle of the shell. These forms are* grouped
together as Monomyaria ; but this is no natural group, since nearly-
related forms (e.g. within the Mudleriacea) may possess either one or
two adductors, and widely different forms (e.g. Tndacna, Anornia,
Miielleria, Aspen/ilium) agree in having only one. The Anomiidce,
Ostreidce, Spondylidce, Limidce, Pectinidce, Avicitlidcc, Miielleridce, etc., are
Monomyarian.

The adductor oftwi (e.g. Pectcn, Ostrca, Nucula) consists of two apparently
different parts, one containing smooth fibres and the other fibres which appear

VII

MOLL USCA—MUSGULA TUEE

125

transversely striated, although their striation does not correspond with that of
Arthropod and Vertebrate muscles.

The pedal musculature, taken as a whole, answers to the
columellar muscle of other Molluscs, especially of the Gastropoda.
It consists of symmetrical pairs of nmscles attached at one end to the
inner surface of the shell on which they leave impressions, the other
ends entering the foot. The correspondence of this musculature with
the columellar muscle of the Gastropoda is best seen by comparing a
Protobranchiate with Patella or Fissurella. In Nucula or Leda, for
example, there is an almost continuous series of muscle bundles
running down to the foot on each side between the anterior and

O

posterior adductors. The two series taken together, seen from above
or below, have an oval outline answering to the horseshoe-shaped or
almost oval form of the section of the columellar muscle in Patella
(Fig. 106) or Fissurella.

In most cases in which the foot is developed, the following muscles on each sido
are distinguished in order from before backwards (cf. Fig. 108) : (1) the protractor

AA

--OB

FIG. 10S.— Pliodon Spekei, from the left (after Felseneer). The shell, mantle, gills, and oral
lobes of the left sides removed. AA, Anterior ; AP, posterior adductor ; OA, anal ; OB, branchial
aperture of the siphon ; V, visceral mass ; p, foot ; 1, protractor pedis ; 2, retractor pedis anterior ;
3, elevator pedis ; 4, retractor pedis posterior.

pedis ; (2) the anterior retractor pedis ; (3) the elevator pedis, and (4) the posterior
retractor pedis.

Where there is a byssus, the posterior retractor becomes the byssus muscle. It
is then usually highly developed, runs far forward, and may break up into several
bundles.

In those cases in which the foot is rudimentary and the byssus wanting, the pedal
muscles degenerate.

In Pcctcn the pedal retractors are asymmetrically attached, i.e. only to the
left valve. The same is the case in Anomia, where the shelly plug which lies in
the byssus notch, of the right valve, and corresponds with the byssus, is attached to
the left (or physiologically upper) valve by two highly-developed retractors. These
two muscles leave scars near that of the adductors. This fact gave rise to the
erroneous opinion that the Anomia were Triiny/iria.

126 COMPARATIVE ANATOMY CHAP.

E. Cephalopoda.

In the Cephalopoda, a cartilaginous endoskeleton is developed.
This not only serves for the attachment of various muscles and
muscular membranes, but is also a protection for important organs,
especially for the central portion of the nervous system and the eyes.
Of the different cartilages forming this endoskeleton the onl}*" constant
one is the cephalic cartilage.

1. Tetrabranehia (Nautilus).

Nautilus possesses only the cephalic cartilage. This is shaped
somewhat like an X, with thick limbs. The oesophagus runs up
between the one pair of limbs, the other pair serving as supports for
the funnel and as surfaces of attachment for its muscles.

The most important of the muscles is the large paired shell
muscle, which corresponds with the columellar muscle of other
Molluscs. It arises from the cephalic cartilage, and runs on each
side into the band (annulus), by which the body of the Nautilus is
attached to the inner wall of the body-chamber (cf. Fig. 32, p. 22), and,
like the band itself, is attached to the shell. The muscle leaves a
deep scar on the shell (the lobate sutural line). From the lateral
edges of the cephalic cartilage, especially that portion of it which
supports the funnel, a broad muscle-band, the museulus eollaris, runs
forward on each side embracing the nuchal part of the body. The
two unite on the neck to form the muscular nuchal plate. The
ventral lower side of the cephalic cartilage serves for the attachment
of the musculature of the tentacles.

2. Dibranehia.

The cartilaginous skeleton is much more developed than in

Nautilus, owing perhaps, to some ex-
tent, to the atrophy of the shell. Fins,
with their supporting cartilages, for
example, are developed only in those
forms with internal, degenerated
shells.

The cephalic cartilage (Fig. 109) is every -
FIG. 100.— Cephalic cartilage of Sepia, where well developed. It encloses all those
1. Central aperture through which the oeso- central portions of the nervous system which
phagus passes ; 2, preorbital cartilage ; 3, are C1-Owded round the cesophagus, being in

chamber for the eye; 4, ««*i^n^ the form of a hoUow ciiculftr capsrde travBraed

auditory capsule.

by the cesophagus. Processes of this cartilage

assist in supporting the eyes, and in conjunction with independent, preorbital
cartilages form a kind of cartilaginous eye socket. A basibrachial cartilage is
found at the base of the anterior arms in some Dccapoda. We have further to

VII

MOLL USC A —MUSCULATURE

127

mention the nuchal cartilage and the
cavity (p. 55). In the diaphragm,
over which the mantle depends,

cartilages for locking the cleft of the mantle
i.e. in the posterior wall of the visceral dome,
there is in the Decapoda a cartilage near the

funnel, the diaphragm cartilage. Finally must be mentioned a dorsal cartilage,

It lies, posteriorly, on the anterior

which is specially strongly developed in
border of the mantle, where the latter pro-
jects over the neck ; it bears the same
relation to the nuchal cartilage as does
the cartilaginous projection on each side of
the mantle to the cup-shaped socket at each
side of the base of the funnel or siphon (cf.
Fig. 80).

In Sepia the dorsal cartilage is continued
in the shape of a cartilaginous rod running
up on each edge of the shell. The inner
edges of these rods have a groove into which
the edge of the shell fits, and thus form a
kind of fold round its lateral edges.

In the Octopoda there is a cartilaginous
band on each side in the dorsal integument
which may correspond with the dorsal carti-
laginous rods in Sepia. It is possible that
the "internal shell" of the only Octopod
in which a shell is found, viz. Cirrho-
teuthis, is not in reality homologous with
the shell of the Decapoda, but corresponds
with the cartilaginous bands of Octopus
fused in the middle line.

The (basipterygial) cartilages, univers-
ally found at the bases of the fins in the
Decapoda, complete the list.

With regard to the musculature
of the Dibranchia, that of the mantle,
the fins, and the arms cannot be
described in detail. We note, how-
ever, that the pallial musculature
is principally attached to the shell
or to the dorsal cartilage, the fin-
musculature to the fin-cartilages, and
the brachial musculature to the an-
terior side of the cephalic cartilage,
and partly to the basi-brachial carti-
lage when such is present.

The remaining musculature can be best explained with the assist-
ance of the accompanying diagram (Fig. 110), which represents the
musculature of Enoploteutlds.

The strong paired depressor infundibuli (1) rises from the shell
on each side (or from the dorsal cartilage), and runs downwards and
backwards to the base of the funnel and to the cartilaginous socket.
From it spring most of the muscles of the anterior wall of the funnel.

FIG. 110.— Diagram of the more important
parts of the Dibranchiate musculature.

Body seen from the left side. i>, Ventral ; <?,
dorsal; o, anterior; p, posterior ; 1, depressor
infundibuli; 2, retractor capitis lateralis; 3,
retractor capitis medianus ; 4, musculus col-
laris ; 5, adductor infundibuli ; 6, shell ; 7,
dorsal cartilage ; S, nuchal cartilage ; 0, cephalic
cartilage ; 10, mantle cavity ; 11, cartilaginous
socket of the locking apparatus on the posterior
wall of the visceral dome ; 12, corresponding
cartilaginous knob on the inner wall of the
mantle, which fits into 11 ; 13, funnel or siphon
(infundibulum) ; 14, diaphragm cartilage.

128 COMPARATIVE ANATOMY CHAP.

The retractor eapitis lateralis (2), which is also paired, rises from
the same point as the depressor infundibuli, runs into the head, and
is attached to the cephalic cartilage. The retractor eapitis medianus
(3), originally paired, but usually become single by fusion, arises at
the posterior (inner) side of the shell, and also runs into the head,
and is attached to the cephalic cartilage.

In the Dibranchia, the first muscles which fuse are the two median retractors of
the head (Onychoteulhis), these then fuse more completely with the lateral retractors
(Ommastrcphes, Scpiotcuthis, Loligo, Srpiola), so that finally (Sepia) the whole of the
musculature running from the shell into the head forms a muscular sheath open
posteriorly. This sheath encloses the lower portion of the visceral cavity, which is
principally occupied by the digestive gland or liver, and thus forms a kind of
muscular hepatic capsule. The posterior opening in this capsule may finally become
completely closed by the depressor infundibuli, in that, on the one hand, its
anterior edges fuse with the posterior and median edges of the capsule, and, on the
other, it sends out numerous muscles to the diaphragm, forming the diaphragma
musculare.

The muscular hepatic capsule, i.e. all the muscles forming it, the
retractors of the head and the depressors of the siphon, may without
doubt be accepted as the homologue of the columellar muscle of other
Molluscs. Like the latter, they run down from the shell or its vicinity
into the head and foot (represented by the siphon).

The adductors of the funnel (5) have still to be mentioned.
They rise from the cephalic cartilage and run upwards and backwards
to the funnel. Finally, the eollaris (4) is a strong muscle which runs
forwards right and left from the wall of the funnel, and is attached
to the lateral edges of the nuchal cartilage. In the Octopoda and Sepiola,
where a pallio-nuchal concrescence (cf. pp. 54, 55) has rendered a
nuchal locking cartilage unnecessary, the eollaris passes uninterruptedly
over the neck like a saddle, forming a closed circle round the nuchal
portion of the body.

XIII. The Nervous System.

(As a general introduction to this section the reader may be referred to pp. 27, 28.)

A. Amphineura.

The nervous system of the Amphineura is very significant from
the point of view of the comparative anatomist. Its most important
peculiarities may be briefly described as follows :—

1. The gang-Home cells are either not at all or not exclusively
localised in definite ganglia.

2. Four nerve cords run through the body from before backward.
These contain not only nerve fibres, but ganglion cells distributed along
their whole length. They might suitably be called medullary cords,
and must be considered as belonging to the central nervous system.

vii MOLLUSCA—THE NERVOUS SYSTEM 129

One pair of these cords run along the body laterally, these are the
lateral or pleurovisceral cords ; the second lie ventrally, and are the
pedal cords. The visceral and the pedal cords of each side unite
anteriorly, and when so united become connected with those on the
opposite side by a transverse commissure, which runs in front of and
over the oesophagus and contains ganglion cells ; this is the cerebral
or upper half of the cesophageal ring. The pleurovisceral cords unite
posteriorly above the rectum, forming a visceral loop. The pedal
cords are connected both inter se and with the pleurovisceral cords by
anastomoses, so that the whole nervous system strikingly recalls the
ladder nervous system of the Turbellaria and Trematoda.

a. Chitonidse (Figs. Ill and 51, p. 40). — The scheme just given
is founded upon the nervous system of Chiton. The typical ganglia of
the central nervous system of the Mollusca are not yet, in Chiton, found
as distinct ganglia united by means of commissures and connectives,
but the ganglion cells are equally distributed along the commissures
and connectives, an arrangement which is probably primitive. The
upper cesophageal ring thus corresponds with the cerebral ganglia and
the commissures connecting them, and in the same way the pedal
cords contain the whole central portion of the pedal nervous system,
and the pleurovisceral cords the central portion of the visceral, pallial,
and branchial nervous systems. Only in one single species of Chiton
(C. rubicundus) two distinct (cerebral) ganglia occur near each other
in the middle line in the upper half of the oesophageal ring.

Looking more closely at the nervous system of the Chitonidse, we have to
observe : (1) the arrangement of the cesophageal ring and the medullary cords ; (2)
the peripheral ganglia ; (3) the nerves of the ladder-like nervous system ; (4) the nerves
running from the central nervous system (cesophageal ring and medullary cords).

1. Form and arrangement of the central nervous system. — The visceral cords
run back one on each side in the latei-al body wall above the branchial groove ; these
two cords unite above the anus. The pedal cords run in the dorsal part of the
pedal musculature somewhat near one another, from before backward, to end without
uniting where the rectum commences. The cesophageal ring consists, in the first
place, of the semicircular portion mentioned above, which, on account of the peculiar
shape of the body of the Chiton, lies in the same plane as the visceral cords. Poste-
riorly, each limb of this semicircle divides up into the pedal and visceral cords. At
the point where the pedal cord rises from the ring, a cord with a thickened base
separates from it and runs inwards ; this, uniting below the mouth with a similar
cord from the other side, forms the lower half of the cesophageal ring. The upper
and lower halves together form the closed cesophageal ring.

2. Besides this central nervous system there are peripheral ganglia connected
with it by nerve cords consisting only of nerve fibres.

(a) The buccal ganglia together form a horseshoe-shaped ganglionic mass below
the oesophagus, which mass is connected on each side by the cerebrobuccal connective
with the thickened portion of the lower cesophageal ring. The buccal ganglionic
mass in C. rubicundus divides into two paired ganglia and one unpaired ganglion
joined to one another by connectives. The buccal ganglia innervate the oesophagus
as far as the stomach and also the oral aperture.

(b) On each side, from the lower half of the cesophageal ring, somewhat further
VOL. II K

130 COMPARATIVE ANATOMY CHAP.

in than the buccal connective, a nerve (the subradular connective) rises and runs

Es

FIG. 111.— Diagram of the nervous system of Chiton siculus (after Bela Haller). The mantle
removed on the right side. In the centre and to the left the upper part of the foot removed, to
expose the pedal nervous system. F, Foot ; K, last gill ; .-1, anus ; 0, upper, U, lower half of the
resophageal ring ; 1, 2, nerves of the cesophageal ring ; r, connective to the anterior visceral ganglia ;
p, connective to the ganglia of the subradular organ n (above on the left) ; Es, pleurovisceral and
pedal cords ; mn, gastric nerve ; So, point of attachment of the sphincter oris ; n (below on the
right), J<i, •/(._>, nephridial nerves ; m, pallial nerves ; p (to the right below), cardial nerves; v, a
dorsal nerve of one of the pedal cords. The commissures between the pedal cords are seen, and
the nerves running outwards from the latter.

forward and inward to the subradular ganglion. This ganglion lies in the sub-

vn MOLLUSCA—THE NERVOUS SYSTEM 131

radular organ which is situated on the floor of the buccal cavity. The two sub-
radular ganglia are united by a short commissure.

(c) Two small gastric ganglia, connected by a fine commissure, lie at the anterior
end of the stomach, and are joined on each side to the anterior end of the visceral
cord by a long connective.

3. The nerves of the ladder-like nervous system. — The two pedal cords are con-
nected by anastomosing commissures along their whole length, but no nerves are
given off' by these commissures to the pedal musculature. In Chiton rubicundus tin-
visceral and pedal cords are united by numerous connectives, which, in other
Chitonida?, appear either to be wanting or to be reduced to one single anterior or
posterior anastomosis.

4. The nerves running from the central nervous system :—

(«) Nerves of the cesophageal ring. — Numerous nerves rise from the upper or
cerebral portion of the cesophageal ring to innervate the cephalic part of the mantle,
the snout, the upper and lower lips, the gustatory buds on the lower wall of the
oral cavity, and the musculature of the buccal mass. The lower portion of the
(csophageal ring, besides the connectives to the buccal and subradular ganglia,
sends off from its median portion another pair of nerves, which run along the base
of the buccal cavity.

(b) Nerves of the pleurovisceral cords. — Each of the pleurovisceral cords
gives off two nerves to each gill. Besides these they send many nerves to the
mantle, and, posteriorly, nerves which enter the body cavity, probably running to
the kidneys and the heart.

(c) Nerves of the pedal cords. — The pedal cords give off on each side seven or
eight nerves outwards to the lateral musculature of the body, and specially numerous
nerves run down from it to the pedal musculature (inner and outer pedal nerves).
These pedal nerves are richly branched, and, anastomosing with one another, form
a complete neural network in the foot.

b. Solenogastres. — The central nervous system of the Solenogastres
differs from that of the Chitonida' principally in a tendency to form
distinct ganglia ; the pedal and pleuroviseeral cords, nevertheless,
still retain their outer coating of ganglion cells along their whole
length. Fig. 112 is a diagrammatic representation of the structure of
the nervous system of Proneomenia Sluiteri. The fused cerebral ganglia
in the middle line are very large. On both the pleurovisceral and
the pedal cords ganglionic swellings can be distinguished : (1) three
pairs of posterior visceral ganglia ; (2) two anterior pedal ganglia.

The posterior visceral ganglia are connected by cords, which run
transversely over the rectum and correspond, to some extent at least,
with the loop by which the two visceral strands in Chiton are united.

The two anterior pedal ganglia are connected by a strong trans-
verse commissure, which may correspond with the ventral half of the
cesophageal ring of Chiton.

Farther, the pleurovisceral cords are joined with the pedal cords,
and the latter are also connected inter se by transverse connections
along their whole length. The pleurovisceral cords likewise are con-
nected by arched transverse commissures.1

1 These connectives and commissures, however, do not seem to run uninterruptedly
from one cord to the other.

132

COMPARATIVE ANATOMY

CHAP.

On each side of the cerebral ganglion, a nerve rises, which runs to
a ganglion below the pharynx and behind the radular sheath, this is
the sublingual ganglion ; this latter is united with the corresponding
ganglion on the other side by a short transverse commissure. These
sublingual ganglia probably correspond with the buccal ganglia of
Chiton.

Dondersia is specially noteworthy because distinct ganglionic swellings occur at
regular intervals along the pedal cords ; this is particularly marked in the anterior
part of the body. The equally regularly repeated transverse commissures joining the
pedal cords, and the connectives between the pedal and visceral cords, start from
these distinct ganglia.

In Lepidomenia hysti-ix, one ganglion occurs posteriorly and one anteriorly in
each longitudinal trunk (whether pleurovisceral or pedal), and each is connected
with a similar ganglion of the opposite side by a transverse commissure. •

In Neomenia and Cha-torli'nii". no connectives between the visceral and pedal

10

FIG. 11-2.— Nervous system of Proneomenia Sluiteri (original drawing by J. Heuscher).
1, Cerebral ganglia ; 2, pleurovisceral cords ; 3, 4, 5, posterior ganglia of the pleurovisceral cords ;
6^ sublingual ganglia; 7, anterior pedal ganglia; 8, right pedal cord; 9, left pedal cord; 10, 11,
strong posterior commissures between the pedal cords ; 12, anterior pedal commissure ; 13, sub-
lingual commissure.

cords have been observed, and, so far as is at present known, in Cha-todcrma, the
commissures between the pedal cords are also wanting. Further, in Cha'toderma,
the visceral and pedal cords of each side unite together posteriorly to form one
single cord, which becomes connected with the similar cord on the other side by a
transverse cord which runs over the cloaca.1

B. Gastropoda.

The nervous system of the Gastropoda is of great interest to the
comparative anatomist on account of the crossing of the pleurovisceral
connectives in the Prosobranchia, which will be further described in this

section.

The nervous system of this class consists typically of those parts
which we have already mentioned in our scheme of the organisation of
the Mollusca, viz. :—

1 For further details see Simrotli's new edition of Bromi's Klassen und Ordnungen
des Thier-reiches, vol. iii.

vii MOLLUSCA—THE NERVOUS SYSTEM 133

1. Two cerebral ganglia near or above the oesophagus, which
are connected by a cerebral commissure.

2. Two pedal ganglia below the oesophagus, connected with each
other by a pedal commissure, and with the cerebral ganglia by two
cerebropedal connectives.

The cerebral and pedal ganglia with the commissures and con-
nectives belonging to them form a ring encircling the oesophagus,
which may be compared with the oesophageal ring of the Annulata
and Arthropoda.

3. Two pleural or pallial ganglia (between the cerebral and
pedal ganglia), which are connected with the cerebral ganglia by two
eerebropleural, and with the pedal ganglia by two pleuropedal con-
nectives.

4. A simple or complex visceral ganglion lying below the in-
testine, united to the pleural ganglia by two pleuroviseeral con-
nectives.

5. A ganglion, which may be called parietal, almost always occurs'
in the course of each pleuroviseeral connective. The parietal ganglion
divides the connective into two parts, an anterior pleuroparietal and
a posterior viseeroparietal connective.

The cerebral, pedal, and pleural ganglia are (with unimportant
exceptions) always arranged symmetrically to the median plane in all
Gastropoda. The pleuroviseeral connectives and their ganglia, how-
ever, are only found in such a position in some Gastropoda, In fact,
only in the Opisthobranchia (including the Pteropoda but excepting Actceon)
and the Pulmonata are they symmetrical, in the sense that the right
connective and its ganglion lie entirely on the right, and the left
connective and its ganglion entirely on the left side of the body.
The Opisthobranchia and Pulmonata are therefore called euthyneurous
Gastropoda.

In the Prosobranchia and Actceon, the pleuroviseeral connectives are
asymmetrical, inasmuch as they cross one another, the connective
springing from the right pleural ganglion running over the intestine to
the left before joining the visceral ganglion, while the connective
from the left pleural ganglion runs under the intestine to the right side
of the body. In consequence of this crossing, the parietal ganglion of
the connective which springs from the right pleural ganglion becomes
the supraintestinal ganglion, which lies on the left side, and the
parietal ganglion of the connective springing from the left pleural
ganglion becomes the infra-intestinal ganglion which lies on the right
side. The ProsobranchiaaxLd Actceon are thus streptoneurous Gastropoda.

The Areas of Innervation of the various Ganglia.

1. The cerebral ganglia innervate the eyes, the auditory organs,
the tentacles, the snout or proboscis, the lips, the motor muscles of the
proboscis and buccal mass, and the body walls lying at the base of the

134 COMPARATIVE ANATOMY CHAP.

snout. Even when the auditory organs are found in close proximity
to the pedal ganglia, or in close contact with them, they receive their
nerves from the cerebral and not from the pedal ganglia.

2. The pedal ganglia supply nerves to the musculature of the foot,
and occasionally to the columellar muscle also (Patella).

3. The pleural ganglia send nerves chiefly to the mantle, the
columellar muscle, and the body walls lying behind the head.

4. The parietal ganglia innervate the ctenidia and osphradium,
and also send some nerves to the mantle.

5. The visceral ganglia supply nerves to the viscera. The con-
nectives and commissures also may give oft" nerves which belong to the
areas innervated by the neighbouring ganglia.

6. The buccal ganglia, which will be described below, innervate
the muscles of the pharynx, the salivary glands, the oesophagus, the
anterior aorta, etc.

A comparison of the typical nervous system of the Gastropoda with that of the
Amphinrura reveals the following homologies : —

1. The cerebral ganglia of the Gastropoda correspond with the cesophageal ring
of Chiton, with the exception of the central portion of its lower half ; and further
with the cerebral ganglia of the Solenogastres.

2. The pedal ganglia of the Gastropoda answer to the pedal cords in the A-HI-
pkineura, concentrated each into a single ganglion. The arrangement in the Dioto-
cardia, which are the more primitive Prosobranchia, is very interesting in this con-
nection ; in the Diotocardia the pedal ganglia are continued posteriorly as two
true pedal cords, which, like those of the Amphineura, are connected by transverse
commissures.

It is more difficult to compare the pleural, parietal, and visceral ganglia of the
Gastropoda with nerves found in the Ai/t/>/<in''nm. The most satisfactory view
seems to be that this whole complex of ganglia, together with its connectives, corre-
sponds with the pleurovisceral cords of Chiton. The areas of innervation coincide,
these being the mantle, ctenidia, osphradia (Chiton?), and viscera.

3. If this last assumption is correct, the pleural ganglion must lie supposed to
have arisen by the concentration into one ganglion of that part of the pleurovisceral
cord of Chiton which contains the pallial ganglionic cells, this concentration having
taken place at the anterior end of the cord, where it leaves the oesophageal ring. If,
then, the two component portions of each side of the ring, the cerebropedal and the
pleural, move further apart, and at the same time the cerebral and pedal ganglia of the
ring become more individualised as ganglia, a double cerebropedal connective comes
into existence on each side. One of these connectives shows no ganglion in its course,
and is the true cerebropedal connective of the Gastropoda. The second, however, has
the pleural ganglion in its course, and from this latter spring the visceral cords ; this
second connective is thus divided into a cerebropleural and a pleuropedal connective.

4. Chiton has numerous gills on each side, each of which receives two nerves
froTii the pleurovisceral cord near it. The Gastropoda have at the most two gills,
one on the right and one on the left. In correspondence with this reduction, the
ganglionic cells of the pleurovisceral cords belonging to the branchial nerves of
Chiton have become concentrated on each side into a single ganglion belonging to the
single gill. The parietal ganglion is thus accounted for. That portion of each
pleurovisceral cord which lies between the pleural and the parietal ganglia becomes
the pleuroparietal connective, which consists of fibres only without ganglion cells.

vii MOLLUSCA—THE NERVOUS SYSTEM 135

5. There is'no nerve in Chiton homologous with the visceral ganglion or ganglia of
the Gastropoda ; this is the chief difficulty in the comparison of the two nervous
systems. In the Amphineura, the pleuro visceral cords unite above the intestine ;
in all other Molluscs the point of junction (which is the visceral ganglion) lies below
the intestine.

In Proneomenia the posterior commissures between the pleurovisceral cords are
merely a more strongly developed part of a general commissural system.

Origin of the Crossing of the Pleuroviseeral Connective
(Chiastoneury) (Figs. 113-116).

Several attempts have been made to explain the peculiar crossing
of these connectives in the Prosobranchia. The one here given is in a
high degree probable if not altogether satisfactory.

We must start with a supposed racial form which was perfectly
symmetrical, even in its nervous system, and possessed an organisation
somewhat like that of our hypothetical primitive Mollusc (p. 26).
Such an organisation agrees in most important points with that of the
extant Chitonidce ; only one gill, however, was present on each side.

Further, the parietal ganglia innervated the gills and the osphradia,
and were thus closely connected with these organs.

The racial form of the Gastropoda may have been surrounded by
a mantle border which widened posteriorly, i.f. covered a somewhat
deeper mantle cavity which contained the pallial complex, viz. the
median anus, to the right and left of which were the ctenidia and
osphradia, and between the ctenidium and anus on each side the
nephridial aperture.

If we suppose this pallial complex to have changed its position,
shifting gradually forward along the right mantle furrow, each cteni-
dium would drag along with it its parietal ganglion. The heart and
its auricles which are connected with the ctenidium would also become
shifted.

As long as the pallial complex had not moved far forward to the
right, the pleurovisceral connectives would not cross, but would only
be shifted to the right (Fig. 114). We find the Tectibranchia among
the Opisthobranchia apparently at this stage, the only difference being
that they have already lost the original left ctenidium and also the
original left auricle (Fig. 43, p. 33).

If the pallial organs are still further shifted forward along the
mantle furrow (Figs. 115, 116) till they come to lie quite ante-
riorly, and once more symmetrically, above and behind the neck, the
original left ctenidium comes to lie on the right, and the original right
ctenidium on the left in the anteriorly placed mantle cavity. The
original right ctenidium has, however, dragged its parietal ganglion
over the intestine to the left side, and the latter becomes the
supraintestinal ganglion. The original left ctenidium, on the
contrary, has dragged its ganglion below the intestine to the right

136

COMPARATIVE ANATOMY

CHAP.

side, and this ganglion becomes the infraintestinal gang-lion. The
pleurovisceral connectives, in which these ganglia lie, now cross and
give rise to the condition called ehiastoneury. The visceral

M.1I3CI

FIGS. 113, 114, 115, IKL— Diagrams to illustrate the shifting forward of the pallial complex
along the right side of the body and the development of ehiastoneury. p, Mouth ; u1<; »////,
iilp, original left cerebral-, pleura!-, and pedal-ganglion ; ulpa, iirpa, original left and original right
parietal ganglion ; nln, original left auricle ; uos, uros, original left and original right osphradinm ;
iUct, -urct, original left and original right ctenidium ; -tnli, base of the mantle ; mr, edge of the same ;
•m, mantle cavity ; c, visceral ganglion ; ve, ventricle; n, anus.

ganglion

as

in which these connectives terminate posteriorly lies
before under the intestine.

It is unnecessary to show in detail how this displacement also affects
the heart and its auricles, the osphradia, and the nephridial apertures.

vii MOLLUSCA—THE NERVOUS SYSTEM 137

Although chiastoneury may be satisfactorily explained by this
theory of displacement, the cause of the displacement itself has still to
be sought (cf. § xiv. p. 149).

Special Remarks on the Nervous System of the Gastropoda.

I. Prosobranchia. (n) Diotocardia. — These are the most primitive Gastropoda.
The ganglia are not yet very distinct, thus recalling the Amphineura. The cerebral
ganglia are connected by two long commissures, the cerebral commissure running
forward over the pharynx, and the labial commissure running under the oesophagus.
The indistinctly separated buccal ganglia together form a horseshoe-shaped figure,
and are united on each side by a connective with the thickened root of the labial
commissure.

The pleural ganglia lie close to the pedal ganglia, so that no distinct pleuro-
pedal connectives can be distinguished. The pedal commissure is very short, and
contains ganglion cells. From each pedal ganglion, a long pedal cord runs back into
the foot ; these two pedal cords contain ganglion cells along their whole length, and
are connected by transverse commissures. These cords and commissures thus exhibit
the same arrangement as in the Amphineura. The pedal cords innervate the mus-
culature of the foot and the epipodium. There is only one indistinct visceral
ganglion, which is joined to the pleural ganglia by two pleurovisceral connectives,
crossed in the typical way.

In FissureUa only does a ganglion occur on the supraintestinal pleurovisceral
connective. In no other Diotocardian is there a ganglion at the point of departure
of the strong branchial nerve from the pleurovisceral connective ; this nerve, how-
ever, forms the branchial ganglion just below the osphradiimi at the base of the
gill. Where a ctenidium, or merely an osphradium, is found on each side, there is a
branchial ganglion close to it ; where only the left (ur) gill is retained (Turlinidcc,
Trochidir), only the left branchial ganglion is found. Since, as a rule, the parietal
ganglia are wanting in the Diotocardia, and the branchial ganglia in the Monotocardia,
the branchial ganglia of the Diotocardia have been considered, with much prob-
ability, as intestinal ganglia, which have shifted away from the pleurovisceral connec-
tives and towards the bases of the gills. As, however, FissureUa possesses both a
supraintestinal and a left branchial ganglion, it would be necessary to assume that
an originally single ganglion had here become divided into two.

The symmetrical pallial nerve is always connected by a pallial anastomosis
with the asymmetrical pallial nerves on the same side of the body. The symmetri-
cal pallial nerve rises out of the pleural ganglion, the asymmetrical nerves out of
the parietal ganglion, or the pleuroparietal connective.

The nervous system of the Neritido' and Hclicinidce are peculiar, in that the supra-
intestinal pleurovisceral connective and its corresponding ganglion are wanting.

Docoglossa. — The only essential difference between the nervous system of Patella
(Fig. 117) and the typical system of other Diotocardia lies in the fact that the
pleural and pedal ganglia are joined by a distinct pleuropedal connective.

(b) Monotocardia (Fig. 118). — The parietal ganglia are always present. The
cerebral commissure is short, and lies behind the pharynx. The labial commissure
is wanting (except in the Paludinidcc and Ampullaridce). The pedal cords and
transverse commissures are wanting (except in the ArchUcenioglossa : Paludinidcz,
Cyclophoridcv, Cypnridce). The number of visceral ganglia varies from one to
three.

The progressive development of so-called Zygoneury is noteworthy. In the
Diotocardia, a pallial anastomosis exists between the symmetrical and asymmetrical

138

COMPARATIVE ANATOMY

CHAP.

pallial nerves on each side. If this anastomosis were to shift along the two pallia!
nerves of one side to their places of origin, i.e. the ganglia from which they spring,
it would become a pallial connective uniting the pleural and parietal ganglia of
the same side of the body. There would thus arise a new accessory plenrointestinal

connective, which would be symmetrical
and not twisted, and thus unlike the asym-
metrical twisted connective already existing.
Zygoneury thus depends on the development
of such a pallial connective. In the large
majority of cases in which it occurs it takes
place on the right side (a few Rostrifera,
viz. some of the Ccrithiidcc, Ampulla /•//</".
Turitcllidoe, Xenophoridce, Struthiolariidoc,
Chenopidce, Strombidce, Calyptra'idcc, and in
all Probosddifera siphonostomata and all
Stcnofflossa). Less frequently, zygoneury
takes place on the left side (Ampullarii<l<r,
a few Crepidulidcc, Naticido:, Lumdlariida:
Cyprccidcc). In other Prosobraneh la then-
is only a pallial anastomosis on each side, as
in the Diotoca rd ia ; the nervous system is
then called dialyneurous.

The progressive concentration of the
central nervous system of the Mouotocardia,
which keeps pace with the development of
zygoneury, must lie emphasised. The con-
nectives uniting the various ganglia con-
tinually shorten, so that at last anteriorly
on the oesophagus there is a collection of
ganglia ; these are the cerebral, pleural.
pedal, infraintestinal, and supraintestinal
ganglia, all lying close together, to which
must be added the small buccal ganglia.
Only the visceral ganglia remain far back
in the visceral dome.

In J\ratica, where the anterior part of the
foot is strongly developed, and is bent back
over the head (Fig. 98), a propedal ganglion be-
comes differentiated from the pedal ganglion.
The nervous system of the Heteropoda
requires fresh investigation. So far as we
at present know, they certainly have crossed
visceral connectives, and are therefore Proso-
branchia, and, as the rest of their organisa-
tion shows, Monotocardia. The cerebral
ganglia and the pedal ganglia (pleuropedal
ganglia ?) are far apart, so that the cerebro-
pedal connectives are very long.1
II. Opisthobranchia. — The nervous system of this order, in which the typical
Gastropodan ganglia are developed, is further characterised : (1) by the absence of

FIG. 117. —Nervous System of Patella
(adapted from figures by Pelseneer and
Bouvier). 1, Cerebral ganglion ; 2, cerebral
commissure ; 3, labial ganglion ; 4, buccal gan-
glion ; i>, cerebropleural connective ; 6, cerebro-
pedal connective ; 7, nervus acusticus ; 8,
auditory vesicle ; 9, plenral ganglion ; 10, pedal
commissure ; 11, right, 12, left osphradium ;
13, visceral ganglion ; 14, supraintestinal gan-
glion ; 15, pedal cords ; 16, indication of an
infraintestinal ganglion.

1 Of. Pelseueer's Introduction i> I'etmh des Mollvsques, 8vo, Bruxelles, 1894, pp.
104, 105.

VII

MOLLUSCA—THE NERVOUS SYSTEM

139

cliiastoneury, i.e. the pleurovisceral connectives do not cross (except in Actu'on) ; and
(2) by a marked tendency to concentration of the ganglia around the posterior end
of the pharynx.

(a) Tectibranchia. — As a rule only the right parietal ganglion is found (in Actceon
the left is also present). A nerve rises from it which innervates the ctenidium, the
osphradium, and the mantle, and forms a branchial ganglion at the base of the gill.
A delicate lower cerebral commissure is often found, which runs along the pedal

FIG. 118.— Nervous System of Cyclostoma elegans (after Lacaze-Duthiers). J. Tentacular
nerve; 2, eye; 3, cerebral ganglion; 4, pedal ganglion; [>, infraintestinal ganglion; 6, visceral
ganglion ; 7, osphradium ; S, supraintestinal ganglion ; 9, auditory vesicle ; 10, plcural ganglion.

commissure below the pharynx, and may be compared with the labial commissure
of the Diotocardia.

As types of the Tectibranchia we may take Bulla as representative of the
Cephalaspidae, and Aplysia as representative of the Anaspidic (Apli/siidcc).

Fig. 119 gives the nervous system of EuUa hydatis ; only three points concern-
ing it need be mentioned: (1) The pleural ganglia have shifted till they lie close
to the cerebral ganglia, the cerebro pleural connectives becoming correspondingly
shortened. (In Actcvoii these ganglia have even fused, and are no longer to
be distinguished externally.) (2) There are three visceral ganglia. (3) The
commissures are comparatively long. (4) The parapodia are innervated from
the pedal ganglia.

In many Cephalaspidce, moreover, no distinct right parietal ganglion exists. It

140

COMPARATIVE ANATOMY

CHAP.

seems to have moved up to the right pleural ganglion, or to have fused with it, so
that the nerve running to the branchial ganglion rises direct from the right pleural
ganglion.

The nervous system of the Ptcropoda thecosomata, which we derive from Cephala-
spidce, bears a general correspondence to that of the latter, especially in the fact
that the pleural ganglia shift near to or fuse with the cerebral ganglia. The
pleurovisceral connectives are so much shortened that the ganglia occurring in
their course lie close to the cerebral and pedal ganglia. There are usually two such

FIG. 1 1'.i.— Nervous System of Bulla hydatis (alter Vayssiere).

I, Bnceal ganglion; 2, cerebral ganglion; 3, pleural ganglion; 4,
pedal ganglion ; 5, part of the right pleural ganglion (?) ; 7, eye ;
8, cerebral commissure ; i>, peilal commissure ; 10, auditory vesicle ;

II, right parietal ganglion ; 1'2, 13, 14, visceral ; 15, branchial
ganglia.

FIG. 1-20.— Nervous System
of Aplysia, diagram, combined
from several sources. 1,
Buccal ; 2, cerebral ; 3, pleural ;
4, pedal ; 5, right parietal ; 6,
visceral ganglion ; "7, osphra-
diuin ; S, genital ganglion ; 9,
branchial ganglion.

ganglia (the right parietal and a visceral ganglion t), less frequently three (two
intestinal and one visceral ganglion ?). The pedal ganglia also innervate the tins,
which correspond with the parapodia of the Cephalaspidce.

Fig. 120 represents the nervous system of Aplysia, one of the Anaspida-. The
two cerebral ganglia have moved close to each other in the middle line. The pleural
ganglia here, unlike those of the Cephalaspidce. lie close to the pedal ganglia, so
that the pleuropedal connectives are much shortened. The pedal commissure is
double, the anterior commissure is, relatively speaking, short and thick, the posterior
long and thin. The long pleurovisceral connectives run back from the pleural

VII

MOLLUSGA—THE NERVOUS SYSTEM

141

ganglia, and enter two ganglia lying side by side ; that to the right represents the

right parietal ganglion, innervating chiefly the gill and osphradium, the nerves

running to these organs forming a ganglion at the base of

each ; that to the left is the visceral ganglion. One of the

nerves which run from the latter forms a genital ganglion at

the base of the accessory glands connected with the genital

organs. In other Anaspidce, such as Notarchus (Fig. 121),

the pleurovisceral connectives are so much shortened that

the parietal and visceral ganglia lie close to the periceso-

phageal group of ganglia, which then consists of two cerebral,

two pedal, and two pleural ganglia, and further, the right

parietal and the visceral ganglia. The two cerebral ganglia

are further connected by a thin hirer commissure. The

parapodia are always innervated from the pedal ganglia.

The nervous system of the Pteropoda ijf/maosomata, which

are nearly related to the Anaspidce, corresponds in all essential

points with the nervous system of the latter, being of the

same type as that of Notarchus. , F'"- 1-j' ~1Jerv°us

System of Notarchus
(b) Nudibranchia and AscoglosBa.— The nervous system punctatus (after Vays-

is here characterised by very great concentration of the Sj^re)i diagrammatic, i,
typical Molluscan ganglia, and by a tendency to the forma- Buccal ; 2, cerebral ; 3,

tion of numerous accessory ganglia (at the bases of the 1'le"™! ; 4, pedal ganglia :
11-1 c ,i • 5, right parietal ganglion ;

tentacles and rhinophores, and at the roots of their nerves, 6' vjsce]!a] ^nc-iFon

in the course of the genital nerves, etc.). The pleural gan-
glion has moved close to the cerebral ganglion, and may fuse with it. The pedal

ganglia have also moved towards the cerebral ganglia
so that now the whole cesophageal complex of gan-
glia lies almost entirely on the dorsal side of the oeso-
phagus. The pedal commissure which runs under
the gullet, and is sometimes double, is thus very
much lengthened. The pleurovisceral connectives
are short, and occasionally enter an unpaired visceral
ganglion, which has also been drawn into the ceso-
phageal complex. This single ganglion of the visceral
connectives may be wanting (Fig. 122) ; in that case
the two visceral connectives appear like a commissure
between the two pleural ganglia runningunder the oeso-
phagus and parallel with the pedal commissure, some-
times even united with it. The fusion of all the
ganglia belonging to the peri-cc-sophageal complex is
carried very far in such animals as Tethys, where the
pleural and pedal ganglia of each side may fuse with
the cerebral ganglion. The pie tiro - cerebropedal
ganglion thus formed shifts towards the dorsal

FIG. 1-2-2. — Nervous System of
Janus (after Pelseneer simplified).
1, Buccal; 2, cerebral; 3, pleural;

middle line close to the similar ganglion of the other

4, pedal ganglia ; 5, commissure be- si(je> w{th which it forms a large supra-oesophageal

tween the two pleura! ganglia, which •.. ... c ,,

ganghonic mass. Its composition out 01 the six

corresponds with the two pleuro- &

visceral connectives of other Mol- typical ganglia can, however, be made out by the
lusca ; 6, pedal commissure ; 7, grouping of the ganglion cells and the arrangements
auditory vesicle ; 8, eye ; 9, ganglion Of the nerve tracts. A nerve leaves this mass on
of the rhinophore. ea(jh sid6) the tw() uniting under the gullet. These

form the pedal commissure, which when closely examined is found to be double. A
third delicate commissure running under the oesophagus connects the lateral portions.

142

COMPARATIVE ANATOMY

CHAP.

of the supra-cesophageal mass, and represents the visceral commissure, in which is

found a small visceral ganglion.

Among the Nudibranchia the two buccal ganglia are always found on the

posterior and lower wall of the pharynx. They are connected with each other by a

buccal commissure, and with the brain by two cerebrobuccal connectives, in whose

course accessory ganglia may be found.

The whole peri-cesophageal complex of ganglia is in the Nudibranchia enclosed in

a capsule of connective tissue.

III. Pulmonata (Fig. 123). — The central nervous system here possesses all the

typical ganglia of the Gastropoda. These, grouped
together as in so many Opisthobranchia and many
Prosobranchia, immediately behind the pharyngeal
bulb, form the peri-cesophageal complex, into which
even the parietal and visceral ganglia have been
drawn. The cerebral ganglia lie close to each other
dorsally, and all the other ganglia, which are also
close together, lie ventrally. The cerebropedal and
cerebropleural connectives are consequently always
easily distinguished. In Testacdla they are even of
some length, in adaptation, no doubt, to the special
shape and the great development of the pharyngeal
bulb. All other connectives and commissures, on the
contrary, are much shortened, so that the ganglia
connected by them lie close together. A visceral
ganglion is always found, and usually also in each
pleurovisceral connective a parietal ganglion. When
an osphradium is present (Basommatyjhora) it is
innervated from the parietal ganglion of the same
side. In Pulmonata with a dextral twist, the osphra-
dium lies on the right, and in those with a sinistral
twist on the left ; in the former the right parietal
the ganglia gauglion is the larger, and in the latter the left.
c ' ' The smaller parietal ganglion may also fuse with
the neighbouring pleural ganglion. Lobes are often
formed in the cerebral ganglia, in which certain
groups of nerves have their origin. The pedal com-
missure is often double. Buccal ganglia are always

found. They lie posteriorly on the pharynx below the cesophagus, and are joined to

one another by the bnccal commissure and to the cerebral ganglia by cerebrobuccal

connectives.

C. Scaphopoda.

The nervous system of the Scaphopoda (Fig. 101, p. 113) is
symmetrical ; the visceral connectives are not crossed. The two
cerebral ganglia lie very near one another in front of (or, if the
intestine is regarded as horizontal, above) the gullet over the snout ;
the two pedal ganglia, close to one another, lie on the anterior side
of the foot, more or less at its centre, and are joined to the cerebral
<mnglia by two long cerebropedal connectives. The two pleural
ganglia lie close to and above the cerebral ganglia, so that the
.cerebropleural connective is very short. The pleuropedal connective

FIG. 123.— Central portion of the
Nervous System of Helix pomatia
(after Bb'hmig and Leuckart), some-
what diagrammatic,
being in reality less
Buccal ganglion ; ~2, optic nerve with
thickened root (3) arising from the
cerebral ganglion (4) ; .3, pedal ; 0,
pleural ; 7, parietal ; 8, visceral
ganglion.

vii MOLLUSC A— THE NERVOUS SYSTEM 143

at once fuses with the cerebropedal, the two entering the pedal
ganglion as one connective. Posteriorly, to the right and left of the
rectum, near the anus, there are two visceral ganglia of the pleuro-
visceral connectives, joined to one another by a commissure running
behind the intestine. There are no special parietal ganglia distinct
from the visceral or the pleural ganglia.

There are four buccal ganglia, two behind the gullet or below it (if the intestine
is supposed to be horizontal), and two lying laterally and anteriorly to (or above)
the muscular mass of the radula. The anterior are connected with the posterior, and
these to the cerebral ganglia by connectives, and the two posterior and two anterior
inter sc by commissures running behind (under) the cesophagus. Nerves run from
the posterior buccal ganglia to the small ganglia of a subradular organ.

D. Lamellibranehia.

The nervous system (Fig. 124), like the whole organisation of the
Lamellibranehia, is perfectly symmetrical, and consists typically of
three pairs of ganglia: (1) the eerebropleural ; (2) the pedal; and
(3) the viseeroparietal ganglia. These three pairs of ganglia lie, as
a rule, far apart, and the connectives uniting them are therefore long.
The two pedal ganglia lie close together, while the two eerebropleural
and the two viseeroparietal ganglia are connected by distinct com-
missures beset with ganglion cells.

1. The eerebropleural ganglia are the result of the fusion of the
cerebral with the pleural ganglia. In the Protobranchia, however, the
pleural ganglia are still distinct, and lie immediately behind the
cerebral ganglia at the commencement of the visceral connectives. In
Nucula, the pleuropedal connectives are distinct for some distance,
and then unite with the cerebropedal connectives. In Solenomya
they still have separate roots, but are otherwise fused along their
whole length with the cerebropedal.

The eerebropleural ganglia are supraoesophageal, and are in
contact with the anterior adductor muscle, when this is present.
They send nerves into the oral lobes, the anterior adductor, and the
mantle.

2. The pedal ganglia lie at the base of the foot.

3. The third pair of ganglia, which correspond with the ganglia
of the visceral connectives in the Gastropoda, lie posteriorly beneath
the rectum, behind the foot, and are generally in contact with the
posterior adductor muscle ; in the Protobranchia, however, they lie
much further forward. Their area of innervation corresponds with
that of the combined parietal and visceral ganglia of the Gastropoda,
for these viseeroparietal ganglia supply with nerves the two ctenidia,
the two osphradia, the posterior portion of the mantle, the posterior
adductor, and the viscera.

The buccal or stomodseal nervous system is much reduced ; this reduction is
connected with the absence of a muscular pharynx and of all buccal armature. The

144

COMPARATIVE ANATOMY

CHAP.

anterior portion of the intestine receives nerves from the visceral connectives. Since
the fibres of these nerves have been proved to originate in the cerebral ganglia, we
may assume that, on the degeneration of the pharynx, the buccal connectives united
with the visceral connectives, so that the intestinal nerves now rise from the latter
and do not come direct from the brain. In the Pholadidcc and Tercdinida: the
visceral connectives are united in front of the visceroparietal ganglia by a second

10

FIG. 124.— Nervous system of Cardium edule (after Drost), seen from the ventral side. The
left mantle (the right in the figure) has been removed and the right bent back ; the foot has been
laid on one side. 1, Oral lobes ; 2, 3, 4, pallia! nerves, running nearly parallel to the edge ; 2, the
nerve of the pallial edge ; 5, mantle ; 6, gill ; 7, point of junction of the principal pallial nerves ; S,
mantle edge of the respiratory aperture ; 9, ditto of the anal aperture ; 10, posterior adductor ; 11,
viscero-parietal ganglion ; 12, branchial nerve ; 13, foot ; 14, pedal ganglion ; 15, left cerebropleural
ganglion ; 10, mouth ; IV, right cerebropleural ganglion ; IS, anterior adductor.

commissure, which runs under the intestine, and may perhaps be considered as a
buccal commissure shifted far back.

The mantle is innervated, as is clear from the above, partly from the cerebro-
pleural, and partly from the visceroparietal ganglia.

The two anterior pallial nerves, which rise from the cerebropleural ganglia, run
back along the edges of the mantle, to join the two posterior pallial nerves which
originate in the visceroparietal ganglia. A nerve thus runs parallel to the edge of
the mantle on each side (nerve of the pallial edge), and like a connective, unites the
anterior cerebropleural ganglion witli the posterior visceroparietal ganglion. This

VII

MOLLUSGA—THE NERVOUS SYSTEM

145

1

pallial nerve gives off branches to the organs at the edge of the mantle and to the
siphons, and is further connected with a rich nerve plexus in the mantle fold, in
which certain connecting nerves, further from the edge of the mantle, but running
parallel to it, are particularly strongly developed. A varying number of small
peripheral ganglia attain development in the pallial plexus and in the siphonal
nervous system.

E. Cephalopoda.

The symmetrical nervous system of all Cephalopoda is marked by
the great concentration of the typical Molluscan ganglia, including
those of the visceral connective.

In the following description of the nervous system, we shall consider the body in
its physiological, not in its true morphological position, i.e. we shall imagine the
pharynx and oesophagus to be running
horizontally as in other Molluscs (cf. p.
36). The true morphological position
will be given in brackets after the con-
ventionally accepted position.

1. Tetrabranehia (Figs. 125, 126).

In the complex of ganglia
which in Nautilus surrounds the
oesophagus behind the great buccal
mass, and which is not yet com-
pletely enclosed in the cephalic
cartilage, the ganglia are not very
distinct from the commissures and
connectives. The cerebral ganglia
(14, in Figs.) are represented by a
broad band-like nerve cord running
over (morphologically in front of)
the oesophagus, and from them run
two ganglionic cords, one anterior
(lower) and one posterior (upper), Flo. 125._NervoussystemofNautilus(after

which pass just below (behind) the Jhering). 1, Buccal ganglion; 2, pharyngeal

CeSOphagUS. The anterior (3) re- ganglia ; 3, pedal commissure ; 4, infundibular

, , -• , i nerve ; 5, nerve in the female for the tentacles of

presents the pedal, and the posterior the posterior and inner lobes. this nerve soon

(15) the combined pleiiral and swells to form a ganglion (cf. Fig. 126); 6, nerves

visceral ganglia.

for the other tentacles; 7' Pedal cord (=pedai

- ganglia) ; 8, auditory organ ; 9, olfactory nerve ; 10,

Ihe cerebral COrd gives rise optic ganglion ;.ll, nerve of the optic tentacles;
laterally to the large Optic nerves 12, connective to the pharyngeal ganglion ; 13,

(each of which at once swells into labiaj. »«**!*"> 14' cerebral c,onl (=cerebral

ganglia) ; 15, pleurovisceral cord.

an optic ganglion), numerous nerves

to the lips, the nerves for the optic tentacles, the auditory and olfactory

nerves, and the cerebrobuccal connectives.

From the pedal cord, nerves run to the tentacles round the mouth
and to the funnel. In the female, the nerves for the inner circle of

VOL. II L

146

COMPARATIVE ANATOMY

CHAP. VII

tentacles come from a braehial ganglion, which, however, does not

supply all the tentacles (Fig. 126, a);
this is joined to the pedal ring by a
brachiopedal connective.

The pleuroviseeral cord gives off
numerous pallial nerves (there is no
stellate ganglion), and two strong vis-
ceral nerves which run near the middle
line accompanying the vena cava, inner-
vate the gills, the osphradia, and the
FIG. 120.— Nervous system of Nautilus, blood-vessels, and form a genital gan-

from the right side. Numbering the same g\[ou hjo-h Up jn the visceral dome.
as in Fig. 125. a, Ganglion for the ten-
tacles of the posterior and inner lobes in The sympathetic nervous system consists
the female. - . . . , . . . ,

of an infra-cesophageal commissure, which rises

from the cerebral ganglion, and passes close under the cesophagus in the musculature
of the buccal mass ; two ganglia, a pharyngeal and a buccal ganglion, are found on
each side in its course.

2. Dibranehia (Figs. 127, 128).

The peri-cesophageal mass of ganglia, comprising the Avhole of the
central nervous system, is entirely enclosed in the cephalic cartilage.
The large typical ganglia are so crowded together that it is extremely
difficult to distinguish them one from another, and the connectives
and commissures are not visible externally. The whole complex has
a continuous cortical layer of ganglion cells.

The more or less distinct separation of each pedal ganglion into
two, one anterior (lower) and one posterior (upper), is characteristic
of the Dibranehia. The former of these is the braehial ganglion, and
innervates the arms, which must be considered as parts of the foot ; and
the latter is the infundibular ganglion, and innervates the siphon,
which may be regarded as the epipodium. This differentiation of the
pedal ganglia can be traced to the great development of that part of
the foot (viz. the arms) which surrounds the head. In the same way
in Natica, where the anterior part of the foot is strongly developed,
and is bent back over the head, a propedal ganglion becomes
differentiated from the pedal ganglion. The braehial ganglia become
joined in the Lilmmcliia to the cerebral ganglia by cerebrobrachial
connectives. In Eledone and Octopus, they are further connected Avith
one another by a thin supraoesophageal commissure.

The pleural ganglia lie laterally in the perioesophageal mass, while
the ganglia of the visceral connectives, i.e. the parietal and visceral
ganglia Avhich lie close together, their connectives having shortened as
much as is possible, form the posterior (upper) portion of the infra-
O3sophageal mass.

The following are the connectives which are revealed by sections
through the peri-oesophageal mass :—

12

18

FIG. 127.— Anatomy of Octopus (after Leuckart and Milne Edwards). The body is cut open posteriorly, the mantle laid
back to the right and left, and the liver removed. 1, Brachial artery ; 2, brachial nerve ; 3, pharynx ; 4, buccal ; 5, cerebral
ganglion ; 6, efferent duct of the upper salivary glands ; 7, funnel ; S, upper salivary glands ; !>, crop ; 10, anus ; 11, afferent
branchial vessel (branchial artery) ; 12, left renal aperture ; 13, efferent branchial vessel (branchial vein) ; 14, gastric ganglion ;
15, left auricle; 16, spiral caecum of the stomach ; 17, renal sac; 18, water canal; 19, ventricle; 20, ovary; 21, rectum ; 22.
efferent ducts of the digestive gland (liver), cut through near its opening into the intestine ; 23, mantle ; 24, stomach ; 25,
right ctenidium ; 20, aperture of the right oviduct ; 27, stellate ganglion ; 28, nerve to the gastric ganglion ; 29, upper salivary
; 30, aorta ; 31, ojsophagus ; 32, optic ganglion ; 33, lower salivary glands.

148

COMPARATIVE ANATOMY

CHAP.

(1) Two cerebro-brachial ; (2) two cerebro-infunclibular ; (3)' two
cerebropleural ; (4) two brachio-infundibular ; (5) two pleuro-infundi-
bular ; (6) two pleurobrachial connectives. The close proximity of
the]' visceral ganglia to the peri-oesophageal mass makes it impossible
any longer to distinguish the visceral connectives.

The cerebral ganglia give rise to the two optic nerves (which soon swell into the

enormous optic ganglia at
the bases of the eyes), the
auditory nerves, the olfac-
tory nerves (which for a
certain distance fuse with
the optic nerves), and the
connectives of the buccal
ganglia.

The brachial ganglia
send off separate nerves to
the arms, which nerves are
connected by a hoop-like
commissure round the base
of the circle of arms. Run-
ning through the arms, the
nerves swell into succes-
sive ganglia which corre-
spond with the transverse
rows of acetabula.

The separation of the
pedal ganglion into a bra-
chial and an infundibular
ganglion can be proved on-
togenetically and anatomi-
cally. There is no such
separation in the male
Nautilus, the brachial and
infundibular nerves spring-
ing from one and the same
ganglion. In Argonauta
(Fig. 128, F) the separation
is not externally visible,
but in Octopus (E) we see
the first traces of it ; in
Si'ina (D), LoUrjo (C), and
Sepiola (B), it becomes
more and more evident, till
finally in Ommatostrcphes

(A) the distinct brachial ganglion has moved away from the infundibular ganglion,

with which it is joined by a slender externally visible connective.

In this same series, the separation of the so-called upper buccal ganglion from

the cerebral ganglion also takes place, the buccal remaining united to the brachial

ganglion by the brachiobuccal connective.

The parietal ganglia give rise to the two large pallial nerves. Each of these runs

backward and upward, and enters the stellate ganglion on the inner surface of the

iO

FIG. 128.— Central nervous system of various Dibranchia,
from the-right side. All the figures after Pelseneer. A, Ommato-
strephes ; B, Sepiola ; C, Loligo ; D, Sepia ; E, Octopus ; F, Argo-
nauta. 1, Cerebral ; 2, pedal ; 3, visceral ; 4, brachial ; 5, upper
buccal ganglion ; 6, infundibular nerve; 7, visceral nerve; 8, optic
nerve cut through; 9, pallial nerve; 10, brachial nerves; and in
Fig. B the pharynx ()>h), and oesophagus (u) are drawn in black.

vii MOLLUSGA—THE ASYMMETRY OF THE GASTROPODA 149

mantle. Numerous nerves radiate into the mantle from this ganglion, one of them,
which runs dorsally, looking like the direct continuation of the pallial nerve through
the ganglion. The pallial nerve often divides into two branches sooner or later after
it has left the parietal ganglion ; one of the branches running to and through the
stellate ganglion, to unite beyond it with the other branch which runs past the
ganglion. The two stellate ganglia are often connected by a transverse commissure.

The visceral ganglia give off, near the middle line, two visceral nerves, which
innervate the rectum, the ink-bag, the gills, the heart, the genital apparatus, the
kidneys, and certain parts of the vascular system. The two genital branches of
these nerves are connected by a commissure.

The sympathetic nervous system consists of a buccal ganglion lying beneath
(behind) the oesophagus in the buccal mass ; this ganglion is joined to the upper
buccal or pharyngeal ganglion by a buccal connective. Two nerves run up along
the oesophagus from the lower buccal ganglion to the gastric ganglion, which lies
on the stomach, and innervates the greater portion of the intestine and the digestive
gland (liver).

XIV. An Attempt to explain the Asymmetry of the Gastropoda.

i.

Chiastoneury, i.e. the crossing of the two pleuro-visceral connectives in the
Prosobranchia, may be explained on the three following assumptions.

1. The ancestors of the Prosobranchia were symmetrical animals ; the mantle
cavity lay behind the visceral dome and in it the pallial complex, that is, the ctenidia,
osphradia, nephridial apertures, genital apertures, and, in the centre, the median
anus.

2. The visceral commissure or ganglion lay beneath the intestine.

3. The pallial complex shifted gradually from behind forward, along the right
side of the body (cf. p. 136).

The position of the pallial complex in the Tectibmnchia, among the Opisthobranchia
on the right side, can also be thus explained. The pallial complex in its forward
movement in these animals has either not yet reached the anterior position or,
having reached it, has shifted back again.1 The visceral connectives are therefore
not crossed.

The above assumptions do not, however, explain—

1. The asymmetry which is brought about in some Gastropoda by the dis-
appearance of one ctenidium, one osphradium, and one renal aperture.

2. The coiling of the visceral dome and shell, especially the dextral or sinistral
spiral twist.

3. The relation existing between the manner in which the visceral dome and
shell are coiled, on the one hand, and the special asymmetry of the asymmetrical
organs (ctenidia, osphradia, nephridia, anus, genital organs) on the other.

4. The cause of the shifting forward of the pallial complex.

2.

It is unnecessary to discuss the first of the above assumptions, viz. that the
ancestors of the Gastropoda were symmetrical animals, since all Molluscs except
the Gastropoda are symmetrical, i.e. the Amphiivura, the Lamcllibranchia, the
Scaphopoda, and the Cephalopoda.

1 See note to § 13, p. 158.

150

COMPARA TIVE ANA TO MY

CHAP.

The assumption that the pallial complex originally lay posteriorly is also well
founded. In all symmetrical Molluscs, the amis lies as the centre of the complex
posteriorly in the middle line, and further, in all symmetrical Molluscs, the nephridial
and genital apertures lie posteriorly at the sides of the anus. When the ctenidia
and osphradia have been retained in symmetrical Molluscs, they lie symmetrically
on the posterior side of the visceral dome. This is the case in the Cephalopoda, and
in the most primitive LamellibraiicMa, the Protdbranchia (Nucula, Lcda, Solenomya),
and even in some Chitonidix, and those Solcnogastrcs which still have rudiments of
gills.

In keeping with the posterior position of the pallial complex, the mantle fold
which hangs down round the base of the visceral dome is, in symmetrical Molluscs,
widest posteriorly where it has to cover the complex ; at this part the mantle
furrow deepens into a mantle cavity.

In connection with the second assumption, it still remains unexplained why in
the Amphineura the commissure between the pleuro-visceral cords runs over the
intestine ; whereas on the other hand, in all other symmetrical Molluscs, the
visceral ganglion lies, as in the Gastropoda, below the intestine.

3.

The third assumption, that the pallial complex has shifted forward, requires
separate discussion.

If the pallial complex did thus shift forward, chiastoneury must necessarily
have taken place ; the original left half of the complex must necessarily have become
the present right half, and vice versa. Further, the right pleuro-visceral connective
would have to become the supra-intestinal connective and the left the infra-intestinal
connective ; the original right parietal ganglion the supra-intestinal ganglion, and
the original left parietal the infra-intestinal ganglion. But iclnj did such a shifting
take place ? We shall here attempt to answer this question.

Cause of the shifting forward of the pallial complex. — We have assumed the
symmetrical racial form of the Gastropoda (with posterior mantle cavity and sym-
metrical pallial complex) to be a dorso-ventrally
flattened animal with a broad creeping sole, a
snout-like head with tentacles and eyes, and a
somewhat flat cup - shaped shell covering the
dorsal side of the body (Fig. 129). It therefore
resembled in outward appearance a Fissurclla, a
Patella, or a Chifmi, if we assume the imbricated
shell of the last to be replaced by a single shell.

The body of such a racial form was only pro-
Fio. 129. —Hypothetical primitive ,, , , , „ mi. i i r

Gastropod, from the side, o Mouth ; tected dorsally by the shell. The hard surface
7,-, head ; sm, shell muscle ; oso, apical along which the animal slowly crept served to
shell aperture ; a, anus ; n, renal aper- protect its lower side, the dorsal shell being
ture ; mh, mantle cavity ; ct, ctenidium ; presse(:i nrmly against the substratum, when

necessary, by the contraction of a powerful shell

muscle (cf. Fig. 106, p. 122). When the shell was thus pressed down, communica-
tion between the pallial cavity and the exterior (for the purpose of inhaling and
exhaling the respiratory water, and ejecting the excreta, excrement, and genital
products) was rendered possible by means of a cleft in the posterior edges of the
mantle and shell.

Unlike their racial form, all known Gastropoda (except those whose body form
has been secondarily modified, generally in connection with the rudimentation of the
shell) are distinguished by the fact that the viscera with their dorsal integumental

vii MOLLUSGA—THE ASYMMETRY OF THE GASTROPODA 151

covering protrude hernia-like in the form of a high spire-like visceral dome, with
which the shell corresponds in shape. The uncoiled shell of every snail is as a
matter of fact spire-shaped.

The development of such a shell and dome has already been recognised as due to
the increased protection needed by the body when the capacity for creeping becomes
developed. The whole of the softer part of the body can be withdrawn into such a
shell, and, further to increase the protection, an operculum is often developed on the
foot for closing the aperture of the shell, when the animal has retired into it. The
shell muscle of the racial form no longer serves for pressing the shell against the
surface on which it rests, but for withdrawing the head and foot into the shell. It
becomes the coliunellar muscle (Fig. 131, sm).

Taking in turn the different stages in the development of the Gastropod

pjo 13^

(Lettering ,n this an,, in the following three
fl the ^^ ^ h] pig_ m)

FIG. 130.— Hypothetical primitive Gastropod,

from above, o, Mouth; vie, til pi, iilf, original
left cerebral, pleura! and pedal ganglia ; ulpa,
wpa, original left and right parietal ganglia ;
ula, original left auricles ; uot, uros, original left
and right osphradia (Spengel s organs) ; ulct, urct,
original left and right ctenidia (gills) ; mb, base
of the mantle ; mr, edge of the mantle ; m, mantle
cavity; v, visceral ganglion; ve, ventricle; o,
anus.

shell, we have as the first and most important its dorsal spire-like prolongation.
In this way the cup-shaped shell of the racial form becomes a high conical shell like
that of Dcntalium.

Such a shell carried vertically by the animal (Fig. 131) would, when the latter is
at rest, be in a state of unstable equilibrium, which would be upset by movement or
by the slightest pressure from without. It is also evident that when the animal is
in motion a vertically placed spire-like shell would be extremely awkward.

If we assume the shell to be carried at some other angle to the body, we have
the following possible positions :—

1. The shell might be carried inclined forward (Fig. 132). Such a position is
the most unfavourable imaginable for locomotion, for the functions of the mouth,
and for the sensory organs on the head.

152

COMPARATIVE ANATOMY

CHAP.

On the other hand, such a position is the most favourable imaginable for the
functions of the organs belonging to the posteriorly placed pallial complex, which
now lie dorsally, since in this position the mantle cavity is subjected to least pressure

<vr>U c(

FIG. 132.

from the viscera and from the columellar muscles. The downward pressure of
the visceral mass which now takes place would tend indeed to widen the cavity.

2. The shell might be carried inclined backwards (Fig. 133). This position is
the most favourable imaginable for locomotion and for the functions of the organs

Fio. 133.

of the head, which would thus be free on all sides. It is, however, the most
unfavourable imaginable for the functions of the organs of the pallial complex,
which now lie beneath the visceral dome. The mantle cavity has to bear the whole
pressure of the visceral mass, and especially that of the columellar muscle ; it would

be squeezed together, so that the
.* circulation of the respiratory water
would be prevented or at least
rendered more difficult, as would
also the ejection of the excreta, ex-
crement, and sexual products.

3. Finally, the shell may be
carried inclined to the right or left
(Fig. 134). This is neither the most
favourable nor the most unfavour-
able position for locomotion, for the
head, and for the pallial complex.
It is an imaginable intermediate
position.

In this position there is no dead point, as shifting of the parts would always be
possible, and the shell be enabled to take up the .position most suitable for locomo-
tion and for the functions of the cephalic organs, and the mantle cavity that best
suited for the exercise of the functions of the pallial complex lying within it.

Assuming that the shell is inclined to the left (Fig. 135), the pressure brought
to bear on the mantle cavity would vary in amount in different areas of that cavity.
It would be greatest on the left side, and would continually decrease towards the

FIG. 134.

vii MOLLUSC A— THE ASYMMETRY OF THE GASTROPODA 153

right. On the left there would be a pressure from the front which would, so to
speak, squeeze out the pallial complex backwards over to the right. It must further
be noted that the point subjected to least lateral pressure and to the greatest down-
ward pull lies on the right,
which has become the upper
side of the visceral dome. At
this point the mantle furrow
will most easily deepen, and
become more spacious. Into
such a deepening the organs of
the pallial complex which are
being pressed from the left
have room to move forward to
the right. Here we have the
first step in the shifting for-
ward of the pallial complex
along the right mantle furrow.
Further, as soon as the least
shifting of this sort has taken
place, the shell and visceral
dome can move slightly from
their present position on the
left, towards that backward
position which we have seen
to be the most favourable im-

aginable for locomotion and for

FIG. 135.— Diagram illustrating the variations of pres-
sure to which the shell and visceral dome are subjected
when inclined to the left. The thickness of the concentric
lines indicates the amount of the pressure, a, Point of greatest
the functions of the cephalic pressure ; l>, point of least pressure. The arrows give the
Ol'rrans direction in which shifting takes place. It is evident that

,, . the left side of the pallial complex is subjected to greater

If we suppose this process

1 pressure than the right.

gradually to be completed, the

shell and visceral dome finally gain the most favourable backward position, and the
pallial complex is gradually shifted forwards along the right mantle furrow. The
pallial complex thus lies anteriorly on the upper side of the visceral dome, which
now points backwards. This anterior position is that of the least upward pressure,
or rather of the greatest downward pull, i.e. it is the point at which the mantle
cavity can most easily deepen and widen, and where the pallial organs can best
fulfil their functions.

The position of the shell and the pallial complex characteristic of the Gastropoda
is now attained, and with it chiastoneury and the inverse position of the organs of
the pallial complex.

4.

The second stage in the development of the Gastropod shell is the coiling in
one plane of the visceral dome and shell.

If the Gastropod visceral dome assumes the most favourable inclined position above
described, it will, under normal conditions, change its conical shape. The side which
lies uppermost will become arched and the lower side concave. This change of form
is caused by the stronger growth of the integument of the visceral dome and mantle
on that side, which, in the inclined position of the visceral dome, is the most
stretched or pulled. The visceral dome also becomes curved in one plane, and the
shell naturally adapts itself to the changes of shape of the dome. Again, the shell
could not remain conical, because a large part of the dorsal integument (base of the
visceral dome) would then be uncovered, and in consequence of the increase of those

154 COMPARATIVE ANATOMY CHAP.

parts of the body not covered by the shell there would come a time when the body
could no longer be completely withdrawn into it.

5.

Before discussing the third stage in the development of the Gastropod shell, we
must consider its growth. This, from a geometrical point of view, is of three kinds :
growth in height, peripheral growth, and radial growth or increased thickness of
the shell wall. This last does not here concern us.

Supposing, for simplicity's sake, the shell to be conical, growth in height occurs
at the base (or aperture of the shell), and takes place by means of continual deposits
of bands of new material at the edge of the aperture, by the growing edge of the
mantle.

Peripheral growth is the enlargement of the circumference of the base or aperture
of the shell.

If the height and the peripheral growth remain uniform round the whole
aperture of the cone (which is assumed to be round), the cone increases without
altering its shape.

If, however, the growth in height is not uniform, but steadily and symmetrically
increases along each side from an imaginary minimum point to a diametrically
opposite maximum point, the peripheral growth, however, remaining uniform, a
spirally twisted hollow cone is produced.

If the minimum and maximum points in this growth continue throughout in
one and the same plane, a symmetrical shell coiled in this plane of symmetry
results.

If, however, as growth increases, the maximum point shifts from the symmetrical
plane, say to the left (the minimum point shifting in the opposite direction to the
right), the maximum and minimum points no longer trace on the spirally coiled
shell straight but spirally twisted lines, and the conical shell is then not coiled
symmetrically in one plane, but asymmetrically in a screw-like spiral. We then
have what conchologists call a dextrally twisted shell.

The growth of the Gastropod shell actually takes place in this last manner.

6.

This, the dextral (or sinistral) coiling of the Gastropod shell, is the last stage to be
discussed. If the visceral dome and shell which are twisted in one plane pass, in growth,
from an incline to the left to a backward incline, this is equivalent to the continual
shifting of the point of maximum growth to the left and that of minimum growth
to the right ; the necessary consequence being a dextral screw-like spiral twist.

It must be borne in mind—

1. That the peripheral growth remains constant, i.e. that the outline of the
growing edge of the mantle remaining uniform, the increasing aperture of the shell
also retains the same form.

2. That the additions to the shell by the mantle edge are made in the form of
bauds of new material, the already formed firm shell not altering in shape.

3. That the growing edge of the mantle, which secretes the shell substance, does
not, in the course of the gradual change from the left to the backward incline, itself
become twisted, but retains its position in relation to the rest of the body. It is
thus only the maximum and minimum points of growth in height which become
shifted along the edge of the mantle.

4. It must be noted that this description of the manner in which a dextrally
twisted shell arose only applies to that stage in the ontogenetic or phylogenetic

vii MOLLUSCA—THE ASYMMETRY OF THE GASTROPODA 155

development of the shell duringwhich its displacement in a backward direction and the
shifting forward of the pallial complex occur. When once the result most favourable
to the animal, i.e. the anterior position of the mantle cavity and the backward
direction of the shell, are attained, further displacement, which would be dis-
advantageous, does not take place. It is, then, not at first sight evident why,
when the need for displacement ceases, its action still continues, i.e. why, though
displacement ceases, the visceral dome and shell continue to grow in a dextral
twist and not symmetrically. This point will be explained below.

For the sake of clearness we have treated separately the three important factors
in the development of the Gastropod shell, viz. (1) the formation of a tall conical
shell, (2) the spiral coiling of the same, and (3) the special manner of coiling in a
dextral twist. In reality these three factors do not denote special stages, but all
operate simultaneously. The continually increasing protrusion of the visceral dome
was accompanied by the dextral twist, as a consequence of the twisting of the
visceral dome from its incline to the left to the most favourable backward incline,
by which the pallial complex was shifted forward.

8.

The results of ontogenetic research favour the theory here advanced. We
have first to note the fact that the anus (the centre of the pallial complex) and the
mantle fold originally lie posteriorly. They come to lie anteriorly in the embryo
not by active shifting, but by the cessation of growth on the right side between the
mouth and anus, and its continuation on the left side. There is, however, no
difficulty in harmonising this ontogenetic method of gaining the object with the
phylogenetic method.

9.

So far we have placed mechanical and geometrical considerations in the fore-
ground. But these necessarily coincide with utilitarian considerations. Every
alteration in the direction we have been considering means an improvement in the
organisation of the animal, an advantage to enable it the better to maintain the
struggle for existence. The formation of a spire-like shell, which has been recog-
nised as the starting-point in the development of the asymmetry of reptant Gastro-
pods, was the only method by which complete protection of the whole body could
be attained, and must therefore be considered to have been advantageous under the
circumstances. We might further conclude this from the fact that the possession
of such a shell actually distinguishes the Gastropoda from the primitive Mollusca,
which the Chitonidcc are rightly considered most nearly to represent.

10.

One apparently important objection to the theory here set forth must be mentioned.
If the first factor in the asymmetry of the Gastropod body is the development of a
high spire-like shell, and if the arrangement of the nervous system is necessarily
connected with the coiling of the shell in a definite direction, how can we account
for forms such as Fissurclla ? This Diotocardian genus actually belongs to the most
primitive Gastropods, because the symmetry of the pallial complex is still retained.
But it possesses an asymmetrical nervous system and the typical chiastoneury of
the Prosobranchia, and nevertheless a flat cup-shaped symmetrical shell. We thus
here have secondary characteristics of the inner organisation combined with an

156

COMPARATIVE ANATOMY

CHAP.

apparently primitive shell. The latter is, however, only apparently primitive, as can
be proved systematically and ontogenetically. The forms most nearly related to
Fissurella, such as the primitive gemis Pleurotomaria (Fig. 136 A), Polytrcmaria
(Fig. 136 B), and Scissurella, have spacious spirally coiled dextrally twisted shells.
In Haliotis (Fig. 136 D) the shell becomes flat and the coiling indistinct, as is also
the case to some extent in Emarginula (Fig. 136 C), till finally in Fissurella (Fig.

FIG. I3i3.— Shells of A, Pleurotomaria ; B, Polytremaria ; C, E, Emarginula ; D, Haliotis ;
F, Fissurella ; G, H, stages in the development of the shell of Fissurella ; I, shell of
the Gastropod racial form, with marginal cleft ; K, the same, with apical perforation ;
L, Lamellibranch shell : M, shell of Dentalium, seen from the apical cleft. The shell clefts
and perforations are black, o, Mouth ; a, anus ; ct, ctenidium.

136 F) it again secondarily becomes flattened or cup-shaped and symmetrical. Fis-
surella even passes ontogenetically through an Emarginula stage, in which the shell
is distinctly spirally coiled (Fig. 136 G, H). We may therefore conclude, with as
much certainty as is possible in morphological questions, that the outwardly sym-
metrical Fissurella descends from forms with high spirally coiled shells. Its return
to a flat symmetrical shell may have been determined, as in the Patellidce, Capulidce,
etc. , by adaptation to certain biological conditions.

11.

The explanation given above seems to throw new light on many as yet unsolved
problems in the morphology of the Mollusca, such as the asymmetry of the pallia!
complex in most Gastropoda. Many Diotocardia, all Monotocardia, all Opistho-
branchia, and all Pulmonata show marked asymmetry in the pallial complex. The
asymmetry consists principally in the absence of one gill, one osphradium, and one
nephridial aperture. The inner organisation also shows reflections of this asym-
metry in the nervous system, and the absence of one kidney and one auricle. On
closer inspection, it is found that it is the original left half of the pallial complex

vii MOLLUSGA—THE ASYMMETRY OF THE GASTROPODA 157

(which in a Prose/branch would lie to the right in the mantle cavity near the anus)
which is wanting. The anus is no longer the centre of the pallial group of organs, but
lies outermost on one side. While in the Prosobranckia, for example, the original
left half of the pallial complex (which would now lie on the right) has disappeared,
those organs of the complex (the original right) which are retained, shift from
the left to occupy the empty space. Consequently, we find the anus no longer
anteriorly in the middle line, but on the right side, close to the extreme right of the
mantle cavity.

But what is the reason of the disappearance of the left half of the pallial complex
in the Monotocardia, Opisthobranchia, and Pulmonata ?

In answering this question we must refer back to paragraph 3, where it was seen
that if the spire-like shell assumes the only possible lateral inclination, the mantle
cavity and the pallial complex within it are subjected to unequal pressure. If the
shell is inclined to the left, the side of the posterior mantle cavity subjected to the
greatest pressure is the left, and the pressure continually decreases towards the right.
These variations of pressure are also retained during the whole time in which the
backward displacement of the shell and the forward displacement of the pallial
complex takes place. In other words, i.e. described in terms of our theory, from the
very commencement of the development of the Gastropod organisation, the original
left organs of the pallial complex were subjected to unfavourable conditions. In
this left -sided compression of the mantle cavity the ctenidium especially would
necessarily be reduced in size and become rudimentary, and might entirely disappear.

As a matter of fact, the original left half of the pallial complex (which would
now lie on the right) has entirely disappeared in many Diotocardia (the so-called Azy-
gobranchia), in all Monotocardia, and in the Opisthobranchia. The fact that the
original right gill, the only one remaining, has also disappeared in the Pulmonata
is accounted for by the change to aerial respiration. It is an interesting fact that
in the Basommatophora the original right osphradium is retained.

If, however, the original left gill did not quite disappear, but only became
smaller, we should have to expect that in such Diotocardia as still possess two gills,
the original left (now the right) would be the smaller. This would be the case at
least in the more primitive forms with shells still twisted. Haliotis and Fissurella
are the only Molluscs to which this applies. In Haliotis, whose shell is still
twisted, the right (originally left) gill is in reality the smaller. But in Fissun'Ua
and Subemarginula, where the asymmetry of the mantle cavity has been secondarily
lost, the inequality in the size of the gills has also disappeared.

12.

Another unsolved problem remains. Why does the shell continue to grow
asymmetrically coiled with a dextral twist, after the cause of this asymmetry, viz.
the change from the incline to the left to the backward incline of the shell, simultane-
ously with the shifting forward of the mantle cavity and pallial complex, has ceased
to act, i.e. when the shell has definitely assumed the posterior, and the pallial
complex the anterior, position ? The explanation of this lies in the asymmetry so
early apparent in the mantle cavity, which from the beginning is more spacious
to the right (now left) than to the left, the consequence being that the left half of
the pallial complex atrophied. This asymmetry of the pallial complex and mantle
cavity remained after the displacements of shell and pallial complex had been
definitely accomplished in the Prosobranchia, i.e. the asymmetrical growth, and
therefore the continuous coiling of the viscei'al dome and shell in a spiral twist,
continued.

In altogether exceptional conditions, which rendered a flat cup-shaped shell

158

COMPAEA TI VE ANA TO MY

CHAP.

useful, the return to symmetry in the pallial complex and mantle cavity or fold
would be advantageous, since then symmetrical growth of the shell could take
place. If the difference between the maximum and minimum growth in height
is but slight the shell would be but slightly coiled, and if the peripheral growth
is pronounced, while the growth in height is insignificant, a flat cup-shaped shell
would result (Haliotis, Emarcjimda, Fissurella, Patella, etc.).

13.

Chiastoneury only takes place when the original right half of the pallial complex
crosses over to the left of the median line anteriorly.

This crossing of the line of symmetry has actually taken place in the Proso-
branchia. The original right gill in them lies quite to the left of the mantle cavity.
In the Azyyobranehia and Monotocardia the hind-gut with the anus has at the same
time become displaced into the right (original left) narrower gill-less half of the
mantle cavity, which, however, is still spacious enough to contain the rectum. The
Prosobmnchia are streptoneurous.

In the Tcctibntnchia and Opisthobranchia the pallial complex is found on the right
side of the body, and has nowhere crossed the median line anteriorly. There is
therefore no chiastoneury among the Opisthobranchia, i.e. their visceral connectives
are never crossed.1

In the Pulmonata the pallial complex has shifted far forward, but it has not
passed the middle line with any organ which, drawing the parietal ganglion and the
visceral connective with it, could have brought about chiastoneury. For the left
(original right) gill, the onlyone elsewhere retained, disappeared (apparently very early)
in the Pulmonata. The osphradium, which is retained in aquatic Pulmonata, is the
original right, and still lies on the right side. In considering the arrangement of
the nervous system, it is really immaterial whether we assume that the hind-gut

has shifted back to the right
secondarily, and the osphra-
dium moved to near the re-
spiratory aperture, or that
the hind-gut never reached
the median line, and that
the osphradium never passed
over it. The Pulmonata are
thus euthyneurous.

14.

We saw, in paragraph
3, that with a strongly de-
veloped visceral dome and
posteriorly placed pallial
complex, a shell inclined
forward or coiled forward is
an impossibility for a rep-
tant Gastropod. But such
a shell is not an impossibility for an animal which does not creep. For example,
iu a swimming animal, whose shell, partly tilled with air, serves as a hydrostatic
apparatus, there is no reason why a much developed visceral dome and shell should

1 Except in Avtueon, an exception which makes it probable that in the Opistho-
tlie pallial complex has secondarily returned from an anterior position.

FIG. 13V.— Nautilus, diagram, do, Dorsal ; ve, ventral ; vo,
anterior; Jti, posterior.

vii MOLLUSGA—THE ASYMMETRY OF THE GASTROPODA 159

not become coiled forward, the original posterior position of the pallial complex
being retained as the most favourable under such circumstances. As an example
of this we have the Nautilus, all Nautiloidca and Amman itidca, with their exogas-
trically (anteriorly) coiled shells and posteriorly placed pallial complexes (Fig. 137).
The coiling of the shell of Spirula forms an exception to that of all other Mol-
ft(sc(t, being endogastric. With regard to this we have to consider first, that the
shell of Spirula is internal and rudimentary, and that the backward coiling does not
in any way affect the posteriorly placed mantle cavity ; and second, that only the
modern genus Spirula has such a shell. The Miocene genus Spirulirostra has its
phragmacone eudogastrically bent but not coiled, and the older Belcmnitidcc never
have either curved or coiled shells. Moreover, the shell of this whole group, being
internal and, as far as the original purpose of a shell, protection of the body, is con-
cerned, rudimentary, does not come under consideration in the present discussion.

15.
In an animal living in mud, like a limicolous bivalve, there appears no reason

oio

FIG. 139.— Hypothetical transition
form between Dentalium (Fig. 13.S)
and the racial form of the Gastropoda
(Fig. 140), from the left side.

FIG. 138.— Dentalium, diagram from
the left side, g, Genital gland ; Jet,
cephalic tentacles.

FIG. 140.— Hypothetical racial form
of the Gastropoda, from the left side.

why the shell should not simply become elongated, and why the mantle cavity and
pallial complex should not retain the posterior position. Dentalium (Fig. 138) is

160 COMPARATIVE ANATOMY CHAP.

distinctly in this condition, being the symmetrical primitive Gastropod adapted to
life in mud, and provided with a turret-like shell and posterior pallial complex.
The perforation at the upper end of the shell, which freely projects from the
mud, is of great morphological importance, corresponding physiologically with the
siphons of the limicolous Lamellibranchia. A comparison between Dentalium and
a Fissurella with its pallial complex twisted back, and with a long and turret-
like shell, is, from our point of view, very appropriate. A Fissurella, so transformed,
would almost exactly resemble the hypothetical symmetrical racial form of the Gas-
tropoda, in which, however, we should have to assume a mantle- and shell-cleft
reaching to their edges (cf. Fig. 136, I).

The anatomy of the Protobranchia, which has recently been more closely studied,
and especially the posterior position of the two gills, the flat sole for creeping, and
the presence of the pleural ganglia, justify us in deriving the Lamellibranchia also
from the racial form of the Gastropoda, in which the cleft edge of the mantle would
correspond with the posterior or siphonal edge of the mantle in the former. This
edge of the mantle, having a similar physiological function, often possesses tentacles,
papillae, etc., in both groups.

Dentalium further fits in with our theory, for the forward curve and the position
of the columellar muscle on the anterior side of the visceral dome which would be
disadvantageous to a freely reptaut, is not so to a limicolous, animal.

16.
The Dextral and Sinistral Twists.

Most Gastropods have the visceral dome and shell twisted dextrally. The direction
of the twist has been determined by the fact that the visceral dome and shell origin-
ally inclined to the left, and then more and more backward, thus pushing the
pallial complex along the right mantle furrow. It cannot be determined why the
incline to the left was originally chosen. The shell might just as well have inclined to
the right at first, and then more and more backward, pushing the pallial complex along
the left mantle furrow. The consequent asymmetry would then have been exactly
reversed. To take, a concrete example : in a Monotocardian, with visceral dome and
shell twisted sinistrally, the original left parietal ganglion would become the supra-
intestinal ganglion on the right. The original right half of the pallial complex
would disappear, and the left half which persisted would lie to the right of the anus
or rectum, which would take up its position to the left of the median line.

Gastropoda with sinistrally twisted shells are actually known, many of them
having the asymmetrical organs in the inverse position which corresponds with this
twist. Such are. among the Prosobranchia, Neptunca contraria, Triforis, and occa-
sional specimens of Buccinum ; among the Pulnionata, Physa, Clausilia, Hclicter,
Ampkidromus, and occasional specimens of Helix and Limmaea. In Bulimus pcr-
versus, individual specimens with either sort of shell are found, with the special
asymmetry of the organs belonging to it.

17.

There are, however, snails whose shells are dextrally twisted, but which possess
the organisation of animals with sinistrally twisted shells. This is the case among
the Prosobranchia in the sinistrally twisted sub-genus Lanistes of the genus Ampul-
laria ; among the Pulmonata, in Choanomphalus Maacki and Pompholyx solida ;
among the Opisthobranchia, in those Ptcropoda which, whether as adults (Lima-
cinidoc) or larvre (CymbuliidoB), have a twisted shell. This fact is entirely against
our theory in explanation of the asymmetry of the Gastropoda, for this theory

vii MOLLUSCA—THE ASYMMETRY OF THE GASTROPODA 161

points to a causal connection between the spiral coiling of the visceral dome and
shell on the one hand and the special asymmetry of the asymmetrical organs on the
other. The above-mentioned exceptions to the rule can, however, be explained as
follows. The spiral of a dextrally twisted shell can by degrees become flattened in
such a way that the shell may be simply coiled in one plane or may nearly approach
that condition. In this case the spiral might again assert itself, but on the side

B

D

E

Fia. 141.— Seven forms of Ampullaria shells (diminished in various degrees), seen in the upper
row from the aperture of the shell, in the lower from the dorsal side. The head, foot, and oper-
culum are arbitrarily drawn merely for the purpose of facilitating a comparison between dextrally
and sinistrally twisted shells.

opposite to that on which the umbilicus originally lay, and in this way a false
spiral might form on the umbilical side and a false umbilicus on the spiral side.

The transition from a dextrally twisted to a falsely sinistrally twisted shell, which
latter was, however, genetically dextrally twisted, is illustrated in Fig. 141 by
means of the shells of seven species of the genus Ampullaria. Ampullaria Swain-
soni Ph ? (G) and A. Gevcana Sam (F) are dextrally twisted with distinctly project-
ing spiral. In A. crocastoma Ph (E) the spiral is flat, in A. (Ceratodcs) rotula
Mss. (D) and A. (Ceratodes) ch-iquitensis d'Orb (C) the spiral is already pushed
through or sunk, yet we find a true umbilicus
on the umbilical side. In A. (Lanist.cs) Bul-
teniana Chemn. (B), and still more in A.
purpurca Jon. (A), the false spiral appears on
the umbilical side, and on the spiral side a false
umbilicus is found.

However plausible this explanation may
appear, it can only be proved to be correct if
it is found that where a spiral operculum
occurs, the direction of its spiral is opposite to
that of the spiral of the shell (Fig. 142, A, B,
C), and the commencement of the spiral is
always turned to the umbilical side of the shell,
operculum, but such occur in the Pteropoda.
VOL. II

FIG. 142.

Lanistes has not a spirally twisted
In those Pteropods which combine a

M

162 COMPARATIVE ANATOMY CHAP.

sinistrally twisted shell with the organisation belonging to a dextrally twisted
(Gastropod, the operculum exactly corresponds with that of a dextrally twisted shell.
In Pcraclis, in the larvae of the Cymbuliidcc and in Limacina retroversa Flemniing,
the operculum (the free surface of which must be viewed) is sinistrally twisted, and
the starting-point of the twist faces the (false) spiral, which in these falsely sinistrally
twisted Gastropods lies in the place of the original umbilicus.

Tli is apparent exception is thus shown to be quite in keeping with the rule above
established.

XV. The Sensory Organs.
A. Integumental Sensory Organs.

In the integument of the Mollusca there are epithelial sensory cells
(Flemming's cells), which vary in number and arrangement, and
may be scattered over large areas. Two kinds of these cells may be
distinguished according to their form. One kind, which is found only
in Lamellibranchs, consists of large epithelial cells with large terminal
plates which form part of the body surface and carry tufts of pro-
jecting sensory hairs (" paint-brush cells," Pinsel-Zellen). The second
kind of cells are found in all classes of Mollusca. They are long,
filiform, or spindle-shaped, swelling at one point where the nucleus
lies. They sometimes carry a tuft of sensory hairs, sometimes none.
Each kind of cell is continued at its base into a nerve fibre, which
runs into the nervous system. A distinct specific function can hardly
be attributed to these epithelial cells. They may respond to very
various stimuli, chiefly mechanical and chemical, and thus may act in
an indefinite way as tactile, olfactory, and gustatory cells.

They may become more specialised in function, when crowded
together in certain areas of the body, and may then represent special
sensory organs. Between the individual cells composing such a
sensory organ, however, other epithelial cells (glandular, ciliated,
and supporting cells) are always found.

1. Tactile Organs.

The tactile function of the integumental sensory cells is likely to
assert itself at exposed parts of the body surface, such as the ten-
tacles, epipodial processes, siphons, at the edge of the mantle in the
Lamellibranchia, and at the edge of the foot, etc. We cannot, how-
ever, assume that even in these places the sensory cells are sensitive
only to mechanical stimuli.

2. Olfactory Organs.
(a) The Osphradium.

As has been proved to be the case in the Prosobranchia, sensory cells
occur scattered among the other epithelial cells throughout the whole

vii MOLLUSGA—THE SENSORY ORGANS 163

epithelial lining of the mantle cavity. Here, as in other parts of .the
body, three kinds of epithelial cells can be distinguished : ( 1 ) undifter-
entiated cells, which may contain pigment, and are usually ciliated ;
(2) glandular cells ; (3) sensory cells. The proportions in which
these three kinds of cells appear varies in different regions of the
mantle. If glandular cells prevail on a certain area, that area assumes
a glandular character, and may even develop into a sharply localised
epithelial gland (e.g. the hypobranchial gland). On the gills, undiffer-
entiated ciliated cells predominate. Where sensory cells predominate
a sensory character is given to the region ; such a region, if sharply
circumscribed, the sensory cells continually increasing in number,
becomes a pallial sensory organ. The gradual development and con-
tinuous differentiation of such an organ may be particularly well
traced in the Prosobranchia, the sensory organ developed being the
osphradium. In consequence of its position in the mantle cavity, and
especially on account of its proximity to the gill, it has been assumed
that its principal function is to test the condition of the respiratory
water, or, in other words, that it is an olfactory organ.

The osphradiuru among the Prosobranchia is least differentiated in the Dioto-
'•'irilia. In the Fissurcllidic it does not exist as a sharply localised organ. In the
Monotocardia it becomes more and more differentiated, and lias a special ganglion,
and finally in the Toxiglossa, it reaches the maximum of its development.

A review of the position and number of the osphradia has already been given in
another place (§ V. p. 71). As an example of the special form and structure of this
organ we select the highly developed osphradinm of a Toxiglossa, Cassidaria
tyrrhcna.

The osphradium of Cassidaria is a long organ, pointed at both ends, which lies
to the left of the ctenidium on the mantle in the mantle cavity. As in other highly
specialised Monotocardia (Fig. 71, p. 73) it looks like a gill feathered on both sides, and
has on that account been regarded and described as an accessory gill. It consists of
a ridge rising from the mantle, which in transverse section is almost square, and
carries on each side 125 to 150 flat leaflets, which stand at right angles to the
surface of the mantle, and are so closely crowded that their surfaces are in contact.
The ridge consists almost exclusively of the long osphradial ganglion. Each leaflet
receives from this ganglion a special nerve, which runs along its lower projecting
edge, and sends off four principal branches into it. In its dorsal pallial side each
leaflet contains blood sinuses, which communicate with a sinus lying above the
ganglion in the ridge.

These principal nerves in the leaflets branch, and their last and finest ramifica-
tions penetrate the supporting membrane between the epithelium and the sub-
epithelial tissues. These become connected with the branches of the interepithelial
ganglion cells, each of which again is connected with a spindle-shaped epithelial
sensory cell. The branched interepithelial cells are connected together by their
processes.

The sensory epithelium above described is developed on the lower surfaces of the
osphradial leaflets, i.e. those turned to the man tie cavity, the indifferent, non-ciliated
cells on these surfaces being filled with granules of yellow pigment, while in the upper
surfaces of the leaflets these cells are devoid of pigment and ciliated. Glandular
cells are also found definitely arranged in the epithelium of the osphradial leaflets.

164 COMPARATIVE ANATOMY CHAP.

The osphradial nerve usually springs from the pleuro- visceral connective (from
the parietal ganglion when this is present) ; in the Lamellibranchia it comes from
the parieto-visceral ganglion. The osphradial nerve is generally a lateral branch of
the branchial nerve.

In the Lamellibranchia, the important fact has been demonstrated that, although
the osphradial nerve comes from the parieto-visceral ganglion, its fibres do not
actually rise from this ganglion ; but they pass along the pleuro-visceral connective
and have their roots in the cerebral ganglion.

(b) Olfactory Tentacles.

Certain experiments, to which, however, some exception might be
taken, seem to show that the large optic tentacles of terrestrial
Pulmonata are also olfactory. It is also generally accepted, though
still not certainly established, that the posterior or dorsal tentacles
(rhinophores) of the Opisthobranchia are olfactory organs. These
rhinophores (Fig. 93, p. 98) often show increase of surface, usually
in the shape of more or less numerous circular lamellae surrounding
the tentacle like a collar. The rhinophores are also often ear-shaped
or rolled up conically. Not infrequently they can be retracted into
special pits or sheaths. They are innervated from the cerebral
ganglion by means of a nerve which forms a ganglion at the base of
each.

At the lateral and lower edge of the cephalic disc of the Cephalaspidce, an organ
which is considered to have arisen by the fusion of the labial and cephalic tentacles,
there are structures which are thought to be olfactory, and which, where most
developed, take the form of several parallel " olfactory lamellae " standing up on
the disc.

(c) Olfactory Pits of the Cephalopoda.

In the DibrancMa there is on each side, above the eye, a pit which
is considered to be olfactory. Its epithelial base consists of ciliated
and sensory cells, and underneath it lies, close to the optic ganglion,
an olfactory ganglion. The nerves running to this ganglion come
from the ganglion opticum, but really originate in the cerebral
ganglion. It looks as if these olfactory organs were the remains of
the posterior tentacles of the Gastropoda, and were comparable with
the rhinophores of the Opisthobranchia. In Nautilus the place of the
olfactory pit is occupied by the upper optic tentacle. We have
already seen that Nautilus still retains true osphradia.

(d) The Pallial Sensory Organs of the Lamellibranchia.

Several Asiphoniata have, in addition to the osphradia, epithelial
sensory organs, which lie on small folds or papillae to the right
and left of the anus, between it and the posterior end of the gill.
These are innervated by a branch of the posterior pallial nerve.

Epithelial sensory organs of various forms (plates of sensory epithelium, sensory
lamella4, or papillae, tufts of small tentacles) are found on the mantle in tin-

vii MOLLUSGA—THE SENSORY ORGANS 165

Siphoniata ; these lie on the retractor muscles of the siphons and at the base of the
branchial siphon. These pallial sensory organs also are innervated by the posterior
pallial nerves, and may correspond with the anal sensory organs of the Asiphoniata.
Their function is unknown, but is supposed to lie analogous to that of the
osphradia.

(e) Olfactory Organs of the Chitonidse.

In the mantle furrow of the Chitonidic there are epithelial sensory
organs which are considered to be olfactory. These are ridges and
prominences with extraordinarily high epithelium, consisting of
glandular cells and thread-like sensory cells. In Chiton Icevis and
C. cajetanus there are, on each side of the mantle furrow, two sensory
ridges extending along the whole length of the row of gills ; one of
these, the parietal ridge, belongs to the outer wall of the furrow,
while the paraneural ridge runs along the base of the furrow, above
the bases of the gills and under the pleuro-visceral cord. The para-
neural ridge is continued a short distance along the inner surface of

O

each gill, so that each gill has an epibranchial sensory prominence. In
front of the first pair of gills and near the last the sensory cells in the
paraneural ridge become far more numerous in comparison with the
glandular cells. Chiton siculus, C. Polii, and Acanthochiton (in which
the numerous gills reach far forward) have no parietal and paraneural
ridges. The sensory epithelium in these animals is confined to two
prominences, paraneural in position, behind the last pair of gills,
and connected with a high epithelium covering the pallial wall of
the most posterior part of the furrow.

All these sensory epithelia seem to be innervated from the
pleuro-visceral cords.

The question as to the relation of these sensory epithelia in the Chitonidcc to the
osphradia of other Molluscs, which here presents itself, is difficult to answer. In
position the osphradia best correspond with the epibranchial prolongations of the
paraneural ridges in Chiton Icevis and C. cajetanus.

3. The " Lateral Organs " of the Diotoeardia.

At the bases of the epipodial tentacles of Fissurella and the
Trochidie, and at the base of the lower tentacles of the epipodial ruff
of Haliotis, and also in other parts near the ruff, sensory organs are
found which have been compared with the lateral organs of Annelids.
They consist of patches of sensory epithelium, which may form
either spherical projections or pit-like depressions. The epithelium
of these sensory organs which lie at the lower side of the bases of the
epipodial tentacles, consists of sensory cells, each of which is provided
with a sensory seta, and pigmented supporting cells. Each of these
sensory organs is innervated by the nerve of the tentacle near it,
which nerve originates in the pedal cord and forms a ganglion in the
base of each epipodial tentacle.

166 COMPARATIVE ANATOMY CHAP.

4. Gustatory Organs.

Folds and prominences found in the mouth in some divisions of
the Mollusca have been taken for gustatory organs, although there are
no physiological and hardly any histological grounds for this opinion.
The existence of so-called gustatory pits on a prominence in the buccal
cavity has been proved only in a few Chitonidce and Diotocardia (Haliotis,
Fissurella, Trochiis, Turin, and Patella). This " gustatory prominence "
(which has been best examined in Chiton) lies on the floor of the
buccal cavit}r, close behind the lip. A few gustatory pits are found
in its epithelium, sunk somewhat below the surrounding epithelium.
They consist of sensory cells with freely projecting sensory cones, and
of supporting cells.

On each side of the mouth in the Pulmonata lies an oral lobe,
and under its deep epithelium, which is covered by a thick cuticle,
lies a ganglion. Smaller ganglia are found in the small lobes at
the upper edge of the mouth. All these ganglia receive nerves which
radiate from a branch of the anterior tentacle nerve. These oral
lobes (Semper's organ) are considered to be gustatory organs.

5. Subradular Sensory Organ of Chiton.

In the buccal cavity of Chiton a subradular organ of unknown
physiological significance has been found. It is described as " a pro-
minence lying below and in front of the radula," and in shape re-
sembles two beans with their concave edges turned to one another,
the ends touching ; the space between them forms a channel into
which a small gland opens. Below this organ lie two ganglia, the
subradular or lingual ganglia (cf. section on the nervous system). The
epithelium of the subradular organ consists of green pigmented
ciliated cells and two kinds of sensory cells. A similar organ occurs
in Patella, but has not been thoroughly examined, and at the same
part in various Diotocardia there is a prominence, which, however, has
no sensory cells. The Scaphopoda also possess a subradular organ.

6. The Sensory Organs on the Shell of Chiton.

There are numerous organs definitely arranged on the shell of
the Chitonidfc which have, no doubt correctly, been considered as
sensory, i.e. tactile organs (Fig. 143). They are called aesthetes, and
lie in pores on the tegmentum (cf. p. 39) ; they are club-shaped or cylin-
drical, and each carries a deep cup-like chitinous cap. Each megal-
sesthete gives off all round numerous fine branches or miersesthetes,
each of Avhich ends in a swelling which carries a small chitinous
cap. The body of the aesthetes consists principally of long cells like
glandular cells ; it is produced into a fibre which runs along the
base of the tegmentum, and from here passes together with the

VII

MOLLUSGA—THE SENSORY ORGANS

fibres of the other aesthetes of the shell-plate, between the tegmentum
and articulamentum to the surrounding pallial tissue, or else pene-
trates the articulamentum.

The significance of the separate constituent parts of the esthetes and their
fibrous strands is not yet certainly known. It is probable that they are innervated
from the dorsal lateral branches of the pleuro-visceral cords. It is even not known
whether the fibrous strands are their nerves, or whether the clear fibres running
through them are long sensory cells whose nuclei may lie between the glandular
cells, and in connection with nerve fibres.

We are perhaps justified in assuming that the {esthetes are merely modifications

/nk

Fio. 143.— Section of the tegmentum of Chiton laevis showing an aesthete (after Blumrich).
ink, Micraesthete ; per, periostracum ; sk, principal testliete ; t, tegmentum; (!:, cells resembling
glandular cells ; hf, clear fibres ; /s, fibrous strand ; c, chitinous cap.

of the spines with their papilla; and formative cells, which are so common in the
integument of the Chitonidac. The chitinous cap would then represent part of the
chitinogenous base of the spine.

The sensory nature of the aesthetes is rendered highly probable
by the circumstance that in a few species of Chit/rn individual megal-
resthetes are transformed into eyes.

Each eye is furnished with a pigmented envelope, which is pene-
trated by the micraasthetes, and outwardly covered by an arched
layer of the tegmentum which forms the cornea. Under this is a
lens, and under this again a cell layer, which is regarded as a retina,
and to which is attached a fibrous strand (optic nerve ?) corresponding
with the fibrous strands of the ordinary aesthetes.

B. Auditory Organs.

All Mollusca except the Amphineura possess auditory organs, which
appear very rarely in the embryo. They take the form of two almost

168

COMPARATIVE ANATOMY

CHAP.

closed auditory vesicles (otoeysts), whose epithelial walls usually con-
sist of ciliated and sensory cells. The interior of the otocyst is filled
with fluid and contains a varying number of otoliths (1 to over 100).
These vary in size, form, and chemical constitution, and in the living
animal oscillate in the fluid in which they are suspended.

The otoeysts are usually found on or near the pedal ganglia, rarely
far from it. It is, however, well established that the auditory nerve
does not originate in this ganglion but in the cerebral ganglion, though
it often runs along close to and even in contact with the fibres of
the cerebropedal connective.

In most cases the otoeysts arise as invaginations of the outer
epithelium. An interesting discovery has recently been made, that in

primitive Lamellibranchs
(Nucida, Leda, Yoldia) each
of the otoeysts even in the
adult still opens by means of
a long canal on the surface
of the foot. In such cases
the otoliths are particles of
sand or other foreign matter
taken in from outside. In
Cephaloppds, the remains of
the canal of invagination is
retained (Kolliker's canal),
but it ends blindly.

The auditory organs are
most highly developed in
those Molluscs which are
good swimmers, especially in
the Cephalopoda and Hetero-
poda. Among these, maeulse
and eristse aeustiese are
developed.

Heteropoda. — The struc-

FIG. 144. — Auditory organ of Pterotrachea (after PI

Glaus). 1, Auditory nerve; 2, structureless membrane; ture of the auditory Organ
3 and 4, ciliated cells; 5, otolith ; 6, auditory cells; 7, of Pterotrachea (Fig. 144),

2?.°rting OI'y which has been thoroughly

examined, is as follows : —

The wall of the otocyst consists in the first place of a structure-
less membrane surrounded by muscle and connective tissue. Inside
the vesicle, which is filled with fluid, a calcareous otolith, built up of
concentric layers, is suspended. The inner surface of the vesicle is
lined by an epithelium, containing three different sorts of cells :
auditory, ciliated, and supporting cells. The auditory cells, which
carry immobile sensory hairs, are found on the wall of the otocyst at
a point (macula acustica) diametrically opposite to the place where the
auditory nerve enters. At this spot there is a patch formed of

VII

MOLLUSOA—THE SENSORY ORGANS

169

numerous auditory cells, and in their midst, separated from the rest by
four supporting or isolating cells, one large central auditory cell.
On the larger remaining surface of the wall of the otocyst, separated
by undifferentiated cells, are found flatter ciliated cells, which carry
very long cilia or setse, exhibiting peculiar movements. They some-
times lie flat along the inner wall of the vesicle, and at other times (it
is said in response to strong auditory stimuli) stand upright, projecting
towards the centre of the vesicle, and supporting the otolith.

The auditory nerve, which enters the otocyst at a point exactly
opposite the central cell, at once radiates in the form of fibres over the
whole wall of the vesicle " as meridians radiate from the pole on a
globe," finally innervating the bases of the auditory cells.

The two otocysts of the Cephalopoda are still more complicated ;
they lie in two spacious cavities of the cephalic cartilage. The sensory
epithelium is here found on a macula acustica and on a kind of ridge,
the crista acustica, which projects inwards. Otoliths are only found
on the macula acustica. The auditory nerve divides into two branches,
one going to the macula, and the other to the crista acustica. Kolliker's
canal, above mentioned, which is internally ciliated and ends blindly,
runs out of the otocyst as the remains of the aperture of the original

mvagmation.

Experiments made on Ceplwlopods have shown that one of the
functions of the otocysts is to regulate the position of the animal
while swimming.

C. Visual Organs.
1. Optic Pits.

These are the simplest form of visual organ. They are cup-shaped
depressions of the body
epithelium, which at the
base of the cup forms the
retina. The depression is
sometimes very shallow, at
other times deep, and like
a wide bottle with a short
narrow neck. The optic
nerve enters at the base of
the depression and spreads
out over it. The epithelial
wall or retina consists, ap-
parently in all Gastropoda,

of two kinds of long thread-

FIG. 145.— Eye of Nautilus (alter Hensen). 1, Optic
cavity (pit) ; 2, layers of rods ; 3, pigment layer ; 4, layer of
visual cells ; 5, layer of ganglion cells ; (5, branches of the
optic nerve.

like cells: (1) clear cells
without pigment, and (2)
pigmented cells. Whether
either or possibly both of these kinds can be considered as retinal cells

170

COMPARATIVE ANATOMY

CHAP.

is still a disputed question. In certain cases it has been proved that
the pigment in the second kind lies peripherally ; the axis is free
from pigment, and may perhaps be considered as the sensitive portion
of the cell. In this case, the clear cells would be undifferentiated
supporting cells, or secreting cells. The retina is covered, on that
side of it which faces the cavity, by a thick gelatinous cuticle, or the
whole cavity is filled by a gelatinous body often called a lens. The
clear or secreting cells have been thought to yield this gelatinous mass,
but there is a tendency to regard them now rather as retinal cells.

Optic pits are, among the Gastropoda, only found in such Dintocardia as show
primitive characteristics, e.g.Haliotidtc, Patcllidc?, 7rocfcidce,DelpJviiw.ilidce, and Stonu/-
tiidcc.

In connection with the claim that JVcc/iVcs (Fig. 145) is the most primitive form
among extant CcpJialopoda, it is interesting to find that both its eyes are optic pits.
Each sensory cell of the retina, i.e. of the epithelial wall of the depression, possesses
a cuticular rod projecting towards the cavity, and a layer of ganglion cells is
intercalated between the ramifications of the optic nerve and the retina.

2. Optie Vesicles or Vesicular Eyes.

Optic vesicles are developed from optic pits both ontogenetically
and phylogenetically by the approximation of the edges of the pit,

which finally fuse. A vesicle is thus
formed, over which there is a continuous
layer of epithelium (Fig. 146). The
outer epithelium is free from pigment
over the eye, and is called the outer
cornea, while the immediately subjacent,
and also unpigmented, epithelial wall of
the vesicle forms the inner cornea. The
epithelial base of the original depression
here again forms the retina ; its cells
contain distinct rods projecting towards
the cavity of the vesicle, which is filled
with a gelatinous mass. The optic
nerve usually swells into a peripheral
ganglion opticum before reaching the
retina.

The tentacular eyes of most Guxtro-

Fio. 140. -Eye of a Pulmonate. 1, J

Outer, 2, inner cornea ; 3, body epithe- pOttd, CXCBpt those IhotOMn/UI which

limn ; 4, vitreous body ; 5, retina ; C,, have Clip-like eyes, are of this simple

ganglion opticum Coptic nerve, character.

3. The Eye of the Dibranehiate Cephalopoda.

This is one of the most highly-developed eyes in the whole animal
It is a further development of the cup-shaped and vesicular

kingdom.

VII

MOLLUBGA-^THE SENSORY ORGANS

171

eyes. In the Tetralrancldate Nautilus, as we have seen, the cup-shaped
eye persists throughout life.

These lower stages (i.e. the cup-shaped and vesicular stages) of the
eye are passed through ontogenetically. First a cup-like depression is
foi-med (primary optic pit), then this becomes constricted to form a
vesicle (primary optic vesicle), the inner wall of which becomes the
retina, while the outer (which corresponds with the inner cornea of
the vesicular eye) becomes the inner corpus epitheliale. This em-
bryonic optic vesicle then becomes further complicated : the integu-
ment over it (the outer cornea of the vesicular eye) rises in the form

FIG. 1'4V.— Development of the eye of the dibranchiate Cephalopoda. 1, Body epithelium,
which becomes the outer corpus epitheliale ; -2, inner wall of the optic depression, which becomes
the retina ; 3, outer wall of the optic vesicle, which becomes the inner corpus epitheliale ; 4, fold
which forms the iris ; 5, fold which forms the secondary cornea ; 6, portion of the lens formed by
the outer corpus epitheliale ; 7, portion of the same formed by the inner corpus epitheliale ; s, rod
layer of the retina.

of a circular rampart, and then grows forward towards the axis of
the eye like a diaphragm, which forms the iris, the aperture left in
the same being the pupil. The integument which spreads out over
the circular base of the iris is in close contact with the inner corpus
epitheliale, and becomes the outer corpus epitheliale.

The inner corpus epitheliale forms towards the cavity of the primary
vesicle an almost hemispherical lens, the outer corpus epitheliale form-
ing a similar lens outwards towards the pupil. The two hemispheres
lie in such a way as to form something like a complete sphere ; its

172

COMPARATIVE ANATOMY

CHAP.

two-fold origin, however, always remains evident, its equatorial plane
being traversed by the double lamella of the corpus epitheliale.

A new circular fold grows over the eye, forming a fresh cavity
over it ; this is the secondary cornea of the dibranchiate eye, which
must not be confounded with the primary cornea of the optic vesicle
here represented by the corpus epitheliale. In most forms the circular
fold (cornea) does not altogether close over the e}^e ; an aperture
remains through which the water can enter the anterior chamber of

II —

11 —

11

FIG. 14S.— Section of the eye of Sepia offlcinalis, somewhat diagrammatic (after Hensen).
1-8, As in Fig. 147 ; 1 + 3, corpus epitheliale ; 9, anterior chamber of the eye opening outward at
10; 11, cartilaginous capsule; 12, ganglion opticum = retinal ganglion; 13, nervus opticus ; •-'",
pigment layer of the retina.

the eye. In some animals, however, the secondary cornea closes com-
pletely.

We thus obtain, ontogenetically, some idea of the general structure
of the dibranchiate eye. A few details of the structure of the adult
eye are given below (Figs. 148 and 149).

1. The retina (Fig. 149) consists of two kinds of cells — (1) pig1-
mented visual or rod cells, and (2) limiting cells. Since the nuclei
of the visual cells form, with relation to the centre of the vesicle, an
outer, and the nuclei of the limiting cells an inner layer, and since,
between these two layers, a limiting membrane traverses the inter-
stices between the retinal cells, the retina appears to be laminated,
whereas it in reality consists of one layer of cells. The rods of the

VII

MOLLUSCA—THE SENSORY ORGANS

173

retinal cells lie on the inner side of the limiting membrane, and are
thus turned to the source of light and at the same ...,..:,.,..,:. .,,...,. .
time to the cavity of the primary vesicle. The
retina is covered on its inner side by a somewhat
thick membrana limitans.

2. The eye is surrounded, except on the side
turned to the surface of the body, by a cartila-
ginous capsule, which resembles the sclerotica in
the vertebrate eye ; this cartilage, where it covers
the retina, is perforated like a sieve, so that the
optic nerves can pass through it.

3. Immediately underneath the cartilaginous
floor of the retina lies a very large ganglion opticum,
in the form of a massive cerebral lobe. From this
rise the nerves which run to the retina through the
perforations of the cartilaginous capsule.

4. The two halves of the lens, which are unequal
in size (the outer being the smaller), consist of
homogeneous concentric laminae.

5. The cavity of the primary vesicle (between
the retina and the lens) is filled with perfectly
transparent fluid.

It has been proved that,, as in the Arthropoda
and Vertebrata, the pigment granules of the rod
cells, which in the dark lie at the base of the cell,
under the influence of light travel towards its free
end.

4. The Dorsal Eyes of Oneidium and the Eyes FlG. i4!-i.-Two retinal
at the edge of the Mantle in Peeten (Fig. cells of a ceptaiopoa,

much magnified (after

150) and Spondylus. Grenacher). 1, Meiu-

lirana limitans ; •>,, pig-

These eyes have been said to resemble vertebrate ment ; 3, secreted
eyes in structure, because in them the visual rods f^^'^^
are turned away from the light, being- directed limiting ce'n ; s, limiting
inwards towards the body.

They are vesicular eyes, but in them it is the
outer wall of the vesicle, that turned to the light, which becomes the
retina, while the inner wall (which in other Molluscs forms the retina)
is a pigmented epithelium. At the same time the outer or retinal wall
is invaginated towards the inner pigmented wall, as is the endoderm
towards the ectoderm in the formation of the gastrula. The conse-
quence of this is, that the cavity which in other Mollusca is filled by the
gelatinous mass (lens) disappears, and the vesicle becomes a flattened
thick-walled plate (Peeten) or cup (Oneidium), consisting of a pigment
layer and a retina. The body epithelium which passes over the eye is
unpigmented and transparent, and here becomes the cornea. Beneath

membrane ; 0, retinal
cell ; 10, nerve fibre.

174

COMPARATIVE ANATOMY

CHAP.

the cornea, within the optic cup or on the plate, lies a cellular lens,
which in the dorsal eyes of Oncidium consists of a few (5) large cells,
but in the pallial eyes of Pecten and Spondylus of very numerous cells.

-ft

Fie. 150.— Section through the eye of Pecten (after Patten). <•, Cornea; I, lens; cp, pig-
mented body epithelium ; <j, layer of ganglion cells ; r, retina ; st, rod layer of the retina ; d, tap-
••tuiii ; f, pigmented epithelium ; /, sclerotica ; «, optic nerve ; n\ and mo> its two branches.

The development of this lens is unknown ; it is perhaps formed by a
thickening or invagination of the embryonic ectoderm which covers
the eye.

In Oncidium, the optic nerve penetrates the wall of the optic cup, as in the
vertebrate eye, to spread out on the inner surface (with regard to the centre of the
vesicle) of the retina, and to innervate the retinal cells.

VII

MOLLUSGA—THE SENSORY ORGANS

175

In Pcctrn, the optic nerve which runs to each eye from the nerve tor the pallial
edge, divides, close to the eye, into two branches. One of these runs to the base of
the optic plate, and there breaks up into fibres, which radiate on all sides to the edge
of the plate, then bend over towards the retina to innervate some of its cells. The
other branch runs direct to the edge of the plate, there bends round at a right angle
and supplies nerves to the rest of the nerve cells. The fibres of this branch are not,
however, directly connected with the retinal or rod cells, as there is a layer of anas-
tomosing ganglion cells interposed between the two. Between the pigmeuted epi-
thelium and the rod layer of the retina, a tapetum lucidum is found, which gives the
eye of the Pectcn its metallic lustre.

Dorsal eyes are found in many species of Oncidium. They lie at the tips of the
contractile papilla? found on the dorsal integument of this curious Pulnwnafe ; on each
papilla three or four such eyes occur. Besides these, Oncidium has the two normal
cephalic eyes usually found in Gastropods.

The pallial eyes of the LamcUibranckiatcs, Pectcn and .Spoiidi/lus, are found in
large numbers on the edge of the mantle, between the longer tentacles, and on the
tips of shorter tentacles. The rods of the retina in Pccten, when fresh, are of a very
evanescent red colour (visual purple ?).

5. The Eyes on the Shell of Chiton.

These have already been described (p. 167). Their morphological significance
cannot be determined as long as their development is unknown and their histological
.structure imperfectly investigated.

6. The Compound Eyes of Area (Fig. 151) and Peetuneulus.

These are found in great numbers at the edge of the mantle, and
are epithelial organs which do not in any way agree in structure with
the other visual organs foiind in
Mollusca, but rather resemble
certain simple Arthropodan
eyes.

In form they resemble an
externally convex shell. The
unilaminar epithelial wall of
the shell passes, at its edge,
into the surrounding pallial
epithelium. In section, its com-
ponent elements appear to be
arranged like a fan ("Facher-
auge "). These elements are of

three kinds: (1) conical visual (adapted from Rawitz). I, Retinal cells with rod-like
Cells, with their bases turned Bodies (2); 3, lament cdls; 4, slender interstitial

outwards; (2) a sheath of six

cylindrical pigment shells surrounding each visual cell. Each group,
consisting of one visual cell and its surrounding pigment cells, may be
considered as a single eye or ommatidium of the simplest structure, in
which the retinula is represented by one single visual cell. (3) Slender,
almost thread-like interstitial cells which stand between the ommatidia.

FlQ 151._Section of the eye of Arca barbata

176 COMPARATIVE ANATOMY CHAP.

7. Degeneration of the Cephalic Eyes.

It is becoming more and more probable that the cephalic eyes of the various
Mollusca are homologous structures, and that they primitively occurred in all forms.
They may, however, under certain biological conditions become rudimentary, and
even disappear, as in boring animals and those living in mud or in the deep sea
and in parasitic Molluscs. The Lamellibranchia and Chitonidw (?) even have
cephalic eyes appearing temporarily during development ; they disappear later,
when, covered by the shell, they are useless. They may be replaced by secondarily
acquired visual organs arising at more suitable parts of the body, and thus we
have eyes 011 the mantle edge in some bivalves and on the shell of some Chitonidce.

XVI. The Alimentary Canal.

The alimentary canal is well developed in all Molluscs, and is
composed of (1) the buccal cavity; (2) the pharynx or cesophageal
bulb; (3) the oesophagus or fore-gut; (4) the mid-gut with the
stomach ; (5) the rectum or hind-gut with the anal aperture. The
mouth originally lies at the anterior, and the anus at the posterior
end or side of the body, the latter in the mantle furrow or cavity.
The former always retains its original position, but the latter, as
central organ in the pallial complex, becomes shifted more or less far
forward along the right (less frequently the left) side, in the mantle
furrow.

When the visceral dome grows out dorsally in such a way that
the longitudinal axis becomes shorter than the dorso-ventral axis, as
is the case in many Gastropods and Cephalopods and in Dentalium, the
mid-gut at least, with its accessory gland, the so-called liver, runs up
into this dome, filling the greater part of it. The intestine then forms
a dorsal loop, consisting of an ascending portion running up from the
fore-gut and a descending portion running down to the anus. In the
Gastropoda, where the anus is shifted more or less forward, the
descending portion bends forward to the right (rarel}' to the left) to
reach it.

Besides this principal visceral loop, which is caused by the
development of the visceral dome and modified by the displacement of
the pallial complex, the intestine, in nearly all Molluscs, forms secondary
loops or coils which add to its length. These loops are found
principally in the tubular portion of the mid-gut which follows the
stomach. They are as a rule most pronounced in herbivorous
animals, Avhich thus have longer alimentary canals than carnivorous
forms.

The large digestive gland, usually called the liver, enters the
stomachal division of the mid-gut. Functionally, this organ only
very slightly corresponds with the vertebrate liver, if indeed it may
be said to correspond at all with that organ. It agrees more nearly

vii MOLLUSCA—THE ALIMENTARY CANAL 177

with the pancreas, and perhaps combines the functions of the different
specialised digestive glands of Vertebrates.

There is a radical difference between Lamellibranchs and other
Molluscs,1 in the fact that in the latter the anterior portion of the fore-
gut which follows the buccal cavity is developed as a muscular pharynx
(oasophageal bulb, buccal mass), and carries at its base on a movable
lingual cushion a file -like organ, the radula, which is beset with
numerous hard teeth composed of conchyolin or chitin. The radula
serves chiefly for mastication, but is sometimes used in seizing, holding,
and swallowing prey.

None of the Lamellibranchs have a pharynx provided with a radula,
they are therefore called Aglossa as opposed to all other Molluscs, which
are Glossophora.

Hard jaws, composed of conchyolin, are almost always found in
varying number and arrangement in the buccal caArity of the Glosso-
phora, but are wanting in all Lamellibranchs.

One or two pairs of glands open into the pharynx in the Glossophora/
these are usually called salivary or buccal glands, although they very
slightly if at all correspond physiologically with the glands so named
in the Vertebrata. Glands may also open into the buccal cavity. The
Lamellibranchs have no salivary glands.

The absence of the pharynx, tongue, jaws, and salivary glands in
the Lamellibranchia is accounted for by their manner of life. They do
not have to seek their food. Some of them are attached and others
feed in the same way as attached animals on small particles suspended
in the respiratory water (animalcula?, microscopic alga?, and particles of
detritus) which are brought to the mouth by means of ciliary move-
ment. These fine particles require no mastication before being
swallowed.

This method of feeding also affects the outer organisation of the
Lamellibranchia, which have lost the cephalic portion of the body with
the tentacles and eyes : Aglossa — Acephala = Lipocephala, and
Glossophora = Cephalophora.

In some Gastropoda (Murcx, Purpura) and in Dentalium there is in
connection with the last part of the hind-gut an anal gland, and in
the Cephalopoda (excepting Nautilus) a gland known as the ink-bag".

The alimentary canal of the Mollusca runs through the primary
and often also through the secondary body cavity, attached in various
ways by fibres or bands of connective tissue. Its walls consist of an
inner epithelium usually to a great extent ciliated, an outer muscular
layer in which longitudinal and circular fibres occur, not always in
regular layers, and, where it passes through the primary body cavity,
an outer envelope of connective tissue.

The pharynx and perhaps sometimes part of the oesophagus, and a
part, in all cases very short, of the hind-gut, arise ontogenetically out of
the ectodermal stomoda?um and proctodseum. But the exact limits

1 For the rare exceptions to this rule, see p. 183.
VOL. II N

178 COMPARATIVE ANATOMY CHAP.

of the ectodermal and the endodermal portions of the intestine are
difficult to determine.

A. Buccal Cavity, Snout, Proboscis.

The alimentary canal has an oral aperture bordered by variously - shaped lips,
and in many Glossoplwra (in nearly all Gastropoda) leads into a vestibule or
anterior cavity roofed over by the lips and lined by a continuation of the outer wall
of the head. The dermal glands are not unfrequently (many Opisfhobranchia and
a few Prosobranchia) more strongly developed on the lips as labial glands. In
many Gastropods, when the lips open, the mouth is able to seize and hold prey like
a sucker.

Where the snout is short it is simply contractile (the Chitonidce., the Dioto-
cardia, most herbivorous Tcenioglossa, and many Pulmonata and Nudibmncliin}.
In this case the parts immediately surrounding the mouth are so strongly con-
tractile that when contraction takes place the mouth is drawn in somewhat so as
to lie at the base of a depression. An exaggeration of this arrangement, combined
with the prolongation of the snout, leads to the formation of the retractile or
proboscidal snout. The snout can in such cases be invaginated from its tip, i.e.
from the oral aperture into the cephalic cavity, the mouth then lying at the base of
the invagination (many Tectibranchia, Capulidoc, Strombidce, Chenopidce, Calyptrceidce,
Cypraeidcc, Lamellariidce, Natiddce, Scalaridce, Solariidce).

Finally, in many carnivorous Prosobranchi-a (Tritoniidce, Doliithe, Cassididn .
Illicit iglossa, and a few Toxoglossa) a proboscis, often very long and enclosed in a
special proboscidal sheath, is developed (Figs. 71 and 152) ; this sheath lies in the
cavity of the head, which is often prolonged like a snout, and may even stretch back
into the body cavity. The oral aperture lies at the free anterior end of the
cylindrical proboscis, and we have to regard the proboscis with its sheath as a very
long snout, the base of which, however, is permanently invaginated into itself. In
this way the proximal portion of the snout forms the permanent proboscidal sheath,
while the distal portion with its terminal oral aperture forms the proboscis. Neither
of these portions can be invaginated or evaginated ; it is merely a zone lying
between them which takes part in the retraction of the proboscis into the body
cavity. This zone, when so invaginated, forms a temporary backward prolongation
of the proboscidal sheath, but when the proboscis is protruded forms the basal
portion of the latter. The permanent portion of the proboscidal sheath is connected
with the wall of the head by bauds which make its evagination impossible, and the
inner .wall of the permanent proboscis is connected by muscles or bands with the
oesophagus lying within it, so that this portion of the organ cannot be invaginated ;
the oral aperture can thus never lie at the base of the proboscidal sheath.

When the proboscis is retracted, there is therefore an aperture at the anterior
end of the snout or the head, which is not the oral aperture, but that of the
proboscidal sheath. When the proboscis is protruded, it projects beyond the
aperture of the sheath and carries at its point the oral aperture.

The proboscis is retracted by means of muscles attached at the one end to the
body wall and at the other to its (invaginable) base. In its protrusion, a flow of
blood towards the snout probably plays the chief part, accompanied by contraction
of the circular muscles of the head and proboscis.

The (carnivorous) Plcropoda gymnosomata also have a protrusible proboscis (Fig.
17, p. 11) provided with so-called buccal appendages. The same is present in the
allied Apl iisiidd', but is weakly developed. The Thecosomata have no proboscis.
The buccal cavity of Dcntalium is noteworthy. It extends throughout the

VII

MOLLUSCA—THE ALIMENTARY CANAL

179

whole length of the freely-projecting egg-shaped snout, which carries leaf-like labial
appendages. On each side of the buccal cavity there is a ponch, the so-called cheek
pouch, which is lined with glandular epithelium and opens into the cavity anteriorly.

FIG. 152.- Diagram of the proboscidal apparatus of the Prosobranchia. A, proboscis
retracted. B, The same protruded, u-c, Cephalic integument ; c, edge of the aperture of tin'
proboscidal sheath ; c-d, immovable wall of the proboscidal sheath ; d-e, movable (evaginable and
invaginable) wall of the same ; e-f, immovable wall of the proboscis ;/, edge of the oral aperture, at
the anterior end of the'proboseis ; g, pharynx ; h, oesophagus ; i, retractor muscle ; k, salivary
glands ; I, cephalic cavity.

An exact comparative investigation of the mechanism of the proboscidal
apparatus, the contractile snout, etc. of the Prosobranchia is still a desideratum.

There are other forms of proboscis, differing greatly from the one just described
(e.g. that in the Tcrebridce).

In the Heteropoda, the head forms a long snout which is often described as a
proboscis. The name is inappropriate, as this snout is not retractile and the mouth
is always found at its anterior end.

180 COMPARATIVE ANATOMY CHAP.

B. The Pharynx and Jaws, the Tongue and Salivary Glands.

The mouth or buccal cavity is followed in all Molluscs except the
Lamellilranchin by the pharynx or cesophageal bulb (buccal mass).
The pharyngeal cavity opens anteriorly into the buccal cavity, and
posteriorly into the oesophagus. The pharynx is characterised by the
possession of (1) jaws, which lie anteriorly at the boundary between
the buccal and pharyngeal cavities; (2) a lingual apparatus at its
base, and (3) salivary glands, which usually open laterally near its
posterior boundary.

1. Jaws are almost universal, and are sometimes, especially in
carnivorous animals, very highly developed ; less frequently they are
rudimentary or wanting. They are hard cuticular formations of the
epithelium of the anterior pharyngeal region, and no doubt composed
of conchyolin or some related substance, in a few cases hardened by
calcareous deposits (e.g. Nautilus}.

The jaws serve for seizing prey or particles of food. The great variations in
number, form, and arrangement of the jaws can best be understood by assuming
that they originally extended completely round the entrance to the pharynx ; and
that of this ring sometimes only upper and lower or sometimes only lateral portions
have been retained.

Such a complete circle of jaws is found at the entrance to the pharynx in some
forms, such as Umbrella and Tylodina (Opisthobranchia).

The fresh-water Pulmonatcs have an upper and two lateral jaws.

Most Prosobranchia and Opisthobranchia have two lateral jaws. These may
approach so near one another as almost to touch (Haliotis, Fissurella). Terrestrial
Pulmonata have an upper jaw and occasionally a weak lower jaw as well.

The jaws are particularly strongly developed in the Cephalopoda, which have an
upper and a lower jaw, the two together resembling in shape the beak of a parrot.

In the Opisthobranchiate family Aphjsiidcc, JVotarchus, Accra, Dolabclla, and
Aplysiella have, besides the lateral jaws, numerous hooks or small teeth on the roof of
the pharyngeal cavity. The hook sacs (Fig. 17, p. 11) of the Pteropoda rjymnoso-
mata, which are wanting only in Halopsyclic, are perhaps to be derived from these
pharyngeal teeth.

The hook sacs are paired dorsal outgrowths of the pharyngeal cavity, which vary
in length and lie in front of the raclula. The walls of the sacs carry hooks project-
ing inward. When the proboscis of these carnivorous animals is protruded, the sacs
are completely evaginated, so that the hooks come to lie outside (Fig. 17, p. 11).

Jaws are wanting or rudimentary in the Amphincura and the Scaphopmln :
among the Prosobranchia, in the Tozoglossa, Pyramidellidce, Eulimidcc, many
Trochidcc, the Heteropoda, and in many Nudibranchia (Tctlnjs, Melibe, Doridupsis,
Phyllidia] ; in the Ascoglossa, and in certain Tcctibranchia (Actceon, Doridium,
Philinc, Utriculus, Scapli«mli'i\ Lobigcr). Among the Pulmonata they disappear
in a series of Tcstaccllidcc, being present in Daudebardia rt/fa, rudimentary in
D. Saulcyi, and wanting in Tcst<io1hi.

2. The lingual apparatus (Figs. 153, 154) is highly characteristic
of all Molluscs except the Lamellibranchia, i.e. of all Glossophora. It
may be said that every animal with a radula is a Mollusc.

VII

MOLLUSOA—THE ALIMENTARY CANAL

181

The ventral and lateral walls of the pharynx are thickened and
very muscular. On the floor of the cavity rises a tough longitudinal
muscular cushion, the tongue. Its surface, which projects into the

0-..J

FIG. 153.— Longitudinal section (not quite median) through the snout of a Prosobranchiate,
to illustrate the pharyngeal apparatus. 1, Dorsal wall of the head ; 2, mouth ; 3, .jaw ; 4,
radula ; 5, lingual cartilage ; 6, muscular wall of the pharynx ; 7, muscles attached at one end to the
pharynx and at the other to the ventral wall of the head (S) ; 9, cavity of the head ; 10, radular
sheath ; 11, oesophagus ; 12, aperture of the salivary gland ; 13, infolding behind the radular sheath.

pharyngeal cavity, is covered by a rough cuticle consisting of chitin
(or conchyolin ?) ; on this basal membrane are found very numerous
hard chitinous teeth, often many thousands, arranged in close transverse

FIG. lo4.— Median longitudinal section through the anterior part of the body of Helix
(after Howes), a; (Esophagus; rd, radular sheath; tic, cerebral ganglion; sU, aperture of the
salivary gland ; oc, muscle mass in the ventral pharyngeal wall ; rd, radula ; hj, upper jaw ; ?j, ?.j,
lips of the oral aperture; im, pharyngeal muscles; rm», retractors of the pharynx; prjl, pedal
gland.

and longitudinal rows. The basal membrane and the teeth together
form the radula of the tongue.

The anterior end of the tongue projects freely into the pharyngeal
cavity, the radula bending down over this end so as to cover for a

182

COMPARATIVE ANATOMY

CHAP.

certain distance its lower surface. Immediate!}' in front of the tongue
there is always a depression in the ventral pharyngeal wall, forming
a sort of pocket. The radula, at the posterior extremity of the
tongue, sinks into a narrow more or less long tube, the radular sheath,
which is an outgrowth of the pharyngeal cavity running downward
and backward. The radula, always lying upon the anterior or
ventral wall of this sheath, which is anteriorly thickened to form
the tongue, extends to the base of the sheath, which is the place
of its formation.

The tongue with, the radula on it is movable, and in most cases its movements can
be compared with those of the cat's tongue when licking, but are usually slower.
This action helps to rasp the food which has been seized, and often also broken up, by
the jaws. The tongue can either move inside the pharyngeal or buccal cavities, or
can be extended to the oral aperture or even protruded more or less far beyond it.

In or under the fleshy tongue, a lingual cartilage is very commonly found,
consisting of two or four or even more pieces. This cartilage forms a support for

the radula, and affords firm points of attachment
for certain muscles belonging to the lingual
apparatus.

The musculature of the pharynx, which can
be separated into bundles or strands, and is often
very complicated, consists first of the muscles
which form the wall of the pharynx, and which,
being principally developed veutrally and laterally
round the radula, determine the special licking
movements of the tongue ; secondly, of muscles
which move the whole pharynx or the whole of the
lingual apparatus, evaginating or protruding them.
The second group consists, speaking generally, of protractors and retractors, attached
at the one end to the pharynx and at the other to the body wall after running
through the cephalic or body cavity. Pressure of blood may also take some part in
the protrusion of the pharynx.

The tongue and radula further often serve for seizing prey (e.g. in the carnivorous
HeteroiHuht, in Testacclla, etc.).

The radula is of great importance in classification. Further details concerning
it must be sought in special works and in text-books of conchology. The points to
be specially noticed are (1) the size and form of the whole radula, (2) the number of
longitudinal and transverse rows of teeth, and (3) the form of the teeth in each of
these rows. As a rule the transverse rows resemble one another, but exceptional
rows differently constituted from those immediately preceding or following them
recur at intervals.

Three kinds of teeth have been, as a rule, distinguished. First, there is usually
a single median longitudinal row of central or rachial teeth. On each side of this
row are several rows of more or less similar lateral teeth or pleurae. Finally, at the
lateral edges of the radnla, there are single or very numerous longitudinal rows of
marginal teeth or uncini.

Dental formulae are used for the radular teeth, in the same way as for the teeth
of mammals ; in these the number of central, lateral, and marginal teeth in a
transverse row are given.

The reader will find the dental formulae of some of the Molluscs in the Systematic
Review.

FIG. 155.— Four transverse rows
of the radular teeth of Cyclostoma
Elegans (after Claparede).

VII

MOLLUSCA—THE ALIMENTARY CANAL

183

The total number of radular teeth varies very greatly, from 16 in Eolis
to 39,596 in Helix GMesbreghti.

As a rule, the teeth are most numerous and finest in herbivorous animals. In
carnivorous Molluscs we have two extremes : (1) great development of the proboscis,
with weak development of the pharynx and radula, and a comparatively small
number of teeth (carnivorous Prosobnouhia) ; (2) absence of a protrusible proboscis,
with great development of the pharyngeal apparatus and the radula, and numerous,
often large, teeth (Hetcropoda, carnivorous Pulmonata and Cephalopoda).

The muscular pharynx is most developed in carnivorous Pulmonates. In these
it may be half (Dandebardia) or even more than half as long (Testacella) as the
whole body, and may occupy a very large part of the body cavity. It is protruded
in such a way that the tongue with the radula occupy the anterior end of the
evaginated pharynx (Fig. 54, A, p. 44).

In very rare cases (apart from the LamelUbrancMa) the radula completely
atrophies ; this is the case in parasitic Gastropoda (Stilifer, Eulima, Thijca, Entn-
<•«/"•/<«)> in tne ConiUwphilidce (Coralliophila, Lcptoconchus, Magilus, Rhizochilus),
among the Nudibranchia in Tetliys and Mtlibc, among the Amphincura in Neomcnia,
and certain species of the genera Dondfrsin and Proncomcnia. In Chcetoderma, a
single tooth of the radula is retained.

Even in certain carnivorous Prosobranchia which are furnished with a proboscis,
the above-mentioned reduction of the whole pharyngeal apparatus goes so far that
the radula disappears (certain species of Tcrebra).

Formation of the Radula.

The teeth of the anterior transverse rows of the radula become worn out by use,
and are continually being replaced by new teeth which are pushed forward. The
formation of new transverse rows of teeth
is constantly taking place at the posterior
blind end of the radular sheath. In Pul-
monata and Opisthobranchia they appear as
cuticular formations secreted by several
transverse rows of large epithelial cells —
the odontoblasts (Fig. 156) ; the basal
membrane which carries the teeth is secreted
by the anterior row or rows, the teeth them-
selves by the posterior rows.

Each group of odontoblasts which has
formed a tooth is not replaced by another,

formative cells of the basal membrane ; <'•,
teeth of the radula ; 8, basal membrane.

FIG. 150.— Longitudinal section through
the posterior end. of the radular sheath of
but continues to produce new teeth behind a Pulmonate (after Rdssler), diagram, i, -2,
those already formed, so that for each longi- 3' 4' Formative cells of the radular teeth ; 5,
tudinal row of teeth there is at the base of
the radular sheath a group of odontoblasts
which has produced all the teeth belonging to that row. A layer of "enamel" is
deposited on the teeth so formed by the epithelial roof of the radular sheath.

In the Chitonidce, Prosobranchia, and Cephalopoda, the odontoblasts are very
numerous narrow cells, which form, at the base of the sheath, a cushion divided into
as many parts as there are teeth in a transverse row of the radula.

The radular sheath in the Pulmonata, Scaphopoda, Opisthobranchia, and Cephalo-
poda is short, and is contained in the ventral and posterior muscular wall of the
pharynx, very seldom projecting posteriorly beyond it ; but in many Prosobranchia
it is long and narrow, and reaches back into the cephalic cavity or even right into
the body cavity. This latter is especially the case in the Diotocardia ; in the

184 COMPARATIVE ANATOMY CHAP.

Docoglossa (Patella) the sheath, which lies above the foot on the floor of the body
cavity, is even longer than the body (Fig. 158).

3. Salivary glands (buccal glands, pharyngeal glands) are universally
found in Glossophora, i.e. in Molluscs which have a pharynx and lingual
apparatus. They are universally absent in Lamdlibranchs. They may
occur in one or two pairs. The posterior or in other cases the only
pair often lies on the wall of the oesophagus, and sends forward two
ducts which enter the pharynx laterally, usually somewhat behind the
point where the radular sheath opens into the pharyngeal cavity.
Very little is known of the function of these glands ; an exact
morphological comparison of the various pharyngeal glands of the
Gastropoda is at present hardly possible.

Amphineura. (a) Chiton. — Two small delicate buccal glands lie
on the roof of the buccal cavity and open into the mouth. They can
therefore hardly be regarded as pharyngeal or salivary glands.

(b) Solenoifastres. — Salivary glands are here found in all genera
except Neomenia, and in Chcetoderma. They are present in some species
but appear to be absent in others. A pair of long glandular tubes
with high glandular cells x and strong muscular walls lie anteriorly under
the intestine and are produced in the form of two narrow ducts, which
enter the pharyngeal cavity on the tongue either separately or through
a common terminal portion. Besides these there is another pair in
some species (Paramenia impexa, Param. palifera, Proneomenia vagans,
Dondersia, flavens) ; the ducts of these open together through an un-
paired terminal portion on the dorsal wall of the pharyngeal cavity,
at the point of a papilla which rises from the base of a pit-like
depression.

Gastropoda, (a) Prosobranehia. — In most cases there is only
one pair of salivary glands. These are usually lobed or branched
glandular masses, which lie, in the Diotocardia, at the sides of the
pharynx, in the Monotocardia, at the sides of the oesophagus. In the
former case, the ducts are short and do not pass through the cesophageal
ring formed by the nerve centres and their connectives and commis-
sures, which in these forms surrounds the anterior end of the pharynx.
In the Monotocardia, the ducts are long, and generally accompany the
oesophagus through the oesophageal ring (which lies behind the pharynx),
and open on the posterior lateral Avail of the latter.

Two pairs of salivary glands are found in certain ]_)inf,>ri/rdia (e.g.
Haliotis, Fissuretta), and further in Patella, the ScalarndoB, lanthinidce,
certain Purpuridce, Muricidce, and in the Cancellariidce.

One of the two pairs of glands in Haliotis is developed in the form
of large lateral glandular sacs covering the pharynx on the right and
left (Fig. 105, p. 121).

In the Amp\dlariida\ the ducts of the salivary glands do not pass

1 This differs smn -what from the description found in Simroth (Bronu's Klasten
und Ordnungen, vol. iii. pp. 183-185).

vii MOLLUSCA—THE ALIMENTARY CANAL 185

through the cesophageal ring, which here, as in the Diotocardia, sur-
rounds the anterior end of the pharynx.

Whereas the salivary glands are, as a rule, branched tubes or
acinose, they are sometimes (Scalariidce, lanthinidce, Cancellanidce)
simply tubular or (Doliidce, Xenophoridce, etc.) sac-like.

The passage of the ducts of the salivary glands through the
oesophageal ring in the Monotocardia may have come about by the
shifting back of the ring along the pharynx from its former position
in the Diotocardia, where it encircles the anterior end of the pharynx
in front of the apertures of these ducts. The salivary ducts would
thus necessarily become surrounded by the ring.

The ducts in the Monotocardia become the longer the further the
nerve ring shifts back from the mouth and pharynx. They are very long-
in animals provided with a protrusible proboscis, where the ring lies
far back on the oesophagus, behind the non-evaginable portion of the
proboscis. The ducts here run along the whole length of the latter.
But in those cases in which the oesophageal ring has shifted back more
quickly than the ducts have lengthened, the glands lie in front of the
ring. In the event of the subsequent lengthening of their ducts, the
glands might stretch back outside the ring. The arrangement of
the glands in the Toxoglossa and Eachiglossa would thus be explained ;
here the greater part of the glands lies behind the ring, although
the ducts are said not to pass through it.

The acid secretion of the salivary glands of certain Prosolrancliiu
(species of Dolium, Cassis, Cassidaria, Tritonium, Murex) and Opistho-
branehia (Pleurobranchus, PleurobrancMdiuni) contains 2'18-4'25 percent
of free sulphuric aeid. These carnivorous animals are able, by means
of their proboscides, to bore into other Molluscs and Ecliinoderms which
are protected by calcareous skeletons. The sulphuric acid in their
glands probably serves for transmuting the carbonate of lime into
sulphate of lime, which can then easily be worked through by the
radula.

(//) Pulmonata. — Two salivary glands (Fig. 157, 10) are always
found, their ducts entering the pharynx to the right and left of the
boundary between it and the oesophagus. The glands lie on the
oesophagus and the anterior part of the stomach in the shape of long,
lobate, jagged leaves. In some cases they are acinose or round and
compact.

(c) Opisthobranehia. — The salivary glands, of which only one pair
is almost always found, here vary in size and shape still more than in
the Pulmonata. These glands, which enter the pharynx, must not be
confounded with other glands which in many Opisthobranehia enter
the buccal cavity, and are sometimes more strongly developed than
the salivary glands.

Dentalium has no salivary glands opening into the pharynx, for
the glandular " cheek pouches " enter the buccal cavity, and two
diverticula which lie further back belong to the oesophagus.

186

COM PA EA TI VE A NA TOM Y

CHAP.

The Cephalopoda have a posterior and an anterior pair of salivary
glands. Were the fore-gut, which here rises vertically in the visceral
dome, to occupy the horizontal position it has in the Gastropoda, the
anterior pair would lie dorsally and the posterior ventrally with regard
to it. The two posterior glands (Fig. 127, 29, p. 147) are always
present (except in Cirrhotcuthis and Loligopsis, in which they are said to
be wanting), and lie on the oesophagus. Each gland has a duct, which
soon unites with that from the other gland, forming a terminal portion
which accompanies the oesophagus through the cephalic cartilage, and
opens above the radula into the pharyngeal cavity. The posterior

FIG. 157.— Alimentary canal of Helix, dissected out and seen from the right side (after
Howes). 1, 3, Tentacles ; -2, constrictor pharyngis ; 4, Irvator pharyngis ; 5, depressor ; 6, pro-
tractor pharyngis ; 7, pharyngeal bulb; 8, radular sheatli ; 9, columellar muscle, divided into a
retractor pedis and retractor pharyngis ; 10, salivary glands ; 11, digestive gland (liver) ; 12, ducts
of the same (gall ducts) partly cut open ; 13, hermaphrodite gland ; 14, stomach cut open, in its
base are seen the apertures of the gall ducts 15 ; 16, mid-gut ; 17, hind-gut ; 18, anus.

glands occasionally (e.g. in Oegcrpsidce) fuse behind the gullet, in which
case the duct is single throughout its whole length.

The anterior salivary glands are specially well developed in the
Octopoda (Fig. 127, 33, p. 147), and lie on the pharynx, into which
they empty their secretions by a duct, which seems always to be
unpaired. In the Decapoda the anterior glands are much smaller or
rudimentary ; they are generally represented by a single gland hidden
within the muscular wall of the pharynx.

Nautilus has no posterior salivary glands, but there are glandular
outgrowths of the pharyngeal cavity on each side of the tongue,
which perhaps correspond with the anterior salivary glands of other
Cephalopoda.

The Cephalopoda ( ? without exception) have an additional acinose

vir MOLLUSGA—THE ALIMENTARY CANAL 187

lingual gland, opening into that part of the pharyngeal cavity which
lies between the tongue and jaws.

The Lamellibranchia, as already mentioned, have neither pharynx,
jaws, tongue, nor salivary glands. In the Nuculidw, however, which
are rightly considered to be primitive forms, the mouth leads into a
widening of the intestine, on each side of which a glandular pouch
opens. These pouches perhaps correspond with the cesophageal
pouches of the Ckitouidce and Rhipidoglossa, which will be described
later.

Natica, which bores through the shells of living Lamellibmnchs and
feeds on their bodies, has a sucker-like organ on its proboscis (Fig. 98,
p. 107). The epithelium of the concave side of this organ, which is
applied to the shell attacked, forms a gland for secreting acid — prob-
ably sulphuric acid — which serves for dissolving the carbonate of
lime of the bivalve shell, which is then at once thrown out in the form
of powdered sulphate of lime.

C. The (Esophagus.

That portion of the intestine which lies between the pharynx (or
the mouth in Lamellibranchs) and the stomach is called the cesophagus,
the stomach being here used as the name of that widening of the

O

intestine into which the gland of the mid-gut opens. It is always
easy to detect the anterior boundary of the cesophagus. In Lamelli-
branchs it lies at the mouth, but in the Glossophora at the posterior and
upper end of the pharynx. The posterior boundary, however, can
often only arbitrarily be defined, as the cesophagus, which is usually
narrow and tubular, often widens very gradually into the stomach,
the structure of its walls at the same time gradually changing. In
other cases, widenings of the alimentary canal occur before the
stomach, and it is difficult to decide whether these are anterior
divisions of the stomach or posterior widenings of the cesophagus.

In Lamellibranchia, terrestrial Pulmonata, most Opisthobranchia, and
the Cephalopoda Decapoda the oesophagus is a simple ciliated tube
running to the stomach, being often provided with longitudinal folds,
and therefore extensible ; in other divisions, however, complications
occur, which are caused by glandular outgrowths or muscular en-
largements.

In a few Solcnogastres (e.g. Proneomenia], on the boundary between the short
cesophagus and the mid-gut, a more or less long blind diverticulum occurs ; this is
single, and runs forward dorsally to the pharynx, and may extend over the cerebral
ganglion to the end of the head.

In Chiton there are two lateral glandular sacs (sugar glands) connected with the
short cesophagus ; their inner glandular walls project into the lumen in the form of
villi, and their secretion changes boiled starch into sugar.

188

COMPARATIVE ANATOMY

CHAP. VII

Similar glands, which communicate with the anterior part of the oesophagus, are
found in the Khipidoglossa (e.g. Haliotis, FissureUa, Turbo). The glandular epithe-
lium in these also projects in the form of villi or folds into the lumen.

The so-called crop of the Docoglossa (Patella) no doubt corresponds with the two
lateral cesophageal sacs in the Chitonidcc and Rhi^ndoglossa. This is a saccular
widening of the oesophagus (Fig. 158, m), which, on account of the constitution of
its walls, has been compared with the psalterium of a Ruminant. A similar widen-
ing of the oesophagus is found in Ct/prceidce and ATaticidce, which must be counted
among the most primitive of the Monotocardia.

In those Monotocardia which are provided witli a proboscis, the length of the
thin oesophagus is in proportion to that of the proboscis.

The mouth lies at the tip of the proboscis, then follows a short and often rudi-
mentary pharynx, and then the long oesophagus, which runs through the whole
length of the non-protrusible portion of the proboscis, passes through the oosophageal

FIG. 158.— Median longitudinal section through Patella (after Ray Lankester). bn; Efferent
branchial vessel ; bra, afferent ditto ; asd, duct of salivary gland »? ; go, amis ; no, right nephridial
aperture; sJ, salivary gland; coi; heart; pc, pericardium; np, kidney; rf, intestine; hp, hepatic
gland (liver) ; v, blood vessel ; in (to the right), border of mantle covering the gills ; r, radular
sheath ; g, gonads ; in, crop ; ph, pharynx ; rtl, radula ; odm, masses of muscle and cartilage of the
lingual apparatus ; o, mouth ; fc, head or snout.

ring, and may be even further prolonged posteriorly. When the proboscis is re-
tracted, the posterior portion of the oesophagus becomes coiled ; when the proboscis
is extended, it lies in the protruded or evaginated basal portion.

Not infrequently in camivoTOUS Monotocardia there is a glandular widening in
that section of the oesophagus which follows the long proboscidal portion. The
oesophagus is most complicated in the fiachiglossa and many Toxoglossa, where this
widening, in the form of a large compact accessory gland, can become separated from
the intestine (Leiblein's gland, poison gland), and where other glands and widen -
ings may occur (Fig. 159). It seems probable that in certain Prosobranchia diges-
tion and resorption takes place even in the fore-gut.

In the Pulmonata and Opisthobranchia, there is sometimes a widening (crop, fore-
stomach) anteriorly to the stomach, and in the same waj' the short cesophagus of
the Scaphopoda has a glandular widening, or two lateral glandular diverticula.

Among the Cephalopoda, the Decapoda have a simple thin tubular cesophagus ;

.«

-L

— 1

FIG. 15".

FIG. 159.— Alimentary canal of Murex trun-
culus (after Bela Haller). 1, Pharynx ; 2, ducts of
the salivary glands (5) ; 3, oesophagus ; 4, 6, and 7,
glands of the fore-gut (8) ; 9, digestive, gland (liver) ;
10, stomach ; 11, hind-gut; 12, gland of the hind-
gut ; 13, anus.

II- -

9---1

Fio. 161.

FIG. 100.— Alimentary canal of Sepia. 1, Jaw ;
2, pharynx ; 3, posterior buccal ganglion ; 4, duct
of the salivary gland (5) ; 6, digestive gland (liver) ;
7, anus; 8, rectum ; 9, efferent duct of the pigment gland (ink-bag), 10; 11, .stomachal ccecum

12,

stomach ; 13, ganglion gastricum ; 14, " pancreatic appendages " of the gall ducts of the digestive gland.

FIG. 1(31.— Sketch of the anatomy of Limacina helicina, from the right side, after removal of the
mantle, heart, and kidney (after Pelseneer). 1, Fin (parapodium) ; 2, foot ; 3, central nervous system
(oesophageal ring) ; 4, oesophagus ; 5, anus ; 6, columellar muscle ; 7, duct of the hermaphrodite gland.
7a ; 8, intestine ; 9 and 10, dental plates of the stomach ; 11, accessory glands of the genital apparatus ;
12, mantle cavity ; 13, seminal groove or furrow.

190

COMPARATIVE ANATOMY

CHAP.

the cesophagiis of the Octopoda, however, is provided with a lateral pouch, the crop

(Fig. 127, p. 147), whose walls are not glandular. This may
serve as a reservoir of food when the stomach is already
full. In Nautilus, the crop is a very large saccular widening
of the oesophagus, larger than the stomach itself.

/.— -18

5—

I-"-

8--

—13

—-10

FIG. 102.— Diagram of the
anatomy of Clio striata,

from the right side ; the
heart, kidney, and mantle of
this side removed (after
Pelseneer). 1, Fin(parapo-
dium) ; 2, aperture of the
penis ; 3, right tentacle ; 4,
genital aperture ; 5, penis ;
6, oesophagus ; V, duntal
plates of the stomach ; 8,
ducts of the gonad ; 9,gonad;
10, intestine ; 11, digestive
gland ; 12, ducts of the same
(cut off) ; 13, accessory
glands of the genital appar-
atus ; 14, mantle cavity ; 15,
terminal portion of the geni-
tal ducts ; 16, central ner-
vous system (ganglion ring);
17, foot ; IS, pharynx.

D. The Mid-gut with the Stomach and
Digestive Gland (Liver).

The oesophagus leads into a wider portion of
the alimentary canal, the stomach. Into this the
ducts of a gland open ; this gland is strongly
developed in nearly all Molluscs, and is usually
called the liver, but may be more appropriately
named the digestive gland, since it in no way
fulfils the functions of the vertebrate liver. As
far as is at present known, it functions rather as
a pancreas, or it combines the functions of the
various digestive glands of the vertebrate intes-
tine, no such thorough division of labour as is
found in the Vertebrates having taken place. The
digestive gland is, in most cases, a richly-branched
tubular or acinose gland, which to the naked eye
appears a compact lobate body of a brown,
brownish-yellow, or reddish colour. Its glandular
epithelium consists of three sorts of cells—
hepatic, ferment, and calcareous cells. In
many Nudibmnchia the gland breaks up into
branching intestinal diverticula, which spread
through the body almost like the gastro-canals or
intestinal branches in the Turbellaria, and run up
into the dorsal appendages of the body (clado-
hepatic Nudibranchia).

Chostoderma, among the Solenogastres, has a
simple midgut diverticulum, which may corre-
spond morphologically with the digestive gland
of other Molluscs ; but in Proneomenia, Neomenia,
etc., the straight mid-gut is provided throughout
its whole length with narrow lateral glandular
sacs arranged closely one behind the other at
right angles to it.

A part of the mid - gut gland (the part
nearest to the point Avhere the duct leaves it)
and the glandular epithelium of the duct may be
specially differentiated in Cephalopoda, and may,
finally, form a distinct system of glands called the
pancreas (Fig. 160).

VII

MOLLUSGA—THE ALIMENTARY CANAL

191

The stomach is not infrequently a lateral outgrowth of the niesen-
teric wall, so that the aperture (cardia) leading into it from the oeso-
phagus and that leading out of it into the small intestine (pylorus)
are more or less near one another. A sort of connection between
these apertures may arise, a ciliated furrow or channel bounded by
longitudinal folds running between them, and in some cases continued
into the adjoining sections of the alimentary canal.

In the Cephalopoda, the duct of the digestive gland (the so-called
hepatic or gall duct) does not open direct into the stomach, but into a
coecal outgrowth of the stomach, the spiral eceeum.

In very many Lamellilmnchia there is a diverticulum of the
stomach which contains within its lumen a rod-shaped gelatinous cuti-
cular formation, called the crystalline stylet. Similar structures occur
in the Prosobranckia, and especially in the Rhipidoglossa and Toxoglossa.

In many Opisthobrancliia, the inner wall of the stomach carries
variously-arranged cuticular teeth, dental plates, jaw plates, etc., which
serve for triturating the food. In such cases the muscular wall of the
stomach is strongly developed.

The stomach is succeeded by a narrower tubular section of the
mid-gut, called the small intestine (intestinum), which usually forms
coils or loops ; these are more numerous in herbivorous or detri-
tivorous than in carnivorous Molluscs.

The stomach, small intestine, and digestive gland, together with
part of the sexual organs, compose the whole or by far the largest
portion of the visceral dome, where this is present.

1. Amphineura.

The Chitonidne show the typical division of the mid-gut into stomach, digestive
gland, and small intestine. The stomach
lies far forward, and has a wide outgrowth
on one side, which is, functionally, a reser-
voir of secreted matter. The cardia and
the pylorus lie near one another. The
digestive gland is paired ; the larger liver
to the right has four apertures, while the
smaller one to the left has only one prin-
cipal aperture into the stomach. The
small intestine is more than four times as
long as the body, and it forms many loops
Avhich are constant in their arrangement.
Chiton feeds on small or even microscopic
alg?e. FIG. 1(33.— Part of a horizontal median

Unlike the Chitoiiidce, the Solcnogastrcs section through Proneomenia Sluiteri.
Show no separation of the mid -gut into Sefa of the first second third and fourth

r\vr\ QT* o v>a 00011 Tivr»ioririnrr i r-<M»i T ha TMrt-Vi^ o »i rt

stomach and small intestine. The mid-gut

order are seen projecting from the right and
left into the lumen of the mid-gut. In the

runs straight through the body, the greater background is the dorsal wall of the gut, with
part of which it tills. The glandular lateral the groove which cuts into the hermaphrodite
coeca found in Ncomenia, Proneomenia, SlaiKl (c/- 1 '• 53i P- 42)-
etc., and called hepatic diverticula, are caused by the projection into the lumen from

192

COMPARATIVE ANATOMY

CHAP.

3--

each side of narrow septa arranged at right angles to the gut, or transversely (Fig. 163) ;
in these septa, muscle fibres run down to the rudimentary foot, and blood lacuna?
abound. In Proneomcnia Sluitcri, septa of the first, second, third, or fourth order can
be distinguished, as seen in the figure. The septa on the right alternate with those
on the left side of the body. In the dorsal middle line the mid-gut forms a narrow
ciliated longitudinal groove which cuts deep into the gonad, cilia are also found on its
medio-ventral surface.

2. Gastropoda.

The digestive gland of the Gastropoda falls into two or more lobes, between
which the stomach and the coils of the small intestine lie embedded. One, two, or

more ducts of the gland may open into the
stomach. The walls of the digestive gland
show the same division into layers as the wall
of the alimentary canal. For details as to the
ferment, hepatic, and calcareous cells forming
the epithelium of the gland, and their physio-
logical constitution, the reader must be re-
ferred to special histological and physiological
treatises.

In the Nudibranchia, as already mentioned,
the digestive gland breaks up into a system of
glandular diverticula (the so-called "diffuse
liver"). The Acolidiadce (e.g. Tcrgipcs) afford
an instructive instance of this. Three diver-
ticula rise from the stomach, two anterior and
lateral, and one posterior and unpaired. These
ramify in the body cavity, and finally send up
their last ramifications or lobes into the dorsal
appendages. The contents of the intestine can
penetrate into these last ramifications of the
" diffuse liver " (Fig. 164).

Further, within the Nudibranchia the break-
ing up of the compact digestive gland to form a
"diffuse liver," i.e. the loosening from one
another, and the spreading out of the glandular tubes which are in close contact in
the compact gland, can be followed almost step by step. In the Tritonidce the gland
is a great compact mass. In other families, such as the Tethymelibidcc, Lomanotida;
Dendronotidic, Bornellidcc, Scyttccidcc, it divides into two anterior accessory livers
and a posterior principal liver, from which diverticula run up into the dorsal append-
ages. Finally, the accessory and principal livers break up into separate "hepatic
branches" (Aeolidce), which in some cases anastomose. The posterior principal
branch of the " diffuse liver" gives off specially numerous lateral branches ; it often
widens out to a pouch, and may then be compared to an extended gall bag, or a
posterior diverticulum of the stomach. In Phyllirho'e, a pelagic form, without
dorsal appendages, the "diffuse liver" is simplified, consisting of four unbranched
blind tubes, the two anterior opening into the stomach separately, the two posterior
entering it together (Fig. 19, p. 12).

The stomach of many Opisthobranchia consists of two divisions separated by a con-
striction. In some forms, such as the Bull idee among the Tcctibmnchia, the Ptero-
poda thecosomata, and the Tdlujmclibida', Bornellidcc, Scylhvidtt.', among the Nudi-
branchia, it is armed with hard chitinous plates, spines, teeth, etc., occurring in
varying number and arrangement on its inner wall (Figs. 161 and 162)

FIG. 1(34.— Alimentary canal of Aeolis
(after Souleyet). 1, Pharynx ; 2, stomach ;
3, branched digestive gland (liver); 4,
anus ; 5, rectum.

VII

MOLLUSCA—THE ALIMENTARY CANAL

193

3. Scaphopoda.

The mid-gut of Dcntalium (Fig. 165) consists of a looped stomachal tube bent

FIG. 165.— Alimentary canal, kidney, and sexual organs of Dentalium, from behind (after
Lacaze-Duthlers and Leuckart combined), a, Mouth ; 6, leaf-like oral tentacles ; c, snout ; d,
entrance to pharynx ; e, pharynx with radula, /; g, hind-gut ; h, right kidney ; i, anus ; k, right
nephridial aperture ; I and q, ducts of the digestive gland, n : m and o, gonad ; n and p, digestive
gland (liver) ; r, left nephridial aperture ; s, left kidney ; t, stomach ; u, pharynx ; v, lobes or sails
on which the filamentous tentacles are placed.

back on itself, and of a small intestine lying in a tangled coil behind the oesophagus.
VOL. II O

194 COMPARATIVE ANATOMY CHAP.

Two digestive glands, lying in the upper part of the body, open through wide
apertures into the stomach. Their form can be gathered from Fig. 165.

4. Lamellibranchia.

In the Lamellibranchia the oesophagus, which lies under the anterior adductor,
widens at the anterior base of the foot to form the stomach. This descends some-
what into the foot. At the posterior base of the stomach lie two apertures ; one of
these is the pylorus, and leads into the small intestine which runs more or less
coiled within the base of the foot ; the other leads into a tubular diverticulum, the
sheath of the crystalline stylet. The large richly-branched acinose digestive gland
(liver) opens through several apertures into the stomach, with which it lies in
the anterior part of the pedal cavity. In PJwlas, Jouannetia, and Teredo, the
stomach has another ccecum besides the sheath of the crystalline stylet. In all
bivalves there is, on the inner wall of the stomach, a gelatinous cuticular structivre
(dreizackiger Korper, fleche tricuspide), which varies in thickness, and is continued
into the gelatinous crystalline stylet. This latter is secreted in concentric layers as
a cuticular structure by the epithelium of the sac in which it lies. A plausible
suggestion has recently been made as to the use of these gelatinous structures, viz.
that they serve for surrounding with a slimy envelope foreign particles, such as
sharply-pointed grains of sand, which enter the alimentary canal with the food ; in-
jury to the delicate walls of the intestine is thus avoided, and the travelling of such
particles along the digestive tract is facilitated. The point of the crystalline stylet
projects freely into the lumen of the intestine. In some forms it does not lie in a
separate sac, but in a groove (Najada, Cardium, Mytilus, Pecten, etc.). The tricuspid
body and the crystalline stylet appear temporarily, and are renewed periodically.
Similar structures have been observed in the stomachs of various Gastropods.
Haliotis has a stomachal ccecum which can be compared with the sheath of the
crystalline stylet.

In the lower Lamellibranchs, the Nuculidce and Solenomyidcc, the stylet is either
very slightly developed or wanting. In the Arcidm also, it is only slightly
developed.

The Scptibranchia (Poromya, Cuspidaria) are distinguished from all other
Lamellibranchia by the absence of coils, and the consequent shortening of the small
intestine (ef. on the intestine of the Lamellibranchia, Figs. 24, 25, 26, 27, 28,
pp. 16, 17/18, 19).

5. Cephalopoda.

The stomach in the Cephalopoda always lies in the dorsal portion of the
visceral dome in the shape of a sac with a strong muscular wall. It always has a
ccecal appendage (stomachal or spiral ccecum, Figs. 166, 160), which varies in shape
and size ; into this the digestive gland (liver) opens. This ccecum is a reservoir for
the secretion of the digestive gland. Food never enters it, there are even valves
at the point of entrance into the stomach, which allow the secretion collected in
the ccecum to pass into the stomach, but prevent the entrance of the contents of the
latter into the ccecum.

In Nautilus, the ccecum does not open into the stomach, but into the commence-
ment of the small intestine, and is in the form of a small round vesicle with lamella*
projecting into its lumen. In Sepia and Scpiola also, the ccecum is more or less
round ; in Rossia, it is slightly developed ; in Loligo and Sepiotevthis, very long and
ending in a point ; in all Ocgopsidce and Octopoda, more or less spirally coiled at the
blind end.

The well-developed digestive gland seems to arise as a paired organ, even when

VII

MOLLUSCA—THE ALIMENTARY CANAL

195

unpaired in the adult. The whole of the much branched gland is surrounded by a
common integument, and it thus outwardly appears to be compact.

The digestive gland of Nautilus consists of five lobes (four paired and one
unpaired), which lie around the crop. They have two ducts, which enter the ccecum
through a short common terminal portion.

In the Dibranchia, also, the digestive gland always lies on the ventral side of the
stomach, close to the ascending oesophagus.
It is undivided, and round or oviform in the
Uctopoda, Oegopsidcc, and Sepiola. In Loli/jo
and Sepioteuthis, it is traversed by the oeso-
phagus and the aorta ; in Enoplotcuihis, its
dorsal half is cut into two points by these
organs ; and the same is the case in fiossia.
In Sepia and SpiruJa, the gland forms two
lateral lobes which are distinct in Sepia, but
connected along the middle line in Spirula.

There are always two ducts (gall ducts)
which rise near the median plane from the
upper part of the gland, and open into the
stomachal ccecum separately or through a
common terminal portion.

The following facts have been ascertained
as to the function of the so-called pancreas
of the Cephalopoda. It is originally a
specially differentiated portion of the diges-
tive gland, and is easily distinguishable in
the Odopoda by its different colour ; it lies
near that part of the gland from which the
ducts spring. In Loligo, the pancreas is
found in the much thickened wall of the
ducts themselves. In this case it consists
of numerous glandular anastomosing out-
growths of the epithelium of the ducts into
their wall. In other Dccapoda, these gland-
ular outgrowths pass from the wall of the
ducts into the surrounding body cavity, and
then each duct appears throughout its whole
length to be covered with acinose or ramified
' ' pancreatic appendages. " The pancreatic
secretion contains diastase, and appears to
carry out only one part of the functions of the
digestive gland, viz. that part which corre-
sponds with the digestive functions of the salivary glands in the higher Vertebrates.

The small intestine, in which among all Molluscs the resorption of the digested
food chiefly (if not exclusively) takes place, is short in the carnivorous Cephalopoda,
and forms several coils only in Trcmoctopv.s molaceus.

FIG. iiio. — Alimentary canal of Loligo
sagittata (without pharynx and salivary
glands) partly cut open (after Gegenbauer).
1, (Esophagus ; 2, probe, inserted into th«
pylorus ; 3, stomach ; 4, stomachal ccecum
with spiral ccecum 5 ; 0, hind-gut ; S, ink-
bag ; 7, aperture of the same into the hind-
gut.

E. Hind-gut (Rectum).

This is generally short in Molluscs. Where it is sharply marked
oft' from the small intestine, it usually dift'ers from the latter in being
thicker and more muscular.

196 COMPARATIVE ANATOMY CHAP.

In the majority of Lamellibranchs, and in nearly all Diotocardia, the
rectum traverses the ventricle ; this fact, with many others, supports
the relationship of these two groups.

In certain Molluscs, viz. the Scaphopoda, a few Prosobranchia
(Muriddce, Pwymridce), and the Cephalopoda, the hind -gut has an
accessory (anal) gland, which is well known in the Cephalopoda as the
ink-bag1.

The rectal gland in Dentalium is a branched acinose gland opening into the hind-
gut, according to one account through six separate ducts, and according to another
through one single duct. Eggs and spermatozoa have been met with in the lumen
of this gland, and it has been supposed that they have been accidentally drawn out
of the mantle cavity by the swallowing-like action of the hind-gut, which has been
observed in Dentalium.

• The anal gland found in some Rachiglossa (Monoceros, Purpura, Murex) is always
dark in colour (brown or violet), and is either tubular with many bulgings of its
wall, or acinose with an axial duct. It always enters the hind -gut near the
anus.

A gland has been found near the rectum in the Pteropoda thecosomata (Clio,
Cavolinia) and the Bulloida, and has been described as an anal gland, but this
requires further investigation.

The ink-bag1 of the Cephalopoda (Fig. 167), which is wanting only
in Nautilus, is a much developed anal gland. It enters the hind-gut
near the anus. The ink or sepia pigment secreted by it consists of
extremely minute particles which are ejected with vehemence from
the bag and discharged through the funnel. The pigment quickly
mixes with the water, and envelops the animal in a pigment cloud,
thus screening it from its enemies.

Form and position of the ink-bag (cf. Figs. 160, p. 189 ; 177, p. 213 ; 178,
p. 214). — The typical position of the ink-bag is in front of the rectum, i.e. in the
loop formed by the intestine in ascending from the mouth and then descending to
the anus. In Spinda, Enoplotcvtliis, and Scpioteuthis, the ink-bag is very small ;
it progressively increases in size in series both of Decapoda and of Octopoda, its
division into a saccular portion and a duct opening into the hind-gut in front of the
anus becoming more and more distinct. In the Octopoda, it lies embedded in the
upper part of the liver within the muscular hepatic capsule (cf. p. 128). It is still
found in this position (between the liver and the rectum) in Sepiola. In other
Decapoda, however, the ink-bag is found shifting higher and higher in the visceral
dome, its duct at the same time increasing in length. Finally, in Sepia (and the
fossil Dibranchia), it is found at the top of the visceral dome, behind the gonad. Its
duct runs along the right side of the hind-gut, bending round somewhat before
reaching the anal section of the rectum so as to enter the latter anteriorly. Onto-
genetically, however, even in Sepia, the ink-bag arises as an anterior outgrowth of
the rectum.

Structure of the ink-bag in Sepia (Fig. 167 A). — The ink-bag in this instance
consists of three parts: (1) the pigment gland which secretes the "ink" ; (2) the
pigment reservoir and the duct, which forms (3) an ampulla with a glandular wall near
its aperture. The pigment gland is a sac at the base of the ink-bag on its anterior
wall (that turned towards the gonad). It projects into the cavity of the ink-bag,

VII

MOLLUSCA—THE ALIMENTARY CANAL

197

which serves as reservoir and duct for the pigment. The latter, after being formed
in the gland, passes through an
aperture in its wall into this
reservoir. The cavity of the
gland is traversed by numerous
perforated and richly vascular -
ised lamellae of connective
tissue, which are inter - con-
nected in such a way as to form
a kind of sponge-like structure.
New lamella? are continually
being put forth by the formative
zone of the gland, which is a
narrowed portion bent back
downwards, while the oldest
lamellae, which lie nearest the
aperture of the gland, become
detached and degenerate. All
the lamellae are covered by a
glandular epithelium and the
formation of the pigment can
be traced in all its stages from
its appearance in the epithelial
cells of the formative zone to
its condition in those of the
oldest lamella1. In the forma-
tive zone, the young glandular
cells are at first colourless. In
the succeeding lamella?, how-
ever, pigment granules increase
in number and from the older
lamellae are emptied into the
cavity of the gland, the epi- FIG. 167.— Morphology of the pigment gland (ink-bag) of
thelial cells then becoming the Cephalopoda (after P. Girod). A, Median longitudinal
, , , 111" section through the ink-bag of an adult. «, Anus ; 1, terminal

i anrl breaking up. portion common to the rectum (2) and the duct of the ink-

Both the gland and the reser- bag ; 3, ampulla ; 4 and 5, sphincter muscles of the ampulla ;
voir are surrounded by a vascul- 6, duct of the ink-bag ; 7, pigment reservoir ; S, opening of
arised integument of connective the I'igment 8land into the reservoir ; 9, portion of the gland
, . . , traversed by lamella- ; 10, formative zone of the lamella',

tissue ; the same integument B_G VariQUS stageg in the development of the pigment

forms the framework of con- giand ; B, anal papilla ; C, invagination in the same ; D, ap-
nective tissue running through pearance of two new depressions at the base of C ; these
the lamellae or trabecula? within inc^ase in depth, the one becoming the pigment gland ft, the
the crl 1 other the rectum 2. In F, the formative zone has appeared

in the gland, in G, the first lamell* and the duct. H, I, K,

. . . ,

Ine ink-bag is further envel-

,.hanges ju the relative positions of the rectum and gland in

oped as a whole in a tough integu- the course of development, seen from the posterior (mantle)
ment consisting of three layers : si(ie- In H. the rectum lies behind the ink-bag. In I, the

(1) an inner "littering silvery latter has shifted' and in K lies behilld the rectui" (°" the

, J mantle side).
layer (argentea), similar to the

corresponding layer in the outer integument ; (2) a central muscle layer (inner
longitudinal and outer circular muscles ; and (3) an external layer of connective
tissue.

The terminal ampulla has, at its two narrow ends, folds projecting inward and
functioning as valves ; it can be closed at these parts by sphincter muscles. The

198 COMPARATIVE ANATOMY CHAP.

ampulla itself also forms longitudinal folds on its inner surface, between which
glandular tubes open.

The anus, in the Cephalopoda, always carries two lateral projecting appendages,
which are often lancet-shaped.

The short and narrow hind-gut of the Solenogastres opens into the
dorsal portion of a cavity, the cloaca, which lies at the posterior end
of the body ; this, again, communicates with the exterior by means of a
ventral and very extensible longitudinal slit. Into this cloaca the
ducts of the genital organs, which are morphologically to be regarded
as nephridia, also open.

In the Lamellibranchia, after the hind-gut has traversed the heart,
it runs straight backward over the posterior adductor, to open through
the anus into the posterior and upper portion of the mantle cavity
(anal chamber).

On the position of the anus, cf. Section V. on the arrangement of
the organs in the mantle cavity.

XVII. The Circulatory System.
A. General.

All Mollusca have a circulatory system ; in some divisions,
especially in the Cephalopoda and some Prosobranchia, this may attain a
high level of complication by the development of a closed arterial and
venous vascular system. The heart, as the central organ of propulsion,
is never wanting. It lies enclosed in the pericardium, a division of
the secondary body cavity ; its primitive position is median, above
the hind-gut. In the Lamellibranchia and Diotocardia, it is traversed
by the hind-gut, in other Gastropoda it lies near it. It is always
arterial, i.e. it pumps the blood flowing from the respiratory organs
back into the body.

In those symmetrical Molluscs in which the dorsal portion of the
body rises as a high visceral dome, the intestine first ascending into
the dome and then descending to the anus, the heart comes to lie
behind the hind-gut (Dentaliwrn, Cephalopoda).

In asymmetrical Gastropoda, its position depends upon that of the
pallial complex. Where the hind-gut and anus have shifted with the
pallial organs to the anterior side of the visceral dome, the heart also
lies anteriorly (Prosobranchia, Pulmonata, a few Tectibranchia).

The heart gives rise, as a rule, to two large arteries (aorta), one
of which runs to the head, the other to the visceral dome, to supply
blood to the viscera. Not infrequently they leave the heart as one
large vessel. Where the circulatory system is not closed, the arteries
sooner or later convey the blood to the primary body cavity or coalom,
i.e. into the lacunar system. The venous blood is sometimes conveyed
along distinct vessels, sometimes in channels without proper walls into

VII

MOLLUSGA—THE CIRCULATOR V SYSTEM

199

the gills, where it becomes arterial and flows back through the auricles
(atria) into the heart.

There is, typically, one pair of auricles, one on each side of the
ventricle. This is the case in all Molluscs provided with two sym-
metrical gills. The arterial blood flows out of the left gill into the left
auricle and thence into the ventricle, and out of the right gill into the
right auricle and thence into the ventricle (Diotocardia, Zeugobranchia,
Lamettibranchia, Cephalopoda Dibranchia}. Again, where a longitudinal

FIG. 16S. — A-H, Diagrams illustrating the relation between the ctenidia, the heart, and
the aorta. A, Chiton; B, Lamellibranchia ; C, Dibranchiate Cephalopoda; D, Tetra-
branchiate Cephalopoda ; E, Prosobranchia Diotocardia Zeugobranchia ; F, Prosobranchia
Diotocardia Azygobranchia ; G, Prosobranchia Monotocardia ; H, Opisthobranchia Tecti-
branchia. 1, Ventricle ; -2, 3, 2c', ~2b, 3n, 3fr, auricles ; 4, vena bran chialis= efferent branchial vessel ;
5, aorta ; 5«, aorta cephalica ; 57), aorta visceralis ; 6, aorta posterior vel superior ; 7, ctenkla.

row of numerous gills is found on each side in the mantle furrow
(Chitonidce), the heart lies posteriorly above the hind-gut, and has one
auricle on each side of the ventricle. This fact appears quite as much
to support the view that one pair of gills and one pair of auricles were
present in primitive Molluscs, as does the arrangement in Nautilus
(Cephalopoda Tetrabranchia) the other view, that there were two pairs
of gills and also two pairs of auricles.

In the majority of Gastropoda, where one of the two original gills
has disappeared, the auricle belonging to it has usually also disappeared.

200 COMPARATIVE ANATOMY CHAP.

The original right gill and right auricle are usually retained in Gastro-
pods with shells dextrally twisted. In Gastropods with a true sinistrally
twisted shell, the left gill and left auricle are retained.

There is, however, a whole division of the Prosobranchia, the
Diotocardia, in which both auricles are retained. It is evident that the
gills are more liable to disappear than the auricles, since in some
groups both auricles remain when one gill has disappeared (for
details see opposite page).

When, in Gastropoda with only one auricle, the pallial complex has
shifted to the anterior side of the visceral dome, the respiratory organs
lie in front of the heart, and the single auricle in front of the ventricle
(Prosobranchia, Monotocardia, most Pulmonata, a few Opisthobmnchid).
In those Gastropoda, however, in which the pallial complex lies on one
(usually the right) side of the body, the gill is placed behind the heart
and the auricle behind the ventricle. This is the case in nearly all
the Opisthobranchia. In a few Pulmonates also, such as Testacella,
Oncidium, etc., the auricle lies behind the ventricle, as a consequence
of special organic modifications.

The blood, or rather the haemolymph, is a fluid rich in dissolved
albumen (hsemocyanine), which assists in nourishing the body and in
respiration. Amceboid cells, the lymph cells or amcebocytes, are
suspended in the hsemolymph. Haemoglobin is occasionally found
dissolved in the hsemolymph or combined with special blood corpuscles.
The lymph cells either become detached from the wall of localised
blood-making glands, which may vary in position, or, in a more
diffused manner, from large vascular areas. They seem, from their
origin, to be cells of connective tissue.

The walls of the heart and of the walled vessels consist of smooth
muscle fibres thickly felted, and (on the heart) of an external endo-
thelium which belongs to the pericardium. An inner endothelium is
wanting, so that the muscle fibres are directly bathed by the blood.

The wall of the ventricle is always more muscular than those of
the auricles. At the point where the auricles open into the ventricle,
valves projecting into the lumen are always found, which, when the
latter contracts, prevent the return of blood into the auricle. Besides
these atrio-ventricular valves, there are occasionally other valves
between the ventricle and the aorta. Valves may also occur in the
peripheral blood channels, when these form contractile enlargements
(e.g. the valve between the branchial heart and the afferent branchial
vessels of the Cephalopoda).

In various Gastropods and in Chiton a network of ganglion cells
and nerve fibres has been found in the wall of the heart, innervated
by two nerves of different origin. The nerve which runs to the
ventricular plexus originates, in the Prosobranchia, in the left parietal
ganglion, that running to the auricle from the left parieto-visceral
connective. Where there are two auricles, they are innervated from
the branchial ganglia.

vii MOLLUSCA—THE CIRCULATORY SYSTEM 201

B. Special.

1. Amphineura.

a. Chitonidse (Polyplacophora). — The heart is symmetrical, with two lateral
auricles.

The ventricle and the two auricles are long tubes. The auricles are in open
communication with the ventricle about the middle of their length. Besides this,
the two auricles pass into one another posteriorly, the posterior end of the ventricle
also opening into them at this point.

The ventricle lies against the dorsal wall of the pericardium, to which it is
attached by a median band of endothelium. The ventricle passes into an aorta
which allows the blood to flow into the cceloru through apertures in its wall. With
the exception of the pedal arteries, the rest of the circulatory system is lacunar ;
there are no vessels with walls of their own.

The venous blood is collected from the lacunar system of the body (primary
ccelom) into longitudinal channels which run on each side under the pleurovisceral
cords. From these channels it flows into the gills, where it becomes arterial, and
returns through other longitudinal channels which run above the pleurovisceral
cords. Two transverse channels in the region of the heart (cf. Fig. 51, p. 40)
convey the arterial blood into the auricles.

The two pedal arteries lie laterally and ventrally with regard to the pedal cords ;
they probably draw their blood from the aorta and pass it on to the lacunar system
of the foot.

b. Solenogastres. — The heart lies above the hind-gut on the dorsal side of the
pericardium. It does not lie freely in the latter, nor is it suspended by an
endothelial band, but simply projects into the pericardium from above, so that only
its under surface is covered by the pericardial endothelium. The presence of two
auricles has not been proved. The rest of the circulatory system is purely lacunar.
Specially large blood channels lie in the depths of the principal septa which project
into the mid-gut, and bulge these out. Large blood sacs are also occasionally found
in folds which project into the pharyngeal cavity from its wall, and there are more
or less large sinuses in the folds, which, in Ncomcnia and Gha'todcrma, project
into the cloaca and may be regarded as gills. In all these parts the intestinal
epithelium separating the sinus from the intestine is ciliated, and respiration no
doubt takes place.

2. Gastropoda.

Relation of the auricles to the ventricle. — The lowest Gastropods, i.e. the
Diotocardia among the Prosobranchia, have a heart with two auricles. This is not
only the case in the Zeugdbrancliia (Fissurclla, Haliotis, etc.), which have two gills,
but also in the AzygobrancMa (Turbinidce, Trochidce, Neritidcc), in which only the
left (originally the right) gill has been retained. No branchial vein then enters the
smaller (rudimentary) auricle on the right, the veins having atrophied with the gill.
In the Zeugobranchia, the long ventricle lies in a line with the hind-gut, which runs
length-wise through it. In the Azygolrancliia, the ventricle lies transversely with
respect to the hind-gut which runs through it, the left auricle lying in front of the
ventricle, and the right auricle behind it. The left branchial vein enters the
anterior (left) auricle. If we suppose the posterior (right) auricle to have disappeared
altogether, as is the case in all other Gastropoda, the heart consists of a ventricle
and one auricle lying in front of it, which receives the branchial or pulmonary vein
from the gill or lung in front of it.

202 COMPARATIVE ANATOMY CHAP.

This serial order of the ventricle, auricle, branchial or pulmonary vein and
respiratory organs is characteristic of the Azygobrancliia, Mbnotocardia, and most
Puhnonata.

The Docoglossa (Patella and allied forms) have only one auricle ; the ventricle
in Patella (not in Acmaxt), however, is divided into two parts.

Among the Monotocardia, only Cyprcea (as far as is at present known) has a
rudimentary right auricle, closed on all sides except at its aperture into the
ventricle.

Among the Pulmonata there are forms in which the auricle lies behind the
ventricle. This must be regarded as a secondarily acquired position, determined by
the shifting back of the anus and the mantle cavity to the posterior end of the body
( Testacella, Oncidium ). In Daudcbardia, the auricle still lies in front of the ventricle ;
nevertheless this genus, like several other shell-less Pulmoiiates, is opisthopneumonic,
i.e. its respiratory network lies chiefly behind the heart. In Testacella, the auricle
also lies behind the heart (cf. p. 77).

In the OpistfwbrancJtia, the auricle lies behind the ventricle ; this is connected
with the position of the gill at the posterior end of the body, or where no true
ctenidium is found, but where respiration takes place by means of anal gills, or
dorsal appendages, or through the integument, with the point of entrance of the
branchial vein into the heart from behind.

In a few Tectibranchia, e.g. Aetmrn, Accra, Gastropteron, the gill lies some-
what far forward, and the auricle is then placed laterally, to the right of the
ventricle rather than behind it.

It is of great importance, with regard to the position of these
organs in the Lamellibranchia, to note the fact that, in many Diotocardia
(e.g. Fissurella, Haliotes, TurUmdce, Trochidcc, Ncritidce, Neritopsidce, etc.)
the ventricle is traversed by the hind-gut, while in all other Gastropods
the intestine merely runs past it.

Circulation, (a) Prosobranchia.— A large vessel, the aorta, springs from tin-
ventricle. This soon divides into two branches: (1) the anterior or cephalic aorta
(A. cephalica), and (2) the posterior aorta (A. visceralis).

The anterior aorta conveys blood to the anterior part of the body (head, pharynx,
proboscis, cesophagus, stomach, copulatory organs) and to the mantle, and gives oft
among others the important arteria pedalis ; this latter soon breaks up into
separate arteries, which run longitudinally through the foot. In some cases the
cephalic aorta is richly branched, breaking up into numerous fine vessels which
spread out in and on the above-mentioned organs ; in others, the arteries, without
branching, open into arterial sinuses. Among these, the large cephalic sinus into
which the anterior aorta opens (e.g. in Haliotis) deserves special mention. Where
the cephalic aorta runs beyond the cesophageal ring formed by the central ganglia
and their commissures, it passes through this ring.

The aorta visceralis supplies the organs which lie in the visceral dome, especially
the digestive gland, the genital glands, and the mid-gut.

The venous blood collects in the lacunar spaces of all parts of the body, and
Hows into a large venous sinus, i.e. into the space in which the stomach, salivary
glands, intestine, digestive -'land, and genital organs lie. This space or primary
body cavity is somewhat spacious round the .stomach, but very limited in the
visceral dome, where the lobes of the digestive gland, the walls of the intestine,
and the genital glands with their accessory parts are so crowded together as to leave
very narrow spaces between them.

VII

MOLLUSGA—THE CIRCULATORY SYSTEM

203

three

The blood passes out of the large venous sinus back into the heart
channels.

1. A large part of it flows through lacuna? or vessels into the paired or unpaired
branchial artery (afferent branchial vessel). In the course of branchial respiration
the blood becomes arterial, and collects in an efferent branchial vessel (cf. section on
the respiratory organs, p. 84), which, as branchial vein, conducts it to the auricle
of the heart. Where there are two gills, there are naturally two branchial arteries
and two branchial veins, the latter conducting the arterial blood to the two auricles.

2. Another part of the venous blood flows through the kidney, then again
collects in lacuna; or vessels which lead to the gills, and finally reaches the heart

FIG. 169.— Circulatory system of Paludina vivipara (after Leydig). The animal is seen from
the left side. 1, Eye; 2, cerebral ganglion; 3, efferent branchial vessel (branchial vein); 4, gill
(ctenidium) ; 5, afferent branchial vessel ; li, kidney; 7, aorta visceralis, winding up close to the
colnmella ; S, ventricle; 9, auricle; 10, aorta cephaliea ; 11, venous sinuses of the body; 12,
auditory vesicle ; 13, pedal ganglion.

through the branchial veins. Less frequently, the venous blood, after passing
through the kidney, enters the auricle more or less directly, i.e. without passing
through the gills, and there mixes with the arterial blood coming from the gills.

3. A certain part of the venous blood, passing by both the kidney and the gill,
flows direct into the branchial veins leading to the auricle.

The arterial blood in the heart is thus mixed with venous blood.

(b) Pulmonata.— (Examples : Helix pomatia, Limax, Figs. 170, 171, 95, p. 100).
The blood vascular system is like that of the Monotocardia. The only important
deviation is caused by the occurrence of pulmonary respiration. Various veins col-
lect the venous blood out of the large body sinus and the lacunar system, and unite
to form one large vein, which accompanies the hind-gut, and, as vena circularis,

204

COMPARATIVE ANATOMY

CHAP.

runs along the thickened edge of the mantle which concresces with the nuchal integu-
ment. From this vein spring numerous venous vessels which spread out on the

FIG. 170.— Pulmonary veins, heart, and arterial system of Helix (after Howes). The mantle
(roof of pulmonary cavity) is cut open and turned back. 1, Pulmonary vein (efferent pulmonary
vessel) ; 2, kidney ; 3, auricle ; 4, ventricle ; 5, rectum, cut through ; 6, hermaphrodite gland ;
7, columellar muscle ; S, aorta visceralis ; !>, salivary glands ; 10, aorta cephalica.

under surface of the mantle, i.e. on the roof of the mantle cavity, and there form a
delicate respiratory network. In this network the blood becomes arterial, and is

Fio. 171.— Vascular system of Limax, aft or drawings combined by Leuckart from Delle Chiaje
and Simroth. The veins carrying the venous blood out of the body into the lungs are black.
-4, auricle ; (', ventricle ; VR, venous circular sinus of the pulmonary cavity ; Ax, aorta cephalica ;
Ay, aorta visceralis ; M, muscular stomach; ZD, hermaphrodite gland; //, digestive gland; /, in-
testine ; AL, respiratory aperture ; A', arteria genitalis.

next conducted through many vessels into the large pulmonary vein (vena pulmon-
aris), which runs back almost parallel to the rectum along the roof of the mantle

vii MOLLUSCA—THE CIRCULATORY SYSTEM 205

cavity, to enter the auricle. The vessels of the respiratory network form projecting
ribs on the surface of the mantle. The pallial epithelium in the mantle cavity is
ciliated.

The efferent pulmonary vessels, which, near the kidney, run along the right side
of the pulmonary vein, first enter the kidney and break into a fine vascular network
before passing into that vein.

The cephalic aorta does not pass through the cesophageal ring, but runs between
the pedal and visceral ganglia ; this is said to be the case in most Opisthobranchla.

In Opisthopneumonic Pulmonata (e.g. Daudebardia, Testacella), in which the
small or rudimentary visceral dome has shifted to the posterior end of the body, and
the organs elsewhere found in the dome (liver and genital organs) now lie in the body
cavity above the foot, and thus in front of the posteriorly placed heart, the posterior
aorta (A. visceralis) is much reduced, but the anterior aorta (A. cephalica) is strongly
developed. The posterior aorta supplies only the posterior lobes of the liver and the
hermaphrodite gland, and the anterior aorta (cephalic aorta, A. ascendens) has thus
to supply the anterior lobes and even part of the genital organs, which usually receive
their blood from the posterior aorta.

In Oncidium, there is an arteria visceralis corresponding with the posterior
aorta, which branches off soon after the aorta leaves the heart, but it here runs
anteriorly.

(c) Opisthobranchia. — Here also the arrangement is essentially the same as in
the Prosobranchia, though modified by the different position of the gills, as has been
already briefly noted.

Gastropteron affords a good illustration of the circulatory system of the Tecti-
branchia. The heart, which is enclosed in a spacious pericardium, lies to the right,
in front of and above the base of the gill. It lies transversely, the larger and more
muscular ventricle to the left, the auricle to the right. Out of the ventricle springs
the aorta, which at once divides into a posterior and an anterior aorta. The anterior
aorta enters the cephalic cavity, giving off as its principal arteries : (1) the artery of
the copulatory organ. (2) The two large pedal arteries, each of which again soon
divides into two branches, viz. (a) an anterior artery, which branches richly in the
parapodia ; (6) a posterior artery, which runs back on each side parallel to the
median line of the foot. (3) The arteries of the cephalic disc. (4) The arteries
of the cesophageal bulb and of the cesophagus. (5) The anterior end of the aorta
itself branches in the tissues surrounding the mouth. The following are the chief
branches of the posterior aorta : (1) The gastric artery. ('2) The hepatic arteries.
(3) The genital arteries. The venous blood flows back from all parts of the body
through richly - branched channels into two large venous sinuses, one of which
represents the cephalic and the other the body cavity. Wide but short vessels
convey the venous blood out of these sinuses into the kidney, which contains a
rich venous lacunar system. From the kidney it flows direct into the afferent
branchial vessel, becomes arterial in the gills, and collects in the efferent branchial
vessel, which, as the branchial vein, soon enters the auricle.

All the venous blood in Gastropteron, therefore, on its way back to the heart,
passes first through the kidney and then through the gill, so that only arterial blood
flows through the heart.

This is, however, not by any means the case in all Tectibranchia. For example,
in Pleurobranchus, a large part of the venous blood passes from a dorsal circular
sinus through a very short but wide passage direct into the branchial vein close to
its point of entrance into the auricle, passing by both the kidney and the gill.

Dorididse. — Without going into details as to the circulatory system of this group,
it may be mentioned that part of the venous blood passes directly through two
lateral vessels into the auricle. Another part flows into an inner venous circumanal

206 COMPARATIVE ANATOMY CHAP.

sinus, which lies at the base of the circle of gills. From this the blood rises into the
gills, becomes arterial, flows back into an outer circumanal vessel, and thence back
through the branchial vein into the auricle (Fig. 93, p. 98).

Nudibranchia. — The heart, enclosed in the pericardium, almost always lies in
front of the centre of the body, in the median line. The aorta, which springs from
the ventricle, divides into an anterior and a posterior aorta, each of which breaks up
into an arterial system, the arteries having walls of their own. The finer branches of
these arteries open into the lacunar system of the body, which occasionally forms
canals resembling vessels, and is connected with the large cephalic and visceral
sinuses. Veins, apparently with walls of their own, run from the lacunar system of
the dorsal appendages or the integument, and carry the arterial blood back to the
auricle. The blood iisually finally enters the heart through three "branchial" veins,
—two lateral and one median posterior, — which open into the posteriorly-placed
auricle.

3. Scaphopoda.

The circulatory system of Dentalium, but for the recently-discovered rudimentary
heart, is entirely lacunar, consisting of systems of canals, sinuses, and spaces, the
special arrangement of which cannot here be described.

The pericardium with the heart lies on the posterior side of the body, dorsally to
the anus. If we imagine the intestine of Dentalium straight and horizontal, the
heart would occupy the typical position on the dorsal side of the hind-gut. It has
no auricles, and is merely a sac-like bulging into the pericardial cavity of its
anterior wall. It is connected by fine slits with the surrounding sinuses of the body.

4. Lamellibranchia.

The Heart. — In nearly all bivalves, the heart, which is traversed by
the hind-gut, possesses two lateral auricles, and lies in a pericardium.

There are, however, isolated exceptions to this rule. In Nucula,
Area, and Anomia, the ventricle lies over (dorsally to) the hind-gut.
This dorsal position must be regarded as the primitive position of the
Lamellibranchiate heart, since the above genera are among the most
primitive bivalves, and, further, since the heart of the A mphineura, the
Scaphopoda, and the Cephalopoda also lies over or behind the hind-gut.
The perforation of the heart by the hind-gut must have arisen by the
bending of the ventricle down round the latter.

The heart in the above-mentioned genera is further distinguished by the fact
that the ventricle is more or less elongated in the transverse direction, its lateral
ends being swollen, while the central part, which lies above the intestine, becomes
narrower and thinner. This modification goes furthest in Area Nocc, where there
seem to be two lateral ventricles unconnected by a central portion. This separation
of the ventricle into two lateral parts has here brought about a separation of the two
aorta. The two anterior as well as the two posterior branches, however, after a
comparatively short separate course, unite to form an unpaired anterior and an un-
paired posterior aorta.

Although these genera have, as a rule, a heart lying above the hind-gut, in some
specialised forms the heart is placed under the hind-gut, e.g. Melcagrina, Ostrea,
Teredo. The cause of this modification must lie in the increasing distance between
the base of the gills and the original region of the heart, the auricles and the ventricle
having shifted with the latter. The auricles, however, no longer lie laterally to the

VII

MOLLUSCA—THE CIRCULATORY SYSTEM

ventricle, but arc drawn down to its lower side, where they grow together, communi-
cating through a more or less large aperture. Pinna, Avicula, and Perna exhibit
the consecutive stages in the displacement of the heart to the lower side of the hind-
gut. The shifting of the gills from the original region of the heart just mentioned
is caused by the shifting forward of the posterior adductor, which grows more and
more massive and finally reaches a median position on the shell valve. It has already
been mentioned that this posterior adductor, by the continuous reduction and final
disappearance of the anterior adductor, becomes the one adductor of the Mono-
myaria.

In Teredo also, the heart lies on the under side of the hind-gut. This is con-
nected with the approximation of the hind-gut with the anus to the mouth dorsally,

ro

an

FIG. 172.

FIG. 173.

Fio. 172.— Transverse section through Anodonta, to illustrate the course of the circulation
of the gills and the kidneys, and the branchial veins (after Howes), br, Gills ; brc, efferent
braiirhial vessel (branchial vein) which opens into the large branchial vein brej, running along the
base of the gills, and here cut through transversely ; pv, pallial vein ; vc, large venous sinus of the
body ; kb, pericarclial gland ; a«i, auricle ; rj, rectum ; v, ventricle ; -n> and rrj, renal vessels ; In'iij,
afferent branchial vessel (branchial artery), running along the base of the gills ; bra, lateral brandies
of the same running in the gills. The veins or sinuses conveying venous blood are black.

Fta. 173.— Another section through Anodonta (after Howes). .Lettering as in Fig. 172. au,
auricle ; sbc, spaces at the base of the gills, bathed by the water and communicating with the
mantle cavity, between the ascending and descending branchial laiiielhe.

while the gills, remaining in their original position, retain the heart on the lower
side of the hind-gut.

Circulation (Fig. 25, p. 17). — The arteries have walls of their own, and branch
into fine vessels, which discharge the blood into the lacunar system of the body.
The venous system seems to have no distinct vessels with walls of their own.
although it forms more or less wide channels resembling true vessels.

An anterior and a posterior aorta spring, as a rule, from the ventricle. The
anterior aorta runs forward above the intestine and breaks up into various arteries.
The arteria visceralis supplies the intestine, the digestive gland, and the genital
gland ; the pedal artery supplies the foot ; the anterior pallial artery spreads out
over the anterior part of the mantle and the oral lobes (labial palps).

208 COMPARATIVE ANATOMY CHAP.

The posterior aorta leaves the ventricle posteriorly and rims along the lower side
of the hind-gut. It soon divides into two large lateral arteries, — the posterior pallial
arteries. The principal branches of the anterior and posterior pallial arteries run
along the free edge of the mantle on each side and then unite, forming together the
arteries of the pallial edge. From the roots of the posterior pallial artery smaller
arteries spring, which supply with blood the hind-gut, the pericardium, the posterior
adductor, the retractors of the siphons, etc. The venous blood is collected out of
the lacunar system of the body through converging channels into one longitudinal
venous sinus ; this lies under the pericardium (Fig. 172).

From this sinus, the greater part of the blood flows through the complicated
system of venous canals in the kidneys, after which it is collected on each side into
a branchial artery which runs along the base of the gills, and thence enters the
two branchial lamellfe. It becomes arterial through respiration in the gills, flows as
arterial blood into a branchial vein parallel with the branchial artery, and thence
into the auricle.

Part of the venous blood, however, passes by direct channels out of the venous
sinus into the branchial artery (passing by the kidneys), and part even flows direct
into the pericardium. In this way some venous blood comes to be mixed with the
arterial blood flowing through the heart from the gills.

Not all Lamellibranchia have an anterior and a posterior aorta springing out
of the heart. In the lower groups of the Protobranchia and Filibranchia there are
numerous forms (Nueula', Solenomya, Anomia, Mytilidce) in which only one anterior
aorta leaves the ventricle ; this soon, however, gives off the arteria visceralis, which
supplies blood to those parts which, in other Lamellibranchia, are fed by the aorta
posterior. In their possession of a single aorta rising from the ventricle, the above
lower Lamellibranchiates agree with Chiton and the Gastropoda. The rise of this
aorta from the posterior end of the ventricle in the Prosobranchia and in most
Pulmonata is a secondarily acquired arrangement, caused by the shifting forward
of the pallial complex.

It must further be noted that in a very specialised bivalve, Teredo, the posterior
aorta fuses with the anterior, and thus the two leave the heart as one vessel.

In those Lamellibranchiates which have siphons, a muscular and contractile
widening occurs in the posterior aorta near the point where it leaves the ventricle ;
this is called the bulbus arteriosus. Its special function is perhaps that of bringing
about pressure of blood, to assist in the extension of the siphons. The backward
flow of the blood into the ventricle in the contraction of the bulbus arteriosus
(systole) is prevented by a linguiform valve which projects from its anterior wall.

5. Cephalopoda.

Heart (Figs. 127, 168, pp. 147, 199, and 174). — We must here again point out the
important fact that Nautilus has a heart with four auricles, while the Dccapoda and
Octopoda a heart with only two auricles. This difference is connected with the
difference in the number of the ctenidia : four in Nautilus (Tetrabranchia), two
in the Decapoda and Octopoda (Dibranchia).

In Nautilus, the heart is an almost square sac drawn out to two points on each
side ; the four auricles which open into the four points of the ventricle are long
tubes, more like widened branchial veins than auricles.

The strongly muscular ventricle of the Dibranchia is almost always elongated
into a tube. In the Octopoda it lies transversely, the two auricles being in the same
plane with the ventricle. In the Ocgopsidce, the ventricle lies along the longitudinal
axis of the body, i.e. it is elongated dorso-ventrally, and the auricles are at right

VII

MOLLUSGA—THE CIRCULATORY SYSTEM

209

angles to it. The heart of the Mijopsidce occupies a position halfway between those
just mentioned.

The heart here described is the arterial heart, which corresponds with the heart
of the other Mollusca. It is called arterial to distinguish it from the venous hearts,
which will be described below.

Circulation. — It is important to note that the circulatory system is at least
partially closed. There is not only a richly-branched arterial, but a richly-branched
venous system, the vessels of which have walls of their own. These two systems
pass into one another in certain parts of the body, e.g. the integument and certain
muscle layers, through a system of capillary vessels. In other parts, however, the
arterial branches conduct the blood into a lacunar system ; when it has become

FIG. 174.— Circulatory system, venous appendages of the nephridial system, and gills of
Sepia officinalis, anterior view (after Hunter). 1, Aorta cephalica ; 2, ctenidium ; 3, vein leading
to the ctenidium ; 4, branchial heart ; 5, appendage of the branchial heart (pericardial gland) ;
6, venous appendages of the nephridial system ; 7, aorta abdominalis ; 8, vena abdominalis ; 9,
lateral veins ; 10, vena cephalica ; 11, auricles ; 12, ventricle (cf. Fig. ISO).

venous, the blood collects out of this into sinuses (especially into a peripharyngeal
cephalic sinus), and flows to the gills through veins with walls of their own.

Two aorta rise from the ventricle : (1) the aorta cephalica, which runs downward
(upwards in the figure) to the head, and (2) the aorta abdominalis, which runs up
towards the apex of the visceral dome. The former is much stronger than the latter.
The aorta cephalica first gives off branches to the mantle and to the anterior wall of
the body, and then provides the stomach, the pancreas, the digestive gland, the
oesophagus, the salivary glands, and the funnel with arteries. After accompanying
the oesophagus, it divides in the head into two branches, which run to the bases of
the arms, and there break up into as many arteria? brachiales as there are arms.

The aorta abdominalis supplies with arteries the hind-gut, the ink-bag, the
genital organs, the dorsal part of the body wall, and the fins, when these latter are
present.

Only in the Ocgopsidce are the aorta limited to the two, above described, springing
from the heart. In the Octopoda and the Myojisidcc, there are other arteries rising out
of the ventricle, and running to the same part of the body as the aorta abdominalis
VOL. II P

210 COMPARATIVE ANATOMY CHAP.

in the Ocgopsidce ; among these are the arteria genitalis, which runs to the genital
glands, and, in the Myopsidcc, a fine vessel called the arteria anterior.

At certain places, the arteries may swell out to form small muscular and con-
tractile widenings, called peripheral arterial hearts.

In the venous system of Sepia, the venous blood in each arm collects (partly
through capillaries and partly through lacuna?) into a vein running down the inner
side of the arm. All the brachial veins convey their blood to a circular cephalic
sinus surrounding the buccal mass, which is the reservoir for collecting the venous
blood from the whole head region. Out of this sinus springs the large vena
cephalica, which runs up along the posterior side of the oesophagus and the liver
into the visceral dome, collecting on the way venous blood from the liver, the funnel,
etc. A little below the stomach it forks, forming the two venae cavse, which open
into the two contractile venous hearts at the bases of the gills. From the upper
part of the visceral dome the blood collects into several abdominal veins, the most
important of which are an unpaired vena abdominalis, opening into the vena
cephalica exactly at the point where it divides into the venae cavae, and two lateral
abdominal veins, which open into the latter near their point of entrance into the
branchial hearts.

In the region of the heart, all these veins carry acinose or lobate appendages
(venous appendages), which are hollow, and communicate at many points with the
veins, so that they are richly supplied with blood. The cavity into which these
appendages project is that of the renal sacs, and the epithelium which covers them
belongs to the epithelial wall of the kidneys (cf. Fig. 186, p. 224). We thus see
that here the blood flowing back from the body has abundant opportunity of giving
off its excretory constituents to the kidneys.

Appendages are found on both the branchial hearts ; these are the pericardial
glands, which will be further described later. The two branchial hearts, by their
contraction, drive the venous blood into the afferent branchial vessel. The blood,
which has become arterial in the gills, flows through the efferent branchial vessel
(the so-called branchial veins) into the auricles of the heart, and thence into the
ventricle (on the branchial circulation, cf. p. 96).

In the Cephalopoda, unlike the other Mollusca, the whole of the blood, in
returning from the body, flows through the gills, so that the heart contains only
arterial blood. By far the greater part of the blood, before entering the gills, comes
into contact with the kidneys in the venous appendages.

In the Octopoda, the venous system shows some not unimportant modifications.
In Octopus, two veins, connected with one another by anastomoses, run along the
outer side of each arm and collect the venous blood. At the bases of the arms these
veins become connected in pairs, and unite later in such a way as to form on each
side a lateral cephalic vein.

These two veins unite to form the large vena cephalica, which runs up in front
of the funnel and behind the oesophagus. The brachial veins do not here, as in
Sepia, convey their blood first to the venous cephalic circular sinus, but are directly
connected with the cephalic vein. A cephalic sinus nevertheless exists in Octopus ;
it is not, however, connected with the vena cephalica, but with a large sinus which
fills the whole visceral dome, and is, in fact, the primary body cavity, in which
the viscera lie bathed by the venous blood. The latter flows out of this large
venous sinus through two wide veins, the so-called peritoneal tubes, into the
upper part of the vena cephalica, near the point where this divides into the two
venae cavse.

Nautilus is chiefly distinguished by the absence of the branchial hearts.
Further, each of the two venae cavae divides into two branches, which run, as
afferent vessels, to the gills.

vii MOLLUSCA— THE BODY CAVITY 211

XVIII. The Body Cavity.
Primary and Secondary Body Cavity, Pericardium, Pericardial Gland.

The Mollusca are said to have a primary and a secondary body
cavity. The former is the system of lacunae and sinuses, into
which the arteries open, and out of which the veins, where these are
present, draw their blood. It has no epithelial walls of its own, its
boundaries are formed by connective, nerve, or muscle tissue, or by
epithelia, which, however, belong to other organs, such as the intestine,
the kidneys, or the body wall.

The so-called secondary body cavity or eoelom is, in most Mollusca,
very much reduced, usually consisting of only two small cavities, the
pericardium and the cavity of the gonads (testes, ovaries, or her-
maphrodite glands). The eoelom is always lined by an epithelium of
its own, the ccelomic epithelium, and corresponds with the true ccelom
of the Annelida, which also possesses such an epithelium. Like the
latter, it is connected, by means of the nephridial funnel, with the
nephridia, which lead to the exterior, and in Molluscs are usually
found only in one pair. A probe can therefore be introduced through
the kidney into the eoelom, i.e. into that part of it which, containing
the heart, is called the pericardium. The germinal layers must be
considered as proliferations of the cuelomic endothelium. The epi-
thelium of the pericardium is, in very many Molluscs, differentiated
into glands, called the pericardia] glands ; these probably may be
classed together with the kidney as excretory.

We should be justified in assuming, a priori, that the lumen of the
genital glands of the Mollusca is part of a true eoelom, and that
the germinal layers themselves, i.e. that complex of cells which yields
the eggs and spermatozoa, are outgrowths of the endothelial wall of
this coelom. Direct support is, however, given to this assumption by
the fact that in the Solenogastres, Sepia, and Nautilus, the sac of the
genital glands is in open communication with the rest of the coelom,
forming, in fact, an only partly distinct division of the same.

In the Solenogastres (e.g. Proneomenia}, the hermaphrodite gland lies above the
mid -gut as a long tube, which in transverse section appears heart- or kidney-shaped,
as its lower part bulges out 011 each side. Its shape is determined by the fact that
the mid-gut forms dorsally a narrow but deep furrow, which cuts into this glandular
tube from below. The tubular gland is divided into two lateral spaces by a partition,
whose endothelial wall is the place of formation of the eggs ; these lateral chambers
may again be traversed by septa, on which the genital products develop. This
division is especially distinct at the posterior part of the tube, the two chambers
being there completely isolated, and entering the pericardium separately as genital
ducts.

If the secondary body cavity of Proneomenia is compared with that of an Annelid,
we find the following differences :

212

COMPARATIVE ANATOMY

CHAP.

Iii Proneomenia, the dorsal vessel is wanting in the region of the mid-gut. The
ccelom is much less spacious, and instead of surrounding the intestine lies only on
its dorsal side. It is developed merely as a hermaphrodite glandular sac, its endo-
thelial wall yielding the genital products.

In the region of the hind-gut, the vessel lying in the dorsal mesentery is developed
as a heart, the ccelom being here represented by the pericardium.

FIG. 175.— Diagrammatic sections through an Annelid (A) and a Solenogastrid (B and C), to
illustrate the relation of the ccelom to the genital glands and nephridia. B, Region of the cloaca ;
C, region of the mid-gut ; 1, dorsal mesentery ; 2, dorsal vessel or heart ; 3, germinal epithelium ;

4, coelom — in B= pericardium, in C= hermaphrodite gland (in the ccelom are genital products);

5, nephridia ; 6, intestine ; 7, cloaca.

The pericardium is connected with the cloaca by two canals ; these may be
considered as the morphological equivalents of nephridia (cf. Fig. 175).

As the genital glands have been recognised as part of the ccelom in the Soleno-
gastres, Nautilus, and Sepia, they must necessarily fall under the same category in
all other Molluscs, even when no longer in direct connection or in open communica-
tion with the same.

In the ChitonidcK, the ccelom is large, and falls into three distinct divisions. One
contains the intestine and digestive gland (liver), which are accordingly outwardly

M

FIG. 170.— Diagrammatic longitudinal section through Chiton, tu illustrate the relation
between the various parts of the ccelom (after Haller). 1-8, Position of the eight dorsal shell-
plates; M, anterior portion of the dorsal integument; L, snout; m, mouth; I, digestive gland-
(liver); d, intestine;/, foot; n, kidney; ji, pericardium; c, portion of the ccelom surrounding the
intestine ; h, heart ; Ip, band connecting pericardium and genital gland ; ijdr, genital gland ; In, band
connecting the genital gland and the posterior portion of the ccelom which surrounds the intestine.

(i.e. on the side turned to the ccelom) covered with an endothelium. The mesen-
teries, however, which originally attached the intestine to the body wall, and
along which the parietal endothelium passed into the visceral endothelium of the
intestine and liver, have disappeared, with the exception of portions retained on
the hind-gut. The two other divisions of the ccelom are : (1) the pericardium, and

VII

MOLLUSGA—THE BODY CAVITY

213

(2) the genital gland. Certain bands, by means of which the three divisions are
connected together, have been regarded as the constricted remains of communications
between the three divisions of the originally single coelom (Fig. 176).

The Cephalopoda may with advantage be considered in connection with the
Amphineura. In Nautilus and the Decapuda (e.g. Sepia, Fig. 177) a spacious
secondary body cavity is found in the dorsal part of the visceral dome. It is incom-
pletely divided by a projecting dorsal septum into two cavities, one lying above the
other ; the lower of these contains, as pericardium, the heart with the arteries and
veins running out of and into it, the branchial hearts, and the pericardia! glands ;
while the upper holds the stomach and the genital glands. This double cavity,

FIG. 177.— Diagram showing the
secondary body cavity of Sepia (after
Grobben). Median longitudinal section
through the body, in which, however, some
organs are represented which, being paired
and symmetrical, do not properly come
into the plane of the section. The outlines
of the coelom are indicated by thicker lines.
1, Female germinal body, with eggs (2) pro-
jecting into the genital cavity (the ovarial
division of the coelom); 3, shell; 4ft, an-
terior portion of the renal sac ; 5, pancreatic
appendage of the efferent duct (bile duct)
of the digestive gland (liver) ; 4«, anterior
venous appendage of the renal system ; G,
aperture (funnel) of the kidney into the
crelom ; 7, outer or pallial aperture of the
kidney ; 8, digestive gland (liver) ; 9,
" head " (Kopffuss) ; 10, funnel ; 11, end of
the oviduct with female genital aperture ;
12, mantle cavity ; 13, mantle ; 14, posterior
portion of the renal sac ; 15, intestine ; 14],
posterior venous appendage of the renal
system (pericardial gland); IS, fold, in-
completely dividing the coelom into an
upper and a lower portion ; 19, stomach ;
20, upper division of the cceloin (principally
genital cavity); 21, pigment gland (ink-
bag); 22, aperture of the oviduct into the
genital cavity ; rf, dorsal ; c, ventral ; n,
anterior; y),»posterior.

which is called the viscero- pericardial cavity, is covered by endothelium, which also
covers the organs within it. It is connected by two ciliated funnels with the two
renal sacs. In Nautilus it also opens direct into the mantle cavity by two canals,
whose apertures lie close to the renal apertures.

While the ccelom in Nautilus and the Decapoda is very spacious, in the Octopoda,
on the contrary, it is very much reduced. It consists merely of a narrow system of
canals, which, however, have thick walls ; this was formerly called the water vascular
system. The organs, which in Nautilus and the Decapoda lie in the coelom, viz.
the arterial heart with its afferent and efferent vessels, the branchial hearts and the
stomach, are no longer found within the body cavity, but outside of it, and are

214

COMPARATIVE ANATOMY

CHAP.

— \ 19

therefore no longer covered with endothelium. Nevertheless this canal system of
the Octopoda shows the same morphologically important characteristics as the ccelom
of the Dccapoda. There are, for instance, on each side three canals which open
together, one entering the renal sac, the second widening round the pericardial
gland (appendage of the branchial heart) to form a flask-shaped capsule, and the
third running to the genital gland to be continued into its wall. [In so far as in the
Octopoda the heart is excluded from the ccelom, which has been reduced to the ' ' water
canal system," the reduction of this cavity has gone further in these Mollusca than
in any others, which all retain at least the heart in one portion of the ccelom, the
pericardium.

In the Lamellibranchia and Gastropoda, the only part of the ccelom retained,
besides the genital glands, is the pericardium. The pericardium and the gonad are,

, however, entirely separated

from one another. In Lamelli-
Iranchs, there is in the peri-
cardium, besides the heart, a
part of the hind - gut which
traverses it ; in the Gastropoda
(except in those Dioiocardia
in which the hind-gut pene-
trates the heart), only this
latter organ. Rarely (e.g.
Phyllirhoe) the auricle also
is excluded from the peri-
cardium.

The pericardial gland is
found in most Mollusca. It
is a glandular differentiation
of the endothelial wall of the
pericardium, and perhaps, as
already suggested, shares the
excretory functions of the
kidney. Its position iu the
pericardium varies, but it
seems in all cases shut off from
the blood vascular system,
with which it is, however,
functionally connected. Its
secretions or excretions must be discharged into the pericardium, and thence out-
wards through the kidney.

Among the Prosobraiichia, in the Diotocardia, the pericardial gland is found on
the auricle, its walls forming dendriform branched outgrowths into the pericardial
cavity, these being covered with pericardial endothelium. Where pericardial glands
are found in the Monotocardia, they lie on the wall of the pericardium itself.
Similar lobate formations occur among the Opisthobrancliia, in Aplysia, and
Notarchus, on the anterior aorta which runs along the pericardial wall ; in Pleuro-
Imiichus and Plcurobranchcca on the lower, in Doridopsis and Phyllideea on the dorsal
pericardial wall. The lateral furrows of the pericardium of Doris form niches, which
may again have accessory niches. These enlargements of the surface of the peri-
cardial epithelium have also been considered as pericardial glands.

Pericardial glands are much more common among the Lamcllibranchia than
among the Gastropoda, but are wanting in the most primitive forms (Nucula,
Solcnomya, Anomia). The gland is usually of a rusty red colour, and occurs in two

FIG. 1VS.— Eledone moschata. This figure corresponds with
Fig. 177 of Sepia (after Grobben). Si, Efferent duct of the
digestive gland; 17a, pericardial gland (appendage of
branchial heart) ; 23, water canals.

the

vii MOLLUSC A— THE NEPHRIDIA 215

forms, consisting either of glandular protrusions of the endothelial wall of the
auricles into the pericardial cavity, or of glandular tubes protruding from the
anterior corner of the pericardium into the mantle (Keber's organ, red-brown
organ). The first form is found specially strongly developed in Mylilus, Lithodomus,
and Saxicava, more or less developed in Dreissena, Unio, Anodonta, Venus, Car-
dium, Scrobicularia, Solcn, Pholas, and Teredo, and more or less rudimentary in
Pecten, Spondylus, Lima, Ostrea. The second form has been observed in Unio,
Anodonta, Venus, Cardium, Scrobicularia, Solen, Pholas, Montacuta, and Dreisscnsia.
Pericardial glands may also occur singly in other parts of the pericardium, as in
Meleagrina (as a projecting ruff in the posterior base of the pericardium), and in
Chania on the ventricle, etc.

The pericardial gland of the Cephalopoda is the so-called appendage of the
branchial hearts. This is a structure connected with the branchial heart, and
covered with peritoneal endothelium, which projects into the viscero-pericardial
cavity, or, in the Octopoda, into a flask-like widening of the water-canal system
(which has been recognised as a division of the ccelom). In Sepia this appendage is
conical. A deep furrow on the surface which projects into the viscero-pericardial
cavity leads into a richly-branched system of canals, the glandular epithelium of
which is a continuation of the peritoneal epithelium. Blood sinuses from the
branchial heart penetrate in between the canals of this system. In other Cephalo-
poda, the pericardial gland varies in form and structure ; details of these variations
cannot, however, be here given. Nautilus possesses two pairs of pericardial glands ;
this fact is connected with its possession of two pairs of gills, with their two pairs
of afferent vessels, and on these the two pairs of pericardial glands occupy positions
corresponding with those of the branchial hearts.

XIX. The Nephridia.
Kidney, Organ of Bojanus.

The organs which serve for excretion are homologous in all
Mollusca.

They consist typically of two symmetrical sacs, which, on the one
hand, open into the mantle cavity, through the two outer renal
apertures, and on the other are connected by two inner apertures
(renal funnels, ciliated funnels) with the pericardium or coelom. The
nephridia always lie near the pericardium. Their walls are richly
vascularised, indeed a large part of the venous blood, in returning
from the body, flows through the renal walls and gives off excretory
matter before it enters the respiratory organs. The renal walls
are traversed exclusively by venous blood.

The nephridia are paired in all symmetrical Molluscs, and also in
those Gastropoda which have paired gills and two auricles (Diotocardia).

In all other Gastropoda, along with the original right ctenidium
(which, in the Prosobranchia, lies to the left), and the corresponding-
auricle, only one kidney (the corresponding one) is retained.

Nautilus, which has four gills and four auricles, has also four
kidneys ; only two of these, however, communicate with the viscero-
pericardial cavity.

216

COMPARATIVE ANATOMY

CHAP.

A relation between the nephridial and genital systems similar
to that found in the Annelida exists in the Sohnogastrida?, the
nephridia functioning as ducts for the genital products, the latter
passing from the hermaphrodite gland (genital chamber of the ccelom)
into the pericardium.

In a few Lamdlibranchia, Diotocardia, and in the Scapliopoda, there
is a relation between the genital glands and the nephridia, the former
opening into the latter ; so that a certain part of the nephridium
functions not only as renal or urinary duct, but also as efferent genital
duct. In all Diotocardia, it is the right nephridium which functions as
genital duct. In the Monotocardia, in which the right nephridium of
the Diotocardiii has atrophied as such, its duct persists as genital duct.
In all other Molluscs the genital ducts are entirely distinct from the
urinary passages.

A. Amphineura.

The kidneys of the Solcnocjastridw and the Chitonidcc differ greatly from one
another in structure.

1. In the Sole nogasfr idee, two canals spring from the pericardium, embrace the
hind-gut, and open into the cloaca beneath it through a common terminal portion

52 1

11

FIG. 179.— Paramenia impexa. Posterior end of the body ; the integument must be supposed
to be removed on the right side, and also a piece of the wall of the right nephridium ; diagram (after
Pruvot). 1, Integument; 2, ovarial portion of the hermaphrodite gland ; 3, testicular portion of
the same, near the point where the latter opens into the pericardium (4) ; 5, glandular appendage
of the right nephridium ; 6, dorsal commissure of the pleurovisceral cords ; 7, organ called the
sensory bud ; 8, aperture of the hind-gut into the cloaca; 9, gill; 10, cloaca; 11, common aper-
ture of the nephridia into the cloaca : 12, lower portion of the nephridium ; 13, upper portion of
the right nephridium, which opens above into the pericardium ; 14, hind-gut.

(Fig, 179). These canals function as ducts for the genital products. It is also
certain that they correspond morphologically with the kidneys of other Molluscs,
even though their excretory activity has not been proved. They are covered with
an extraordinarily deep epithelium of long filiform glandular cells.

In some Solcnogastridce, an accessory gland opens into each nephridial canal.

2. In the Chitonidce, the strongly-developed paired nephridia function exclusively
as excretory organs.

Each nephridium (Fig. 180) consists of a wide canal shaped like a long Y,

VII

MOLLUSCA—THE NEPHRIDIA

217

the diverging portions being directed backward, and the undivided portion
forward. These Y-shaped kidneys run longitudinally along each side of the body
through its whole length. One of the paired limbs of the Y opens outward into
the posterior part of the mantle cavity, the other into the pericardium, which also
lies in the posterior part of the body. In this way the pericardia! and outer aper-

1!

FIG. ISO.— Nephridial and genital systems of Chiton, diagrammatic, from above, after the
figures and accounts of various authors. 1, Mouth ; 2, gills ; 3, unpaired portion of the nephridium
which runs forward, with its lateral branches ; 4, gonad ; 5, efferent ducts of the gonad ; (3, portion
of the "nephridium running to the outer aperture (10) ; 7, portion running to the reno-pericardial
aperture (9) ; 8, genital apertures ; 9, reno-pericardial funnel ; 10, nephridial aperture ; 11, peri-
cardium, indicated only in outline ; 12, anus.

tures of the kidney lie near one another. The third limb of the Y ends blindly
anteriorly. Secondary lobules or lobed canals open into all the three parts of the
kidney, and are specially abundant in its anterior portion. Except in the terminal
portion of the efferent branch, the epithelium of the limbs as well as that of the lobes
is cubical and ciliated.

B. Gastropoda.

1. Prosobranchia. (<i) Diotocardia. — Among all the Gastropoda, Fissurella
alone possesses a symmetrical excretory apparatus, in the sense of having two

218

COMPARATIVE ANATOMY

CHAP.

nepliridia opening into the mantle cavity to the right and left of the anus. The
left nephridium is, however, much reduced, while the right, which is strongly
developed, sends its lobes everywhere into the spaces' between the lobes of the liver,
the intestine, and the genital organs. There are no reno-pericardial openings. The
genital gland does not open direct into the mantle cavity, but through the right
kidney.

In Haliotis, Turbo, and Trochus, both nepliridia are present. The left nephri-
dium has, however, almost entirely lost its excretory function, but is still connected
both with the pericardium and the mantle cavity. It is called the papillar sac, its
walls projecting into its cavity in the form of numerous large papillte. The blood
lacunas which penetrate into the papilla; communicate direct with the auricles, and
are thus supplied with arterial blood. In these lacuna of the papilla? a crystalloid
substance (albumen ?) is deposited. It has been thought that these papillar sacs
serve as reservoirs of nutritive material (in the form of the crystalloids just men-
tioned), and when needed yield it up to the blood.

The right nephridium is exclusively excretory in function. It is divided into
two lobes, one behind the other, which communicate by means of a wide aperture ;
the anterior lobe lies under the floor of the mantle cavity, bulging it upward. A
spongy network, covered with excretory epithelium, rises from part of its wall into
the cavity of the nephridial sac. The meshes of the network are penetrated by a
system of vessels with walls of their own. Nearly all the venous blood, before

reaching the gills, passes through the
vascular system thus developed on the walls
of the kidneys. The right nephridium is
in no way connected with the pericardium.
The Neritidae have only one nephridium
to the right of the heart, which opens
through a slit in the base of the mantle
cavity. The renal sac is traversed by trabe-
culae, many of which reach from one wall to
the other, forming a spongy structure. The
trabeculse are covered by a glandular epithe-
lium on the surfaces turned to the spaces of
the sac.

Patella (Fig. 181) still has two nepliridia,
both functioning as excretory organs. The
apertures lie at the two sides of the anus.
FIG. 181.— Diagram of the two nephridia The right kidney is, however, much larger
of Patella (after Lankester). l-sa, Anterior h ^ left_ T, both l[& ^ the ^ of
and upper lobe of the large right kidney l:sl ; J

fcri, lower sui.visceral ; ksp, posterior lobe of the pericardium, but there are no reno-peri-
the same ; /, subanal tract of the large right cardial apertures. The internal structure
kidney ; g, anal papilla with the portion of the of the right kidney is spongy, but the left
rectum which runs to it ; It, papilla with the formg a s;mpie cavity, into which folds
aperture of the left kidney (which is not h u A lacunar

drawn) ; /, the same of the right kidney ; /,

pericardium, indicated by a dotted outline ; without special walls traverses the trabecular
the existence of the reno-pericardial aperture network of the right kidney, but is com-
flgured near/, is now denied. pletely cut off from its cavity ; the venous

blood from the body passes through this

system before entering the gills. The lacunar system of the left kidney communicates
directly with the auricle.

In Haliotis and Patella also the genital products pass, as in Fissurella and the
Diotocardin generally, out of the genital gland into the right kidney, and are ejected
through the right renal aperture.

vn MOLLUSC A— THE NEPHRIDIA 219

(b) Monotocardia. — The Monotocardia have only one nephridium functioning as
an excretory organ, viz. the left of the Diotocardia. This takes the form of a sac
lying immediately below the mantle cavity on the right side of the pericardium,
directly under the integument. It is generally found to the left of the hind-
gut ; less frequently (Cassidaria , Tritoniidce) the kidney is traversed by the rectum,
or the latter runs forward below it. The slit-like pallial aperture of the kidney,
however, is always found to the left of the hind-gut, quite at the base of the mantle
cavity. This position of the kidney, and especially of its outer apertures, had
already led to the assumption that the Monotocardian nephridium corresponds with
the left kidney of the Diotocardia, before this fact was established. The assump-
tion was all the more plausible because of the occurrence of a gland called the
anal kidney in a few Monotocardia (e.g. Dolium) ; this gland opens to the right
near the anus, and might represent the right kidney of the Diotocardia.

The kidney is always connected by means of a canal (the reno-pericardial canal)
with the pericardium.

Lamella? or trabecula?, covered with the glandular epithelium of the kidney,
project inward from the lateral walls of the renal sac. These are especially
strongly developed in fresh - water Prosobranchia (excepting Valvata), traverse
the whole kidney, and impart to it a spongy structure. The venous blood always
flows through the whole of the glandular part of the kidney, either in special
vessels or in lacunae, before passing on to the gills ; but an open communication
with the renal cavity is never found.

In the Tcenioglossa Proboscidifera the kidney forms two lobes similar in struc-
ture. In Natica and Cyprcca the lobes begin to differ, and among the Stenoglossa
this difference becomes more and more marked in a way which need not here be
described.

In Paludina and Valvata the kidney no longer opens into the posterior base of
the mantle cavity, but is continued as a urinary duct (ureter), which runs forward
in the mantle and opens at its edge.

The above-mentioned theory that the single kidney of the Monotocardia corre-
sponds with the left kidney of the Diotocardia has recently been ably opposed,
another theory being put forward in its place. Attention is specially drawn to the
fact that in the Diotocardia the left kidney is always the smaller, that in Patella it
is shifted to the right side of the pericardium, and that in Haliotis, Turbo, and
Trochus (as papillar sac) it is not excretory in function. In Haliotis, Turbo,
Trochus, and Patella the lacunar system developed in the wall of the left kidney is
in direct communication with the auricles.

In most Monotocardia there is a differentiated part of the kidney, viz. that
which is called the nephridial gland. This consists of two principal parts : (1)
canals, covered with ciliated epithelial cells and opening into the kidney. These are
merely protrusions of the renal wall, which project into the organ ; their epithelium
is a continuation of the renal epithelium. (2) Between these canals, the organ is filled
with cells of connective tissue and muscles, and contains blood lacuna?, one of these
being specially large and communicating with the auricle. This latter portion of
the organ perhaps plays the part of a blood-forming gland.

This nephridial gland may perhaps be the persistent excretory portion of the lost
nephridium, i.e. the right of the Diotocardia. The duct of this lost nephridium is
now known to persist as genital duct. As we saw above, all Diotocardia discharge
the genital products through the right nephridium.

2. Pulmonata (Fig. 182). — The Pulmonata have only one kidney, which lies
in the mantle at the base of the pallial cavity, between the rectum and the peri-
cardium. The renal sac is of the so-called parenchymatous type, the excretory
epithelium of its wall projecting into the cavity in the form of numerous

220

COMPARATIVE ANATOMY

CHAP.

folds and lamella} in such a way as to leave hardly any central free space. The
kidney always communicates by means of a ciliated canal (renal funnel or renal
syringe, " Nieren-Spritze ") with the pericardium. The position of the kidney and
the morphology of the urinary duct have already been explained (pp. 74-78).

3. Opisthobranchia— Tectibranchia. — Only one kidney is found in the usual
position on the right side of the body, with the pericardium in front of it and the
hind-gut behind it. It is of the parenchyniatous type, and is connected by a
ciliated canal with the pericardium. It opens at the base of the gill in front of the

anus.

In the Pteropoda the delicate-walled kidney is not parenchyniatous, but is a

---c

Km. ls-2.— Neparidium and pericardium of Daude-
bardia rufa, from above, diagram (after Plate). 1, Peri-
cardium ; '2, reno-pericardial aperture (renal funnel) ; 3,
nephridium ; 4, primary ureter ; 5, rectum ; 6, seeon.lni v
ureter (t/. Fig. 74, p. 77).

FIG. 183.— Nephridium of Bornella (after Hancock).
1, Kidney ; 2, part connecting it with the reno-pericardial
aperture (pyriform vesicle, renal syringe) ; 3, part of the
pericardial wall ; 4, ureter ; 5, nephridial aperture.

simple hollow cavity lined with epithelium, and always communicates with the
pericardium, against which it lies.

Nudibranchia (Fig. 183). —The kidneys of the Nudibranchia are strikingly
different in form from those of the Tectibranchia. The unpaired kidney is here
somewhat similar to the paired kidney of the Chitonidcv. It is a somewhat wide
tube (renal chamber) traversing the cavity of the body, to a greater or less extent ;
branches entering it from all sides. This tube is connected at one end with the
pericardium by a duct (renal syringe, pyriform vessel), which varies in length, and
at the other opens outward through a ureter at the base of or near the anal
papilla.

It is said that Pleurubrancluva, a Tcctibmnchiate, from which the Nudibranchia
may perhaps be derived, possesses a Nudibranchiate kidney.

In Pln/llirhoe, the urinary chamber has no branchings ; it runs back from the

VII

MOLL USC A—THE NEPHRIDIA

221

pericardium as a simple median tube. Anteriorly it is connected with the peri-
cardium by a funnel, and near the middle communicates with the exterior by means
of a lateral urinary duct (Fig. 19, p. 1'2).

C. Scaphopoda (Fig. 165, p. 193).

Dentaliwm has a pair of symmetrical kidneys, one on each side of the hind-gut.
Each nephridium consists of a sac provided with short diverticula. The two nephri-
dia are connected by a tube above the anus, and open into the mantle cavity by
two apertures at the sides of the anus. If, as maintained by all authorities, there
are no reno-pericardial apertures, the Scaphopoda would be the oiily group of Molluscs
in which these apertures are entirely absent. Apart from the symmetry of the
kidneys, a fact to be specially noted is that the genital products pass out of the
genital gland into the right kidney (either by the bursting of the wall between the
two organs or through an aperture), and only reach the exterior, i,i\ the mantle
cavity, through the right renal aperture.

It must, further, be noted that near the anus on each side, between it and the
renal aperture, a pore, the water-pore, occurs, the function of which is still doubt-
ful. If these pores really lead into the blood lacunar system of the body, as was
formerly maintained, and is still held to be possible, this would be the only known
case of the direct imbibition of water into the blood.

D. Lamellibranchia.

The nephridium (organ of Bojanus) is always paired and symmetrical, and lies
below the pericardium and in
front of the posterior adductor.
Each nephridium is tubular or
sac-like, opening at one end
through a funnel into the peri-
cardium, and at the other into
the mantle cavity. This com-
munication of the kidney with
the mantle cavity always takes
place above the cerebrovisceral
connective.

The lowest Lamellibranchia
(Protobranchia, Nucula, Li-iln.
8olenomya) are distinguished
in two ways. (1) Eacli nephri-
dium is a simple tube, with a
free cavity not traversed by
trabeculfe or lamella?. This
tube consists of two portions
which unite posteriorly at an
angle ; the anterior end of one

FIG. 184.— Transverse section through the body of Ano-
donta, showing the pericardium, thf heart, and the kidneys,
combined and diagrammatised from figures by Griesbach.

of these portions enters the Not au the parts represented occur on the same section. 1,

pericardium through the renal Pericardium; 1, ventricle ; 3, auricles; 4, hind -gut; 5,
funnel, the other end opens
into the mantle cavity. (2)

The paired genital "lands do lnantle cavity throu*h the nePhridial ai.t-i-ture ; 10, genital

aperture ; 11, base of the foot,
not open outward directly, but

enter the kidneys near their pericardial funnel — a fact which is very important in

venous sinus ; 6, reno-pericardial aperture (funnel) ; 1, renal
sac or cavity ; 8, vestibular cavity, which at 9 enters the

222 COMPARATIVE ANATOMY CHAP.

connection with the arrangement in the Solcnogastridcc, the lower Prosobranchia (i.e.
the Diotocardia), and the Scaphopoda.

In other Lamellibranchia also there is a relation between the genital glands and
the kidneys. In the Pectinidcc and the Atwmiidrc the genital gland opens into the
kidney, but near its outer aperture. In Area, Ostnea, Cydas, and Montacuta, the
kidney and the genital gland open on each side into the base of a common depres-
sion (urogenital cloaca) ; in all other bivalves the outer nephridial and genital aper-
tures are separate.

The simple structure of the Protobranchiate kidney becomes complicated in
other Lamellibranchia in the following manner :—

1 . That portion of the renal tube which opens outward forms an external cavity
(vestibular cavity, external sac) ; this cavity lias no excretory epithelium ; it
encircles the outer side of the pericardial portion of the kidney, the renal sac ( Fig.
184). The latter alone is developed as an excretory organ. Folds or trabecula?,
covered with glandular epithelium, project inward from its walls, forming a paren-
chymatous or spongy structure. The renal sac is connected with the pericardium
by means of a nephridial funnel of varying length.

•2. The two renal sacs communicate freely in the median plane. The connecting
part is widest in the most specialised bivalves (Pholadaera, Myacea, Anatinacea,
Septibmnchia).

In Anomia, where all the parts are asymmetrical, the two kidneys, which do
not communicate with one another, are also asymmetrical.

Venous blood flows through the kidneys on its way to the gills. The afferent
renal vessels seem to have walls of their own, but the efferent vessels appear to be
lacunar. Open communication between the blood vascular system and the kidneys
is nowhere found.

E. Cephalopoda.

(Of. Figs. 185, 186, and the sections on the ccelom and the blood
vascular system, pp. '213 and 208).

The Cephalopoda have two (Dibranchia) or four (Tetrabrancliia) spacious sym-
metrical renal sacs, in the posterior and upper part of the visceral dome. These
communicate in the typical way at the one end with the ccelom, and at the other
with the exterior (mantle cavity). Only one of the two pairs of kidneys in Nautilus,
however, possesses ccelomic funnels.

The large veins returning from the body to the heart run along the anterior wall
of the urinary sac. These veins bulge out into the cavity of the sac to form the
venous appendages already mentioned. The epithelium of the urinary sac which
covers these appendages is no doubt the seat of the excretory function. The excretory
matter is discharged into the urinary sac (the wall of which is otherwise smooth),
and passes out thence through a ureter of varying length into the mantle cavity.
The renal aperture is found on the median side of the base of the gill, and in Nautilus,
the Oegopsidce, and Scpioteutliis among the Myopsidce, it is simple and slit-like ; in the
other Myopsidae and in the Octopoda, however, it lies at the end of a renal papilla
which projects freely into the mantle cavity.

The two renal sacs in the Octopoda are entirely distinct. Near the point where
the renal sac passes into the ureter lies the renal funnel, which corresponds witli
the pericardial aperture of other Molluscs, and which here leads to the ccelomic
.cavity, now reduced to the "water vascular system."

In the Dccapoda, the two renal sacs communicate with one another in the median
plane. In Xcpia, there are two points of communication, one above and the other
.below. The lower junction is bulged out to form a large sac, which rises towards

VII

MOL L USC A — THE NEPHRIDIA

223

the apex of the visceral dome on the anterior side of the paired renal sacs (<•/'. Fig.
177, p. 213). The veins returning from the body to the heart run in the partition
between the unpaired anterior and the paired posterior sacs, and may here bulge out
to form venous appendages, not only posteriorly, i.e. into the cavities of the two
paired renal sacs, but also anteriorly, into that of the unpaired connecting sac. Xear

FIG. 185.— Renal sac, ccelom, genital organs, etc., of Sepia. A, female ; B, male. The
visceral dome is seen from behind ; the mantle, the body wall, the ink-bag, and in A the hind-gut
and the nidamental gland are removed (after Grobben). a, Heart ; li, genital vein ; c, genital
artery ; d, stomach ; e, female germinal body ; /, aperture of the oviduct in the ovarial cavity ;
rj, oviduct ; 7i, unpaired anterior renal sac ; i, abdominal vein ; t, appendage of the branchial
heart (pericardia! gland); I, branchial heart; m, paired posterior renal sac; n, gill; o, canals of
the c«;loiii leading to the kidneys ; j>, gland of the oviduct ; q, female genital aperture ; r, renal
aperture. In B, 1, testes ; 2 (the indicator points rather beyond the right, place), aperture of the
male germinal body into the genital cavity or capsule ; /, aperture' of the seminal duct into the
male genital capsule ; 3, section of the coalom containing the vas deferens (peritoneal sac) ; 5, anus ;
6, rectum ; q, male genital aperture.

the point where each renal sac is produced into the ureter, the reno-pericardial canal
springs from it, opening into the secondary body cavity which contains the heart,
and corresponds with the pericardium of other Molluscs.

The form of the renal sac is at least partly determined by the form and position
of the surrounding viscera, the stage of maturity of the genital organs, and the
different shape of these organs in the two sexes. All viscera which press against the
renal wall from without, bulging it inward, are naturally covered at the points of

224

COMPARATIVE ANATOMY

CHAP.

contact with the epithelium of the renal sacs. The same is the case with all organs
which, like the stomach, the gastric ccecum, and the efferent ducts of the digestive
glands in the Decapoda (Sepia), apparently lie inside the spacious renal sacs. These
organs really lie outside of them, being only suspended into them, like the intestine
of an Annelid, which apparently lies within the body cavity, but is entirely separated
from it by the peritoneal endothelium.

It has been already mentioned that only one of the two pairs of renal sacs of
Nautilus, viz. the upper pair, has reno-pericardial apertures. This fact was

arvo

FIG. 186.— Diagram showing the posterior paired renal sacs of Sepia officinalis, and the vein
running along its anterior wall with its venous appendages, from behind (after Vigelius). vc,
Vena cava ; r«.o, right nephridial aperture ; i/j, left reno-pericardial aperture, the outlines of the
secondary body cavity are indicated by a dotted line ; nj, vena genitalis ; rvc, right branch of the
vena cava ; fpil, right pallial vein ; m, right vena abdominalis ; rim, ruin nf the ink-bag ; vns, left
vena abdominalis ; cv, section of the secondary body cavity (capsule of the branchial heart), which
surrounds the branchial heart cl>, and the appendage of the same (pericardial gland) x; vps, left
pallial vein; Ire, left branch of the vena cava cephalica ; rm, left vena genitalis; rpr, secondary
body cavity (viscero-pericardial sac) ; y, left reno-pericardial aperture (renal funnel) (</. Fig. 174).

brought forward in support of the view that the two pairs of renal sacs arose by
the division of one single pair, corresponding with that of the Dibranchia. Accord-
ing to this view, the lower pair of gills, and the two auricles are also to be considered
to be new acquisitions. Indeed, the whole question of the original metamerism of the
Molluscan body, which has so often been asserted, rests on very weak foundations.
It gains no support from the Chitonidc?, where, in spite of large numbers of pairs of
gills, only two auricles occur, and where no relation exists between the number of the
shell plates and that of the gills.

vii MOLLUSCA— GENITAL ORGANS 225

XX. Genital Organs.

A. General.

In treating of the genital organs of the Mollusca, we shall have to
consider — (1) the gonads or germinal glands, those most important
organs, in which the reproductive cells (eggs and spermatozoa) are
formed; (2) the duets through which these cells reach the exterior;
and (3) the eopulatory organs.

1. The gonads or germinal glands have already, in Section XVIII. ,
been recognised as completely or incompletely demarcated portions
of the secondary body cavity, and have been described in their
relation to the other divisions of that cavity.

The gonads are paired and symmetrical in the Lamettibranchia and
Solenogastres, occurring in one pair. In all other Mollusca, only one
unpaired gonad is found. In very rare cases, such as that of some
hermaphrodite Lamellibranchs, which will be described later, there are
two pairs of gonads, one female and one male.

The sexes are- separate, among the Amphineura in the Chitonidce
and Chcetoderma, in many Lamellibranchs, in the Scaphopoda, among the
Gastropoda in the Prosobranchia (excepting a few Marseniadcc and the
Falvata), and in all Cephalopoda. Hermaphroditism prevails among
the Amphineura in Proneomenia, Neomenia, and allied forms ; in many
Lamellibranchs, among the Gastropoda in the Pulmonata, Opisthobranchia,
and in the Prosobranchiate family of the Marseniadce.

In hermaphrodite animals, it is the rule that the same gland, the
hermaphrodite gland, produces both eggs and spermatozoa, but in
exceptional cases there are in the same individual distinct male and
female gonads (testes and ovaries). This is the case, as already
mentioned, in certain bivalves, viz. the Anatinacea and the Septi-
branchia, which possess two testes and two ovaries.

Position of the gonads. — The long tubular hermaphrodite glands
of the Solenogastres, which are separated from one another by a median
septum, lie in the anterior prolongation of the pericardium, over the
intestine. In the Chitonidce, the gonads are found in a similar
position, but are not in open communication with the pericardium.
In the Gastropoda they lie in the visceral dome, usually in its upper-
most part, between the lobes of the digestive gland. Where the
visceral dome has disappeared, the gonad with the intestine and the
digestive gland shift back into the primary body cavity above the
foot. The gonads in the Scaphopoda occupy a position similar to that
of the Gastropodan gonads, lying dorsally in the high visceral dome,
above the anus and the kidneys. The same is the case in the
Cephalopoda. The paired much-lobed genital glands of the Lamelli-
branchia lie in the typical position in the primary body cavity, above

VOL. II Q

226 COMPARATIVE ANATOMY CHAP.

the muscular part of the foot, between the coils of the intestine.
They may lie behind the " liver," or else, passing between its lobes,
spread out at the sides of and below the kidney.

The epithelium which lines the gonads is, morphologically, the
endothelium of the secondary body cavity. The reproductive cells
may either be produced from any part of the epithelium of the gonad,
or from definite areas of this epithelium (Cephalopoda), which areas
may then be called germinal epithelium or germinal layers. It may
then appear as if the germinal gland lay in or on a special

sac.

whereas this sac is, in reality, the gonad itself, and the germinal
gland is only the much-developed germinal layer of the gonad.

The ripe reproductive cells become detached from their place of
formation, and fall into the cavity of the gonad, i.e. into a part of the
secondary body cavity, from which they pass out in various ways.

2. The gonads either have separate ducts (Chitonidce, Monotocardia,
Pulmonata, Opisthobranchia, Cephalopoda, many Lamellibranchia) or they
utilise the nephridia as ducts. In the latter case the genital products
either pass direct into the kidney, and reach the exterior through
the nephridial aperture (all Diotocardia, the Scaphopoda, and many
Lamellibranchia}, or they first pass into the pericardium, and then are
ejected through the nephridia (Solenogastres). Where the gonads
open into the kidneys, their apertures may lie in various parts of
these organs ; either in the proximal part, which communicates with
the pericardium by means of the renal funnel, and is usually widened
into the renal sac, or in the distal part (ureter) which opens externally,
or into a shallow urogenital cloaca.

The gonads therefore open into :—

a. The pericardium (Solenogastres).

b. The proximal or pericardial part of the kidney.
r. The distal part or ureter of the kidney.

(1. The urogenital cloaca. Or :—

e. They open externally, quite apart from the kidney.

Paired gonads have paired ducts (Solenogastres, Lamellibranchia).
Where there is a single unpaired gonad, there is either a single
efferent renal duct, or a single renal duct is made use of (Gastropoda,
Scaphopoda, Cephalopoda, etc.); the duct is then always asymmetrical
and usually lies on the right side. A paired duct, belonging to an
unpaired genital gland, is, however, found in the Chitonidce and in
many Cephalopoda.

When the genital glands have special efferent ducts, various
sections of the latter may be differentiated into accessory sacs and
glands, copulatory apparatus, etc., which, especially in the Pulmonata,
Opisthobranchia, and Cephalopoda, transform the ducts into a very
complicated apparatus. In males, this complication arises through
the development of copulatory organs, and of special glands which
form the capsules of the spermatophores, and of seminal vesicles, etc. ;
in females, through the development of albuminous glands, shell

VII

MOLL USCA— GENITAL ORGANS

227

glands, receptacula seminis, vagina, etc. Since, in hermaphrodite
Molluscs, both kinds of complication occur simultaneously in the
same genital apparatus, the most complicated arrangement is found
in the (hermaphrodite) Pulmonata and Opisthobranchia.

3. Copulatory organs are wanting in many Molluscs, such as the
Antphiitcuni (see below), nearly all Diotocardia, the Scaphopoda, and
all Lamellibranchia. They are present in the Monotocardia, the
Pul'tnonata, Opisthobranchia, and Cephalopoda. In the Gastropoda, in
the nuchal region, to the right, there is a male apparatus, consisting
sometimes of a freely projecting muscular penis, sometimes of an
organ which can be protruded or evaginated through the genital
aperture. In the Cephalopoda, this is a definite arm in the male,
which is specially modified (hectoeotilised), sometimes in a very
remarkable manner, and which plays a more or less important part in
copulation.

B. Special.

n. Gonads. (1) Amphineura. — The long hermaphrodite gland of Proncomenin
and allied forms has been called paired. As a matter of fact it is divided into two
more or less distinct lateral tubes, by a median much-folded septum. In the lower
portion of each tube, that

which lies next the intestine, % ,**• , ,1

the germinal epithelium pro-
duces spermatozoa, in the
upper portion eggs. Pos-
teriorly, these tubes sepa-
rate for a certain distance,
and open as a pair of dis-
tinct ducts into the anterior
end of the pericardium.

The male or female gonad
of the Cliitonidce lies as a
long unpaired sac on the
dorsal side of the intestine, in
front of and partly under the
pericardium. In the ovary,
numerous pear-shaped tubes
(Fig. 187) project from the
epithelial wall into the
cavity. Eacli of these tubes FIG. 1ST.— Section through the wall of the ovary of Chiton
is a stalked follicle, with egg (diagram after Haller). 1, Eggs at different stages of ilevelop-
cells surrounded by follicular ment ; 2> Serininal epithelium ; 3, egg sac or tubes ; 4, follicular
cells. These follicles are ePithelium '• 5> e^' tube after the discliarse "f the egg-
found in all sizes and at all stages of development. Each egg is at first a simple
ovarial epithelial cell, which is distinguished by its size from the surrounding
epithelial cells. As it grows and becomes more and more rich in yolk, it sinks
down under the ovarial epithelium, bulging out this latter towards the ovarial
cavity, and thus forming a young follicle. The wall of the pear-shaped testicle
also i-ises into its cavity in the form of numerous folds, in which the epithelium
becomes inultilaminar, and produces the mother cells of the spermatozoa.

The fact that the gonad of Chiton has two ducts makes it probable that it was

228 COMPARATIVE ANATOMY CHAP.

originally paired. The two ducts, i.e. the two seminal ducts in the male and the
two ovarial ducts in the female, open into the mantle furrow on each side, somewhat
in front of the reual aperture (Fig. 180, p. 217).

(2) Gastropoda. — The gonads of the Prosobranchia offer but few points of
interest to the comparative anatomist. In the Pulmo-nata and Opisthobranchia, the
germinal gland is a hermaphrodite gland, in which spermatozoa and eggs are
produced simultaneously. This gland is much lobed, or else consists of numerous
converging diverticula ; the spermatozoa and eggs arise intermingled on the walls,
become detached at one of the stages of their development, and then lie free in the
cavity of the gonad. The same applies to the large hermaphrodite gland of the
Tectibranchia, which varies much in its outer form. It lies in the posterior part of
the body, on the digestive gland, penetrating at times between its lobes ; it is itself
more or less lobed, its lobes consisting of secondary lobes (vesicles or acini). In
all these acini, spermatozoa and eggs are simultaneously produced. It is only in
the Pleurobranchcea and allied forms that the parts of the gland which produce
spermatozoa and those which produce eggs are localised ; this arrangement resembles
that in the Nudibranchia, which will presently be described. The constituent lobes
or vesicles are either male or female, the former producing only spermatozoa, the
latter only eggs. This is the arrangement found also in some Nudibranchia
(Am})horina, Capcllinia), but in most Nudibranclis the male and the female
germinal regions become separated in such a way that the terminal acini yield eggs
only, but open in groups into lobes of the gland which produce only spermatozoa.
Each lobe has its duct ; these ducts, uniting together, finally form the duct of the
hermaphrodite gland. This gland thus forms an extensive organ spread out in the
larger posterior part of the primary body cavity ; where there is a compact diges-
tive gland it covers this organ. Pltyllirhoe has 2 to 6 (usually 3) separate globular
acini whose long and thin ducts combine to form a hermaphrodite duct (Fig.
195, p. 238).

The hermaphrodite gland of the Pteropoda ( Tectibranchia natantia] always lies
in the upper (dorsal) portion of the visceral dome ; it is sometimes acinose and
sometimes consists of converging tubular follicles or of lamina} closely crowded
together. The eggs are always produced at the peripheral part of the acini, tubes,
or lamella-, while the spermatozoa arise in the central parts, near the ducts. These
two parts are generally separated by a membrane, which the eggs have to break
through to reach the hermaphrodite duct. The Pteropoda are protandrously
hermaphrodite, i.e. the spermatozoa are produced before the eggs, an arrangement
found in many hermaphrodite Molluscs.

(3) Scaphopoda. — The gonad (testis, ovary) in these animals is a long spacious
sac, provided with lateral diverticula ; it lies above the anus, rising high up into the
visceral dome along the posterior side of the body. In the Solenopoda (Siphono-
dentalium, etc.) a large part of the gonad stretches into the mantle. In young
animals, the gonad is closed on all sides, but in adults its wall appears to fuse with
the right kidney, and in the partition wall so formed an aperture arises which
establishes communication between the gonad and the right nephridium.

(4) Lamellibranchia. — The gonads are here found in the form of much-branched
tubular or lobate masses lying on each side in the primary body cavity, surrounding
and partly penetrating between the other internal organs. In some cases (Anomiidce,
Mytilidir), the gonad on each side stretches into the mantle. In others (Axinus,
Montacuta), it bulges out the body wall in such a way that branched outgrowths,
containing the germinal tubes, project from the body into the mantle cavity.

In most Lamellibrauchia the sexes are separate, but hermaphroditism sometimes
occurs. There are (1) whole groups of bivalves which are hermaphrodite ; e.g. the
most specialised forms, such as the Anatinacea and Scptibranchia ; (2) families

vii MOLLUSC A— GENITAL ORGANS 229

with a few hermaphrodite genera : Cyclns, Pisidium, Entovalva ; (3) genera (Ostrcea,
Pccten, Cardium) with a few hermaphrodite species ; (4) occasional cases of herma-
phroditism in species the sexes of which are usually separate : Anodonta. The
hermaphroditism of the Lamellibranchia is, however, always incomplete in the sense
that the spermatozoa and the eggs do not ripen simultaneously.

In the Anatinacea and Septibranehia, there are on each side entirely separate male
and female gonads, whereas all other hermaphrodite Lamellibranchs have a her-
maphrodite gland on each side.

(5) Cephalopoda. — The sexes are always separate in this class. It has already
been mentioned that the germinal sacs form a part of the secondary body cavity, witli
which they are in open communication.

One single unpaired gonad is always found, lying in the uppermost part of the
visceral dome. It is a variously-formed sac (peritoneal sac or genital capsule), lined
on all sides by an epithelium often to a great extent ciliated, which is in reality the
peritoneal epithelium of the secondary body cavity. The whole of the epithelium
covering the wall of the gouad is not, however, germinal, but only that on its anterior
side (that turned to the shell). The germinal layer here forms what may be called, in
the narrower sense, the ovary or the testis, which is then said to be contained in a
peritoneal sac or an ovarial or testicular capsule, or else to project into or be suspended
in such sac or capsule. The whole apparatus is really a gonad, in which the places
of formation of the reproductive cells are localised on the anterior wall.

From this it is clear why the testes and ovaries do not appear to possess efferent
ducts of their own, but to empty their products into their respective capsules, these
products passing out into the mantle cavity through the ducts of these capsules
(oviducts and seminal ducts). Since, however, the entire germinal sac corresponds
with the genital gland of a Gastropod or a Lamellibranch, the reproductive pro-
ducts in reality merely fall into the cavity of this gland (the testicular and ovarial
capsules), and pass out through the ovarial and seminal diicts, which exactly corre-
spond with the same ducts in the Gastropoda, Lamellibranchia, and Chitanidce.

The genital cavity has also another means of communication with the exterior,
since, in the Cephalopoda, it is in open communication with the remaining part of
the secondary body cavity, whether the latter forms a viscero-pericardial cavity (Deca-
poda) or is reduced to the "water canal system " (Octopoda). This latter part of the
body cavity again is connected, by means of the nephridia, with the mantle cavity.

In this way, the genital cavity communicates with the mantle cavity directly by
means of the oviduct or seminal duct, and indirectly through (1) the viscero-pericardial
cavity or the " water canal system," and (2) the nephridia. This second way of
communication, however, is never used for discharging the genital products.

The female germinal layer or ovarial layer (the ovary in the narrower sense) is
always found on the anterior wall of the gonad, and varies considerably in^structure (Fig.
188). We can always distinguish (1) the eggs, and (2) the ovigerous wall. The former
are stalked, and project from the wall into the cavity of the gonad (the cavity of the
ovarial capsule). The largest and oldest eggs are covered by a follicular epithelium,
and this latter by the general epithelium of the wall of the gonad, which also covers
the stalk. Eacli egg has a separate stalk. The youngest eggs are mere prominences
on the wall, which in the process of growth acquire a stalk, by means of which they
remain connected with the wall from which they project. This arrangement is
exactly like that in the Chiton. When the eggs are mature, the follicle bursts, they
fall into the genital cavity, and thence reach the exterior through the oviduct.

In Nautilus (Fig. 188, A) and Eledoiie the whole wall of the gonad, with the
exception of the posterior surface, can produce eggs ; these stand out from it all over
on simple stalks. In Argonauta (Fig. 188, B] and Trcmoctopus also, the whole ova-
rial capsule except the posterior wall produces eggs, but the egg-bearing region (to

230

COM I' A UA TI I 'E ANA TOMY

CHAP.

obtain increase of surface) projects into the genital cavity in the form of numerous
dendriform processes, the eggs being attached by simple stalks to the stems and
branches. In Parasira (Tremodopus) catenulata there is a central region containing
more than twenty large "egg trees " surrounded by a circle of smaller "trees." On
the anterior wall of the gonad in Octopus there is a single but very richly-branched
"egg tree" (0). In Sepia, Sepiula, and JKossia the egg-bearing surface bulges out
in the shape of a ridge on the anterior wall of the gonad. This ridge, in Loh'go.
becomes a narrow fold, the free edge of which is produced into filaments, which carry
on all sides simply -stalked eggs. In the Oegopsidcc (Ommastrephes, Fig. 188, D,
Onychoteuthis, Thysanoteuthis) the region which carries the eggs is only attached by
its upper and lower ends to the wall of the gonad, and forms an otherwise free spindle-
shaped body traversing the genital cavity, and beset all over with stalked eggs.

In Octopus and Eledone all the eggs in a given ovary are found at the same
stage of maturity.

A peculiar transformation of the follicular epithelium takes place in the ovarial
eggs of the Cephalopoda when nearly mature. An extraordinary increase of surface
occurs in the shape of numerous folds, which run longitudinally along the egg, either
reticulating or remaining parallel to one another, and projecting far into the yolk

FIG. ISS.—A-D, Four diagrams of the female gonads of the Cephalopoda. .4, Nautilus type.
B, Argonaut type. C, Octopus type. D, Ommastrephes type. 1, Aperture of the oviduct into
tlie gonad ; •_', cavity of the gonad (a section of the secondary body cavity) ; 3, egg-carrier.

of the egg which they surround. This arrangement may be connected with the
nutrition of the egg.

The male germinal layer (germinal body, or testis in the narrower sense) is a
variously-shaped (often globular or oviform) compact organ, which usually lies free
in the genital cavity, suspended to its anterior wall by a thin ligament (mesorchium)
in which the genital artery runs. The germinal body is everywhere covered with
epithelium, which is continued over the mesorchium into the epithelium of the wall
of the gonad (endothelium of the testicular capsule). On the surface of the germinal
body which is turned away from the mesorchium, there is a funnel-shaped depression
(Fig. 189, A) ; towards this, from all sides, the tubular testicular canals which form
the male germinal body converge, in order to open into it. In these testicular canals,
between which there is a slight framework of connective tissue, the spermatozoa are
produced, and are passed on to the genital cavity through the depression into which
all the canals open ; they reach the exterior by means of the seminal duct. The
testicular canals originally possess a multilaminar germinal epithelium, which yields
the spermatozoa, and which passes at the common aperture into the outer epithelium
of the germinal body, and so into the epithelium of the germinal sac.

This description applies to the male germinal body of most Cephalopoda. In

VII

MOLLUSCA— GENITAL ORGANS

Loliijo (B), however, the funnel-shaped depression into which all the testicular
canals open is replaced by a longitudinal furrow, into which these converging canals
open. In Sepia (C), the germinal body lias no ligament, but lies immediately in
front of the anterior wall of the gonad, and is thus outside the genital cavity. The
germinal body here has a central channel towards which the radially arranged
seminal canals converge from all sides, and which they enter. This channel, again,
opens through an efferent duct into the genital cavity, from which the spermatozoa
are conducted to the exterior by the seminal duct.

The spermatozoa of the Mollusca are of the common pin shape. In many species
of Prosobranchia two different forms of spermatozoa, the hair-shaped and the vermi-
form, occur in one and the same individual. This phenomenon has by some been
taken as an indication of developing hermaphroditism, and by others as pointing to
a former hermaphrodite condition ; in the first case the vermiform spermatozoa
would be the eggs beginning to form, in the second the rudiments of eggs. There
is, however, no solid foundation for either of these views.

With regard to the question whether the hermaphrodite or the dioecious condition
is the original condition, the latter alternative may be considered as the more prob-
able. Of the five classes of the Mollusca, two, the Scaphopod,a and the Cephalopoda,

Fio. 189.— A, H, C, Three diagrams of the male gonads of the Cephalopoda. .1, ordinary
type. /;, Loligo. <', Sepia. 1, Seminal duct ; -2, cavity of the gonad ; 3, space into which all the
canals of the testis open, and which itself opens into the cavity of the gonad, in .Sc/nc, by means
of a canal (4); 5, suspensor of the male germinal body, attaching it to the'anterior wall of the
gonad.

are altogether dioecious. Among the Amphineura, the Chitonidcn, which most
recent observers hold to be less specialised than the Solenogastres, are sexually
separate. Among the LamellibrancMa, the sexes are separate in the Protobranchia,
which are rightly considered as primitive forms ; and most other bivalves are also
dioecious. Among the Gastropoda, the sexes are separate in the Prosobranchia,
especially in the Diotocardia, which are universally considered to be the lowest and
least specialised Gastropods.

b. The ducts. — The manner in which the sexual products are conducted to the
exterior in the Amphineura, tica.phojwda, and LameUibranchia need not again be
discussed, as it has already been described in the general part of this section, and in
the section on the nephridial system. We thus have now only to treat of the very
complicated ducts of the Gastropoda and the Cephalopoda.

(1) Gastropoda. — It has been seen that in all Diotocardia (Haliotis, Fissurclla,
Patella, etc.) the genital products are ejected through the right kidney. In
the Monotocardia, the right kidney has atrophied as such, but, according to the
most recent investigations, its duct persists as genital duct. In the Pulinonata
and Opisthobranchia, the genital aperture is no longer in the mantle cavity,

232 COMPARATIVE ANATOMY CHAP.

but has shifted far forward along the right side of the neck, probably in con-
nection with the development of the copulatory apparatus. The position of this
aperture is thus not necessarily affected by any further displacement of the pallial
complex, or indeed of the whole visceral dome, which explains the fact that, in Daude-
bardia and Testacella, the common genital aperture, and in Oncidium, the male
aperture, lies far forward on the right side of the body, although the pallial complex
lias shifted completely to the posterior end of the body.

In the Opisthohrandiia also, the single or (secondarily) double genital aperture
lies to the right in front of the anus and even in front of the kidney. This position
seems inexplicable except by the supposition of a shifting back of the pallial com-
plex in which the genital aperture, emancipated from the complex, took no part,
thus coming to lie in front of the shifted anal and renal apertures.

Monotocardia. — Unlike the Diotocardia, which, with the exception of the Neritidce,
have no copulatory organs, the Monotocardia possess a penis, which, however, does
not lie in the mantle cavity where the genital aperture originally lay. It would be
unable to function in this position, and is therefore placed on the right side of the
head or neck (Fig 71, p. 73), and forms a freely projecting, extensible, muscular
appendage, which often attains a considerable size. The male genital aperture,
however, in very many, perhaps in most, Monotocardia, remains in its original posi-
tion in the mantle cavity, to the right, near the rectum. In such cases, a ciliated
furrow runs forward on the floor of the respiratory cavity, along the right side of the
neck, to the base of the penis, to the tip of which it is continued as a deep groove.
This furrow conducts the semen to the penis from the genital aperture. In some
cases the furrow closes, and forms a canal ; the penis then becomes tubular, and the
seminal duct enters into it. The genital aperture is thus shifted far forward from
its original position. The seminal duct, which arises from the testis, usually forms
coils as it runs along the columellar side of the shell. The vas deferens has no
special appendages, although it may widen into a vesicle at some point in its course.

In the female, the genital aperture remains in the mantle cavity, lying to the
right near the rectum, behind the anus. The duct remains, as a rule, more or less
simple ; it is divided into the following consecutive sections: (1) an oviduct, rising
from the ovary, which may bulge out to form one or more receptacula seminis ;
(2) the uterus, a wider section with thick glandular walls, in which the eggs arc
provided with albumen and a shell : (3) a muscular sheath, the vagina, which leads
to the outer genital aperture. In Pahidina, there is a special albuminous gland
opening into the oviduct.

In hermaphrodite Prosobranchia (Valvata, a few Marscniadce, e.g. Marsenina,
Onchidiopsis) a hermaphrodite gland is found. This gland gives rise either to one
duct, which divides later into a vas deferens and an oviduct, or to a vas deferens and
an oviduct which are from the first distinct. The vas deferens runs to the penis as
in the males of dioecious Prosobranchiates ; the oviduct runs to the female genital
aperture. Both these ducts are, -owing to the occurrence of accessory glands, etc.,
more complicated than in other Prosobranchiates.

Opisthobranchia and Pulmonata. — The ducts in these orders are extremely
complicated, both by division into many consecutive sections and by the develop-
ment of various accessory organs.

In the following descriptions of several types of genital ducts only the most
important points can be mentioned. We give first the type of duct commonly
found in the Cephalaspidce (Tectibranchia).

1st Type. — The hermaphrodite gland has a single undivided efferent duct,
opening out through a single genital aperture. From this aperture the fertilised
eggs pass out direct, but the spermatozoa pass into a ciliated seminal furrow which
runs along in the mantle cavity, and by which they are conducted to the penis.

VII

MOLLUSC A— GENITAL ORGANS

233

This lies more or less far forward in front of the genital aperture, near the right
tentacle.

If we imagine the testis of a male Monotocardian transformed into a herma-
phrodite gland, and the vas deferens into a hermaphrodite duct, the above condition
would be realised.

Gastropteron may be chosen as a good example of this arrangement (Fig. 190),
which is further found in other
Cephalaspidce (Doridium, Pkiline,
Scaphander, Bulla) and all Ptcro-
poda.

The hermaphrodite gland or ovo-
testis, which lies between the lobes
of the liver in the posterior part of
the body, gives rise to a herma-
phrodite duct, which, after a long
coiled course, enters a short but
much widened terminal section
known as the uterus or genital
cloaca. This cloaca opens outward
in front of the base of the gills
through the genital aperture. Into
the cloaca open : (1) the common
efferent duct of two glands, one
of which, the albuminous gland,
supplies the egg with albumen,
while the other, the nidamental or
shell gland, yields its outer pro-
tective envelope ; (2) the duct of
a globular vesicle (receptaculum
seminis, Schwammerdam's vesicle),
which receives the spermatozoa dur-
ing copulation. From the genital
aperture, which has a more or less
median position on the right side of
the body, the seminal furrow runs
forward to the penis. The latter is
enclosed in a special sheath, out of
which it can be protruded, and into FIG. 190.— Genital organs of Gastropteron Meckelii
which it is withdrawn by means of (after Vayssiere). The penis and the seminal furrow

a retractor muscle. A gland called a,re not dra™' V Coml"°n genital aperture ; o genital

cloaca ; 3, albuminous gland ; 4, nidamental gland ; 5,

3 prostata opens into the penis, hermaphrodite duct; 6, hermaphrodite gland; 7, re-
The penis itself lies on the right ceptaculum seminis.
anteriorly, 011 the boundary between

the head and the foot. When it is at rest its sheath lies in the cephalic cavity,
near the buccal mass.

The very complicated ducts of Aplysia and Accra do not essentially differ from
that above described. The hermaphrodite duct, on reaching the region of the
albuminous gland, coils back upon itself, the ascending and descending portions of
this coil surrounding the albumen gland with their spiral coils. The penis has no
prostata.

2nd Type. —The hermaphrodite gland gives rise to a hermaphrodite duct, which
soon divides into two parts, the vas deferens or seminal duct, and the oviduct. The
former runs to the male copulatory apparatus, the latter to the female genital

234 COMPARATIVE ANATOMY CHAP.

aperture. The male aperture and the penis lie in front of the female, far forward on
the head or neck ; the two apertures are quite distinct, and both lie on the right.

This second type may be deduced from the first, if we assume not only that the
common duct of the hermaphrodite gland divided into a male and a female duct,
but also that the seminal furrow closed to form a canal in continuation of tin-
male duct.

When the duct of this second type split into a male and a female duct, the
accessory organs also so divided that the male opened into the vas deferens, the
female into the oviduct.

To this type belong, among the Pulmonata, the JBasommatophora, a few species
of Daudcbardia (D. Saulcyi, in which the two apertures lie close together), the
Oncidia, and Taginulidce. In both these latter groups, the female aperture has
followed that part of the pallial complex which shifted to the posterior end of the
body, and lies near the anus. The male aperture has, however, retained its anterior
position on the head, behind the right cephalic tentacle. The two apertures thus
lie at the opposite ends of the body. Among the Opisthobranchia, this second type
is exemplified in Oscinius (Tectibranchia).

Taking Limnsea stagnalis and Oncidium as examples, we find in the former
(Fig. 191) that the hermaphrodite gland which lies embedded in the "liver," high
up in the visceral dome, gives rise to a thin hermaphrodite duct ; this soon divides
into a male and a female duct. The male duct first widens into a flattened sac,
then into a large pear-shaped glandular vesicle (prostata). From this vesicle it runs
as a long thin vas deferens through part of the pedal musculature, and finally enters
the male copulatory apparatus, which is, in fact, merely the widened muscular and
protrusible end of the vas deferens. A small penis tube is first formed by the vas
deferens, and this projects on a papilla into a subsequent larger tube (the penis
sheath), which is evaginated during copulation. Protractors are attached to the
sheath, and retractors to the small tube ; the latter alone with its papilla enters the
vulva during copulation.

An albuminous gland opens into the female duct immediately after its separation
from the male duct. It then forms a uterus consisting of wavy folds, and is continued
into a large pear-shaped body as oviduct, the narrow end of which is the vagina and
leads to the female genital aperture. The oviduct receives a lateral accessory gland
called the nidamental gland, and the vagina the efferent duct of the globular
receptaculum seminis.

In Oncidium celticum (Fig. 192) the hermaphrodite gland and female accessory
glands lie in the posterior part of the body, between the lobes of the liver and the
coils of the intestine. From the gland rises a hermaphrodite duct, which at one
point carries a small lateral caecum, and opens into an irregularly-shaped organ, the
uterus. Within the uterus two projecting folds border a channel ; if these folds
become apposed, the channel becomes a tube. This channel runs from the point of
entrance of the hermaphrodite duct to the point where the seminal duct leaves the
uterus, and serves for conducting the semen. The remaining wider portion of the
uterus serves as oviduct and egg-reservoir, and carries a large csecal appendage ;
the ducts of the two much-lobed albuminous glands also enter the uterus.

A comparison of Limncm and Oncidium shows that in the latter the male and
female ducts separate from one another further back than in the former. The vas
deferens in <><i<-i<linin is only incompletely separated as a groove in the uterus. Its
differentiation into a separate duct takes place here, as in terrestrial Pulmonates, at
the distal end of the uterus. The thin seminal duct (vas deferens) passes into the
body wall to the right, and runs forward along the longitudinal furrow between the
foot and back, passing again at the anterior end of the body into the primary body
cavity, where it forms numerous coils, and finally enters the copulatory apparatus.

VII

MOLL USC A— GENITAL ORGAN*

235

This apparatus, in Limncea, consists of a large evaginable terminal widening, into
which the vas deferens projects in the form of a papilla. Blood pressure causes the
penis sheath or prseputium to l>e evaginated through the genital aperture, into which

FIG. 101.

13 UN-

1lv

FlO.

FIG. 191.— Genital organs of Limnasa stagnalis (after Baudelot). 1, Male genital aperture ;
2, larger penis tube (penis sheath) ; 3, protractors ; 4, smaller penis tube ; 5, vas deferens ; 0, pro-
stata ; 7, flattened widening of the vas deferens; 8, hermaphrodite duct ; 0, hermaphrodite gland ;
10, part of the digestive gland (liver) ; 11, albuminous gland ; 12, nidamental gland ; 13, uterus ;
14, pear-shaped body ; 15, receptaculum seminis ; Hi, vagina ; 17, female genital aperture.

FIG. 192. — Genital organs of Oncidium celticum (combined from the figures of Joyeux-
LafTuie), somewhat diagrammatic ; only part of the vas deferens is drawn. 1, Male genital aperture ;
2, penis sheath (prjfputium) ; 3, penis papilla ; 4, vas deferens ; 5, uterus, the seminal furrow in.
the uterus is indicated by dotted lines ; 6. civcum of the uterus ; 7, oviduct and vagina ; 8, csecal
appendage ; 9, receptaculum seminis ; 10, female genital aperture ; 11, albuminous glands ; 12, caecum
of the hermaphrodite duct, 13 ; 14, hermaphrodite gland.

it is again withdrawn by means of a retractor. In other species of Oncidium, the
copulatory apparatus is complicated by the occurrence of accessory penis glands and
variously-shaped cartilaginous armature.

236

COMPARATIVE ANATOMY

CHAP.

The oviduct which separates from the vas deferens at the end of the uterus is also
a vagina. It is a simple tube which opens outward to the right near the anus
through the genital aperture. Near the middle of its course it is joined by the
stalk-like duct of a globular vesicle, the receptaculum seminis (bursa copulatrix), and
by a long glandular cascal appendage.

3rd Type. — We find this in the Stylommatophora among the Pulmonala, and also

Ps

3D.

PIG. 193. —Anatomy of Helix pomatia (after Leuckart, Wandtafeln). The shell is removed
and the mantle hud back to the left, the organs of tin* visceral dome and head are isolated and
separated. To the left (in the figure) are the genital organs. L, Digestive gland (liver) ; Zd, her-
maphrodite gland; ./, intestine; N, kidney; V, ventricle; M, fore-stomach; F, foot; A, anus;
Al, edge of the mantle near the respiratory aperture ; Mr, retractor muscle ; G, cerebral ganglion ;
Fl, flagellum ; Sk, cesophageal bulb (pharynx) ; P, penis ; R, retractor of the tentacle ; Ps, dart sac ;
AD, digitate glands ; Vd, vas deferens ; X, lateral bulging of the stalk of the receptaculum seminis
(Rs); Od, portion of the uterus belonging to the oviduct ; Eil, albuminous gland ; Zg, hermaphrodite
duct.

in all Nudibrancihia and a few Tcdibranchia (e.g. Pleurobranchcea). The herma-
phrodite gland gives rise to a hermaphrodite duct, which, as in the second type,
sooner or later divides into a male and a female duct. These, however, do not open
out through distinct apertures, but again unite to form a common atrium genitale or
a genital cloaca. This third type may be deduced from the second by suppos-
ing that the male and female apertures became approximated, and finally opened
together.

VII

MOLLUSCA— GENITAL ORGANS

237

Helix pomatia and Pleurolranchcea Meckelii afford good examples of this
arrangement.

Helix pomatia (Fig. 193). — From the hermaphrodite gland a hermaphrodite
duct, in zigzag coils, passes into the long folded uterus. The straight band which
passes along the folds of the uterus is that portion of it which belongs to the seminal
duct ; the folds belonging to the female ducts. The seminal channel, however, is
merely a furrow within the uterus, divided from the cavity of the latter by two
projecting folds, the edges of which become superimposed. A longitudinal glandular
band, which is regarded as a prostata, accompanies this duct. At the point where
the hermaphrodite duct passes into the uterus, the large linguiform albuminous
gland opens into it. At the end of the uterus, the male and female ducts become
entirely distinct. The thin vas deferens runs in coils to the copulatory apparatus,
which again opens into the genital cloaca. The copulatory apparatus consists of a
protrusible penis ; at the point where the vas deferens enters this organ, the latter
carries a long hollow appendage, the flagellum, the glandular epithelium of which
perhaps yields the substance of the spermatophoral capsules. At the same point a
retractor muscle is attached to the penis. The
short oviduct widens before opening into the
genital cloaca. The widened portion has the
following appendages: (1) a long stalked pear-
shaped receptaculum seminis, lying close to the
uterus, — the stalk has a lateral bulging, which
is sometimes rudimentary ; (2) two tassel-shaped
organs, the digitate glands, the milky secretion
of which contains calcareous concretions, and
no doubt assists in the formation of the outer
envelope of the egg; (3) the dart sac, which
lies close to the cloaca, and contains a pointed
calcareous rod, the spiculum amoris, which is
thrust by each individual into the tissue of the
other as an excitant during copulation.

The common outer genital aperture lies in
the nuchal region behind the right optic-
tentacle.

Pleurobranchsea Meckelii (Fig. 194). — The
hermaphrodite duct, which rises from the gland,
forms a long ampulla or widening, and then
divides into a male and a female duct. The
vas deferens runs in coils to the penis sheath,
which it enters, coiling up in it almost like a
watch-spring, and then forms the evaginable
widened end portion which is called the penis,
and which can be invaginated by a retractor l, Common genital aperture ; 2, penis

muscle. The oviduct has a shorter course, and sheath : 3> Penis: 4> "tractor muscle of

_. , ,. iii the same ; 5, vas deferens ; 6, nidamental

receives the short efferent duct of a globular gland . T> albuminous gland. s> genital

receptaculum seminis. The widened terminal cioaca ; 9, oviduct ; 10, receptaculum
portion of the oviduct (the vagina), which enters seminis ; 11, widening and csecal appen-

the genital cloaca with the penis, receives the tlase of the oviduct = r2> hermaphrodite

,, i -j 0.111 duct; 13. hermaphrodite gland,

ducts of the albuminous and nidamental glands

(shell and slime glands) ; the second of these may be regarded as the homologue of
the digitate gland of Helix.

There is a general agreement between the ducts of the Nudibranchia and those
just described ; in details, however, extraordinary variety prevails. The male and

FIG. 194.— Genital organs of Pleuro-
branchaea Meckelii (after Mazzarelli).

238

COMPARATIVE ANATOMY

CHAP.

female ducts nearly always unite in the base of a genital cloaca, which often lies
anteriorly on the right, on a papilla. The male and female apertures are rarely
separate ; when they are so, they lie close together (cf. Fig. 195 of Phyllirhoe).
The penis is often armed in various ways.

The important subject of the mutual relations of the three types of genital ducts
in hermaphrodite Gastropoda has been much discussed, but no satisfactory con-
clusion has been reached. Ontogenetic research has been appealed to so far in vain.

It is thus not at present known whether the
^ single hermaphrodite duct has arisen by the

fusing of separate male and female ducts, or
whether the separate ducts have come into exist-
ence by the splitting of an originally single
hermaphrodite duct. The difficulty is increased
by the fact that the genetic significance of the
hermaphrodite gland is uncertain.

Fertilisation is mutual in hermaphrodite
Gastropods. It is, however, certain that, in the
Pulmonata at least, when copulation does not
take place, self- fertilisation can occur. The
hermaphrodite, duct not infrequently carries one
or two lateral creca or vesiculse seminales, in
which an animal can store up its own sperm to
be used in fertilising its own eggs if cross-fertil-
isation does not take place. The eggs and the
.sperm are often not ripe at the same time.

(2) Cephalopoda. — Although the gonad in all
extant Cephalopoda is unpaired, the ducts are
originally paired in both sexes. In Nautilus,
the Oegopsidcv, and the Odopoda, there is one
pair of ducts in the female ; but in the males a
paired seminal duct occurs only in Nautilus and
Pliilonexis carcnce (Tretiwctopus). In Nautilus,
in which both sexes possess paired ducts, the
left duct is in both cases rudimentary and no longer functions. It is the so-called
pear-shaped vesicle, which is attached on one side to the heart and the lower end
of the gonad, and on the other opens into the mantle cavity at the base of the
lower gills.

Where only one duct is retained, it is, in both sexes, the one on the left, as in
Loligo, Sepia, Scpiola, Ilossia, ScpiotnitJiis, C'hiroteuthis, Cirrhoteuthis, etc.

The genital ducts rise on the wall of that part of the secondary body cavity which
is known as the genital cavity (peritoneal sac, genital capsule), and open into the
mantle cavity at the sides of the anus, between the nephridial aperture and the base
of the gills.

Male ducts, seminal duct. — In the more complicated form of male duct,
such as that of Sepia (Fig. 196), four principal divisions may be distinguished.
From the testicular capsule rises a vas deferens, which runs along in close coils, and
then widens into a vesicula seminalis, the highly developed and much folded epi-
thelium of which plays an important part in the formation of the spermatophores.
The vesicula seminalis is continued as a thin vas deferens to the last division, the
spermatophoral pouch (Xeedham's pouch), which serves as a reservoir for the

FIG. 195.—- Genital organs of Phylli-
rhoe (ai'tiT Souleyet). 1, Vas deferens ;
•2, penis ; 3, oviduct ; 4, male, 5, female
genital aperture ; 0, vagina ; 7, herma-
phrodite gland ; 8, hermaphrodite duct ;
0, receptaculum seminis.

VII

MOLL USC A— GE NIT A L ORGA

239

spermatophores. This pouch is flask-shaped and projects freely, with the end which
corresponds to the neck of the flask, at which the male genital aperture lies, into
the mantle cavity. The vas efferens receives (1) the short duct of an oviform gland,
the prostata, and (2) a simple, lateral, non-glandular caecum. The prostata takes
part, like the vesicula seminalis, in the formation of the spermatophores. The
prostata, caecum, and vesicula seminalis, in their natural position, form a coil,

ll

Fin. 196.

FIG. 190.— Male genital organs of Sepia officinalis. 1, Genital aperture ; -2, spermatophoral
pouch ; 3, vas efferens ; 4. cwcum ; 5, prostata ; (J, canalicule, opening into that part of the body
cavity which surrounds the male duct ; T, vesicula seminalis ; S, 9, vas deferens ; 10, gonad, a
portion of the posterior wall is removed, the genital cavity is revealed, and on its anterior wall is
seen the aperture of the male germinal body (12) ; 11, aperture of the seminal duct into the genital
cavity.

Fio. 197.— Male genital organs of Octopus vulgaris (after Cuvier). 1, Penis ; •_', muscle, cut
through ; 3, spermatophoral pouch ; 4, vesicula seminalis ; 5, prostata ; (3, vas deferens ; 7, opened
genital cavity, on whose anterior wall the testicular canals of the germinal body (8) are seen ;
9, aperture of the seminal duct into the genital sac.

which lies in a special division of the secondary body cavity, the peritoneal sac. It
is remarkable that the vas deferens is in open communication with this peritoneal
sac by means of a narrow tube.

The male efferent apparatus of Octopus (Fig. 197), as compared with that of
Sepia, is distinguished chiefly by the absence of a separate vas en'erens. The long

240

COMPARATIVE ANATOMY

CHAP.

vesicula seminalis opens into the large prostata near the point where the latter enters
the spermatophoral pouch. This point lies, not in the base, but in the neck of the
pouch, where the latter is produced into the long fleshy penis, the point of
which projects into the mantle cavity. The penis is provided with a lateral
csecum.

It has already been mentioned that, as far as we know at present, only two living

ventral.

SfT.

dorsal.

Fio." 198.— Female genital organs ol Sepia officinalis (chiefly after Brock). The mantle
cavity is opened, the posterior integument of the visceral dome removed, the ink-bag laid some-
what to one side, and the oviduct uncovered. The complex of organs thus exposed is seen from
behind. 1, Funnel ; 2, edge of the aperture of the funnel ; 3, cartilaginous locking apparatus ;
4, left ganglion stellare ; 5, glandular terminal portion of the oviduct with the female genital
aperture ; 6, left lateral lobe of the accessory nidamental gland ; 7, gland of the oviduct ; S, left
gill ; 9, oviduct filled with eggs which are seen through its wall ; 10, left nidamental gland ; 11,
mantle ; 12, ovarial sac, opened from behind, the stalked ovarial eggs are seen on its anterior wall ;
13, ink-bag (pigment gland) ; 14, stomach ; 15, right nidamental gland ; 10, central portion of the
accessory nidamental gland ; 17, right lateral lobe of the same ; 18, right gill ; 19, right renal
aperture ; 20, anus.

Cephalopods, Nautilus and Philonexis carcnce, have paired male ducts. In
the left duct is rudimentary. Whether the two ducts of Philonexis carence correspond
with the two ducts which we may assume that the Cephalopoda originally possessed
is very doubtful. The two vasa deferentia of Philonexis, which arise out of the
testicular capsule, and differ considerably in structure, unite together later, and

vii MOLLUSC A— GENITAL ORGANS 241

both lie on the left side. It is also remarkable that the spermatophoral pouch has
two apertures, and that there are thus two genital apertures.

Female genital organs— Sepia (Fig. 198). — The complicated female efferent
apparatus consists of two entirely distinct parts, opening separately into the mantle
cavity: (1) an unpaired oviduct (to the left), the position and aperture of which
correspond with those of the seminal duct in the male ; and (2) the nidamental
glands. The two large nidamental glands are pear-shaped organs, lying just
beneath the integument in the posterior part of the visceral dome, symmetrically, at
the sides of and anterior to the descending efferent duct of the ink-bag. They open
into the mantle cavity at their ventral ends. Each gland appears symmetrically
divided by a series of glandular lamellaj, traversing it from side to side. The spaces
between the lamella open into the central slit-like duct ; this structure is to be seen
even on the exterior of the gland. Besides these two nidamental glands there is an
accessory nidamental gland lying below and in front of the former. It is of a
brick-red colour, and consists of a central part and two lateral lobes. It consists of
numerous coiled glandular canalicules, which open into a glandular area in the
mantle cavity. This glandular area forms a depression between the central and
lateral lobes. As the aperture of the large nidamental gland also lies in this
depression, the secretions of the two glands here mingle.

The oviduct which rises from the ovarial sac is, during the reproductive season,
so full of eggs, that it becomes much distended, especially at the part which opens
into the ovarial sac. Before this duct opens outward into the mantle cavity at the
same point and in a similar manner as the seminal duct in the male, it becomes con-
nected by means of a freely projecting portion with a doubly-lobed or heart-shaped
accessory gland, the gland of the oviduct, which repeats the structure of the nida-
mental gland. The terminal portion also (from the point of entrance of this gland
to the aperture of the oviduct) is glandular, two symmetrical rows of perpendicular
glandular leaflets projecting from its wall into its lumen.

The secretions of the nidamental glands, accessory nidamental glands, and the
glands of the oviducts yield the outer envelopes of the ovarial eggs.

Nidamental glands occur, among the Cephalopoda, (1) in the Tetrabranchia
(Nautilus) ; (2) in the Dibranchia, among the Decapoda, in the Myopsidce (Sepia,
Scpiola, fiossia, Loligo, Sepioteuthis, etc.); in a few Oegopsidec (Ommastrcphes,
Onycotcuthis, Thysanoteuthis). They are wanting in the Octopoda and in some
Oegopsidce (Enoploteuthis, Chiroteuthis, Owenia).

Nautilus is distinguished from all other living Cephalopoda (1) by the possession
of only one nidamental gland, and (2) by the fact that this gland does not lie in the
visceral dome but in the mantle.

Accessory nidameutal glands are found only in the Myopsidcs. The two glands
are either separate (Rossia, Loligo, Sepioteuthis) or fused together (Sepia, Scpiola).

Glands of the oviduct occur in all Cephalopoda, but vary in position and in
structure.

Outgrowths of the oviduct, which function as receptacula seminis, occasionally
occur (Tremoctopus, Parasira).

In all Cephalopoda, certain quantities of spermatozoa are collected in extremely
complicated envelopes, the spermatophores. The substance of these large fila-
mentous spermatophores is yielded by the prostata and the vesicula seminalis, but
the mechanism by which so complicated a case is produced is still unknown.
When touched, or when they reach water, the spermatophores burst at definite
points, and scatter their contents. At the reproductive season the spermatophoral
pouch is entirely filled with spermatophores. In Philonexis carence, however, only
one very long sperniatophore is produced.

VOL. II R

242

COMPARATIVE ANATOMY

CHAP.

c. The copulatory apparatus — Hectocotylisation in the Cephalopoda. — The

copulatory apparatus of the Gastropoda, and the penis which projects into the
mantle cavity in certain Cephalopoda have already been described.

One of the most remarkable and enigmatical phenomena in connection with
the Cephalopoda is their hectocotylisation. This consists in the transformation of
one of the oral arms of the male into a copulatory organ and
spermatophore-carrier. This arm is said to be hectocotylised ;
during copulation it becomes detached, and finds its way into
the mantle cavity of the female.

Typical hectocotylisation (Fig. 200) is found only in the
Oetopodan genera Argonauta, Philonexis, and Trcmodopus. In
Tremoetopus and Philoncxis (Parasira) the third arm on the
right is the one transformed, in Argonauta the third on the left.
The arm is at first enclosed in an outwardly pigmented sac (Fig.
200 A), when this bursts, the arm becomes free, and then its
special form can be recognised (B). The folds which formed
the sac bend back so as to form a new sac, which receives the
spermatophores and is now inwardly pigmented. An aperture
leads from this sac into a seminal vesicle inside the hectocotylised
arm ; this vesicle is continued into a long thin efferent duct,
which runs the whole length of the arm and opens outwardly
at its end. The end of the arm is transformed into a long
filamentous penis, which at first is also enclosed in a special sac.
When the penis is evaginated the sac remains as an appendage
at its base.

The spermatophores then pass from the pigmented sac into
the seminal vesicle, and are ejected through the efferent duct
which opens at the tip of the penis.

It is probable that Cephalopods grasp one another, during
copulation, with their arms, in such a way that their mouths face
each other. In this position the hectocotylised arm of the male
becomes detached, and in some way or other forces its way
into the mantle cavity of the female. Detached arms are often
found in the mantle cavity of the female, as many as four have
been found at one time.

We still do not know (1) how the hectocotylised arm fertilises
the eggs of the female, or (2) how the spermatophores reach the
hectocotylised arm.

The males and females in the above-mentioned genera differ
from one another, apart from the sexual dimorphism caused by
the development of the hectocotylised arm. The males are
much smaller, and in Argonauta the female only has a shell.

It is very probable that the detached hectocotylised arm
can be replaced by a new one.

Although a true hectocotylised arm, which can be detached,
is only developed in the three genera above mentioned, it has
been proved that in all other Cephalopoda (even Nautilus, cf. p. 117), a certain arm
or portion of the head in the male is in some way modified, differing in some
(often unimportant) manner from the other arms. Such an arm is said to be
hectocotylised, and it is assumed that it plays some part in copulation, although
its exact function is unknown. In Sepia and Nautilus it is even difficult to
imagine what part it can take in copulation. The constant occurrence of

remarkable as

Fio. 199.— Spermato-
phore of Sepia (after
Milne Edwards), c,
Outer case ; 6, inner
case ; c, spermatozoal
sac ; d, e, f, g, various
parts of the ejacula-
tory apparatus.

hectocotylised arm is the more

it is by no means always

a
the

VII

MOLLUSC A— GENITAL ORGANS

243

same arm that is thus transformed. In the Octopoda, as a rule, it is the third
on the right side, but in the Octopodan subgenus Scceiirgus and in Argonauta

FIG. 200.— Male of Argonauta argo (after H. Miiller). (Female, Figs. 35, 3(3, pp. 24, 25.) A,
With the hectocotylised arm enclosed in the sac (d). B, with the arm free, a, Funnel ; b, edge of
the mantle fold ; c, left eye ; d, sac ; dlt hectocotylised arm ; c, mouth.

it is the third on the left. In the Decapoda the hectocotylised arm is generally
the fourth on the left, but in the genus Enoploteuthis it may be the fourth on

FIG. 201. — Hectocotylus of Philonexis
(Octopus) carense (after Leuckart). <i,
Spermatophoral pouch ; b, seminal vesicle ;
c, efferent duct of the same ; d, appendage
= remains of the penis sac ; e, penis ; /,
.sucker.

the right, or even in one and the same species of Ommastrephes, it is sometimes
the fourth on the left and sometimes the fourth on the right. In Sepiola and

244

COMPARATIVE ANATOMY

CHAP.

Rossia, it is the first ami which is hectocotylised. Finally, both the arms of one
pair may be thus transformed ; in Idiosepiou and Spimla this is the case with the
fourth pair, in Rossia with the first.

The difference in size between the male and the female, which has been mentioned
as occurring in those forms which have true hectocotylised arms, is also found,
though not to the same degree, in many other Cephalopoda, in which the male is
slightly smaller than the female.

XXI. Parasitic Gastropoda.

1. Thyca ectoconcha (Fig. 202) is a Prosobranchiate Gastropod which is parasitic
on the Star-fish Linckia multiforis. The chief points in its organisation are shown
in Fig. 202, a longitudinal section in which, however, several organs which lie
laterally to the section are also represented. The organisation of the Gastropod is as
yet little influenced by its parasitic manner of life. It possesses a shell, shaped

cec

ml

FIG. 202.— Longitudinal section through Thyca ectoconcha (after P. ami F. Sarasin). Some
organs not actually belonging to the section are included, cer, Cerebral ganglion ; cl, alimentary
canal ; fl, folds ; /«, foot ; 1; gill ; I, liver ; ml, mantle ; oc, eye ; ot, otocyst ; pcd, pedal ganglion ; pr,
proboscis ; s/j false foot ; si, oesophageal bulb ; vl, cephalic fold.

somewhat like a Phrygian cap. In the mantle cavity lies the gill. The alimentary
and nervous systems also are in no way remarkable. It has eyes and auditory
organs, and a short powerful snout, and muscular cesophageal bulb, which penetrates
the Star-fish between the calcareous parts of its integument into the tissues. There
is no radula. The base of the snout is surrounded by a muscular disc consisting
of an anterior and a posterior part. This disc, the so-called false foot, is the
grasping organ by which the animal attaches itself to the integument of its host
so firmly that it cannot be torn away without injury. The rudiment of a foot
(fs) occurs without an operculum.

2. The Gastropodan organisation is somewhat more strongly modified in
Stilifer Linckice (Fig. 203), which is parasitic on the male Linckia. The whole
body of this parasite penetrates into the calcareous layer of the integument of the
host, on which it raises pathological globular swellings, and further causes the
peritoneum to bulge inwards towards the body cavity. The parasite communicates

VII

MOLLUSCA— PARASITIC GASTROPODA

245

with the outer world only by means of a small aperture at the tip of the swelling.
The parasite, thus established in the integument of its host, is surrounded on all
sides by a fleshy envelope (sm). This envelope is only broken through by an
aperture at the point where the apex of the dextrally twisted shell lies ; this
aperture corresponds in position with the aperture above mentioned as occurring
at the tip of the pathological swelling. This envelope is called the false mantle,
and corresponds morphologically with the false foot of Thyca, much increased in
size and bent back on to the shell. There occur besides a true mantle, a gill, a
rudimentary foot without an operculum, eyes, auditory organs, and a typical
Prosobranchiate nervous system. The development of the remarkable false mantle

f

PIG. 203.— Longitudinal section through Stilifer Linckiae (after P. and F. Sarasin). lc,
Buccal ganglion ; W, blood sinus ; ccr, cerebral ganglion ; d, alimentary canal ; /s, foot ; k, gill ; ?,
liver ; ml, mantle ; n, proboscidal nerve ; oc, eye ; ot, otocyst ; ped, pedal ganglion ; pr, proboscis ;
sm, false mantle ; sub, subintestinal ganglion ; sup, supraintestinal ganglion.

no doubt signifies that, although the animal is embedded deep in the integument
of the host, communication with the exterior is retained. Water for respiration can
enter and flow out of the mantle cavity, and the fecal masses and genital products,
and perhaps also the larvae can pass into the cavity of the false mantle and be
ejected through its aperture. The sexes are separate. The snout has lengthened
into a very long proboscidal tube which pierces the tissues of the integument of the
Star-fish, which are rich in blood, and draws from them the necessary nourishment.
Both cesophageal bulb and radula are wanting.

3. The two parasites just described are typical Gastropods, and are easily
recognised as such when carefully examined ; there are, however, two other
parasitic Gastropods in which the typical organisation is so much modified that it

246

COMPARATIVE ANATOMY

CHAP.

would be difficult to recognise them as Gastropods, or even as Molluscs, were it not
proved that the larvae of one of these forms at least are distinctly Gastropodan
larvag. The incomplete state of our knowledge of the development of these two
parasites, and the absence of any transition forms between them and the typical
organisation, make them very difficult to understand.

Entocolax Ludwigii inhabits endoparasitically the body cavity of a Holothurian
(Myriotrochus Rinkii), one end of its vermiform body being attached to the body
wall of its host. Its organisation, a scheme of which is given in Fig. 205, can be
best studied with the help of some hypothetical transition forms, through which a

-nt

•MX

Fio. 204.— A, B, C, D, Hypothetical transition stages between Thyca and Stilifer on the
one side and Entocolax (Fig. 20.j) on the other (after Schiemenz). u, Anus ; f<l, pedal gland ;
7, liver (digestive gland) ; Id, hepatic intestine ; M, mouth ; mag, stomach ; o, ovary ; of, aperture of
the false mantle ; sf, false foot ; sw, false mantle ; u, uterus ; w, body wall of the host.

Gastropod of the type of Thyca or Stilifer might pass in developing into an
endoparasitic parasite like Entocolax. Fig. 204 A shows the first stage, which still
much resembles Thyca, and is still ectoparasitic ; Fig. 204 B, C, D are further
stages in development. In proportion as the animal becomes endoparasitic, and
gives up its relations to the external world, do the sensory organs, the shell, and the
mantle cavity with the gill disappear. The stomach, as a separate section of the
intestine, degenerates, the digestive gland (liver) becomes a simple unbranched
diverticulum of the intestine, which loses the rectum and anus. All organs for the
purpose of mastication at the anterior end of the alimentary canal are lost. The

VII

MOLLUSC A— PARASITIC GASTROPODA

247

false mantle becomes larger and larger, and envelops the small visceral dome, which
gradually becomes rudimentary, and finally contains merely the genital organs. At
the stage D the whole animal already projects freely into the body cavity of the
host, attached to its wall by a displaced portion of the false foot, and connected with
the exterior only by the aperture of the false mantle. If this last means of
communication with the exterior is also abandoned, i.e. if the whole false mantle
with its aperture becomes enclosed in the body cavity of the host, we have a form
corresponding with the endoparasite Entocolax Ludwigii (Fig. 205). In this form,
the cavity enclosed by the false mantle, into which the ovary and its receptacula

FIG. 206.

-m,

FIG. 20£

FIG. 205. —Entocolax Ludwigii, sketch
after Voigt. Lettering the same as that in
the preceding figure.

FIG. 206.— Entoconcha mirabilis, sketch
by Schiemenez (after Baur). Lettering as

in Fig. 204. liod, Testes?

seminis open, serves as a receptacle for the fertilised eggs, which were found in it in
their first stage of segmentation in the one (female) specimen discovered.

Entoconcha mirabilis, an endoparasite which has been found in a Holothurian,
Synapta digitata, is even more deformed than Entocolax. The body of this parasite
is a long vermiform coiled tube, attached by one end to the intestine of the host,
while the rest of the tube floats freely in the body cavity of the latter. Its
organisation has as yet been imperfectly investigated. Fig. 206 is a very simple
diagram, which is introduced for comparison with Fig. 205 of Entocolax. It is
impossible to say how far such a comparison, which the lettering is intended to
facilitate, is justifiable. Up to the present time, no aperture leading from the ovary
into the brood-chamber, which is thought to be the cavity of the false mantle, and is

248 COMPARATIVE ANATOMY CHAP, vn

filled with embryos (not represented in the figure), has been observed. In a widening
of the tube near its attached end, a number of free " testicular vesicles" have been
found, but their real significance can only be discovered by further research.

The embryos found in the brood cavity of Entoconcha have the same general
structure as Gastropodan larva1. They have a spirally twisted shell, into which
the body can be withdrawn ; an operculum, a small velum, the rudiments of two
tentacles, two auditory vesicles, [a foot, and an intestine, which, according to one
observer (the most recent), consists of only a mouth, pharynx, oesophagus, and the
rudiment of a liver, but according to an older authority is complete. There is,
further, a branchial cavity with a transverse row of long cilia. Nothing further is
known of the development and life history of Entoconcha.

Some details of parasitic Lamellibranchiate larvae (Union/idee) will be given in the
section on Ontogeny.

XXII. Attached Gastropoda.

Of the several forms of attached Gastropods known, only Vcrmctus, whose inner
organisation has been carefully investigated, can be shortly described in this place.
Vcrmctus has a shell which, instead of being coiled like the well-known shell of the
snail, is a calcareous tube, which rises freely from the bottom of the sea, to which its
tip is cemented. This shell is very like the calcareous tubes of tubicolous worms
such as Serpula. The larva of this form, however, possesses a typically coiled shell,
and even the young animal, after it has attached itself, lias such a shell. In the
course of growth, however, the coils become loosened, and the shell finally grows out
as a tube.

The typical organisation of the Monotocardian ProsobrancMatcs, to which Vermctus
belongs, is little affected by the attached manner of life. The visceral dome, like
the shell, is much elongated and almost vermiform. The intestine, the circulatory
system, the kidneys, the mantle, the gill, and the nervous system are typically
developed. The sexes are separate, and copulatory organs, which could not be
used by attached animals, are wanting. The head is well developed, and the
pharynx well armed. When the animal is slightly irritated, it is said not to
withdraw at once into its shell, like other Gastropods, but to bite. The foot has
the form of a truncated cylinder, and is directed anteriorly, ventrally to the head.
It cannot, of course, function as a locomotory organ, but carries the operculum for
closing the shell, and, by means of the pedal gland, secretes mucus. Vermctus is
said to produce great quantities of this secretion, which it allows to float in the
water for a time like a veil, and then swallows together with all that has become
attached to it. In this way it fishes for the small organisms which form its food.

XXIII. Ontogeny.

A. Amphineura.

1. Ontogeny of Chiton Polii (Fig. 207). The egg possesses little nutritive yolk.
The segmentation is total and somewhat unequal ; a coelogastrula is formed by
invagination.

(a) The blastopore of the gastrula larva marks its posterior end. .'» A pair of
endoderm cells near the dorsal edge of the blastopore are specially large. A
longitudinal section shows two dorsal and two ventral ectodermal cells with larger

FIG. 207.— Development of Chiton Polii (after Kowalevsky). A-F, Six stages in the develop-
ment of the gastrula into the young Chiton ; sections nearly median. Cr, frontal section through
stage C, oblique, from the upper part of the velum to the blastopore. H, I, K, L, transverse
sections of four stages of development behind the mouth. 1, Blastopore ; 2, archeuteron or
midgut ; 3, mesoderm ; 4, ectoderm ; 5, velum or preoral ciliated ring ; 6, stomodaeum or oesophagus ;
7, mouth ; 8, radular sac ; 0, body cavity ; 10, pedal gland, in I oesophagus ; 11, foot ; 12, anus
with proctodamm ; 13, cerebral ganglion ; 14, pretrochal tuft ; 15, pleurovisceral cords ; 1C, pedal
cords ; 17, mantle furrow ; IS, eye ; c, shell ; ci-c7, the seven shell plates first formed.

250 COMPARATIVE ANATOMY CHAP.

nuclei ; these belong to a double row of cells on which is developed the preoral
ciliated ring which, in Molluscs, is called the velum (Fig. 207 A).

(b) At a later stage, the blastopore appears shifted somewhat towards the ventral
side, and an inward growth of ectodermal cells begins at its edge ; this is the
commencement of the formation of the ectodermal stomodaeum. At the posterior
and upper edge of the blastopore, there is, in the figure, a cell lying between the
endoderm and the ectoderm ; this is, no doubt, a mesodermal cell (B).

(c) The larva elongates ; a distinct stomodaeum (embryonic oesophagus), leading
through the blastopore into the archenteron, is formed by the continuous growth
inward of the ectodermal cells ; this organ becomes shifted still further forward
along the ventral surface (C).

(d) Fig. 207 G is an oblique section from an anterior upper to a posterior lower
point through a slightly older larva, which shows the stomodseum, and, at the sides
of the blastopore, the first mesoderm cells. These are probably derived from the
endoderm, and are symmetrically placed at the two sides of the blastopore.

(e) A median section through the next stage (D) shows no mesoderm cells as yet
in the median plane. The mouth, however, appears shifted forward along the
ventral side as far as the ciliated ring or velum, the double row of cells in the latter
being very clear.

(/) Transverse section of an older stage (H). The mesoderm cells have increased
in number, and are arranged in two groups at the sides of the stomodceum, between
the ectoderm and the endoderm.

(g) At a later stage, a longitudinal section of which is given in Fig. 207 E, the
principal feature is a stronger development of the mesoderm, in which a space, the
body cavity, now appears. A bulging backward of the stomodseum forms the first
rudiment of the radular sac. Behind the mouth, a sac-like depression is formed,
evidently by the ectoderm ; this has been called the pedal gland, although it has
not yet been discovered what becomes of it in the adult animal.

(h) When the body cavity forms, the cells of the mesoderm become divided into
two layers, the inner visceral layer becoming applied to the intestine, and the outer
parietal layer to the ectoderm (c/. Fig. 207 I). In the transverse section, we see,
deep down in the ectoderm, the first rudiments of the pleurovisceral cords. The
pedal cords arise in the same way, and anteriorly, in the cephalic area, which is
encircled by the preoral ciliated ring, the rudiments of the supra-cesophageal
central nervous system form as a neural plate, i.e. as a thickening of the ectoderm,
which carries a tuft of long cilia.

(i) At later stages (F, K, L), the central nervous system with the pleurovisceral
and pedal cords become detached from the ectoderm and take up their mesodermal
position. The rudiments of seven shell-plates appear on the back as cuticular
formations ; the eighth only appears later. A posterior invagination of the ectoderm
represents the rudiment of the proctodeum (the embryonic hind-gut with the anus).
The first teeth appear in the radular sac. The whole of the cephalic area and the
region of the foot become covered with cilia. On the dorsal ectoderm, on the parts
that are not covered by the shell-plates, the first calcareous spines appear. In the
posterior part of the body, a great accumulation of mesodermal elements evidently
marks the position of a formative mesodermal zone.

At this stage, the larva leaves the egg envelope, and swims about freely, and, on
the degeneration of the ciliated ring, sinks to the bottom transformed into a young
Chiton. During this last transformation two lateral larval eyes appear on the
anterior ventral side of the body. The development of the circulatory system, the
nephridia, the genital organs and the ctenidia has not been followed.

2. Solenogastres. — The ontogeny of this order is as yet only known through
a very incomplete account of the development of Dondcrsia banyulensis. The

VII

MOLL USCA— ONTOGENY

251

segmentation is unequal and total, and takes place through the formation of
micromeres. The process of gastrulation seems to occur in a manner half way
between epibole and invagination. The blastopore marks the posterior end of the
larval body, which is divided by two circular furrows into three consecutive regions.
The anterior region consists of two circles of cells, and evidently corresponds with
the pretrochal area. It is partially ciliated, and carries in the middle a group of
longer cilia, one of which is sometimes to be distinguished from the rest as a flagellum.
The second region, which consists of a single row of cells, carries a circle of long
cilia, and evidently represents the velum. The third region consists of two rows
of cells carrying short cilia ; the second row edges the blastopore. At an older
stage, the posterior part of the larva appears to be withdrawn into an invagination
of the anterior part. The whole or by far the greater part of the body of Dondersia
is said to be produced from this posterior part (the "embryonic cone") alone. On
this embryonic cone, there appear, first of all, on the two sides of the middle line,
three pairs of consecutive imbricated spiculre, still retained in their formative cells.

Fig. 208. — Dondersia toanyulensis. A, Larva 30 hours old. B, Larva 100 hours old. C, Young
Dondersia immediately after transformation (7th day), after Pruvot.

They soon break through these cells, and their number is increased by the appear-
ance of new ones anteriorly. The embryonic cone lengthens, becomes curved ven-
trally. The anterior part of the body with the velum and the pretrochal area becomes
reduced and finally appears as a sort of collar at the anterior end of the body. The
larva sinks to the bottom, and throws off the whole anterior part of the body with
the velum and the pretrochal area. Such throwing off or resorption of parts of the
body which have been of great physiological importance in the larva is very common
in the animal kingdom ; see sections on the ontogeny of the Worms (e.g. Nemertina,
Phoronis, etc., vol. i. p. 272), of the Arthropoda (Metamorphism of Insects, vol. i.
p. 490), and of the Echinodcrmata.

On the dorsal region of the young Dondersia, seven consecutive, imbricated, but
only slightly overlapping, calcareous plates can now be distinguished, consisting
of rectangular rods lying close alongside of one another (Fig. 208, 0). This
fact is very significant with regard to the shell of the Chiton, which in the adult
consists of eight, but in the larva of only seven plates. If it could be proved that
the Solcnogastridce pass through a Chiton stage, the view that they are more
specialised animals than the Polyplacophora, and are to be derived from Chiton -like
forms, would receive almost decisive support.

Besides the seven dorsal calcareous plates, the young Dondersia has numerous

252

COMPARATIVE ANATOMY

CHAP.

circular calcareous spicules, covering it laterally ; the ventral side is, however,
naked. A mouth is still wanting, the eudodermal mass is not yet hollow, and on
each side, between the endoderm and the integument, there is a solid mesoderrnal
streak.

B. Gastropoda.

As a type of the development of the Gastropoda, we may take Paludina vivipara
(Figs. 209 and 210), the ontogeny of which has recently been again very carefully
investigated. Development here takes place within the body of the mother. The
egg is comparatively poor in yolk. A ccelogastrula is formed by invaginatiou, the
blastopore of which marks the posterior end of the germ, and becomes the anus.
No proctodieum is formed. The whole of the intestine from the stomach to the

A

FIG. 209.— Development of Paludina vivipara (after v. Erlanger). A and B, Stage after
gastrulation, with the rudiments of the mesoderm and the ccelom as outgrowths of the archenteron.
A, Median optical longitudinal section. B, Horizontal optical longitudinal section. C, Horizontal
optical longitudinal section through the embryo, after the entire separation of the ccelomic sac
from the intestine. D, Sagittal optical longitudinal section through an embryo, in which the
mesoderm has broken up, the cells becoming spindle-shaped. 1, Velum ; 2, segmentation cavity ;
3, archenteron ; 4, coeloin ; 5, blastopore ; 6, mesoderm cells ; 7, shell-gland .

anus proceeds from the endoderm. The mesoderm arises as a ventral hollow out-
growth of the archenteron, which soon becomes constricted from the intestine, and
lies between the intestine and the ectoderm in the segmentation cavity as a vesicle
with two points directed forward (Fig. 209 A, B, C). This vesicle spreads out to
the right and left dorsally round the intestine, finally closing round it dorsally. Its
outer wall of cells, which becomes applied to the ectoderm, forms the parietal

vii MOLLUSCA— ONTOGENY 253

layer, while its inner wall, which is applied to the intestine, forms the visceral
layer of the mesoderm. The cells of the mesoderm soon become detached from one
another (Fig. 209, D) ; they assume the spindle shape and finally fill the segmenta-
tion cavity like a network.

In the meantime the velum has appeared, and, between it and the anus, the shell-
gland forms. The cesophagus arises as an invagination of the ectoderm, which soon
becomes connected with the midgut. By the addition of a paired primitive kidney,
the typical Molluscan Trocophora is formed ; this at first is quite symmetrical,
the anus lying posteriorly in the middle line.

After the development of the cesophagus, a mass of mesoderm cells collects on

each side of and below the intestine, this mass soon becoming hollow. In this way

two mesodermal sacs are formed which approximate towards the middle line till

they touch, and then fuse to form one sac, the double origin of which is still, for a

time, evidenced by the presence of a median septum. The sac which thus arises is

the pericardium. Fig. 210 A shows a somewhat further developed embryo seen

from the right side. Below and behind the mouth are seen the projecting rudiment

of the foot, on which to the right and left the auditory vesicles have arisen as

invaginations of the ectoderm. In the pretrochal area, protuberances to right and

left represent the rudiments of the tentacles, at the bases of which the eyes have

appeared as ectodermal pits. The shell gland has secreted a shell. The greater

growth of that side of the body which is covered by the shell has caused a bending

by which the anus is shifted towards the ventral side. Immediately behind the

anus, the ectoderm bulges out to form the rudiment of the mantle fold, so that the

anus comes to lie in a shallow depression, the rudiment of the pallial or respiratory

cavity. It is important to note that at this outwardly symmetrical stage, the mantle

cavity and the anus lie posteriorly. The fore-gut (oesophagus) has greatly lengthened.

The digestive gland has grown out from the stomach ventrally in the form of a

wide sac, but is still connected with the latter by a wide aperture. The pericardium,

in which the septum is still visible, has already somewhat shifted from below the

stomach to its right side. The rudiments of the definite nephridia next form in

the following way (Fig. 210, D). In each division of the pericardium (the left

division being smaller than the right) the wall bulges out ; the right outgrowth

becomes the secreting portion of the permanent kidney ; the left degenerates, but

must be regarded as a temporarily appearing rudiment of the (original) left kidney.

The mantle cavity, which lies beneath the pericardium, presses into it to the right

and left in the form of two projections. The right projection, continuing to grow,

becomes connected with the rudiment of the right kidney and forms its efferent duct.

The left projection does not grow further, nor does it become connected with the

rudiment of the left kidney.

A further stage is depicted from the right side in Fig. 210 B. The following are
the most important alterations. The optic pit has become constricted into an optic
vesicle. The mantle fold has grown further forward, and has become deeper to the
right. The undivided pericardium has shifted altogether to the right of the
stomach, and lies above the rectum, which bends forward and downward. The
body is already asymmetrical.

At the following stage (Fig. 210, C) the posterior and dorsal region of the body
rises distinctly from the rest as a visceral dome ; the shell covering this part of the
body has increased considerably in size. The mantle fold has become much broader,
and the mantle cavity much deeper ; the latter now lies chiefly on the right side of
the body. The looping of the intestine is far more marked. On the posterior and
dorsal side of the pericardium, the pericardial wall sinks in the form of a channel,
which soon closes and forms a tube ; this is the rudiment of the heart. The two
apertures of the tube, where the wall of the heart passes into that of the pericardium,

254

COMPARATIVE ANATOMY

CHAP.

communicate with the body cavity. The heart tube becomes constricted in the
middle, the anterior division forming the auricle and the beginning of the branchial
vein, the posterior, the ventricle and the rudiment of the body aorta.

//=:

Pio. 210.— Development of Paludina vivipara (after v. Erlanger) A, Right aspect of an
embryo, in which the pericardium is divided into two parts by a septum. B, The same of a some-
what older embryo, with an undivided pericardium. C, The same of an older embryo, in which the
first rudiment of the heart has appeared. D, Ventral aspect of the posterior end of an embryo, in
which the asymmetry of the visceral dome begins to appear. The anus is still median, but the
mantle cavity is already deeper on the right (the left in the figure). 1, Velum ; 2, mid-gut ; 3,
digestive gland (liver) ; 4, pericardium ; 4« and 4b, divisions of the same formed by a septum ; 5,
free edge of the shell ; 6, shell groove ; 7, anus ; 8, mantle cavity ; Sb, base of the mantle cavity =
base of the mantle fold ; 0, free edge of the mantle ; 10, foot ; 11, auditory organ ; 12, oesophagus ;
13, cephalic tentacle ; 14, eye ; 15, efferent duct of the (originally) right nephridium ; I5b, rudi-
mentary efferent duct of the (originally) left nephridium ; 10, primitive kidney ; 17, rudiment of
the heart ; 18«, right nephridium ; ISb, rudimentary left nephridium.

Fig. 211 A shows a somewhat older embryo which already resembles in form the
adult animal. The velum is reduced, and a ventral bulging of the anterior division
of the oesophagus represents the rudiment of the radular sac. The ventricle and

VII

MOLL USC A—ONTOGENY

255

the auricle are distinct. An ectoderrnal depression on the foot forms the operculum.
The mantle cavity which lies on the right side, and into which the rectum opens,
now also stretches to the left on the anterior and dorsal side of the sharply
demarcated visceral dome. The gill appears in the form of a protuberance on the

no.

lib

ila

lib

PIG. 211.— Development of Paludina vivipara (after v. Erlanger). A, An embryo in which
the first rudiment of the gill has appeared. B, A nearly mature embryo. Both are seen from the
left side. Lettering as in Fig. 210. In addition, 17a, Auricle ; 17b, ventricle ; IS, nephridium ; if),
rectum; 20, rudiment of the radular sac; 21, rudiment of the gill; 22, osphradium (Spengel's
organ) ; 23, rudiment of the genital duct ; 24, rudiment of the gouad ; 25, operculum.

inner surface of the mantle cavity, and the osphradium at the left of the gill as an
ectodermal protuberance.

Fig. 211 B finally shows ns an embryo in which the mantle cavity has assumed
the anterior position on the visceral dome. The ctenidium and osphradium have
developed further. The velum is very much reduced and can only be seen in
sections. This stage is important on account of the appearance of the rudiment of
the genital organs, which is identical in the two sexes. A depression of the (meso-

256

COMPARATIVE ANATOMY

CHAP.

dermal) pericardial wall, which becomes separated from the pericardium, forms the
rudiment of the gonad, while an ingrowth from the base of the mantle cavity runs
towards this, and is the (ectodermal) rudiment of the genital duct. The latter
arises on one side of the anus, just as the efferent duct of the permanent kidney
rises on its other side ; this ontogenetic fact confirms what was stated above (p. 219)
that the genital duct of the Monotocardia corresponds with a part of the right (which
originally, and in the young embryo, is the left) kidney of the Diotocardia (apparently
wanting in the Monotocardia].

The vascular system arises very early in the form of spaces between the mesoderm
and ectoderm or entoderm, round which the mesoderm cells grow, and which become
secondarily connected with the heart.

All the ganglia of the nervous system, the cerebral, pleural, pedal, parietal, and
visceral ganglia arise separately as ectodermal thickenings, which become constricted
off from the ectoderm by delamination. They only secondarily become connected
with one another through the growing out of the nerve fibres. The parietal ganglia
arise to right and left in the middle region of the body, but soon become shifted by
the displacement of the organs of the visceral dome, one above the intestine and the
other below it. The rudiment of the visceral ganglion is said to appear dorsally to
the hind-gut and to move later to its position beneath the same.

The observations on the development of Paludina vivipara, here briefly described,
are in many ways of great importance, and confirm in the most unmistakable
manner the results arrived at by comparative anatom)'. The following are specially
noteworthy.

1. The manner in which the pericardium originates favours the opinion that it
is a secondary body cavity. It is important to note that the pericardium is at

first paired, being divided into
two lateral halves by a septum,
which afterwards disappears.

2. The fact that the gonad
arises as an outgrowth of the
pericardium, confirms the view
arrived at by comparative
anatomy, that the genital cavity
also is a secondary body cavity.

3. The anus and the mantle
cavity originally lie symmetri-
cally at the posterior end of
the body, but, through asym-
metrical growth, come to lie
first on the right side of the
visceral dome, and finally on

_ its anterior side.

9 S 7
PIG. 212.— Larva of Oncidium celticum, from the left

The development of other

l, Cerebral ganglion ; 2, edge Gastropods cannot here be de-
scribed in detail. We refer the

side (after Joyeux Laffuie).

of the mantle ; 3, rudiment of the gonad ; 4, larval shell-
muscle ; 5, hind-gut ; 6, rudiment of the digestive gland ;
7, auditory organ ; 8, pedal ganglion ; 9, foot ; 10, oesophagus ; reader to the bibliography _ at

11, eye ; 12, branched muscle cells of the velum ; 13, velum.

the end. As a rule, nutritive
yolk is present in larger

quantities than in the viviparous Paludina, in which the small provision of
yolk is evidently connected with the favourable conditions of nutrition of the
embryo.

The blastopore generally corresponds in position with the future mouth ; it often,

VII

MOLL USGA— ONTOGENY

257

perhaps usually, remains open ; notwithstanding this, the cesophagus arises by the
sinking in of ectoderm cells.

Paludina is, as far as is known, the only Mollusc in which the mesoderm
originates as an outgrowth of the archenteron. This fact is no doubt connected
with its poverty in nutritive yolk. In other Gastropods, the mesoderm arises in the
manner already described for other Molluscs, as two large symmetrical primitive
cells, at the posterior edge of the blastopore ; these cells look more like endodermal
than ectodermal cells, and soon pass into the segmentation cavity.

A Veliger larva, i.e. a Trochophora with Molluscan characteristics, always forms
(1) a dorsal shell gland with the embryonic shell,
and (2) a ventral rudiment of the foot.

The outward appearance of the Veliger larva,
however, varies much in different groups, the
variations being connected with the manner of
life and of feeding of the embryo.

In the marine Gastropods, i.e. in the majority
of the Prosobrauchia (including the Heteropoda),
the Pulmonate genus Oncidiuiii, and all Opistho-
branchia, the embryo leaves the egg envelope early,
as a free-swimming Veliger larva. In all these
forms, the preoral ciliated ring is well developed.

The ectodermal floor of the ciliated ring usually FIG. 213. — Larva of Cymbulia
bulges out anteriorly, so that the cilia appear to (Pteropod), from the left side (after
be carried by a distinct circular ridge. This ridge Gegenbaur). i, Velum ; 2, shell ; 3,
, , , ,, ~ , , c . parapodia (fins); 4, foot with oper-

even grows out laterally to form a lobe or varying .* „,

size, which carries at its edge long and strong

cilia, and is occasionally itself produced into an upper and a lower lobe. This is
the true velum of the free-swimming Gastropod larva, and is its only organ of
locomotion. It is internally traversed from wall to wall by contractile mesoderm
cells (muscle cells), which make it highly contractile. In the older larvre, the head
with the velum can be withdrawn into the shell.

, It is probable that the velum of the larva also serves for respiration, and perhaps
for bringing about a circulation of the body fluid by means of its contractility.

The embryos of fresh-water and terrestrial Gastropods, where these animals are
not viviparous, remain longer in the egg, and leave it only after their transformation
into young Gastropods, the larval organs (the velum, the primitive kidney, the
cephalic vesicle, and the pedal vesicle or podocyst) having degenerated within the
egg envelope. Even in these forms, the mass of nutritive yolk contained in the egg
is not very great, but there is a large quantity of albumen stored up within the egg
capsule, which serves as food for the developing embryo ; this is either absorbed
through the body wall or swallowed. The egg capsules are always large, in some
cases (in tropical terrestrial Gastropods) as large as the egg of a small bird ; but
their size is not, as in the Cephalopoda, determined by that of the egg contained,
but by the quantity of albumen in which the small egg is embedded. The mature
egg capsule contains a young Gastropod of considerable size with a well-developed
shell.

In terrestrial and fresh-water forms, the velum is not needed as a locomotory
organ, and is therefore reduced to a single ring of cilia or to two lateral ciliated
streaks. It is entirely wanting in the embryos of a few terrestrial Gastropod snails.
The respiratory and circulatory functions, which were originally merely accessory
functions of the velum, here become of greater importance. The nuchal region
becomes much bulged forward, and forms a cephalic vesicle (Fig. 214), which is
sometimes very large, and undergoes regular pulsations. The posterior division
VOL. II S

258

COMPARATIVE ANATOMY

CHAP.

of the foot, in the same way, is often

Fio. 214.— Embryo of Helix Waltoni (4 mm.
long), from the right side (after P. and F.
Sarasin). 1, Cephalic vesicle ; 2, upper (optic)
tentacle ; 3, eye ; 4, lower tentacle ; 5, oral
lobe ; 6, sensory plate ; 7, podocyst.

Ill the larva of the gynmosomatou
rings are developed on the body.

widened into a pulsating pedal vesicle or
podocyst. Towards the end of larval life
the cephalic and pedal vesicles and other
similar "larval hearts" degenerate.

The embryonic shell is either retained
throughout life or is thrown off at an early
stage, and replaced by the rudiment of the
definitive shell. Even a second temporary
shell occasionally attains development.

It must again be noted that shell-less
Gastropods, to whatever natural division
they belong, pass through a typical Yeliger
stage, and at the older Yeliger stage have
a distinctly demarcated coiled visceral
dome, with a corresponding shell, and
usually an operculum on the metapodium.
s Pteropoda three postoral accessory ciliated

C. Scaphopoda.

Ontogeny of Dentalium. — Segmentation, in these animals, leads to the formation
of a cceloblastula, from which a coslogastrula arises by imagination. The blast opore
at first lies posteriorly on the ventral side, but
gradually shifts, as in Chiton, more and more
forward along the ventral side. The stomodseum
arises as an ingrowth of the ectoderm, the blasto-
pore nevertheless remaining open. A typical
Molluscan Trocophora is developed, although no
primitive kidney has been found. The velum is
a thick ridge round the body of the long oviform
larva. This ridge consists of three rings of very
large ectoderm cells, each row carrying a circle
of long cilia. The shell gland spreads out at an
early stage, its lateral edge soon growing out
ventrally and posteriorly as the mantle fold.
The free edges of the two folds fuse at a later
stage below the body. The anus forms very late.
The development of the cerebral and pedal
ganglia and of the auditory organ has been
specially carefully observed. On the ventral

side of the pretrochal area, in front of the velum

hours old, posterior and lower aspect

and behind the tuft of cilia, two symmetrical (aftei. Kowalevsky). i, Cephalic tuft ;

invaginations of the ectoderm form the cephalic 2, rudiments of the cerebral ganglion

sacs or tubes. These become constricted from the (cephalic tubes); 3, velum, consisting

ectoderm at a later stage, their lumen gradually of three rows of cilia; 4, mouth, hidden

,„..,. . .,.,"-. ,, under the ridge of the velum ; 5, mantle
narrows and finally disappears, while their walls

Fio. 215.— Larva of Dentalium, 37

become thick and multilaminar by the con-

tinuous growth of the cells. The two cell masses which thus arise become connected
in the middle line above and below the cesophagus, and form the cerebral ganglion.
The otocysts arise at the base of the pedal rudiment on each side as ectodermal
epithelial pits, which soon become detached from the ectoderm in the form of
epithelial vesicles, Immediately beneath these auditory vesicles, certain ectoderm

VII

MOLL USO A— ONTOGENY

259

cells sink below the surface, and form on each side an ectodermal cell mass, which
becomes detached from the rest of the ectoderm, sinks into the mesoderm of the
foot, and fuses with the similar mass on the other side to form the pedal ganglion.

D. Lamellibranchia.

1. Development of Teredo (Figs. 216 and 217). Segmentation is here total and
unequal. The gastrula, formed by epibole (Fig. 216 A, B) consists of (1) two
large endoderm cells (macromeres), a thick cap of ectoderm cells (micromeres)
closely covering these, and
two symmetrical primitive
mesoderm cells of medium
size at the posterior edge of
the blastopore. The blasto-
pore closes from behind for-
ward, the ectodei'm cells by
continual division growing
entirely round the endoderm
cells ; during this process the
two mesoderm cells become
covered by the ectoderm and
come to lie between the latter
and the endoderm (Fig. 216
C). Somewhat anteriorly on
the ventral side, a depression
of the ectoderm forms a pit,
the stomodseum (D). The
ectoderm separates off from S7\ O |
the two - celled mesoderm,
thus giving rise to a seg-
mentation cavity, or primary
body cavity. A double
preoral ciliated band is
formed (D, E). The two
large endoderm cells, by
fission, produce other smaller
cells. Cilia appear over the
whole surface of the germ,
with the exception of the
posterior dorsal surface, where
the ectoderm cells, which
have become cylindrical, sink
in to form the shell gland (F).
The latter secretes the first
rudiment of the shell in the
form of a simple cuticular
membrane. The endoderm

•u

FIG. 216.— A-G, Stages in the development of Teredo (after
Hatschek). A, C, D, E, F, G, from the right side, B in optical
horizontal section. 1, Ectoderm ; 2, macromeres = endoderm
cells ; 3, primitive mesoderm cells ; 4, segmentation cavity ;
5, stomodseum (oesophagus) ; 6, mouth ; 7, preoral ciliated
band ; 8, shell gland ; 9, shell ; 10, larval muscle cells ; 11,
cephalic plate with tuft ; 12, anal invagination, anus ; 13,
endodermal mid-nut.

cells begin to collect to form

the intestinal wall. After the formation of the first rudiment of the shell, the shell
gland flattens and spreads out ; its edge can still be found as a ridge running under
the edge of the shell. The endoderm now forms a large globular hollow mid-gut,
into which the oesophagus breaks through. Each of the primitive mesoderm cells

260

COMPARATIVE ANATOMY

CHAP.

has given rise to two or three smaller cells. The thin cuticular shell becomes
bivalvular by the appearance of a mediodorsal boundary line.

A further stage is distinguished first by the appearance of a small posterior
ectodermal invagination, the proctodseum, which produces the rectum and anus.
An ectodermal thickening, the neural plate, appears in the pretrochal area, carrying
three flagella. Some of the mesoderm cells become muscle cells (Fig. 216 G).

The next stage may be called that of the Trochophora larva. This larva differs
from a typical Annelidan Trochophora only by possessing a shell, which now covers
the greater part of the body, and by a mantle which appears, at first, posteriorly, and

10

12 2

15

FIG. 217.— Older Larva of Teredo, from the right side (after Hatschek). Lettering as in Fig.
216. In addition, 14, rudiment of the digestive gland (liver) ; 15, preoral ciliated baud (velum) ;
16, postoral ciliated baud ; 17, primitive kidney ; 18, auditory vesicle ; 19, rudiment of the pedal
ganglion ; 20, rudiment of the gill ; 21, mesodermal streak.

then at the sides, as a fold, and continues to grow from behind forward. The region of
the body which lies behind the cephalic area has spread out 011 each side to form a
broad fold, which becomes outwardly applied to the shell. The neural plate has
become multilaminar, and the proctodteum has broken through into the mid-gut.
The primitive mesoderm cells have given rise to a short mesoderm streak on each
side. At the anterior end of each mesoderm streak a somewhat long body, the
primitive kidney, has formed ; this contains a channel which opens externally, and
whose lumen is ciliated at a later stage. The rudiment of the digestive gland
appears in the mid-gut as a paired semi-spherical outgrowth. The body is no longer
ciliated all over ; cilia are retained only on the neural plate and in the anal region.
The double preoral ciliated band now becomes very distinct, and a postoral band is

vii MOLLUSGA— ONTOGENY 261

now added. The region between the two ciliated bands also carries cilia and forms
an adoral ciliated zone.

A further stage of development is depicted in Fig. 217. The rudiment of the
pedal ganglion can be recognised as an ectodermal thickening on the ventral side,
and that of the gill as a thick epithelial ridge. The stomach has formed a csecum
posteriorly, and the narrow mid-gut has formed a loop. The two auditory vesicles
containing otoliths have arisen between the mouth and anus as ingrowths of the
ectoderm which have become detached. The mesoderm consists of branched muscle
ceils, branched cells of connective tissue, the primitive kidney and the still
undifferentiated cells of the mesoderm streaks.

The ectodermal thickening, which represents the rudiment of the pedal ganglion
at a later stage, becomes rounded off and detached from the ectoderm, at the same
time becoming surrounded by the cells of the mesoderm streak, which have rapidly
multiplied, and which unite in front of it to form a median mass of cells. This
median mass of mesoderm cells increases greatly by rapid division, bulging forward
the ectoderm in the anterior ventral region, and thus forming the rudiment of the
foot. In the growing branchial fold slits occur, a single slit appearing first, and
another soon following in front of the first. The further development of this
larva is unknown.

The development of other marine bivalves runs very much the same course as
that of Teredo, the same larva being formed. The ciliated band is very strongly
developed in all marine bivalves (Teredo, Ostrca, Modiolaria, Cardium, Montacuta,
etc. ), and is generally carried by a collar-like expansion of the integument, or velum,
which is often divided into two lateral lobes. The velum, which on account of the
band of strong cilia it carries is the locomotory organ of the free-swimming larvse of
these Lamellibranchs, can be protruded out of and withdrawn! into the shell.

Among fresh-water Lamellibranchs there is one form, Drcisscnsia polymorpha,
whose larva is free-swimming and carries a well-developed velum. This form is
said to have migrated from salt to fresh water in (geologically speaking) recent
times.

Special arrangements are found among the other fresh-water forms. The eggs of
Pisidium and Cyclas, for instance, develop in special brood capsules in the gills of
the mother animal, and leave these as young bivalves. The Trochophora stage is,
nevertheless, passed through, but the velum, not being used for locomotion, remains
rudimentary.

2. Ontogeny of Cyclas cornea (Figs. 218 and 219). — We shall here only mention
the points in which the development of Cyclas differs from that of Teredo, and
describe such observations as complete those made on the latter. The blastula
consists of a cap of small cells (ectoderm cells) and a floor made of three large cells,
one very large primitive endoderm cell and two symmetrical primitive mesoderm
cells. The primitive endoderm cell yields through fission a disc of endodermal cells.
The two primitive mesoderm cells become overgrown by the ectoderm cells, and
thus reach the segmentation cavity. The endoderm invaginates in such a way that
a slit-like blastopore arises, which reaches from the region of the future mouth to
that of the future anus. This blastopore closes completely. The oesophagus arises
as an ingrowth of the ectoderm. A Molluscan Trochophora is formed with a
shell gland, a rudimentary foot, a mid-gut, a stomach, anus, primitive kidney, and a
neural plate. The velum is reduced to a ciliated area lying at the sides of the
mouth (Fig. 218) ; this reduction is connected with the fact that the Trocho-
phora of Cyclas is not a free-swimming larva, for the eggs of Cijdas pass through
the whole course of their development within the gills of the mother animal.
Above the neural plate the ectoderm cells are large and flat, and form a projecting
cephalic vesicle. The mesoderm consists of (1) scattered cells, which lie under

262

COMPARATIVE ANATOMY

CHAP.

the ectoderm of the cephalic cavity, in the foot and on the intestine (especially on
the oesophagus, where they are already changed into muscle cells) ; and (2) two
mesoderm streaks lying at the sides of the intestine. The pedal ganglia arise
together with the paired rudiment of the byssus gland, as thickenings of the
ectoderm at the posterior end of the foot. The auditory vesicles originate as
ingrowths of the ectoderm. The mantle forms by degrees from behind forward as a
ridge, which grows more and more ventrally downwards. At the same time the

22 25

". k $y 2o-2M

\ a />//•

23

6

FIG. 21 S.— A-D, Four stages in the development of Cyclas cornea, from the right side (after
Ziegler). 1, Membranous shell ; 2, rectum ; 3, anus ; 4, free edge of the mantle ridge or fold ;
5, rudimentary byssus cavity with gland ; (i, rudiment of the pedal ganglion ; 7, foot ; 8, velar
region; 9, oesophagus ; 10, stomach; 11, calcareous shell; 12, pericardium; 13, kidney; 14, rudi-
ment of the gonad ; 15, edge of the membranous shell ; 1(5, edge of the calcareous shell ; 17, rudi-
ment of the gill; IS, byssus thread; IP, visceral ganglion; 20, posterior adductor; 21, glandular
part of the kidney ; 22, lateral wall of the pericardial vesicle ; 24, median wall of the same ;
25, digestive gland (liver) ; 26, cerebral ganglion ; 27, mouth ; 28, auditory vesicle.

shell gland, which at its edge secretes the delicate shell membrane, spreads out and
becomes flattened. Beneath the shell-membrane the rudiments of the permanent
shell valves are produced from two small round areas lying to the right and left of
the dorsal middle line (B). The digestive gland (liver) develops from two lateral
globular outgrowths of the wall of the stomach. The gonads arise from cells of the
mesoderm streaks, which are larger than the rest and also in other ways differen-
tiated, so that they can very early be distinguished. In the anterior and dorsal

VII

MOLL USO A— ONTOGENY

263

part of each mesoderm streak a group of cells surrounds a cavity, which at first is
very small, but becomes continually larger. The two vesicles thus formed, the
cavities of which represent the secondary body cavity, form the pericardium.
Behind these the mesoderm cells collect in such a way as to form on each side a
strand, which becomes hollow ; this is the rudiment of the nephridium, which at
once becomes connected with the pericardial vesicle, and, growing further towards
the ectoderm, soon opens outward. The two pericardial vesicles lengthen posteriorly
and upward, each becoming divided into two parts, one lying behind the other, by a
constriction, the parts still communicating dorsally with one another (Fig. 219 A).
The two double vesicles grow towards one another above the rectum, and finally
fuse in the dorsal middle line (B). In a similar manner they fuse below the
rectum. The inner wall of the pericardial vesicle becomes the wall of the ventricle
(C), and its lateral wall becomes

that of the auricle. At the points
where the lateral vesicles were con-
stricted lie the slits through which
the auricles communicate with the
ventricle, and the atrioventricular
valves.

The visceral ganglion arises at
the posterior end of the mantle
furrow from an ectodermal thicken-
ing. The pleurovisceral connectives
form, in all probability, throughout
their whole length, through con-
striction from the ectoderm. The
gill arises on each side as a fold on

A

FIG. 210.— A-C, Diagrams illustrating the develop-
ment of the pericardium and heart of Cyclas cornea
(after Ziegler). 1 and 2, The two lateral pericardial
vesicles ; 3, rectum ; 4, pericardial cavity ; 5 and 6,
invaginations of the lateral walls of the pericardium =
rudiments of the two lateral auricles ; 7 and 8, median
walls of the two lateral pericardial vesicles, in B partly
fused to form a median septum (above and below the
intestine), which in C has disappeared ; 9, rudiment of
the ventricle.

the dorsal edge of the inner surface

of the mantle. It develops from

behind forward. In the contrary

direction furrows form on the

branchial fold, commencing from below upwards ; these are found on the inner as

well as the outer surface, and exactly correspond. The inner furrows join the outer

right through the gill, and thus give rise to the branchial slits.

3. The development of the Unionidse (Anodonta, Unio) is much influenced by
the parasitic manner of life of the larva. The fertilised eggs reach the outer leaf
of the gill of the female, and there run through the first stages of their develop-
ment. Segmentation leads to the formation of a cceloblastula, in which the rudi-
ment of the shell gland appears very early as an incurved plate of large and
high cells of the blastula wall. The archenteron forms by imagination at a very
late stage ; this is no doubt connected with the later parasitism of the larva.
Before this iiivagination occurs the mesoderm has begun to form ; its two primi-
tive cells lie in the blastoccel at the part where, later, the enteric invagination
appears.

The embryo known as Glochidium parasiticum has, in the last stage of its
development, which is passed through in the gills of the mother animal but within
the egg-shell, the following structure (Fig. 220). It is bilaterally symmetrical, and
has a bivalve shell. Each valve has, at its ventral edge, a triangular process,
the exterior of which is beset with short spines and thorns. Between the two valves,
which are markedly concave, lies the soft body, which lines the shell internally in
such a way that its ventral epithelial layer might, incorrectly, be called a mantle.
It may be called the false mantle. If this false mantle is examined from below,
when the shell is open, it is seen to have on each side four sensory cells furnished

264

COMPARATIVE ANATOMY

CHAP.

B

with long sensory hairs ; three of these cells lie near the shell process, and the fourth
near the middle line. Between the two more median sensory cells a long adhesive
filament projects from the opening of the gland which secretes it. Behind this
gland are found — (1) the oral sinus; (2) a small prominence, the pedal swelling;

(3) the ciliated lateral pits,
one on each side ; and (4)
furthest back of all, the
ciliated shield or patch.
Between the mantle and
shell the embryonic adduc-
tor runs across from the one
valve to the other. Besides
these are only found a few
isolated muscle fibres, and
the rudiment 'of the mid-
gut, the latter as an epi-
thelial vesicle, which be-
comes entirely separated
from the ectoderm, and in
no way communicates with
the exterior.

The embryo at this stage
leaves the gills, at the same
time emerging from the egg
shell. Its adhesive filament
floats in the water. If a-
passing fish comes in contact
with such an embryo, the
latter can, by closing its
shell, attach itself by means
of the triangular processes
mentioned above, to its in-
tegument, into which the
spines on these processes
penetrate. The embryo of

Anodonta attaches itself
Fio. 220.— Glochidium larva of Anodonta, from the outer leaf

of the gill of a female. A, from below, the shell being open chiefly to the fins, that ot

(after Schierholz). B, in optical transverse section (after Unio to the gills of the fish.

Flemming). 1, Sensory seta; 2, adhesive filament; 3, shell- The epithelium of the part

process ; 4, false mantle ; 5, lateral pits ; 6, oral sinus ; 7, pedal Qj- ^.jle gg^ attacked "rows

swelling ; 8, ciliated patch ; 9, embryonic adductor ; 10, shell. . , ,

very rapidly in such a way

as in a few hours to surround the parasite completely. The embryonic false mantle
grows out from each valve of the shell as a fungus-like body to penetrate the tissues
of the host, and probably serves for nourishing the embryo. During this endo-
parasitic life, which lasts for several weeks, the transformation of the embryo into
the young Mussel is completed. In the course of this process of transformation
some larval organs are resorbed, and also serve for nutrition ; first the sensory cells
disappear in this way, then the gland of the adhesive filament with the remains of
the filament itself, then the adductor, and finally the false mantle. The rudiment
of the definitive mantle and shell then appear. The vesicular mid-gut joins the oral
sinus ; the pedal swelling grows into the linguiform foot, and, in this, the rudi-
mentary byssus gland appears as an ingrowth of the epithelium. The rudiments
of the inner branchial leaves, the digestive gland, the nephridium, the heart, the

vii MOLLUSC A— ONTOGENY 265

cerebral, pedal, and visceral ganglia, and the auditory vesicle appear during the
parasitic stage, in the same way as in other Lamellibranchs.

During the last week of parasitic life the capsule formed by a growth of the
tissue of the host which surrounds the embryo becomes thinner ; the parasite breaks
through it, and falls to the bottom of the water as a young Mussel. The only
organs still wanting are the genital organs, the outer leaves of the gills, and the oral
lobes.

E. Cephalopoda.

Tetrabranchia. — Nothing is known of the development of Nautilus.

Dibranchia. — The egg is usually very large, and contains, like that of the sharks,
reptiles, and birds, a great quantity of nutritive yolk. It belongs to the telolecithal
meroblastic type, and is enclosed in a capsule. A number of such capsules may
become cemented together to form strings. The partial segmentation takes place
at the animal pole of the egg, and leads to the formation at that point of a germinal
disc (blastoderm).

Ontogeny of Sepia. — The blastoderm grows so very slowly round the yolk, that
long after all the outer organs of the embryo are quite recognisable in the region
of the original germinal disc, the opposite pole is still occupied by the yolk. The
germ lies in such a way that the centre of the germinal disc or animal pole is placed
dorsally, and corresponds with the uppermost point of the visceral dome of the adult
animal, while the mass of nutritive yolk lies ventrally.

1st Stage (Fig. 221 A). — In the centre of the germinal disc there appears an
oval-rhombic bulging ; this is the rudiment of the visceral dome and the mantle.
On each side of this there arises a bean-shaped prominence, the rudiment of the eye.
Behind the eye, on each side, a long narrow ridge runs backward in a curve ; about
half way down this ridge a small prominence, the rudiment of the funnel cartilage,
forms close to its outer side. The part of the ridge lying in front of this prominence
becomes the muscle which runs from the funnel to the nuchal cartilage ; the
posterior part (which lies behind the rudiment of the visceral dome and mantle)
forms the paired rudiment of the funnel itself. Between the two rudiments of
the funnel two other prominences rise symmetrically — the rudiments of the gills.
A pit in the centre of the rudiment of the visceral dome has been indicated as the
rudiment of a shell gland (?).

2nd Stage (Figs. 221 B and 222 A). — The rudiments just described stand out
more prominently. On the outer and posterior sides of the rudiments of the funnel
the rudiments of the two posterior pairs of arms first appear as prominences, then
those of the third and fourth pairs. The first indications of the head are seen in
the form of a large double swelling on each side, the outer and anterior part of
which carries on each side the rudiment of the eye. The embryo becomes covered
with cilia. At the extreme anterior end the mouth appears in the middle line,
forming the opening of the oesophagus, which begins to sink inwards.

3rd Stage (Fig. 221 C). — The whole embryo has become more arched dorsally,
and more marked off from the yolk. On the latter, the blastoderm, which consists
of two layers, an external ectoderm and an internal yolk membrane, has spread out
further towards the ventral (vegetative) pole of the egg. At the posterior edge of
the rudiment of the visceral dome, the mantle fold has grown out forward in such a
way as to form a small mantle cavity, which already partly covers the rudiments of
the gills. In the space between the rudiments of the funnel and the gills the
proctodseum has formed by invagination, and its aperture, the anus, can be seen.
The rudiment of the fifth pair of arms appears.

4th Stage (Figs. 221 D and 222 F, G). — The visceral dome projects further,

266

COMPARA TIVE ANA TOMY

CHAP.

and has a free mantle edge all round its base. The gills have shifted further
into the mantle cavity, which is now larger, and lies posteriorly. The rudiments
of the funnel also now lie close to the mantle, and are so approximated pos-
teriorly as nearly to touch. The rudiments of the arms have shifted from
behind further forward round the rudiments of the head. As the whole embryo
projects more distinctly from the yolk, the rudiments of the arms shift nearer to

B

C

D

E

FIG. 2-2\.— Ontogeny of Sepia (utter Koelliker). A-E, Five stages of development. The free
surface of the germinal disc which lies on the yolk is seen, its centre corresponding with the dorsal
point of the visceral dome of the adult Srjiin. The anterior side of the embryo lies lowest in the
figures. «, Visceral dome with mantle ; ft, rudiment of eye ; c, rudiment of gill ; d, halves of the
funnel ; c, rudiment of the funnel cartilage belonging to the apparatus for closing the mantle ;
/, peripheral part of the blastoderm, which, growing all round the yolk, forms the yolk-sac ;
17, mouth : it, posterior cephalic lobe ; i, anterior cephalic lobe ; 1;, anus ; 5, anterior or first pair
of arms ; 4, 3, •!, 1, second, third, fourth, and posterior pairs of arms.

one another and under the rudiments of the head. The anus is already covered by
the mantle fold.

5th Stage (Figs. 221 E and 222 B, H).— The arms shift still nearer to one
another (i.<: towards the axis of the embryo), grouping below the rudiments of the
head (which have become fused), and form a somewhat narrow circle on the ventral
side in such a way that, when the embryo is seen from the dorsal side, some of them
are hidden by the head. As a consequence of this the embryo, which is already
recognisable as a young Sepia, now becomes sharply constricted from the yolk
beneath it. The free edges of the rudiments of the funnel fuse and move to a position
within the mantle cavity.

6th Stage (Fig. 222 C). — The rudiments of the head and arms have now
assumed the typical position to form the "head" (Kopffuss). The embryo is now
altogether distinct from the yolk, to which it merely hangs instead of, as before,
lying upon it. The blastoderm finally grows round the yolk and so forms a yolk
sac. At first this sac is four or five times the size of the embryo, but in proportion

VII

MOLL USCA— ONTOGENY

267

FIG. 222.— Various stages in the development of Sepia (after Koelliker). A, B, C, D,
Anterior view ; E and F, from the left side ; G and H, from behind. Lettering as in Fig. 221. In
addition: d, rudiment of the funnel-nuchal muscle (collaris) ; rfj, paired rudiment of the funnel
proper ; p, yolk ; aj, edge of the mantle ; t, optic imagination (?) ; -«, region of the shell ; q, edges
of the two rudiments of the funnel bent round ; r, fins. In G the mantle fold is raised up in H
cut off.

268 COMPARATIVE ANATOMY CHAP.

as the latter grows at the expense of the yolk and develops further, the sac becomes
smaller, so that when the embryo is hatched the size of the yolk-sac is only one third of
that of the young animal (Fig. 222 D). It must further be mentioned with regard to
the yolk sac that it is at no time in communication with the intestine. As the
embryo becomes constricted from the yolk the latter divides into two parts — an inner
part, lying inside the embryo, and an outer part, filling the sac. These two parts
are connected by means of the stalk of the yolk sac, which projects downward from
the "head." The yolk within the embryo is divided into three unequal parts, the
largest of which fills the visceral dome ; another mass of considerable size fills the
"head," and these two masses are connected with a smaller portion lying in the
nuchal region.

Loligo and Argonauta have a smaller yolk sac, round which the blastopore
grows at an earlier stage than in Sepia. The yolk sac of Argonauta is entirely
taken into the body before the latter has completely closed ventrally.

The quantity of nutritive yolk is still less in a Cephalopod (Ommatostrephes ?),
the spawn of which floats in the sea. Segmentation is in this case also partial and
discoiclal, but the blastoderm almost completely encloses the yolk before any organ
develops on the germ, and no external yolk sac is formed.

The results of the investigations hitherto made with regard to the germinal
layers, the development of the inner organs, and the inner differentiations of the
outwardly visible organs are so contradictory and in many cases so incomplete,
that no description of them is here attempted. Further investigation is much
needed. The development of the eye has already been described (p. 171), and that
of the hind-gut and ink-bag was illustrated (p. 197).

Two important facts in the ontogeny of the Dibranchia should be noted. (1)
In considering the arms as parts of the foot, it is important to notice that they
arise behind the rudiments of the head, and only secondarily come to lie round and
below the latter. The mouth, at quite a late stage, lies at the anterior end of the
circle of arms (Fig. 222 C). (2) The funnel consists of two separate lateral rudi-
ments, the free edges of which fuse secondarily. This point is important in connec-
tion with the separation of the two lobes of the funnel, which lasts throughout life
in Nautilus. For the view of the funnel as epipodium, </. p. 116.

The fact that the velum is wanting in the Cephalopod embryo must also be
noted. The absence of this organ is explained by the direct development of the
Cephalopoda within the egg capsule at the expense of a large quantity of nutritive
yolk.

Investigations as to the development of the shell, and as to the nature of the
organ which has been called the shell gland, are much needed.

XXIV. Phylogeny.

No actual points of connection between the Molluscan phylum and any other
division of the animal kingdom have as yet been found ; the origin of the Mullusca
is therefore merely a matter of speculation. The present writer favours the view
that the Mollnsca descended from animals like the TurlcUaria, which had become
differentiated from the modern Platotlrs by the acquisition of a hind-gut and a heart,
and the (at least partial) transformation of the genital cavity into a secondary
and primitively paired body cavity. There is a striking agreement in the nervous
system of the lower Molluscs (Chiton, Solcnoyastres, and in some respects the Dioto-
cardia) and that of the Platodcs ; in both there is a ladder-like nervous system
with the principal trunks beset along their whole length with ganglion cells ; the

vii MOLLUSC A— LITERATURE 269

pleurovisceral cords answer to the lateral trunks of the Platodcs, and the pedal
cords to the ventral longitudinal trunks of the latter. If such a hypothetical racial
form were to secrete a dorsal shell, perhaps at first in the form of a thick cuticle
containing calcareous particles, a typical Molluscan organisation would be produced.
The development of a shell would deprive the greater part of the surface of the body
of its original respiratory function, and would lead to the formation of localised
gills. By means of the development of a mantle fold these delicate-skinned organs
could be brought under the protection of the shell. The musculature on the dorsal
side, which the shell covered, would disappear, and with it the dorsal longitudinal
nerve trunks. The musculature on the ventral side, which was already strongly
developed in the Planaria, would become strengthened in the development of the
foot with its sole for creeping. A part of the dorsoventral musculature would be
changed into the shell muscle.

In this derivation of the Moll n sea their characteristic larval form might be
explained, without any need for tracing it to the Annelidan Trocophora, in the
following way. It would correspond to a Turbellarian larva (Miiller's Polyclade
larva, etc. ), on to which certain Molluscan characteristics such as the shell gland,
the shell, the anus, and the foot had been shifted back. The preoral ciliated band
(the velum) of the Molluscan larva would correspond with the same structure in
the Turbellarian larva. The primitive kidney of the former would answer to a
simplified water vascular system, while the permanent nephridia as ovarial and
seminal ducts might be homologised morphologically with the ducts of the
genital products in the Turliellaria.

Review of the most important Literature.

Comprehensive Works. Text-Books. General ^Works. Investigations
treating of all or several Classes.

Boll. Beitrdge zur Vcrgleich. Histologie des Molluskentypus. Arch, fur mikr. Anat.

Supplcmcntband. 1869.
H. G. Bronn. Die Kl,asscn und Ordnungcn dcs TJiicrreiches. 3 Bd. Malacosoa.

I. Malacozoa accphala. 1862. II. Malacozoa cephalophora, von W. Keferstein.

1862-1866. (New edition now appearing, v. Simroth. )
G. Cuvier. M&nwires pour servir a Fhistoirc et a I'anatomie dcs Mollusqucs. Paris,

1817.
G. B. Deshayes. Traite elementaire de Conchyliologic. 3 vols. Paris, 1839-1857.

Histoire naturelle des Mollusqucs (Exploration de I'Algfric). 1848.
Eydoux and Souleyet. Voyage autaur du monde sur la corvette la Bonitc. Histoire

naturcllc. Zoologic. Paris, 1852.

Paul Fischer. Manuel de Conchyliologic et de Paleontologie conchyliologique. His-
toire naturelle des Mollusqucs mvants ctfossiles. 2 vols. Paris, 1887.
H. von Jhering. Vcrgleichcndc Anatomic des Ncrvensystems, und Phylogenic der

Molluskcn. Leipzig, 1877.

Keber. Bcitragc zur Anatomic und Physiologic der Weichthicre. Kijnigsberg, 1851.
E. Ray Lankester. Mollusca. Encyclopaedia Britannica. 9th edit. Vol. XVI.

1883.

R. Leuckart. Zoolocjischc Untersuchungen. Heft 3. Giessen, 1854.
Pelseneer. Introduction d I'etudc dcs Mollusqucs. Bruxelles, 1894.
Poli. Tcstacca utriusque Siciliae corumque historia ct anatome. 3 Bd. 1791-1795.

270 COMPARATIVE ANATOMY CHAP.

H. Simroth. Uebcr ciniyc Tayesfrayen dcr MalacozooJoyic, Muptsachlich Convcryen-
zcrschcinuntjc a b/'fre/end. Zcitschrift Natwrwiss. Halle. 62 Bd. 1889.

The early parts of the new edition of vol. iii. of Bronn's Klassen und

Ordaunyen des Thierreichs.

J. Thiele. Die Stammesverwandschaft der Mollusken. Ein beitray zur Phylogenieder
Tliirre. Jenaische Zeitschr. f. Naturwisscnsch. 25 Bd. 1891.

S. P. Woodward. A Manual of the Mollusca. 4th edit. 1880.

Amphineura.

L. Graff. Anatomic des Clmctodcrma nitidulum. Zeitschr. f. tcissen. Zooloyie.
26 Bd. 1876.

- Neomenia und Chaetoderma. Zeitschr. f. u-issen. Zooloyie. 28 Bd. 1877.

B. Haller. Die organisation dcr Chitoncn der Adria. Arb. aus dem Zool. Instil, in
JJ'ien. I. Theil. 4 Bd. 1882. II. The it. 5 Bd. 1883.

G. A. Hansen. Anatom. Beskrivelsc of Chcetoderma nitidulum. Nyt maya~. for
nat-urvidcnskab. Vol. XXII. 1877.

- - Neomenia, Proneomciiai vnd Chcetoderma. Bergen Mus. Aarster. f. 1888.

J. Heuscher. Zur Anatomic tend Histoloyie ron Proneomenia Sluiteri. Jena. Zeitzeh.

Vol. XXVII. 1893.
A. A. W. Hubrecht. Proncomcuin S/i/i/< ri. Niederl. Arch. Zool. Suppl. I. 1881.

- A Contribution to the Morpholoyy of Amphineura. Quart. Journ. Mia:
Science. Vol. XXII. 1882.

- Dondcrsia f estiva yen. et spec. nov. Danders Fcstbundel. Nederl. Tijdschr.
Gcncesk. 1888.

J. Koren and D. C. Danielssen. l)<^,-,-iptions of new species belonging to the

ttolenapus, with some observations on their organisation. Ann. Nat. Hist. (5).

Vol. III. 1879.
A. Kowalevsky and A. F. Marion. Contributions a I'histoire des Solenogastres ou

Aplacophores. Ann. Mus. H. N. Marseilles. Tome III. 1889.
A. Th. v. Middendorff. 7>iV/-«>' zu einer Malacozooloyia rossica. I. Seschrcibtnii/.

und Anatomic neuer oder far Russland neucr Chitoncn. Mem. de I' Academic

St. Petersboury. Tome VI. 1849.
G. Pruvot. Sur I' organisation dc quclqucs Ntomtniens des cotes de France. Arch.

Zool. cxper. 2U. Vol. IX. 1891.
A. Sedgwick. On certain paints in the anatomy of Chiton. Proceed. Roy. Soc.

No. 217. Dec. 1881.
Simroth. See aliove.
T. Tullberg. Neomenia, a new genus of invertebrate animals. Mvenska Vet. Al-«<l.

Hundl. Bd. III. 1875.
Axel Wiren. Studien iiber die Solenogastres. I. Monographic des Chcetoderma niti-

dulum Lort'n. Kongl. Sv-cnska Vetcnskaps - Akademicns Handlinyar. Vol.

XXIV. Stockholm, 1892.
In addition, works of Van Bemmelen, Call, Pelseneer, etc.

Gastropoda.

Alder and Hancock. A monograph of the British Nudibranchiate Mollusca.

London, 1850-1851.
R. Bergh. Beitraye zu einer Monographic der Polyccraden, I., II., III. fer-

handl. der k.' k. Zooloy. Botan, Gesellscliaft zu Wien. 29, 30, 33 Bd. 1879-

1883.

vn MOLLUSCA— LITERATURE 271

R. Bergh. Ueber die Verwandtschaftsbezichungcn dcr Onchidien. Morph. Jahrb. 10

Bd. 1884.
Report on the Nudibranchiata of the " Challenger" Expedition. ('hall. Report

Zool. Vol. X. 1884.

Die Titiscanien, cine Famil/r il/'r r/iipii/ni/fossoi Gastropodcn. With three

Plates. Morphol. Jahrb. 16 Bd. 1890.

Die Marseniadcn. Zool. Jahrb. 1 Bd. 1886 ; compare also Semper's 7iV /*<•,/

im Archipcl dcr Philippincn. 2 The if c. Wissensch. Jtesultate. Suppl. Heft 3.

1886.
Die cladohcpatischcn Nudibranchicn. Zool. Jahrb. Abth. fiir Systematise,

5 Bd. 1891.

— Die cryptobranchiaten Dorididcn. Zool. Jahrb. Abth. fiir Systematic, fi
Bd. 1891.

In addition, numerous monographs of various families, genera, and species
of the Opistliobranchia in various journals.

J. E. V. Boas, Spo/ia atlantica. Bidray til Ptcropodernes Morfoloyi og SystemaMk
samt til Kundskabcn om deres yeografiske Udbredclse. Danske Vid. Sclsk.
Skr. (6). 4 Bd. 1886.

— Zur Systematik und Biologic dcr Ptcropodcn. Zool. Jahrb. 1 Bd. 1886.

L. Boutan. Rcchcrchcs sur V anatomic et le devcloppcment de la Fissurcllc. Arch.

Zool. c.qrtr. (2). Tome III. 1886.
E. L. Bouvier. Systeme ncrvcux, morphologic, rjenerale ct classification des Gastro-

podcs prosobranches. Ann. des Sciences Nat. (7). Tome III. 1877.

E. Claparede. Anatomic und Entwickelungsgeschichte dcr Neritina fluviatilis.

Mailers Archil-. 1857.
P. Garnault. Rcchcrchcs anatomiqucs et histologiqucs sur le Gyclostoma clegans.

Arch. Soc. Linn. Bordeaux, 1887.

C. Gegenbaur. Unfersuchungen ilbcr Pteropoden und Hetcropodcn. Leipzig, 1853.
R. J. Harvey Gibson. Anatomy and physiology of Patella vnlgata. Part I. Ana-
tomy. Trans. Roy. Soc. Edinburgh. Vol. XXXII. 1887.
B. Haller. Uittcrsuchungcn ilbcr marine Shipidoglossen. I. Stitdie. Morph. Jaltrb.

9 Bd. 1883. //. Studie. 11 Bd. 1886.

Die Morphologic der Prosobranchier, yesammclt auf ciner Erdumsegehmg

durch die Konigl. ital. corvette " Vettor Pisani." I. Morph. Jahrb. 14 Bd.

1888. //. ibid. 16 Bd. 1890.
Huxley. On the morphology of the cephalous Mollusca as illustrated by the anatomy of

certain Hctcropoda and Ptcropoda, etc. Philos. Transactions. 1853.
J. Joyeux-Laffuie. Organisation et d&vcloppcment de I'Oncidie (OncJiidhun cclticum

Guv.). Arch. Zool. exper. Tome X. 1882.
H. de Lacaze-Duthiers. Histoirc ct monographic du Plcurobranchc orange. Annalcs

des Sciences Nat. (4). Tome XI. 1859.

Memoire sur la Pourpre. Annalcs des Sciences lf«f. (4). Tome XII.

1859.

— Memoire sur le systems nerveux de I'Haliotide. Ann. des Sciences Nat.
(4). Tome XII. 1859.

Memoire sur I'anatomie et rcmbryoyenic des Vermets. Ann. des Science*
Nat. (4). Tome XIII. 1860.

— - Histoirc de la Testacella. Arch. Zool. exper. (2). Tome V. 1888.
A. Lang. Vcrsuch ciner Erklarung dcr Asymmetric der Gastropodcn. Viertel-
jahrsschrift d. Naturf. Gescllsch. Zurich: 36 Bd. 1891.

F. Leydig. Ueber Paludina vivipara. Zeitschr. f. u-iss. Zool. 2 Bd. 1850.
Milne Edwards. Note sur la classification naturelle des Mollusques Gastropodes.

Ann. des Sciences Nat. 1848.

272 COMPARATIVE ANATOMY CHAP.

G. Moquin-Tandon. Recherches anatomiques sur I'ombrelle dc la nuditcrraniic. Ann.

dcs Sciences Nat. (5). Vol. XIV. 1875.

H. Miiller and C. Gegenbaur. Ueber Phyllirhoe bucephalum. Mull. Arch. 1858.

A. Nalepa. Beit rage zur Anatomic der Stylommatophorcn. Sitz.-Bcr. Akad. JVicn.

87 Bd. 1883.
Nordmann. Monographic de Tcrgipes Edwardsii. Mim. Acad. Imp. St. P&ersbourg.

Tome IV. 1843.
J. Paneth. Be it rage zur Histotngie der Ptcropoden und Hetcropodcn. Archiv. fur

mUcrosk. Anat. 24 Bd. 1884.
J. I. Peck. On the anatomy and histology of Cymbuliopsis calccola. 4 Plates.

Studies Biol. Labor. Johns Hopk. Univ. Vol. IV.
Paul Pelseneer. Report on the Ptcropoda collected by H. M.S. " Challenger" during,

etc. Parts I. II. III. Chall. Report Zool. Part LVIII. 1887 ; Part LVI.

1888 ; Part LXV. 1888.

— The cephalic appendages of the gymnosomatous Pteropoda, and especially
of Clione. Quart. Jount. Microsc. Science (2). Vol. XXV. 1885.

L. Plate. Studien iiber opisthopneumone Lungcnschneckcn. I. Daudebardia und
Testacc.lla. Zool. Jahrbiicher. Abth. fur Anatomic und Ontogenie. 4 Bd. 1891.

Quatrefages. M&moire sur les Gastropodcs phlebcnt&res. Ann. des Sciences Nat.
Tomes III. and IV. 1844 and 1845.

Rang and Souleyet. Histoire naturclle dcs Mollusques Pt6ropodcs. Paris, 1852.

B. Sharp. Beitrdge zur Anatomic von Ancylusjiuviatilis (0. F. Mull] und Ancylus

lacustris (Geoff roy). Inaug. -Dissert. "Wurzburg, 1883.

H. Simroth. Ueber die Bewegung und das Beivegungsorgan des Cyclostoma elegans
und der einheimischen SchnecTcen uberhaupt. Zeitschr. f. Wiss. Zool. 36 Bd.
1881.

— Vcrsuch einer Xaturgcschichte der deufschcn Nacktschnecken und Hirer
curopiiischen Venrandtcn. Zeitschr. f. wiss. Zoologie. 42 Bd. 1885.

In addition, many other treatises on the Pulmonata in various journals.
S. Trinchese. Mater iali per la fauna mur/fiimi IfuJiana. Aeolididae c famiglie

aft in. Attiaccad. Ltiicci (3). Mem. Arol. XL 1883.
Troschel. Beitrdge zur Kenntniss der Pteropoden. Arch. f. Naturg. Tome XX.

1854.
M. Vayssiere. Recherches anatomiques sur les Mollusques de lafamille des Bullides.

Ann. Hist. Nat. Zool. (6). Tome IX. 1880.

Piceherches anatomiques sur les genres Pclta (Runcina) ct Tylodina. Ann.

dcs Sciences Nat. (6). Tome XV. 1883.

Recherches zoologiques et anatomiques sur les Mollusques opisthobranches du

golfc de Marseilles. Part I. Tect (branches. Ann. Mus. Hist. N. Marseilles.
Tome II. Mem. 3. 1885. Part II. Ibid. Tome III. Mem. No. 4. 1888.

Nicolas Wagner. Die u'irbellosen Thierc dcs weissen Mceres. 1 Bd. Leipzig. Fol.

1885.
H. Wegmann. Contribution a I'histoirc naturclle des Haliotides. Arch. Zool. cxpe'r.

(2). Tome II. 1884.
Note sur T organisation de la Patella vulgata L. Recueil. Z. Suisse. Tome

IV. 1887.
Emile Yung. Contributions a I'histoire physiologiquc, de, Vescargot (Helix pomatia).

M6m. Cour. Acad. Bclg. Tome XLIX. 1887.

Scaphopoda.

Herm. Fol. Sur I'anatomie microscopiquc du Dcntalc. 4 PL Arch. Zool. cxpir. (2).
Tome VII. 1889.

vii MOLLUSC A— LITERATURE 273

H. de Lacaze-Duthiers. Hisioire de I' organisation et du developpement du Dcntale.

Ann. des Sciences Nat. (4). Tomes VI., VII., and VIII. 1856-57-58.
L. Plate. Bemerkungen zur Organisation der Dentalien. Z. Anzeiger. 11 Jahrg.

1888. Ueber das Hers der Dentalien. Ibid. 14 Jahrcj. 1891.
M. Sars. Om Siphonodentalium vitreum, etc. Christiania, 1861.

Lamellibranchia.

Ernst Egger. Jouannetia Cumingi Son. Einc morphol. Untersuchung. Arbeit.

Zool. Instit. Wurzlurg. 8 Bd. 1887.
Garner. On the anatomy of the Lamellibranchiate Conchifera. Transact. Zool. Soc.

London, Vol. II. 1841.
H. de Lacaze-Duthiers. M6moire sur I' organisation de I'Anomie. Ann. des Sciences

Nat. (4). Tome II. 1854.

Morphologic des Acephales. 1 Mem. Anatomic de I'Arrosoir (Aspergillum

dichotomum). Arch. Zool. ci-per. (2). Tome 1. 1883.

Leydig. Anatomic und Entwickelung von Cyclas. Muller's Archiv. 1835.

H. A. Meyer and Moebius. Fauna der Kieler Bucht. Leipzig, 1865.

Paul Pelseneer. Report on the anatomy of the deep-sea Mollusca collected by H.M.S.

"Challenger." Report Chall. Zool. Part 74. 1888.

Contribution a I'etude des Lamellibranchcs. Archives de Biologic. Tome XI.

1891.
A. de Quatrefages. Memoire sur le genre Tarct. Ann. des Sciences Nat. (3). Tome

XL 1849.

Cephalopoda.

A. G. Bourne. Tlte differences between the males and females of the pearly Nautilus.

Nature. Vol. XXVIII. 1883.

J. Brock. Studicn ubcr die VerwandtscIiaftsi'crJialtnisse der dibranchiaten Cephalo-
poda n. Erlangen, 1879.

— J'ersuch einer Phylogenic der dibranchiaten Ccphalopoden. Morph. Jahrbuch.
6 Bd. 1880.

Zur Anatomic und Systematik der Ceplialopodcn. Zeitschr. f. wiss. Zool. 36

Bd. 1882.

Delle Chiaje. Memorie su' Cefalopodi. Memorie sulla storia e notomia degli ani-

mali scnza vertebre del regno di Napoli. Napoli, 1829.
Ferussac and d'Orbigny. Histoire naturelle gdndrale et particuliere des Cephalopodes

acetabulifercs vivants etfossiles. Paris, 1835-1845.
Leon Fredericq. Recherches sur la physiologic du Poulpe commuii (Octopus vulgaris).

Arch. Zool. expdr. Tome VII. 1879.
Carl Grobben. Zur Kenntniss der Morphologic und Verwandtschaftsi'crhdltnisse der

Cephalopoden. Arb. Z. Inst. Wien. 7 Bd. 1886.

B. Haller. Beitrcige zur Kenntniss der Morphologic Nautilus Pompilius. Zool. d.

Semon's Forschungsreise in Australicn. Vol. V. 1895.
H. von Jhering. Ueber die Verwandtschaftsbczichungcn der Cephalopoden. Zeitschr.

/. wiss. ZooJ. 35 Bd. 1880.

Van der Hoeven. Beitrag zur Kenntniss von Nautilus. Amsterdam, 1856.
Will. E. Hoyle. Observations on the anatomy of a rare Cephalo'pod (Gonatus Fabricii).

Proc. Z. Soc. London, II. 1889.
H. Miiller. Ueber das Mdnnchen von Argonauta argo und die Hcdocotylcn. Zeitschr.

f. wiss. Zool. 1855.
R.Owen. Memoir on the pearly Nautilus, etc. London, 1832.

VOL. II T

274 COMPARATIVE ANATOMY CHAP.

R. Owen. Description of some new and rare Cephalopoda. Trans. Zool. Soc. London.

Vol. II. 1841.

Cephalopoda. Todd's Cyclopaedia, etc. Vol. I. London, 1836.

— Supplementary Observations on the Anatomy of Spirilla australis Lam. Ann.

of Nat. Hist. (5). Vol. III. No. 13. 1879.
J. B. Verany. MoUusqucs mediterraneans observes, dicrits, figures ct chromolitho-

graphies d'apres le vivant. Part I. Cephalopodes de la Mediterranec. Genes,

1847-1851.
Verany and Vogt. Memoires sur Us Hectocotyles, etc. Ann. des Sciences Nat.

Tome 17. 1852.
W. J. Vigelius. Untcrsuchungen an Thysanotcuthis rhombus Frosch. Ein Beitrag

zur Anatomic, der Gephalopoden. Mitth. Zool. Station zu Neapel. 2 Bd. 1880.
F. Ernest Weiss. On some Ipigopsid cuttle-fishes. Quart. Journ. Micr. Science.

Vol. XXIX. 1889.

Treatises on Single Organs or Groups of Organs, — Integument, Mantle, Shell,

Oral Lobes.

Fe"lix Bernard. Recherchcs sur les organes palltaux des Gastro}}odcs prosobranches.

These. Paris, 1890. Also in Annales des Sciences Nat. VII. 1889.
F. Blochmann. Ueber die Driisen des Mantelrandes bci Aplysia und verivandten

Formcn. Zcitschr. f. wisscnsch. Zool. 38 Bd. 1883.
Jos Blumrich. Das Integument der Chitonen. Mit ciner Vorbemerkung von Prof.

Hatschek. Zcitschr. f. wiss. Zoologie. 52 Bd. 1891.
Bowerbank. On the structure of the shells of molluscous and conchifcrous animals.

Trans, of Micr. Soc. I. London, 1844.
W. Carpenter. On the microscopic structure of shells. Report British Assoc. 1843

and 1847. London, 1844 and 1848.
E. Ehrenbaum. Untersuchungen uber die Structur und Bildung der Schale der in

der Kieler Bucht hdufig vorkommenden Muschcln. Zcitschr. f. wiss. Zool.

41 Bd. 1884.
P. Girod. Recherchcs sur la peau des Cephalopodes. Arch. Zool. experiment (2).

Tome I. 1883.
H. Meckel. Mikrographie einiger Driisenapparate der niederen Thierc. Mutter's

Archiv. 1846.
Felix Miiller. Ueber die Schalenbildung bei Lamcllibranchiaten. Zool. Beitrdgc

Schneider. 1 Bd. 1885.
R. Owen. On the relative positions to their Constructors of the chambered shells of

Cephalopods. Proc. Zool. Soc. London. 1878.
Bernhard Rawitz. Der Mantelrand der Accphalen. 1. Ostracca. Jenaischc

Zcitschr. f. Naturwiss. 22 Bd. 1888. 2. Arcacca, Mytilacca, Unionacca. Ibid.

24 Bd. 1890.
G. Steinmann. Vorlaufigc Mitthcilung uber . die Organisation der Ammonitcn.

Bcr. Nat. Gcs. Freiburg. 4 Bd. 1889.

G. Steinmann and L. Doderlein. Elemente der Palciontoloyie. Leipzig, 1890.
Johannes Thiele. Die Mundlappcn der Lamcllibranchiaten. Zcitschr. f. wiss.

Zoologie. 44 Bd. 1886.
T. Tullberg. Studien uber den Bau und das Wachsthum des Hummcrpanzcrs und.

ilf r MoUuskcnschalen. Konyl. Svcnsk. Vetensk. Akad. Handling. 19 Bd.
1882.

Karl A. Zittel. Handbuch der PaJdontologie. I. Abth. Palaozoologie. II. Band.
Mollusca und Arthropoda. Miinchen and Leipzig, 1881-1885.

vii MOLLUSCA— LITERATURE 275

Musculature, Foot, Pedal Glands, the Taking in of Water,

Th. Barrois. Los glandes du pied et les pores aquiferes des Lamcllibranches. Lille,

1885.
J. Carriere. Die Driisen im Fusse dcr Lamellibranchiaten. Arbeit, aus d. zool.

List it ut JFiirzburg. 5 Bd. 1879.
Die Fussdrusen der Prosobranchier und das Wasscrgcfass-system der Lamcl-

libranchier und Gastropoden. Archiv. f. mikrosk. Anatomic. 21 Bd. 1882.
C. Grobben. Zur Morphologie dcs Fusses der Hcteropoden. Arb. Zool. lust. Wien.

7 Bd. 1887.
A. Fleischmann. Die Beivegung des Fusses der Lamettibranchiaten. Zeitschr. f. wiss.

Zoologic. 42 Bd. 1885.
Georg Kalide. Beitrag zur Kenntniss dcr Musculatur der Hcteropoden und Ptero-

podcn, zugleich ein Beitrag zur Morphologie des Moll ' uskenfusses. Zcitschr.f. wiss.

Zool, 46 Bd. 1888.
J. H. List. Zur Kenntniss der Driisen iin .Fusse von Tethys ftmbriata L. Zcitschr.f.

wiss. Zool. 45 Bd. 1887.
Paul Pelseneer. Sur la valeur morphotogique des bras et la composition du systeme

nervcux central dcs Ctyhalopodcs. Arch. Biol. Tome VIII. 1888.

Sur I'epipodium des Mollusques. Bull. Scicntif. France ct Belg. Tome 19.

1888. Tome 22. 1890. Tome 23. 1891.
Bernhard Rawitz. Die Fussdriise der Opisthobranchier. Abhandl. Akad. Berlin.

1887.
Ludwlg Reichel. Ueber die Bildung dcs Byssus dcr Lamellibranchiatcn. Zool.

Bcitrage, Schneider. 2 Bd. 1888.
P. Schiemenz. Ueber die Wasseraufnahme bci Lamellibranchiaten und Gastropoden

(ciiischliesslich Ptcropodcn}. Milth. Zool. Station Neapel. 5 Bd. 1884. 2

Theil. 7 Bd. 1887. This paper . contains the further literature on this

subject.
Jap. Steenstrup. Hectocotijlus dannelsen has Octoiwds, etc. K. Dansk. Vidcnsk.

Selskabs Skrifter. 1856.

Nervous System.

L. Bohmig. Beit-rage zur Kenntniss des Centralnervensystcms einiger pulmonaten

Gastropoden. Inaug.-Diss. Leipzig, 1883.
E. L. Bouvier. Systeme nervcux, morphologic generale et classification des Gastropodcs

prosobranchcs. Annales des Sciences Nat. (7). Tome III. 1887.
Louis Boutan. Contribution d I'etude de la masse nerve-use vent rale (cordons palleo-

visceraux} et de la collerette de la Fissurelle. Arch. Zool. cxperim. (2). Tome

VI. 1889.
J. Brock. Zur Neurologic, der Prosobranchier. Zcitschr. f. wiss. Zool. 48 Bd.

1889.
0. Biitschli. JBemerkungen uber die wahrschcinlichc Herleitung der Asymmetrie der

Gastropoden, spec, der Asymmetrie im Nervensystem der Prosobranchier. Morph.

Jahrb. 12 Bd. 1886.
Cheron. Recherches sur le systeme nerveux des Ciphalopodcs dibranchianx. Annales

des Sciences Nat. (5). Tome V. 1866.
Karl Drost. Ueber das Nervcusystem und die Sinncserjithelien der Hcrzmuschel

(Cardium edule), etc. Morph. Jahrbuch. 12 Bd. 1886.
Duvernoy. Memoires sur le systeme nerveux des Mollusques actphales. MAmoires

de I'Acad&mic des Sciences. Tome XXIV. 1854.

276 COMPARATIVE ANATOMY CHAP.

B. Haller. Zur Kcnntniss der Muricidcn. Eine vergl.-anat. Studic. I. TJieil.

Anatomie dcs Nervensysterns. Denkschr. math.-naturw. Klasse Akad. Wissensch.
Wien. 45 Bd. 1882.

— Untersuchungen uber marine Ehipidoglossen. II. Textur des Central-nerven-
systems und seiner Hiillen. Morpli. Jahrb. 11 Bd. 1885.

H. von Jhering. V&rgleichende Anatomie des Nervensystcms und Phylogenie der

Mollusken. Leipzig, 1877.
Lacaze-Duthiers. Du systeme ncrrcux dcs Mollusques gastropodes pulmones aqua-

tiques. Arch, dc Zool. cxp. Tome I. 1872.
Bernhard Rawitz. Das centrale Nervensystem der Acephalen. Jenaische Zeitschr. f.

Naturwiss. 20 Bd. 1887.
Paul Pelseneer. Sur la valeur morphologique des bras et la composition du systeme

nervcux central des C6phalopodes. Arch. Biol. Tome VIII. 1888.

— Rccherches sur Ic systeme nerveux des Pteropodes. Arch. Biol. Tome
VII. 1887.

C. Semper. Ueber Sehorgane vom Typus der JVirbelthieraugen. Wiesbaden, 1877.
H. Simroth. Das Fussnervensystem von Paludina mvipara. Zeitschr. f. wiss. Zool.

35 Bd. 1880.

Ueber das Ncrveusystcm und die Bcwcgung der dcutschen Binncnschnccken.

Progr. d. Ilcalschule. 2 Ordni/nrj. Leipzig. No. 503. 1882.

J. W. Spengel. Die Gcruchsorgane und das Ncrvcnsystem der Mallusken. Zeitschr.
f. wiss. Zool. 35 Bd. 1881.

Sensory Organs.
F61ix Bernard. R,echcrches sur les organcs palleaux des Gastropodes prosobranches.

Ann. des Sciences Nat. (7). Tome IX. 1890. Containing researches on the

osphradia of the Gastropoda.
J. Brock. Ueber die sogcnannf.cn Amjcn ro>i Trlilacnn und das I'orl.-n //////<•// ro/i

PseudoMoropliyllkorpern im Gcf ass- system der Muschdn. Zeitschr. f. iciss. Zool.

46 Bd. 1888.
0. Biitschli. Notiz zur Morphologic des Aurjes der Muscheln. Fcstschr. 5QQ-jahriff.

Bestehens der Piuperto-Carola dargeb. r. Nat.-Med. Ver. Heidelberg. Xat. Theil.

1886.
Justus Carriere. Die Sehorgane der Thicre vergleichend - anatomisch dargestcllt.

Miinchen and Leipzig, 1885.

Ucler MoUuskenangcn. Arch. f. mikrosk. Anat. 33 Bd. 1889.

C. Claus. Das Geharorgan der Seteropoden. Arch. f. mikrosk. Anat. 12 Bd. 1875.
P. Fraisse. Ueber Molluskcnaugen mit embryonalcm Typus. Zeitschr. f. wiss. Zool.

35 Bd. 1881.
W. Flemming. Ci//''/^nchungen uber Sinnesepithelien der MnJln*ki'n. Arch. f. mikr.

Anat. Tome ATI. 1870.
H. Grenacher. Abhandlungen zur vergleichenden Anatomic des Auges. I. Die

Retina der Ccphalopoden. Abhandl. Xaturf. Gesellsch. z. Halle. 16 Bd. 1884.

//. Das Auge der Heterojjodcn. Ibid. 17 Bd. 1886.
V. Hensen. Ueber das Auge einigcr Cephalophoren. Zeitschr. f. wiss. Zool.

15 Bd. 1865.
C. Hilger. Beitrdge zur Kenntniss des Gastropodcnauges. Morpli. Jahrb. 10 Bd.

1884.
Lacaze-Duthiers. Otocystm on capsules /m/Htircs des Mollusques (Gastropodes).

Arch. d. Zool. cxp. Tome I. 1872.
E. Ray Lankester and A. G. Bourne. 'Jn the existence of Spengel's olfactory organ

and of paired genital ducts in the pearly Nautilus. Quart. Jour. Micr. Science.

Vol. XXIII. 1883.

vii MOLLUSC A— LITERATURE 277

F. Leydig. Ueber das Gchororgan der Gastropoden. Archiv. f. mikrosk. Anatomic.

7 Bd. 1871.
H. N. Moseley. On the presence of eyes in the shells of certain Chitonidce and on

the structure of these organs. Quart. Journ. Micr. Science (2). Vol. XXV.

1885.
Ph. Owsjannikow and Kowalevsky. Ueber das Centralorgan und das Gchororgan

der Cephalopoden. St. Petersburg, 1867.
W. Patten. Eyes of Molluscs and Arthropods. Mitth. Zool. Stat. Ncapel. 6 Bd.

1886.
Paul Pelseneer. Sur I'ceil de quclqucs MoUusques gastropodcs, et Les organcs dcs

scns^chcz Ics MoUusques. Annales Socie'te: Beige Microsc. (Mdmoires). Toine XVI.

1891.

Rawitz. See above under heading Integument.
P. B. Sarasin. Ucbcr drei Sinncsorganc und die Fussdruse cinigcr Gastropoden.

Arbeit. Zool.-zoot. Inst. IVurtzburg. 6 Bd. 1883.

B. Sharp. I>H the visual organs in Lamcllib ranch iata. Mitth. Zool. Stat. in Ncapel.

5 Bd. 1884.
D. Sochaczewer. Das Ricchorgan dcr Landpulmonaten. Zcitschr. f. iviss. Zool.

35 Bd. 1880.
J. E. Tenison-Woods. On the anatomy and life history of Mollusca peculiar to

Australia. Proc. Roy. Soc. N.S. Wales. Vol. XXII. 1889.
Johs. Thiele. Die abdominalcn Sinncsorgane dcr Lamellibranchier. Zeitschr. f.

wiss. Zool. 48 Bd. 1889.

Intestine, Ink-bag.

D. Barfurth. Ueber den Bau und die Thdtigkcit der Gastropodcnlebcr. Arcliiv.

f. mikr. Anatomic. 22 Bd. 1883.
Th. Barrois. Lc stylet crystaUin dcs Lamellibranches. Revue biol. du Nord de la

France. Tome II. 1890.
Em. Bourquelot. Rechcrches sur Ics jMnomenes de la digestion chcz Ics MoUusques

cfyhalopodcs. Arch, de Zool. cxp. (2). Tome III. 1885.
J. Frenzel. Mikrographie der Mitteldarmdruse (Leber) der Mollusken. I. Allgc-

mcinc Morphologic und Physiologic dcs Druscncpitliels. Nova acta Acad. Goes.

Leop.-Carol. 48 Bd. 1886.
Heinrich Maria Gartenauer. Ueber den Darmkanal einiger cinhcimischcn Gastro-

podcn. Inaug-Diss. Strassburg, 1875.
Patrick Geddes. On the mechanism of the odontophore in certain Mollusca. Trans.

Zool. Sue. London. Vol. X. Part II. 1879.
Paul Girod. Rechcrches sur la pochc du noir des Cephalopodes dcs cotes de France.

Arch, dc Zool. cxperim. Tome X. 1882.
Macdonald. General classification of the Gastropoda. Trans, of the Linn. Soc. of

London. Tome XXIII. 1860.
Panceri. Gli organi e la secrezione dell' acido solforico nei Gastropodi con un appen-

dice, etc. Atti della R. Accad. delle scienze fisiche. Tome IV. 1869.
Bossier. Die Bildung dcr Radula bei den ccphalophorcn Mollusken. Zcitschr. f.

wiss. Zool. Bd. XLI. 1885.

C. Semper. Zumfcineren Bau der Molhiskenzunge. Zeitschr. f. iviss. Zool. 9 Bd.

1868.

H. Troschel. Das Gebiss der SchnecJcen. 1 Bd. Berlin, 1856-1863.

W. J. Vigelius. Vergleichcnd-anatomische Untcrsuchungen iiber das sogenanntc Pan-
creas der Cephalopoden. Verhandl. k. Akad. H'/'toisch. Amsterdam. Dccl 22.
1881.

278 COMPARATIVE ANATOMY CHAP.

Respiratory Organs, Circulatory System.

Felix Bernard. Rcchcrches sur les organcs palUaux des Gastropodes prosobmnchcR.

These. Paris, 1890.
Bojanus. Ueber die Athcm- und Kreislaufswerkszeuge der zweischaligcn Musclieln.

Isis, 1817, 1820, 1827.
L. Cuenot. Etudes sur le sang et les glandes lympJiatiqucs dans la stric animale.

2Partic. Invertebres. Arch, de Zool. experim (2). Vol. IX. 1891.
Carl Grobben. Uder den bulbits arteriosus und die Aortcnklappen der Lamelli-

branchiatcn. Arbeiten a. d. Zoologischen Institute der Universitdt Wicn. 9 Bd.

1891.
W. A. Herdmann. On the structure and function of tJie ccrata or dorsal papillae in

some Nudibranchiatc Mollusca. Quart. Journ. Microsc. Science. Vol. XXXI.

Part I. 1891.
L. Joubin. Structure et developpcment de la branchie de quclques Cephalopodcs des

cotes de France. Arch, de Zool exp&rim. (2). Vol. III. 1885.
Langer. Ueber das Gcfdss-systcm der Teichmuschcl. Denkschriften der Wiener Aka

dcmif. 1855 and 1856.
A. Menagaux. Eccherclies sur la circulation des Lamellibranches marins. Bcsancon,

1890.
K. Mitsukuri. On the structure and significance of some aberrant forms of Lamelli-

branchiate giUs. Quart. Journ. Microsc. Science. N. S. 21. 1881.
H. L. Osborn. On the gill in some forms of prosobranchiate Mollusca. Stud. biol.

labor. J. Hopkins Univ. Vol. III. 1884.
R. Holman Peck. The structure of the Lamcllibranchiate gill. Quart. Journ. Micr.

Science. Vol. XVII. 1877.
C. Posner. Ueber der Bau der Najadenkicmc. Arch. f. mikrosk. Anat. Bd.

XI. 1875.

Secondary Body Cavity, Nephridia, Genital Organs.

Baudelot. Rcchcrchcs sur I'apparcil g6ner. des Mollusques gastropodcs. Ann. Sci.

Nat. (4). Tome XIX. 1862.
Th. Behme. Beitrdge sur Anatomic und Entwickdunyscjeschichtc des Harnapparates

der Lungenschnccken. Arch. Naturg. Jahrg. 55. 1889.
J. Brock. Ueber die Gcschlcchtsorgane der Cephalopodcn. Erster Beitrag. Zeitschr.

f. wiss. Zool. 32 Bd. 1879.
J. T. Cuningham. Tht renal organs (ncphridia) of Patella. Quart. Journ. Micr.

Science. Vol. XXIII. 1883.

— Note on the structure and relations of the kidney in Aplysia. Mith. Zool.
Station in NcapcL 4 Bd. 1883.

R. von Erlanger. On the paired Nrjiltriilfn of the Prosobranchs, the homologics of
the only remaining ncphridium of most Prosobranchs, and the relations of the
ncphridia to the gonad and genital duct. Quart. Journ. Micr. Science. Vol.
XXXIII. 1892.

C. Grobben. Morpholog. Studicn ubcr den Harn- und Gcschlechtsapparat, sowie die
Leibcshohle der Ceimalopodcn. Arb. Zool. Inst. Wicn. 5 Bd. 1884.

Ueber die pcricardialdrusc der Lamellibranchiatcn. Ein Bi'itrui/ ~vr Kcnnt-

niss der Anatomic dicscr MoU.iiskenklasscn. Arb. Zool. Inst. Wicn. 7 Bd.
1888.

— Die pcricardialdruscn der Gastropodcn. Arbeit. Zool. Inst. der Univ. Wien.
9 Bd. 1890.

vii MOLLUSCA— LITERATURE 279

A. C. Haddon. On the generative and urinary ducts in Chiton. Proceed. Royal

Dublin Soc. (2). Vol. IV. 1885.

B. Haller. Beitrage zur Kenntniss dcr Niere der Prosobranchicr. Morph. Jahrb.

11 Bd. 1885.
A. Hancock. On the structure and homologies of the renal organ in the Molluscs.

Trans, of the Linn. Soc. Vol. XXIV.
P. P. C. Hoek. Lcs organcs de la generation de I'huitre. Tijdschr. Ncderl. Dicrk.

raven. Suppl. D. 1. 1883.
H. von Jhering. Ueber den uropneustischcn Apparat der Hcliccen. Zeitschr. f. wiss.

ZmL 41 Bd. 1884.
J. Kollmann. Ueber Verbindungen zioischcn Colom und Nephridien. Baseler

Festschrift zum Wurtzburger Jubilaum. 1882.
A. Kowalevsky. Ein Beitrag zur Kenntniss der Excretionsorgane. Biol. Centralblatt.

9 Bd. 1889.
E. Ray Lankester. On the originally bilateral character of the renal organs of

Prosobmnchia, and on the homologies of the yolk sac of Cephalopoda. Ann. of

Nat. Hist. (5). Vol. VII. 1881.
— Observations on the Pondsnail, etc. Quart. Journ. Micr. Science. Vol. XIV.

1874.
E. Ray Lankester and A. G. Bourne. On the existence of Spengel's olfactory organ

and of paired genital ducts in the pearly Nautilus. Quart. Journ. Micr.

Science, Vol. XXIII. 1883.
G. F. Mazarelli. Intorno all' anatomla dell' apparato riproduttore delle Aplysiai del

golf<> <H Napoli. Z. Anz. 12 Bd. 1889.
. Intorno all' apparato riproduttore di alcuni Tectibranchi (Pleurobranchcea,

Oscanius, Accra). Zool. Anz. 14 Jahrg. 1891.
O. Niisslin. Beitrage zur Anatomic und Physiologic der Pulmonatcn. Habilita-

tionsschrift (Carlsruhe). Tubingen, 1879.
R. Owen. On the external and structural characters of the male Spirilla australis.

Proceed. Zool. Soc. London. 1880.
Remy Perrier. Eecherchcs sur I' anatomic ft Vhistologie du rein des Gaslropodes

prosobranchcs. Annales des Sciences Nat. (7). Tome VIII. 1890.
Walter Rankin. Ueber das Bojanus'schc Organ der Teichmuschel (Anodonta cygnea

Lain.) Jenaische Zeitschr. fur Naturwissensch. 24 Bd. 1890.
A. Schmidt. Dcr Gfeschlechtsappamt der Stylommatophoren, etc. Abh. des Nat.

lrf reins fur Sachsen und Thuringen. 1 Bd. 1885.
P. StepanofiF. Ueber Geschlechtsorgane und Entwickelung von Ancyhis fluviatilis.

St. Petersburg, 1886.
W. J. Vigelius. Bijdragc tot de Kennis van hct cxcrctorisch Systecm der Cephalopoden.

Acad. Procfschrift. Leiden, 1879.
Ueber das excretionssystem der Ceplmlopoden. Niederl. Arch. f. Zool. 5 Bd.

1880.

Parasitic Gastropoda.

Albert Bauer. Beitrage zur Naturgeschichte dcr Sijnapta. III. Die Eingcweide-

schncckc in der Lcibeshohle der Synapta digitata. Nova Acta Academ. Cces.

Leap-Carol. Tome XXXI. 1864.
Max Braun. Ueber parasitische Schnccken. Zusammenfassender Bericht im Cen-

trulbl. f. Baktcriologie u. Parasitenkunde. 5 Bd. 1889.
Johannes Miiller. Ueber Synapta digitata und die Erzeugung von Schnecken in

HoJothurien. Berlin, 1852.
Paul and Fritz Sarasin. Ueber zwei parasitische Schnecken. Ergcbn. Naturw. Forsch.

auf Ceylon in 1884-1886. 1 Bd. Wiesbaden, 1887.

280 COMPARATIVE ANATOMY CHAP.

P. Schiemenz. Parasitische Schneckcn. Kritisches Rcferat. Biol. Ccntralblatt.

9 Bd. 1889-1890.

Walter Voigt. Entocolax Ludwigii, tin neucr scltsamer Parasit aus einer Holothurie.
Zeitschr. f. wiss. Zool. 47 Bd. 1888.

Ontogeny.

F. Blochmann. Ueber die Entivickelung von Neritina fluviatilis, Miill. Zeitschr.

f. wiss. Zool. 36 Bd. 1881.

— Beitrdge zur Kenntniss dcr Entivickelung der Gastropoden. Zeitschr. f. wiss.
Zool. 38 Bd. 1883.

W. K. Brooks. The development of the Squid (Loligo Pealii, Lesueur). Annivers.

Mem. Boston Soc. Nat. Hist. Boston, 1880.
R. von Erlanger. Zur Entwickelung von Paludina vivipara. I. and II. Mor-

pholoyisches Jahrbuch von Gegenbauer. 17 Bd. 1891.
Hermann Fol. Etudes sur le devclopjxment des Mollusques. I. Sur le dtveloppement

des Pteropodes. Archives de Zool. experim. Tome IV. 1875. //. Sur If

developpeinent embryonnaire et larvaire des Hetcropodes. Tome V. 1876. ///.

Sur le devcloppemcnt des Gastropodes pulmone. Tome Till. 1879-1880.
H. Grenacher. Zur Entwickelungsgeschichte dcr Ccphalopoden, zugleich ein Bcitrag

zitr Morphologic der hoheren Molluskcn. Zeitschr. f. wiss. Zool. 24 Bd. 1874.

A. C. Haddon. Notes on the development of Mollusca. Quart. Journ. Micr. Science.

Vol. XXII. 1882.

B. Hatschek. Ueber Entwickelungsgeschichtc von Teredo. Arb. a. d. Zool. Instit.

Universitdt Wien. Tome III. Heft 1. 1880.
R. Horst. Embryogtnic de I'huitre. Tijdschr. Ncdcrl. Dierk. Vcr. Suppl. Deel. 1.

1884.

Development of the European, Oyster. Quart. Journ. Micr. Science. Vol.

XXII. 1882.

A. Kolliker. Entu-ickelungsgeschichte der Cephalopoden. Ziirich, 1884.
A. Kowalevsky. Etude sur Tembryogenie du Dentalc. Annales du Musee d'histoirc

naturelle de Marseilles. Zoologie. Tome I. 1883.

— Embryogenie du Chiton Pol it (Philip2n) avec quclqucs rcmarqucs sur le
devcloppement des autres Chitons. Ann. Mus. N. H. Marseille. Tome I. No. 5.

A. Krohn. Beitrdge zur Entwickelungsgeschichte der Ptcropoden und Heteropoden.
Leipzig, I860.

E. Ray Lankester. On the developmental liistory of the Mollusca. Philos. Transact.
London. 1875.

Observations on the development of the Cephalopoda. Quart. Jour. Micr. Science.

Vol. XV. N.S. 1875.

S. Loven. Beitrdge zur Kenntniss der Mollusca acephala lamellibranchiata. Stock-
holm, 1879.

J. Playfair MacMurrich. A contribution to the embryology of the prosobranch
Gastropods. Stud. Biol. Lab. J. Hopkins Univ. Arol. III. 1886.

William Patten. The embryology of Patella. Arbeit. Zool. Inst. Wien. 6 Bd.
1885.

G. Pruvot. Sur le developpcment d'un Solenogastre. Comptes rend. Paris. Tome

CXI. 1890.

Carl Rabl. Ucbcr die Entivickclung der Tellerschnecke. Morph. Jahrb. 5 Bd. 1879.
— • Die Ontogenie der Sussivasscrpulmonaten. Jenaische Zeitschrift. 9 Bd.
1875.

— Ueber die Entwickelungsgeschichte dcr Malcrmuschcl. Jenaische Zeitschrift.

10 Bd. 1876.

vii RHODOPE VERANII 281

W. Salensky. lOtudes sur le developpcment du Vermct. Arch. Biol. Tome VI.

1837.
— Beitrdge zur Entwickelungsgeschichtc der Prosobranchicr. Zeitschr. f. wiss.

Zool. 22 Bd. 1872.
P. B. Sarasin. Entwickelungsgcschichte der Bithynia tentaculata. Arb. Zool-Zoot.

Instit. Wiirtzburg. 6 Bd. 1882.
Paul and Fritz Sarasin. Aus der Entwickclungsgeschichte von Helix Waltonii.

Ergcbn. Nat. Forsch. Ceylon, 1884-1886. 1 Bd. Wiesbaden, 1888.
P. Schiemenz. Zusammenfassende Darstcllung der Beobachtungen von Eisicj,

Rouzaud, Jourdain, Brock, etc., ilber die Entwickelung der Genitalorgane der

Gastropodcn. Biol. Centralblatt. 7 Bd. 1888.
C. Schierholz. Uebcr Entwickelung der Unianiden. Denkschr. Akad. Wien. 55

Bd. 1888.
F. Schmidt. Beitrag zur Kenntniss der postcmbryonalen Entwickelung der Najaden.

Arcltiv. fiir Naturgeschichte. 51 Jahrg. 1885.
M. Ussow. Untersuchungen uber die Entwickelung der Cephalopodeii. Arch. Biol.

Tome II. 1881.
L. Vialleton. Recherches sur hs premiere phases du developpement de la Seiche.

Annul. S'c. Nat. (7). Tome VI. 1888.
Wladimir Wolfson. Die embryonale Entiuickelung des Lymnceus stagnalis. Bullet.

Acad. Imp. Sc. St. Petersbourg. 26 Jahrg. 1880.
H. E. Ziegler. Die Entwickelung von Cyclas cornea, Lam. Zeitschr. f. wiss. Zool.

41 Bd. 1885.

Appendage.

Rhodope Veranii.

This small animal (circ. 4 mm. in length) is long and spindle-shaped, and out-
wardly bilaterally symmetrical. The body epithelium is ciliated all over. There
is a dermo-muscular tube, inside which, embedded in the connective tissue (paren-
chyma), are found numerous irregularly shaped calcareous particles.

Alimentary Canal. — The mouth lies at the anterior end of this canal, and leads
into a wide buccal or cesophageal cavity, into the first part of which two acinose
salivary glands open. A radula and jaws are wanting. A narrow oesophagus con-
nects the cesophageal cavity with the tube-like mid-gut, which runs through the
whole length of the body. The midgut possesses a well-developed muscular wall, and
is continued anteriorly, above the point where the oesophagus enters it, iu the form
of a diverticulum, which runs forward over the brain. There is no separate digestive
gland. The right side of the mid -gut gives rise to a short, thin, ciliated rectum,
which runs through the posterior third of the body, and opens through the anus to
the right.

The nervous system consists of two pairs of ganglia lying so close together above
the oesophagus as almost to form one mass, and of one infra-ossophageal ganglion,
which lies somewhat asymmetrically to the left. The two ganglia of each of the
upper pairs are connected by transverse commissures, and the posterior dorsal pair
with the lower ganglion by two connectives which embrace the (esophagus. Two
lateral nerves which run backward are the most strongly developed. They arise out
of the posterior upper pair of ganglia, close to which lie a pair of eyes and a pair of
ciliated auditory vesicles, each of the latter containing an otolith.

Genital Organs. — Rhodope is hermaphrodite. The gonads consist of about 20
follicles which lie ventrally in the median and posterior thirds of the body ; the
anterior follicles produce eggs and the posterior spermatozoa. The ducts of all the

282 COMPARATIVE ANATOMY CHAP.

follicles are said to unite to form a common duct. If this is really the case, then
the gonadial follicles together form a hermaphrodite gland. The hermaphrodite
duct, which runs forward, is said to divide into an oviduct and a vas deferens.
The latter leads to the muscular penis, which can be protruded from the male genital
aperture on the right anteriorly. With the oviduct are connected a receptacu-
lum seminis and a gland (albuminous or nidamental gland). The female genital
aperture is said to lie on the right side, behind, and distinct from, the male
aperture.

A differentiated blood vascular system has not been found. A well-developed
body cavity is, however, present, filled with colourless nutritive fluid, in which blood
corpuscles are suspended.

Special respiratory organs are wanting.

The nephridial system has been described as follows. To the right, in front of
the anus, between the latter and the genital aperture, lies the outer nephridial
aperture. It leads through a short ciliated canal into a spacious renal chamber,
which is a widening of a longitudinal canal. The renal chamber bulges out at
several points to form short cteca. Into this chamber nine or ten small flask-like
organs open ; these resemble the excretory ciliated cells of the Platodes, inasmuch
as "flames"1 arise at the base of each flask, the neck of which opens into the
chamber.

Development is direct. At no stage are there any indications of a shell gland, a
shell, or a foot.

Systematic Position. — Rhodope is by some classified among the Turbellaria
(near the Hhabdoccelidce), by others among the Mollusca (near the Nudibranchia],
while others again are inclined to see in it a transition form between these two
phyla.

There is apparently only one single point to support the theory of the relation of
Rhodope to the Turbellaria, viz. the presence of the ciliated excretory cells in the
nephridial system. On the other hand, the derivation of the nephridial system of
Rhodope, with its renal chamber and aperture to the right, from that of the Nudi-
branchia appears far more probable than its derivation from the water vascular
system of the Platodes. The presence of a rectum and anus, and of an infra-ceso-
phageal ganglion (pedal ganglion), is difficult to reconcile with a relationship to the
Turbellaria. The occurrence of an infra-cesophageal commissure in one isolated case,
that of Microstoma linearc (cf. vol. i. p. 166), is hardly a convincing argument. The
genital apparatus of Rhodo2)e is much nearer to the Nudibranchiate than to the
Turbellarian type.

There are, no doubt, serious obstacles in the way of those who seek to establish
the relationship of these animals with the Mollusca. The chief of these is the want
of a heart and the entire absence of a shell and a foot, even in the embryo. The
question to be decided is whether it would be possible for a Mollusc which had lost
foot, gills, and shell (e.g. Phyllirhoe) by the further loss of the heart, so far to depart
from the typical organisation of the Mollusca, that these organs would not appear,
even temporarily, in the course of development. If this question is answered in the
affirmative, then the asymmetry of Rhodope, and especially the position of the
genital, nephridial, and anal apertures on the right side, which entirely agrees
with their position in the Nudibranchia, affords strong support to its claim to be
related with the Mollusca.

The view that Rhodope is a transition form between the Turbellaria and the
Mollusca need hardly be treated seriously.

Cf. vol. i. p. 152, where flame cells are described.

vii RHODOPE VERANII— LITERATURE 283

Literature.

L. von Graff. Ueber Rhodope Veranii. Koell. ( = Sidonia elcgans, M. Schulze).

Morph. Jahrbuch. 8 Bd. 1883.
A. Koelliker. PJioclope, nnovo genere di Gastropodi. Giornale dell' Istituto R. Lom-

bardo di scienze e.c. Tome 16. Milano, 1847.
S. Trinchese. Nuovo osservazionc aulla Rhodopc Veranii. Koell. Rendic. dell'

Accad. di Napoli. 1887.

CHAPTER VIII
SEVENTH RACE OR PHYLUM OF THE ANIMAL KINGDOM

ECHINODERMATA.

THE Echinodermata are, as a rule, essentially radiate in structure.
The}7, however, always deviate from strict radial symmetry in minor
points, both in the skeletal system and in the arrangement of
the inner organs ; sometimes they may become almost bilaterally
symmetrical. The Echinodermata possess a skeleton of calcare-
ous matter deposited in the deeper connective tissue layers of
the integument. This skeleton is in texture a fine rigid sponge-
work. It consists either of microscopically small isolated calcareous
bodies (Holothurioidea) or of larger plates which often carry spines, and
are connected together1 either movably or immovably (other Echino-
derms). The coelom is spacious. Thei'e is a blood vascular system.
The intestine, which is provided with a mouth and anus, is completely
separated from the coelom. The Echinodermata possess a peculiar sys-
tem of canals or tubes— the water vascular system. This system, on
the one hand, takes in water from the exterior through a stone canal
(sometimes several such canals are present), which primitively opens
outwards, and, on the other hand, sends out terminal canals to ex-
ternal extensible apperflages arranged in the radii or ambulacra.
These are the ambulacral feet or tentacles, which in free forms serve
principally for locomotion, but also for respiration ; in attached forms,
for respiration, and also perhaps for conducting food. The sexes are
almost always separate. Development is accompanied by metamor-
phosis. The larvae are free-swimming and pelagic ; they are bilater-
ally symmetrical, with ciliated bands, generally produced on processes.
The Echinodermata are exclusively maiine, and contain a great number
of fossil forms ; certain extinct types attained a great development
during the palaeozoic age.

The race of the Echinodermata is divided into five classes — Holo-
thurioidea, Eehinoidea, Asteroidea, Ophiuroidea, and Pelmatozoa.

CHAP, vin EGHINODERMATA— SYSTEMATIC REVIEW

285

Systematic Review.

CLASS I. Holothurioidea.

The body is elongated along its principal axis ; it is cylindrical or vermiform. It
shows more or less distinct bilateral symmetry. The integument is soft or leathery,
and contains irregularly arranged, generally microscopically small, calcareous bodies.
The mouth lies at the oral (anterior) end of the principal axis of the body, and is
surrounded by feelers. The anus lies at the apical (posterior) end of the principal
axis. Ambulacral or tube-feet are either present or wanting. An external madre-
porite is usually not found.

ORDER 1. Actinopoda.

All the outer appendages of the water vascular system arise from the radial
canals, and take the form of feelers round the mouth
and of tube-feet (and ambulacral papilla2) in other
parts of the body ; such feelers are always present,
the feet and papillae, however, may be wanting.

Family 1. Aspidochirotse.

Tube -feet present. Mouth often more or less
ventral in position. Body usually shows distinct
flattening of the ventral surface. 18-30 peltate
tentacles. Tentacular ampullae well developed.
Stone canals often numerous. Retractor muscles
wanting. Respiratory trees present. Cuvier's
organs often present. Mullema, Holothuria,
Stichopus,

Family 2. Elasipoda.

Tube-feet present. Mouth more or less ventral
in position. Body almost always distinctly flattened
on the ventral surface. 10, 15, or 20 tentacles, more
or less peltate in shape. Stone canal always single,
and not infrequently in direct communication with
the exterior through the integument. Retractor
muscles wanting. Respiratory trees wanting or
quite - rudimentary. Cuvier's organs wanting.
Sub-fam. Psychropotidse : Psychropotes (Fig. 223),
Benthodytes. Sub-fam. Deimatidse : Z>cima, Pan-
nychia, Lactmogone. Sub-fam. Elpidiidae : Elpidia,
Kolga, Pcniagone.

Family 3. Pelagothuriidse.

Tube-feet wanting. Mouth and anus terminal.

PIG. 223.-Psychropotes longi-

Body cylindrical ; round the crown of tentacles it cauda (after Thgei). 1, Oral ten-
widens out into a thin disc, the edge of which is tacle; 2, mouth; 3, 4, 8, ambulacral
produced into long rays. 13-16 tentacles. Re- appendages of the (ventral) trivium ;

tractor muscles wanting. Neither respiratory trees, 5' anus ; 6' dorsal aPPend*f with

J its two posterior processes (7).

nor ciliated organs, nor Cuvier s organs present.

Calcareous bodies altogether wanting. Pelagic, swimming by means of the disc.
Single genus and species : Pelagothuria natatrix (Figs. 224 and 225).

286

COMPARATIVE ANATOMY

CHAP.

FIG. 224.— Pelagothuria natatrix (after Ludwig), completed ; from above. 1, Body ; 2, anus.

FIG. 225.— Pelagothuria natatrix (after Ludwig) ; front view, i.e. from the oral pole. 1, Mouth
2, oral tentacles ; 3, disc ; 4, canals of the disc.

VIII

ECHINODERMA TA —S YSTEMA TIC RE VIE W

287

Family 4. Dendrochirot*.

Tube-feet present. Mouth dorsal or terminal. Anus also often dorsal. Body
cylindrical, or pentagonal, or with a distinctly marked creeping sole. 10-30 arbor-

FIG. 227. — Psolus ephippifer,
young female, from the dorsal side
(after Theel). 1, Oral valves; 2,
anus.

FIG. 226.— Cuoumaria planci (original). 1, The
two smaller ventral oral tentacles ; 2, mouth ;
3, anus.

escent tentacles, often of unequal size. Tentacular
ampullre not distinct. Not infrequently more than
one stone canal. Retractor muscles well developed.
Respiratory trees present ; Cuvier's organs only
occasionally found. Cucumaria (Fig. 226), Thyone,
PhyUophorus, Colochirus, Thdelia, Psolus (Figs. 227
and 228), Rhopalodina.

Family 5. Molpadiidse. FIG. 228. — Psolus ephippifer,

female, dorsal aspect (after Theel).

Tube-feet wanting. Mouth terminal. The pos- i, Oral valves, opened ; 2, anus ;
terior end of the cylindrical body often narrowed to 3, oral tentacles ; 4, dorsal cal-
a shorter or longer tail-like piece, which is more or careous scales,
less distinct from the trunk. 15 tubular or digitate

tentacles normally present. Tentacular ampullae present. A single stone canal.
Retractor muscles distinct only in the genus Molpadia. Respiratory trees present.
Cuvier's organs almost always absent. Moljjadia, Caudina, Trochostoma, An-
kyrodcrma.

288

COMPARATIVE ANATOMY

CHAP.

ORDER 2. Paractinopoda.

Only some of the outer appendages of the water vascular system arise from the
radial canals, the rest from the circular canal, and the only form taken by them is
that of tentacles round the mouth.

Family 1. Synaptidse.

Tube-feet wanting. Mouth terminal. Body cylindrical, more or less elongated
and vermiform. 10-27 feathered or digitate tentacles. Stone
canals occasionally numerous. Retractor muscles sometimes
present. Respiratory trees and Cuvier's organs wanting. Sexual
glands often hermaphrodite. Synapta (Fig. 229), Chirodota,
Myriotrochus. 1

FIG. 229.— Synapta
digitata (original).

CLASS II. Echinoidea (Sea-urchins).

The body of these Echinoclerms is covered by a usually firm
but sometimes flexible test, which contains the ccelomic cavity
and the viscera. The test varies in shape, from spherical to a
form which is flatly compressed in the direction of the principal
axis. It consists of numerous pentagonal or hexagonal closely
contiguous plates, -which, arranged in meridional rows, form five
ambulacral and five interambulacral areas. It is covered by the
outer layer of the integument, and carries spines articulating
with it. At the apical pole there is a system of plates, consisting
of five basal plates, five radials, and the anal plate. The mouth
is usually in the middle of the oral surface, less frequently shifted
towards the edge in what is called the anterior direction. An
anus is always present, either at the apical pole or at some part
of the posterior interambulacral area. The apertures of the
madreporite lie in the apical system, generally in one of the basal
plates ; they are connected not only with the stone canal but
with the so-called dorsal organ. The ambulacral vascular system
has outer appendages developed as tube-feet and gills. Mouth
with or without teeth. In the former case a complicated
masticatory apparatus is developed within the test for the move-
ment of the teeth ; the muscles moving this apparatus are
attached to a perignathous apophysial ring developed at the
edge of the oral aperture of the test (i.e. round the peristome).
Sexually separate or hermaphrodite. The genital ducts open
externally through pores in the basal plates or outside these
latter. Development direct (with care of the brood), or with
metamorphosis (free-swimming larvse).

SUB-CLASS 1. Palseechinoidea.

Either only one row or more than two rows of plates in each
interambulacral area. Two or more meridional rows of plates in
each ambulacral area. Plates of the test do or do not imbricate.

Oral aperture of the test (with peristome) in the middle of the oral surface. Jaws

1 The arrangement of the classes and families of the Holothurioidea by Ludwig in
Bronn's Klassen und Ordnungen des Thiemichs, 1892, is here followed.

VIII

EGHINODERMA TA—8 YSTEMA TIG RE VIE W

289

present. Anal area either within the apical system, or outside it, in the posterior
interambulacral area. Palaeozoic forms.

Order 1. Bothriocidaroida,

Regular Echinoidea, with a more or less spherical, firm test. In each inter-
radius there is only one meridional row of plates ; in each ambulacral area there are
two. Anal area, with anus within the apical system. Mouth in the centre of the
oral surface. Bothriocidaris.

Order 2. Perischoechinoida.

More than two meridional rows of plates in each inter-
Two or many meridional rows in each radius. Test thick and rigid, or

Regular Echinoidea.
radius.

FIG. 230.— Palseechinus elegans M'Coy
(after Baily).

Fir,. 231.— Tiarechinus princeps Laube (after Loven).
1, Genital aperture ; 2, anus; 3, basal; 4, radial; 5, ambu-
lacrum ; 0, the 3 upper plates of an interambulacrum.

thin ; in this latter case more or less imbricated. Jaws present. Fam. Archseo-
cidaridse : Lcp-idocentrus, Archccocidaris ( = Echinocrinus), Palccechinus (Fig. 230)
Fam. Melonitidse : Melonites.

Order 3. Flesiocidaroida.

Test small and rigid, almost hemispherical. Apical system very large, with
large united basal plates and central anal area. Ambulacra narrow, with two meri-
dional or vertical rows of plates. Interambulacra with one single peristome plate,
followed by three plates separated by vertical sutures. Tiarechinus (Fig. 231).

Order 4. Cystocidaroida.

Test irregular (exocyclic), spherical or ovoid, thin and flexible. Madreporite
central. Ambulacral areas narrow, with two vertical rows of plates. Interambu-
lacral areas broad, with numerous vertical rows of scale-like movable plates. Anus
in the posterior interambulacrum above the ambitus. Echiuoeystis ( = Cystocidaris).
VOL. II U

290 COMPARATIVE ANATOMY CHAP.

SUB-CLASS 2. Euechinoidea.

Echinoidea with two vertical rows of plates in each ambulacra! and iu each
interambulacral area. Mouth on the oral side, rarely shifted towards the edge
(anteriorly). Teetli and jaws present or wanting. Anus either within the apical
system, or outside it, i.e. somewhere in the posterior interradius.

Order 1. Cidaroida.

Mouth central, anus within the apical system. No external gills. With jaws
and almost perpendicularly placed teeth. Perignathous apophysial ring interrupted.
Both the ambulacral and the interambulacral plates are continued over the peri-
stome on to the oral area as far as the mouth. On the oral area they are imbricated.
Ambulacra narrow. Large principal and small accessory spines. Spha?ridia want-
ing. Cidaris.

Order 2. Diadematoida.

Mouth central, anus within the apical system. So-called internal gills well
developed, or rudimentary, or wanting. With external gills, and incisions in the
peristome. With jaws and teeth. Perignathous circular apophysial ring closed.
Only the ambulacral plates are continued over the peristome on to the oral area,
where they often appear as separate buccal plates. Spline ridia present.

Sub-Order 1. Streptosomata.

Test more or less flexible, with inner dorso- ventral longitudinal muscles. Botli
external and internal gills present. The ambulacral plates (and only these) are
continued over the peristome on to the oral area. Fam. Echinothuridae : /VA///-
echinus, EchinotJmrin, PJwrmosoma, Asthenosoma.

Sub-Order 2. Stereosomata.

Test rigid, without internal longitudinal muscles. External gills present, in-
ternal gills rudimentary or wanting. The ambulacral plates on the oral area are
replaced by isolated buccal plates. Fam. 1. Saleniidae : Peltastes, Salcnia (almost
exclusively fossil). Fam. 2. Hemicidaridae : Hcmiddaris, Acrocidaris, Gonio-
Piigus, etc. (fossil). Fam. 3. Aspidodiadematidse : Aspidodiadema. Fam. 4.
Diadematidse : Diadema, Diplopodia, Pcdina, Echinothrix, Astropyga, Codeclriiius.
Orthopsis, Pcronia, Echinopsis, etc. (fossil and extant). Fam. 5. Cyphosomatidse.
Cyphosoma, etc. (almost exclusively fossil). Fam. 6. Arbaciidae : Arbacia, Echi-
nocidaris (Fig. 232), Ccelopleurus, Podocidaris (extant and fossil). Fam. 7. Tem-
nopleuridse : Glyphocyplius, Tenmopleurus, etc. (extant and fossil). Fam. 8.
Echinometridge : Echinometra, Parasalcnia, etc., Spongylocentrotus, Splwereclvinus
(mostly extant). Fam. 9. Echinidae : Echini/ x, 7T<av>/""'"N/'-s Tt-iinfustcs (extant
and fossil).

Order 3. Holectypoida.

Mouth central. Anus outside of the apical system in the posterior interradius
(exocyclic). With external gills. Only one pair of pores or a single pore on each
ambulacral plate. Jaws weak ; teeth perpendicular ; both jaws and teeth may be
wanting. Sphwridia present, (a) Ambulacral apophyses present : Holectyinis,

VIII

ECHINODERMA TA —SYSTEM A TIC RE VIE W

291

'i-, etc. (principally fossil), (b) Ambulacral apophyses rudimentary or want-
ing : Discoidca, Conoclypeus (fossil).

Order 4. Clypeastroida.

Mouth central or sub-central. Anus outside of the apical system in the posterior
interambulacrum. With external gills. With tentacle pores in the interradii.

FIG. 232.— Echinocidaris (Arbacia) pustulosa, from tlie apical side (original). The spines have
been removed from part of the shell. 1, Interambulacrum ; '2, ambulacrum.

More than one pair of pores on each ambulacra! plate. Tentacles differ in one and
the same animal. Teeth usually almost horizontal, rarely vertical. The jaws lie
above the apophysial ring, which is interrupted. Spha?ridia present.

The test is seldom much arched ; it is usually more or less flattened, and often
even disc-like. It often has many incisions and perforations, and is usually bilater-
ally symmetrical. Its dorsal wall is connected internally with its ventral wall by
means of pillars, needles, septa, etc. Basal plates of the apical system fused. The
ambulacra form petaloids in the apical region.

Fain. 1. Fibulariidse : Echinocyamus, Fibularia, etc. (extant and fossil). Fam.
•1. Clypeastridae : Clypeaster (Fig. 233), etc. (extant and fossil). Fam. 3. Laganidse :
tnit: (extant and fossil). Fam. 4. Scutellidse. In all the genera of this family

292

COM PA RA TI VE A NA TOM Y

CHAP.

FIG. 233.— Clypeaster sp., test from the apical side (original).

Fn:. 234.— Scutella sexforis, test from the apical side (original).

vui EGHINODERMATA— SYSTEMATIC REVIEW 293

the shell is very flat : Scutelfa (Fig. 234), Eckinodiscus, Eni-npf, J/<7//7« (Fig. 235),
Rof.ula, Arachnoides, etc. (extant and fossil).

FIG. 235.— Mellita testudinata? from the oral side (original).

Order 5. Spatangoida.

Mouth central, sub-central, or on the anterior edge of the oral surface of the test
Anus outside the apical system, in the posterior interradius. External gills, jaws.
teeth, and perignathous apophysial ring wanting. Sphaeridia present. The ambu-
lacra generally form apical petaloids. The test is bilaterally symmetrical, arched,
often heart-shaped.

Sub-Order 1. Cassiduloidea.

Fam. 1. Echinoneidse : Echlnoconus, Echinoneus, Oliyopyyus, Ecki-itolrissus, etc.
(extant and fossil). Fam. 2. Cassidulidse : Cassidulus, Catopyyus, Ctypeus, Pygurus,
Echiiiohi.tnpas, etc. (mostly fossil). Fam. 3. Collyritidse : CoUyrites, Disaster, etc.
(fossil). Fam. 4. Plesiospatangidse : Eolainpas, Archiacia, etc. (fossil).

Sub-Order 2. Spatangoidea.

Fain. 1. Anan-chytidse : Echinocorys, Holastrr, Hemiimfwstcs, Cardiaster, Urc-
Cystech/inus, Calymne, etc. (the last three genera extant, the rest fossil).
Fam. 2. Spatangidse — Group 1, Adetes : Isastcr, Echinospatagus, ffeterolampas,
ijus, etc. (almost exclusively fossil) ; Group 2, Prymnadetes : Hemiastcr,
Linthia, ScMzastcr (Fig. 236), Agassizia (extant and fossil) ; Group 3,
Prymnodesmia : Micraster, Brixxtts, tipiiffuii'/oiiiorpha, Brissopsis, Spatangus, Palccop-
ir'tixfi-x (Fig. 237), EehinoeardiuHi, LnVfitin, etc. (extant and fossil) ; Group 4,
Apetala : Genieopatag^is, Palceobrissus, Accstc, Aerope, etc. (extant and fossil).

294

COMPAEA TI VE ANA TOM Y

CHAP.

Fiti. -236. — Schizaster lacu-
nosus ? from the apical side
(original). The spines, and the
protuberances on which they
stand, are not depicted. 1, Tin-
anterior unpaired ambulacrum ;
•2, the right anterior ambulacrum ;
3, fasciole ; 4, the right posterior
interainlmlacrnm ; 5, the right
posterior ambulacrum ; 6, the
unpaired posterior interambula-
crum ; 7, anal region.

Fni. ^37.— Palaeopneustes
Murray! (after Agassiz).
from the oral side. 1, The
anterior ambulacrum ; 2, 3.
the anterior right and the
posterior right ambulacra ;
4, peristome ; 5, anal region.

VIII

ECHINODERMA TA—S YSTEMA TIC RE VIE W

295

Fam. 3. Leskiidre : Faheostoma (extant). Fam. 4. Pourtalesiidse : Pourtalesia
(Fig. 238), Spatagocyslis, Echinocrcpis (extant).1

cvtv

FIG. 238.— Pourtalesia Jeffreys!, from the side (after Loven). The smaller tubercles are not
depicted, np, Apex ; os, oral pole ; ««, anal region. The numbers are explained in the text, in the
section on the perisomatic skeleton of the Echinoidea, p. 342.

CLASS III. Asteroidea (Stelleridea), Star-fish.

Echinodermata, with body flattened in the direction of the principal axis, the
radii being produced laterally into longer or shorter arms. The arms are usually five
in number, but their number may be increased to forty or more. They are not dis-
tinctly marked off from the central part of the body (the disc) ; and besides the radial
blood vessels, nerves, and ambulacral vessels, diverticula of the intestine and con-
tinuations of the genital organs run into the ccelomic cavities of the arms. The
body is usually covered with calcareous plates, but is flexible. The calcareous
plates carry spines, and often pedicellarite as well. Along each arm runs a ventral
furrow, within which there is a longitudinal row of paired ambulacral plates. The
consecutive pairs are movably articulated with one another. Besides these, there
are, on the arms, adambulacral, inframarginal, supramarginal, and dorsal plates.
The ambulacral grooves run from the central month on to the arms, and along these
on their oral (ventral) side, below the ambulacral plates, to their tips. The tube-
feet (tentacles) rise from the base of this groove, to which they are limited. Anus
apical (i.e. in the centre of the upper side), rarely wanting. Madreporite also on the
apical side of the disc. The sexes are separate. Development is in most cases with
metamorphosis (free-swimming pelagic larvfe) ; when the brood is protected develop-
ment is direct.

SUB-CLASS 1. Palseasteroidea.

Palaeozoic Asteroidea, in which the ambulacral plates in the two longitudinal
rows in each arm, at least in the middle of the arm, are arranged alternately (not
opposite or in pairs). Aspidosmna, Palceaster, Palccocoma, etc. (all Palaeozoic forms).

' The classification of the Echinoidea here given is after Martin Duncan,
af tit? Genera and Great Groups of the Echinoidea. London, 1889.

Revision

296

COMPARATIVE ANATOMY

CHAP.

SUB-CLASS 2. Euasteroidea.
Asteroidea with paired, i.e. opposite ambulacra! plates or "vertebrae."

Order 1. Phanerozonia.

Asteroidea with large, strongly developed, marginal plates. The inframarginal
and supramarginal plates are closely fitted together. Papulse (branchial vesicles)
only occur on that surface of the body which is surrounded by the supramarginal

FIG. 230.— Ctenodiscus procurator (after Sladen), from the oral side. A Gastropod in thf
.stomach is visible through tlie mouth.

plates, i.e. on the apical or upper side. Ambulacral plates broad. In each ambu-
lacral furrow there are two longitudinal rows of tube-feet. The adambulacral plates
are prominent in the oral skeleton. Where pedicellarise occur they are sessile.

Fain. 1. Archasteridse : Pamrchastrr, Dytastcr, Pluton aster, Pscudarchaster,
Archaster, etc. Fam. 2. Porcellanasteridae, the centre of the apical system pro-
duced into a more or less long outgrowth : Porcellanaster, Hui>lnil<i*f<-i-, OV)/or//.xr?/.x-
(Fig. 239), etc. Fam. 3. Astropectinidse, without anus and usually without pedicel-
lariae : Astropecten, Eathybiastcr, Ilyastcr, Luidia, etc. Fam. 4. Pentagonasteridse :
Pentagonaster, Astroganium, Nectria, Calliaster, Stellaster, Goniodiscus, Mimaster, etc.
Fam. 5. Antheneidse : Anthenca (Fig. 240), Gunirtstci; etc. Fam. 6. Pentacerotidse :

viir

ECHINODERMA TA— SYSTEM A TIC RE TIE W

29'

Pentaccros, Aniphiaster, Culcita, Astcrodiscus, etc. Fam. 7. Gymnasteriidse :
nastcria, Tyla.stcr, Asteropsis,
Marginaster, etc. Fam. 8. As
terinidse : Ganeria, Asterina,
l'ni ni Ipi-s, etc.

Order 2. Cryptozonia.

Asteroidea, in which the mar-
ginal plates are indistinct and
more or less rudimentary in the
adult. The supramarginal plates
are often separated from the in-
framarginal by intermediate plates. The
papulae are not limited to the apical sur-
face, but often occur also between the
marginal plates and on the oral (lower)

FIG. '240. — Anthenea tuberculosa, Gray?
juv. (after Sladen). 1, Snpramarginal plates ;
'2, pedicellarise ; 3, madreporite ; 4, anus.

FIG. 241.— Cnemidaster Wyvillii (after Sladen). dc, Dorsocentral ; /•, radials ; bn, basals ;
sm, supramarginals ; d, dorsals ; t, terminals.

surface of the body. Ambulacral plates narrow, closely crowded. Tube-feet often
in four rows. In the oral skeleton the ambulacral or interambulacral plates are
prominent. Pedicellarife sessile or pedunculate.

298 COMPARATIVE ANATOMY CHAP.

Fam. 1. Linckiidse : Chactnsfe r, OpMdiaster, Linckin, Metrodira, etc. Fam. '2.

FIG. -24-2. — Hymenaster caelatus (after Sladen), with arms bent back.

Zoroasteridae : Zoroaster, Cncmidastcr (Fig. 241). Fam. 3. Stichasteridse : Sti-
i; etc. Fam. -4. Solasteridse : Sof aster, C'rossastcr, Corftkrastrr, etc. Fam. 5.

Fie. -_'4o.— Hymenaster nobilis (al'ter Sladen), t'nuu the mal side, v natural :<ize.

Pterasteridse, with brood cavity on the apical side of the disc : Pteraster, Retastcr,
Ifi/nii-iKiftfcr (Figs. 242 and 243), Myxastct; Bi'nthastrr, Pi/thonastcr, etc. Fam. 6.

VIII

ECHINODEJRMA TA—SYSTEMA TIC RE VIE U'

Echinasteridse : Acanthastcr (numerous arms), Mithrodia, Cribrclla, E<:liinnxt<'i-,
/Ww.sYiv, etc. Fam. 7. Heliasteridse, with numerous short arms : HrJiiiati'i: Fani.
8. Pedicellasteridae : Pcdice/fantn\ Fam. 9. Asteriida, tube - feet in four rows :
Asterias, Uniophora, Coronastcr, etc. Fain. 10. Brisingidse, with numerous very
long arms, marked off from the small disc : Brisinga, Labidiastcr, etc.1

CLASS IV. Ophiuroidea.

Echinodermata flattened in the direction of the principal axis of the body, the
radii' of which are produced into five long, round, simple or much branched slender
arms. The arms are sharply marked off from the central part of the body, and do
not contain either caeca of the intestine or extensions of the genital organs. The

Fin. -244.— Ophiolepis elegans, Liitken (after Lyman). ds, Dorsal shields ; ss, lateral shields ;
dc, dorsoceiitral ; ib, inf'rabasal ; IIK, basal ; r.s, radial shields ; r, radial.

axial part of the arms is occupied by a longitudinal row of vertebral ossicles, articu-
lated together, and consisting of two fused lateral ambulacra! plates or ossicles.
The body is usually covered with calcareous plates. On the arms we can distinguish
a longitudinal row of ventral shields on the oral side, two longitudinal rows of

' The classification of the two orders of the Euasteroidea is that of W. Percy Sladen,
/d'jiort on the Asteroidea collected by H.M.S. C'htdlenycr. London, 1889.

300

COMPARA TIVE ANA TOMY

CHAP.

lateral spine-bearing shields, and a longitudinal row of dorsal shields. On the
apical surface of the disc larger radial shields are found at the sides of the bases of
the arms : thus ten in all. On the oral side of the disc there are five interradial
plates which are distinguished by their great size ; these are the buccal shields. One
of these plates is at the same time the madreporitic plate. Mouth at the centre of
the lower side. Anus wanting. The ambulacral tube-feet appear on each side on
the arms between the ventral and lateral shields. On the lower side of the disc, close
to the bases of the arms laterally, there are in all ten or twenty slit-like apertures—
the bursal apertures. These lead into blind sacs projecting into the ccelom ; these
are the bursfe, which serve for respiration and for the reception and ejection of the
genital products. Development direct (viviparous and with care of the brood), or
with metamorphosis (free-swimming pelagic larvae).

Order 1. Ophiurse.

Arms unbranched, movable in the horizontal plane, usually distinctly plated.
Buccal shields, one of them at the same time the madreporitic plate, distinctly
developed.

Fam. 1. Ophioglyphidae : npliim-ci, Pectliiura, Ophivlcpis (Fig. 244), Ojrfiiozona,
Ophwgfypha, Opliioctcn, Opltiomusium. Fam. 2. Amphiuridse : OpJn'acfis (Fig. 245),

— 2

. B£L_3

>— C >-£/ /

FIG. 24'j.— Ophiactis poa, Lym. (after Lyman). Disc ami basal portions of *he arms ; from the
oral side. 1, Ventral shields ; 2, spines on the lateral shields (4) ; 3, tentacle scales ; 5, lateral
buccal shields ; ti, bursal apertures ; 7, buccal shields ; 8, first ventral shield of the arm ; 9, torus
angularis ; 10, oral papilla1.

. luijih iui'c, Ophiocnida, OpJiincoiiu>, Ophiacantha, Ophiothrix. Fam. 3. Ophio-
myxidae, disc and arms covered by a thick naked integument : Ophiom/yxa,

eimjalc.

VIII

ECHINODERMA TA—X YSTEMA TIC RE VIE Jl '

301

Order 2. Euryalse.

Anns simple or branched, can be rolled up vertically towards the mouth. Only
rudimentary shields are found below the soft but thick outer integument. Without
spines. In forms with unbranched arms there are usually 5 buccal shields, one of
which is the madreporitic plate. Most of the forms with branched arms have no

/(HPr &

V& -V iUL \ -,

,gpL

f

^^K\ M or/' -^^?

FIG. 246.— Astrophyton Lincki (Miiller and Troschel), from the oral side (original).

distinct buccal plates. There is then either a single madreporite in an oral inter-
brachial area or else there are 5 interbrachial madreporites.

Single Fam. Astrophytidse : Astropliyton (Fig. 246), Gforgonocephalus, Euryale,
Trichastcr (arms slightly and only at their tips, dichotomously branched), Astrodon
(the same), Astrocnida (the same), Astroporpa (arms undivided), Astrogomphus (the
same), Astrochde (the same), Astrotoma (the same), Astroschema (the same), Ophio-
creas (the same), etc.1

1 For a more recent classification of Ophiuroidea, see F. J. Bell, I1 roc. Xool. Soc.
London, 1892, pp. 175-183.

302 COMPARATIVE ANATOMY CHAP.

CLASS Y. Pelmatozoa.

Echinodermata which are either permanently or temporarily l attached by the
centre of the apical surface, so that the oral surface (with the mouth, as a rule, in its
centre) looks upward. The body is usually raised upon a jointed stem attached to
it at the apex. An axial canal, in which are blood vessels and nerves, runs through
the stem. This stem is sometimes found only in the young, the bod}' becoming
detached later, and further in a few attached forms no stem at all is developed.
The apical system of plates consists of 5 basals and 5 radials, to which 5 infra-
basals and a varying number of interradials are often added. The plate in the
embryo Antctlon, which becomes fixed to the ground and is subsequently lost, is
called " dorsocentral," and is supposed to belong to the apical system. The number
of the principal rays is rarely 4 or 6. The plates just mentioned form a cup
(dorsal cup), which either simply carries or else more or less completely encloses the
visceral mass. The cup carries jointed appendages, — arms or pinnulte or both.

The oral side (in these animals turned uppermost) is often provided with f» oral
plates, which surround or cover the central mouth, and it may further be protected
in very various ways by radially and interradially situated plates (ambulacrals,
interambulacrals, and orals), which together form the tegmen calycis. Or again this
cover of the calyx may be either naked or set with very small isolated calcareous
pieces. The anus lies usually at the end of a longer or shorter tube, excentrically in
an interradius of the tegmen, occasionally, however, at the boundary between the
cup and the tegmen. The circumcesophageal canal of the water vascular system does
not communicate direct with the exterior. The radial canals of this system run
into the arms. Each of the latter has a food groove on its oral (uppermost) side.
The tube-feet, which rise from the edge of this furrow, are tentacular, and do not
serve for locomotion, but for respiration, and possibly for conducting food.
Development, so far as is known, with metamorphosis.

SUB-CLASS 1. Crinoidea.

Pelmatozoa with long usually branched arms. The arms are jointed, the con-
secutive ossicles being connected by muscles and bands. The arms can be expanded,
and closed up together, or again can roll up orally. They may carry jointed,
unbranched appendages, the pinnuhe, which are probably modified branches. The
nervous system is generally said to be "double," i.e. there is an abactinal and an
oral system. The abactinal nervous system consists of a central portion lying in
the apex of the dorsal cup and of radiating strands which run through the skeletons
of the stem, the arms, and the pinnuhe. The oral nervous system consists of a
circumoral nerve ring, and radiating strands which run into the arms through the
epithelium at the base of the food grooves, and which branch with the arms. The
food grooves of the arms pass at their bases on to the tegmen, running in it to
the central mouth. Ambulacral tentacles may be wanting. The circular canal of
the water vascular system is connected with the body cavity by means of several
stone canals, and the body cavity is in open communication with the exterior by
means of water pores. The mouth is in the centre of the tegmen (exc. Actino-mctrtt).
The sexual organs extend right into the basal parts of the arms, and even into
their pinnulre. In pinnulate crinoids. so far as is known, however, the genital
products only ripen in the pinnule.

1 There is, however, no evidence to show that Mursupites was attached even in the
larval stasje : unlike Antedonidse, it has no trace of a stem.

viii ECHINODEBMATA— SYSTEMATIC REVIEW 303

The old division into PalcKOcrinoidm and Ncocrinoidca seems artificial ; that here
adopted also cannot be considered as definitive.1

Order 1. Inadunata.

Calyx comparatively small ; dorsal cup with monocyclic or dicyclic base ; the
basals in the former, and infrabasals in the latter case may be fused to 4, 3, 2, or
1. The only other plates in the apical capsule are 5 radials. In the posterior
interradius there are very often 1-3 asymmetrically placed anal plates, but no plates
in the other interradii.

The tegmen calycis varies. In some Inaduttata (Larvifuniua) there are 5 large
oral plates, which, rising at the edge of the calyx directly above the radials, form a
closed pyramid covering the food grooves of the disc, and the mouth. In many
other forms the orals (which may be partly resorbed) lie at the centre of the tegmen
calycis. The posterior oral plate is often larger than the others, and is shifted for-
ward between them. The ambulacra appear at the surface of the tegmen calycis be-
tween the oral plates and the edge ; they are bordered on each side by rows of small
lateral pieces, the ambulacral groove being also roofed in by small covering pieces.
Plates of various shapes, size, and arrangement are found in the interambulacral
regions. In the posterior ambulacral region the tegmen calycis often bulges out in
the form of a plated sac, the so-called ventral sac (Fistulata) ; this varies in form
and size (sometimes reaching beyond the arms), and may sometimes have contained,
besides the rectum, a large part of the body cavity. The anus lies at its tip or on its
anterior side.

Arms free, i.e. not included in the dorsal cup (hence the name Inadunnta),
simple or branched, with or without pinnulre. The food grooves of the arms are
roofed in by two or more rows of alternating, wedge-shaped, interlocking, ambulacral
plates ; these plates could probably be erected.

Almost exclusively palreozoic forms.

A. Monocyclica.
With monocyclic basis (without infrabasals ; several radials often horizontally

A.

PIG. 247.— Haplocrinus mespiliformis (after Wachsmuth ami Springer). A, from the anal
side ; B, from the oral side. 1, Orals ; 2, oral pole ; 3, anus ; 4, radials ; .0, right posterior infer-
radial or radianal ; (3, basals ; 7, first brachial ; 8, facet for attachment of the arm.

bisected). Haplocrinus (type of the so-called Larviformiu, without anal plate) (Fig.

1 Classification chiefly after the recent works of Waclismuth and Springer and Her-
bert Carpenter. See Bibliography, p. 551,

304

COMPARATIVE ANATOMY

CHAP.

247). Hctcrocrinus, Hcrpctocrinus, Calccocrimis, Catillocrinus, Pisocrinus, Hybocrimis,
locrinus, Symbathocrinus, Bclemnocrinus, Gastrocoina (?}, Cupressocrinus.

B. Dicyclica.

With dicyclic base (with infrabasals). Fam. Dendrocrinidse : Dendrocrinus,
Iliuiwcrinus, Potcriocrinus. Fam. Decadocrinidse : Botryocrinus, Barycrinus,

, i-X

mm

CO

FIG. 248.— Encrinus lilliformis (original).
Ci, c->, Costals or primibrachials ; /', railials ;
vo, stem ; p, pinnulaj.

FIG. L'40. — Cyathocrinus
longimanus (after Angelin).
pr, Ventral sac ; !, place where an
arm-branch has been removed ;
r, radials ; 6«, basals ; ib, infra-
basals ; col, stem ; x, anal plates :
co, costals or primibrachials.

Atelcstucrinus, Dccadocrinus, Graphiocrinus, Encrinus (Fig. 248), (without anal
plates, ventral sac reduced to a short cone, Trias), Cromyocrinus, Ayassizo-
crinus. Fam. Cyathocrinidse : Cyathocrinus (Fig. 249), Gissocrinus, Lccythocrinus.
Hypocrinus.

The genus Marsupitcs from the Chalk, and the following extant families are
perhaps to be classed near the Inadunata ; in these latter five large separate orals
occur, the ventral sac being reduced to an anal tube, and no anals appearing in the
dorsal cup. Hul ophite (Fig. 250) (Lias, to present time), Hyocrinidic (Fig. 251) (Lias,
present time), Bathycrinidcc (extant).

VIII

EGHINODERMA TA —S YSTEMA TKJ RE VIE II '

305

FIG. 250.— Holopus Rangi d Orbigny, fruin the trivial
side (after P. H. Carpenter).

K -3

no
PH J

| 1
S,2

M .S
fl C5

rt *~

3 -f

(D •-

^3 ~f-

S u

« -?
« 5

OT ^~*

S g

B§

o s

>>^

w s
I g

.-* ^

m
II

II'. ' . : > i " \ .'. * -i ,-*]"' ^

-"•'Hi"V;
- if

!
i . ' • ;

O <M

^ Ji"

'rt

t-t

O

'f \ W!

I
1

tr|«-;;l\:

• i
f^lf

; I

si

«

VOL. II

306 COMPARATIVE ANATOMY CHAP.

Order 2. Camerata.

Plates of the calyx firmly connected by means of sutures. The apical capsule
shows a tendency to develop a very rich system of plates, incorporating the
proximal brachials to a greater or lesser extent. These brachials are connected
together in the interradii by interradial plates, which vary in number, and to which,
in the anal interradius, special anal plates may be added. In those cases in which
the arms are incorporated in the calyx to such an extent that they branch in the
latter before they become free from it, their branches may be connected by inter-
calated plates. Each of the five radials is usually followed by two brachial plates,
formerly called 2nd and 3rd radials. The tegmen calycis is richly plated with firmly
connected pieces, and is often much arched, forming a so-called vault. The mouth,
which lies in the centre of the tegmen, is covered with five firmly united oral
plates ; the hindermost of these, which is often the largest, projects in between
the four others. The ambulacra, with their lateral and covering plates, are mostly
not visible from outside, as the interambulacral plates which border them laterally,
and which are often very numerous, close together over them by means of processes,
and thus cover them externally. The ambulacra, in their course on to the bases of
the free arms, divide as many times as the arms have already divided on the
dorsal cup. The interradials of the dorsal cup often pass, without any sharp
boundary, into the interradially arranged interambulacrals of the tegmen calycis.
The subcentral (less frequently central) anus, which is surrounded by firm anal
plates, is either sessile or else comes to lie at the tip of a chimney -like prolongation
of the tegmen ; this anal tube, formerly thought to be a proboscis, may project
beyond the arms. Arms branched ; in adults, almost without exception, the
brachials become arranged in a double row with primitive articulation, and pinnules
closely folded together. Dorsal canals (in the brachials) have never been observed.
Exclusively paheozoic forms.

Family 1. Reteocrinidse.

Apical capsule, with monocyclic or dicyclic base. Four or five basals. Inter-
radial and interaxillary regions deeply sunk, plated with a large number of irregular
immovable pieces, which are continued on to the interambulacral areas of the tegmen
calycis. Posterior interradial region broader, and divided by a perpendicular row
of somewhat large anal plates. Anus subcentral. Arms composed of a single row of
calcareous joints. Pinnules strong. Ectcocrinus. Xenocrinus.

Family 2. Rhodocrinidse.

Apical capsule with dicyclic base. The circle of the five radials interrupted by
that of the five first interradials, which are in direct contact with the basals.
Interradial area plated with regular definitely arranged pieces. Posterior interradial
area differs but slightly. Tegmen calycis thickly plated. The plating of the apical
interradial region passes without break into that of the tegmen calycis. Ambulacra
not externally visible. Orals often indistinct. Anus subcentral. Ilhoducrinu*.
Gilbertsocrinus, Rhipidocrinus.

Family 3. Glyptasteridse.

Base dicyclic. With the exception of the first anal plate, which is in contact
with the posterior basal, the interradials do not touch the basals. Interradial
region of the apical capsule and tegmen calycis as in the Jikodocrinida: Oral plates
distinct. Anus subcentral. Glyptctster.

VIII

ECHINODERMATA—SYSTEMA TIC REVIEW

307

Family 4. Melocrinidse.

Base monocyclic, 3-5 basals. The basals in contact only with the radials.
Interradial areas of the apical capsule with numerous large regularly arranged plates.
Plates of the tegmen calycis often small and regular. Orals distinct. Anus sub-
central. Mi'locriiius (Fig. 252), Marietcrinus, Glyptocrinus, Stclidiocrinus.

Family 5. Actinocrinidae.

Base monocyclic, 3, rarely 4, basals. The first anal plate rests upon the circle of
basals ; the first interradials otherwise being in contact only with the circle of

vk

ir
ir

P~F:

CO

FIG. 252.— Melocrinus typus, Br.
p, Pinnulw ; '(/•, arms ; c?i, distichals ;
Cj, Co, first and second costal ; r, radial ;
ba, basal ; co, stem ; ir and id, inter-
radials.

Fie. 253.— Batocrinus pyriformis,
Shum. (after Meek and Worthen).
vk, Ventral capsule ; br, arms ; p, pin-
nulfe ; di, distichals ; cj, c-t, costals ;
r, radials; bit, basals; co, stem; ir,
interradials ; t/.br, points of insertion
of the anus.

radials. Tegmen calycis usually much arched, consisting of numerous firmly
connected plates, some of which at least are large, arranged in definite order. The
ambulacra of the tegmen calycis with their skeleton hidden, or only visible in forms
with flat tegmina. Anus subcentral. Orals usually distinct. Carpocrinus, Agarico-
crinus, Pericchocrinus, Mcgistocrinus, Actinocrinus, Tcleiocri/ius, Steganoerinus,
Aiiqihoracrinus, Physetocrinus, Strotocrinus, Batocrinus (Fig. 253), Eretmocriiim,
Dorycrinus.

Family 6. Platycrinidae.

Base monocyclic, 3 basals, which are unequal. Anal and interradial plates not
in contact with the basals. The very large radials together with the basals form

308

COMPARATIVE ANATOMY

CHAP.

almost the whole of the apical capsule. Each radial is connected with a short and
small costal plate. The various brachials which follow (distichals, palmars, etc.)

are free, i.e. belong to the freely out-
standing arms. In each interradius
there are at least three interradials,
which, however, appear more or less
shifted on to the oral side. In the
proximal (apical) interradial ring there
are no special anal plates, this ring
consisting in each interradius of 3-5
transversely placed plates, the central
one being the largest. Orals large.
Tegmen calycis mostly much arched.
The ambulacra and their covering
plates often appear at the surface.
Anus subcentral. Platycrinus (Fig.
254), Marsupiocrinus, J&udadocrinus.

Family 7. Crotalocrinidae.1

Base dicyclic. The apical capsule
consists exclusively of the typical
plates of the apical system (infrabasals,
basals, and radials), to which is added
an anal plate. The brachials of the
separate rays (to the fourth order)
firmly united by sutures. Arms very

"cE-*^ mobile, uniserial, long and much

Fio. 254.-Platycrinustriacontadactylus (after Branched; branches free or connected
M'Coy). dl, Distiehals ; c, costals ; /•, radial ; IM, together in such a way as to form a net-
basal; co, stem ; <Y, interradials; vk, ventral capsule, work around the calyx; this network

is either continuous or else divided into

five leaf-like lobes corresponding with the rays. Arms and their branches traversed
by large axial canals. Tegmeu calycis flat, richly plated with distinct orals, inter-
radials, and anals ; ambulacra externally visible, with large rigid covering plates,
which combine with the other plates to form the solid tegmen. Anus subcentral.

(This family is distinguished from all other Camerata by the presence of axial canals,
and by the mobility of the free joints of the arms.) CrutaJocrimis, Etvallocrimis.

Family 8. Hexacrinidse.

Base monocyclic. 2 or 3 basals. The first anal plate rests on the circle of
basals, and resembles the radials in shape. In other respects like the Platycrinidce.
ffexacrinus, Talarocrinus, Dicliocrin vs.

Family 9. Acrocrinidse.

Base monocyclic. 2 basals, separated from the radials by a broad zone of
small plates arranged in circles round the basals ; these form the largest part of
the apical capsule. Each radial is followed by 2 costals. The radials and

1 This family, originally placed near Cyathocriiius, was referred by Wachsmutli and
Springer, first to the Ichthyocrinoida? and then to the Camerata ; Bather, however,
would refer it to its original position in the Inaduiiata.

vin ECHINODERMATA— SYSTEMATIC! REVIEW 309

costals of the 5 rays laterally distinct. Interradials in two circles ; in the first
circle there are two plates to each interradius, and in the second circle only one,
which, however, is larger than the former two. Posterior interradius much larger,
with twice as many interradials, between which there is, further, an intercalated
vertical row of anal plates. Acrocrinus.

Family 10. Barrandeocrinidse.

Ba.se monocyclic. 3 basals. The first anal plate rests on the circle of basals.
The interradials rest upon the sloping oral ends of the radials. Arms bent back on
the calyx, fusing laterally with one another by means of their pinnulse in such a
way as to form a firm envelope around the calyx. Barrandcocrinus.

Family 11. Eucalyptocrinidse.

Base monocyclic. The apical capsule consists of 4 basals, 5 radials, 2x5
costals, 2x10 distichals, 3x5 interradials, and 1x5 interbrachials. No anal
plates. The tegmen calycis consists of 5 large interradials, 5 large and 10 small
interbrachials, the oral plates, and two other plates lying further up towards the
apex. Anus shifted quite to the centre. The plates of the tegmen form 10 niches ;
in the bases of these niches ambulacral grooves (two in each) run to the bases of
the 10 pairs of arm-branches, which are received into the niches. Eiicalyptocrinus,
Gallicriiius.

Order 3. Articulata (Ichthyocrinidae).

Skeleton flexible. Anal plates often occur in the posterior interradius of the
calyx. Base dicyclic. Three infrabasals of unequal size, which are usually hidden
by the uppermost joint of the stem. Radials perforated, with one or more
primary brachials. The circle of the combined radials and primary brachials is
closed, or else interrupted by one or more plates in each interradius. The brachials
of the first, second, and often also of the third order are incorporated in the calyx.
The radials and the separate brachials are articulated together. Arms uniserial.
Pinnulre appear to be wanting. Interradials irregular and varying in shape, size,
;ind arrangement, inconstant (may be either present or wanting in one and the
same species). In the posterior interradius there is often one asymmetrical plate.
Tegmen calycis only known in a few forms, soft and flexible, the plates lying in it
not being firmly fused together. Five separate orals of unequal size grouped round
the open mouth, the posterior oral being the largest. Ambulacra with their cover-
ing plates appear at the surface. Between them, there are interambulacral plates
which are occasionally distinguished by their remarkable size. Interambulacral
areas often sunk. Food grooves of the arms enclosed by movable covering plates.
A plated process (anal tube with anus ?) is found excentrically in the posterior
interradius of the tegmen.

Fam. Ichthyocrinidae — Palaeozoic forms : Ichthyocrinus, Forbcsiocrimis, Cleio-
i- rin us, Taxocrinus (Fig. 255), etc.

The unstalked genus Uintacrinus, from the upper Chalk, and the extant unstalked
genus Thaitmatocrinus (Fig. 256), ought probably to be classed here. In the latter
the uppermost ossicle of the stem is retained as centrodorsal. The dorsal cup
consists, apart from the centrodorsal, of 5 basals, 5 radials, and 5 interradials,
which last rest on the circle of basals, and alternate with the radials. Tegmen
with central open mouth, which is protected by a pyramid of 5 large separate
orals. Between the orals and the edge of the calyx (or the oral edge of the
interradials of the dorsal cup) the tegmen is covered with small irregular plates

310 COMPARATIVE ANATOMY CHAP.

indistinctly arranged in two to three rows. The anal interradial carries a short

-r

FIG. 255. — Taxocrinus multibrachiatus, Ly.
and Cass. if, /Yj, and iY.>, Interradials ; di, dis-
tichals ; 6«, basals ; ib, iiifrabasals ; co, stein ; r,
radials ; cj, &>, and 03, primary brachials.

FIG. 250.— Thaumatocrinus renovatus.
P. H. C. (after P. H. Carpenter). Calyx
from the anal side, cj, Co, and eg, Primary
brachials ; /•, radials ; en, points of insertion
of the cirri; oc/, centrodorsal ; lY, inter-
radials ; in, interradialia analia ; p«, proces-
sus analis ; ta, tubus analis ; j>, pinnula?.

jointed appendage,
arms with pinnula?.

Besides this there is a short anal tube. Five unbranched

Order 4. Canaliculata.

Calyx symmetrically five-rayed. Base dicyclic, the infrabasals usually not
separate, but atrophied or fused with the proximal columnal 5 basals, occasionally
not externally visible. Each radial is followed by 2 costals. Anal plates always
wanting (hence the regularity of the calyx). Interradials with few exceptions
wanting. Arms simple or divided (one to ten times). Tegmen calycis usually flat,
with open mouth and ambulacra appearing at the surface. Orals rarely present.
Tegmen calycis often plated with small loose-lying plates. Stem present either only
in young forms or also in adults. Basals and radials perforated by dorsal canals.
To this order belong, besides Mesozoic and Tertiary forms, most of the extant
Crinoids.

Family 1. Apiocrinidse.

Calyx consists of 5 basals of equal size, 5 radials and 2x5 primary brachials.
Distichals may also take part in its formation. Interbrachials and interdistichals
may occur. Tegmen flexible, with small plates. Arms more or less branched, con-
sisting of a single row of joints. Stem without cirri, usually expanding in its
proximal region to the same width as the calyx, but not containing the viscera.
Jurassic, to present. Apiocrinvs, Miffi'i-icrintis, and the extant Calamocrinus.

Family 2. Bourgueticrinidse.

Calyx consists of 5 basals and 5 radials. Brachials connected in pairs by
syzygial sutures. Five orals in the tegmen calycis. Interambulacral region other-
wise not plated. Ambulacra with covering plates, but without lateral plates. Stem,

vin ECHINODERMATA— SYSTEMATIC REVIEW 311

with root-like processes at its bajg, or with irregularly arranged cirri ; its proximal

P

:•

-co

FIG. 257.— Metacrinus Murray! (after P. H. Carpenter). Most of the arms and the larger part
of the stem broken off. p, PimmUe ; ci, cirri : ng, node.

ossicle usually enlarged. Upper Jurassic, Chalk, Tertiary, Recent. Rhizocrinus,
Jjouryiieticrinus.

312

' '( i M PARA TIVE ANA TOMY

CHAP.

1

,/

Ki . , - -

V -l 'i\

FIG. 259.— A, Cystoblastus Leuchtenbergi.
1, Interradial ; 2, 3, radial; 9, basal; 10, infra-
basal ; 8, anus ; 6, genital aperture. B, From
the oral side (after Volborth). 4, Mouth ; 5,
ambulacrum. Fig. 295, p. 332, shows the apical
side.

Frci. -2:^. Antedon incisa (after P. H.
Carpenter). I. Arms; 2, cirri.

FIG. 260.— Protocrinus oviformis. Eichwald
(after Volborth). •_'. Anus; 1, third aperture : 3,
ambulacrum.

VIII

ECHINODERMA TA— SYSTEM A TIC RE I 'IE Jl'

313

Family 3. Pentacrinidse.

Calyx small as compared with the stem and the arms ; it consists of 5 basals
and 5 radials. (In the genus Extract- inus the infrabasals are separate). Rays
divided one to ten times. Stem surrounded at intervals by whorls of cirri. No
root-like processes on the stem. One or more free primary brachials. Orals wanting
in the adult. Trias, to Recent. Pcntacrinus, Metacrinus (Fig. 257), E.'-tracrinus,
Balanocrinus.

Family 4. Comatulidse.

Adult free, larva stalked. The calyx is closed apically by the uppermost ossicle
of the larval stem, which is fused with the larval infrabasals ; this ossicle carries cirri
and becomes detached from the rest of the stem. It is called " centrodorsal."
The basals are externally visible, or else form an internal hidden rosette. Five or
ten simple or branched rays. The radials of the radial circle are usually followed,
in forms with divided arms, by two fixed primary brachials. Interradials wanting.
Orals wanting in the adult. Atclecrinus (basals externally visible), Eudiocrinnx,
Antcdon (Fig. 258), PromacTiocrimis, Actinometra (the only Crinoid genus with
excentric mouth). Since Jurassic times, many living species.

SUB-CLASS 2. Cystidea.

Body (calyx) oviform or spherical, plated with numerous very variously shaped
pieces, which are rarely quite regularly, and often irregularly arranged ; stalked, sessile,
or (rarely) free. Arms in many cases unknown, perhaps wanting in many forms ; when
present, weakly developed, resembling pinnules, and rising near the mouth. Food

i z

Fio. -2(<l.— Orocystis Helmhackeri,
Baur (after Barrande). 1-3, Ite F,«. oo2.-Agelacrmus cmcmnatensis.

three apertures.

grooves, arranged irregularly on the calyx, radiate from the mouth. At some dis-
tance from the mouth a second aperture (anal aperture), and between the two a third
aperture of unknown significance. Double pores or "pectinated rhombs " on some or
all of the plates. Palneozoic Pdmatozoa, whose organisation is still little understood.

Order 1. Cystocrinoidea (cf. the section on the perisomatic skeleton of the
Cystidea) : Porocrinus, G'aryocrinus, Echinocncrinus, Cystoblastus (Fig. 259 A and B).

Order 2. Eucystidea : Protoerinus (Fig. 260), Gflyptosphcerites, 0/v«v/.s'//.v (Fig.
261), Echinosphcera, Aristocystis, Ascocystis, Wesi-fr*, A n<- fact- in us (Fig. 262).

314

COMPARATIVE ANATOMY

CHAP.

SUB-CLASS 3. Blastoidea.

Armless Pelmatozoa, either pear-shaped, club-shaped, oviform, or spherical.
Body usually regularly radiate. Base monocyclic. Three basals, oue small and

FIG. 2(53.— Pentremites, from FIG. 2t>5.— Codaster bilobatus. M'Coy, from the oral

the side, without pinnules. 1, side, (after Etheridge and Carpenter). 1, Hydrospiiv

Interradial = deltoid ; 2, 3, radials ; slits; 2, lateral plates; 3, ambulacral groove; 4, mouth;

4, basal ; 5, ambulacrum : f<. spir- 5, radial ; ti, suture between two radials; 7. anus; S, inter-

acle. radial ; 0, ridge on an interradial.

Fio. -264. — Granatocrinus Norwood!
(after Etheridge and Carpenter): fn.m Fro. 266.-OrophoorimiB Stelllfonnls (after Ethe-

the apical side, with stem. ridge ^'l Carpenter) ; from th,- oral side. 1. Lateral

plates ; 2, covering plates of the ambulacra ; 3, hydro-
spire slits ; 4, anus ; f>. ambulacral groove ; 0, points
of attachment of the pinnules.

two larger. Five radials, more or less deeply cut out for the reception of the five
ambulacra. Five interradials lying above the five radials, and surrounding the

VIII

ECHINODEKMA TA—S YSTEMA TIC RE I 'IE Jr

315

peristome. One of these is perforated by the anus. The ambulacra are bordered
along each side by a single or double longitudinal row of jointed pinnule-like
appendages. Ambulacra with lateral and accessory lateral plates. In each ambu-
lacrum, under the lateral plates, there is a lancet-like piece, which is penetrated
lengthwise by a canal, and in which a radial ambulacral vascular trunk probably
ran. Ten groups of " hydrospires " on the radials and interradials. Peristome
covered by small plates, which are continued into the covering plates of the ambu-
lacra. For details cf. the section
on the Skeletal System, p. 328.
Palreozoic forms.

Order 1. Regulares.

Stalked Blastoids with sym-
metrical base. The radials resemble
one another, as do the ambulacra.

Fam. 1. Pentremitidae : Poi-
ti'i'iititcs (Fig. 263), Pnitn niitlili'n,
McsolJastus. Fam. 2. Troosto-
blastidae : Ti'oostucr tints, Metablas-
t/in, etc. Fam. 3. Nucleoblastidse :
Elceocrinus, SchizoWctstus, Cr>tjif»-
lilastus. Fam. 4. Granatoblastidse :
Gfranatocriniis (Fig. 264), Helem-
b/astus, Fam. 5. Codasteridse :
Cadaster (Fig. 265), PJumnschf*,,!,, Fl0' 2"37.-Astrocrinus Benniei (after Etheridge
„ , -, . i . ,„. ami Carpenter). 1, 4, 5, Interradials or deltoid plates ;

Cryptoschisma, Orophocrinus (Fig. 2> radials. 6j the inodi(1eil radial . 3> amblllaci.um . ,,

266).

the moditiod ambulacrum ; 7, basal ; 8, notch-like sinus.

Order 2. Irregulares.

Unstalked Blastoids, in which one ambulacrum with its radial is differently
developed from the rest.

Single family, Astrocrinidse : K/ei'tkerocrinus, Astrocrinus (Fig. 267), Pente-
phyllum.

I. General Morphology of the Eehinoderm Body.

The body of most Echinoderms, superficially observed, appears to
be of strictly radiate structure, but more careful examination reveals
that even in apparently perfectly radiate forms, e.g. regular Sea-urchins
and Star-fish, strict radiate symmetry is not found either in the
external or in the internal organisation ; in the latter, indeed, the
asymmetry is evident. Nevertheless, in order to facilitate a simple
description of the position and arrangement of the organs, terms are
habitually used which assume a strictly radiate structure. For the
purposes of description we may imagine the Eehinoderm body to
be spherical or egg-shaped. Two poles may be distinguished in it,
At the oral, adaetinal, or ventral pole there lies, in most Echinoderms,
the oral aperture, while at the other apical, abaetinal, or dorsal
pole in many forms is found the anal aperture. The line which
connects the oral and apical poles is called the principal axis.

316

COMPARATIVE ANATOMY

CHAP.

Round this principal axis many important parts of the body are
grouped in a radiate manner. The typical number of the rays is,
with few exceptions, five. In the Echinoderms, as in the radiate
Coelenterates, rays of the first, second, and third order may be distin-
guished. The radii or radial regions of the first order, in which the
principal organs lie, are called perradii, ambulaeral radii, or simply
radii. The five radii of the second order, Avhich regularly alternate
with these five principal radii, are the interradii or interambulaeral

FK;S. -_'(i8 ami 209.— Representatives of the principal divisions of the Echinodermata. In
Ki,L'. -68, in the morphological position ; in Ki^. 2i.iO, in the natural position with regard to the
sea-floor. A, Holothurian. B, Sea-urchin. C, Star-fish. D, Crinoid— a, Apical pole ; o, oral
pole ; an, anus.

radii. The far less important ten radii of the third order, each of
which lies between a perradius and an interradius, may be called
adradii. Between the two poles, at right angles to the principal axis,
we have the equator. In those Echinoderms which are provided with
large skeletal plates, the body and skeleton is further divided into two
zones, separated from one another by the equator ; these are the oral,
adaetinal, or ventral zone, and the apical, abaetinal, or dorsal zone.
In the centre of the former lies the mouth.

via EGHINODERMATA— MORPHOLOGY OF XKELEToX 317

While these terms facilitate the morphological description of the
body they do not take into account its position in the water, or

with regard to the sea-tioor, which is assumed to be horizontal.
Thus the normal position of the Star-fish and Sea-urchin is such that
the oral zone is directed downwards and the apical zone upwards ;
while the very reverse is the case in the Crinoids, where the oral
zone faces upwards and the body is attached to the substratum by a
stem which is inserted at the apical pole. In the Holothurians, again,
the principal axis of the body lies parallel to the substratum, and the
oral pole forms its anterior, the apical pole its posterior end.

For particulars as to the form of the body and the external
organisation of the various classes and orders of the Echinodermata,
cf. the Systematic Review, and also specially the two sections which
treat of the skeletal and ambulacral systems.

II. Morphology of the Skeletal System.

Meaning of the Most Important Lettering of the Figures.

a Apical pole.

am Ambulacral plates.

an Anus or anal area.

ap Ambulacral pores.

B Buccal plates.

In Basals.

In- Bracliials, arms.

t\ First costal or primibrach.

r, Second costal or primibrach.

ca Points of insertion of the cirri.

cd Centrodorsal.

ce or c Central plate.

ci Cirri.

co Column, stem.

cpa Covering plates of the ambulacral

grooves.

D Dentes, teeth.

dc Dorsocentral.

<li Distichal or secundibrach.

ds Dorsal shields.

i/o Genital aperture.

•ia Interambulacral plates.

inn Anal interradials or anals.

ib Infrabasals.

id Interdistichals or intersecundi-

brachs.

it- Interradials.
in Madreporite, pore - openings of

the stone canal.
a Nodal columnal.
a Oral pole, mouth.
in' Orals, or mouth-plates.
j> Pinnules.
pa Anal.
rs Radial shields.
/• Radials.
ss Lateral shields.
t Terminals,
to Anal tube or ventral sac.
i'k Tegmen calycis.
1-5 Interradii or interambulacral

areas of the Echinoidea.
I-V Radii or ambulacral areas of the

Echinoidea.

(In many of the diagrams of the apical system of various Echinoderms the
infrabasals are dotted, the basals shaded with concentric lines, and the radials
marked black. The brachials of the Crinoids are shaded with radial lines. )

Introduction.

The extensive comparative and ontogenetic researches which have
been made on the Echinoderms have shown that it is to some degree

318

COMPARATIVE ANATOMY

CHAP.

probable that certain pieces or plates of the skeleton are homologous
in all the divisions. It may be assumed that these plates composed
the primitive Echinoderm skeleton. From this primitive arrange-
ment, the skeletons of all known Echinoderms, whether extant or
extinct, appear to be derived, on the one hand, through the loss of
certain pieces of this primitive skeleton, and, on the other hand, by
the acquisition of new or secondary pieces of varied form, number, and
arrangement.

The hypothetical primitive Echinoderm skeleton consists of two
principal groups or systems of plates : (1) the oral, and (2) the
apical.

The oral system has five interradially placed oral plates,
arranged radially round the oral pole. This oral system develops

round the left ccelomic vesicle,
out of which the oral portion
of the coelom rises.

In the apical system the
following plates occur: (1) a
^^ central plate at the apical

"^mS^^^J^^y^ Pole; (2) a circle of five
Mw/P^plBr \>J3 radially placed plates, the
«|Kf^f3||P-^ infrabasals; (3) alternating
\^K. ^h^^JV^^i^ with these, five interradially

placed plates, the basals ; and
(4) around these, five radially
placed plates, the radials.
The apical system develops
on the right ccelomic vesicle,
from which the apical portion
of the ccelom is derived.

The stalked larva of
Antedon (Fig. 270) has
retained this system of plates
less altered than in any
known Echinoderm. It has all the typical pieces of the oral and
apical (or aboral) system of plates.

All the skeletal plates of the Echinodermata consist of carbonate
of lime. Their microscopical structure is very characteristic, so that
small fragments can at all times be recognised and distinguished
from similar fragments belonging to the skeletons of other animals.
The structure is a sponge-work ; and thin sections of the skeletal plates
or of the microscopic calcareous bodies appear to be perforated in a
lattice-like manner. The finer structure, especially of the spines of
Sea-urchins, is of great systematic importance.

FIG. 270. — Diagram of the apical system of the
Antedon larva, combined from various stages. Ex-
planation of the lettering on p. 317. The number of
infrabasals is here shown as 3, but these are produced
by fusion of 5, which number has also been seen.

viii ECHINODERMATA— MORPHOLOGY OF SKELETON 319

A. The Apical System (Calyx).
I. Echinoidea.

The part of the test which in the Sea-urchins is formed by the apical
system varies greatly in size. In the older and more primitive forms,
the regular Echinoidea, it is still somewhat extensive as compared with
the rest of the test (Fig. 271), but in modern, especially irregular forms
(Clypeastfidce and Spant&ngidce), it continually diminishes in relative
size till it is nothing more than a minute region at the apical pole. It
is possible to deduce the apical system of the Echinoidea directly from
the hypothetical primitive form by the help of certain Saleniidcc (Fig.
272). It is true that in the apical system of this family, as in that

*• Z

FIG. '27-2.— Salenia sp. Apical sys-
tem (after Lov&n). For lettering see
(f 0- Q 5 P- 317.

FIG. -JVl.— Tiarechinus princeps, Laube (after Lov&n).
1, Genital aperture ; 2, anus ; 3, basal ; 4, radial ; 5,
ambulacrum ; 6, the three upper plates of an inter-
ambulacrum.

of all other Echinoidea, the infrabasals are entirely wanting, but all
the other typical plates are present : i.e. a central plate, and round it
five basals, and outside these, alternating with them, five radials. In
the right posterior interradius each of the three plates, the central
and the two basals, is incomplete at the point where they meet. A
circular region, the anal region, in which the anus lies, is thus formed.
The anus, therefore, here lies asymmetrically in the apical system, and
this is the case in most Palceechinoidea and in most regular Euechinoidea.
According to the universally accepted terminology, it lies in the right
posterior interradius.

The typical system above described for the adult Saleniidce, has
been found to be repeated in very young specimens of other Sea-
urchins examined for this purpose (Echinus, Fig. 273 ; Toxopneustes,
Fig. 274).

Apart from these cases, where a primitive condition is shown by

320

COMPARATIVE ANATOMY

CHAP.

the presence of the central plate, riiost regular Echinoidea show the
following typical composition of the apical system : In the centre of

ra

FIG. L'73. — Echinus sp. (1 to 2 mm. long).
Apical system (after Loven). For lettering see
IP. 317.

the system lies the anal area, with a few large, or many small, calcareous
plates. A central plate cannot be distinguished. Within the anal

area lies the anal aperture,
usually excentric, less fre-
quently central. Round the
anal area are found the
circles of plates present in
all Echinoids, viz. the proxi-
mal circle of five basal
plates, and the distal circle
of five radial plates (Fig.
275). One, or several, or
even all of the radials may,
however, become wedged in
between the basals apically,
and finally may take part
in the limitation of the anal
area.

The ontogeny of To.ro-

Fn.;. ^75. — Toxopneustes drcebachiensis, O.F.M. ^WM.s/W shows that there is
Apical system of the adult (after Loven). For lettering at first jn fae amj area of
see p. 317. „ . . .,

very young Echinoidea one

large central plate (Fig. 274). Near this central plate, which ceases
to grow and degenerates, accessory plates appear. Among these

sfc

Fio. 274. — Toxopneustes droebachiensis.
O.F.M. (10 mm. long). Apical system (after
Loven). For lettering see p. 317. sh, Tubercles
carrying spines ; ap, anal plates.

via ECHINODERMATA— MORPHOLOGY OF SKELETON 321

accessory plates, which continually increase in number, the anal
aperture then forms, somewhat excentrically (Fig. 274). After a time
the central plate can no longer be distinguished from the accessory
plates.

As a rule, i.e. in the greater number of eases, the basals and
radials attain, in Echinoidea, the following special significance :

1. Each basal is perforated by a large pore or hole, through
which one of the five genital glands opens externally. On this
account the basals of the Echinoidea have long and almost universally
been called genital plates.

2. Each radial is also perforated by a narrow canal, which opens
at its surface through a single (rarely double) aperture. In this
canal lies the terminal tentacle of the water vascular system, the
frequently pigmented end of which projects somewhat beyond the
aperture. Since these collections of pigment were formerly considered
to be eyes, the plates (radials) carrying them were called the ocular
plates.

3. The fine, and usually very numerous, apertures of the water-
vascular system perforate one of the five basals (genital plates), which
becomes the madreporite (m). This is the right anterior plate.

It must, however, be noted that (1) the genital apertures are not
necessarily connected with the basal (genital) plates. The latter
must not be regarded as terminal appendages of the genital ducts,
but as independent portions of the test. For (a) the basals are solid
when first developed, and are only perforated by the genital pores
after the genital ducts have completely developed ; (b) the genital
apertures in some Echinoidea lie outside the basals. For example,
among the Clypeastroida, in some species of the genera Laganum, Encope,
Mellita, etc., the genital pores lie outside the apical system, between
its edge and the first two interradial plates which border it ; in
Clypeaster rosaceus, they lie in the five sutures between the interradial
plates, and are separated from the apical system by two or three pairs
of interambulacral plates (Fig. 277) ; and further, in another true
Echinoid, Goniopygus, the genital apertures lie interradially quite outside
the whole apical system. (2) The madreporite, through which water
flows into the stone canal, is not necessarily exclusively connected with
the right anterior basal (genital) plate. On the contrary, the neigh-
bouring genital plates, indeed, all the five plates, and in isolated cases,
even the neighbouring interradial plates of the corona may be perforated
by the afferent ducts of the stone canal. In Palceechinus each basal plate
is perforated by three pores, which are perhaps apertures of the stone
canal, perhaps genital apertures, or else partly the one and partly the
other. In no case, however, do the madreporic apertures extend to
the radials (ocular plates).

In the Echinoidea the primitive character and especially the
radiate structure of the apical system may be more or less strongly
modified. The original cause of such modification is principally to

VOL. II Y

322

COMPA If A TIVE ANA TOMY

CHAP.

be sought in the shifting" of the anus and anal area out of the
apical system into the posterior interradius ; by this shifting the anus
may come to lie at any point between the (aboral or dorsal) apical
system and the oral (or ventral) area. In its posterior and downward
shifting the anus thus does not carry the apical system with it, but
the latter remains on the dorsal side, although it is often shifted
somewhat excentrically anteriorly, rarely posteriorly. The whole body
is then bilaterally symmetrical, and when seen from above it is oval or
heart-shaped, etc., in outline. The line connecting the mouth with the
anus, which in the regular endoeyelie Echinoidea altogether or nearly

FIG. 276.— Holectypus clepressus, Cot-
teau. Apical system and neighbouring
parts of the perisome (after Lov&n). For
lettering see p. 317.

FIG. 277.— Clypeaster rosaceus, L. Apical
system and neighbouring parts of the perisome
(after Lov&n). For lettering see p. 317.

coincides with the vertical (principal) axis, now becomes the more in-
clined, i.e. approaches the more nearly to the horizontal, the further the
anal aperture is removed from the apical system into the posterior
interradius, and is shifted on to the under side (into the oral or actinal
region). Those Echinoidea in which the anal aperture has been
shifted outside the apical system are called exoeyelie or irregular.

Among the Palceechinoidea the genus Echinocystls (Cystocidaris)
alone is exocyclic. It appears that in this form the whole apical
system consisted merely of one madreporic plate.

Among the Euechlnoidea the three orders Holectypoida, Clypeastroida,
and Spatangoida are exocyclic.

a. Holectypoida (Fig. 276). In consequence of the wandering of the anus out
of the apical system, the posterior basal plate has lost its genital aperture, probably
in connection with the disappearance of the related genital gland (the place of which
has been taken by the rectum) ; in Cunocli/pcus and G(ifi'rnji//:/iis this plate has even

via ECHINODERMATA— MORPHOLOGY OF SKELETON 323

altogether disappeared. In some more recent species of the genus Holectiji>nx
the genital pore of the posterior basal plate reappears secondarily. The space in
the apical system, vacated by the anal area, is occupied by the madrepore (the
right anterior basal plate), which greatly increases in size, or else all the five basal
plates shift together towards the apical pole, the pores of the stone canal being then
distributed over several or all of them. The plates of the apical system may fuse
to a greater or less extent.

l>. Clypeastroida (Figs. 277 and 278). The whole apical system is here extra-
ordinarily reduced in extent ; it is, indeed, very minute. All the five basals are fused
together, and sometimes also fused with the radials. At least four genital pores are
retained. Where there are only four, it is always the posterior which is wanting.
The pores of the stone canal open in very various ways in the region of the fused
basals. Many scattered pores are sometimes found, or one single large pore, or the
pores open into irregular pits or grooves. In the family of the Clypeastridm the

FIG. 27S.— Laganum depressum, Less. Apical FIG. 279.— Apical system and neigh-
system and neighbouring parts of the perisome (after bouring parts of the perisome of Meoma
Loven). rp, Ocellar pores in the fused radials ; inp, ventricosa, Lamk. (after Loven). For
pores of the madreporite in a branched furrow. For lettering see p. 317.
lettering see p. 317.

genital pores have wandered out of the apical system ; they lie either at its edge, or
further removed in the sutures between the (paired) rows of interradial plates.

c. Spatangoida. The apical system of these exocyclic Echinoids is much reduced
in extent, although not so much so as that of the Ctypeastroida. It varies much in
detail, and in a few extreme forms (e.g. Pourtalesia) the primitive condition is to a
very great extent obliterated and destroyed.

1. In many geologically ancient forms the influence of the wandering of the anus
out of the apical system is seen in the disappearance of the posterior basal plate
(together with the genital pore belonging to it), and in the absence of a central plate.
The other basals and the radials (each of which has its pore) have all shifted,
and occupy an area which is sometimes circular or almost regularly pentagonal,
sometimes lengthened in the direction of the plane of symmetry. In the latter case
(Fig. 283), the two middle radials touch in the median line, separating the two
anterior from the two posterior basals. The apertures of the stone canal are found
in the right anterior basal, which is occasionally somewhat enlarged.

2. In most recent fossil forms and in the great majority of the extant Spatangoida,

324

COMPARATIVE AXAToMY

CHAP.

when their development is also taken into account, we rind the following condi-
tions. The posterior basal plate, again appears, but never has a genital pore. The
central plate also reappears. The apertures of the stone canal spread out from the
right anterior basal towards the centre, /.<'. on to the central plate. From this they

FIG. 280.— Apical system
of Abatus oavernosus,
Phil, (after Loven). red-
lettering see p. 317.

Fin. 281. — Apical system of
Spatangus purpureus. Hi; nun. in
size (after Loven). For lettering
see p. 317.

pass on to the posterior basal plate, and the three plates fuse together. The sutures
between them disappear, and so a large central madreporic plate is formed, which
in very many forms shows a tendency to increase in size and to spread out in the
direction of the posterior interradius, and thus more or less to press asunder the two

FIG. 282.— Apical system and neighbour-
ing parts of the perisome of Micraster coran-
guinum (after Loven). For lettering see p. 317.

Fie. 2S3.— Apical system and neigh-
bouring parts ol the perisome of
Holaster suborbicularis, Defr. (after
Loven). For lettering see p. 317.

posterior radials (Figs. 279-281). The genital aperture on the right anterior basal
may disappear, in which case only three genital pores remain. In isolated cases,
the left anterior basal plate may also lose its genital pore.

3. A method of dissolution of the apical system, unique among the Echinoidae,
is found in many Co/fi/ri/tilc' (Fig. 284). If we imagine that the elongated apical
system, described under heading 1 (Fig. 283), becomes very much more elongated
in the direction of the plane of symmetry, and breaks into two groups, one anterior
and the other posterior, we have the condition in these animals. The anterior
group contains the four basals, of which the right anterior is the madreporitic
plate, and three radials, viz. the anterior unpaired, and the right and left anterior.

viii ECHINODERM ATA— MORPHOLOGY OF SKELETON

325

The posterior group consists of the two posterior radial plates ; the posterior
unpaired (fifth basal) plate is wanting. The anterior group is separated from the
posterior by a row of plates which belong to the right and left posterior interradii, as
can be seen by comparing the figures. This arrangement is found in no other Echi-
noid. As in all Echinoids, however, the radials remain connected with the apical
ends of the five double rows of ambulacral plates, so that these latter divide in a

f m

w

v

g

V

FIG. 285.— Apical system and neighbour
ing part of the perisome of Pourtalesia
Jeffreysi, W. Th. (after Loven). For letter-
ing see p. 317.

FIG. 284.— Apical system
of Collyrites elliptica, Lam.
(after Loven). For letttrin.
see p. 317.

remarkable manner into three anterior ambulacra (trivium), the anterior unpaired
and the anterior right and left, and two posterior ambulacra (bivium).

4. The apical system is most of all reduced and obliterated in the peculiar
Spatangoid family of the Paurtalcsiidcc (Fig. 285). Let us take as an example P.
.A/7/r//.s-/. The whole system, which is irregularly pentagonal in outline, is shifted
forward, and separated from the apical ends of the two posterior ambulacra by
the uppermost plates of the posterior unpaired and of the right and left posterior
interradii. It, almost certainly, consists of four basal plates, each perforated by a
genital pore, but fused together into one single piece in which no suture can be seen.
In the central and anterior portion of this plate lie the scattered pores of the stone
canal. No radials can be recognised.

326

COMPARA TIVE ANA TOMY

CHAP.

Although there are good palseontological reasons for the generally accepted belief
that all known exocyclic (irregular) Echinoidea are descended from endocyclic
(regular) forms, it has been conjectured that these latter may themselves have had
exocyclic ancestors (which, indeed, are unknown to us). Thus the modern Sjiutnn-
cjoida and <J!iij>mstroida, for example, by the position of the anus in the posterior
unpaired interradius, may secondarily have attained a primitive condition. The
anus would then have wandered first from the posterior unpaired interradius to the
centre of the apical area, and then, in the exocyclic forms known to us. have shifted
back again in the same direction. This suggestion, which is of special significance
with reference to the primitive Pcfmatozoa, receives some (not vei-y satisfactory)
support from the fact that in the very old family of the Sn/cniidcc among the regular
Echinoidea, the anus lies at the posterior edge of the apical system in the oldest
forms, but during geological development approaches more and more near the centre
of the system, near which it is found asymmetrically (posteriorly to the right) in
the modern forms.

II. Asteroidea.

The typical plates of the apical system are not present in most
adult Star-fish, or at any rate cannot be made out among the numerous
calcareous pieces embedded in the dorsal area of the disc. There are,

however, exceptions to this rule. For
instance, in species of the genera Pinfn-
gonaster, Tosia, Astrogonium, Stellaster,
Ned rut, Ferdina, Pentaceros, Gymnasteria,
Sajtastcr, Ophidiaster, Zoroaster, the central
plate, the five basals and the five radials
can still be more or less clearly recognised
in the adults. Occasionally (in species
of Pentagonaster, Ghjmnasteria, Pentaceros,
and many Goniasteriike) there are even to
be found plates which in position corre-
spond with the infrabasals. The whole
apical system is specially well developed
in young specimens of the deep-sea Star-
fish Zon nister fulgcns (Fig. 286). The
aperture of the stone canal lies in the

in "young 7pecTmen" oTVoroaster right anterior interradius, outside the
fuigens (after siaden). For lettering basal ; the anus in the right posterior

interradius, inside the basal. In all

Asteroids, the madreporic plate and anus lie in these interradii of the
apical region (cf. the Echinoidea, Figs. 272-275).

The typical apical system can also be proved ontogenetically in
Star-fishes, even in forms in which it is absent or unrecognisable in
the adult. Five basals, a central plate and five radials are actually
among the first plates formed in the embryo Star-fish, in the very
order in which they are here named, though always after the terminals,
presently to be described, which appear first of all. Small plates,
appearing radially within the circle of basals, have been considered to

FIG. 2S<3.— Apical system of plates

vin EGHINODERMATA— MORPHOLOGY OF SKELETON 327

be infrabasals. This view is, however, not certain, because other
new and also radially arranged plates may be added to these, which
may thus also themselves possibly be accessory structures.

III. Ophiuroidea.

In this class, the plates of the apical system do not appear in the
embryo in exactly the same order as in the Astcroidca. First the five
radials and the central plate form, and, somewhat later, between the circle
of radials and the central plate, the five basals and the five infrabasals
appear. In many Ophiuroidea, an embryonic condition of the apical
system is retained in the adult, the central plate being surrounded by
the circle of five radials, while the basals and infrabasals are wanting

FIG. -.'.ST.— Plates of the apical system of the disc of
Ophiomusium validum (after P. H. Carpenter). For

lettering sec j>. 317.

FIG. 288.— Apical system of a
young Amphiura squamata (after
P. H. Carpenter). For lettering
see p. 317.

(species of the genera Ophioglypha, Ophiomastix, Ophiopyrgus, Ophiura,
Hemipholis, Ophioceramis, Ophiopholis, Ophiotrochus). In many others,
however, there are, besides the radials, the five basals, which may
vary greatly in size (species of the genera Ophiogli/pha, Ophiomastix,
Ophiomusium, Ophiura, Ophiopholis, Ophiozona, Ophiadis, Ophiokpis).
In Ophiomitra cxigua there is only the central plate with five basals
around it. In some Ophiuroidea a complete apical system is developed,
infrabasals being added to the basals, the radials and the central
plate (isolated species of Ophioceramis, Ophioglypha, Ophiozona, Ophio-
musium (Fig. 287), Ophiokpis). In very many Ophiuroidea the
calcareous plates developed at the apical surface of the disc are so
numerous that it is then impossible to recognise among them the
typical plates of the apical system. The adult Ophiuroidea have no

328 COMPARATIVE ANATOMY CHAP.

anus. The apertures of the stone canal are not found on any of the
apical plates, but ventrally, on one of the oral shields.

IV. Pelmatozoa.

In no other class of the Echinodermata do the plates of the apical
system form so large a part of the skeleton of the body wall (apart
from the arms) as in the Pelmatozoa, The body of these Echinoderms
consists of a central calyx, which contains the viscera, and usually
carries jointed appendages, radially arranged at its edge ; these are
the arms and pinnulse. Typically the Pelmatozoa are attached to
the sea-floor by their apical poles, with or without the intervention
of a stem ; in some the stem becomes separated from its attachment
(Pentacrinus), and may dwindle in size (Millericrinus), or may be present
only in the embryonic stages (Antedon), or there may be no trace of
either stem or attachment (Marsupites). The oral side of the calyx
(and also of the arms) is thus turned upwards, while the apical side
of the calyx (the dorsal cup) is turned downwards and either
surrounds the viscera like a bowl or carries them like a dish. The
plated test of this bowl or dish consists exclusively, or for the greater
part, of the plates of the apical system : the basals and the radials,
to which infrabasals may be added. The anal aperture always lies
interradially, usually on the oral side of the body and not con-
nected with the apical system.

Sub-Class 1. Crinoidea.

There are a good many Crinoids in which the apical system is
completely developed. The five radials and the basals are constant,
although the latter may be hidden. The infrabasals are inconstant.
The Crinoids in which the latter are present are said to have a
dicyclie base, those in which they are absent have a mono-cyclic
base.

A central plate has been observed in the larva of A)if<.'J«ti. It
occurs at the distal or root end of the larval stem, and ultimately
becomes severed from the animal.

The part taken by the plates of the apical system in the construc-
tion of the apical capsule varies greatly. In the stalked larva of
Antedon they alone form the skeleton of the apical side of the calyx ;
although an anal interradial has a transitory existence. The same is
the case also in many other adult Crinoids, which in this respect show
a primitive or an embryonic character (many Ii'mdniuifn. larrifonn'm
and many Inadunata fistulata, Encrinux, Marsupites, ll<>l<>pus, Hyocrinus,
Bathycrinus, and a few Canaliculata : Rhizocrinus, Pentacrinus).

In most Crinoids, on the other hand, the plates of the typical apical
system, i.e. the infrabasals (where these occur), basals and radials do
not form the whole skeleton of the apical capsule, but only a certain

viii ECHINODERMATA— MORPHOLOGY OF SKELETON

329

(often even very small) part of it ; other plates take part in its
structure, as we shall see more in detail when describing the peri-
somatic skeleton. The border of radials round the apical capsule
becomes more or less markedly disturbed by the appearance of

FIG. 289.— Apical system of Cyatho-
crinus. For lettering see p. 317. inn, Anal
interiadial.

Fin. 29U.--Marsupites ornatus. Plates of
the dorsal cup. For lettering see p. 317.

special " anal plates " in the posterior unpaired interradius ; these
specialised anals occur very frequently in palaeozoic Crinoids (Fig. 291).

The Crinoids with dicyelie base (with infrabasals, Figs. 289
and 290) are: (a) most Iitaduimta ; (l>) among the Camerata, the
families of the Rdeo<-riui<l<i' p. p.,
Bhodocrinidce, Glyptasteridce, and
Crotalocrinidce ; (c) the Articulata
(Ichthyocrinidce) ; (d) the Canalicu-
lata, in which, it is true, the infra-
basals are often either fused with
the uppermost joint of the stem
or atrophied, at least in the adult ;
such are conveniently termed
Pseudomonocyclic.

The Crinoids with monoeyclie
base (without infrabasals, Fig. 291)
are, apart from a few Inadunata,
the Camerate families of the
Melocrinidce, Actinocrinida>, Platy-
crinidce, Hexacrinidce, Acromnidce,,
JBarmndeocrinidce, Eucalyptocrinida.'.

Instead of the typical five infrabasals and five basal s there are
very often found four, three, or even only two plates in these rings ;
this is especially the case in extinct Crinoids belonging to the orders
Inadunata, Camerata, and Arficuhita. The plates are then almost

FIG. 201.— Actinocrinus proboscidalis. Plates
of the dorsal cup. For lettering see p. 317.

330

COMPARATIVE ANATOMY

CHAP.

always of unequal size, and it appears not unlikely that the reduction
of their number was caused by the fusing of neighbouring plates.
These characteristics necessarily destroy the strictly radial symmetry
of the dorsal cup.

Still further fusions may occur (among the Canaliculate').

The relative sizes of the plates of the infrabasal, basal and radial
circles vary greatly, but this is of no great interest to the comparative
anatomist.

Sub-Class 2. Blastoidea.

The Blastoidea are palaeozoic Pelmatozoa, whose stalked armless
body very often has the appearance of a bud (Fig. 263, p. 314).
Seen from the side, the body is an oval, truncated sometimes at
the apical, sometimes at the oral end. Seen from the oral or
aboral pole, its outline is in by far the greater number of (regular)
forms regularly pentagonal with rounded projecting angles, some-
times not unlike a short -

** armed Star- fish (Figs. 265

and 266, p. 314). In the
irregular Blastoids, on the
contrary (Eleutherocrinus, As-
trocrinus, Fig. 267, p. 315),
the radiate structure is dis-
turbed by the modified form
of one of the ambulacra.
The outline of the ovoid
body of Eleutheromnus, seen
from the apical or oral pole,
is irregularly pentagonal,
with three shorter and two
longer sides, the latter be-
longing to the left posterior
and the unpaired posterior
interradii. In Astrocrimi*.
the body is flattened in the
direction of its principal axis,
FIG. i».i:>.- Apical system of Pentremites. m/-M,A\is and, when seen from the oral

passing through mouth and anus;* the smaller;?/, and } j , } fc

2, the two larger basals ; ir, interradials ; r, radials. l J

metrically four-lobed, the

lobes being of unequal size. The largest of the lobes lies diametrically
opposite the abnormally shaped ambulacrum, which is on the smallest
truncated lobe. The two other middle-sized lobes are almost alike in
form (Fig. 267, p. 315.)

The whole body of the Blastoids is plated. The test consists,
apart from the ambulacra, of three circles of plates (Fig. 292), two of
which belong to the typical apical system of the Echinodermata, while

vin ECHINODERM AT A— MORPHOLOGY OF SKELET»X 331

the third consists of perisoraatic plates, which, in all probability,
correspond with the primary interradii of the Crinoids.

The first circle at the apex is that of the (interradial) basal plates.
There are always three of these, one smaller and two larger of equal
size, as also occurs in the Crinoids. The monocyclic base of the
Blastoidea is thus symmetrical. But the line of symmetry (the so-
called dorsal axis), which passes between the two larger plates and
through the small unpaired plate, does not coincide with the
symmetrical (ventral) axis of the body, which passes through the
mouth and the anus, the latter lying in the posterior inter-radius on
the oral surface. The smaller unpaired basal plate lies in the left
anterior interradius. If we imagine the two larger basal plates cut
into two similar parts by radial lines of division, we obtain the five
equal-sized, strictly radially arranged, and interradially placed basals
of most other Echinoderms. The uppermost ossicle of the stem
is inserted at the point where the three basals of the Blastoids meet.

The circle of the basals is immediately surrounded by that of the
radials. The typical number of five is always retained in these,
which, in regular Blastoids, are strictly radiate in
their arrangement. These are called fork-pieces,
because each of them is produced upwards, i.e.
orally, in the shape of a tuning-fork, the two
limbs holding between them the distal end of an 1 ,
ambulacrum. The radials form a closed circle,

their lateral edges being contiguous.

The third circle of plates is in immediate
contact with the radials, and surrounds the peri-
stome. It consists of five interradial plates,
which, in regular Blastoids, are strictly radial ;
these are the interradials or deltoid plates.
These plates do not form a closed circle, as they £ &

are separated from one another by the five Flo 203 _Eieuthero-
ambulacra. The apical edges of each deltoid crinus cassedayi, from
plate rest on the oral edges of the contiguous the apical si,u- (after Ethe-

* , . . ,. T! ,. i • ridge ami Carpenter).

forks of two consecutive radials or fork pieces. nn.bb, Axis passing through
The relative sizes of the basals, radials, and inter- the mouth ami the anus ;
radials of the Blastoids vary greatly (<•/. figures). *, the smaller y and *, the

,., ,.,.,.. .11 two larger basals; r, radials;

One oi the five interradials, which is distinguished «„, anai side.
as the posterior, is perforated by the anus.

In the irregular Blastoids (Fig. 293), which are without stems,
all the plates of the regular forms are found, but are, naturally,
irregularly developed. The radial which supports the modified ambu-
lacrum is smaller than the other radials and differently shaped. It
appears shifted quite on to the oral surface. At the same time, the
pair of basals (y and z) which flank this radial are much prolonged
orally as narrow plates.

It cannot at present be decided whether there are skeletal pieces

332

COMPARA TI I 'E ANA TOMY

CHAP.

in other Echinodermuta homologous with the interradials of the
Crinoidt'tt and the Blastoidea. In all endeavours to answer this
question the following plates should be kept in mind : in the
Ophiuroidea the interradially placed plates between the circle of radials
and the oral side (Fig. 287, p. 327), and among the Echiiioiilm, in
Tiarechwa* (Fig. 271, p. 319), the central of the three interradial plates
of an interambulacral area.

Sub-Class 3. Cystidea.

The spherical, pear-shaped, egg-shaped, or cup-shaped body of the
Cystidea is also enclosed in calcareous plates. In one of the principal
groups, that of the Eucystidea, the plating consists of numerous con-
tiguous plates arranged without any recognisable order. In this case

-i/l?

FIG. 204.— System of plates of the apical capsule
of Caryocrinus ornatus, spread nut (after Hall). For
lettering see p. 317.

FIG. 205.— Cystoblastus Leuchten-
bergi, from the apical side. 11, Point
of insertion of the stem ; 8, anus ;
10, infra basals ; 12, pectinated rhonil'-..

a typical apical system of plates cannot be distinguished. In the
other principal group, the Cystocrinoidea, certain forms of which show
near relationship to the Crinuiilt'ii and Blastoidea, the test consists
of a relatively small number of plates, and a true apical system can
be found round the apical pole.

The forms assumed by this apical system may be grouped around
two central types : Caryocrinus and Eehinoencrinus. The group
(.'((/•t/ucriiiiix (Cwylocrinus, Hemicosmites, Juglandocfrinus) has its plates
arranged in six rays ; while the group Eehinoencrinus (Callocystis,
Lepadocrinus, Apiocystis, Cystolhistiis, Gh/ptocystis, Pleurocystis, Prim»-
c//.sfa>, Pseudocnmis, etc.) shows the typical five-rayed arrangement of
the plates. In both groups the base is dicyclic, i.e. there is a circle
of infrabasals inside the circle of basals.

Caryoerinus, six-rayed (Fig. 294). — The circle of infrabasals
consists of four plates, two larger (which are contiguous) and two
smaller. Each of the two larger plates is double. Outside the circle
of the infrabasals lies a closed circle of six interradial basals, and

VUI

EGHINODEEMATA— MORPHOLOGY OF SKELETON

this is surrounded by a closed circle of six radials. These plates,
together with two accessory plates (interradials ?), form the whole
test of the cup of the attached Caryocrinus, from the point of
insertion of the stem to the base of
the arms. The anus lies excen-
trically on the oral surface, in the
(interradial) prolongation of the
suture between the two larger infra-
basals ((/. Figs. 294, 295).

Eehinoenerinus, five-rayed (Fig.
29G). --The circle of infrabasals
consists of four plates, one large
posterior plate and three smaller
ones. The larger plate is double
(i.e. consists of two fused plates).
Outside the circle of infrabasals
comes the closed circle of the five

basalS, and OUtside this that of the CU.P °f Eehinoenerinus armatus, spread

(after Forbes). For lettering see p. 317.

five radials, between which acces-
sory pieces are intercalated, the homologies of which cannot be made
out. The anus lies posteriorly to the right. In Cystdblastus the
radials, like the radials or fork-pieces of the Blaxtoidea, have deep
incisions on the oral side for the reception of the ambulacra (cf. Fig.
259, A and B, p. 312, and Fig. 295).

i. Jiiii.— System of plates of the dorsal

out

B. The Oral System of Plates.

In certain Echinodermata (Pelmatozoa and Ophiuroidea) there is a
system of plates surrounding the oral (ventral, actinal) pole, and thus
diametrically opposite to the apical system. This system develops
round the left ccelomic vesicle of the larva in a Avay similar to that in
which the apical system develops round the right vesicle. The oral
system is, however, much simpler than the apical, and consists of one
single circle of five plates (less frequently six, in the six-rayed arrange-
ment of the whole system) ; these plates, placed interradially, corre-
spond in the oral system with the basal plates of the apical system,
and are called oral plates.

In our considerations of this oral system AVC again find the best
starting-point to be the stalked larva of Aniedon (Pcntacrinus stage).
In a young stage of this larva the oral surface of the calyx appears
vaulted over by a roof closed on all sides. The surface of the calyx
thus forms the floor, and the vault the roof, of a closed cavity, which
is called the oral or tentacular vestibule. At the centre of the floor
the oral aperture breaks through, connecting the intestine with the
vestibule. The mouth is thus at this stage not connected with the
exterior. The fifteen primary tentacles, which rise on the disc of
the calyx, also cannot project externally, but are covered over by the

334

COMPARATIVE ANATOMY

CHAP.

roof of the vestibule. This roof is formed of five interradial lobes,
supported by five interradial skeletal plates, the oral plates. An aper-
ture only arises secondarily at the apex of the roof, and the five oral
lobes separate in such a manner that the tentacles can project through
the clefts between them. The mouth is now in open communication
with the exterior.

At first the five oral plates rest directly on the oral edges of the
basal plates of the apical system. But in proportion as the calyx
increases in size, and the arms grow out, the distance between the
basals and the newly-formed radials, which support the arms, on the
one hand, and the oral plates on the other, becomes greater and greater,
since the latter remain at the centre of the tegmen calycis, surround-
ing the mouth. There thus arises, between the bases of the arms and
the circle of the oral plates, which in comparison with the continually

growing calyx be-
comes more and more
insignificant, a cir-
cular zone, the peri-
pheral zone of the
tegmen calycis. The
food grooves running-
out from the mouth,
passing between the
five oral lobes, tra-
verse this peripheral
zone of the tegmen
to the bases of the

FIG. :2<i;.— Haplocrinus mespiliformis (alter Wachsmuth and arms. Tins peri-
Springer). A, From the anal side ; B, from the oral side. 1, Orals; p]ieral Z0ne COlltillU-
2, oral pole ; 3, anus ; 4, radials ; 5, inferradial ; 6, basals ; 7, tirst ••, . . .
brachial ; 8, point of attachment of the arm. <^V increases 111 S1Z6,

while the central part,

surrounded by the five oral lobes, does not grow further, and forms
an ever-diminishing central region of the tegmen calycis. Finally, the
oral plates, with the lobes, are entirely resorbed, and the minute
central zone can no more be distinguished ; the whole oral surface of
the Antedoii calyx is a free disc, by far the greater part of which has
been formed outside the base of the oral pyramid. In the centre of
this oral disc the mouth lies uncovered, and on the surface of the
disc the food grooves are visible running out radially to the bases
of the arms.

Among the immense array of forms comprised under the crinoids
we find a few groups with five oral plates forming, as in the larva of
Antedon, the whole skeleton of the tegmen calycis. In the Inadunata
larviformia, type Haplocrinus (Fig. 297), there is actually a closed
pyramid of five oral plates, which, at the edge of the calyx, rest on
the radials of the dorsal cup. Only at the bases of the arms do
the five oral plates separate to form five radial apertures, through

VIII

EGHINODERMATA— MORPHOLOGY OF SKELETON

335

which the food-grooves pass out on to the arms. The posterior oral
plate is somewhat larger than the others, and has a perforation
which may be the anus (?).

The same condition is found in the extant genera Holopus and
Jfi/ocrinus (Fig. 298), the extant unstalked genus Thaumatocrinus,
and the extant canaliculate
genus Rhizocrinus. All these
genera possess five oral plates,
which, however, are separate,
and do not form a closed
pyramid ; the mouth, there-
fore, is in open communica-
tion with the exterior between
them. Compared with the 8-
larva of Antedon and with
ffaplocrinus, Holopus shows
the most primitive (or em-
bryonic) condition, since in
it the oral pyramid is large,
covering nearly the whole of
the tegmen, so that between

its base and the edge of the Pl°- 20S.— Hyocrinus Bethellianus (after P. H.
rnlvv milv a vprv small npri Carpenter). Tegmen calycis. 1, Axial canal of the

caiyx only a very small pen- brachials . 2> extension of body cavity in the anu .

pheral Zone remains. In 3, food groove of the arm ; 4, smaller plates of the

HlfOCrimiS (Fia;. 298) also, and tegmen; 5, orals; 6, anal cone; 7, oral edges of the

radials

Thaumatocrinus the orals are

still of considerable size, but the peripheral zone, which is beset with
small closely -crowded plates, is somewhat broader than in Holopus

(about one-fifth the diameter
of the whole tegmen). In
Rhizocrinus lofotensis the orals
are smaller, and in Rhizocrinus
Rawsoni they are almost rudi-
mentary, so that the zone
which surrounds them forms
the greater part of the
tegmen.

In the Cyathocrinidce (Tn-
adunata Jistulata), five large
plates can sometimes be dis-
tinctly made out in the centre
of the plated tegmen ; some-
FIG. 299. — system of plates of the tegmen of times, however, irregular

Platycrinus tuberosus (after Wachsmuth and pieces are found in their
Springer). For lettering see p. 317. ^^ wh(jn they ^ ^

tinct, the posterior plate is the largest, and is sometimes shifted
anteriorly between the others. In all cases they cover the mouth in

336 COMPARATIVE ANATUMY CHAP.

such a way as to hide it. These plates are by some regarded as
orals.

In the Camerata (Fig. 299) five supposed oral plates (or) can almost
always be distinguished in the centre of the richly and rigidly plated,
often highly arched, tegmen. They close together firmly over the
mouth. The posterior oral is larger than the rest, and presses in
between them.

As far as is known, in the Articulata (Ichthyocrinoidea) also, five
orals can be distinguished at the centre of the richly but loosely
plated tegmen. But, in this case, they are separate, and surround an
open mouth. The posterior plate is larger than the rest.

In the Canaliculate (with the exception of the above-named genus
RTiizocrinus} the orals are altogether wanting in the adult.

la the Blastoidea the oral region is covered by a roof consisting
of numerous small plates usually without definite arrangement, which
are continued as covering plates over the ambulacra. In a few forms,
however, and especially in Stephanocrinus, five orals can be made out.
In Stephanocrinus these five interradial orals, resting on the inter-
radials (i.e. the deltoid pieces), form a closed pyramid over the oral

region.

In many Cystidea, also, the mouth is arched over by an oral
pyramid. In CyatJiocystis, the five oral plates forming this pyramid
are more or less equal in size, but in species of the genera Sphceronis,
Glyptosphcera, and Pirocystis the posterior oral is, as in so many
Came rat it, larger than the rest. In the six-rayed Cystid Cari/m-rimix
this latter is the case, one of the six orals having shifted from behind
forward between the other five, which surround it symmetrically.

In the Ophiuroidea, on the oral (lower) side of the disc, there is
in each interradius a plate, usually distinguished by greater size. One
of these plates, which are called buecal shields (Fig. 245, p. 300),
is, as madreporite, perforated by the pores of the water-vascular
system. In the pentagonal larva of Amphiura these buecal shields
appear at the edge of the oral side. They have been homologised,
probably correctly, with the orals of the Pelmatozoa.

In the Asteroidea, on the lower surface of the disc at the edge of
the mouth, in each interbrachial region, there occurs a skeletal plate
of very various shape, which is called the odontophore (Fig. 310, p.
352). These plates, which might be described as the proximal OP
basal plates of the interbraehial system, may correspond with the
orals of the Pelmatozoa and the oral shields of the OphiuroiiJeti, although
they may be pushed below the surface by the oral plates (the first
pairs of adambulacral plates), and are usually completely covered
externally. They arise early in the larva of Asterias (after the five
terminal plates, the five basals, the apical central plate, the ten oral
ambulacral plates, and twenty other ambulacra! plates are formed),
interbrachially between the oral ambulacrals.

Orals have not been discovered in the Eehinoidea. Whether

vin ECHINODERMATA— MORPHOLOGY OF SKELETON 337

certain pieces of the calcareous ring of the Holothurioidea corre-
spond with the orals of other Echinoderms cannot at present be
determined.

C. The Perisomatie Skeleton.1

All those skeletal pieces which protect the body, between the apical
and the oral systems, taken together, form the perisomatie skeleton
of the Echinodermata. It is obvious that the extent of the periso-
matie skeleton must vary inversely with that of the polar (apical and
oral) systems. AVhere the polar systems form only a small part of
the body wall the perisomatie skeleton is the more strongly developed,
and rice versa. In the Blastoidea, for
example, nearly the whole of the test
is formed by the polar systems
(especially the apical), while in most
Echinoidea, Asteroidea, and Ophiuroidea,
the perisomatie system covers nearly
the whole body. Where the equatorial
zone of the body is produced into
variously shaped branched or un-
branched arms, as in most Pelmatozoa,
Asteroidea, and OpMuroidcu, the skeleton
of these arms is exclusively formed by
perisomatie pieces. It is at present
impossible to prove any definite
homologies between the parts of the
perisomatie systems throughout the
Echinodermata.

I. Holothurioidea.

as

FIG. 300.— Microscopic calcareous bodies
of Holothurioidea. 1, Anchor and anchor
plate of Synapta inhserens, O. F. M. ; 2,
"stool" of Cucumaria longipeda, Semp ; 3,

In the cutis of the Holothurioidea,
well in the body wall as in the
wall of the tentacles, ambulacra, tube- c™ciform b°dy of Cucumaria crucifera

,. n , , , .,, , Semp; 4, rod from one of the tube-feet of

feet, and ambulacra! papillae, there are sticopus japonicus ; 5, supporting plate
found enormous numbers of micro- from one of the tube-feet of stychopu*
scopically minute calcareous bodies of w°nicus =„ 6' " sto°i " of Hoiothum

, £ . J . Murray i ; 7, rod from the ventral ambula-

definite shapes (b Ig. 300). ihese give cial appendages of Oneirophanta mutabUis,

the integument a firm and rough Th6el ; s> latticed hemisphere of Colochirus

ooTisist-pnrv Thpir nrinriml sio-nifi cucumis- SemP: 9. "wheel" of Acantho-

icy. principal signm- trochus mirabili8> Dan. and Kor.

cance may well be that of protection.

These small calcareous bodies may be called, according to their shapes,
"anchors," "wheels," "rods," "anchor plates," "crosses," "lattices,"
"stools," "buckles," "biscuits," "cups," "rosettes," etc.

1 On the author's use of the term " perisomatie," see footnote, p. 362.
VOL. II 7

338 COMPARATIVE ANATOMY CHAP.

The shape and method of association of these bodies is of importance for
classification, especially for distinguishing one species from another. Nearly all
their various forms can be traced back, in a way which cannot here be further
described, to a common form, viz. to a very short rod, which tends to branch
dichotomously at each end. In some Dendrochirotce (Psohts, ThccUa, etc.) the
calcareous bodies upon the (physiologically) dorsal side of the body attain a
specially large size (1 to 5 mm.), so that the back appears to the naked eye to be
covered with scales or plates (Fig. 228, p. 287).

In the l>i 'in/ /•"! -Inrotcc an anterior part of the body, the proboscis, is invaginable.
At the posterior boundary of this proboscis (when evaginated) five calcareous oral
valves are found in a few genera. When the proboscis is invaginated these come to
lie close together in the form of a rosette, which protects the aperture. In Psolus
these five oral valves are placed interradially, and each is a large triangular calcareous
plate (Fig. 228, p. 287) ; in Colochirus, Adinocucumis, etc., they are arranged radially
and consist of compact masses of calcareous granules and ambulacral papillre. In
many Aspidochirota and Dfinlrm-liiriita radially or iuterradially arranged anal
valves (anal plates or anal teeth) also occur round the anus.

II. Eehinoidea.

The skeleton of the Ecliinoiilca forms a plated covering called the
test, which encloses the viscera. The greater part of this test is
composed of the plates of the perisomatic system, since, as a rule, the
plates of the apical system (the central plate, the periproctal plates,
the basals and radials) occupy but a small, and even sometimes a
minute, area at the apical pole. There are, however, exceptions to
this rule, e.y. the Triassic genus Tiarechinus, in which a considerable
portion of the test is formed by the plates of the apical system (rf.
Fig. 231, p. 289).

The form of the shell is thus, as a rule, in the Eehinoidea, deter-
mined by the perisomatic skeleton. The horizontal outline of the
shell, i.e. the outline seen when an Echinoid shell is viewed from the
oral or the aboral pole, is called the ambitus. This ambitus in
regular EcMnoids is, as a rule, strictly circular, or else pentagonal with
rounded corners ; less frequently it is oval, in which case the greatest
diameter of the ambitus need not coincide with the symmetrical axis.
In iri't-t/ult.rr Eehinoidea the ambitus is symmetrical, being generally
elliptical (lengthened from before backward), or else egg- or heart-
shaped.

In all ErJtiitn iiirit, except the Spatangoida, the mouth lies at the
centre of the oral surface of the test ; in the Spatangoida it has shifted
anteriorly on this surface. The mouth, however, always remains the
centre round which the plates of the perisomatic skeleton are
grouped.

We have already seen that in regular endocyclic forms, the anus
lies in the centre of the apical system, but in exocyclic forms it leaves
the apical system and enters the posterior interradius, where it may
approach the ambitus, or even cross it on to the oral surface, always,
however, remaining in the posterior interradius.

vin ECHINODERMATA— MORPHOLOGY OF SKELETON 339

The whole perisome, from the mouth to the apical system, falls
into two sections: (1) a small portion surrounding the mouth, the
peristome or oral area ; and (2) the larger remaining portion be-
tween the peristome and the apical system, the corona. In the peri-
stome the skeletal pieces are usually loosely embedded near one
another, or imbricate one with the other, remaining movable one
against the other. Sometimes the peristome is membranous, without
skeletal pieces. In the corona the skeletal pieces are usually firmly
connected with one another by means of sutures, like the plates of
the apical system, together with which they form a rigid test. In
dead Echinoidea, and in nearly all fossil forms, this test remains
intact, while the skeleton of the peristome falls to pieces, and is
therefore rarely preserved.

The perisomatic skeleton in all Echinoidea consists of two systems
of plates, which run from the apical system over the ambitus to the
mouth as ten meridional zones : five of these zones or systems of
plates are placed radially, and these are called the ambulacra. These
five zones, on which the tube-feet rise, are always in contact with the
five radial (ocular) plates of the apical system, so that each ambulacrum
touches an ocular plate with its apical end. The ambulacral plates
are perforated for the passage of the ambulacral vessels, which serve
for swelling the tube-feet. The five other zones or systems of plates
are interradially placed, and are called interambulaera or interambu-
lacral plate systems. They alternate regularly with the ambulacra.

Considering the perisomatic skeleton of the Echinoidea more closely, the follow-
ing special points are worth attention.

(«•) The Number of the Vertical or Meridional Rows of Plates in the Ambulacra
(radii) and Interambulaera (interradii).

In all EuecJiinoidea (from Devonian times up to the pi-esent), the corona consists
of twenty meridional rows of plates, ten of which united in pairs belong to the
ambulacral system, and ten also in pairs to the interambulacral system. Five
double rows of ambulacral plates thus regularly alternate with five double rows of
interambulacral plates.

In the exclusively Palaeozoic PalarrJ/ iimi'/ca, the number of meridional rows of
plates in both ambulacra and interambulaera varies. The number of rows in all
the five ambulacra and in all the five interambulaera of individuals of one and the
same species is, however, always the same.

In the ambulacra, however, the number of rows of plates in the Palaeechinoidea
is usually two. The Melon it ida: (Fig. 301) form the only exception, having four to
ten rows in each ambulacrum.

In the interradii, on the other hand, the number of rows of plates varies.
Bothrioddaris has only one single row of plates in each interradius. In all other
Palct'cchinoidea there are more than two (3-11) rows of plates in each interradius
(Fig. 230, p. 289). The interesting genus Tiarcchiiius (Fig. 231, p. 289) is dis-
tinguished by the great simplicity of its interradial system of plates ; in each inter-
radius there are only four plates, a single one at the edge of the peristome — the
large marginal plate of the peristome — and three intercalated between this and the

340

COMPARATIVE ANATOMY

CHAP.

•UYl

adjoining apical system, these plates being separated by meridional (perpendicular)

sutures.

The plates of the Echinoidea are most frequently pentagonal. In the two per-
pendicular rows of an ambulacrum or
an interambulacrum the consecutive
plates usually alternate in such a way
that the suture between the two rows
forms a zigzag line. The sutures
between the plates, which lie one
below the other in a row, usually run
horizontally (Fig. 232, p. 291).

(l>) The Pores perforating the Plates
of the Ambulacral System.

As a rule, in the Echinoidea, the
pores occur in pairs. These double
pores occur only on the ambulacral
plates. One double pore belongs to
each ambulacral foot.1 From the
ampulla, under the test (at its inner
side), two canals run out, which,
running separately through the plate,
unite at the base of the tube-foot to

form a single canal, which runs through the foot and ends blindly at its tip.
Originally, there was only one pair of pores on each ambulacral plate. Where two
or more pairs occur on one plate, the plate can be proved to be composed of just as
many fused plates as there are pairs of pores. Primary plates are such as reach
from the lateral edge of a two-rowed ambulacrum as far as the median suture
between the two rows of ambulacral plates. Half plates are such as do not reach
the suture, and included plates such as do not reach the edge of the ambulacrum.
Isolated plates reach neither the edge nor the median suture of the ambulacrum.

Besides the double pores there are, in the Clypeastroida and Spatangoida, single
pores as well, to which small tentacles belong. The arrangement of these pores
varies, and they are often not confined to the ambulacra, but are also found on the
interradii, especially on the oral surface. Occasionally they are scattered, often in
grooves, the so-called ambulacral grooves, which radiate out from the peristome,
and may stretch more or less far towards the ambitus or even beyond it, and may
be more or less branched.

FIG. 301.— Apical system and adjoining peri-
some of Melonites multipora, Norw. (after Meek
and Worthen). For lettering see p. 317.

(<•) The Symmetry of the Echinoid Shell.

The test of the regular Echinoids (Cidaroida, Diadi'iiinioida, and most
Palcccchinoidca), viewed superficially, appears to be strictly radiate. The anal area
lies at the apical, and the oral area at the diametrically opposite oral pole. All the
ambulacra and interambnlacra appear similar one to the other, and the ambitus,
with few exceptions, is circular or regularly pentagonal with rounded corners. In
the Holcctypoida also the test, as a rule, appears radial, with regard both to the
circular (or regularly pentagonal) form of the ambitus and to the similar develop-
ment of the ambulacra and interambulacra. The peristome occupies its place at the
centre of the oral surface. Notwithstanding this, the longitudinal axis and the

1 For the different forms and arrangements of the.se feet or tentacles, cf. section on the
ambulacral system, p. 416 et seq.

viii ECHINODERMATA— MORPHOLOGY OF SKELETON 341

plane of symmetry can be recognised in the Holedypvida at the first glance, because
the anal area lias shifted out of the apical system, and into that interradius which is
called the posterior interradius. The same is the case in the Clypeastroida, and, in
a still higher degree, in the Spatangoida. In the Clypeastroida the peristome with
the mouth still remains in the centre of the oral surface, or only very slightly shifts
away from this position. But the ambitus is no longer circular or regularly
pentagonal ; its outline appears symmetrically lengthened or shortened in the
direction of the longitudinal axis, in such a way that, even in a superficial view, the
plane of symmetry is discoverable. Apart from the fact that the posterior inter-
radius is at once recognisable by the anus lying in it, it is often further distin-
guished in the Scittcllida'- by a perforation through the test (lunula), which never
occurs in the other interradii. Further, in the SciiMlido:, the bilateral symmetry
is often distinctly indicated by the number and arrangement of the radial lunulre,
or of the marginal incisions (Figs. 233-235, pp. 292, 293).

The bilateral symmetry, which is most pronounced in the Spatangoida, culmi-
nates in the remarkable family of the Pourtalesiidce. The ambitus, which varies
greatly in details, is frequently egg-shaped, or heart-shaped, and in Pourtalesia
rtask-shaped. Not only does the anus always lie somewhere in the posterior inter-
radius, but the oral area also shifts from the centre of the oral surface, moving more
or less far along this surface anteriorly. In the Cassidulidce all the transition stages
between a central and a frontal position of the oral area occur. Since the mouth,
with the oral area, always forms morphologically the centre of all the systems of radii,
in shifting anteriorly it necessarily draws along with it the systems radiating out
from it. We shall return later on to the dissimilarity in the ambulacra, and especially
to the abnormal development of the anterior ambulacra, and consequent formation of
the bivium and trivium, to the special form of the peristome of the Spatangoida, etc.

The apical system also does not always remain at the dorsal centre of the test,
but shifts more or less far forward (less frequently backward), and the highest point
of the test may then come to lie in front of (less frequently behind) its central
point (Figs. 236-238, pp. 294, 295).

We have seen that in exocyclic Echinoidea (in which the anal area lies some-
where in the posterior interradius) the longitudinal axis and the plane of symmetry
can easily be made out even in a superficial examination, they can also be dis-
covered by careful observation, even in regular endocyclic Echinoidea, which are
apparently strictly radiate. When describing the apical system, the constant
relation of the outer apertures of the pores of the stone canal to the right anterior
basal plate, was pointed out. These relations never quite disappear, and where
the apical system is retained, they define with certainty the longitudinal axis
and the plane of symmetry.

Further, even where the apical system has not been retained, it is always
possible, as has been proved by a very careful investigation of the Echiiioid test,
to determine the longitudinal axis and the plane of symmetry by the definite and
constant arrangement of the plates of the test, both in regular and irregular endo-
cyclic and exocyclic Echinoids. This constant relation of the plates to one another
is expressed in Loven's law.

Let the test of any Spatangoid be laid with the dorsal (apical) side on a perpen-
dicular surface, in such a way that the mouth is directed upward, and the posterior
unpaired interradius (between the bivium) downward. Let the five ambulacra be
then marked with the figures I, II, III, IV, V (Fig. 302), starting from the left
lower ambulacrum (the right posterior of the animal) and proceeding in the direc-
tion in which the hands of a watch travel. Two plates of each ambulacrum, the
so-called marginal peristome plates, take part in forming the boundary of the
peristome. The first marginal plate which is met with in each ambulacrum, when

342

COMPARATIVE ANATOMY

CHAP.

moving from left to right may be marked a, the second b, and these letters may
further indicate the rows to which these plates belong. In this way we can
name each of the ten ambulacral plates bordering the peristome. Examining
these ten plates carefully, we see that those indicated by la, Ha, lllb, IV«, \b

n

IS"

FIG. 302.— Kleinia luzonica (Gray). Apical system, spread out (after Loven). fa, Fascioles.
Further explanations in the adjoining text.

are larger and possess two pores each, while the smaller plates I//, lib, Ilia, l\/>
and V<f have only one pore each. Only the ambulacra I and V, i.e. the two
posterior ambulacra, are thus bilaterally symmetrical, while the two (paired)
anterior ambulacra II and IV, and the two rows of plates of the anterior unpaired
ambulacrum III, are asymmetrical. This law holds for all Echinoidea (not only
for adults but for their young stages also) ; the plates la, Ha, lllb, IV«, V6 are

vin ECHINODERMATA— MORPHOLOGY OF SKELETON 343

Fio. 303.— Toxopneustes drcebachiensis juv., 4 mm. in diam. The whole system of plates
spread out in one plane (after Loven). B, Peristome plates. D, teeth.

marked by common characters, and are distinguished from the plates U, lib, Ilia,
l\b, Va, which also resemble one another. These different characters are, it is true.
often not very evident.

344

COMPAEA TI VE ANA TOM Y

CHAP.

As a further example, let us take the test of a young Toxopneustes drcebachiensis,
4 mm. in diameter (Fig. 303). If we examine it we shall find that, of the ten
ambulacral plates bordering the peristome, five, belonging to different ambulacra,
are of greater size (consisting each of three primary plates), and show three double
pores, while the five others are smaller (consisting of but two primary plates) and are
perforated by only two double pores. We can place the test in only one position, viz.
that given in the figure, in which the formula la, II«, lllb, I Va, V&, and Ib, lib, Ilia,
IVb, Va holds good. In this we see that a median plane, corresponding with that of
the irregular Echinoidea, can be established also for the regular Echinoidea. The
accuracy of this law can be proved by investigating the position of the madreporite.
In the above'case this actually lies in the right anterior basal plate between the radii
II and III.

Loven's law also applies to other plates besides the ambulacral marginal plates of
the peristome.

It may be remarked in passing here that the system of marking above described
can be used for naming all the plates of the Echinoid test ; in this way we have
the ambulacra I-V, the ambulacral rows of plates la, Ib, lire, lib, Ilia, lllb, IV«,
l\b, Va, and Vfc, and in the apical system the radials I-V. If we mark the inter-
radii (interambulacra) 1 -5, starting from the one lying to the left of ambulacrum I,
and proceeding in the direction of the hands of a watch (viewing the test orally),
we get the interambulacral rows of plates la, Ib, 2a, 2b, 3a, 36, 4a, 4&, 5a, 5b, and
the basals 1-5. The madreporite lies in basal 2. The consecutive plates, counting

along each row of ambu-
lacral and interambulacral
plates, start; from the edge
of the oral disc.

The arrangement of
plates revealed by Loven's
law, taken together with
the special position of the
madreporite, and with the
excentric position of the
anus in the anal area of
the regular Echinoids, show
us that, strictly speaking,
no Echinoid is either radi-
ally, or bilaterally, sym-
metrical.

(</) The Relation of the
Ambulacral and In-
terambulacral Plates
to the Peristome.

Three cases must be
distinguished.

1. The plates, both of
the ambulacra and of the interambulacra, are continued in a modified form over the edge
of the peristome, and on the peristome itself, towards the mouth (Cidaroida, Fig. 304).
2. Only the ambulacral plates are continued on to the oral integument (Diculc-
matoida), forming either several concentric rings of plates (Streptosomata, Echino-
thuridif), or as five pairs of plates lying isolated in the integument, the so-called
buccal plates (Stercosomata).

FIG. 304. — Peristome and. neighbouring parts of the test
of Cidaris hystrix, Lamk. (atu-r Loven).

viii EGHINODERMATA— MORPHOLOGY OF SKELETON 345

3. Neither the ambulacra! nor the interambulacral plates are continued on to the
peristome (ffolcctypoida, Clypcastroida, Spcdanyoida).

Among the Palccechinoidea also there are forms in which the perisomatic plates
reach as far as the mouth ; in Lepidocentrus, indeed, they do this in such a way as
to make it impossible to distinguish the coronal from the peristomal plates.

Apart from the peristome plates just mentioned, the oral area is beset all over
with small irregularly arranged calcareous bodies.

With regard to the number of coronal plates which border the peristome (mar-
ginal plates of the peristome), it is to be noted that in regular Echinoidea (Cidaroida,
Diadematoida), and even in most Holedypoida, ten pairs occur, five ambulacral and
five interambulacral. There are, however, certain Holectypoida in which, in one or
several interradii, only a single marginal plate occurs. In the Clypcastroida (Fig.
306) and Sjmtangoida (Fig. 302) the peristome is, as a rule, bordered by five pairs of
ambulacral and five single interambulacral marginal plates. Exceptions to this rule
are found in the Spatangoid division, the Cassiduloidea, where, for example, among
the Echinoneidce, EcJiinoneus and Amblypygus have two marginal plates in their
second and fourth interradii and only one in the others.

(e) Manner in which the Skeletal Plates are Connected.

In most Euechinoidea the plates of the skeleton, at least those of the corona, are
firmly and immovably connected together by means of sutures, and thus form a
rigid test. This is not the case in very many Palceechinoidea, and among the

FIG. 305.— Oral area of Cidaris papillata, Leske, from within (after Loven).
apo, Perignathous apophyses.

Euechinoidea in the Diadematoid Echinotkuridcc ; also, as far as the skeleton of the
peristome is concerned, in the Cidaroida (Fig. 305). The edges of the plates here
overlap, i.e. they are imbricated. In the Echinothuridcc the plates are divided from
one another by strips of imcalcified connective tissue, which, to some extent,
allow the test to change its shape. The imbrication of the ambulacral plates is in
a direction opposite to that of the interambulacral. Viewing the test from without,

346

COMPARA TIVE ANA TOMY

CHAP.

the imbrication of the ambulacra is adoral, i.e. the oral edge of each plate overlaps
the apical edge of the next in order below it, whereas, in the interambnlacra, the
imbrication is apical. Lateral imbrication also occasionally occurs.
Slight imbrication is also found in certain Spatangoida.

(/) Special Modifications of the Ambulacra.

In all Echinoidea, in which the mouth remains at the centre of the oral surface,
the five ambulacra are alike in length, breadth, and in the arrangement of their

!\r

• • f '

Fio. 30(3. — System of plates of a Clypeastroid (Encope Valenciennes!, Agass.), sp

(after Loven).

out

pores, prominences, etc. They only vary in length when the apical system, towards
which, radiating from the peristome across the ambitus, they converge, is shifted
from the centre of the apical hemisphere to a somewhat anterior (less frequently
posterior) position. If the test of such an Echinoid, in which the ambulacra

vni EGHINODERMATA— MORPHOLOGY OF SKELETON 347

are of unequal length owing to the shifting of the apical system, be viewed from
the oral side, the ambulacra still form a regular, or almost regular, five -rayed
star round the central oral aperture or peristome. Where, however, as in the
Spatangoida, the peristome with the mouth has moved from the centre of the oral
surface (on which the Echinoids creep), and is shifted more or less anteriorly, and
finally, where in the Pourtalesia it comes to lie quite on the anterior ambitus, the parts
taken by the five ambulacra in the formation of the oral surface are necessarily very
different. The unpaired anterior ambulacrum (III) and the two anterior and
lateral ambulacra (II and IV) shorten and form an ever smaller portion of the
whole ambulacral area of the oral (ventral) surface, in proportion as the peristome
with the mouth shifts forward. They form together the trivium. Conversely, the
two posterior radii at the same time lengthen and form an increasingly large
portion of the ambulacral area of the ventral surface. They form the bivium. The
length of the ambulacra of the trivium and the bivium in the apical direction is of
course determined by the position of the apical system. If this system shifts for-
ward, the trivium is shortened apically ; if backward, the ambulacra of the trivium
(especially the anterior unpaired ambulacrum) are lengthened, while those of the
bivium are shortened. This grouping of the ambulacra into an anterior trivium
and a posterior bivium is especially clear on the apical surface of those Spatangoida
which have a diffused apical system, e.g. the Colly ritidcc and Pourtalesiidce (cf. pp.
324, 325). Since the apical ends of the ambulacra are always in contact with the
radial plates of the apical system, and since, further, in the diffused apical system
the two posterior radials I and V, which are separated from the anterior, are
shifted posteriorly, the apical ends of the two posterior ambulacra (the bivium) are
also necessarily separated from the three anterior ambulacra (the trivium) by a con-
siderable space (Fig. 284, p. 325).

In the Palceechinoidea, and among the Euedvinoidea in the Cidaroida, the Diaili'-
matoida, nearly all Holectypoida,, and many Spatangoida, the ambulacra throughout
their whole courses have a similar structure, and are similarly provided with pores.
In the Clypcastridce and many Spatangoida, however, the ambulacra are modified on
the apical side in a characteristic manner ; they arepetaloid, each ambulacrum forming
a petalodium (Figs. 233, 234, p. 292 ; 236, p. 294, and 306). Such a petalodium
arises by the divergence of the two rows of large double pores of each ambulacrum
from one another immediately on leaving the apex, and their reapproximation and
junction before they reach the ambitus. The two rows of pores of each petalodium
make a figure like a lancet-shaped leaf, and the five petaloids together form round
the apex a graceful rosette of leaves, which recalls the petals of a flower. Ou the
remaining plates of the ambulacra, />. those not forming the petalodium, the pores
are single and small ; they are, further, few in number and scattered. Between
the regular ambulacra and those which have apical petaloids there are many transi-
tion forms, occurring often within one and the same family. One of these transi-
tions is specially frequent ; the two rows of pores of a petalodium do not unite
at their oral ends but remain open. The ambulacra are then called sub-petaloid.
Such petaloids are often very long.

The petaloids often sink in (Fig. 236, p. 294), and then, not infrequently, serve
as brood cavities, or marsupia, for containing the young.

Just as the ambulacra occasionally form petaloid rosettes round the apical system,
so, in the family of the Cassidulidce (sub-order Cassiduloida of the order Spatan-
f/oida), can they form rosettes of so-called phyllodes round the peristome (Fig. 307).
The five phyllodes, in which the well-developed double pores lie thickly crowded
together, sink in, while the five interradial marginal plates of the peristome between
them are contrariwise bulged out. The five interradial cushions form, together with
the five radial phyllodes, what is called a floscelle.

348 COMPARATIVE ANATOMY CHAP.

The anterior unpaired ambulacrum in many exocyclic Echinoidca differs greatly,
both in shape anil in the number, arrangement, and form of its pores, from the other

FIG. 307.— Oral psrisome of Cassidulus pacificus, Ag., with the live phyllodes (after Loven).

four. This variation in the anterior ambulacrum is found almost exclusively in the
order Spatniiyoida, especially in the Cassiduloid family Plesiospatangidce and in the
sub-order Spatanrjoidea (here especially, and, to a very marked degree, in the family
of the SjKitu ii'ii'lK:}.

((/) Special Modifications of the Interradii.

We can here only point out certain conditions occurring in the order Spatan-
i in Ida.

In the sub-order Spatangoidea an extraordinary asymmetry of the two
posterior interradii 1 and 4 prevails (cf. Fig. 302, p. 342). The right posterior in-
terradius 1 is always so modified near the peristome that two plates fuse, thus con-
trasting with the left posterior iuterradius, which remains only slightly if at all
modified. This fusion takes place either between the second and third plates of the
row la, or the two second plates of rows 1« and 1&, or the second and third plates
of row & and the second plate of row a. In the last case, the second plates of the
two rows of iuterradius 4 are also fused.

Since, in the Spatanyoida, the peristome, with the mouth, is shifted forward on
the oral surface, the posterior unpaired iuterradius occupies a considerable portion of
the ventral surface (and this is also the case in the Cassiduloidca with mouth shifted
forward). It is often somewhat bulged out, and the region occupied by it on the
oral side is known as the plastron. It takes part in the limitation of the peristome
by means of a single crescent-shaped plate, which is known as the labrum in those
forms which have a projecting under-lip to the transverse peristome (cf. Fig. 302,
p. 342). In many Spata/nyoida the labrum is followed posteriorly by two large sym-
metrically arranged plates (sternum), which again are followed by two smaller but

vin EGHINODERMATA— MORPHOLOGY OF SKELETON 349

still not insignificant plates (episternum). The test is then amphisternal. In
other forms, however, the arrangement of the plates on the plastron (apart from the
labrum) approaches the usual arrangement, i.e. the plates of the two rows alternate
more or less distinctly, so that the median suture which divides them forms a zig-
zag line. This arrangement, as compared with that first described, is older and more
primitive. The test is then called meridosternal.

In most Clffljeastridcc the interamlmlacra are interrupted, i.e. they do not run
continuously from the apical system to the peristome, but, near the latter, are
crowded out by the broad plates of the ambulacra which touch one another inter-
radially, so that the five interradial marginal plates of the peristome are completely
isolated from the remaining portions of the interambulacra (Fig. 306). Not infre-
quently, the paired interambulacra are interrupted and the unpaired posterior inter-
ambulacrum is uninterrupted.

(7t) Form of the Peristome.

In most Echinoidea, i.e. in those in which the peristome retains its central posi-
tion, its shape is pentagonal, or decagonal, or round, less frequently oval or oblique,
or quite irregular, often with branchial incisions. But where the peristome is
shifted anteriorly, as in the sub-order Spatangoidea, the peristome is transverse and
crescent-like, with depressed anterior upper-lip and raised posterior under-lip. The
peristome, however, is always central in the embryo, and is originally pentagonal.

(i) Ornamentation.

The outer surface of the plates of the Echinoid test are beset — in many different
ways, which are of importance in classification — with numerous larger or smaller pro-
minences, granules, etc., on which spines and peclicellariae are planted.

In the sub-order Spatangoidea, narrow, finely gramilated streaks or bands run, in
definite arrangement, along the surface of the test, and carry small rudimentary
spines or pedicellarise. These are called fascicles or Semites (Fig. 302, p. 342). The
following systematically important forms of fascioles are to be distinguished :—

1. The peripetaloid fascicle encircles the apical rosette of petaloids.

2. The lateral or marginal fascicle runs round the shell near the ambitus.

3. The lateral subanal fascicle branches off from the peripetaloid fascicle and
runs below the anus.

4. The subanal fascicle forms a ring below the anus (between the latter and
the peristome). They may give off anal branches which run up on each side of the
anus, and occasionally unite above it to form an anal fasciole.

5. The internal fascioles run around the apex and the anterior ambulacrum.
The tentacles and plates in those regions which are encircled by the internal

and subanal fascioles are modified.

One very varied form of ornamentation of the Echinoid test, which arises early
during postlarval development, is due to the deposit of calcareous substance on the
plates, and is known as epistroma.

(k) Marginal Incisions or Perforations.

These are often to be found in the flat disc-shaped test of the Scidcllida', in
some or all of the ambulacra, and not infrequently also in the posterior inter-
ambulacrum. The edge of the shell is at first entire, but during growth marginal
indentations and incisions make their appearance, and these may close to form per-
forations (lunulse). (Figs. 234, 235, pp. 292, 293, and 306, p. 346.)

350

COMPARATIVE ANATOMY

CHAP,

(1) The Perignathic Apophysial Girdle (Figs. 308, and 348, p. 402).

In all E?liinoid?a in which the mouth is armed with five teeth, moved by a com-
plicated masticatory apparatus, i.e. in all Echinoidea except the Spatangoida and a
few Holcctypoida, processes, directed apically inwards, are found at the peristomal
edge of the test ; these serve for the attachment of the muscles and bands of the
masticatory apparatus. They either consist solely of the ambulacra! or iuter-
ambulacral marginal plates of the peristome bent round inwards, or else a few of the
plates next in order also take part in their formation.

These processes may be divided into those which rise on the ambulacral marginal
plates, and those which rise on the interambulacral marginal plates. The former

may be called the ambulacral apophyses,
the latter the interambulacral apophyses.

The apophysial circle is closed or inter-
rupted. In the former case, which is best
illustrated by the Diitdcmatoida (Fig. 308, A),
an apophysis rises on the peristomal margin
of each ambulacral area on each side of the
ambulacral suture. The two apophyses of
one and the same ambulacrum usually unite
at their free ends, which project into the
body, in such a way as together to form a
kind of arch ; this is called an auricle, and
affords passage for some of the important
organs (for the trunks of the radial ambulacral
vessels, of the nerves, etc.). There are thus,
Fie, 308. -The perignathic apophyses in all) ten ambulacral apophyses, which may

of a radius and of the two neighbouring ., • ,- c ' • , mi

unite in pairs to form hve auricles. The
mterradn of various Echinoidea. A,

Diadematoid. The apophyses of the interambulacral apophyses project less far into
ambulacral plates (nm) form true auricula; the interior of the body. The two apophyses
(.<»/•)• B, Cidaroid. Apophyses are formed, of one au(j the same interambiilacriim together

form a ridge which runs along the edge of

not by the ambulacral but by the iuter-
ambulacral plates, forming what are callnl
false auricles, in C (also a Cidaroid) the.se the penstome, and connects two neighbour-
interambulacral plates have fused. ing auricles ; these ridges are generally fused

with one another and with the auricles.

Such a closed apophysial ring, which rises on the edge of the peristome and pro-
jects into the body, may be compared to a circular wall with high arched gateways
at five radially arranged points. The five arched gateways would represent the
auricles, i.e. the five pairs of ambulacral apophyses, and the circular wall would
be formed of the five pairs of interambulacral apophyses.

In the Cii/ttmii/i/. (Fig. 308, B and C) the apophysial ring is interrupted. The
.•iinliiilacral apophyses are wanting, but the interambulacral apophyses are all the
inure strongly developed, and form ear-shaped processes. The two apophyses of an
mterainbulacrum are connected by a suture at their bases, but diverge at their tips.
When the two interambulacral apophyses standing at the sides of an ambulacrum
approximate above it (the ambulacrum), but without fusing, a false auricle may
be formed.

The ambulacral apophyses are also wanting in a few Holcctypoida; where they
are present, they do not unite in pairs to form auricles.

In all Clypeastroida, the apophysial ring is interrupted, and consists either of
ambulacral or of interambulacral apophyses.

vin ECHINODERMATA— MORPHOLOGY OF SKELETON 351

III. Asteroidea.

Here also the perisomatic portion forms by far the greater part
of the whole skeleton. Only in a few forms does the apical system
constitute a distinctly appreciable element in the skeleton. Further,
the oral system also, even if we include, besides the orals (odonto-
phores, proximal plates of the iuterbrachial system), the terminals, as
radials belonging to the oral system, forms but a very small fraction
of the whole skeleton.

The skeleton of the Asteroidea is distinguished from that of most
Echiuoidea by its mobility. It is not a rigid capsule, but its principal
plates are articulated one with another, and are movable one upon
another by means of muscles. The arms can bend upwards and
downwards, and also occasionally, to a certain degree, laterally (in the
horizontal plane). The ambulacral furrows may be deep, or shallow.
The disc is sometimes shortened in the direction of the principal axis,
i.e. flattened.

In the perisomatic skeleton of the Asteroidea three principal parts
may be distinguished : (1) the ambulacral, (2) the interambulacral,
and (3) the accessory.

(a) The Ambulaeral Skeleton.

From the free end, or tip, of each arm or ray a large median groove
runs on the oral side to the centre of the disc, and here runs into the

Fie. 309. —Transverse section through the brachial skeleton of Astropecten aurantia-
CUS (Gray) ; original. For lettering see p. 317. «t, Supports of the ambulacral plates or supra-
umbulacral plates ; ad, adambulacral plates ; jj,Vpaxilla'. ; 1, position of the radial canal, etc. ;
•2, ampulla ; 3, ambulacral feet.

mouth. In the base of this ambulaeral furrow rise the ambulacral,
or tube-feet in two or four longitudinal rows (Figs. 239, 243, pp. 296,
298, and 343, p. 396). The plates of the ambulacral skeleton, which

352

COMPARATIVE ANATOMY

CHAP.

may be compared with vertebrae, and are the principal pieces of the
skeleton, form a long roof over the ambulacral furrow, which opens
downwards. In a transverse section through the arm of an Asteroid
(Fig. 309) Ave see that the roof of the furrow invariably consists of
four skeletal pieces. Two of these pieces— the ambulaeral ossicles
(am) — form the greater part of the roof. They lie symme'trically to
the median plane of the arm, and articulate with one another along
the ridge of the roof. The two other skeletal pieces — the adambu-
laeral ossicles (ml) — meet the diverging edges of the ambulacral
ossicles, and so lie at the edge of the furrow, or, in other words, at
the lower lateral edges of its skeletal roof.

The general form of the ambulacral ossicles is that of transversely elongated
clasps. They are arranged in two longitudinal rows in close proximity to one
another, and in this way form the roof, which arches over the groove along the whole
of its course, from the tip of the arm to the mouth.

In the Euasteroidea (to which sub-class all recent forms belong) the ambulacral
ossicles of the two rows are arranged in pairs, each ossicle on one side of the roof

FIG. 310.— Scheme of the oral skeleton of the Asteroidea, from the inner side (after Ludwig).
or, Oral plate (odontophore) ; MI, first lower transverse muscle of the [ambulacra! furrow ; Mf, the
interradial muscle ; I-VI, first to sixth ambulacral ossicles ; 1-6, first to sixth adambiilacral
ossicles ; a, b, c, d, e,f, apertures for the ampull<e of the tube-feet.

corresponding with one on the other side. In the Palccastcroidca, on the contrary,
the ossicles alternate, at least in the middle part of the arm.

The (smaller) adambulacral plates usually alternate regularly with the ambulacral
plates.

We must here emphasise the important fact that the ambulacral ossicles of the
Asteroidea lie much deeper than the skeletal pieces of the same name in the
Echinoidea. In the latter class they are quite superficial, the radial trunks of the
water vascular system, as well as the radial nerves and the spaces of the schizocoel,
are to be found on their inner side ; whereas, in the Asteroidea, these organs lie on
the outer side under the ambulacral roof. Of the whole ambulacral vascular system
only the ampulla? lie on the inner sides of the ambulacral ossicles, i.e. that turned
towards the general body cavity.

Between every two consecutive ambulacral ossicles there is one (and never more
than one) aperture for the passage of a tube -foot. The number of ambulacral
ossicles in a row thus always corresponds quite accurately with the number of the
tube-feet on the same side of the ambulacral furrow.

Each aperture for the passage of a tube-foot normally lies in the corner between

viii ECHINODERMATA— MORPHOLOGY OF SKELETON 353

two ambulacral ossicles and an adambulacral ossicle (cf. Fig. 310). In those
Asteroids which have four longitudinal rows of tube-feet, however, these apertures,
at some distance from the mouth, alternate regularly in such a way that the
laterally placed aperture of one interstitium is followed by a more median aperture
in the next interstitium, the next again being lateral, and so on. The connecting
line between the apertures of one and the same side of an ambulacrum in this case
forms a zigzag, the angle of which is the more pointed the narrower the ambulacral
ossicle. The consequence of this is, that the tube-feet which stand in the corners of
the zigzag line appear arranged in two rows, that is, in the whole ambulacrum, in
four rows.

The oral aperture, which always lies in the centre of the ventral
surface of the disc, and into which the ambulacral furrows of the arms
converge, is surrounded by a circle of firmly connected calcareous
pieces, the external edges of which are in immediate contact with the
ambulacral and adambulacral ossicles. This circle forms the oral
skeleton of the Asteroidea. It is extremely probable that its separate
pieces (which in the five-rayed forms number thirty, and in forms with
a greater number of rays are six times as numerous as the rays) are
merely the transformed and more firmly connected proximal ossicles
of the ambulacral and adambulacral rows. In this case, in each ray
or arm, the first two pairs of ambulacral and the first pair of adambu-
lacral plates of these rows (in Ct&nodiscus, the first three ambulacral
and the first two adambulacral pairs of plates) would take part in the
formation of the oral skeleton. The oral skeleton is ambulaeral (in
many Cryptozonia} or adambulacral (in the Phanerozonia and some
Cr//pto20)iia), according as the ambulacral or the adambulacral portions
of the circle project the further into the oral cavity.

(//) The Interambulaeral Skeleton.

This comprises the ambitus, i.e. the whole surface of the body
between the oral (or ventral) and the apical (or dorsal) regions, on both
of which, however, interambulacral plates may be found. The inter-
ambulacral skeleton thus forms the lateral walls of the arms. The
pieces constituting it are called marginal plates, and are arranged in
each lateral wall in two rows, one above the other. The upper row
consists of the supramarginal (Fig. 309 sm) and the lower of the infra-
marginal (in/) plates. It only rarely happens (e.g. in Lnidin) that
the marginal plates agree in number and length with the ambulacral
ossicles. The marginal plates, which in the order Phanerozonia are
large and well developed, become reduced in that of the Criiptozonia,
being difficult to distinguish externally. They may be altogether
wanting, or else represented merely by microscopically small rudi-
ments. The row of inframarginal plates may be separated from that
of the adambulacral ossicles by a row of small intermediate plates. In
the same Avay a row of small intermediate plates may be intercalated
between the two rows of marginal plates.

VOL. II 2 A

354 COMPARATIVE ANATOMY CBAP.

(f) The Accessory Skeletal System.

In this system may be included all those plates or ossicles which occur in those parts
of the body not covered by the ambulacral and marginal systems. This accessory system
is very variously developed, and a comparative study of it cannot here be under-
taken. The plates differ greatly in size, form and ornamentation, and arrangement,
sometimes being scattered or lying loosely near one another, or else closely approxi-
mated, sometimes imbricating or reticulating by means of anastomoses of skeletal
pieces.

Not infrequently either the whole, or parts, of the accessory skeleton are reduced.
It is often covered by a considerable layer of integument, and is difficult to dis-
tinguish externally. Its plates may diminish greatly in size, even becoming micro-
scopically small, but they are rarely altogether wanting.

Three sub-divisions of the accessory skeleton may be distinguished : —

1. The dorsal, abactinal, or apical accessory system, when present, consists
of skeletal plates developed in the dorsal integument of the disc and in the arms.
We have seen above that in the Asteroidea the apical system only rarely takes any
recognisable part in the formation of the dorsal skeleton. There are, nevertheless,
forms (e.y. Cnemidaster) in which the large and distinct plates of the apical system
form almost the whole of the dorsal protection of the disc.

2. The ambital accessory system consists of the iutermarginal plates already
mentioned as occasionally being intercalated between the supra- and the infra-
marginal rows of plates.

3. The ventral, actinal, or oral system in the same way consists of the already
mentioned intermediate plates which may occur between the inframarginal and the
adambulacral plates. It is most developed in those forms in which the disc
increases in size at the expense of the arms. i.e. in forms whose outline is more or less
pentagonal. The ventral accessory plates then fill up the larger or smaller triangular
regions between the ambulacral furrows on the lower side of the disc.

Finally, two other skeletal systems which occasionally occur in the Asteroid
body must be mentioned.

In a certain number of Star-fish each ambulacral ossicle is connected by a skeletal
plate, or more rarely by a row of two to three firmly united plates, through the
body cavity, with a marginal plate of its own side, or else with a laterally placed
accessory plate. These simple or compound skeletal pieces, which are limited to
the arms, and which here correspond in number with the ambulacral ossicles, are
called supports to the ambulacral ossicles or supraambulacral plates (Fig. 309 so).

The other skeletal system, which occurs especially in Asteroidea with large discs,
but is altogether wanting in many forms, is called the interbrachial system. It
continues the divisions between the arms, either completely or incompletely, into
the interior of the disc, and consists either of interbrachial walls, running from the
oral to the actinal skeleton, or of interbrachial chains of skeletal plates descending
vertically to the oral skeleton. In each interradius a proximal plate of this inter-
brachial skeleton, however, always enters into closer relations with the oral skeleton.
These plates are the orals, already mentioned in the section on the oral system.

At the free end of each arm in every Asteroid there is to be
found a single median skeletal plate, which is sometimes of consider-
able size and distinctly visible, sometimes small and inconspicuous ;
it carries on its lower side a visual organ. These plates are called
ocular plates or terminals. According to recent investigations they
develop very early (apparently first of all the plates) over the left

viii ECHINODERMATA— MORPHOLOGY OF SKELETON 355

coelomic vesicle. They must thus belong to the oral system, and
perhaps, in this system, correspond with the radials in the apical
system.

In the development of the Asteroidea the formative centre of each
newly appearing plate in a radius of the perisomatic system is always
immediately proximal to the ocular plate of the arm. At these
points new plates continually appear between those last formed and the
ocular plates, which thus always remain at the free tips of the arms.

(d) Comparison of the Perisomatic Skeleton of the Asteroidea with that of

the Echinoidea.

The ocular plates (terminals) of the Asteroidea bear to the newly appearing
plates of the perisomatic skeleton relations altogether similar to those which the
radials (also "oculars ") of the apical system in the Echinoidea bear specially to the
ambulacral plates. Since it has not been proved that the radial plates of the
Echinoidea arise over the right ccelomic vesicle, it is possible that they, although
lying high up at the apex, belong genetically to the oral system, and correspond
with the terminals of the Asteroidea. The radials should then not be represented in
the apical system of the Echinoidea.

In a comparison of the skeletons of the Echinoidea and the Asteroidea we should
then have to suppose that in the former the ambulacra have been lengthened round
over the ambitus to the apex ; and that, further, the body took on the form
of a pentagonal pyramid, by the abbreviation of the arms and the elongation of
the principal axis of the body ; and that, therefore, the whole region occupied by
the accessory skeleton of the Asteroid has disappeared. The marginal plates of the
Asteroid would then correspond with the interambulacral plates of the Echinoid,
and the adambulacral ossicles of the former with the ambulacral plates of the latter.
A comparison of the ambulacral plates of the Echinoid with the plates of the same
name in the Asteroid is rendered difficult by the difference of position of the two,
the former being superficial, epiambulacral, and epineural, and the latter deeper,
subambnlacral, and subneural. The ambulacral ossicles of the Asteroid would thus
not be represented in the skeleton of the Echinoid.

IV. Ophiuroidea.

(a) Skeleton of the Arms.

The brachial skeleton of the Ophiuroidea consists typically of six
longitudinal rows of plates, a dorsal row (dorsal shields), a ventral
row (ventral shields), two lateral rows (lateral shields), and a double
row of internal ossicles lying in the axis of the arm. This system is
jointed, or segmented, in quite a regular manner — one dorsal, one
ventral, one axial piece and two lateral pieces together composing
a skeletal segment (Fig. 311).

The external pieces together form, in each arm, a jointed tube,
which determines the shape of the arm. Most of the lateral shields
carry spines ; on each shield there are usually four of these, one
above the other, so that each longitudinal row of shields is aj-med
with four longitudinal rows of spines. The tube -feet emerge at

356

COMPARATIVE AXAToMY

CHAP.

regular segmental intervals through apertures which lie on each side
between the ventral shields and the lateral shields belonging to them

etc

a.m>

— te

FIG. 311.— Transverse section through the arm of an Ophiurid (after Ludwig). Diagram.
.«, Lateral shields ; <?.<, dorsal shields; <:l, cavity of the arm (co-loin); uc, spines; nw, the ambu-
lacral plates (vertebra-) ; ./, loop of tentacle canal in the groove on the distal face of the ossicle (c/.
next fig., A 4) ; tie, tentacular canal of the radial vessel (m) of the water vascular system ; t< .
feeler (tentacle) ; n; radial pseudohaemal vessel ; rn, radial nerve strand ; te, ventral shield.

772

FIG. 312.— Vertebral ossicles (ambulacral plates) of Ophiarachna incrassata (after Ludwig),
to show the articulating prominences and depressions, etc. A, Three vertebral ossicles from the
side. B, Vertebral ossicles from the proximal (adoral), and C, from the distal (aboral) side. D, Three
vertebral ossicles from the ventral side, pr, Proximal ; di, distal ; m, radial trunks of the water
vascular system; rn, radial nerve trunk; n; radial pseudohsemal canal. 1, Point at which the
branch of the radial water vascular trunk running to the tube-foot passes out of the substance
of the vertebral ossicle at its distal side ; 2, point where this branch re-enters the ossicle ;
4, channel between these two points, which receives the loop of the branch belonging to the
tube-foot; 3, depression for the lower intervertebral muscle; 5, channel for the radial water
vascular trunk ; 6, depression for the tube-foot ; 7, channel for the branch of the nerve running to
the tube-foot ; 8, pseudohaemal vessel to the same ; 9, nerve branch to the same ; 10, brand) of the
water vascular system to the same, which at 12 passes into the substance of the ossicle, and at 13
out of the latter and into the tube-foot ; 11, point at which the nerve branch (14) running to the
upper intervertebral muscle, enters the vertebral ossicle.

(cf. Fig. 245, p. 300). At the edge of these apertures there are smaller
spines or scales.

The axial double plates are called vertebral ossicles, a very suit-

VIII

ECHINODERMATA— MORPHOLOGY OF SKELETON 357

KU

y.9

am, \

FIG. 313.— Ophiohelus umbella, Lym. A macerated

able name, since they play a part altogether similar to that of the
vertebrae of the axial skeleton in Vertebrate animals. In a large
majority of cases the two lateral portions of a vertebral ossicle are
fused in the median plane in such a way that no sutures are now to
be seen. These ossicles, however, arise ontogenetically as two, at first
entirely distinct, lateral pieces, which only fuse later. There are,
further, certain deep-sea Ophiuroidea (Opkiokelus, Fig. 313) in which
each vertebral ossicle consists, even in the adult, of two distinct slender
pieces, articulated one with the other.

The vertebral ossicles fill up the greater part of the skeletal tube
formed by the dorsal, ventral, and lateral shields. Between them and
the tube, in dried skeletons,
only small spaces are to be
found, which dorsally contain
continuations of the body
cavity of the disc, while ven-
trally they contain the radial
water vascular trunk, the
radial nerve cord, the epineural
canal, and the pseudohremal
vessel. The lateral branches
of the radial vessels of the
water vascular system, before

entering each tube-foot, pass joint from near the tip of an arm, from the dorsal side

, , i . -I ,1 (after Lyman). ds, Dorsal; ss, lateral shield; am,

through, 011 each Side, tne ambulacrai ossicles ; apa, hook spines.

substance of the vertebral

ossicle of the corresponding segment, nearer the distal than the
proximal end of the ossicle. The consecutive vertebral ossicles of the
arms articulate one with another, and are connected by means of four
intervertebral muscles. The contraction of the two upper inter-
vertebral muscles brings about the upward curving, and the contrac-
tion of the two lower, the downward curving, of the arms. The
horizontal (lateral) movement is brought about by the contraction of
the upper arid lower muscles of the same side. The vertical movement
of the arms is very slight in true Ophiuridse, whereas in the Eunjalidce
the arms can be completely rolled up orally (cf. Fig. 246, p. 301).

Small accessory plates may occur in addition to the dorsal shields. The super-
ficial brachial skeleton is much reduced in the Astrophytuicv (Euryalidae) and the
Ojih iomyxidcc, and the arms are, in these animals, covered by a soft integument, in
which only small skeletal pieces occur. In other forms the brachial skeleton is so
covered by an integument, often containing small embedded skeletal pieces, that it
is either partly or altogether invisible externally.

At the distal end of each arm in the Ophiuroidea there is, as in
the Asteroidni, an unpaired median terminal, which surrounds the
tip of the radial water vascular trunk (the terminal tentacle) in the
form of a short skeletal ring. Since, in the A*f> /'<>i</<'", the terminal

358 COMPARATIVE ANATOMY CHAP.

plate receives the terminal tentacle in a channel on its ventral side, it
is important to note that " the terminal plate in the Ophiuroidea also
originally forms a channel opening downwards, and only later closes
to form a ring."

The relation between the terminals and the developing brachial
skeleton is the same in the Ophiuroidea as in the Asteroidea. The
oldest skeletal segment is the one lying most proximally (orally), and
of the following segments the more distal are always the younger.
The plates which compose each newly appearing skeletal segment
always arise at the end of the arm, on the proximal side of the ter-
minal, which thus remains at the extreme tip of the arm.

When we consider the paired elements of the vertebral ossicles and the relative
positions of the skeletal plates and the water vascular system, we are able to estab-
lish the following homologies between the components of the brachial skeletons of
the Ophiuroidea and Asteroidea.

OPHIUROIDEA. ASTEROIDEA.

The two lateral halves of the vertebral Ambulacral ossicles.

ossicles.

Lateral shields. Adambulacral ossicles.

Ventral shields. Not represented.

(/>) The Oral Skeleton.

The most important and constant plates of the oral skeleton, in
the Ophiuroidea, as in the Asteroidea, are the specially modified
proximal plates of the brachial skeleton. The most satisfactory view
which has been propounded as to the morphological worth of the
oral skeleton is that it consists essentially of the ambulacral ossicles
(the halves of the vertebral ossicles), adambulacraJ ossicles (lateral
shields), and ventral shields of the first and second proximal skeletal
segments of the arms.

If we look at the oral region of any Ophiuroid from without, i.e.
from the free oral surface of the disc, or from within, i.e. after removal
of the apical cover of the disc and the viscera, we see the mouth in the
centre of the disc as a rosette-like, or star-shaped, aperture. The slits
arranged radially round the centre are called the bueeal fissures. Be-
tween them lie the triangular oral-angles (Figs. 245, p. 300, and 314).
Five pairs of large plates form the frame surrounding the mouth ;
these are the oral-angle plates (Fig. 314). At the interradial
angle of each of these, i.e. the angle which projects towards the
centre of the oral aperture, two neighbouring angle plates meet.
Each angle plate has, on the side facing a buccal fissure of the oral
aperture, two depressions for receiving the first tube-feet which have
shifted into the oral aperture, and are known as oral tube-feet, or oral
tentacles. There are often in addition, in the dorsal side (that facing
the body cavity) of the circle of oral-angle plates, two circular furrows

vin EOHINODERM ATA— MORPHOLOGY OF SKELETON 359

or channels, one of winch receives the nerve ring and the other the
water vascular ring.

In Astropliyton part of the water vascular ring is entirely enclosed
within the oral-angle plates.

Closer examination reveals the fact that each oral-angle plate
consists of two fused plates, a proximal and a distal. The former,

ta

Fin. 314.— Oral skeleton of the Ophiopya longispinus, Lym., from within; above, an inter-
railial region of the cover of the disc, rs, Radial shields; <u«, vertebral ossicle; ami, peristonial
plates ; pteb, depressions for the oral tentacles ; am-i+adi, oral-angle plates ; fb, bnrsal apertures ;
tn, torus angularis ; D, teeth ; ibi; interbrachial region ; sgc, bursal scale ; yp, genital plate
(after Lyman).

directed towards the centre of the mouth, fuses with the corre-
sponding piece of its associated oral-angle plate, the two forming the
oral angle. The distal plate at its distal end is in contact with a
corresponding plate on the opposite side of the buccal fissure. The
former of these constituents of each oral-angle plate is regarded as an
adambulacral plate of the first brachial segment, taking part in the
formation of the oral skeleton, while the distal plate is regarded as an

360 COMPARATIVE ANATOMY CHAP.

ambulacral ossicle of the second skeletal segment. It is the latter
which is provided with furrows for the nerve and the water vascular
rings, and with depressions for the oral feet (two on each piece).
The distal portions of each pair of oral-angle plates, which together
border a buccal fissure, would thus correspond with the lateral halves
of a brachial vertebral ossicle, not fused together.

In viewing the under (oral) side of the disc of an Ophiuroid (Fig.
245, p. 300) we can easily recognise the interradially placed bueeal
shields (scuta buccalia), which are usually large, and have already
been mentioned as belonging to the oral system. At the sides of
each buccal shield, between it and the neighbouring oral-angle plates,
lie two skeletal plates, which are known as lateral bueeal shields
(scutella adoratia). That these last-mentioned plates belong to the
same row as the adambulacral plates (lateral shields) of the arms can
generally easily be seen. They are the adambulaeral plates of the
second segment taking part in the formation of the oral skeleton.
The third pair of adambulacral plates is thus the first pair of lateral
shields in the arm.

Again vieAving the oral skeleton from the dorsal or apical side (Fig.
314), we see that above the ten oral-angle plates lie ten other plates,
which usually to a greater or lesser extent roof over the Avater vascular,
and the nerve furrows. These, the peristomal plates, thus lie upon
the inner sides of the oral-angle plates, i.e. the sides facing the body
cavity. The peristomal plates belonging to two neighbouring radii
meet interradially, and may fuse together to form single plates. The
two peristomal plates belonging to one and the same radius may, in
the same way, touch one another (in Avhich case the ten plates
together form a closed circle), or their radial ends may remain more
or less apart. Accessory peristomal plates sometimes occur ; in other
cases these are altogether Avanting. The peristomal plates are con-
sidered to represent the ambulaeral ossicles (halves of the verte-
bral ossicles) of the first segment of the oral skeleton, a view
which does not appear to be certainly established, chiefly because they
are in no Avay connected with the tube-feet. The two pairs of tube-
feet of each radius of the oral skeleton, as has been pointed out, belong-
to its tAvo oral-angle plates.

At the distal end of each of the oral slits radially, vieAved from
without, there is, in many, indeed, in most Ophiuroidea, a plate Avhich
also takes part in the limitation of the oral cavity (Fig. 245, p. 300).
This plate can at once be recognised as the most proximal plate in
the row of ventral shields. It is the ventral shield of the second
segment of the oral skeleton. The lateral shields belonging to them
are the lateral bueeal shields.

In a row Avith, but dorsally to, this ventral shield, Avithin the buccal
fissure, there is a second plate (Avhich, hoAvever, may occasionally be
wanting) ; this varies greatly in size and form, and is to be regarded
as the ventral shield of the first segment of the oral skeleton.

VIII

EGHINODERMATA— MORPHOLOGY OF SKELETON

361

The following table embodies this view of the oral skeleton, viz.
that it consists of modified pieces of the first two skeletal segments of
the radii (arms).

Skeletal Segment of the arm. '

1st (Proximal) Segment of tin-
oral skeleton.

The two halves of the The distal portions of The two peristoinal

vertebral ossicle (ambu- the two oral-angle plates plates of a radius (Fig.

lacral plates) (Figs. 311 ! belonging to a radius 314
and 314 am). ' (Fig. 314 «»i2 + «e?i).

The two lateral shields The two lateral buccal The proximal portions

of the oral - angle plates
belonging to a radius -(Fig.
314

(adambulacral plates) shields of a radius (Fig.
(Figs. 311 ss, and 245, 4, , 245, 5, p. 300).
p. 300).

The ventral shields Externally visible ven- Inner ventral shield of
(Figs. 245, 1, p. 300, and tral shield of each radius the oral skeleton.
311 bs). of the oral skeleton (Fig.

, 245, S, p. 300).

Accessory Parts of the Oral Skeleton.

At each oral angle (at the point where two neighbouring oral-angle plates meet
interradially), on the side facing the buccal cavity, there lies a vertical row of small
skeletal pieces, which may fuse together to form the torus angularis (Fig. 386, tit,
p. 486). This carries the teeth (D) which project into the buccal cavity. The oral-
angle plates themselves carry, at the edges which are visible externally, i.e. from
the ventral side, small spine-like skeletal pieces. Of these those which project into
the buccal fissures are called oral papillse ; while those which rise at the tips of the
oral angles, and are turned to the axis of the buccal cavity, are called dental
papilla. Consequently, in each oral angle, the teeth above mentioned lie dorsally
to the dental papillre.

Accessory Skeletal Plates of the Disc.

Lower side. — The pieces already described as appearing at the surface on the lower
side of the disc, and which belong to the oral system (oral shields) or to the oral
.skeleton (oral-angle plates, lateral buccal shields, ventral shields), hardly ever form
the whole ventral carapace of the disc. On the contrary, between the roots of the
arms (interbrachially or interradially) these plates leave free spaces (Figs. 245, p. 300,
and 314 ibr) ; these are often triangular, and are sometimes covered with plates which
vary in size and number, and frequently imbricate, or else they consist of soft integu-
ment with small skeletal granules scattered through it. These interbrachial regions
of the disc may be armed with spines of varying length.

On either side of the root of each arm, on the ventral surface of the disc, there
are one or two fissures or slits ; if two, one proximal and the other distal. These
bursal apertures (Figs. 245, 246, pp. 300, 301, and Fig. 314) lead into the bursie,
which will be described later. The aclradial edge of each of these slits is usually
supported by a single skeletal piece, the genital plate, while the interbrachial edge
is plated with a row of scales, which is directly continued into the plating of the
neighbouring interbrachial region.

Upper (apical) side of the disc.— It follows from what was said above (p. 327)

362 COMPARATIVE ANATOMY CHAP.

that the apical system, whether complete or incomplete, forms, in many Ophiuroidea,
even in adults, the greatest, or at any rate a considerable, part of the dorsal carapace
of the disc. Those regions which are not covered liy the apical system are plated
by the perisomatic skeleton. The plates of this skeleton vary much in size, form,
number, and arrangement, and not infrequently, especially in cases where the apical
system does not consist of large distinct plates, the dorsal integument of the disc is
soft, and only provided with scattered skeletal pieces, which are sometimes micro-
scopically small.

Ten large perisomatic plates appear most constantly (even more constantly than
any of the circle of plates of the apical system) ; one pair of these lies near the base
of each arm. These are called the radial shields (Figs. 244, p. 299, and 314 /•«),
and are often present even when there are no large plates in the rest of the dorsal
carapace of the disc. Sometimes the radial shields, covered with a soft integument,
reach from the base of the arm to near the centre of the disc, their presence being
then outwardly marked by a graceful rosette formed of five pairs of radial ridges.

V. Crinoidea.

(Cf. the apical and oral systems of this class, pp. 328-333).

The perisomatic skeleton of the Crinoidea consists of: (1) The
perisomatic skeleton of the calyx ; (2) the skeleton of the arms and
pinnulae ; (3) the skeleton of the stem.

(K) The Perisomatie 1 Skeleton of the Calyx.

In this are included all the skeletal pieces of the calyx, which do
not belong either to the apical system (central, infrabasals, basals, and
radials) or to the oral system (orals).

In the young stalked larva of Antedon the skeleton of the calyx
has no perisomatic pieces ; it consists exclusively of the typical plates
of the oral and apical systems (Fig. 270, p. 318).

The only forms in which this stage persists throughout life are
those of the type Inadunata larviformia, e.g. genus ll<ij>}un-inn*
(Fig. 297, p. 334). _

In all other living and extinct Crinoidea a perisomatic skeleton is
developed, although it varies in extent to an extraordinary degree.

This skeleton may consist of very various systems, and may be
developed both in the dorsal cup and the tegmen.

a. One, or several, or even many, pieces may appear only in the
posterior or anal interradius, especially in the dorsal cup supporting
or bordering the anus. These anals, which characterise the posterior
interradius, more or less markedly disturb *he regularly radial
structure of the calyx.

//. In all the five interradii one or many pieces may occur, both

1 Mr. F. A. Bather (Nattiral Science, vol. vi. pp. 418, 419 : 1895), in reviewing the
original German edition of this work, adduces strong reasons against the use of the term
"perisomatic" as here employed by the author. The term is, nevertheless, retained in
tliis English version because its exclusion seemed to necessitate the entire rearrangement
of this Section V.— TR.

vin ECHINODERMATA— MORPHOLOGY OF SKELETON 363

in the dorsal cup and in the tegmen. They are called interradials.
In the tegmen they develop in the zone between the orals and the edge
of the calyx, and belong to the interambulacral system of plates.
Usually, only the interradials of the dorsal cup are so called, although
they are not infrequently continued, between the bases of the arms,
into the interradial system of plates of the tegmen without any sharp
boundary.

<: The proximal portions of the arms may, to a greater or lesser
extent (to their first, second, etc. divisions), be taken into the calyx,
in which case the skeletal segments of the arms (brachials) become
perisomatic plates of the dorsal cup, and are known as fixed brachials
(primary, secondary, etc., formerly called radials of 1st, 2nd, etc.
orders). (For the meaning of these names, see below, the section on
the brachial skeleton, p. 370.)

'/. Just as interradials may appear in the dorsal cup between the
five radials and the fixed branchials of the five radii, so the branches
of each arm incorporated into the calyx may themselves be connected
by intercalated plates. Those which lie between brachials of the
second order are then called interdistiehals or interseeundibraehs,
those between brachials of the third order (after the second forking)
interpalmars or intertertibrachs, etc.

When more than five free arms rise from the edge of the calyx, i.e.
when some length of the arms and their branches is incorporated into
the calyx, the food-grooves running over the tegmen from the mouth
to the periphery divide dichotomously in such a way that the number
of grooves ultimately corresponds with that of the free arms. The
regions between the branches of the five primary radial food-grooves
are as a rule also plated with small interambulacral pieces.

e. The food-grooves running over the tegmen from the mouth
to the bases of the arms Arery often have a skeleton of their own,
which may be continued into the ambulacral furrows of the arms and
their branches. This ambulacral skeleton may consist of lateral
plates (which border the furrow laterally) or of covering" plates (which
cover the furrows, changing them into passages or tunnels), or of both
these sorts of plates. Subambulacral plates may also occur.

Special Remarks on the Perisomatic Skeleton of the Crinoid Calyx.

In the Inadunata larviformia (Type : Haplocriiius) there is no perisomatic
skeleton of the calyx. This latter consists exclusively of the plates of the apical
and oral systems (five basals, live radials, three of which are transversely divided,
and five orals).

The first perisomatic plate of the calyx occurs in related forms, interradially, in
the radial circle, and rests upon the posterior basal ; it is the anal.

As a type of the Inadunata fistulata, we shall first select Cyathocrinus. In the
dorsal cup only one perisomatic plate appears, which is found resting on the
posterior basal, between the two posterior radials (Fig. 289, p. 329). The apical
capsule thus altogether resembles that of the Larviformia. The tegmen calycis,
on the contrary, shows an entirely different condition, which, however, may vary

364

COMPARATIVE ANATOMY

CHAP.

considerably in the different species, and indeed in different individuals of one
and the same species. The orals now no longer occupy the whole of the tegmen,
but are supposed by some writers to be represented by certain plates which occur
at its centre, and vary in number and regularity, often being replaced by small
irregularly arranged perisomatic plates. The mouth is always hidden beneath
them. From these supposed orals the five ambulacral furrows run over the tegmen
to the bases of the five much branched arms. Each furrow is covered or bordered
by two or four rows of alternating covering plates. Interradially there are 5 plates
(deltoids), the edges of which meet beneath the ambulacrals and form the floors of
the furrows. They sometimes appear at the surface for a certain distance between
the ambulacrals ; in other cases, they are even here covered by more or less numerous
interambulacrals.

In some species of Cyathocrinus, and in many related Inadunata the posterior or
anal interradial area bulges out to form a ventral or anal sac, which is sometimes
cylindrical, sometimes club-shaped or bladder-like (Fig. 315). This anal sac, besides

the hind-gut, probably contained a large part of
the body cavity. It is covered with numerous
plates arranged in vertical rows. The plates
of the neighbouring rows alternate in some
species. The anus lies sometimes near the tip
of the sac, sometimes on its anterior side, and is
often encircled by special plates. The anal sac
may attain such dimensions that it is as long
as, or even longer than the arms. The first
tendency to the formation of such an anal sac is
met with in Hijbocrinus, in which the posterior
interradial region of the tegmen is somewhat
though still only slightly bulged.

The Inadunata so far mentioned are palaeo-
zoic forms. From them certain more recent
types may be derived. In Encrinus (Trias) the
anal sac has again become a short cone. In
forms closely related to this genus, and in
Marsupitcs (Chalk), the anal pieces as well
have disappeared, so that, while the base is
dicyclic, the regularly radial dorsal cup consists
only of the plates of the apical system, periso-

FIG. 310.— Cyathocrinus longimanus,
after Angelin, frum the anal side, after
removal of the greater part of the arms.
pr, Ventral sac ; x.
In, basals.

anal plate; r.radials; matic Pieces beinS> in tllis system, altogether
wanting.

The same is the case in the dorsal cup of the

family Holopidce (Lias to present time), Hyocrinida' (Lias to present time), Botliit-
crinidce (present time). In the tegmen calycis of these forms we first notice that the
large anal sac of the Cyathocrinidce is reduced to a small anal tube. In Holujms,
between the base of the open oral pyramid and the edge of the calyx, there is only a
very narrow zone beset with irregular perisomatic plates. This zone is still wider in
ff>/ocrinus (cf. Fig. 298, p. 335), and is thickly covered with numerous small plates.
Between the ambulacra! furrows lie the interambulacral plates ; the furrows, im-
mediately on emerging from between the oral plates, are bordered and covered by
lateral and covering plates. In the posterior interambulacral area, near the edge of
the tegmen, sometimes excentrically, there rises the short conical plated anal tube,
with the anus. In Bathycrinus, where the orals are either wanting or reduced, the
interradial region is either naked or plated with small pieces. The ambulacral
furrows have lateral plates only. The anus lies on a very short papilla-like anal cone.

vin ECHINODERMATA— MORPHOLOGY OF SKELETON

365

The Canaliculate/,, like the more recent Inaihtnufa (Lias to present time), are
distinguished by the regular radial structure of the dorsal cup, in which interradials
only exceptionally occur, and special plates in the posterior interradius (anals) never
occur. Very often (Apiocrinus, Ithizocrinus, Aitt':i!<n>i'/n-) two or more brachial
plates following the radials of the calyx are incorporated as "fixed brachials" into
the dorsal cup.

In connection with the tegmen calycis, it must be noted that among the Can ili-
culata, orals appear in the adult only in Bhizocrinus. As a rule, the tegmen calycis
is plated in the interambnlacral regions with numerous loosely connected skeletal
pieces, which vary in size according to the species and genus. These small plates
are perforated by pores. This skeletal covering is not infrequently continued on to

fa fa,

Fie. 316.— Tegmen calycis of Metacrinus angu-
latus, P. H. Carp, (after P. H. Carpenter), n, Mouth;
///•, arms ; ;i, pinnula? (both broken oft") ; tu, anal tube,
near which there is a second abnormal tube to] ; <•;/",
covering plates of the ainbulacral furrows.

FIG. 317.— Actinometra strota, P. H.
Carp, (after P. H. Carpenter). Tegmen
calycis. o, Mouth; ««, anus; «/«, food
grooves of the arms ; afil, the same of the
tegmen ; plt two pinnuhe, which take the
place of one of the two posterior anus.

the bases of the arms, and occasionally runs out between these apically in such a
manner as to be visible in the interradii of the dorsal cup.

The ainbulacral furrows of the tegmen calycis are rarely open, but usually covered
with covering plates and often bordered by lateral plates (Fig. 316). Occasionally
the mouth also may be covered with perisomatic plates, but it is usually open.

The anal tube in the posterior interradius varies in size and in its position within
this interradial area. Its plating agrees with that of the interambulacral area on
which it is found.

The interambulacral areas may also be naked, i.e. covered with integument
containing only very small calcareous corpuscles.

Actinometra is the only recent Crinoid in which the mouth is found
placed quite excentrically (anteriorly) on the tegmen, and the anus,
which lies in the enlarged posterior interradial area, comes to lie

366

COMPARATIVE ANATOMY

CHAP.

almost centrally (Fig. 317). In consequence of this shifting the
ambulacra are, of course, very unequal in length.

The (paleozoic) Camcmta are distinguished by the tendency to strong development
of the perisomatic skeleton in the calyx, and by the plates being so firmly inter-
connected as to form a rigid test. In the formation of the dorsal cup, the
bases of the arms are incorporated to a certain extent in such a manner that their
lower brachials become fixed plates of the cup. In the five interradii of the dorsal
cup, interradials appear, to which, in the posterior interradius, special anal
plates are often added. In those cases in which the arms take part in the formation
of the capsule beyond their first branchings, interdistichals, etc., may connect the
branches firmly together.

The tegmen calycis also consists of plates which are usually very numerous and
firmly connected together. Just as the mouth is always covered by characteristically
arranged, closely fitting, orals, so also the ambulacral furrows are never open, but are
always arched over by large covering plates, some of which may be distinguished by

A B

flf

FIG. 318.— Actinometra (after P. H. Carpenter). Diagrams to illustrate the courses of the fuml
grooves over the tegmen calycis. Ai-EL>, the directions of the rive pairs of anus. In the centre the
anal tube.

their greater size. In the older forms, the tegmen is, as a rule, rather flat, and the
covering plates of the ambulacral skeleton appear at the surface. In the course of
the geological development of the palteozoic epoch, however, the tegmen bulges out
more and more, and finally forms a high, firmly plated "vault" or dome (Figs.
253, 254. pp. 307. 308), which, immediately behind its centre, may be prolonged
to form a tube, often of greater length than the arms, with the anus at its tip.
Where such a highly arched dome is developed, the interambulacral plates, which
border the ambulacral furrows, send out processes over the- latter. The processes
(which are closely joined to one another) from one side of the ambulacra meet and
become firmly connected with those from the other, so that the ambulacral furruws
with their skeletons are completely arched over, and are not externally visible.

(This condition was until quite recently wrongly explained as follows. The
Camerata possessed an inner, naked, or merely loosely plated tegmen, in which the
ambulacra ran from the mouth in the centre to the periphery, and this tegmen
was arched over by a firmly plated vault in .such a way, that between the tegmen
and the vault there was a free space. )

The interradial plates of the tegmen are often continued directly, i.e. without a
boundary line, into the interradial plates of the dorsal cup.

The anus, surrounded by special plates, lies in the posterior interradius.

viii ECHINODERM ATA— MORPHOLOGY OF SKELETON 367

a. The apical capsule or dorsal cup. — In Platycrinus, the dorsal cup (cf. Fig.
254, p. 308) still consists exclusively of the plates of the apical system (three basals
and five large radials). The arms are free from their bases. A plate which is found
in each interradins, between the bases of the free arms and between the radials, may
be considered to belong almost as much to the tegmen as to the dorsal cup. In
Hexacnnus the radial structure of the apical capsule is essentially disturbed by the
appearance of an anal plate, which presses in between the two posterior interradials
in the posterior interradius, and to which, in the direction of the tegmen, two or
three other anals may be added. Further, in each radius, the one small primary
brachial plate present has become a fixed plate of the apical system. As a further

FIG. 31',!.— Gilbertsocrinus tuberculosus, Hall (after Wachsmuth and Springer). Tin- system
of plates of the dorsal cup and of the interradial appendages IB. Bo, point of attachment of the
arms ; B/, commencement of the free portions of the arms. For other lettering see p. 317.

example we may take Dimcrocrinus (Glyptasteridce), in which the dorsal cup is still
more complicated. In each radius the radial is followed by two primary brachials,
which are incorporated into the dorsal cup. In each case the second of these
brachials is followed by two or three secondary brachials, which are also fixed in
the dorsal cup, the last of them carrying a free arm. In each interradius there are
several interradials ; first a large plate which lies between the primary brat-Mais, and
then two more lying at the level of the secondary braehials. The posterior interradius
is broader than the others. The first plate here lies between the radials, and
agrees with them in size, then follows a second row of three plates, and, orally
from these, various small plates which lead over on to the tegmen calycis. Inter-
distichals may also occur. Mclocrinus (Fig. 252, p. 307) and bonier/ mi*, etc. agree
with Dimcrocrinus in these points.

In Gilbertsocrinus (Elaidui-finidcc) also, the two primary brachials and the two or
three secondary brachials are incorporated into the dorsal cup (Fig. 319). In each of

368

COMPARATIVE ANATOMY

CHAP.

the five interradii there are several (twelve) interradials, the arrangement of which
is shown in the figure. The anal interradius is hardly distinguishable from the
other interradii. The distichals or secondary brachials are connected by smaller
interdistichals.

The perisomatic skeleton of the dorsal cup of Adinocrinus (Fig. 291. p. 329)
is very like that of Gilbertsocrinus ; but the anal interradius is much larger than
the others, and its plates are divided into two lateral groups by the intercalation of
a vertical row of anal plates. This is also the case in Batocrinus (Adinocrinidcc}.
Here, however, not only the 5x2 primary brachials and the 10x2 secondary
brachials, but also the 20 x 2 tertiary branchials are incorporated into the dorsal
cup. In Sf rotocrinus (regalis) an extreme form is found (Fig. 320). The calyx

FIG. 320.— Strotocrinus regalis (after Wachsmuth and Springer). The apical border. The
conical portion of the dorsal cup is broken away (as far as the distichals ill) and shows the
teginen with the aims, the mouth and the food grooves. The dotted lines denote the manner of
branching of the tixed arms, an, anus ; &/, fixed joints of the arms, which form the border ; B/, the
free arms which run out from the edge of the border ; to, interambulacral region of the teginen
calycis ; am, ambulacra ; ;-/, fixed pinnulie.

is very large. The dorsal cup consists of a small conical portion above the stalk,
followed by a border spread out horizontally. In each radius each radial is followed
by two primary brachials. The second costal is in each case followed by the two
secondary brachials (making ten in all). Up to this point the above mentioned
plates, together with the apical system, form the conical part of the dorsal cup.
The plates which follow form the horizontally expanded border. Each distichal is
followed by a principal row of (six) plates, which runs radially to the edge of the
border, where the last plate carries a free arm. Accessory rows branch alternately
from these principal rows, three on each side. These also run to the edge of the
border, and the last plate of each row carries an arm-branch. Seventy free arm-
branches in all thus rise from the edge of the border. In the interradii, in the
interdistichal regions, and between all the further branches of the fixed arms, inter-

vm ECHINODERM AT A— MORPHOLOGY OF SKELETON 369

am-

Cfl

radials, interdistichals. etc. are found binding the bracliials into the rigid horizontal
border. Their number and arrangement are best elucidated by the figure. The anal
interradius is not distinguished from the others in any marked manner.

b. The Tegmen calycis. — The tegmen of Marsupiocrinus (ccelatus) is only slightly
vaulted. It is plated with numerous small, firmly connected pieces (Fig. 321).
Among these, we can easily distinguish the covering plates of the ambulacra, which
thus here come to the surface, and
can easily be distinguished from
the somewhat larger interradial
and interambulacral plates. In
the centre of the tegmen lie the
five orals, arranged in the manner
which is characteristic of the
C'amerata, and behind these,
subcentrally, the anal aperture,
surrounded by special plates.

If the ambulacral covering
plates are larger and more massive,
as in many species of the genus
Platycrinus, it is then more
difficult to distinguish the inter-
radial plates of the tegmen from
them.

The genus Agaricocrimis
affords examples of the specially
strong development of single covering plates of the ambulacral skeleton, which are
called radial dome plates. The tegmen is highly vaulted.

An extraordinarily highly vaulted tegmen is found in the Actinocrinidce
(Actinocrfnus, Batocrinus, Figs. 253, 254, pp. 307, 308). It is regularly and firmly

an.

PIG. 321.— Tegmen calycis of Marsupiocnnus coelatus
(after Wachsmuth ami Springer). <//•, Orals; KM, am-
bulacra ; cp, covering platas of the ambulacral furrows ;
ia, interambulacral region.

FIG. 322.— Part of the dorsal cup of Forbesiocrinus, spread out. For lettering see p. 317.
In addition, IO, one of the four similar interradial regions ; iA, the deviating anal interradial
region ; pal, palmars.

plated with large strong plates more or less equal in size. Nothing can be seen
of the ambulacral skeleton externally, it having been pressed down, or rather, over-
grown, as already described, by the interambulacral plates. In the posterior inter-
VOL. II 2 B

370 COMPARATIVE ANATOMY CHAP.

radius, immediately behind the centre of the tegmen, this dome is produced still
further into a long tube similarly plated ; this is the anal tube, on the tip of which
lies the anus.

The At-ticulata, so far as the perisomatic plates of the calyx are concerned,
agree with the Camcrata in that the ossicles of the arms are sometimes incorporated
into the dorsal cup as far as to their second or third divisions (Fig. 322), the
primary, secondary, and often also the tertiary brachials becoming fixed plates
of the dorsal cup. The number of the brachials in each arm and its branchings
varies. Three primary brachials are often found in each radius. But these fixed
brachials are not, as in the Camernta, rigidly connected inter sc and with the radials,
but are articulated. The spaces on the dorsal cup between the radii and between
their branchings are filled either with quite small, loose, and irregular calcareous
corpuscles or scales, or with small, definitely arranged plates (interradials, inter-
distichals, etc.). In the posterior interradius there are often, in addition, special anal
plates frequently asymmetrically arranged.

The tegmen calycis of one .species of the genus Taxocrinus is" wrell known. The
radii and their branchings are bulged out while the interradii are depressed. From
the central mouth, which is open and surrounded by five orals, the five ambulacra!
furrows run out, dividing dichotomously in correspondence with the branching of
the arms. Each ambulacral furrow has a fioor of two longitudinal rows of sub-
ambulacral plates, is bordered by lateral plates, and closed in by two longitudinal
rows of covering plates. The covering plates in the two rows are alternately
arranged, their interlocking forming a zigzag line, and it is very probable that they
were movable, i.e. that they could be raised and depressed. The interambulacral
regions contain a large number of small, loose, irregular plates. In the posterior
interradius, at the edge of the tegmen, there is a plated process (anal tube ?).
For Thaumatocrinus, see "The Systematic Review" (p. 309).

(/>) The Braehial Skeleton.

The calyx of the Crinoidea carries at its edge (on the boundary
between the tegmen calycis and the dorsal cup) five arms, which
are rarely simple, but usually branched, and in the living animal are
beautifully extended. The arms can be made by stimulation to
fold together over the tegmen. They are found in this position
also in dead animals, and therefore almost always in fossilised
individuals.

The arms, which contain important inner organs, are supported
by a special brachial skeleton. This consists of consecutive calcareous
pieces, the brachials, which are either firmly connected or articulated
with one another. The brachials are deepened on their oral side,
that which is directed upwards when the arms are spread out, to form
a more or less deep longitudinal groove along the arms and all their
branches ; this is the ambulaeral groove. In the base of this groove
lie the most important inner organs of the arms (radial canals, water
vessels, outgrowths of the body cavity, etc.). The soft integument
which covers these organs, and stretches over the ambulacral grooves
of the brachial skeleton is in its turn depressed to form a channel.
These integumental channels, which accurately correspond with the
ambulacral grooves of the skeleton, are called food grooves. At the

vni ECHINODERMATA— MORPHOLOGY OF SKELETON 371

bases of the free arms they pass into the ambulacral grooves or food
grooves of the tegmen calycis, which run to the mouth.

The arms, when divided, as they normally are, usually branch
dichotomously ; occasionally, however, they give off alternating
branches, which may again branch alternately. In most Crinoids the
arms and their branches carry, at the sides of the ambulacral grooves,
closely crowded and alternating processes ; these are rod-shaped, and end
in a point, and are known as pinnules. The skeleton of these pinnules
resembles that of the arm, and, like the latter, is jointed. The pinnules
may best be regarded as the ultimate branchings of the arms, and it is
very probable that in the numerous pakeozoic Inmhintifa that have no
pinnules the last branches of the arms fulfilled their functions.

The brachial skeletons in the Orinoidea are always direct continuations of the
radials of the apical capsule. The first plate which follows the apical radial radially
must be considered, morphologically, as a brachial or ossicle of the arm, although
it is only rarely^.;/, in the Inadiuiata) a free ossicle. The terms introduced to
denote the various orders of brachials are almost as numerous as the writers them-
selves. It is the clearest plan to speak of them as brachials of the first, second,
third, etc. orders, or as primary, secondary, tertiary, etc. brachials. Such a plan
was, however, soon found too cumbrous for practical purposes, and was supplanted
by the terms costals, distichals, palmars, and postpalmars. To these terms, how-
ever, considerable exception may be taken, and it seems simplest to adopt the intel-
ligible and congruent terms primibrachs, secundibrachs, tertibrachs, etc., which are
capable of indefinite extension, and are readily symbolised as IBr, IIBr, etc.

It has already been pointed out, in the section which treated of the perisoniatic
plates of the calyx, that brachials are incorporated into the dorsal cup in many,
indeed, in the great majority of Crinoids. We can accordingly distinguish free
brachials from fixed brachials, the latter being those which have become peri-
somatic plates of the dorsal cup. The first brachials to be so incorporated are
naturally the primibrachs, the next the secundibrachs, the tertibrachs may also then
become fixed. In describing the skeleton in detail, therefore, the terms fixed primi-
brachs, fixed secundibrachs, etc. are used, and the number of these plates in each
arm is given. The arrangements found in the various divisions of the Crinoidea
in this respect have been already briefly described in the preceding section. That
of the Inadunata is the simplest, since in them the arms are free from their very
bases (hence the name), the first primibrach being a free ossicle of the arm ; the most
complicated condition is that of certain Cameriita (Actinocrinoidca, etc.), in which
the brachials of several orders are incorporated into the calyx, and being connected
by interradials, interdistichals, etc. lend to the dorsal cup its rich plating.

In branched arms those joints above which the divisions or branchings take place
are called axillary, f.y. we have an axillary costal, axillary distichal. or, as they
may more simply be called, primaxil, secundaxil, etc. (lax, Ilax, etc.).

With regard to the distribution of the pinnula1, it is the rule, at least in modern
Crinoids, that the axillary joints never carry pinnula?, and that where two joints
are connected by syzygial sutures or by ligaments, pinnulte are also wanting on the
lower or proximal joint.

There are three different ways in which the free brachials may be arranged. Tin-
arms may consist of a single row of joints, the brachials being superimposed in a
single series with parallel surfaces of contact (uniserial). Again, the joints may
" alternate," if they are wedge-shaped, and if, in the row, the thick and the thin sides

372

COMPARATIVE ANATOMY

CHAP.

of the wedges regularly alternate. Or again, the joints may be arranged in two series
or rows, the contact-surfaces of the one row alternating with those of the other, and
the two rows themselves interlocking along a zigzag line (biserial).

The Articulata, many Canaliculate/,, and the recent Inadunatn have the joints
of their arms arranged in single rows. This condition has been proved to be
ontogenetieally and phylogenetically primitive, i.e. for the palaeozoic InaiJumitu
and the Camerata. The majority of pakeozoic Inadunata have uniserial arms,
but towards the end of the palaeozoic period forms appeared with alternating rows.
(e.g. Pofcriocrinus), and finally some genera in which the brachials may be biserial
at the tips of the arms (Ihipachycrinus, Erisocrinus, Hydreionocrinus).

Most of the Camerata (an order limited to the palaeozoic age) have biserial
arms. But by far the greater number of the Lower Silurian species have uniserial

bf bf

FIG. 323.— Part of the arm
Of a Crinoid. Diagram showing
the transition from the uniserial,
through the alternating, to the
biserial arrangement of the
brachials.

Fir;. 3-24.— Part of the disc formed by the arms
of Crotalocrinus rugosus (alter Wachsmuth and
Springer). 2, The trabecula- connecting the arms ;
///•, the arms with the covering plates (opn) over their
food grooves ; in 3 these covering plates are removed.

arms. In the Upper Silurian, however, but few forms persisted with such arm*.
and they are found side by side with species and genera with alternating, or with
two rows of, brachials.

In Crinoids whose arms have two rows of joints, the uniserial and the alternate
stages are passed through ontogenetieally. It must, further, be specially emphasised
that not a single case is known of arms being formed of two rows of brachials
throughout their whole length, i.e. from the radials of the calyx to their tips. At
'their bases the arms always, for a certain distance, have a single series of brachials,
then they have alternating brachials, and finally two rows. The transformation of
the uniserial arm into an alternate, and finally into a biserial one commences,
ontogenetieally and phylogenetically, at the tip of the arm, and proceeds from that
point towards the base.

The food grooves of the arms resemble those of the calyx. They are sometimes
naked and open, and at others provided with a variously developed ambulacra!
skeleton, consisting either only of lateral plates, or of lateral and covering plates.
Subambulacral plates may also occur in the floor of the food grooves, dividing them
from the subjacent organs of the ambulacra! furrows of the skeleton (body cavity of
the arms, genital strands, pseudohaemal canals, etc.). Where covering plates are
present there are two rows which alternate and interlock in such a way as to form a

vnr EGHINODERM AT A— MORPHOLOGY OF SKELETON 373

median xigxag line. Tliese plates can be raised and depressed in the living animal ;
when they are raised the food groove is open, when depressed, it is shut.

An altogether peculiar arrangement is found in the arms of the genus Crotuln-
crinus (Upper Silurian, England. Sweden), which is thought by some to belong to
the Caincriitii. The free arms branch extraordinarily frequently, the separate branches
being closely crowded together, and forming together a wide expanded coherent
disc round the calyx, resembling the fully open corolla of a flower. As many as 500
to 600 branches may in some forms reach the edge of this disc (C. rurjosus, Fig. 324).
Each ossicle of the arms has two lateral processes, which become connected with
similar processes of the corresponding ossicles of the neighbouring arms or branches,
so that the disc formed in this way by the skeleton of all the free arms is lattice-like.
At definite distances from the calyx the brachials are of equal length, so that they,
as well as the sutures which lie between the consecutive brachials, seem to be arranged
in regular concentric rings round the calyx. The whole braehial disc was very flexible,
and could be rolled up over the calyx from its periphery. In C. pulchcr, the brachial
disc falls into five broad radial lobes, which, when the disc closes over the calyx, over-
lap like the petals of a bud. The food grooves are covered by double longitudinal
rows of alternating covering plates.

(c) The Stem (Columna).

The great majority of Crinoids are attached to the bottom of the
sea by means of a jointed stem. Among recent Crinoids only the Ante-
doniihfi and Thaumatocrinus are, in the adult condition, non-pedunculate
and unattached. The stalked condition is undoubtedly the more
primitive, for (1) the Crinoids show very markedly the habitus
characteristic of many attached animals, and (2) all free and unstalked
Antedonidce pass through an early stalked and attached stage. The
stem, which varies greatly in length and thickness, consists of a series
of calcareous ossicles one above the other, the uppermost of which is
connected with the centre of the apical system, and carries the calyx
with its arms.

The ossicles of the stem (columnals) vary greatly in shape. They may be flat and
disc-like, or long and cylindrical ; sometimes they are gradually thickened towards
each end in such a way as to resemble dice-boxes. Further, the columnals in
different parts of one and the same stem may be very different. The external outline
of the ossicles in transverse section is sometimes pentagonal, sometimes round,
rarely elliptical. They are connected with one another more or less firmly by sutures,
or else are movably articulated. The stem throughout its whole length is pene-
trated by a central canal (axial canal), which thus runs through all the consecutive
columnals. Within this canal run the ccelomic canals (continuations of the chambered
organ) and nerves. The size of the canal in transverse section differs as much as
its shape. The outline of its section seems most frequently to be pentagonal or quinque-
lobate, but it is not infrequently round. Occasionally also the central canal is sur-
rounded by five narrower peripheral canals.

New ossicles are added, as the animal grows, at the upper end ; at first they are
small and flat, and often concealed within the stem. The most constant place of
their appearance is between the uppermost columnal and the base of the calyx.
New ossicles may, however, also be intercalated between two already formed ossicles,
but this almost always takes place at the upper end of the stem. In a growing stem
the ossicles in the upper part vary greatly in length, the shortest being the youngest.

374

COMPARATIVE ANATOMY

CHAP.

At definite intervals the stem may carry whorls of so-called
cirri. These are jointed processes of the stem, pointed at their tips,
and perforated by a longitudinal canal which communicates with the
central canal of the stem (Figs. 257, 258, pp. 311, 312).

The cirri are, as observations on living animals have shown, very mobile. Five
cirri, as a rule, belong to one whorl, being inserted 011 the five sides of the nodal
ossicle. Between two consecutive nodes there are a varying number of columnals
which do not carry cirri. These together form an internode. Whereas in the
Iiuiclunnta, ArticiiJutii, and 6''///' /•"/" cirri are, as a rule, wanting, or only present
at the lower part of the stem, in the CanaJictihtta (Pcntacrinidce) nodes are found
along the whole length of the stem between the consecutive internodes. In the

A B

FIG. 326.— Diagram to elucidate Wachsmuth and Springer's rule. A. Cnnoid with dicyclic
base. B, Crinoid with nioiiucyclic base. Fur lettering see i>. ::17.

recent species of Pcntacrinus and Mi'tacrimis, each nodal ossicle is connected with
the next ossicle of the internode below it by a syzygial suture.

Peculiar relations exist between the stem and the base of the apical capsule,
according to the " rule of Wachsmuth and Springer," given in the diagram Fig. 325.
In Crinoids with dicyclic base (i.f. where the base consists of basals and infrabasals,
with pentagonal stem and five-rayed central canal, the five edges or angles are interradi-
ally arranged, while the five rays of the central canal and the five cirri of each whorl
are radially arranged. In Crinoids with monocyclic base (i.e. where the base consists
exclusively of the basals, Fig. 325 B) the reverse is the case. In those Crinoids which
possess cirri, and in which the stem and central canal are not round, the character
(monocyclic or dicyclic) of the base of the calyx can be determined — apparently with
great certainty — from an examination of the stem. This is of importance in forms
in which the infrabasals are very small, or, being covered by the uppermost joint
of the stem, are hidden, or when they occur only in a young stage. Such forms are
said to be constructed on a dicyclic plan, and have been called " pseudo-monocyclic. "
It is possible that certain genera in which Wachsmuth and Springer's rule appears
to be violated may eventually be proved pseudo-monocyclic. Meanwhile, however,
the rule is not absolutely universal.

The lower part of the Crinoid stem is called the root. It serves,
in various ways, to attach the body to the sea floor. If the latter

viii ECHINODERMATA— MORPHOLOGY OF SKELETON 375

be muddy or sandy, the base of the stem puts out lateral branches,
the so-called root-cirri, the numerous ramifications of which penetrate
the sea floor in all directions. The end of the stem itself may at
the same time branch like the root-cirri. When the sea floor is
rocky, the root-cirri spread out more horizontally, accommodating
themselves to the surface to which they are attached, and becoming
cemented to it at their ends by means of a calcareous secretion.

Further, it is almost certain that individuals of some species, e.g. Pentacrinidce
and some Palaeozoic crinoids, are capable of free locomotion when the stem is either
voluntarily or accidentally
broken. Such locomotion is
no doubt chiefly promoted by
the movements of the arms,
the cirri serving rather for
attachment.

In Hoi opus (Fig. 250, p.
305) the stem is wanting.
The calyx, which resembles
a reversed cone in shape, is
cemented to the substratum
by means of an irregularly
expanded calcareous mass.

The Antedonidtx are
stalked and attached
only in the young stage.
The larval stem re-
sembles that of the
Bourgueticrinidse (Fig.
326) ; the cirri are
developed only on its
uppermost ossicle, on
which five radially ar-
ranged cirri first appear,
then five interradially

-la

arranged. At a certain

FIG. 326. — Several stalked young stages of Antedon
phalangium (A) ; Antedon spec. (B) ; Antedon tuberosa (C) ;
and Antedon multispina (D), after P. H. Carpenter. For

<•!<', Points of attachment of the cirri.

T~ . . , aim iiuuuuoii mm

stage, differing in the lettering see p. 317.
different species, the
calyx together with the uppermost columnal (i.e. the one carrying cirri),
which is fused with the infrabasals to form the centrodorsal, separates
from the stem, the latter remaining attached to the place where it was
fixed. Above the cirri already formed, i.e. between them and the base
of the calyx, new whorls of cirri continuously appear on the centro-
dorsal, which constantly increases in size, so that we are tempted to
consider this piece as part of the stem, consisting entirely of nodal
ossicles fused together without intervening internodes.

The Comatulidae can both swim by the rowing movements of their arms, and
creep by means of the cirri and of the arms. They can, further, anchor themselves
by their cirri, the arms being then directed upwards.

376 COMPARATIVE ANATOMY CHAP.

(d) The Manner of Connection between the Skeletal Pieces.1

Under this head we have to consider the method of connection between the
ossicles of the arms and of the pinnulse, between the plates of the apical capsule, and
between the coltunnals.

Perhaps the clearest view of the great diversity which prevails in this matter is
obtained by assuming that the plates composing an echinoderm skeleton develop
in a stroma of connective tissue fibrils ; all the plates might thus be supposed to have
been originally but loosely united by such fibrils.

This condition persists in what is known as the loose suture. The ossicles of the
pinnules in many living Crinoids are united in this way.

From this loose suture we can obtain all the man}- variations which are found
either in the direction of greater rigidity or of greater flexibility.

In the former case we have :—

1. The close suture, also known as synostosis, in which the connecting fibres
are short, and " the joints closely and immovably fitted together, though they can be
separated by the action of alkalies," e.g. the radials of Antedon.

2. The syzygy, which is a special case of close suture, viz. that in which, if
pinnules or cirri are carried by the ossicles, the lower one loses its pinnule or cirrus.

The two components of a syzygy are termed epizygal and hypozygal.

3. Anchylosis, in which the two plates or ossicles are immovably cemented
together by an unbroken deposit of calcareous substance, which, however, is less
solid than that of the plates themselves (e.g. the basals of Bathycrinus and the
radials of Rhizocrinus).

On the other hand we have the modifications in the direction of greater flexibility
which lead up gradually to the development of true muscles, the original undiffer-
entiated fibrils becoming muscularly contractile. Such muscular articulations may
indeed be very highly specialised, with interlocking ridges and teeth on the opposed
facets of the ossicles.

As all these different modifications pass into one another by imperceptible stages,
it is not always easy to say whether in any special case we have to do with muscular
articulation or with a less specialised form of connection. It now seems probable
that many of the fibrous connections which were at one time thought to lie only
elastic fibres are really muscles.

For instance in the arms of living Crinoids, a pair of muscles on the oral side are
counteracted by a pair of bundles of " elastic " fibres on the dorsal side. The action of
the muscles in contracting rolls the arms up orally, and on the muscles relaxing, the
" elastic " fibres expand the arms again. Now it is clear that if these fibres were
thus simply elastic, Crinoids would die with their crown of tentacles expanded,
whereas the reverse is the case.

Again, the cirri are actively movable, often (e.g. in Pentacrinus) more so than
the arms, although no muscular articulations occur, in them.

From these facts it is rightly argued that the fibrils in these cases, though differ-
ing histologically from the true muscles, are yet to some extent muscular.

That all these various connections are in reality derivations from some common
primitive form of connection, we gather from the fact that in the stems of Crinoids
we may have anchylosis, close suture, syzygy, loose suture, and true muscular
articulation.

1 This passage was rewritten in accordance with the ten our of Mr. Bather's criticism
in Natural Science, vol. vi. pp. 420, 421. — TR.

vni ECHINODERMATA— MORPHOLOGY OF SKELETON 377

(c) The Nerve Canals of the Arms and of the Apical Capsule.

(Figs. 327-330).

The skeletal joints of each arm (the brachials) are perforated by
an axial canal, which is continued to the tip of the arm, and into the
pinnulse. Where the arms fork or branch in various ways, the axial
canal does the same. It contains nerve strands, and may thus be
suitably considered as a nerve canal. This canal is continued right into
the base of the dorsal cup, perforating the radials, basals, and also the
infrabasals, when present. All the nerve canals, and the nerves within
them, converge towards the apex of the calyx, where, either in the base
of the dorsal cup itself (surrounded by the basals in stalked Crinoids), or
partly enclosed within the centrodorsal (Antedonidce), lies the central
organ of this nervous system, which surrounds the so-called five-
chambered sinus in the shape of a cup or capsule. From this point
the already mentioned central, or axial, canal runs through all the
ossicles of the stem, giving off lateral branches to the cirri.

The nerve strands arise from this apical central nervous system at five interradial
points. The five interradial strands fork either in the basals or in the radials.
Within the radials each branch of a strand becomes connected with a branch of the
neighbouring strand, and from these radially arranged points of junction the radial
nerve strands originate which pass from the radials into the costals, and are
continued into the ossicles of the arms and of their branchings. Within the circle
of radials there is, besides, a circular commissure between the radiating nerve
strands ; the following diagrams illustrate the courses taken by these.

In the Pcntacrinidce, Encrinidce, and Antcdoiudcc, where the nerve strands
divide in the first axillaries, there is a peculiar chiasma nervorum brachialium,
which is shown in the diagrams.

In Encriiius, and it is said also in Pcntacrinus, the nerve strands which run
through the ossicles of the arms are double. But whereas, in Encrinus, they run
separately, and are enclosed in separate canals, in Pcntacrinus they lie in a common
canal.

Many palaeozoic Crinoids, and above all the Camcrata (with the exception of the
Crotalocrinoids), appear to have no differentiated nerve canals in the arms.

(/) The Water Pores.

In the Canaliculate (e.g. Pcntacrinus, Antedon, Adinoiiictra} the
tegmen calycis, whether naked or plated, is perforated by so-called
water pores, whose significance will be discussed more in detail
later on.

If the tegmen is plated, many or all of the plates of the interambulacral areas
are perforated by one or more such pores. In Pcntacrinus dccorus, as many as twenty
pores are found on one plate. The total number of pores varies greatly in different
genera and species. In Antedon rosacea it has been estimated at 1500, and in other
forms may be even greater. The pores are usually limited to the tegmen, and are
least plentiful in the posterior interradius. They may, however, also occur on the

FII;. 327.

FIGS. 327-330. — Diagrams to illustrate
the course of the axial canals and the
nerve strands within them in the dorsal
cup and the first brachial joints of En-
crinus (Fig. 327, after Beyrich), Rhizocrinus
lofotensis (Fig. 328, after P. H. Carpenter),
Antedon rosaceus (Fig. 329) and Bathy-
crinus aldrichianus (Fig. 330, after P. H.
Carpenter). In Fig. 327 only the distal ends
of the interradial canals are represented.
The parts which are transversely streaked
run superficially on the inner side of the
basal platt >.

FIG. 330.

CH. viii ECHINODERMATA— MORPHOLOGY OF SKELETON 379

edge of the calyx between the bases of the arms, and in tho genus Adinoni' //•".
where they are chiefly developed near the ambulacral furrows, they have occasionally
been observed on the lowest pinnula? as well, and even on pinnulre in the middle
or towards the ends, of the arms.

In lihizocriiom, in each interradius of the tegnien calycis there is only one water
pore perforating the oral plate. In Hyocrinus the anal oral plate is perforated by
two pores ; the other oral plates either have one pore each or else none at all.
Further, in this genus, 2 to 7 pores occur in the plates of the interamlmlacral arras
lying between the oral pyramid and the edge of the calyx, except in the posterior
interambulacral area, where they are wanting.

It is impossible to decide with certainty whether the pores which, in certain
Camerata (Actinocrf-n ida:, Melocrinidce, Khodocrinidce), occur at the edge of the calyx at
the bases of the arms (and correspond with the arms in number) are the homolcgues
of these water pores. The same applies to the slit-like pores which are supposed by
some writers to perforate the edges of the plates of the ventral sac in the Inadiin'it"
fistulata (along the sutures), and to the pores which are found along the brachial
furrows in the Inadunata larviformia. These pores may, in some cases, have
been connected with hydrospires (see the next two sections). In many I-iiadnmilii
there is what appears to be a true madreporic plate in the posterior interradius of
the tegmen.

VI. Blastoidea.

One part of the perisomatic skeleton of the Blastoidea has already
been described in connection with their apical system. There are
flve interradial or deltoid plates, which surround the oral region
(peristome), and radiate out from it (Fig. 331). These deltoid plates
do not form a closed circle, i.e. do not touch one another laterally,
but are separated from one another by the proximal portions of the
five ambulacra.

In describing the rest of the perisomatic skeleton, which, apart
from the stem, consists exclusively of the skeleton of the ambulacra,
it is useful to select a few typical forms.

(a) The Ambulacral Skeleton.

1. Pentremites. — Fig. 263, p. 314, shows a representative of
this genus in profile, Fig. 331 from the oral side. The five ambulacral
regions, or ambulacra, together form a five-leaved rosette round the
peristome (Fig. 331, A, B, C, JJ, E}. They are separated from one
another by the five (interradial) deltoid plates (3). The larger, distal
part of each ambulacrum fits in between the two limbs (10, 106) of
the forked radial s.

On the egg- or pear-shaped body the ambulacra stretch as far as
to the equator, or even further, towards the apical pole.

The skeleton of each ambulacrum, when most complete, consists
of the following parts :—

(a) One lancet plate (6).

(/>) One lower lancet plate (12).

380

COMPARATIVE ANATOMY

CHAP.

(c) Two rows of side plates (5).

(d) Two rows of accessory side plates (8).

(e) Two rows of pinnulre (2).

(/) Two groups of folds of hydrospire pouches (13).
(g) A double row of covering plates (1).

D

FIG. 331.— Diagram of the organisation of a Pentremites (nn^ina!). A, B, C, D, E, the live
ambulacra. A, Ambulacrum with covering plates (1) and extended pinnula- (2). B, Ambulacrum
with depressed pinnuhc. C, Ambulacrum after removal of the pinnuhe and covering plates. D,
After the further removal of the side plates and accessory side, plates (except 3). E, After removal
of the lancet plate as well. In the centre is seen the mouth with the spiracles surrounding it, in
the posterior interradius the anus. 1, Covering plates ; 2, pimiulse ; 3, deltoid plates ; 4, their
sloping ambulacral edges ; 5, side plates ; 6, lancet plate ; 7, pores ; S, outer or accessory side
pieces ; 9, furrow of unknown significance on eacli side plate ; 10, radials = fork plates : 11. aperture
of the ambulacral canal ; 12, lower lancet plate ; 13, hydrospire folds.

Passing over the covering plates, which are very rarely retained, we have, at the
centre of each ambulacrum, a skeletal plate about half as wide as the ambulacrum
itself, and more or less similar in shape. This is the so-called lancet plate (Fig.

vni ECHINODERMATA— MORPHOLOGY OF SKELETON 381

331, 6). On its outer, i.e. oral surface, this plate shows a more or less deep longi-
tudinal furrow, which gives off alternating lateral furrows to right and left. This
longitudinal furrow of the lancet plates is universally considered to have the same
significance as the food grooves on the tegmen calycis, and on the arms, of Crinoids.
Each lancet plate is longitudinally pierced by a canal, the so-called ambulacral
canal.

The space on each side between the lancet plate in the middle, and the lateral
margin of the ambulacrum (which latter is formed by the sloping edges of a deltoid
plate and of one limb of a fork plate (radial)), is occupied (a) by a longitudinal row
of large side plates (5), and (b) by a longitudinal row of smaller accessory side
plates (8). The number of side plates and accessory side plates corresponds with
that of the lateral branches of the median ambulacral furrow on the lancet
plate. Each side plate consists of a narrow portion which is directed towards the
edge of the ambulacrum, and of a broad portion which is in contact with the lancet
plate. The broad parts of the consecutive side plates in a row are in contact with
one another, but between the narrow parts of these same plates are spaces, in each
of which lie an accessory side plate, and a hydrospire pore (7). This latter leads
below the surface to the hydrospire pouches under the ambulacrum. The hydro-
spire pores, the accessory side plates, and the narrow ends of the side plates follow
one another regularly in this order.

The margins of the ambulacra carry thin long jointed appendages,
the pinnules (2), which may be compared with the structures of the
same name in the Crinoids.

The pinnules are retained only in rare cases, and are then depressed from the
two sides orally over the ambulacral area (ambulacrum B, Fig. 331). There can,
however, be no doubt that they could be raised and opened out (ambulacrum A).
The number of the pinnules corresponds with that of the side plates in a longitudinal
row, as also with that of the accessory side plates and of the hydrospire pores. The
points of attachment of the pinnulte lie between the consecutive pores. - Each
pinnule consists of a large number of skeletal pieces, which, near the base, alternate
in two rows, but above this are arranged in one row.

If the lancet plate of an ambulacrum is removed (E, Fig. 331),
the (smaller and thinner) lower lancet plate (12), which lies close below
it, appears at the surface.

This plate resembles the lancet plate in shape. If the pinnulse, the side plates,
and the accessory side plates be removed, the edges of the plates which border the
ambulacrum (deltoid plates, limbs of the fork pieces) are seen sloping towards the
floor of the ambulacrum. These sloping edges carry a longitudinal row of transverse
ridges, alternating with depressions, into which latter the narrower (outer) portions
of the side plates fit. On each side, between the lower lancet plate and the sloping
lateral walls of the ambulacrum, some of the parallel clefts and folds of the hydro-
spire pouches (13) are seen lying parallel to the longitudinal axis of the ambulacrum.
At the central or proximal end of the ambulacrum (i.e. near the peristome), the
interradial deltoid plates meet, and are joined by a radial suture. An aperture in
this suture, the ambulacral aperture (11), leads to the interior of the calyx. Through
this ambulacral aperture, the ambulacral canal, which traverses the lancet plate
longitudinally, is connected with a circular canal which surrounds the oesophagus.

In closest proximity to the peristome there are five large inter-

382

COMPA RATI VE A NA TOM Y

CHAI>.

radial apertures, the so-called spiracles. Each of these apertures
leads into the hydrospire pouches in such a way that the halves of
two adjoining ambulacral areas have one common spiracle for their
hydrospire pouches.

Each spiracle forms a depression in the central part of the corresponding deltoid
plate, and is further bordered liy the proximal side plates, and by the proximal
ends of the lancet plates. Occasionally, each spiracle is more or less distinctly
divided into two by a vertical median ridge (septum) projecting into it from the
deltoid plate. In the posterior interradins, the spiracle is confluent with the
anus.

The hydrospires (Fig. 332) are calcareous pouches or tubes
united in groups.

A group of such calcareous pouches is arranged symmetrically on each side of
the middle line of, and in close connection with, each ambulacrum. The pouches,

Fir;. 33:.'.— Section through an ambulacrum of Pentremites, diagrammatic. 1, Deltoid, while
lower down is the radial ; •_', hydrospire pore ; 3. accessory side plate ; 4, side plate ; 5, lancet plate,
with its ambulacral canal 7 ; 0, covering plates ; 8, common channel, into which the hydrospire
pouches (11) enter at !• ; 10, base of a pinnula ; 11, hydrospire pouches ; 12, lower lancet plate.

which hang down into the cavity of the calyx, are parallel to one another, and
stretch from the distal end of the ambulacrum to its proximal end, as far as to the
spiracle, through which they open externally. In addition to this opening, each
hydrospire pouch possesses a slit-like aperture extending along its whole length, in
the ambulacral area. These hydrospire folds are hidden, lying partly under the side
plates and partly under the lancet plate, After the removal of these latter pieces,
as we have seen, they appear at the surface. They vary in number from three to
nine. The concealed hydrospire canal (8), into which, on each side, the hydrospire
pouches open through their clefts, communicates with the exterior by means of the
hydrospire pores already mentioned.

The hydrospire pouches or tubes have thus a double manner of communication
with the exterior, \iv.. through the five or ten spiracles round the mouth, and
through the numerous hydrospire pores at the lateral edges of the ambulacra.

In certain species, the peristome was overarched by a roof of covering plates,
for the most part irregularly arranged (Fig. 331, 1) ; at their centre, five oral plates

NX

vin ECHINODERM ATA— MORPHOLOGY OF SKELETON 383

can sometimes be distinguished. The covering plates, which are rarely retained
complete, are occasionally continued on to the food grooves of the ambulacra, where
they are arranged in t\vo longitudinal rows. Perhaps they could be raised and
depressed ; if not, it is difficult to see how the food groves with their lateral furrows
could function : the lateral furrows would then have to pass under the covering
plates so as to be in communication with the principal groove. In rare cases the
covering plates even spread sideways over the spiracles.

•2. Codaster (Fig. 265, p. 314). — The arrangements here differ considerably from
those of Pentremites, just described. The food grooves are deeply sunk into tin-
lancet plates, which are hollowed out on each side for the reception of the side plates.
Spiracles are wanting. A certain number of the hydrospire clefts which run parallel
to the ambulacrum always appear at the surface of the calyx, laterally to the
ambulacrum (Fig. 333). These clefts run at right angles across the suture between

FIG. 333.— Transverse section through an ambu-
lacrum of Codaster (after Etheridge and Carpenter),

diagram. 1, Deltoid plate or possibly a radial ; 2, ainbu-
lacral canal ; 3, food groove ; 4, lancet plate ; 5, side plate ;
ii, apertures of tlie hydrospire pouches ; 7, lower lancet
plate ; S, hydrospire pouches.

Fie. 334.— Eleutherocrinus Cas-
sedayi.from the oral side (after Ethe-
ridge and Carpenter). aa-l>l>, Axis,
passing through the mouth and anus ;
y, radials ; iY, iuterradials ; rj, the
radial of the differently shaped am-
bulacrum ; :, y, the two larger basals.

the radial and the deltoid plates. One or more hydrospire clefts may be covered by
the side plates of the ambulacra. At those sides of the two posterior ambulacra
which are turned towards the anus, the clefts are altogether wanting.

3. In Orophocrinus (type 0. stcllif minis, Fig. 266, p. 314) there are no hydro-
spire pores on the ambulacra, but depressions occur between the consecutive side
plates for the reception of the bases of the pinnules. The hydrospire clefts lie quite
hidden below the surface within the ambulacra! sinus, covered by the lower lancet
plate. The spiracles, on the contrary, which are ten in number, appear at the sides
of the ambulacra as long wavy slits. The two spiracles of the posterior interradius
are distinct from the anus. The ambulacra, round the mouth at least, are covered
by covering plates.

4. The Irregulares (Astrocrinus and Eleutherocrinus) are chiefly distinguished
by the fact that one of the four ambulacra is developed quite differently from the
others (Figs. 334, and 267, p. 315).

384 COMPARATIVE ANATOMY CHAP.

The bundles of hydrospire tubes or pouches of the Blastoidea
have been compared with the " bursae " of the OpMuroidca. They are
said to have served, like these latter, for respiration and for the
ejection of the genital products. The similarity of these apertures is
specially marked if we compare Orophocrinus with an Ophiurid.

(b) The Stem.

With the exception of the genera Pentephyllum, Eleutherocrinus,
and Astrocrinus, which, at least in the case of all known adults, were
non-pedunculate, the Blastoidea were attached to the substratum by
a jointed stem without cirri (Fig. 264, p. 314).

VII. Cystidea.

The study of the skeleton of this ancient class, which is limited to
the palaeozoic age, has no very great comparative interest. The organi-
sation of the very heterogeneous groups which are classed together
under this heading can only to a small extent be understood from
their skeletal remains. According to the structure of the skeleton, we
can perhaps distinguish two principal groups : the Cystocrinoidea, whose
skeleton consists of comparatively few definitely arranged plates, and
which in some forms approximate the Crinoidea ; and the Encyst i<l en,
whose skeleton consists of a very large number of plates, showing no
definite recognisable order.

It is characteristic of most Cystidea, that all or some of the
plates of the skeleton are perforated in various ways by pores, which,
however, never seem to establish communication between the interior of
the calyx and the exterior. It is difficult to ascertain the significance
of these pores. They could not serve for the passage of the ambu-
lacral feet, since the pore canal, as already said, does not stand in
direct communication with the interior of the calyx. It is now pretty
generally accepted that, as the water passed through them, they served
for respiration. The following principal forms of pores can be
distinguished : —

1. Scattered single pores.

2. Scattered double pores (the pores being united in pairs) (Fig.
260, p. 312).

3. Double pores arranged in rhombs. In this case, the two pores
of a double pore are found on two neighbouring plates, and are con-
nected by a furrow or a canal, which sometimes runs at the outer
and sometimes at the inner side of the plate. This canal or furrow
lies at right angles to the suture between the two plates, and
the suture itself lies diagonally to the rhomb formed by the pores.
Such pore -rhombs may occur on all the plates of the Cystid
test, or again may occur singly. In the latter case, the two halves

VIM ECHINODERMATA— MORPHOLOGY OF SKELETON 385

of a rhomb are not infrequently divided by a smooth intermediate
region.

Since it is difficult or even impossible to give any comprehensive description ot'
the perisomatic skeleton of the Cystidea, it is advisable to treat a few of the best
known forms separately.

Cystocrinoidea.

Porocrinus is a form which only differs essentially from a simple Crinoid of the
order Inadunata in the presence of the pectinated rhombs.

Caryocrinus. — The calyx is almost pear-shaped, and is carried by a long stalk
perforated by a wide axial canal. At the edge of the calyx there are 6 to 13
thin, jointed, biserial arms, each provided with a furrow on the oral side. The
hexagonal dorsal cup, apart from two plates lying interradially in the circle
of the radials, consists exclusively of the already described plates of the apical
system (four infrabasals, six basals, six radials). The tegmen calycis is formed of a
great number of plates, and at their centre six orals, which show the characteristic
arrangement before described (Fig. 294, p. 332), completely cover the mouth. The
ambulacral furrows are not externally visible. An excentrically placed aperture, roofed
over by a pyramid formed of six triangular plates, is regarded as the anus. The
pectinated rhombs are found on all the plates of the dorsal cup, and on them alone.
The two pores of a double pore are connected by a canal on the inner side of the plate.

Echinoencrinus. — The almost egg-shaped calyx consists of the already described
plates of the apical system (four infrabasals, live basals, five radials) arranged in five
rays, and five other perisomatic plates in contact with the circle of the radials, and
partly pressing in between them. At the oral pole there is a depression, round which
short uniserial tentacles rise, and in the centre of which the star -shaped oral
aperture lies. The anal aperture has shifted across the equator of the calyx
apically, and lies to the right posteriorly above the basals. Three pectinated
rhombs are developed, whose position is shown in Fig. 296, p. 333. The calyx is
carried by a short thick stem, the ossicles of which are perforated by a wide canal.

Cystoblastus (Figs. 259, p. 312, and 295, p. 332) shows some resemblance to
the Blastoidea. The egg-shaped or spherical calyx consists of sixteen plates and the
ambulacra. Of these sixteen plates, which are arranged in four circles, fourteen
belong to the apical system (four infrabasals, five basals, and five radials). The
radials exactly resemble the radials (fork plates) of the Blastoidea. Each radial
clasps an ambulacrum between its two limbs. Between two neighbouring radials,
in four of the interradii, a lancet-shaped plate, which is keeled in the middle,
presses in, recalling the deltoid plates of the Blastoidea. In one interradius this
plate is wanting, so that here the neighbouring radials touch one another. At the
centre of the rosette formed by the five ambulacra lies the mouth, and from it a
furrow runs to each ambulacrum, passing down its whole length and dividing it into
two lateral halves. The principal groove of each ambulacrum gives off alternating
lateral furrows, which end in distinct pits (pores ? depressions for the reception of
pinnuke ?). At the base of the calyx there are two pectinated rhombs (cf. Fig.
295, p. 332). Further, the forks of the radials appear transversely striated by
numerous parallel pore clefts, and a similar striation also occurs on each side of the
rib or keel on the four deltoid plates. Perhaps every two neighbouring rows of pore
clefts, belonging, however, to different but adjoining plates, together formed a kind
of pectinated rhomb. A large aperture half way up the calyx is regarded as the
anal aperture, and a smaller, in an angle between two ambulacra, as the aperture of
the water vascular system. This is, however, uncertain. The arms are unknown,
VOL. II 2 C

386 COMPARATIVE ANATOMY CHAP.

and so is the stem, though the latter was present, since it is easy to distinguish a
depression at the apex marking its point of attachment.

Eucystidea.

Protocrinus (Fig. 260, p. 312).— The calyx is non-pedunculate, with somewhat
flattened apical side ; otherwise it is almost spherical. It consists of numerous
pentagonal or hexagonal bulging plates, each of which is provided with several
double pores. At the oral pole lies the oral aperture, from which five long ambu-
lacral furrows radiate, giving off, here and there, lateral furrows. At the end of each
lateral furrow there is a depression on a prominence. At these points, small anus
or pinnuliB were perhaps once articulated. The arnbulacral furrow and the mouth
are roofed over with covering plates. The anus lies excentrically in an interradius,
roofed over by a pyramid of valves. Between anus and mouth there is a small
third aperture. The stalked genus Glyptosphcerites is related to Protocrum.*.

Orocystis (Fig. 261, p. 313).— The almost egg-shaped body is tessellated with a
considerable, number of plates which are usually hexagonal, and are all provided
with pectinated rhombs. The pores are on prominences, arranged as in the figure.
A stem was present, but has never been met with attached to the body. On the
oral side of the body there are two principal apertures, the mouth and anus, lying
on chimney-like prominences, and there is also an additional third aperture. The
area around the mouth is never preserved intact ; probably the mouth was
surrounded by a few tentacles. In the genus Ecliinospluzrd, very common in
the Lower Silurian, the spherical test is formed by a large number of pentagonal
or hexagonal plates, all of which have pectinated rhombs. In each rhomb the pores
lying on the opposite sides of the suture, between the two plates, are connected
in pairs by canals. The oral aperture lies on a chimney-like or conical promi-
nence, surrounded by two to four long or short arms. At some distance from the
buccal cone lies the anus, covered by a pyramid of valves. Between the month
and the anus, but a little to one side, is a third smaller aperture. In Aristucysti*
there are two smaller apertures between the mouth and the anus, one of
which, near the anus, perhaps represents the genital aperture which, in other
Cystids, is possibly confluent with the anus. In Ascocystis, the body, which was
evidently richly plated, is prolonged like a tube ; at the pointed apical pole it was
attached by means of a stem ; at the oral side it is truncated, the oral disc being
surrounded by as many as twenty-live biserial, unbranched arms. The structure of
the disc surrounded by these arms has not yet been made out with certainty.

Mesites.— The body is spherical, and was probably stalked. The test" consists
of numerous plates provided with double pores, and showing no definite arrange-
ment. Five furrows run from the oral pole in meridians
towards the apical pole. Each furrow is covered by a double
row of contiguous, so-called ambulacral plates, and is thus
Fi.; 335 - Trans- converted into a closed canal. Between the consecutive (am-
verse section through blllacral) plates, pores lead into this canal, while on the plates
an ambulacrum of themselves circular areas can occasionally be made out, which
Mesites. have been regarded as points of attachment of pinnulse. A

groove runs along the middle line of each double row of plates ;

this is open along most of its length, but, at the oral pole, small plates form a
slanting roof over it.

We thus distinguish, in each radius, an external groove which runs upon the
ambulacral plates, and an internal canal, which runs below these plates and above
the perforated plates of the test (Fig. 335).

viii EGHINODERMATA— SPINES, ETC. 387

In an interradius on the oral side of the body, nearer one of the ambulacra than
the other, lies the anus, which can be closed by valves.

Mcfiites shows a certain similarity with the Palseechinoidea. By arbitrarily
assuming that an ambulacral vessel ran in the canal below the ambulacral plates.
and that ambulacral feet passed out through the pores between these plates, stress
was laid upon the agreement thereby established between the Echinoids and Mesites
in the position of the radial water vascular trunk on the inner side of the ambu-
lacral plates. But (1) it is quite uncertain that the ambulacral vessel really lay
in this canal and not in the outer channel, (2) the ambulacral feet, in Echinoids.
pass thro ugh the ambulacral plates and not between them as in Mesites, and (3) it is
not at all certain that the pores of Mesites really served for the passage of ambu-
lacral feet.

Agelacrinus (Fig. 262, p. 313). — The body resembles a more or less flat, round
disc, attached to a rirm object (e.g. the shell of a Brachiopod). The test is formed
of numerous, irregularly arranged scale-like plates, which more or less imbricate
with one another. In the middle of the free (oral) side of the disc lies the mouth,
covered with plates, from which five curved ambulacral furrows radiate, covered
with double rows of alternating plates. The plates of each double row form a
tunnel, raised above the level of the disc, and, between the plates, apertures may
occasionally be observed which are supposed to have served for the passage of
ambulacral feet. In one of the interradii, between two ambulacra which converge
so as to form a ring, lies the anus, arched over by a pyramid of valves.

Just as Mesites has been regarded as nearly related to the racial form of the
Eel duo idea, so Agelacrinus (with the related genus Edrioastcr) was said to be nearly
connected with that of the Asteroidra. But it appears hardly conceivable that the
almost rigid skeleton of the attached, sessile Agelacrinus could give rise to the
richly jointed and movable skeleton of the Asteroid. In the Asteroids, it is not
the ambulacral feet but the canals connecting them with the ampullae which pass
out between the ambulacral plates, and the radial water vascular trunks lie outside
of the latter. The double rows of covering plates in Agelacrinus cannot thus be
compared with the double rows of ambulacral plates in the Asteroid.

In conclusion, it may here lie remarked that structures resembling the
pectinated rhombs of the Cystidea occur in many fossil Crinoids and Echinoids.
These are parallel striae on the skeletal plates, which run transversely over the
suture dividing two neighbouring plates, and together form a rhomb -like figure.
In young and fossil Echinoids, it is principally the plates of the apical system
which are ornamented with such striated rhombs.

D. The Spines and their Derivatives — The Sphseridia and the

Pedieellarise.

1. The Spines.

The test of the Echinoidea, and the plated test of the Asteroidea and
Ophiuroidea carry large or small, variously shaped spines or processes,
which also vary in number and arrangement. The knowledge of
the structure, shape, size, and arrangement of these rigid processes
of the body (aeanthology) is of importance for classification. We
must here confine ourselves to the most important points, referring
the reader for further particulars to systematic works.

a. The Spines of the Eehinoidea, which we shall first consider

388 COMPARATIVE ANATOMY CHAP.

merely as parts of the skeleton, occur in all forms. They are found
in definite arrangement over the whole test, on both the ambulacral
and the interambulacral plates, but usually in greater numbers on the
latter than on the former.

The spines are usually slender and pointed, but may also (e.g. the principal
spines of certain Cidaridce) be club-, egg-, plate-, or oar-shaped, etc. ; or again in
other cases they may look like fine setee. The skeleton of the spine shows the
same microscopic lattice-like structure which characterises all parts of the Echiuo-
derm skeleton. Transverse and longitudinal sections of the spines reveal specific
developments of this structure, the lattice-work being closer in some cases and
opener in others, so that a careful examination of the structure of an isolated spine,
taking into account certain possibilities of error, may suffice to determine the species
of a specimen. The spines are mostly solid, less frequently hollow (e.g. in the
Scutcllidcc).

The spines are movably articulated with the test. Each spine
rises from a wart-like prominence of a test plate, which is called a
tubercle.

Large, strong spines rise from large tubercles, and small spines from small
tubercles, so that an examination of the tubercles on the test of an Echinoid which
has lost its spines leads to some conclusions as to the nature of its former spinous
covering. The tests of the Clypcastroida and Spatangoida, for instance, have very
small tubercles, with which the small, inconspicuous, seta-like spines of these orders
correspond. The regular Echinoidea have powerful spines and large tubercles. In
the Cidaroida, especially, by the side of numerous small tubercles carrying small
s] lines, there occur, in the interradii, a smaller number of remarkably large tubercles,
which carry either very long and strong or else shorter but very massive spines,
(Fig. 336).

Most spines are in some way ornamented, by ribs, thorns, etc.

In describing the various parts which can be made out in a
spine, and in that portion of the test plate which surrounds its base,
we shall select a principal spine of Dnrocidaris pap-illatu (Fig. 337).
The spine consists of a shaft and a socket, the latter articulating
with the tubercle of the test plate. The shaft thins away near the
socket to form a neck which, again, is separated from the socket
by a projecting circular ridge or cushion.

Each tubercle rises from a mound formed by the bulging of a
round, smooth area, the edge of which is surrounded by a circle of
smaller tubercles, which carry smaller spines and pedicellarice (Fig.
336).

The socket, where it is in contact with the tubercle, has a pit,
and a similar pit is found at the centre of the tubercle itself. Within
these two corresponding pits runs an axial band consisting of elastic
fibres, which fastens the spine to the tubercle, and passes at its two
ends into the organic substance of the spine and tubercle.

The base of the spine is surrounded by a double fibrous envelope.
The inner envelope consists of elastic fibres, the outer of muscle

VIII

EOHINODEBMA TA— SPINES, ETC.

389

fibres, which latter bring about the movements of the spine on the
tubercle. Both the muscular, and the elastic, fibres are attached on the
one hand to the socket (below the circular ridge or cushion), and on
the other to the area surrounding the tubercle. They pass into the
organic substance of the skeleton.

The spine is covered from tip to base (i.e. to the neck) by a very
hard and thick calcareous layer, the cortical layer, which, in the

Fin. 33C>. — Part of the
surface of the test of
Cidaris tribuloides, Ag.,

near the ambitus, to show
the tubercles and ambu-
lacral pores up. ia, Interam-
bulacral row of plates ; am,
ambulacral row of plates.

FIG. 337.— Large spine of a Cidaris, dia-
grammatic (essentially after Prouho). 1, Cor-
tical layer ; 2, middle layer ; 3, medulla ;
4, neck ; 5, integument ; 6, pedicle ; 7, axial
band ; 8, muscle ring ; 9, circular ganglion ;
10, ligamentous envelope ; 11, tubercle on
the test ; 12, test.

development of the test, is the last part deposited, and determines
the ornamentation of the spine.

At first the body integument covers the whole spine, its epithelium
being provided with cilia. But when the spine has attained its
definitive size, and the cortical layer has been formed, the integument
dies away on those parts which are covered by that layer, and is only
retained round the base of the spine.

At this base, about half way' up the muscular envelope, under the
surface of the epithelium, lies a nerve ring with scattered ganglion
cells ; this runs right round the base of the spine, and innervates its
muscles.

390

COMPA RA TI VE ANA TOMY

CHAP.

The structure of all Ecliiuoid spines resembles that just described, except that
the pits in the socket (or acetabulum) and on the tubercle are usually wanting,
and with them also the axial ligament.

The small spines of the Cidaroida are protective. They surround the anal
aperture, the genital apertures, and the pores of the radials (or ocular plates) ; on the

interambulacra, they surround the bases of the prin-
cipal spines like a circular palisade, and on the am-
bulacra they are arranged in two longitudinal rows.
They can be raised, and inclined towards one another
over the part to be protected. The smaller spines
have no cortical layer and no nerve ring at their
bases. They are always covered by the ciliated
integument, which at the tip of the spine carries
sensory (tactile) hairs. Each small spine carries at
its base, on the side turned away from the part to lie
protected, a whitish, transparent, ampulla - shaped
swelling, which seems to be caused by the presence of
glandular cells in the epithelium. The secretion of
this glandular cushion may perhaps be poisonous.

In Ccntrostcphani/.s Ivngispinus, round the anus,
certain short spines of a lilac colour occur ; in the
living animal, these constantly rotate, the tips of the
spines describing a circle. In the epithelium of these
spines there are sensory prominences, and at their
bases the characteristic circular ganglion. The fibres
of the muscular envelope are transversely striated.

In Podocidaris, immovable spines without any
articulation are found, principally on the apical side
of the shell.

The poisonous spines of Asthenosoma urens
(EchiiiotJinrid). This Echinoid is much feared by
fishermen and divers, on account of the acute pain
produced by contact with its body. Spines, whose
ends are swollen into shiny blue heads (Fig. 338),
seem to form the principal part of its poison apparatus.
These poison spines are arranged in regular bands on
FIG. 338.— Spine with poison the interambulacra, but in the larva are found scattered
head of Asthenosoma urens oyer Qther ag weU_ The axig of the ine is

(after P. and F. Sarasm), diagram-

matic. l, The tip of the spine ; occupied by a hollow calcareous rod with an extremely
2, poison sac ; 3, epithelium of fine point ; this rod, throughout the greater part of
the poison head ; 4, muscles of the its length, is perforated by pores arranged in longi-
same ; 5, lower terminal fascia of tudiual rows . at its filie tip> however, there are only
the poison head, penetrating the ., „,, ni.ji.-i

spine ; 6, longitudinal rows of a few sma11 Pores or eyes" The SWollen liead wlllch
pores in the shaft (7) of the spine, surrounds the tip of the spine contains a somewhat

large poison sac, with an aperture at its tip, through

which the spine may protrude. The epithelium lining the sac passes, at the
aperture, into the outer epithelium of the head. The poison sac and the part of
the spine which runs through it are filled with a clear fluid with floating vesicles
(cells and remains of cells), yielded by the epithelium of the sac. The sac and its
envelope of connective tissue are surrounded by a powerful muscular capsule, most
of whose fibres are attached on the one hand to the sac, and on the other to the part
of the spine lying below it. The contraction of these muscles causes the sharp tip
of the spine to protrude through the aperture of the sac. Perhaps, at the same

VIII

ECHINODERMA TA— SPINES, ETC.

391

time, the poison is squeezed into the spine through the lower pores in that part of
it which lies within the sac, and is squirted out through the few pores or eyes at
its tip.

On the fascicles of the Spatanyoida, whose course has already been described
(p. 349), there are exceedingly numerous very small, granular tubercles carrying small
seta-like spines, thickened at the tip ; these are sometimes articulated, sometimes
immovable. Such clavulse are covered by a ciliated integument, which very
probably contains sensory cells.

b. The Spines of the Asteroidea. — The Asteroid body is also
usually covered with spines and papillae. The form and arrangement
of these vary so greatly that they cannot here be more fully described.
We must refer the reader to the principal systematic works for details.
Their finer structure is

almost entirely unknown,
and we have hardly any
knowledge of the positions
of the sensory organs and
glands which almost cer-
tainly occur.

The spines are often firmly
connected with the skeletal
plates of the body wall, from
which they rise. Spines occur
most constantly at the edges of
the ambulacral furrows, border-
ing them like a palisade. They
are not infrequently movable ;
they can be erected, and inclined
over the furrow for its protection
(Fig. 243, p. 298).

Many Phancrozonia, and
especially the Astropectinidcc,
are characterised by short cal-
careous pillars rising from the integument, which, on their terminal flat surfaces,
caiTV a usually circular group of small, thickly crowded spines, prominences, or
papilke. These structures are called paxillse (Fig. 309, p. 351).

c. Spines of the Ophiuridse. — In the Ophiuridce, it is principally or
exclusively the lateral shields which carry spines, in a manner already
described (p. 355).

These spines are mostly large, slender, and pointed, and are occasionally provided
with thorns. In the genera Ophiomastix, Astroschciiia, and Opliiocrcas club-shaped
spines occur, together with the ordinary kinds. Over the ends of these spines the
epithelium is thickened, and contains glandular and sensory cells. In Ophiopteron
elcgans, numerous small spines of peculiar structure are found on the dorsal side of the
disc. A short stem divides into six long pointed branches, which are connected
by a thin, soft membrane in such a manner as to form a kind of funnel. The whole
structure somewhat recalls an umbrella turned inside out. In the same species,
each lateral shield carries, besides a hook and a thorny spine, ten long, slender

FIG. 33!>. — Three brachial joints of Ophiopteron
elegans, from the middle part of the arm, lower side (after
Ludwig). Is, Ventral shields ; te, tentacle ; ss, lateral
shields; 1, hook; 2, thorny spine; 3, supporting rods of
the tins.

392 COMPARATIVE ANATOMY CHAP.

spines arranged in a row, which runs up from the ventral to the dorsal side of the
arm ; these spines are connected by a thin, transparent membrane in such a way as
to form a sort of fin (Fig. 339). On the first three free joints of the arm, the fin of
one side of the arm passes into that of the other side dorsally. We are justified
in assuming that the animal is able to swim by means of these large fins on
the arms.

The genera OphiothoHa and Ophiohelus are distinguished by peculiar umbrella-
shaped spines. A stem with a swollen, button -like base, articulating Avith a
tubercle, carries at its tip a circle of recurved spines, which, during life, are covered
by a common integument. These are found either in groups near the base of the
ordinary brachial spines, as in OphiothoHa, where they first appear at some distance
from the disc, or else replacing the ordinary spines near the end of the arm, as in
Ophiohchts.

Function of the spines.— The fact that the spines serve princi-
pally for the protection of the body is at once evident, especially when
they are provided with poison glands.

In response to stimuli, the spines become erect. In Diadcma sctosum, which is
very sensitive to light, the long spines turn threateningly towards a hand which
approaches them from any side. The spines of most Echinoids further serve for
locomotion, moving in a co-ordinated manner. This has been directly proved in
the Cidaridcc, Arbacia, Echinus, and Spatangus, etc. In the first of these forms,
the long (principal) spines are indeed the chief, or the only, locomotory organs, and
are used as stilts. Many Echinoids, e.g. Dorocidaris, Arbacia, Spatangus, if laid.
on their backs, can turn themselves over again by the help of their spines.

It has also been proved that the spines may serve for seizing prey and for
forwarding it to the mouth. Several spines incline towards the prey, seize it with
their tips, and pass it on to the next group in the oral direction, and so on towards
the mouth. Compare with this the rise of pedicellarise in Asteroids, p. 394.

2. Modified Spines.

«. The Sphseridia of the Eehinoidea. — These are small spherical
or ellipsoidal bodies, which, by means of a short stalk, articulate with
a prominence of the test, and are inclined sometimes in one direction,
sometimes in another. They either project freely, or else rise from the
base of a pit-like depression of the test (Fig. 340). This depression
may more or less completely close over the sphseridium. We are
here reminded of the various forms of acoustic tentacles in the
Medusae, which sometimes rise freely from the body, sometimes from
the base of a pit, sometimes on the walls of closed vesicles, which latter
come into existence through the concrescence of the edge of the pit
above the tentacle. Here, however, we have to do not with tentacles,
but evidently with modified spines.

Sphseridia occur in all Eehinoidea except the Cidaroida. They are found only
on the ambulacra, and always on the peristomal plates, although in many forms
they are not limited to these, the area in which they occur stretching out in the
direction of the ambitus, or even beyond it. The number and arrangement of the
spha?ridia vary greatly in different groups.

VIII

ECHINODERMA TA—SPH^RIDIA

,393

Structure of the Sphseridia (Fig. 341). — The sphseridui consist
(1) of a very firm and hard transparent calcareous sphere, which is
concentrically laminated, and does not show the lattice-like perforated
structure of the rest of the skeleton, and (2) of the calcareous stem,
which is perforated like a sponge, and is generally continued into the
interior of the sphere. The calcareous sphere perhaps answers to the
cortical layer of a large spine of a Cidaroid (cf. Fig. 337). Not infre-
quently the head is traversed by a canal which opens at its free end.

The sphseridium is covered by a ciliated epithelium which is often
pigmented ; the waving cilia are very long at the base of the stem,
but gradually diminish in length towards the head. The sphseridia,
like the spines, are surrounded at the base where they articulate with
the tubercle, by a muscular envelope and by a, circular ganglion, the

FIG. 340. — Portion of an ambulacrum
bordering the peristome in Echinocidaris
nigra, Mol. (after Loven), magnified. 1, Splwri-
dium iu its niche ; 2, ambulacral double pore ;
3, edge of the peristome.

FIG. 341.— Longitudinal
section through a sphasri-
dium, diagrammatic. 1, Cal-
careous mass of the sph:rri'.lium :
2, epithelium ; 3, calcareous
stem of lattice-like structure ;
4, muscle envelope ; 5, circular
ganglion ; (i, tubercle ; 7, test.

latter lying within the epithelium, which is here specially thickened.
The hair-like cells of this circular thickening of the epithelium are
probably for the greater part sensory.

The sphaeridia have always been claimed as sensory organs, and,
on account of their usual position near the mouth, as gustatory or
olfactory organs. They have also been thought to be auditory, or
organs for the appreciation of the movements of the water. They
also remind us of organs adapted for appreciating the position of the
body in the water.

h. The Pedieellarise. — These are small seizing organs which rise
from the integument. They occur in all Echinoidea, most Asteroidea,
and a few Ophiuroidea in very varying number and arrangement, and
iti many different forms, between which, again, there are transition
forms. They must be considered as modified spines, or groups of
spines. In one and the same species various forms of pedicellaria?,

394 COMPARATIVE ANATOMY CHAP.

definitely arranged, may occur. It is very probable that many of the
different forms of pedicellariae, within certain divisions, have developed
independently out of spines.

1. The pedicellarise in the Ophiuroidea. — In Trichastcr clegans, from about the
thirty-sixth tentacle pore of an arm onwards, the two tentacle papillae are replaced
on the adoral side of each pore by two hooks movably articulated on a stem. This
stem also is articulated with a ventral lateral process of the corresponding brachial
vertebral ossicle. The skeleton of this apparatus consists of three pieces, belonging
to the stem and to the two diverging hooks. The hooks do not move against one
another, the planes of their movement being nearly parallel. On one side a flexor,
and on the other an extensor muscle connects each hook with the stem. In
Astrophyton also, similar pedicellarite are found, and in Opliiothrix frag His the end
of the arm is beset with movable hooks provided with flexor and extensor muscles.
Similar hooks occur, further, on the lateral shields of the arms in certain species of
Gorgonocephalus.

2. The pedicellarise of the Asteroidea (Fig. 342). — In some groups, e.g. the
Astcrinidcc, Solastcridcc, and Plerastcridcc, the pedicellarire are altogether wanting :
in the Astropectinidce they are only very rarely found.

In the simplest cases, groups of small spines may function as pedicellarise. The
spines of such a group are movably connected with the body, and may be arranged
either in two opposite rows of four to five spines each, these rows approximating or
diverging ; or else at definite points of the body three or four spines stand close
together, forming, when they incline towards one another, a three- or four-sided
pyramid. Two spines even may form a group. For instance, on the dorsal surface
of Aster in a, yibbosa, spines are found sometimes isolated, sometimes united in larger
or smaller groups. Among these groups there are couples connected at the base by
a transverse muscle, and such spines can move towards one another more energetic-
ally than those of the other groups (Fig. 342, A to F).

In the above cases, we have to a great extent to do with commencing or rudi-
mentary pedicellarise, and we recognise, in the larger and smaller groups of spines,
the material out of which pedicellarife with two, three, or four forceps may be
developed. (Cf. also what has been remarked on p. 392 on the spines of the Echinoidea
as organs for the seizing and conveying of prey to the mouth, and p. 390 on the
smaller spines of the Cidaroida.)

The true pedicellarise of the Asteroidea usually have two blades or valves, less
frequently three. Stalked and sessile pedicellarise may be distinguished.

a. Sessile pedieellarise (Fig. 342, G). — The two blades rise
directly from the integument. Each consists of a calcareous piece
determining its shape, which may be long or short, broad or
narrow, pointed or blunt, flat or spoon-like. The two skeletal pieces
are directly articulated with a skeletal plate of the integument. In
Gymnasteria carwi/em, for example, numerous double-bladed pedi-
cellariae rise at the edge of the ambulacral furrow. The two blades
are connected at their bases in a manner illustrated in the figure, by a
transverse muscle, the adductor muscle. Further, each blade at its
outer side (i.e. at the side turned away from the axis of the pedicel-
laria) is also connected with the subjacent calcareous plate of the
integument by an opening muscle (abductor). The bases of the

VIII

ECHINODERMA TA —PEDICELLARIA^

395

pedicellate are further attached by a strong elastic fibrous band to
this"same plate.

li. Stalked pedieellarise (Fig. 342, H, K). — Each pedicellaria
rises from a short, soft stalk ; the blades, of which there may be two

R B

K

FIG. 342.— Pedicellariae of Asteroids. A, B, C, D, E, F, Pseudo- or commencing pedicellarkc of
various species. G, Sessile pedicellaria from the edge of the ambulacral furrow of Gymnctiti'fiti
en i- in ifera (after Cue'not). H, Stalked straight pedicellaria diagrammatised (after Cuenot). J,
Basal portion of a stalked crossed pedicellaria of Afti'fCMtnthwn rnbcns (after Perrier). K, A
similar pedicellaria of Asteracanthion glacialis (after Cuenot). 1, Calcareous blade of the forceps ;
•2, basal piece ; 3, occlusor muscle ; 3i, axial muscles of the blades ; 4, opening muscle ; 5, axial
band ; 0, epithelium ; 7, body wall ; 8, stem.

or three, articulate with a basal skeletal piece. The double-blade* 1
(didaetyle) pedicellarise are either straight (forfteiform) or crossed
(foreipiform). Both kinds may be found in one and the same
animal.

We select for description Asterias (glacialis), one of the Asteroid*
most richly provided with pedicellarise, whose arrangement is specially
interesting.

A. glacialis has three kinds of pedicellarise, straight, crossed, and
three-bladed.

The crossed pedicellarise are found in very great numbers, thickly crowded
together on a soft cushion, which surrounds the base of the spines, and into which the
latter can be withdrawn (Fig. 344).

396

COMPAEA TIVE ANA TOM Y

CHAP.

The straight pedicellarise are far less numerous, and are found scattered over the
integument either singly or in groups.

The three -bladed pedicellate are always found entirely isolated, and may be
altogether wanting in some individuals.

Structure of the straight pedicellarise (Fig. 342, H). — Each of the two blades
consists of a hollow toothed skeletal piece, which articulates with a common basal

8

' 2

^~

3

*--^

L

' i

~ (.
y

7 (.
//

? t

iL

i

FIG. 343. -A portion of an arm of Asterias stichantha, Sladen, from the lower side (after
Sladen). 1, 2, 3, 4, The four longitudinal rows of ambulacra! feet ; 5, forticiform pedicellaria< ; li,
adambulacral spines ; 7, papulae ; 8, inframarginal spines ; 9, forcipiform pedicellarut at the
outer bases of these latter.

piece. Two muscles serve for opening the pedicellaria, the outer side of each blade
being attached by a muscle to the basal piece. The blades are closed by means
of two muscles which run from the inner sides of their bases to the basal piece,

FIG. 344.— Asterias (Stolasterias) volsellata.— Adainbulaeral plates and neighbouring portion
of the oral integument of an arm. /, Straight ; fc, crossed pedicellarke on a cushion at the base of
:t large spine (ac) ; n.ci, smaller spine (after Sladen).

and also perhaps by means of two muscles which, lying for the greater part
within the calcareous blades, run from their tips to the basal piece. Each pedi-
cellaria is surrounded by a layer of connective tissue, and covered by body epithelium,
in which glandular cells are scattered.

Structure of the crossed pedicellarise (Fig. 342, K). — A crossed pedicellaria
is not unlike a forceps with short handles. It also consists of three pieces, the

VIII

ECHINODERMA TA —PEDICELLA RIM

397

two limbs of the forceps, and an intermediate or basal piece, upon which the
blades move. Each limb of the forceps consists of the blade and the stalk or
handle. The two limbs cross at the two sides of the intermediate piece like the two
parts of a pair of pincers or scissors. When the two handles are approximated, the
pincers close, when they are drawn apart, they open. The opening and closing of
the pedicellarue is effected by means of six muscles. Two small muscles, running
from the outer side of the bases of the blades to the basal piece, by their contraction,
open the forceps ; while two pairs of muscles close it. One of these pairs runs
within the blades to the basal piece, and the two muscles of the other pair run
transversely from the handles of the two limbs to the basal or intermediate piece.
An axial strand of elastic fibres run from the stalk of the pedicellaria to the base

FIG. a-i.O.-Pedicellariaa of Echinoids. A, Four-bladed pedicellaria of Schizaster canali-
ferus (after Koehler). B, Glandular pedicellaria with glandular sacs on the stalk,
Sphaerechinus granularis. C, Longitudinal section through a decalcified tridactyle
pedicellaria of Centrostephanus longispinus (alter Hamann). 1, Adductor muscle ; •_', nrrvr ;
3, elastic column ; 4, calcareous rod ; 5, longitudinal muscle fibres.

of the forceps. This strand divides into two branches, which embrace its handles.
The fibrous strands of the individual pedicellarise penetrate the cushion, which sur-
rounds the base of the spine (Fig. 344), and finally break up into fibres which become
closely matted together. The whole cushion consists of thickly interwoven fibres of
connective tissue and muscle. Muscle fibres run down from the calcareous piece of
the spine into the cushion, in which they are lost. By means of these muscles, the
cushion can be drawn up the spine, like a sort of sheath. The pedicellaria:', like
the cushion from which they rise, are covered by a markedly glandular epithelium.

The three-bladed pedicellarire, apart from the number of their blades, agree in
structure with the straight two-bladed pedicellaria3.

3. The pedicellarise of the Echinoidea (Figs. 345 and 346). — Pedicellaria' occur
in all Echinoids on the integument, between the spines, and in one and the same
species two or more forms of them may be found. The special arrangement of the
various forms of pedicellarite on the body (whether occurring on the ambulacral or
on the interambulacral areas, and whether orally or apically), their distribution,
number, and very varied form cannot here be described in detail, but must be
sought for in systematic works.

398

GOMPAEA TI VE A NA TOMY

CHAP.

The pedicellari;e of the Echinoiclea are always stalked, and three-, less frequently
two-, or four-, bladed. Two principal forms may be distinguished : seizing pedi-

cellariae (Fig. 345 A and C) and glandular pedi
cellarise (Fig. 345 B and Fig. 346).

a. The seizing pedicellariae. — The form of tin-
blades varies greatly in details. They are some-
times long and slender (p. tridactylre, tetradactylse),
sometimes spoon-shaped and toothed (p. ophioce-
phalre, sen buccales, seu triphyllae), or in other
cases broadened out like leaves (p. trifoliate .
Each blade is always^ supported by a skeletal
piece, which determines its general shape and the
^pi'cial form of its teeth, hooks, etc. The stalk
also is always supported by an axial calcareous
rod, which sometimes penetrates the whole of its
basal half (p. tridactyLe), sometimes only reaches
a short way into the base of the stalk.

The tridactyle pedicellariae of Centroste-
phanus longispinus (Fig. 345 C) will serve to
illustrate the structure of the seizing pedicellariae.
The three slender blades are connected at their
bases, and on the sides turned to the axis of the
whole forceps, by three transverse adductor
muscles, each of which is attached on the inner
(axial) sides of two neighbouring blades. The
three muscles together form a triangle. These
adductor muscles are counteracted by opening
muscles, which run down on the outer sides of the
bases of the blades longitudinally. A nerve enters
each blade, running towards its tip, and innervat-
ing the musculature and epithelial sensory cells.
The inner surface of each blade is ciliated. Within
the stalk, the supporting calcareous rod reaches
Fio. 340.— Organisation of a glan- only half way up, ending in a knob. The con-
dular pedicellaria of Sphserechinus tinuation of the calcareous rod is formed by an
granularis, section. 1, Distal tactile elagtic m wMch consists of modified connective
prominence ; 2, aperture of the gland

of the forceps; 3, proximal tactile tissue, and is enveloped m a sheath of longitudinal
prominence; 4, adductor muscle; muscle til ires. This arrangement makes it possible
5, ski-letal piece of the forceps; G, epi- for the distal portion of the stalk with the head
theliu.u of th, forceps; 7, cavity of to lieml in various directions, and even to bend
the gland of the forceps ; 8, epithelium . ,, , , ., , . ,. ,,,,

of the same; 9, muscle layer of the "ght back .ul'on the basal Portlon- AUien tlie
forceps gland ; 10 and 11, opening muscles which bring about such movement are
muscles ; 12, nerve ; 13, calcareous rod relaxed, the distal part resumes the upright

by means of the elastic pillar it

in the stalk ; 14, aperture of the stalk
irland (lii) ; l.'j, epithelium of the gland.
(The distal tactile prominence here
represented is wanting in this species.)

position

contains.

The adductor muscles of these pedicellarise

consist of transversely striated muscle fibres ;
consequently these tridactyle pedicellariae are very active seizing organs.

b. The glandular pedicellarise have been carefully investigated, up to the present
time, onlyin a small number of Echinoids (Spkcereckinus ijranidaris, Eclinm* aciitus,
E. melo, Dorocidaris papiUata, Strninjiifin; •ii/i-ntim /iriihis, Ecliinocardium flavcscens),
but it is probable that in time they will be found more widely distributed. In
general structure they resemble tin- ordinary seizing pedicellarise possessing three

vin ECHINODERMATA—PEDICELLARLi: 399

seizing blades, which are opened and closed by special muscles, but the fibres of the
adductor muscles are not transversely striated. In the stalk, the axial calcareous rod
is continued as far as the three-bladed head, an arrangement which greatly decreases
the mobility of this kind of pedicellaria.

The most distinctive characteristic of these pedicellarise, however, is the
presence of a large glandular sac in each blade. This glandular sac, which, as is
shown by recent discoveries, consists of two fused sacs, causes each blade to be pear-
shaped. It is covered by a thick glandular epithelium, and has a muscular wall of
its own, in which the (smooth) fibres run in circular layers. This muscular wall, no
doubt, serves for pressing out the slimy, and probably poisonous, secretion, through
the aperture which lies near the tip of the blade. This aperture appears in most
cases to lie on the outer side of the blade.

At the base of each blade, on its inner side, the epithelium is thickened to form
a tactile prominence or cushion, which (besides cilia) carries immovable sensory
hairs. In Echinus acutus, besides the basal, or lower, tactile prominence, there is,
on each blade, a distal or upper prominence, which also lies on the inner side of the
blade (Fig. 346).

Numerous nerves pass from the stalk of the pedicellaria into its head and its
blades, so as to innervate the muscular and the sensory cells.

In a few Echinoids, glands also occur on the stalk of the pedicellaria ; such
glands are specially strongly developed in SphcerecMnus yranidarla. These
glands, three in number, encircle the stalk of the glandular pedicellari;e (p. gemmi-
forrnes), about half way up. Each gland is a large vesicle with an aperture, through
which, on stimulation, a slimy secretion is discharged. The wall of the vesicle con-
sists of glandular epithelium within a muscular layer. The three glands cause large
vesicular swellings on the stalk of the pedicellarhe on which they occur ; they are
covered by unditt'erentiated outer body epithelium.

If we imagine that, in such pedicellarise provided with stalk glands, the distal
portion of the stalk above these glands degenerated, or was no more developed, we
should have the form of pedicellaria which is called p. globifer. Such globifers,
occasionally still provided with rudimentary seizing forceps, have been discovered in
Centrostephanus loiujixjiimis and SjiJtn /•ic/tiiius granularis, side by side with ordinary
pedicellarife. They are capable of pendulous movements.

The function of the pedicellarise has not yet been satisfactorily decided. The
view that, in Echinoids, they play some part in locomotion, has recently been
decidedly opposed, and it has been maintained that Echinoids move exclusively by
means of their ambulacral feet and spines. It has further been asserted that the
pedicellarise lay hold of foreign objects, algse, etc., and hold them fast on the upper
side of the body in order to hide it, but this view also has been opposed, the function
therein ascribed to pedicellarife being claimed for the tube-feet. Such a function could,
in any case, only be accessory. Another view is that the pedicellariie serve for the
holding of prey, and for carrying it to the mouth. In the Asteroidea, however,
which take the food in large pieces (Fish, Crabs, Mussels, Snails, Echinoids, etc.),
they could not well play this part.

The most probable view is that the pedicellarife are protective organs, and fulfil
the function of cleaning the spine-covered body. They clear away foreign bodies.
Small animals which come into contact with the body are seized, and enveloped in
the slimy secretion of the epithelium, or in the possibly poisonous secretion of the
specialised glands of the pedicellario?, and held until they are dead, and then
" thrown overboard." In this way Echinoids and Asteroids may protect themselves
from animal and vegetable growths, parasitic or otherwise. This would explain the
astonishing cleanness of most members of this group in spite of their spinous

400 COMPARATIVE ANATOMY CHAP.

E. The Masticatory Apparatus of the Eehinoidea.
(Aristotle's Lantern.)

In all Eehinoidea, with the exception of the Spatangoida, and per-
haps of a few Holectypoida, the mouth, which lies at the centre of the
peristomal area, is armed with five hard and pointed, interradially
arranged teeth. These teeth are approximated or moved apart by
means of a complicated masticatory or jaw apparatus lying within
the test, and resting on the peristome. This apparatus is known as
the lantern of Aristotle, and is of considerable size ; it is covered on
all sides by a closely applied integument, the lantern membrane, a
continuation of the peritoneum. The spaces within the masticatory
apparatus are completely separated by this membrane from the spacious
body cavity within the test.

The masticatory apparatus resembles a pentagonal pyramid, the
base of which is directed upwards, i.e. projects into the cavity of the
test, while the tip, formed by the five teeth, lies in the mouth. Its
axis is traversed by the oesophagus. It consists essentially of skeletal
pieces, muscles, and ligaments.

a. The skeleton of the masticatory apparatus (Fig. 347) is
composed of twenty-five pieces (including the teeth), radially grouped
around the oesophagus ; some of these pieces have received very un-
suitable names. There are five teeth, five pairs of jaws (alveoli),
five "sickles" (falees), and five radii or rotulse. The "sickles"
may be named intermediate plates, and each pair of jaws forms a
" pyramid."

The principal part of the framework of the masticatory apparatus is formed by
the five interradially placed pairs of jaws. These determine the conical or pyramidal
form of the whole framework. The two pieces of each pair are firmly connected with
one another on the outer side of the framework by a vertical interradial suture, and
together form a hollow triangular pyramid, the fifth part of the whole pyramidal
framework. Each single pyramid thus has one outer and two lateral surfaces. The
five single pyramids are in contact with one another along these lateral surfaces,
which lie radially to the axis of the whole framework. The edges along which the
lateral surfaces come in contact are the axial edges, i.e. those turned towards the oeso-
phagus. The suture, which divides each single pyramid into two halves or jaws, runs
down the outer surface, exactly halving it. The walls of each single hollow pyramid
are, however, not complete : (1) the two lateral surfaces do not quite meet along
their inner edges, but there is a slit left between them (Fig. 347, E) ; (2) the basal
wall (that turned upward) is wanting ; when the soft parts are removed an aperture
is found here, the foramen basale, which leads down into the cavity of the pyramid ;
(3) a large incision (foramen externum) is found, at the base of the outer wall, and
is either confluent with the foramen basale or is separated from the latter by an
;irch, the arcus.

The single pyramids (or pairs of jaws) are the supports and carriers of the teeth.
Each tooth is a long, slender, and hard skeletal piece, curved, so that its convex side
faces outwards ; it traverses the cavity of the pyramid, and projects beyond it at both

viir ECHINODERMATA— ARISTOTLE'S LANTERN 401

ends. The lower end, which projects beyond the tip of the pyramid, is short and
pointed, and forms the externally visible part of the tooth lying in the mouth. The
upper end, which is directed aborally, is called the root of the tooth, and projects
considerably beyond the foramen basale, it is usually coiled inwards (towards the
axis of the masticatory framework). The growth of the tooth no doubt takes place
principally at this root end. On its inner side, the tooth usually has a longitudinal
ridge, the carina, and on its outer side is firmly attached to the outer wall of the

PIG. 347.— Masticatory apparatus of an Echinus, original. A, In profile. B, From the apically
directed basal side. C, External view of a single pyramid. D, Side view of the same. E, Internal
view of the same. F, Tooth. 1, Arcus ; 2, intermediate plate ; 3, freely projecting portion
of the teeth ; 4, median portion of a tooth ; 5, upper portion of the same ; 6, the limbs of a forked
radius (7) ; 8, single pyramid in situ.

pyramid which it traverses, in such a way that it cannot move by itself, but only
with its pyramid.

The fine structure of the teeth differs essentially from that of the other skeletal
pieces of the body (cf. on this subject the special treatises mentioned in the Biblio-
graphy).

The intermediate plates are five more or less flat, oblong, skeletal masses lying
on the base of the masticatory apparatus, like the spokes of a wheel round its central
axis. Each of these intermediate plates rests on the bases of the two contiguous
lateral walls of two pyramids [or pairs of jaws], and therefore between two foramina
basalia.

Finally, lying apically upon these falces, are the five forked radii, which are also
arranged like the spokes of a wheel. Each radius consists of a slender central
stalk, and of two peripheral diverging prongs, and each is bent downwards in such a
VOL. II 2 D

402

COMPARATIVE ANATOMY

CHAP.

way that its prongs point down towards the peristome over the edge of the base of
the pyramid (Fig. 348, 5).

b. Muscles and ligaments of the masticatory apparatus (Fig. 348).— We must
here refer back to what has been said of the perignathous apophysial ring, for the
masticatory apparatus and the apophysial ring are. physiologically, closely connected.
The most important muscles and bands of the masticatory apparatus connect its
component pieces with the apophysial ring, and the latter must be regarded merely
as a folding inwards of the edge of the peristome, which has come into existence

im.

9

Fir.. 348.— Masticatory apparatus of an Echinoid (Toxopneustes) in its natural position at the
centre of the oral side of the shell, which has been broken off, original. 1, Root of the tooth ; 2,
intestine ; 3, accessory intestine (?) ; 4, axial sinus with stone canal ; 5, forked radii ; G, arcus of a
single pyramid ; 7, muscles of the forked radii ; S, perignathous apophysis (auricula) ; 0, ligaments of
the forked radii ; 10, adductor muscles of the teeth ; 11, opening muscles of the teeth ; 12, radial
canal of the water vascular system ; 13, ampulla; ; im, interambulaerum ; am, ambulacrum. The
delicate transparent lantern membrane which covers the whole of the masticatory apparatus is
not represented.

for the insertion of the masticatory muscles. The two apparatus are either absent or
present simultaneously.

Round the masticatory apparatus, ten thin ligaments (9) connect the forked radii
with the interradial apophyses of the perignathous girdle. The two bands which
belong to each fork are attached to the prongs, and continue their lines down-
wards to the apophysial girdle ; they are inserted into the two neighbouring inter-
radial apophyses near the interradial sutures.

The two bands proceeding from each radial fork thus diverge downwards, and
the two proceeding from each interradial apophysis of the perignathous ring diverge
upwards.

These bands appear merely to serve for the attachment of the masticatory appa-

vin ECHINODERMATA— CALCAREOUS RING 403

ratus, and for its maintenance in the upright position over the oral area. Further
investigations, however, must decide whether the bands consist only of elastic fibres,
or whether muscle fibres also occur in them.

The adductor muscles of the teeth (musculi adductores dentium, 10).— These
are present in five interradial pairs ; they are strongly developed as broad bands.
The two muscles of a pair are attached, above, along the outer edge of the arcus of the
pair of jaws (pyramid) to which they belong ; and below, along nearly the whole
length of the corresponding interambulacral apophysis of the peiignathous ring. If
these muscles contract, the upper ends of the pair of jaws (or pyramids) are drawn
outwards and downwards, forcing the lower ends, with the teeth, inwards, i.e.
towards the centre of the mouth. In other words, the externally visible pointed
lower ends of the teeth are pressed together.

The opening muscles of the teeth (musculi abductores dentium sive dilatatores
oris, 11). — These are five radially arranged pairs of muscles, which run horizontally.
The two muscles of each pair are attached on the one side to the inner surface of the
ambulacral apophyses (auricula?), and, on the other, to the halves of the jaws nearest
them, close to the ends which point downwards. These muscles counteract the
adductor muscles ; when they contract, the lower ends of the five pairs of jaws, and
with them the tips of the teeth, are moved, centrifugally, towards the auriculae. The
teeth move apart, and the mouth opens.

The intermediate jaw muscles (musculi intermaxillares) connect the apposed
lateral surfaces of the five pyramids with one another. The five pyramids close
firmly together, when these muscles, which together act like a kind of sphincter,
contract.

The muscles of the forked radii (7) lie on the upturned base of the masticatory
apparatus, forming together a pentagonal ring by connecting the five handles of the
forks for about half their length. As to the function of these muscles, we can
only imagine that they depress the whole masticatory apparatus by their contraction,
and thus cause the oral integument to project conically, especially if the adductor
muscles of the teeth contract at the same time. It is well known that Echinoids are
assisted in locomotion by the bulging forward of the tooth-carrying portion of the
oral area, which is supported by the masticatory apparatus.

In the Clijpcastroida, the frequently asymmetrical masticatory apparatus is more
or less flattened, usually indeed quite flat. The teeth are not vertical, but slope
towards one another quite obliquely, or are even arranged horizontally. The radii
are wanting, and the intermediate plates are rudimentary.

F. The Calcareous Ring of the Holothurioidea.

In the Holothurioidea, the oesophagus is surrounded by a circle of
ten calcareous skeletal pieces (Fig. 349, 3 and 13), five of which are
radial and the other five interradial. This calcareous ring protects
the nerve ring at its inner side. For a certain distance it supports
the radial water vascular trunks and the tentacular vessels, and may
indeed be regarded as the inner skeleton of the oral region of the
body. The five longitudinal muscles or pairs of muscles of the body,
and, where such are present, the five retractor muscles of the oral
region, are attached to this ring, i.e. to its radial portions. The
calcareous ring is altogether wanting in the remarkable free-swimming
form Pelagofhwria (Fig. 224, p. 286).

404

COMPARATIVE ANATOMY

CHAP.

The form and size of the calcareous ring and its separate parts vary greatly. The
radials are often lengthened backwards (apically) into two prongs of varying length,
between which the radial water vascular trunks run.

It not infrequently happens that the separatejparts become partially or altogether
broken up into single pieces, which are connected together like a mosaic.

The number of pieces in the ring may increase or decrease. Where there are
more or fewer than ten pieces, it is always the interradials which either increase or

diminish in number. This is comprehen-
sible when we remember that the longi-
tudinal muscles of the body are attached
to the radials.

The interradial portions are wanting in
species of the genera Phyltoporus, Cucum-
aria, and Trochostoma, and in many Elasi-
PIH/II, especially in the whole family of the
Elpidiidae.

More than ten pieces are found in many
Syncqitidce, viz. in nearly all those forms
which possess more than ten tentacles.
The number of extra interradials then
usually corresponds with that of the super-
numerary tentacles.

Six - rayed specimens] of Cucumaria
Planci have been described, whose cal-
careous ring consists of six radials and six
interradials.

The ring which is originally radiate may
become bilaterally symmetrical. Its plane
of symmetry then agrees with the general
plane of symmetry of the body, and passes
through the fifth interradius (the so-called
dorsal interradius in which the genital
aperture lies) and the central (first) radius
of the ventral side. The symmetry is
determined either by the fact that the
portions of the ring on the ventral side
differ in form, size, and manner of connec-
tion from those on the dorsal side, or else
canals of the water vascular system ; 13, inter- .

radial pieces of the calcareous ring ; 14, one of by the presence of a larger number of such
the two small ventral tentacles. portions, in consequence of an increased

number of interradials in definite sym-
metrical interradii. For instance, Synapta digitata has seven interradials, one each
in the mediodorsal and in the two ventral interradii, and two each in the dorso-

lateral interradii.

The portions of the calcareous ring are more or less closely united together
by means of connective tissue (never by means of muscles) ; in some cases they are
firmly fused together.

Structures corresponding to the calcareous ring of the Holothurioidea have long
been sought for in the other classes of the Echinodermata. It was thought that
in the Echinoidea it might perhaps be represented either by the teeth or by the
perignathous apophysial ring, or in certain parts of the masticatory apparatus.

The homology of the teeth of the Echinoidea with the calcareous ring of the

FIG. 340.— The (Esophagus and half the
oral tentacles of a dendrochirote Holo-
thurian (after Ludwig). 1, Genital aperture ;
•2, genital duct; 3, radial pieces of the cal-
careous ring ; 4, retractor muscles ; 5, madre-
porite ; 6, stone canal ; 7, dorsal mesentery ;
8, intestine; 9, Polian vesicles; 10, circular
canal ; 11, continuation of the radial calcareous
pieces ; 12, proximal portions of the radial
canals of the water vascular system

vni ECHINODERMATA—ENDOSKELETON, ETC. 405

Holothurioidea is no longer maintained. The two structures are altogether differently
related to the nervous and water vascular systems.

The homology of the calcareous ring with the perignathous apophysial ring of
the Ecliinoidea is equally doubtful. The radials of the calcareous ring were in this
case compared with the auriculae (ambulacral apophyses). But each auricule is
paired and consists of two processes, or folds, of the edge of the peristome, which
may or may not be connected together by an arch ; the radials, however, are
from the first unpaired. Only the arches of the auriculae could be compared with
the radials ; the arch, however, is not a single plate, but is formed by apposition
of the two neighbouring ambulacral apophyses of one and the same ambulacrum.

The comparison of the calcareous ring with the masticatory apparatus or
Aristotle's lantern of the Echinoidea still remains. The five radials have been
compared with the five fork pieces, and the five interradials with the five arches
of the pairs of jaws (pyramids) of the lantern. This comparison is in many ways
plausible, but here, as before, many difficulties appear when the subject is carefully
investigated. The arches of the jaws are paired structures, and cannot therefore be
compared with the interradials, which are from the first unpaired. Moreover, it
is very doubtful whether they represent independent skeletal pieces ; they appear
rather to be merely muscular processes of the halves of the jaws. Further, the
sinews which proceed from the forks of the radial fork pieces are attached to the
perignathous apophysial ring interradially (i.e. to the interambulacral apophyses),
while the muscles which are attached to the radials of the calcareous ring of the
Holothurioidea run strictly radially.

G. Further Deposits of Calcareous Matter.

Deposits of calcareous corpuscles and masses may occur in the
connective tissue of the walls of various internal and external organs,
especially in the ambulacral and alimentary systems. These will be
considered in connection with the systems to which they belong.

We shall here only mention certain calcareous deposits in the
Clypeastroida. An endoskeleton is here formed. On the oral, as well
as on the apical, inner surface of the test, needles, pillars, lamellae,
etc. rise, sometimes only at the edge, sometimes over large areas.
These may traverse the whole depth of the test, connecting its oppo-
site walls. They more or less completely separate the ambulacral
structures from the other internal organs, such as the intestine, the
genital organs, etc., and may in some cases attain such great develop-
ment that, as in Encope, they form a sponge-like or cellular calcareous
framework throughout the whole interior of the test, in which larger
spaces are left for the masticatory apparatus, the intestine, the
ambulacra, etc. Not infrequently, the ambulacral vessels are completely
vaulted over by deposits of calcareous matter.

H. Concluding Remarks on the Section on the Skeletal System.

In the above section, I have adopted the views of those investigators whose wide
and for the most part difficult researches have convinced them that at least the
plates [of the apical and oral systems (the central, infrabasals, basals, radials, and
orals) are homologous throughout the whole group of the Eehinodermata. These
plates therefore must be ascribed to the common racial form. But these plates are

406 COMPARATIVE ANATOMY CHAP.

in reality only characterised according to their position in the adult animal, whether
radial or interradial, apical or oral, and according to the place of their first appear-
ance (above one or the other ccelomic vesicle). They have no other distinctive
characteristic by which, for example, a radial could be recognised throughout the
class of the Echinodermata. It is therefore still possible that such correspondence
may be merely superficial, merely the expression of the radiate structure so common
among Echinoderms. There is nothing astonishing in the fact that the skeleton
of a radiate animal commences at the poles either with radially or with interradially
arranged plates. Such correspondence, then, as far as it goes, is described as homo-
logy. But nothing is really gained by insisting that such and such Opium-ids
"possess infrabasals, " because the system of plates commences at the apex with five
radial plates, which are followed by another outer row of radial plates. Is it, after
all, certain that the infrabasals are wanting when the skeletal system at the apex
begins with interradial plates (which are on that account called basals) ?

III. The Outer Morphology of the Holothurioidea.

The Holothurioidea form an exception to the rule which applies to all other
Echinoderms that the outer form of the body is accurately reproduced in the test of
skeletal plates. This test gives us, as a rule, exact information as to the position
of the outer apertures of the internal organs, and as to the relation of the radii
or ambulacra, to the interradii or interambulacra. But in the Holothurioidea-, in
whose integument only microscopically small and isolated calcareous bodies occur,
this is not the case. Having treated of the external morphology of the Eehinoidca, the
Asteroidea, the Ophiuroidea, and the Pelmatozoa in the section on the skeletal system,
we must now give some account of the outer morphology of the Holothurioidea.

We shall begin with those forms in which the body, elongated in the
direction of the principal axis, is, in section, circular or pentagonal with
rounded edges (cf. for example, Cucumaria Planci, Fig. 226, p. 287). At
the oral pole of the principal axis (i.e. in the Holothurioidea, anteriorly)
lies the mouth, surrounded by feelers ; at the opposite, apical (posterior)
pole, the anus. Along the body, from before backward, run five ridges,
corresponding with the radii, and causing the pentagonal form of the
transverse section. On each edge there are two longitudinal rows of
tube-feet.

Careful examination shows that the radiate structure of Cucumariu
is even externally disturbed by certain characters which make it bi-
laterally symmetrical. There is only one genital aperture, at the oral
margin of an interradius, which we will arbitrarily call the dorsal
interradius. Further, of the ten oral feelers, two adjacent feelers
are much smaller than the rest. They lie exactly opposite the genital
aperture, and distinguish the middle ventral radius. The plane which
passes through the dorsal interradius and the middle ventral radius, in
the direction of the principal axis (i.e. longitudinally through the body)
is the plane of symmetry.

If the animal is opened, it is seen that this external symmetry
corresponds with an internal symmetry ; the anterior limb of the
intestine is attached by a mesentery to the body Avail in the dorsal
interradius. The stone canal and the genital glands lie in the dorsal
interradius, and the Polian vesicle in the middle ventral radius.

viii EGHINODERMATA— MORPHOLOGY OF HOLOTHURWIDEA 407

The application of the terms " ventral " and " dorsal," in very
many Holothurioidea, is fully justifiable, since there arises, parallel
to the principal (longitudinal) axis, a flattened creeping sole, along
the middle of which the above-named ventral radius runs, while the
middle of the vaulted dorsal surface opposed to this creeping sole is
occupied by the middle dorsal interradius.

The radii and interradii now become arranged in such a way that
three radii (one middle and two lateral) belong with their ambulacra!

rdd

idd

\Bimum

\Triniun

idr>

FIG. 350.— Diagrammatic section illustrating the symmetry of the Holothurioidea. De-
velopment of the Mviumand the trivium (mainly after Ludwig). imd, Medio-dorsal interradius ;
ids, left dorsal ditto; isc, left ventral; idv, right ventral; idd, right dorsal interradius; rds, left
dorsal radius ; rsc, left ventral ditto ; rmv, medio-ventral ; nlv, right ventral ; rdd, right dorsal
radius; mi, anterior or dorsal mesentery; m2, middle or left; m3, posterior or right mesentery;
'ili ^2i *3> first, second, and third, or anterior, middle, and posterior limbs of the intestine ; vd and
vv, dorsal and ventral intestinal vessels ; bd and 1m, right and left branchial tree (aquatic lung) ;
go, gonad ; dg, genital duct ; a, body cavity.

feet to the creeping sole, and form the trivium ; while on the dorsal
surface two radii (one right and the other left) form the bivium. Two
interradii, on the other hand, belong to the creeping sole, and three
to the dorsal surface (cf. the diagram, Fig. 350).

The creeping sole usually runs along over the whole length of the body, less fre-
quently (Psolus, Psolidium} it is limited to a circumscribed region between the
anterior and posterior ends.

The difference between the ventral and the dorsal side (the trivium and the

408

GOMPARA TIVE ANA TOMY

CHAP.

bivium) becomes still more accentuated by the different development of the ambu-
lacral feet in the two regions. On the ventral side, these feet are altogether or princi-
pally locomotory tube-feet (ending in suckers), on the dorsal side (the bivium) they are
exclusively or chiefly non-locomotory papilla (with more or less pointed ends). This
difference between the dorsal and ventral tube-feet is found both in those forms in
which the ambulacral feet, limited to the radii, are arranged in one or more longi-
tudinal rows, and in those in which they are also found in the interradii and arranged
irregularly.

In the genus Psolus, the distinction between dorsal and ventral, and consequently
the bilateral symmetry of the body, becomes still more marked by the entire absence

an

era....

rmv '

a go an

i imd

aa-

TO.S--

FIG. 351.— Derivation of Rhopalodina (A) from an ordinary Holothurian (B) (after Ludwig).
C, Ideal intermediate form, rds, rvs, rmv, left dorsal, left ventral, merlioventral railius ; imd,
mecliodorsal interradius ; o, mouth ; mi, anus ; go, genital aperture ; CM, water vascular ring.

of ambulacral appendages on the bivium. The tube-feet of the middle ventral radius
may also be wanting in some species of this genus.

Where ventral and dorsal are sharply distinguished, the mouth and the anus
tend to shift on to the ventral side.

The condition of the genus llhopalodina (Fig. 351) is quite peculiar.
The body is pear-shaped, and produced into a long stalk. At the end
of this stalk, close to one another, lie the mouth and the anus, and
between them the genital aperture. On the swollen portion of the
body there are ten double longitudinal rows of ambulacral feet, so that
it appears as if Rhopalodina possesses ten radii, whereas it in reality
possesses only five. To obtain this condition we have to imagine (1)
the body of an ordinary dendrochirote Holothurian bent upward an-
teriorly and posteriorly, arid (2) the approximation of the anus and
mouth by the great shortening of the dorsal interradius. The accom-

VIII

ECHINODERMATA— RADIAL ORGANS

409

panying diagram will help to make this clear. The genus Ypsilo-
thurix seems to have become fixed while in the act of being similarly
modified.

In the genus Psychropotes (Fig. 223, p. 285) the dorsal surface is
prolonged beyond the anus into a long caudal appendage directed
posteriorly. Peniagone is distinguished by an anteriorly inclined comb,
rising transversely from the neck. On the swimming disc of Pelago-
thuria, cf. Figs. 224 and 225, p. 286.

IV. Position and Arrangement of the Most Important Organs

in the Radii.

The position and arrangement of the organs in the radii can best
be explained by describing cross -sections. In the Asteroidea, the

FIG. 352.— Transverse Section of a radial region of the body wall of a Holothurian,
partly diagrammatic. 1, Endothelium of the body cavity ; 2, circular musculature ; 3, longitudinal
musculature ; 4, motor nerve ; 5, radial water vascular canal ; 6, radial blood lacuna ; 7, radial
I'idfte of the deeper oral nervous system ; 8, ampulla ; 9, cutis ; 10, epidermis ; 11, tube-foot canal
of the vascular system ; 12, tube-foot ; 13, nerve of the same ; 14, vessel of the same ; 15, radial
nerve strand of the superficial oral nervous system ; 10, epineural canal ; 17, peripheral nerve ; 18,
pseudohaemal canal.

Ophiuroidea, and the Crinoidea, in which the body is produced radially
into arms, the sections to be described will be those of the arms ; in
the Holothurioidea and Ecliinoidea the sections are of a radial region of
the body wall.

Holothurioidea (Fig. 352). — In a transverse section through a

410

GOMPARA TIVE ANA TOMY

CHAP.

radial region of the body wall of an adinopodan Holothurian we find,
proceeding from without inwards :—

(a) The outer body epithelium (10).

(b) The cutis, or the connective tissue layer of the body wall, with

the calcareous corpuscles (9).

(c) The epineural canal (16).

(<1) The radial nerve trunk of the superficial oral system (15).
(e) The radial nerve trunk of the deeper oral system (7).
(/) The subneural pseudohaemal canal (18).
(g) The radial blood lacuna (radial blood vessel) (6).

FIG. 353.— Transverse Section through a radial region of the body wall of an Echinoid, partly
diagrammatic. 1, Ampulla, traversed by muscle filaments; 2 and 3, the two canals traversing the
test and connecting the ampulla and the tube-foot canal (5) ; 4. circular nerve in the terminal disc
of the tube-foot ; 5, tube-foot canal ; 6, nerve of the tube-foot ; 7, integumental nerve ; S, calcareous
substance of the ambulacral plate ; 9, nerve plexus in the depths of the body epithelium ; 10, suture
between t\vi> platt-s of the two contiguous rows of ambulacral plates; 11, body epithelium; 1:2,
epineural canal ; 13, endothelium of the body cavity ; 14, pseudoluvmal canal ; 15, radial blood
vessel ; 16, radial canal of the water vascular system ; 17, radial nerve strand ; 18, lateral canal of
the radial canal of the water vascular system to the ampulla.

(}i) The radial canal of the water vascular system (5), and the tube-
foot canal branching from it transversely (11), and finally
also the ampulla of the tube-foot (8).

(i) The circular musculature of the body (2).

(/•) The longitudinal musculature of the body (3).

(1) The endothelium of the body cavity (1).

The figure also illustrates the relation of a tube-foot to its canal
and ampulla.

This description does not apply to the Paractinopoda (Synaptidce) in
so far as, in these latter, the radial canals of the water vascular system
are altogether wanting.

VIII

ECHINODERMA TA—RADIA L ORGANS

411

Eehinoidea (Fig. 353). — In a transverse section of an ambulacral
area we find : —

(a) The outer body epithelium (11).

(!>) The cutis, almost entirely calcified, as ambulacral plates (8).

(c) The epineural sinus (12).

(d) The radial nerve trunk (17).

(e) The subneural sinus — pseudohaemal canal (14).

IB

Fia. 354.— Transverse Section through the arm of an Asteroid, diagrammatic. 1, Ridges of
the deeper oral nervous system ; 2, radial canal of the water vascular system ; 3, continuation of the
axial organ in the arm ; 4, radial nerve ridge of the superficial oral system ; 5, pseudohstmal canal ;
(3 and 7, branches of the pseudohsemal system running to the tube-feet ; 8, pedicellaria ; 9, spine ;
10, genital aperture ; 11, branchial vesicle (papulla) ; 12, sessile pedicellaria ; 13, continuation of the
body cavity into the branchial vesicle ; 14, brachial diverticulum of the stomach ; 15, circular sinus
of the schizoco;! round the branchial vesicle ; Iti, supramarginal plate ; 17, inframargiiial plate ; IS,
adambulacral plate ; 19, marginal canal of the pseiulohsenial system ; 20, canal connecting it with
the body cavity ; 21, endothelium of the body cavity ; 22, genital sinus of the co?lo7n ; 23, ^nna'l
(ovarium) ; 24, mesenteries of the diverticula of the stomach ; 25, ampulla canal of the water vas-
cular system ; 26, ampulla ; 27, tube-feet canal ; 28, upper and lower transverse muscles of the
ambulacral skeleton ; 29, motor branches of the deeper oral nervous system ; 30, ambulacral plates ;
31, brachial cavity (ccelom) ; 32, apical longitudinal muscle of the arm ; 33, nerve ridge of the apical
nervous system.

(/) The radial blood vessel (15).

(g) The radial canal of the water vascular system (16).

(li) The endothelium of the body cavity (13).

The figure at the same time illustrates the relation of the tube-feet
to their ampulla?, the double pores, etc.

Asteroidea (Fig. 354). — In a transverse section through the lower
(oral) wall of an Asteroid arm we find, from without inwards :—

412

COMPARATIVE ANATOMY

CHAP.

(a) That the body epithelium stretching across the ambulacral fur-
row is thickened into the longitudinal ridge which stands up
in the base of the furrow, and here contains :

(ft) The radial nerve strand (lying in the epithelium itself) (4).

We find further : (c) below the latter, to the right and left, the
trunks of the deeper oral nervous system (1).

(d) The radial pseudohpemal canal (5), which is divided into two
lateral portions by a vertical septum.

FIG. 355.— Transverse Section through the arm of an Ophiuroid, diagrammatic. 1, Ambu-
lacral tentacle ; ~2, its water vascular canal ; 3, epineural circular canal at the base of the tentacle ;
4, circular ganglion at the base of the tentacle ; 5, ventral shield ; 6. radial epineural canal ; 7, radial
nerve trunk of the superficial oral nervous system ; 8, continuation of the axial organ in the arm (?) ;
9, radial trunk of the deeper oral nervous system; 10, radial pseudolui-mal canal; 11, peripheral
branch of the radial nerve trunk ; 12, spine ; 13, lower (oral) intervertebral muscle cut across ; 14,
lateral shield; 15, vertebral ossicle; 1(3, upper (apical) intervertebral muscle; 17, dorsal canal of
the brachial cavity (co-loin) ; IS, ciliated strip of endothelium ; 19, dorsal shield ; '20, radial canal of
the water vascular system ; 21, lateral portions of the brachial cavity, which are segmentally re-
peated ; 22, branch of the water vascular system running to the tentacle ; 23, ganglion at the base
of the spine ; 24, motor branch of the nerve (of the deeper oral system).

(e) The radial canal of the water vascular system (2), with the
canals of the tube-feet branching from it. (All these are
separated from one another by thin layers of connective
tissue.)

(/) The ambulacral plates (30), with the transverse muscles which
connect them (28).

(y) Still further in, and projecting into the body cavity, are the
ampullae (26) of the tube-feet.

VIII

ECHINODERMATA— RADIAL ORGANS

413

(A) The endothelium of the body cavity (21).

The figure also shows the relations of the ampullae to the tube-feet

and their canals, and the organs of the apical side of the arm

of an Asteriod.

Ophiuroidea (Fig. 355). — In a section through an arm, proceeding
from the lower (oral) to the upper (apical) side, we find :—
(«) The body epithelium.
(b) The ventral shield (5).

9

FIG. 356.— Transverse Section through the arm of a Crinoid, diagrammatic. 1, Radial nerve
trunk of the superficial oral nervous system ; 2, radial pseudolueinal canal ; 3, radial canal of the
water vascular system ; 4, the paired deeper longitudinal nerves of the arms ; 5, 7, and 11, the three
radial sinuses of the brachial ccelom ; 6, genital sinus with genital rachis ; 7 (see 5) ; S, nerve
trunk of the apical nervous system ; 9, end of the nerves at the surface ; 10, branch connecting 4
and 8 ; 11 (see 5) ; 12, tentacle nerve ; 13, tentacle canal of the water vascular system ; 14, sensory
cone on the tentacle ; 15, food groove of the arm.

(c) The radial epineural canal (6).

(d) The radial nerve trunk of the superficial oral nervous system (7).

(e) The radial nerve trunk of the deeper oral system (9).
(/) The (subneural) radial pseudohsemal canal (10).

(g) The radial canal of the water vascular system (20).

(h) The calcareous mass of the vertebral ossicle (15), which is

traversed by the tentacle canals (22), and the intervertebral

musculature (16 and 13).

414 COMPARATIVE ANATOMY CHAP.

(i) The endothelium of the body cavity.

(k) The much diminished body cavity itself (enterocoel, 17 and 21).

(/) The dorsal (apical) body wall, which in this connection is of no
further interest.

Crinoidea (Fig. 356). — On the section of the arm of a Crinoid,
proceeding from the oral to the apical side, we find : —

(a) The body epithelium covering the food groove.

(6) Deep in this epithelium, the radial nerve trunk of the super-
ficial oral system (1).

(c) Below the epithelium (not invariably present) a small schizo-

coel canal (pseudohsemal canal (2).

(d) The radial canal of the water vascular system (3).

(e) At its two sides, the paired subepithelial longitudinal nerves of

the arms (4).

(/) The three radial sinuses; viz. two paired sinuses (5 and 11),
separated by a vertical septum (the so-called ventral or sub-
tentacular canals), and a third unpaired sinus (7), the
dorsal canal, separated from the first two by a horizontal
(transverse) septum.

All these parts lie embedded in somewhat sparse connective tissue.
In the middle between them run :—

(g) Tiie narrow genital sinus (G), with the genital tube (rachis)
within it.

(h) The skeletal ossicle of the arm, or (according to the plane of
the section) the apical and oral muscles and bands, uniting
the ossicles.

(i) In the centre of the joint we find the section of the nerve canal
(axial canal) with the radial trunk of the apical nervous
system (8) which it encloses.

The figure also shows the tentacles, and the nerves which connect
the paired radial nerves of the oral with the radial trunk of the apical
nervous systems.

V. The Integument.

The integument of the Echinodermata consists of (1) the uni-
laminar body epithelium which covers the whole body with its processes
and appendages, and (2) a strong subjacent connective tissue layer
(the eutis or eorium) of mesenchymatous origin, in which the various
skeletal structures develop. The cutis forms by far the largest part
of the body wall. Internally, it is either directly lined by the endo-
thelium of the body cavity or else is separated from the endothelium
by musculature (Holothuria, Asteroidea).

(1) The body epithelium. — («) This is distinct from the subjacent cutis in the
Echinoidca, Asteroidea, many Holothurioidea, and on the oral surface of the disc and
arms of the Crinoidea ; also in the Euryalcc.

In the OpJiiuroidca (excluding Euryalcc), and on the apical side of the disc and

vm ECHINODEEMATA—THE INTEGUMENT 415

arms of the Crinoidca, there is no sharp line of distinction between the body
epithelium and the cutis. Such a distinction is, however, demonstrable in very young
stages. In later stages of development the elements of the two forms of tissue seem
to mingle, and skeletal substance forms right up to the surface of the integument.

In many Holothurioidea also, the body epithelium, as such, is very indistinct.
In Cucumaria, for example, the cutis appears at the surface of the integument, and
the body epithelium is found in the form of nests of cells scattered within the peri-
pheral layer of the cutis. Each cell sends a thin process to the surface of the
integument.

(b) The body epithelium is usually covered by a cuticle of varying thickness.

(f) The body epithelium is ciliated over the whole surface of the body in the
Asteroidea and Echinoidea, but in the Crinoidca only in the food grooves.

The integument of the Ophiuroidea, Crinoidca (with the exception of the food
grooves), and Holothurioidea, is non-ciliated.

(d) The body epithelium of the Astcroidca is rich in glands. The glands are
usually unicellular (goblet glands, granular glands, etc. ), and remain on the level of
the epithelium. In Echinastcr sepositus, large multicellular glands are also found,
whose pear-shaped or spherical bodies dip down into the cutis. In the integument
of the Holothurioidea also glands have been described, and it will probably be dis-
covered that certain epithelial cells of the Echinoidea are of a glandular character.

(e) The integumental pigment may belong to the epithelium as well as to the
cutis ; it not infrequently occurs in both layers.

(/) Epithelial sensory cells, ganglion cells, and nerve fibres will be described in
another place.

(2) The cutis of the Echinodermata is always very thick, although it shows
extraordinary variations in this respect according to the genus and species. It
everywhere consists (a) of a ground- or intercellular substance of gelatinous or carti-
laginous consistency, and (b) of the nucleated connective tissue cells which secrete
this ground-substance and are embedded in it ; these cells are spindle-shaped, star-
shaped, etc. There are, further, (c) in all Echinodermata, granulated plasm cells
or wandering cells (amcebocytes) similar to those which are to be found in different
body fluids. These can move, like amcebte, in and through the different tissues.

In Holothurioidea, these wandering cells may collect in such quantities in the
deep looser layer of the cutis, as to form a distinct layer (Wanderzellenschicht).

The calcareous skeleton of the body wall of the Echinodermata always lies in
the cutis, whether it consists, as in the Holothurioidea, of isolated calcareous cor-
puscles, or, as in other Echinoderms, of larger plates of lattice-like or spongy struc-
ture. In sections through the decalcified body wall, the spaces in which the skeleton
lay are visible. In other words, the connective-tissue fills up all the spaces in the
spongy calcareous skeleton. Since the wandering cells can travel to the surface
through these spaces, they may play an important part in the nutrition of the soft
parts which lie at the surface of the skeleton, especially in Asteroids and Echinoids.

It appears that even the intercellular substance may occasionally become differ-
entiated into fibres, which, however, are difficult to distinguish from the fibrous
processes of the connective-tissue cells.

Where two skeletal plates are united by a suture, this suture is formed of thickly
crowded parallel fibres, which connect the ground-substance of one plate with that
of the other.

416 COMPARATIVE ANATOMY CHAP.

VI. The Water Vascular System.
(System of the Ambulacral Vessels : Hydroccel.)

This is a system of canals filled with fluid, the arrangements of
which may be generally described as follows.

An outer aperture, the madreporite, leads first into a vesicular
section of the ccelom, the madreporitie ampulla. This again is con-
nected by means of a stone canal (so called because that portion of its
wall which consists of connective tissue is often calcified) with a ring"
canal which surrounds the oesophagus. Into the madreporitie ampulla
there opens, further, the axial sinus of the body cavity, which follows
the stone canal in its course, and surrounds a lymphatic gland, the
axial organ.

The water vascular ring1 may carry various accessory structures,
whose principal function seems to be that of lymphatic glands, and
which are known as Polian vesicles, Tiedemann's bodies, etc.

From the ring canal there run out into the radii of the body,
either in the body wall or in close contact with it, as many radial
canals as there are radii (usually therefore five). The radial canals
send off, on each side, tube-feet canals, which run into outer append-
ages of the body wall, ending blindly at their tips. These extensible
appendages are usually present in great numbers, and serve either as
tube-feet for locomotion (Holothurioidea, some Echinoidea, Asteroidea),
and are then provided with a terminal sucker, or as tentacles, ten-
tacular gills, etc., for tactile purposes, for respiration, and for conduct-
ing food (some Echinoidea, Ophiuroidea, Crinoidea). In connection with
the tube-feet canals, tube-feet ampullae are very often found (Holo-
thurioidea, Echinoidea, Asteroidea) ; these are accessory contractile
vesicles, which serve for the swelling of the tube -feet. Special
valves are so arranged as to prevent the flowing back of the water
vascular fluid into the radial canals (Fig. 352, p. 409).

The chief departures from this general description met with in
the five classes of Echinoderms, affect the madreporite, the madre-
poritie ampulla, and the stone canal. These will be described in
detail later on.

Structure of the wall of the water vessels. — Lining the lumen of the vessels,
there, is generally found, first of all, a ciliated epithelium. This is followed, in
most parts (always in the ambulacral appendages), by a longitudinal muscle layer.
Outside this latter lies a layer of connective tissue, and, outermost of all, there is
almost always an external ciliated epithelium. On the ambulacral appendages (the
tube-feet and tentacles) this last is nothing more than the external body epithelium.
But in those parts of the water vascular system winch project into or lie in the
body cavity, it is the endothelium of the ccelom. This outer epithelium of the
water vascular system is rarely altogether wanting ; it is, however, absent in such
parts of the system as run embedded in the body wall. A circular musculature is
seldom found ; it only occurs locally.

vni EGHINODERMATA— WATER VASCULAR SYSTEM 417

Calcareous corpuscles may be formed in the connective, tissue layer of the wall in
other parts of the water vascular system besides the stone canal. Such calcification
always takes place in locomotory tube-feet.

The fluid contained in the water vascular system is sea-water
with traces of albumen (in a 0'5 - 2 per cent solution). Floating in
this fluid are found amoeboid cells (lymph bodies) and coloured cor-
puscles often united into small lumps. The fluid occasionally appears
of a pale yellow, or reddish, colour.

The origin of this fluid is a question of frequent recurrence.
The view which still appears best supported is that sea-water flows in
through the madreporite and the stone canal, but an exactly opposite
view has also been maintained. The observations made on this
subject appear to contradict one another, it being very difficult to
carry on investigations in a decisive and satisfactory manner.

A. Madreporite and Stone Canal.

1. Holothurioidea (Fig. 357). — The condition which must be con-
sidered as the original is that in which only one stone canal occurs ;
this is attached to the dorsal mesentery (cf. p. 406), and its
madreporite lies mediodorsally in the integument, and its pore
canal or canals open outward direct.

Such a condition is found in the adult only in certain Elasipoda
and in Pelagothuria.

In the large majority of Holothurioidea, the stone canal loses all
direct communication with the exterior, while at its distal end,
Avhich now lies in the body cavity, a new inner madreporite forms,
throug-h whose canals communication is established between the
stone canals and the body cavity.

In a comparatively small number of Holothurioidea (never in
MolpadiidcB and Elasipoda} the number of stone canals increases (the
single canals usually shortening at the same time), and may finally
become very great (over 160).

The inner madreporite is found in the form of a variously shaped swelling on
the stone canal, which is often S -shaped or spirally coiled. Only the primary stone
canal is connected with the dorsal mesentery ; this is never the case with accessory
canals. These latter float freely in the body cavity, and this is also the condition
of the primary stone canal of the Aspidochirotce, which has lost its connection with
both the body wall and the mesentery.

More than one canal is found in only a very small number of forms even among
the Synaptidcc, the Dendrochirotce, and Aspidochirotce. The number of accessory
canals varies greatly in different forms ; it does not seem to be of systematic import-
ance, since it varies in individuals of one and the same species. It is probable that
the accessory canals, ontogenetically, bud off secondarily from the water vascular
system, whereas the dorsomedian stone canal arises primarily from the canal which,
in the larva, connects the hydroccel with the exterior.

Branched stone canals, with a madreporite at the distal end of each branch,
occur in Synapta Beselii, Jag ; and Thyonc chilensis, Semp.

VOL. II 2 E

418

COMPARA TIVE ANA TOMY

CHAP.

The madreporite of the primary (mediodorsal) stone canal. The simplest, and
no doubt also the most primitive condition is found in Pelagothuria and in certain
Elasipoda, e.g. species of the genera Scotoplancs, Kohja, Parelpidia, Elpidia, Pcni"-
(jonc, and Benthodytcs. In these the stone canal opens simply through a single
mediodorsal pore, which lies in front of the genital aperture (Fig. 357 A). In other
species of these genera and in species of Psych routes, Lcctmogone, llyodccmon, more
than one madreporite pore is found, their number varying, according to.the species,
from two or three to fifty or more (Fig. 357, B). In other cases (species of the Elasipod
genera Irpa, Elpidia, Oneiroplianta, Oiyihnurgus, Benthodytcs, and the Molpadiidan
genera Trochostoma and Anfci/rodcrma) the distal end of the stone canal still remains

FIG. 357.— Diagrams illustrating the various relations existing between the stone canal
and the madreporites in the Holothurioidea. 1, Body wall; 2, commencement of the radial
i-anal ; 3, oesophagus ; 4, dorsal mesentery ; 5, stoue canal ; G, outer madreporite ; 61, inner madre-
porite ; ~, genital aperture ; S, genital duct ; y, water vascular ring ; 10, Polian vesicle.

embedded in the body wall, but it has lost the pore or pores which formed the com-
munication between it and the exterior. New pores therefore arise laterally at the
distal portion, which still lies in the body wall, and these now open communication
between the lumen of the stone canal and the body cavity, and make this widened
part of the stone canal into an inner madreporite (Fig. 357, C). Other Molpadiidcc
and the Synaptidce and Dendrochii-ota differ from these last only in the fact that in
them the stone canal has become entirely detached from the body wall (Fig. 357, D).
In the Aspidochirota, which also possess an inner madreporite, the latter appears
complicated, in that its pore canals do not open direct into the lumen of the stone
canal, but first into a collecting cavity, which in its turn communicates by means of
an aperture (occasionally through several) with the lumen of the stone canal.

2. Eehinoidea (Fig. 358, 33). — In the Echinoidea, so far as is

vni EGHINODEEMATA— WATER VASCULAR SYSTEM

419

FIG. 358.— Diagram of the organisation of a regular Echinoid. Section in the direction of
the principal axis. The surface of the section lies interradially on the left and radially on the
right. The left half is incomplete. 1, External gill (this would not exactly come into this section,
since there are five pairs of interradially placed gills) ; 2, seizing pedicellaria ; 3, oral integument ;
4, tooth ; 5, mouth ; 6, cushion of connective tissue ; 7, nerve ring of the superficial system ; S,
deeper oral nervous system ; 9, radial epineural canal ; 10, radial blood vessel ; 11, arch of the
ambulacral apophysis (auricula) ; 12, spliseridium in its niche ; 13, radial canal of the water vascular
system; 14, radial nerve trunk (of the superficial oral system) ; 15, radial pseudolmemal canal; l(i,
circular ganglion at the base of the spine; 17, spine; IS, glandular pedicellaria; 19, ambulacral
tube-feet with terminal disc ; 20, 21, ambulacral tentacles (without terminal disc) ; 22, terminal
feeler or tentacle emerging through the pore in the radial (ocular) plate ; 23, apical (genital) ring
sinus ; 24, perianal sinus of the ccelom ; 25, anus ; 26, sinus, into which a process (27) of the axial
organ projects ; 27, aboral process of the axial organ ; 28, madreporite ; 29, genital aperture on the
genital papilla ; 30, genital duct ; 31, madreporitic ampulla, into which the stone canal and axial
sinus enter from below ; 32, axial organ ; 33, stone canal ; 34, part taken by the blood lacuna in
the formation of the Polian vesicle ; 35, root of the tooth ; 36, muscle of the forked radii (Fig. 348, 7)
cut through ; 37, arch (arcus) of a jaw pyramid of the masticatory apparatus ; 38, lantern mem-
brane ; 39, ligament of a forked radius ; 40, adductor muscle of the teeth ; 41, interambnlacral
apophysis ; 42, general body cavity (coelom) ; 43, pyramid ; 44, peripharyngeal sinus, lantern sinus
of the crelom ; 45, part taken by the water vascular system in the formation of the Polian vesicle ;
46, circular vessel of the blood lacunar system ; 47, water vascular ring ; 48, oesophagus ; 49, axial
sinus of the coelom ; 50, hind-gut ; 51, perirectal sinus of the coelom ; at 52 and 54 the section is not
quite radial, so that it does not, as at 22 and 56, take in the radial canal of the water vascular
system, but passes transversely through its lateral canals which lead to the ampulla' ; in 53 the
plane of the section lies still more to the side, so that the ampulla is taken in (cf. Fig. 353) ; 57,
abductor muscle of the teeth ; 58, Stewart's organ ; 59, muscles between the pyramids ; 60, inter-
mediate plate ; 61, forked radius ; 62 and 65, intestinal vessels ; 63, accessory intestine ; 64, prin-
cipal intestine. The accessory intestine in reality runs on the axial side of the principal intestine.

420 COMPARATIVE ANATOMY CHAP.

yet known, there is always only one stone canal, and it always com-
municates with the exterior by means of the pores of the madreporite.
This communication is, however, by no means direct. The pores of
the madreporite first lead into a small cavity lying below it, the
madreporie ampulla, into which opens, on the one hand, the ascend-
ing stone canal, and, on the other hand, the axial sinus of the entero-
coel, to be described later. The stone canal, on leaving the ampulla,
traverses the body cavity, following the axial sinus Avith its lymph
gland, and runs down to the water vascular ring, which in the
Cidaroida and Clypeastroida encircles the oesophagus immediately above
the masticatory framework (Fig. 358), but, in the Spatangoidct, imme-
diately above the mouth. In both the former groups the stone canal
is short and more or less straight, but in the Spatangoida it is very
long and runs in coils.

On the possibly great morphological importance of the ampullae, cf. the section
on Ontogeny.

Echinocyamus pusillus, a Clypeastrid, shows an embryonic condition in the
adult in that the madreporite has only one pore. All other Echinoids, examined
with reference to this point, possess as adults several or numerous pores. The
number of pores increases with age and growth.

The pore canals which traverse the madreporites may anastomose with one
another. They may enter the ampulla through several inner pores or else through
one common inner aperture. In the Spatangidcc, they traverse the substance of a
large skeletal process (apophysis) of the madreporite, which projects into the cavity
of the test.

The condition of the stone canal in the Spatangoida deserves further investiga-
tion, since the observations hitherto recorded contradict one another. According to
one account, the stone canal (in EcMnocardium) breaks up into branches on its way to
the water vascular ring, these branches communicating with the axial blood lacuuar
system. According to another, it ends blindly (in Spatangus purpurcus), and the
water vascular ring is said in no way to communicate openly with the apical stone
canal. A canal, however, runs from the water vascular ring towards the stone canal,
without reaching it. The existence of any kind of communication with the lacunar
system is emphatically denied by those who hold this latter view.

3. Asteroidea. — In all Asteroids the madreporite is external,
and takes the form of a skeletal plate, which is perforated by many
pores, and always lies on the apical side of the disc interradially.
The stone canal, within the axial sinus and attached by a band to its
wall, descends direct to the water vascular ring which surrounds the
oesophagus, and enters this ring interradially. The wall of the stone
canal is generally highly calcined, and its lumen is divided in a more
or less complicated manner into shelves, niches, etc., by projecting
folds which frequently branch. It not infrequently happens in
Asteroids that there is more than one stone canal and madreporie
plate. For example, all Asteroids which reproduce asexually (i.e. by
division) possess more than one canal.

The relations of the madreporite to the axial sinus are interesting.
Not all the pores of the madreporie plate open into the stone canal ;

vin EGHINODERMATA— WATER VASCULAR SYSTEM 421

some of them open direct into the axial sinus. No direct communica-
tion between the stone canal and the axial sinus is found in adult
animals.

The madreporite appears externally marked by furrows radiating from the centre
to the periphery (Fig. 359). In the bases of these furrows lie the apertures of the
pores. The pore canals, which run through the substance of the
madreporite to the stone canal, anastomose in definite ways, which
cannot here be described.

The increase of surface of the inner wall of the stone canal
(Fig. 360) is of some interest. As the middle layer of connective
tissue takes part in the formation of the folds projecting into the
lumen, these folds may calcify. The simplest condition is found in
the Echinasteridce and Asterias tenuispina, where a projecting longi-
tudinal ridge is formed on the inner wall of the stone canal (Fig. quarter of the
360, A). In Asterina the free edge of this fold splits up into two madreporic
diverging lamellse, in such a way that the transverse section is Y- or £epacanthio^
anchor -shaped (B). The lamellae may become coiled (species of rul)ens (;lfter
Asterias, Pentaceros, Gymnasteria, C). Occasionally the ridge traverses Ludwig).
the whole lumen of the canal as a septum (D), and then carries on
each surface a coiled lamella (species of Astropcctcn) . The whole lumen, further,
may be traversed by septa which, in transverse section, form a network (Luidia,
Culdta, species of Astropcctcn and Ophidiaster, F).

Number of the stone canals and madreporitic plates. — Several madreporites
and stone canals (two to five and more) are not infrequently found in individuals

FIG. 359.— A

FIG. 300.— A-F, Transverse sections through the stone canal of various Asteroids. 1, Sus-
pensor of the stone canal to the wall of the axial sinus ; 2, endothelium of the axial sinus ; 3, inner
epithelium of the stone canal ; 4, connective tissue portion of the wall.

with six, seven, or more arms, belonging to species which normally have five arms.
There are, however, some species (having normally five or more arms) which habitu-
ally possess more than one madreporite (Asterias capensis, A. polyplax, Ophidiaster
Germani, Acanthaster echinites and A. Ellisii). On the other hand, the species of the
genera Solaster, Heliaster, and Luidia, which normally have numerous arms, possess
only one madreporite. When more than one madreporite is present they lie, as a
rule, in different interradii. Cases have, however, been observed in which two stone
canals occurred in one and the same interradius, and even in one and the same axial
sinus.

422

COMPARATIVE ANATOMY

CHAP.

4. Ophiuroidea. — In this class, as a rule, one single madre-
porite with one pore aperture and a single stone canal are present.
The pore aperture is not found, as in Asteroids and Echinoids, on the
apical side of the body, but, in adult Ophiuroids, on the oral side of
the disc, asymmetrically in an interbrachial area, and on that edge of
the oral shield which is turned to the bursal aperture. This oral
shield thus becomes the madreporic plate. The pore aperture leads
first into an ampulla (Fig. 361, 3), which probably corresponds with
the axial sinus of the Asteroidea and Ecliinoidea. Into this ampulla
the stone canal which descends from the water vascular ring opens.

FIG. 361. — Stone canal and neighbouring parts of Amphiura squamata, diagrammatic
vertical section through the madreporic interradius of the disc. 1, Water vascular ring ; 2,
stone canal ; 3, ampulla ; 4, madreporic canal ; 5 and 7, axial sinus (?) ; 6, circular genital sinus ;
8, axial organ ; 9, genital rhachis ; 10, bursal pouch ; 11, oral wall of the intestine ; 12, peristomal
sinus ; 13, interradial muscle ; 14, circular nerve ; 15, teeth ; 16, mouth ; 17, oral surface of the disc.

A large part of the ampulla lies on that side of the stone canal which
is turned towards the mouth. In consequence of the position of the
pore aperture, the stone canal, which rises out of the water vascular
ring interradially, runs in a downward (oral) direction.

The diagram (Fig. 361) illustrates in detail (1) the relation of the stone canal to
the axial sinus ; (2) the manner in which the former enters the madreporic ampulla,
the cylindrical epithelium of the former being directly continued into the tessellated
epithelium of the latter ; (3) the opening outward of the ampulla through a madre-
poric canal.

It appears that in many species of the genera Amphiura, Ophiolepis, Ophio-
plocus, Ophionereis, and Ophiocnida, several or many pore apertures occur at the
edge of the oral shield. This is certainly the case in many Astropliytidcc. In
Trichaster, however, only one pore aperture is present ; but this and the stone canal
belonging to it are repeated in each interradius.

In Ophiactis virens also, which reproduces itself asexually by division, several
(as many as five) stone canals occur in the adult in different interradii. In young
individuals only one is found.

VIII

ECHINODEEMATA— WATER VASCULAR SYSTEM

423

5. Crinoidea. — Adult Crinoids have at least five, and usually
many more or even very numerous stone canals, all of which open
into the body cavity. Communication between the exterior and the
body cavity is brought about by
at least five ciliated pores (Kelch-
poren) in the tegmen calycis ;
their number is generally far
greater, and may mount up to a
thousand. Each single pore
corresponds with a madreporite
with one pore canal. We must
not therefore compare the calyx
pores of a Crinoid collectively
with the numerous pores of a
madrepoi'ic plate. Originally,

no doubt agreed with the number
of stone canals. In cases in

FIG. 362.— A stone canal and pore of the calyx

the number Ot pores On the Calyx of Rhizocrinus lofotensis, diagrammatic (after

Ludwig). Interradial section in the neighbourhood
of the mouth. ] , Tegmen calycis ; 2, calyx pore ;
3, aperture of the stone canal into the body cavity ;
which both Structures are Very 4, intestinal epithelium ; 5, intestinal cavity ;

6, coelom ; 7, stone canal ; 8, ring canal ; 9, circular
nerve ; 10, cesophageal epithelium ; 11, oesophagus ;
12, connective tissue.

numerous, however, no such
relation can be established.

In many inadunate Crinoids
(cf. p. 303) a madreporite occurs in the posterior interradius of the
tegmen.

Eliizocrinus lofotensis and Actinocrinus verneuilianus have only five
interradial stone canals and five interradial pores in the calyx. The
openings of the stone canals into the body cavity lie directly below
the pores belonging to them.

For the number and arrangement of the calyx pores, cf. the section on the Tegmen
Calycis of the Crinoids, p. 377.

B. The Water Vascular Ring and its Appendages.

1. Holothurioidea. — The water vascular ring always encircles the
O3sophagus behind (i.e. apically to) the calcareous ring. In all Holo-
thurioidea without exception it carries Polian vesicles. As a rule,
only one Polian vesicle is present.

These pear-shaped or tubular cnsca of the water vascular ring, which project
freely backward into the body cavity, vary greatly in size. In extreme cases they
may be half as long as the body.

In the Molpadiidce, and among the Elasipoda in the Psychropotidcc and Dcimatida:,
more than one vesicle has never yet been observed, and, in the Elpadiidce, there
is, normally, only one. In other divisions, a varying number of species, greatest in
the Synaptidce, have more than one Polian vesicle. In all such species, however,
there was originally only one vesicle. Where accessory vesicles occur they vary
greatly in number, and appear to have very slight, if any, systematic significance.

424 OOMPABATIVE ANATOMY CHAP.

Where only one Polian vesicle occurs, it lies in the left ventral
interradius, very seldom in the left dorsal interradius.

"Where two or more vesicles occur, they are also mostly found in the ventral
region of the circular canal.

The walls of the Polian vesicles correspond in structure, essentially, with those of
the ring canal. Cells belonging to the inner epithelium become amceboid and break
away from the wall. These are said to become the lymph cells of the water vascular
system.

2. Eehinoidea. — In the Spatangoida (which have no masticatory
apparatus) the ring canal encircles the oesophagus immediately above
the mouth. In other Eehinoidea, however, it is pushed up by the
masticatory apparatus which intervenes between it and the mouth.
The canal therefore surrounds the oesophagus at the point where this
latter emerges from the lantern. The ring canal, as well as its
accessory structures, nevertheless, lie within the lantern membrane,
which envelops the whole masticatory apparatus. The circular vessel
(the lacunar ring) is in close contact with the canal (Fig. 358),

In the Spatangoida and some Clypeastridce (Ecliinocyaimis pusillu*),
the ring canal has no accessory structures. In the Stereosomata, on the
contrary, it has, in each interradius, a small outgrowth, which ramifies
and intertwines with similar ramifications of the circular blood vessel
to form together a spongy body, which is known as the Polian vesicle,
and is regarded as a lymph gland. This, which, in the Stereosomata, is
confined to certain localised interradial points, occurs in the Cidaro'uln,
certain Clypeastroida (e.g. Pcrondla orbicularis), and the Streptosomata,
along the whole course of the canal, so that the intertwining of the
appendages of the ring canal and of the circular blood vessel gives
rise to a spongy ring.

An intermediate stage is found in EcMnodiscus Mforis (Chjpeastroid), in which
the interradial spongy bodies of the circular canal are longer than in the Stereosomata,
but long radially arranged tracts are still left free, the canal at these parts retaining
its simple lumen.

3. Asteroidea. — The circular canal which surrounds the mouth,
following the inner outline of the oral skeleton, here has two kinds
of appendages : Tiedemann's bodies and the Polian vesicles, both
of which lie interradially. Tiedemann's bodies appear to occur in all
Asteroids, whereas the Polian vesicles are wanting in some families,
c.ij. the Asteriidce, Echinasterida>, and Lincldidce.

Tiedemann's bodies (Fig. 363, 7) are small tufts of tubules, closely crowded
together, their walls of connective tissue being fused with one another. These
tubules, which open into the circular canal, are lined internally with a cubical
epithelium, and contain within their lumen bundles of cells which have broken
away from the wall. These cells, the protoplasm of which contains pigmented
concretions, become the amceboid lymph cells which float in the fluid of the water
vascular system. They give the Tiedemann's bodies their more or less distinct
coloration.

viii EGHINODERMATA — WATER VASCULAR SYSTEM

Two Tiedemann's bodies usually occur in each interradius ; the interradius
containing the stone canal not infrequently, however, forms an exception to this
rule, only one such vesicle being present in it (Asteriidce, Ecliinasteridce, Linckiidce,
Astcrinida:, Culdtidcc}. If the circular canal is viewed internally in the position
shown in Fig. 363, this body lies to the right of the stone canal.

The Polian vesicles (Fig. 363, 6) are large structures with long stalks, and to
them, as to Tiedemann's bodies, the function of lymph glands has been attributed.
In the Asterinidcc, Culdtida;, Luidia, and several species of Astropecten, one

JO

10

Fia. 303.— Circular canal, Polian vesicles, Tiedemann's bodies, and ampullae of the water
vascular system of Asterina gibbosa (after Cudnot). Seen from within, i.e. from the body cavity.
1, Mouth at the centre of the oral membrane ; 2, stone canal ; 3, axial sinus ; 4, transverse muscles
of the ambulac.ral plates ; 5, ambulacral plates ; 6, Polian vesicles ; 7, Tiedemann's bodies ; S, circular
canal ; 9, blood vascular ring (?) ; 10, ampulla?.

vesicle is found in each interradius. Only in the interradius containing the stone
canal is it wanting, or else (in species of Astropecten) two are here found instead of
one. Astropecten aurantiacus has two to four (usually three) Polian vesicles in each
iuterradius (even in the stone caual interradius). The wall of this vesicle, proceeding
from without inwards, consists of: (1) the ciliated endothelial covering ; (2) a layer
of connective tissue in which run the longitudinal muscle fibres ; (3) a circular
muscle layer, and (4) the inner epithelium, whose cells lie in the interstices of a net-
work of connective tissue.

4. Ophiuroidea. — The water vascular ring here possesses one
Polian vesicle, which functions as lymph gland in each interradius
except that of the stone canal. The structure of the wall of this
vesicle resembles that in the Asteroidea, the longitudinal musculature,
however, seems always, and the circular musculature frequently, to be
wanting. The canals to the first two tube-feet arise directly from the
circular canal, commencing usually as a common canal which forks
later, but occasionally the canals are distinct from the first.

Ophiactis wrens (Fig. 364) occupies an exceptional position among the Ophiuroidea,
being capable of asexual reproduction by means of fission. This form not only has,
as already mentioned, several stone canals, but in each interradius two to three

426

COMPARATIVE ANATOMY

CHAP.

Polian vesicles and, besides (an altogether unique condition), six to fifteen long thin
accessory vessels in each interradius, which are hollow and end blindly ; these
encircle the intestine and in sexually mature animals penetrate between the genital
organs. The walls of these vessels, which are filled with blood and lymph corpuscles,
and communicate with the circular canal, consist, from without inwards of: (1) the

FIG. 364.— A portion of the disc of Ophiactis virens in horizontal section, somewhat diagram-
matic (after Cu^not). 1, Oral tentacles ; 2, tooth ; 3, circular canal ; 5, section of the stomachal
sac ; 6, Polian vesicles ; 7, accessory vessels of the circular canal ; S, stone canal.

endothelium of the body cavity ; (2) a thin layer of connective tissue ; (3) the inner
epithelium. This altogether peculiar development of the water vascular system in
Ophiactis virens is considered to be connected with the absence of bursfe which serve
for respiration, Ophiactis standing alone among the Ophiuroidea in having no such
structures. This peculiar development of the water vascular system is said to be a
supplementary means of respiration.

5. Crinoidea. — The circular canal which surrounds the mouth has
here no accessory structures except the stone canals. It is provided
with longitudinal muscle fibres, which are connected with the epithelial
cells (epithelial muscle cells). As in the radial canals, muscle cells
also occur transversely traversing the lumen of the canal. The
circular canal gives off canals direct to the five groups of tentacles
which surround the mouth.

C. The Radial Canals, the Canals of the Tentacles and Tube-feet ;
the Tentacle and Tube-feet Ampullae.

1. Holothurioidea. — The Holothurioidea fall into two very
distinct groups, the Si/napt/ihc being- distinguished from all
other members of the class by the fact that, in adults, neither
tube-feet, tube -feet canals, ampullse, nor any traces of radial
vessels are found. The Synaptidce (Pamdinopoda) have only oral

vni EGHINODERMATA— WATER VASCULAR SYSTEM 427

tentacles and tentacle canals, the latter springing- directly out of
the circular canal.

The arrangement in all other Holothurioidea (A dinopodci) may be
described as follows. There are five radial canals, and never more.
The tentacle canals never spring directly from the circular canal, but
arise out of the radial canals. The tentacles are to be regarded as the
first (modified) tube-feet, and the tentacle canals as the first tube-feet
canals.

The canals of the tube-feet and tentacles are usually connected
with ampullse.

Actinopoda (Fig. 365). — From the circular canal the radial canals run forward
(anteriorly) along the cesophagus towards the mouth, passing the axial surface of
the calcareous ring (i.e. between it and the cesophagus). They then pass, together
with the radial nerves on whose inner side they lie, through the incisions or apertures
belonging to them in the ring, and run backwards (aborally) in the body wall,
outside the circular musculature, and end blindly near the anus.

In some rare cases, where the ventral surface is sharply distinguished from the
dorsal, and the dorsal ambulacral appendages, i.e. those of the bivium, have entirely
disappeared, the corresponding dorsal radial canals are said also to be wanting. In
a few isolated forms the central radial canal of the ventral side (i.e. of the trivium)
is also said to be wanting.

The tentacle canals branch off from their radial canals just above the calcareous
ring. Their number corresponds with that of the tentacles to which they run.
These canals are often connected, at the anterior edge of the ring, with tentacle
ampullae (Fig. 365, 18). These latter are tubular outgrowths, which vary greatly in
size, stretching back over the outer surface of the calcareous ring, and for the most
part projecting freely into the body cavity. Where such ampullae occur, all the
canals without exception are provided with them. They are entirely wanting in the
families of the Elasipoda and Dendrochirota, but occur normally in the Synaptidce,
Molpadiidce, and Aspidochirota. In Pelagothuria, branches run through the peculiar
swimming disc (cf. p. 286), radially, and reach even to the tips of its processes.
They are evidently to be regarded as modified tentacle ampullae.

The canals of the tube-feet branch off alternately from the radial canals. As a
rule, a separate canal runs from the radial canal to each foot ; but in some cases
(Holothuria tululosa) one canal, by branching, runs to several (4-6) tube-feet. In
the Molpadiidce, and the above-named Holothurian, it is said that there are tube-feet
canals which end blindly, and thus have no tube-feet answering to them. Except
in the footless Molpadiidce and the Psychropotidce, the tube-feet canals are connected
with egg-shaped, often somewhat long and occasionally branched ampullae. These
either lie as covered ampullae outside of the circular musculature of the body wall or,
as free ampullae, press in between the transverse musculature into the body cavity.

At the point where the ampulla opens into the tube-foot canal, but in that part
of the latter which comes from the radial canal, there is a valve, similar to that
found in Asteroids, which will be described later. This valve is arranged in such a
way as to prevent the return of the fluid into the radial vessel, either from the foot
or from the ampullfe. Valves are also found in the tentacle canals.

The walls of the ampullae, resemble those of the Polian vesicles in structure.
The radial canals and their branches are chiefly distinguished by the fact that the
longitudinal musculature is only developed in the outer part of the walls.

Paractinopoda. — The tentacle canals here spring directly out of the circular
canal, and nearly always agree in number with the tentacles. At the level of the

428

COMPARATIVE ANATOMY

CHAP.

calcareous ring, in each tentacle canal a muscular membrane forms a semilunar
valve which projects from the wall with its concave side directed forwards (orally).

FIG. 365.— Section through the oral region of an Actinopod, in the direction of the principal
(longitudinal) axis. On the right, the plane of the section is radial ; on the left, almost interradial.
1, Cutis ; 2, body epithelium ; 3, oral tentacle, cut oft' ; 4, water canal of the oral tentacle ; 5, bin. id
vessel of the oral tentacle ; 6, tentacle nerve ; 7, circular nerve ; S, oral portion of the coeloinatic
perioesophageal sinus; 9, mouth; 10, oesophagus; 11, perioesophageal sinus; 12, interradial
portion of the calcareous ring ; 13, water vascular ring ; 14, blood vascular ring ; 15, ventral
intestinal vessel ; 16, intestinal epithelium ; 17, Poliau vesicle ; IS, ampulla of the oral tentacle ;
19, endothelium of the body cavity ; 20, circular musculature of the body wall ; 21, body cavity ;
22 and 26, radial blood vessels ; 23, radial nerve trunk of the superficial system ; 24, radial epineural
canal ; 25, radial perihtemal canal ; 27, radial canal of the water vascular system ; 28, longitudinal
muscles ; 29, commencement of the radial canal of the water vascular system ; 30, radial portion of
the calcareous ring ; 31, retractor muscle ; 32, dorsal intestinal vessel.

This valve prevents the water vascular fluid flowing back out of the tentacles into
the circular canal.

The wall of the tentacle canals consists, from without inwards, of : (1) the eudo-
thelium of the body cavity'; (2) a longitudinal muscle layer ; (3) a layer of connective
tissue ; (4) a circular muscle layer ; (5) an inner epithelium.

vin ECHINODERMATA— WATER VASCULAR SYSTEM 429

2. Echinoidea (Fig. 358, p. 419). — In the Spatangoida, where a
masticatory apparatus is wanting, the radial canals, on leaving the
circular canal which surrounds the mouth, are already in their
respective radii, and commence at once to send out branches right
and left to the tube-feet. In other Echinoidea, however, the radial
canals have to descend from the circular canal which encircles the
O3sophagus above the lantern to the peristome, and thus, after rising
out of the circular canal, have first to run under the intermediate
plates and over the intermaxillary musculature. They emerge at the
periphery of the lantern and then descend on its outer side, i.e.
outside the intermaxillary musculature to the peristome. Having
reached this latter, they first give off a branch which runs in the oral
region towards the mouth. They then pass through the auriculae,
in order to run up radially towards the apex, on the inner side of
the test, and in the middle lines of the ambulacra. They end
blindly in the pores of the radial plates of the apical system.

While running up on the inner side of the test, the radial canals
give off alternating lateral branches, each of which enters an ampulla
(cf. Fig. 353, p. 410). The ampulla, which projects into the body
cavity, is itself connected, by means of one or two canals, with the
cavity of a tube-foot or tentacle, which latter projects freely on the
outer side of the test. The ambulacral plate at such a point is per-
forated by either a single or a double pore, according as the canal to the
tube-foot is single or double (cf. the section on the Skeletal System).

In all Echinoidea, the tube-feet in young animals are all alike, and each is
connected with its ampulla by a single pore through the test. This may be
considered to be the primitive arrangement. Tube-feet with single pores are found
in adults in a few Spatangoida : in the Pour talesi idee, in the Ananchytidan genera
Urecliinus, Cystechinus, Calymne, in the Spatangoid genus Palaeotropus and the
Cassidulid genus Ncolanyms.

In all other Echinoids, the tube-feet or tentacles have double pores. In the
regular Echinoids (Cidaroida, Diadcmatoida], only double pores are found : but in
the Clypcastroida and the Spatangoida, only the pores of the petaloids are double :
those on the remaining ambulacral regions being single.

The ampulla; are delicate structures which vary in shape. In cases in which
they, like the tentacles to which they belong, stand at some distance from one
another, they are pear-shaped or spherical ; but where they, like the tube-feet, stand
in compact rows in the ambulacral meridians, as in the regular Echinoids and in the
petaloids of the irregular forms, they are lengthened out horizontally and flattened
vertically (dorso-ventrally). The walls of the ampullse, from without inwards,
consist of: (1) a ciliated endothelium ; (2) a layer of connective tissue, containing
occasional embedded calcareous corpuscles ; (3) a circular muscle layer ; (4) an inner
ciliated epithelium. The lumen is traversed from wall to wall by fibres, which are
probably muscular. In several Echinoids, at the points where the lateral canals of
the radial canals open into the ampullse, valves have been observed.

The branches of the radial canal which run in the oral integument supply the
tube-feet or tentacles occurring in this region.

3. Asteroidea. — The radial canals, in this class, run along the

430 COMPARATIVE ANATOMY CHAP.

bases of the ambulacral furrows of the arms, outside the ambulacral
plates. At the tips of the arms they end blindly in the terminal
ocular tentacles. In their courses, consecutive widenings and
narrowings are not infrequently found, these correspond with the
segmentation of the arm, but are never very marked. Each radial
canal gives off — at regular intervals, which correspond with the
skeletal segments, and at opposite points to right and left — canals to
the tube-feet. At the point where such a canal opens into the tube-
foot, a second canal, the ampulla canal, branches off from it. This
canal rises up between two consecutive ambulacral plates to widen out
above these latter into an ampulla which projects freely into the body
cavity (Fig. 354, p. 411).

This ampulla is single in all young Asteroids and many adults (Linckiidcc,
Echinasteridcc, Asteriidce, Luidia). In other Asteroids (Astropcdinidce excluding
Luidia, Asterinidce, Pentacerotidix, e.g. Culcila) two separate ampullae occur to
each tube-foot in the adult.

Valves are found at the points where the canals of the tube-feet
open into the radial canal. A muscular membrane, resembling a
truncated cone with the base attached horizontally round the wall of
the canal, projects into the lumen directed towards the foot. This
valve prevents the fluid pressed out of the ampulla from returning
into the radial canal, either because the membrane is able by muscular
action to close the aperture, or because the pocket surrounding this
projecting membrane is swelled up by pressure of water from the
foot or ampulla, and so closes the valve.

4. Ophiuroidea. — The first point to be noted with regard to
the Ophiuroidea is that they have no tube-feet ampullae.

The radial trunks of the water vascular system run in the arms
between the ventral shields and the vertebral ossicles. At the tip
of the arm each trunk ends in a small terminal tentacle. Regularly
consecutive and distinct widenings are found in their courses
corresponding with the regular segmentation of the arms. Between
every two of these consecutive widenings, the radial canal is provided
with a single layer of band-like circular muscle fibres. A narrow
tube-foot canal runs off to right and left from each widening, running
either straight into its tentacle or first forming a V-shaped loop, which
ascends apically into the calcareous mass of the vertebral ossicle. At
the point where the tentacle canal enters the tentacle, the lumen of
the former becomes much widened, and a valve occurs (similar to that
described in the Asteroid*), which prevents a flowing back of the
water vascular fluid out of the tube-foot into the radial canal.

The first two pairs of canals to the tube-feet or tentacles (the so-
called oral tentacles) come direct from the circular canal.

5. Crinoidea. — Tentacle ampullae are wanting-. The radial
canals lie close under the food grooves of the disc, of the arms, and of
the pinnulae, whose courses they exactly follow, so that they branch just

vin ECHINODERMATA— WATER VASCULAR SYSTEM 431

as often as do the arms and their food grooves. Their course is more
or less markedly zigzag, and they give oft* at the angles thus formed
(i.e. alternately) lateral tentacle canals. Each of these latter runs to
a group of three small tentacles at the edge of the food groove, and
here divides into three canals, which enter the three tentacles and
form their cavities.

Tentacle canals are wanting in all cases where food grooves are
wanting, which is the case in Actinometm over a great part of the
arms, and in some species of Antedon in certain proximal pinnules of
the arms.

All authors agree in maintaining that the inner epithelium of the water vascular
system in the Crinoids differs from that in all other Echinoderms in not being
ciliated. A band of longitudinal muscle fibres runs in the wall of the canals along
the side turned to the food groove. The lumen of the canals is at certain points (i.e.
where the tentacle canals branch, or at the commencement of these canals) traversed
by muscular fibres. This arrangement perhaps fulfils the function of the valves
found in other Echinoderms.

D. The Ambulaeral Appendages.
(Tube-feet, Tentacles, Feelers, Ambulacral Papillse, etc.)

1. Holothurioidea. — The following facts require first of all to be
emphasised.

a. In all Holothurioidea, a smaller or greater number of ambulacral
appendages (10-30) are developed as tentacles near the mouth.

b. The Synaptidce and Molpadiidce have no ambulacral appendages
except these tentacles.

f. In all other Holothurioidea besides the tentacles there are
tube-feet (and papillae) varying greatly in number (often very
numerous), in structure, and in arrangement.

d. These tube-feet (and papillae) are found either only on the
radii, one or two or more longitudinal rows being arranged in each
radius, or else they are distributed, usually in an irregular manner,
over some or all of the interradii. The arrangement of the tube-feet
is not of great systematic importance, since even within one and the
same genus (e.g. Cucumaria), all the intermediate stages between a
strictly radial and an altogether scattered arrangement can be
observed.

e. Where the ventral and the dorsal surfaces are distinctly differ-
entiated, the ambulacral appendages are developed on the ventral side
(in the trivium), normally as loeomotory tube-feet with sucking discs
supported by perforated plates : on the dorsal side, on the other
hand, they take the form of conical non-locomotory papillae, which
have either a rudimentary perforated plate at the narrow tips or none
at all.

432 COMPARATIVE ANATOMY CHAP.

No very sharp distinction between tube-feet and papillae is, however, possible,
either with regard to their distribution, their form, or their structure.

With regard to the number of the tentacles, the following numbers seem to
prevail in the different families : 20 in the Aspidochirotos, 20 in the sub-family
Deimatidce of the Elasipoda, 15 in the Molpadiidcc, 13-16 in the Pelagothuridce, 12
in the Synaptidcc, and 10 in the Dendrochirotce and in the sub-family Elpidiidw
of the Elasipoda.

With regard to form : the tentacles are feathered (Molpctdiidce Synaptidce, Fig.
229, p. 288), dendriform (Dendrochirotce, Fig. 226, p. 287), and shield - shaped
(Aspidocliirotcc, Elasipoda}. In the latter, the disc or shield, the edge of Avhich may
be more or less deeply indented, is carried by a stalk.

The size of the tentacles has already been sufficiently indicated in the systematic
review.

The relation between the arrangement and size of the tentacles on the one hand
and the symmetry of the rest of the body on the other is interesting. In the
Dendrochirotce (cf. Fig. 226, p. 287), of the ten tentacles, the two ventral are almost
always distinguished by being much smaller than the rest.

In many species of Myriotrochus, Synapta, and Chirodota with twelve tentacles,
these are distributed symmetrically as follows : three occur in each of the two
dorsal interradii, and two in each of the three ventral interradii.

The tentacles may be swelled and extended ; and on the; other hand they can be
withdrawn into the body cavity together with the surrounding anterior part of
the body, although not invaginated like the tentacles of a Gastropod.

2. Echinoidea. — Ambulacral feet are developed in all Echinoids
without exception. In early youth, they are always found to resemble
one another, and in both the Echinidce and the Pourtalesiidcc this
is still the case in adults, the former having tube-feet with terminal
suckins discs and the latter tube-feet with rounded ends. In most

O

Echinoidea, on the contrary, more or less marked polymorphism
occurs, division of labour taking place between the ambulacral
appendages of one and the same individual.

This polymorphism is not very striking in the regular Echinoidea,
e.g. the Cidaroida, Eclunotlmriil*/', Diadematidce, Arbaciidce, Echinometridce,
etc. In these, the ambulacral appendages appear, as a rule, in three
different forms : (1) as loeomotory tube-feet with terminal or sucking
discs; (2) as tactile or branchial tentacles without terminal sucker;
and (3) as oral or sensory feet with bi-lobate terminal disc.

All these tube-feet are connected by means of double pores with their ampullae,
which lie within the test. Without detriment to their principal function, they
may all act as respiratory organs, since the presence of the double pore allows of
a circulation of the ambulacral fluid between the inner ampulla and the outer
ambulacral foot ; the fluid in the foot takes in oxygen, carries it back into the
ampulla, and gives it off through the wall of the ampulla to the fluid in the body
cavity.

The loeomotory tube-feet are found on the oral hemisphere of the body, but
may also occasionally occur on the apical hemisphere as well.

The tactile or branchial tentacles are limited to the apical hemisphere. They
are specially suited for respiratory purposes when the ampulla? are large, and have
thin and delicate walls containing no calcareous corpuscles.

viii EGHINODERMATA—WATER VASCULAR SYSTEM 433

The oral tube-feet (always ten in number ?) surround the mouth, and especially
when food is being taken in, are subject to active swinging or pulsating movements;
without, however, touching the food. They have been regarded as olfactory or
gustatory organs.. They seem to be wanting in the Cidaroida and the Echinutlnn-iiiii' ;
on the other hand they occur in those Echinidce which otherwise possess only one
sort of tube-feet, viz. those with sucking discs.

The polymorphism of the ambulacral appendages is much more
marked in the Cli/pcastroidit, and the Spitamjoirfa. It must first be
noted that the ambulacral appendages, in those apical regions of the
ambulacra which are known as petaloids (cf. p. 347), serve for respira-
tion (ambulaeral gills). They seem to be peculiarly fitted for this
activity by the delicacy of their walls, the want of calcareous corpuscles,
the increase of surface obtained by branching", the possession of
double pores (whereas the ambulacral appendages in other parts of the
body have single pores) and by the size of their ampullae.

In the Clijpeastroida, besides the ambulacral gills of the petaloids, three kinds of
appendages have been observed : (1) the ordinary slender tube-feet, with rounded
terminal knobs, scattered on the test ; (2) sessile knobs, with deep sensory

FIG. 300.— Longitudinal section through an ambulacral brush of a Spatangoid (after Loven
and Hamann). 1, Body epithelium ; 1, supporting rod ; 3, supporting plate of the terminal disc ;
4, septa ; 5, canal of the water vascular system ; 0, longitudinal muscles ; 7, nerve ; 8, circular
muscle fibres.

epithelium (sensory tentacles) ; (3) short, thick tube-feet with truncated ends : these
occur between the ordinary feet on the oral side, and are perhaps locomotory.

Among the Spatangoida, the polymorphism of the ambulacral appendages is very
marked in all the divisions except in the Echinonchhe ; it reaches its highest point
in the families of the Spatangidce and Apctala.

The ambulacral gills of the four paired petaloids have been described above.
We note first the characteristic ambulacral brushes which occur in the Spatangoida
more or less near the mouth and the anus, and in the Cassiduloida on the phyllodes
(<•/. p. 347). The terminal plate or disc of an ordinary tube-foot (Fig. 366) is here
extraordinarily widened, and carries a number (usually large) of club-shaped or
conical, solid appendages, each of which is supported by a calcareous rod. These
ambulacral brushes are said to play an important part in the taking in of food by
VOL. II 2 F

434 COMPARATIVE ANATOMY CHAP.

stirring up the sand. On other parts of the ambulacra, slender tentacles without
prehensile discs occur, to which a tactile function has been ascribed. Still more
interesting are the ambulacral appendages of the anterior unpaired ambulacrum.
which are certainly to a still higher degree tactile organs. These vary in shape ; in
all young Spatanyoida and many adults they are distinguished by their remarkable
size, and help to emphasise the bilateral symmetry of the whole body. In Sjinfa/tyiix
and other genera they end in a flat disc, the edge of which is drawn out into short,
solid, knobbed processes, which are supported by calcareous rods. The whole
terminal disc thus looks like a beautiful rosette.

As compared with the ordinary tube-feet, truly gigantic proportions are attained
in the genera Aceste and ^-Eropf- by the ambulacral appendages, which lie in the
depressed anterior ambulacrum within the peripetaloid fasciole. They are found
only in small numbers, yet in their contracted condition they almost completely fill
the depression from whose base they rise. Their ends are provided with large discs.

Turning to the finer structure of the ambulacral appendages of the Echinoidea.
their wall is found to consist of the typical layers. In the locomotory tube-feet of
the regular Echinoids, the inner layer of the connective tissue is specially modified
as an elastic membrane, with circular fibres. Calcareous corpuscles are wanting
only in the respiratory tentacles of the apical surface of the body. In all other parts
of the body they are found in great numbers in the stalk, while in the terminal discs
of the tube-feet they take the form of delicate, circular, terminal plates usually com-
posed of several pieces. The whole surface of the tentacles, except that of the terminal
discs, is ciliated. A nerve enters each tube-foot, running in the epithelium until
near the tip, where it forms a lateral ganglion, which can be externally recognised as
a swelling. From this ganglion, the terminal apparatus of the tube-foot is innervated.
In tube-feet with terminal discs, the epithelium at the edge of the disc is differen-
tiated as a deep sensory epithelium, and within it runs a basal nerve ring, connected
with the lateral ganglion by two nerves. Where the tentacles end in knobs (tactile
tentacles and sessile knobs of the Clypeastroida) or carry knobbed processes on their
terminal discs (ambulacral brushes, rosette-like tube-feet of the anterior unpaired
ambulacrum of the Spain ,/,/,,;, In these knobs are caused by the great thickening of
the sensory epithelium. In the ambulacral gills of the Clypeastridce (Eclti-niM.-yiimnx.
Echinodiscus) the epithelium occasionally thickens to form sensory papillae. The
sensory epithelia appear everywhere to carry stiff sensory setre or hairs.

The lumen of the tube-feet at its central and basal parts is not infrequently
traversed by transverse muscle fibres ; occasionally it appears to be double — a septum
consisting of transverse bands continues for some distance into the tube-foot, the
partition in the test between the two apertures of the double pore. The cavity of the
large terminal discs of the paint-brush tentacles in the Spatangoida is traversed by
concentric, much-perforated septa (Fig. 366). The ambulacral gills and their ampulla'
are traversed by bands arranged radially around an axis, in such a way that the.
fluid contained is forced to circulate at the periphery.

3. Asteroidea. — The ambulacral appendages always take the form
of tube-feet, and stand in two or four longitudinal rows in the ambu-
lacral furrows which run from the mouth to the tips of the arms. It
has already been pointed out (p. 353) that the tube-feet, even when
apparently present in four rows, in reality belong only to two. That
there are only two rows is very clear in young animals. In young
Asteroids all the tube-feet are alike ; all have conical ends, with
rounded tips. This is still the case in many adult Asteroids

VIII

ECHIXODESMA TA — V 'A TER 1 'A SCULAIl S FS TEM

135

pecten, Lui'/in, etc.), while in very many other genera (<'.</. A*t<'ri<ix,
tfulasti'i-} a slight difference in shape occurs. Only the tube-feet near
the ends of the arms retain the primitive shape, while a well-developed
disc is found on all the rest. The former then function chiefly as
tactile tentacles.

The wall of the tentacles shows the typical layers. In the tactile tube-feet, the
epithelium at the conical end is much thickened, and contains very numerous
sensory cells. Within the epithelium, a layer of nerve fibres is developed ; these
run from the base to the tip of the foot. The layer is specially strongly developed

t

,x

i \ WN&rftStL/ v' r^ty

r.

FIG. 367.— Portion of the disc of Hemipholis cordifera, from the oral side (after Lyman). 1,
oral tube-feet ; 2, buccal shields ; 3, jaw = oral-angle plates ; 4, lateral buccal shields ; 5, first ventral
shield ; 6, oral integument, lip ; 7, spines on the marginal plates ; S, retracted tentacle ; t>, tentacle
scale ; 10, ventral shields ; 11, extended tentacle ; 12, tentacle pore ; 13, torus angularis ; 14, tc.-tli.

within the terminal sensory epithelium. A similar deep sensory epithelium, consist-
ing of sensory, supporting, and glandular cells, also covers the sucking discs of the
other tube-foot ; these discs have a depression at their centres, while, round their
ed<?es the nerve tissue lying within the epithelium becomes thickened into a nerve
ring.

From the centre of each sucker, radial muscle fibres run out towards the periphery
and are attached round the ambulacra! canal which ends below the sucker. These
muscles, by their contraction, cause the sucker to adhere. They are entirely dis-
tinct from the longitudinal muscles of the appendage, which explains the fact that
a tube-foot sticking to an object may be cut oft' without becoming detached.

This applies almost equally well to the sucking discs of the tube-feet of the
Echinoids.

436 COMPARATIVE ANATOMY CHAP.

4. Ophiuroidea (Fig. 367). — In the Ophiuroidea, the ambulacral
appendages have no locomotory significance ; they resemble tentacles,
and never have suckers. Locomotion is caused by the jointed arms
themselves. The tentacles are always strictly segmentally arranged, i.e.
one pair of tentacles occurs on each brachial segment. Each of these
emerges through an aperture between the ventral shield and the lateral
shield of a segment. The tentacles are not infrequently covered with
a large number of sensory papillae. A nerve, coming from the basal
circular ganglion and running in the deeper portion of the layer of
connective tissue, traverses the tentacle from base to tip.

It has already been mentioned above that the first ten pairs of
tentacles (i.e. the first two pairs of each arm) have shifted, as oral
tentacles, to a position round the mouth, and receive their canals direct
from the circular canal.

5. Crinoidea. — The small tentacles which form the ambulacral
appendages of this class have already been sufficiently noticed (p. 414).
They never possess suckers, and have no locomotory function, but
simply serve for respiration, and for conducting food to the mouth.

VII. The Ccelom.

fThe Enteroccel, the true or secondary Body Cavity.)

All those cavities of the body which are derived from the entero-
ecelomie vesicles of the larva are considered to belong to the ccelom,
which is lined throughout with endothelium, usually developed as
ciliated epithelium. The ecelomie fluid exactly resembles in constitu-
tion the water vascular fluid already described. The crelom is, how-
ever, except at one single point to be mentioned later, altogether
separate from the ambulacral vascular system.

The ccelom is never found as a single cavity, but is always divided
into several cavities, which may be entirely distinct one from the
other. The largest of these cavities is that which contains the viscera,
and which may be termed simply the body cavity.

The body cavity is most spacious in the Echinoidea and the HolotJ/it-
rioidea ; in these forms it occupies almost the whole cavity of the test,
or the sac- or tube-shaped body. In the disc of the Asteroidea it is
somewhat less spacious, and is very limited in that of the Ophiwroidea,.
In the Crinoidea, it is traversed by a more or less strongly calcified
network of connective tissue.

Where the body is drawn out into arms in the radii, the body
cavity runs into these, and forms the braehial cavities. The brachial
cavities in the Asteroidea are very spacious, but are much narrowed in
the Ophiuroidea and the Crinoidea, owing to the great development of
skeletal plates (vertebral ossicles, joints) in the arms.

A special section of the ccelom, the pericesophageal sinus (peri-
pharyngeal sinus) encircles the oesophagus or pharynx. In the Echi-

viii EGHINODERMATA—THE CCELOMIC CAVITIES 437

noidea, this is quite cut off from the body cavity. The membrane
which separates the body cavity from the perioesophageal sinus is
called, in those Echinoids which are provided with a masticatory frame-
work (Cidaroida, Diadematoida, and Clypeastroida), the lantern mem-
brane. This membrane entirely covers the lantern on the side turned
to the body cavity.

In many Echinoids, this part of the cceloni protrudes externally
at the edge of the peristome, forming the outer gills : in others,
the lantern membrane bulges out into the body cavity, and forms
Stewart's organs.

In the Holothurioidea and Echinoidea, the hind-gut is surrounded by
a small coelomic sinus, the perianal sinus.

In the EcJiinoidea, Asteroidca, and Ophiwroidea, a part of the coelom,
cut off from the rest, runs from the region of the madreporite inter-
radially to the circular canal of the water vascular system. This is
the axial sinus in which the stone canal runs. It contains also a
lymph gland, the so-called ovoid gland or axial organ.

The axial sinus, in the Ecliinoiih, is in open communication with
the ampulla which lies beneath the madreporite. Eecent ontogenetic
researches have shown that this ampulla also is of enteroccelomic origin,
and is thus a section of the coelom. Since the stone canal opens into
the ampulla, an open communication exists at this point, and at this
alone, between a closed division of the ccelom (the axial sinus) and a
section of the water vascular system (the stone canal).

A. The Body Cavity.

1. Holothurioidea. — The spacious body cavity of the Holothurioidea
is divided up by the mesentery which attaches the intestine to the
body wall. This mesentery may be described as having three parts
corresponding with the three sections of the intestine, viz. the dorsal
mesentery, belonging to the first section of the intestine which runs
backward ; the left dorsal mesentery, belonging to the second
section, which bends forward ; and the right ventral mesentery,
belonging to the third section; which runs backward to the cloaca.
All three parts of the mesentery lie interradially.

The so-called ciliated urns or funnels (Fig. 368) are found only
in the Syna^ti<ln'. These are funnel-, cup-, or slipper-shaped organs,
each of which is attached to the body wall or the mesentery by a stalk,
and hangs down freely into the body cavity. They are specially
numerous to the right and left of the dorsal mesentery. In L'lti roil ,4,1.
many funnels have one common stalk, and so form trees of ciliated
funnels.

The funnel consists of three layers, an outer endothelial tessellated epithelium, a
middle and extremely thin layer of connective tissue, and an inner layer of columnar
epithelium, which lines the lumen and carries long cilia. Towards the stalk
the lumen is closed, but it is open towards the body cavity.

438

COMPABA TIVE A XA TuMY

CHAP.

The vigorous movements of the ciliated urns no doubt serve to promote the
streaming and circulation of the fluid in the body cavity.

2. Echinoidea. — In the Echinoids, the body cavity is partitioned
in a manner similar to that described for the H<ilntlturini<l<:n, by mesen-
teries which follow the intestine
in its windings (see p. 480), and
attach it to the inner surface of
the test. The genital orsrans

O O

are also attached to the test by
mesenteries. In regular Echin-
oids, the mesenteries are much
perforated, but are only slightly,
if at all, broken through in the
Spatangoida.

In this latter order, where
the mesenteries have to carry
the heavy intestine, filled with
sand, they are specially strong
and tough. The coils of the
intestine are here also united
inter se by mesenteries. Special
bands attach the intestine to the
apical and oral poles of the test,

PIG :;r,s. -Ciliated urns of a Synaptid (after internal processes Or apophyses
Cuenot). 1, mesentery; _, circular muscle layer .. . , IIP

of the body wall ; :.;, ciliated inner epithelium" of being Sometimes developed tor

HIH urn : 4. mdotheiium of the body cavity ; the attachment of the bands.

T™ such apophyses are found
at the apical pole, at the end of
the stone canal, and a third not infrequently occurs at the peristome,
in an interradius.

The axial sinus, with the axial organ and the stone canal, is at-
tached by bands on the one hand to the apical pole, and on the
other to the oesophagus.

For a description of the calcareous pillars, septa, etc., which, in the Ct>/j>C((Sti-ii/n:
traverse the cavity of the test, see p. 405.

In the fluid of the body cavity in Echinoids there are found, besides blood cor-
puscles, great numbers of spermatozoa-like cells, with long flagella in vigorous move-
ment. These may set up currents in the fluid of the body cavity.

•3. Asteroidea. — The body cavity of the disc is not spacious, the
greater part of it being filled by the large digestive sac. Mesenteries
are wanting in the greater part of the intestine, or are only developed
as isolated filaments or strands of connective tissue. In the peripheral
portion of the disc, radially placed bands or septa traverse the body
cavity vertically in the interradii, connecting the dorsal (apical) body
wall with the ventral (oral) wall.

E~CHINOD£RMATA—THE OCELOMIC CAVITIES 439

The lymph gills, branchial vesicles or papulse, which are only
found in the Asteroids, deserve attention. These are small vesicular
bulgings of the body wall, which occur in great numbers between the
skeletal plates. On these bulgings, the body wall is very thin, and, in
order to facilitate osmosis, devoid of calcareous deposits. It consists
of layers similar to those found in other parts of the body : an outer,
strongly ciliated and glandular epithelium ; a middle layer of con-
nective tissue, containing longitudinal and circular muscle fibres ; and
an inner ciliated epithelium, which is nothing else than the endothelium
of the body cavity. The cavities of the papula? are merely diverticula
of the body cavity which bulge out the much-thinned body wall.

The papula' are sensitive, and contract at the slightest touch.

In Asteroids with very thick body wall, diverticula of the body cavity force their
way into it, branching on their way to the surface. On reaching this latter, each
branch enters a branchial vesicle.

In certain forms, a single di vertical nm traversing the body wall supplies a whole
group of branchial vesicles.

A constant streaming to and fro of the body fluid can easily be observed in the
branchial vesicles.

Each branchial diverticulum of the body cavity is surrounded, in the connective
i isMu- layer of the body wall, by a circular lacuna.

The branchial vesicles occur both on the arms and on the disc. In the PJiaiiero-
:<>iiin they are found only on the upper side of the body ; in the Cryptozonia, on the
rniitraiy, they occur on the sides of the arms as well, and on the lower (oral) side of
tin' body.

4. Ophiuroidea. — The body cavity of the disc is much limited
by the digestive sac and the bursse. Filaments and bands of con-
nective tissue, covered with endothelium, traverse it at irregular
intervals, and connect the viscera with the body wall.

.">. Crinoidea. — The body cavity of the calyx is almost completely
filled with bands, trabecula?., filaments, etc., of connective tissue covered
with endothelium, which together form a spongy network, and often
become calcified. In this network a sac -like membrane becomes
differentiated, which divides the body cavity into a central and a
peripheral space. The central space, which contains the intestine, is
known as the peri-intestinal cavity ; the peripheral, as the subtegu-
mentary cavity. The peri-intestinal cavity, again, contains another
separate portion of the coelom, the axial body cavity, round which the
intestine coils itself. This latter cavity encloses the genital stolon
and communicates, on the one hand, with the five chambers of the
chambered organ which lies in the apex of the calyx, and through it
with the ccelomic canals of the stalk and the cirri ; and on the other,
with the oral or subtentacular canals of the arms. The peri-intestinal
cavity, on the contrary, is continued into the dorsal or apical brachial
canals.

440 COMPARATIVE ANATOMY CHAP.

B. The Brachial Cavities.

1. Asteroidea. — The body cavity of the disc is produced in the
form of large cavities which run along the arms to their tips. The
body cavity is thus found throughout the whole of that part of the
arms which is enclosed by the skeletal plates, and contains (1) the
ampullae of the water vascular system, (2) the two radial creca of the
stomach, and finally (3) a part of the genital glands as well. The
radial caaca of the stomach (two of which occur in each arm) are each
attached to the dorsal wall of the arm by two suspensors, which run
in the longitudinal direction, so that, above each caecum there lies a
coelomic canal whose walls are formed (1) dorsally by the brachial
wall, (2) ventrally by the wall of the caecum, and (3) laterall}' by the
two suspensors (cf. Fig. 354, p. 411).

2. Ophiuroidea (Fig. 355, p. 412). — The vertebral ossicles occupy
so large a part of the transverse section of the arm, that only a very
small space is left for the brachial ccelom. This latter is found as a
flat cavity below the dorsal wall of the arm. It is divided into con-
secutive chambers, which agree in number with the segments of the
arm ; these chambers are incompletely separated from one another by
transverse, vertical, calcareous septa, which connect the vertebral
ossicles with the outer skeletal plates of the arm. These septa leave
a mediodorsal space free, through which all the chambers are in open
communication with one another. This is the " dorsal canal " of
authors.

The endothelium of the brachial cavity is thickened in the dorsal middle line,
and carries specially strong cilia. There is thus in each arm a longitudinal ciliated
band or streak, which, when it reaches the disc, passes into the ordinary endo-
thelium. The activity of these strong cilia originates and maintains the circulation
of the body fluid in the brachial cavity. Occasionally the ciliated streak is deepened
into a groove.

3. Crinoidea (Fig. 356, p. 413). — In this class also the brachial
cavity is much reduced by the strong development of the skeletal
joints of the arms ; unlike that of the Ophiuroidea, however, it is
found on the ventral side of the arm. The cavity is divided by a
horizontal longitudinal septum into two canals, one lying above the
other ; both of these run through the arm and its branches as far as
to the tips of the pinnulaj. The dorsal (apical) ccelomic canal is
known as the dorsal, and the ventral (oral) canal as the ventral or
subtentaeular canal. The latter, on reaching the disc, enters the
axial co?lom, the former enters the peri-intestinal cavity. The ventral
canal itself, again, is divided by a vertical longitudinal septum into two
lateral canals.

The endothelium of the dorsal canal occasionally (especially in the pinnulse)
shows small sac-like bulgings, which are known as ciliated baskets or sacs. The

VIII

ECHINODEEMA TA — THE CCEL OMIC <. 'A } '1 TIES

441

floors of these sacs consist of flat cells, but the epithelium round their openings into
the dorsal canal is much thickened, and provided with large cilia. These ciliated
sacs undoubtedly serve the same purpose as the ciliated bands in the brachhl
cavities of the Ophiuroidca and the urns of the S

C. The Perioesophageal Sinus.

1. Holothurioidea (Fig. 3G5, p. 428). — Between the mouth and
the water vascular ring the oesophagus is surrounded by a membrane-
like sheath, in such a way that between it and this membrane a
narrow cavity is left ; this is the pericesophageal sinus, which is a
section of the coelom. The radial canals of the water vascular system
run along its outer side, and in the radial direction it is traversed by
numerous bands and filaments which are attached on the one hand to
the oesophageal wall, and on the other to the outer membrane of the

!=* Ra?

— 7

-10

FIG. 309.— Median section through the oral region of Spatangus purpureus (after Cuenot).
Ig, Posterior unpaired interradius ; Em, anterior unpaired radius. 1, Radial canal of the water
vascular system ; 2, radial blood vessel ; 3, radial pseudohaenial canal ; 4, radial nerve trunk ; 5,
radial epineural canal ; (}, test ; ~, body epithelium ; S, mouth ; '.», epineural circular canal ; 10,
circular nerve ; 11, endothelium of the body cavity ; 12, pericesophageal sinus ; 13, water vascular
ring ; 14, blood vascular ring ; 15, oesophagus ; 16, membrane which separates the pericesophage.il
sinus from the body cavity; 17, septum, which separates the pseudoha-mal canal from the periieso-
phageal sinus.

sinus. These are only wanting in the part near the mouth, which is
thus distinct, and is known as the peribuceal sinus. The perioeso-
phageal sinus is usually in open communication with the common body
cavity by means of a varying number of apertures in its outer mem-
brane. In Cucumaria there are five such apertures, which are large
and interradial in position. In the Elasipoda alone is the periceso-
phageal sinus completely separated from the general body cavity by
an uninterrupted outer membrane, which runs from the water vascular
ring direct to the body wall.

2. Echinoidea. — In the Spatangoida (Fig. 369) a membrane,
which runs out horizontally from the commencement of the intestine
to surround the circular canal of the water vascular system, completely
separates a very small perioesophageal sinus of the ccelom from the
spacious body cavity. In those Echinoids which are provided with a
masticatory framework, this latter develops within this sinus, which

442 COMPARATIVE ANATOMY CHAP.

thus attains a large size. The membrane which completely separates
the perioesophageal sinus from the general body cavity then becomes
the lantern membrane, surrounding the masticatory apparatus on all
sides, from the point where the intestine leaves the lantern down to
the perignathous apophysial ring. The greater part of the periceso-
phageal sinus (Fig. 358, 44, p. 419) is filled by the masticatory frame-
work, the remaining space being traversed by trabeculse, bands, etc.
All the radial organs arise from within the pericesophageal sinus,
running in it as far as to the auricuhe.

The outer gills and Stewart's organs of the Echinoids are develop-
ments from the pericesophageal sinus.

The outer gills (Fig. 358, 1) consist of five pairs of branched
appendages, which rise at the periphery of the oral region, at the
inner edge of the oral integument, and project freely outward. One
pair of such gills occurs on each interradius. The peristomal edge of
the disc is indented at certain points for the reception of the gills, so
that their presence or absence may be determined by the presence
or absence of these incisions on the test of either extant or fossil
Echinoids. The gills are hollow outgrowths of the oral integument,
their cavities being direct prolongations of the pericesophageal sinus,
and thus in open communication with this latter ; the body fluid of
the sinus can thus enter the gills and flow back into the sinus again.
The wall of such a gill consists of a deep outer epithelium provided
with long cilia, a central layer of connective tissue with calcareous
corpuscles and lacunae, and an inner ciliated covering of endothelium.

External gills occur in most endocyclic (regular) Echinoids. They
seem to be wanting only in the Cldaroida.

Stewart's organs. — Just as external bulgings of the oral integu-
ment lead to the formation of external gills, internal bulgings of the
lantern membrane into the body cavity give rise to Stewart's organs.
These are delicate-skinned vesicles or tubes, which vary greatly in size.
Five are usually present, projecting into the body cavity from the
edge of the (apically directed) base of the masticatory framework,
immediately below the fork plates or radii between these and the
falces or intermediate plates (cf. p. 400). The cavity of the vesicle
is a diverticulum of the perioesophageal sinus.

The Stewart's organs of the Ciilaroirla are large, and are usually beset with
secondary outgrowths. Since the Cidni-i.iiiln possess no gills, it has been conjectured
that Stewart's organs fulfil the functions of the absent external gills, and they have
therefore been called internal gills. It is, however, very difficult to demonstrate
with any certainty the respiratory significance of these organs.

In the Ei-li i until n i'i<lii<, although external gills are present, Stewart's organs may
attain a gigantic size (Asthenosoma vf/r/fs, Fig. 37<ij. They here fill the greater
part of the body cavity, and, it has been conjectured, serve in this case to prevent the
collapse of the flexible test at the time when the genital products are ejected. On
the other hand they may be quite vestigial, or even altogether absent (Fhormosoma). '

1 Cf. Bell. Ann. Man. X- H. vol. iv. 1S89, p. 437.

via EGHINODEBMATA—THE CCELOMIO VAVITIK* 443

In certain C'lypeastroida (J-Jc/u'nntfitti'i/s hit'ori*, l'< ro i>,/Ja orbiculnris] small inter-
radial, thin-walled, vesicular outgrowths of the lantern membrane, on the base of
the lantern itself, have been called Stewart's organs. Two of these orvur in each

Kn.. ;;,-(].— Viscera of Asthenosoma (a ft IT F. ami P. Sarasin). 1, Gonads ; J. cuii>ti •n.-rimi in a
Stewart's organ ; 3, Stewart's organ ; 4, muscle lamellie ; 5, radial canal of tin- wati-r vascular
sy~u-in; 6, tip of a Stewart's organ; 7, forked radius of the masticatory iVaiin-wnrk ; s and
'.i. upper and luwi-i' coils of the intestine ; 10, Polian vesicle ; 11, intestine.

interradius, but are generally wanting in that interradius in which the intestine in
ascending lies upon the lantern.

3. Ophiuroidea. — Two circular membranes, one above the other,
found round the oesophagus, and traversing the body cavity, connect
the (esophagus with the oral skeleton. They thus cut oft' two very
small perioesophageal sinuses from the general body cavity.

444 COMPARATIVE ANATOMY CHAP.

D. The Perianal Sinus.

In the Holothurioidea (with the exception of the Synaptidce) and in
the Echinoidea the end of the hind-gut is connected with the neigh-
bouring body wall by means of a circular membrane, which cuts oft'
a small perianal sinus from the general body cavity. If the circular
muscle fibres, with which the walls of this sinus are abundantly sup-
plied, contract, they act as sphincters, and close the anus. In the
regular Ecli/mii'l*, below the perianal sinus, there is a second closed
sinus surrounding the hind-gut ; this is the periproetal sinus.

E. The Axial Sinus.

1. Asteroidea. — In the madreporitic interradius the general body
cavity is traversed by a large, vertical, flattened tube, with tough, flat,
radially arranged lateral walls. This tube connects the region of the
madreporite with the ventral body wall. The cavity of the tube is an
enclosed portion of the true body cavity, and in a young stage is in
open communication with the enterocoel ; it is known as the axial
sinus (sac- or tube-like canal, sac hydrophorique). It surrounds and
contains (1) the stone canal, which runs down from the madreporitic
plate to the circular canal ; and (2) the axial organ (dorsal organ,
heart, pseudo-heart), which is attached to its Avail by means of a
mesentery. The tough wall of the axial sinus consists of the follow-
ing layers: (1) the ciliated endothelium of the body cavity (on the
side facing that cavity) ; (2) longitudinal muscle fibres ; (3) connective
tissue ; (4) the inner ciliated epithelium lining the axial sinus. The
axial sinus opens dorsally into the aboral circular canal (circular
sinus) of the genital system.

2. Ophiuroidea (Fig. 361, p. 422). — In consequence of the shifting
of the madreporite on to the oral side, the stone canal which springs
out of the circular canal bends outward and downward. At its distal
end it is connected with a small coelomic sinus, usually called an
ampulla, which lies on the side turned to the centre of the disc, and
itself opens outward through the water vascular pore. This sinus no
doubt corresponds with the axial sinus of Asteroids. Another sinus
accompanies the stone canal on the side turned to the periphery of the
disc, and opens into the circular canal of the genital system. That
portion of the wall of this second canal which is in contact with the
stone canal is developed as the ovoid gland.

3. Echinoidea (Fig. 358, p. 419). — The axial sinus, which runs
up from the circular canal to the apex and is accompanied by the
stone canal, is here almost completely filled by the large axial
organ. It is completely cut oft' by a septum from a spacious sinus
which lies near the ampulla, and into which a process of the axial
organ projects. The two communicate only in an early stage of

vin ECHINODERMATA—THE AXIAL ORGAN 445

development. Further, the communication which always primitively
exists between the axial sinus and the aboral circular sinus of the
genital system is interrupted in adult Echinoids, the only known
exception to this rule being EcMnocyamus pusillus.

1. Crinoidea. — In the Comatulidoe, an axial section of the body
cavity round which the intestine becomes coiled is said to exist. In
other Crinoids, such an axial sinus seems to be wanting, or else is filled
with connective tissue. There is, in adults, no connection between this
sinus and the stone canals. In the direction of the principal axis,
however, the axial sinus is traversed by a dorsal (glandular) organ,
which, although of somewhat different structure, no doubt corresponds
with the axial organ of other Echinoderms. This homology seems to
be established by the fact that the dorsal organ of the Crinoids shows
relations to the genital system similar to those of the axial organs in
the Echinoidea, the Asteroidea, and the Ophiuroidea to the same
system.

5. Holothurioidea. — In this class there is no axial sinus cut oft'
from the general body cavity.

F. The Axial Organ.

(Dorsal Organ, Heart, Pseudo -heart, Kidney, Plastidogenic Organ, Ovoid Gland,

Lymph Gland.)

No other organ of the Echinoderm has given rise to so many contra-
dictory statements as the axial organ. The names given in the above
heading, which have been gathered from various authors, show what
different functions have been ascribed to it.

According to the most recent anatomical and ontogenetic researches,
the following points may be stated with some degree of certainty.

(a) The axial organ lies on or in the axial sinus.

(b) It is developed from the endothelium of the body cavity, and,
during early ontogenetic stages, it grows out in the shape of pro-
cesses, strands, or tubes, which, at certain definite points of the body,
become the gonads (ovaries and testes).

(c) Further, in the adult animal, the axial organ is in most cases
still connected Avith the genital system, functioning, however (at least
in the Asteroidea, Ophiuroidea, and Echinoidea), in all probability, as a
lymph gland.

The Holothurioidea appear to possess no axial organ.

In the Asteroidea, Opliiuroidea, and Echinoidea, the axial organ
consists of a network of connective tissue, in whose meshes (embedded
in blood plasm) round cells lie, which, by continual division, yield
lymph corpuscles.

1. Asteroidea. — The axial organ lies in the axial sinus, to the wall of which it
is attached by a mesentery. Below the madreporite it sends off a process into a
small cavity, which is completely cut off from the axial sinus. Further, it pro-

44C, COMPARATIVE ANATOMY CHAP.

trude.s at certain points through the wall of the axial sinus into the general body
cavity.

2. Ophiuroidea. — The axial organ is developed out of the wall of the sinus which
accompanies the stone canal on the side which is turned towards the periphery of
the disc ; and, further, out of that portion of the wall which is in contact with the
stone canal. It projects as a somewhat massive body into the sinus, occupying
almost its entire lumen (cf. Fig. 361, 8, p. 422).

3. Echinoidea (Fig. 358, 32, p. 419). — The axial organ lies in the axial sinus,
which it almost completely fills, and to the wall of which it is attached by means of
numerous strands. It sends off a process into the sinus which lies below the madre-
piirite near the ampulla, this process perforating the wall which divides this sinus
from the axial sinus.

4. Crinoidea (Fig. 384 cij>, p. 482). — The axial organ, which is in this class
known as the genital stolon (or the glandular organ or the dorsal organ), is here
differently constructed. It originates as a thin strand in the axis of the chambered
organ, then ascends direct through the axial section of the body cavity of the calyx
towards the mouth, widening in the first part of its course, and then again narrowing.
It consists of a complex of much-twisted canals with narrow lumina, which are en-
closed in a stroma of connective tissue ; the lumina of these canals may disappear,
and they may become strands. They are lined with cylindrical epithelium. In the
axis of the chambered organ, the axial organ consists merely of a few very thin strands
or canals, but when it leaves the chambered organ and ascends into the body cavity,
the canals swell and branch, so that their number increases till the middle of the h<>dy
cavity is reached. Then their number again decreases, one canal after the other
ending blindly. Finally, in the oral region, the axial organ consists of only a few
strands, which most probably are continued into the genital tubes or strands of
the arms. Such a connection has at least been demonstrated in the young Anfi'dnn.
and it has been further found that, ontogenetically, the genital tubes bud out of tin-
axial orsan.

G. The Chambered Sinus of the Crinoidea and its continuation in

the Stalk and in the Cirri.

Quite in the apex of the calyx, and in Antfd.ou enclosed in the
centrodorsal, there is a cavity containing the apical parts of the
axial organ. This cavity is of enterocoelomic origin. It is divided,
by means of five radially arranged partitions of connective tissue, into
five chambers, which are covered on all sides by epithelium. This is
the "chambered organ" (Fig. 384 ch, p. 482).

In stalked Crinoids the chambered sinus is continued into the
stalk, forming in it a canal. This is also divided, by five radially
arranged partitions, into five sub-canals (the continuations of the five
chambers of the sinus) arranged round a common axis. This common
axis is probably formed by a continuation of the axial organ.

In each of the whorl joints of the stalk in those Crinoids which
have cirri, the five-fold canal widens to form a kind of repetition of the
chambered sinus, giving off into each of the cirri a lateral canal, which
runs through its whole length, and is divided into an upper and a lower
canal by a horizontal partition reaching inwards as far as to the
common axis of the stalk canals.

viii ECHINODERMATA—THE PSEUDOHSEMAL SYHTEM 447

In the Comatulidce, in consequence of the absence of a stalk, the
arrangement is somewhat modified. We must imagine, however, that
only the internodes of the stalk are wanting, just as many whorl
joints being fused with one another and with the centrodorsal as there
are whorls of cirri. In this way the chambered sinus is enlarged by
the addition of the widened portions of the canal which were origin-
ally present in the whorl joints. The cirrus canals now rise direct
from the chambered sinus.

As in the cirri of the stalked Crinoids, so also in those of the
Comatulidce, the canal is divided by a horizontal partition, and in these
latter animals also the partition of each cirrus canal is continued as far
as to the axis of the chambered sinus formed by the axial organ. The
chambered sinus of the Comatulidce, in sections taken in the direction
of the principal axis, appears therefore to be divided by these parti-
tions into as many spaces, consecutively superimposed, as there are
consecutive whorls of cirri.

The five chambers of the chambered sinus, in close contact with
the axial organ, are produced orally for a short distance as ever-
narrowing canals accompanying the axial organ, and then end blindly.

The system of the chambered sinus is, therefore, in adult Crinoids,
completely cut off from the rest of the ccelom.

For the relations of this sinus to the apical nervous system, see the section deal-
ing with this latter, p. 460.

VIII. The Pseudohsemal System.
(Radial Sinuses and Circular Sinus of the Schizoccel, Subneural Canals.

The pseudohsemal system consists of canals which are closely
connected in the same manner in all Echinoderms with the oral
nervous system. As radial pseudohsemal canals, they accompany
the radial nerve trunks as far as to the ends of the radii ; and as
a pseudohsemal ring they accompany the nerve ring in its course
round the cesophagus. They always lie on the inner side of the nerve
trunks (that turned to the body cavity), between these trunks and the
water vascular trunks. The radial pseudohsemal canals give oft" side
branches which accompany the nerves of the tube-feet to their bases.

The pseudohsemal canals are filled with a fluid which resembles
the ccelomic fluid. The intimate relation subsisting between these
canals and the oral nerve ring and the radial nerve trunks makes it

o

probable that they are specialised for the nourishment of these nerves.
It has also been conjectured that they, together with the epineural
canals, which we shall describe later, are of essential service in pro-
tecting the nerve trunks from being pressed or torn.

In the Holothurioidea and Echinoidea the pseudohsemal system is
closed on all sides ; in the Asteroidea and Ophiuroidea, on the contrary.

448 COMPARATIVE ANATOMY CHAP.

it communicates, by means of numerous apertures, with the general
body cavity, and at one point of the pseudohsemal ring, in the madre-
poritic interradius with the axial sinus.

The pseudohsemal s}*stem, in the Ophiuroidea and Asteroidea, is
said to arise, ontogenetically, as a cleft in the connective tissue
(mesenchyme), and thus to be a sehizoecelomie structure. It is,
however (as has been proved in the Holothurioidea), lined with endo-
thelium. Such a lining, in the case of a sehizoecelomie cavity is, how-
ever, so incongruous, that we are justified in desiring further evidence.
(For the position of these canals, see Figs. 352-356, pp. 409-413.)

Special. — In the Holoth urioidca the oral pseudoh;emal ring is, in the Paractinopm 1, i
(Stjnaptidce), separated from each of the radial pseudohfemal canals by a septum.
The pseudohffiinal canals stretch only a short way backwards. In the Actinopoda,
they run the whole length of the body, but are said also to end blindly at both ends,
and the pseudohfemal ring is said to be wanting. The same is the case with the
well-developed radial pseudohfemal canals of the Ecldnoidca. In the Crinoidca the
canals are certainly very much reduced, their existence is altogether denied by some
authors. The pseudohsemal canals of the i^t/tini-oidea give off lateral branches at
regular segmental intervals ; these branches ascend to the brachial cavity (dorsal
canal), and open into it. In the Astcruidi:a, both the circular and the radial vessels are
divided into two by a longitudinal septum. In the radial canals the septum is vertical,
in the pseudohremal ring it runs slantingly, dividing it into an outer and lower and
an inner and upper canal. The latter, the inner and upper canal, in the madreporitic
interradius, communicates with the axial sinus ; the former, the outer and lower
canal, is in open communication with the body cavity of the disc by means of five
ascending, interradial, lateral canals. At regular intervals, between every two
consecutive tube-feet, each radial pseudohaemal canal is connected with two marginal
canals which run longitudinally at the edges of the ambulacral furrow. Each tube-
foot receives two canals from the pseudohwrnal system, which run to its tip ; one of
these comes from the radial canal, and the other from the lateral canal. The lateral
canal, further, sends off at each corner between two consecutive ambulacral plates
and the contiguous adambulacral plate, a lateral branch, which runs up between
these plates. This lateral branch opens into the brachial cavity.

Two specially interesting facts deserve notice : (1) the mesentery by means of which
the axial organ is attached to the wall of the axial sinus is continued into the septum
of the pseudohfemal ring, and through this into the septum of the radial pseudo-
hfemal canals ; and (2) the axial organ, although in a reduced condition, may even
be produced along a greater or smaller portion of these septa. These facts throw
further doubt upon the sehizoecelomie nature of the pseudohfeinal canals.

IX. The Epineural System.

In the Holothurioidea, the E<-]iiii<>i<l«i, and the 0].>lt'h<r<>i<l<:n the oral
nervous system is accompanied by canals known as the epineural
canals, which run between it and the adjacent body epthelium.
This epineural system thus repeats on the outer side of the oral
nervous system the pseudoh?emal system which accompanies the
nerves on their inner side, and, like the latter system, it consists of
an oral circular canal, and radial canals. In the Asteroidea and

viii EGHINODERMATA— BLOOD VASCULAR SYSTEM 449

it, where the oral nervous system still lies in the epithelium,
the epineural canals are wanting1. This fact is in harmony with
what we know of the formation of the epineural canals, which, as has
been proved in the Ophiuroidea, arise through the sinking of the nerve
trunks below the surface — the nerve trunks rising ontogenetically
in the epithelium. These originally epithelial trunks become covered
by two lateral integumental folds, which finally grow together over
them in such a way that between them and the integument, which
now forms a continuous cover over them, a cavity, the epineural canal,
remains. The Synaptidce also have no epineural canals. This is
probably connected with the special way in which their (subepithelial)
nerve trunks develop ontogenetically.

An epineural circular canal is wanting in the Holothurioidca, and, in the
Eclutioidca, the circular canal is not in communication with the radial epineural
canals. A small epineural (periambulacral) space occurs in connection with the
development of a circular ganglion at the base of each tentacle in the Opkiuroidea.

X. The Blood Vascular or Laeunar System.

Within the connective tissue of various parts of the body in most
classes of the Echinodermata there occurs a strongly developed system
of very small spaces or lacunse, which open into one another, and
sometimes form at the surface of different organs a fine, close, membrane-
like network of lacunse. In other cases, they coalesce to form bundles
of canals, running in definite directions, and anastomosing with one
another. This lacunar system was formerly universally called a blood
vascular system, and may still be allowed to retain this name,
although a regular circulation in definite directions of the fluid
it contains has in no single ease been demonstrated.

The intercommunicating lacunse, of which the blood vascular
system consists, have no walls of their own, and no endothelial
lining, and their arrangement in networks or plexuses, which are
sometimes spread out flatly, and at others thickened into " vas-
cular trunks," is entirely confined to the Eehinodermata.

A localised propelling apparatus is wanting. What was for-
merly called the heart has nothing to do with the blood vascular
system, but is the axial organ.

The fluid (blood) contained in the blood vascular system re-
sembles that in the body cavity and in the water vascular system,
but contains much more albumen in solution. In sections of fixed
and stained animals, a vessel can easily be distinguished from the
other almost empty cavities of the body by the considerable quantity
of coloured coagulum contained in its spaces. The solid constituents
doating in the blood are the same as in the body cavity and in the
ambulacra! vascular system.

Leaving altogether on one side the Oji/iinrniika and the Ast&roidea,
in which the existence of a blood vascular system is still doubtful,

VOL. II 2 G

450 COMPARATIVE ANATOMY CHAP, vin

the following may be considered as the chief constituents of the Wood
vascular system in the Echinodermata : ( 1 ) a blood vascular network
in the intestinal wall, whose evident function is to absorb the digested
and dissolved nourishment as albuminoid out of the intestinal wall ;
(2) two large vascular trunks, which, arranged on opposite sides of
it, accompany the intestine in its course ; these conduct the blood,
which has become enriched with albuminoid in the vascular network
of the intestine, to other parts of the vascular system ; (3) a blood
vascular ring, which surrounds the mouth or oesophagus, and into
which the two intestinal vascular trunks open ; (4) five radial blood
vessels, which run, like the water vascular and the nerve trunks, in
the radii • (5) a vascular network on the surface of the gonads
(genital glands) ; (6) a vascular network on the surface of the axial
organ.

The connective tissue surrounding the vessels may, at various
points of the vascular system, become modified for yielding corpuscles
to the blood (lymph glands).

Contractions, very irregular and indistinct, have been observed
only in the vascular trunks of the intestine of the Holothurioidea.

1. Holothurioidea (Fig. 371). — The simplest arrangement of the
blood vascular system in this class is found in the Paractinopoda
(Synaptidce). The intestinal lacuna? pour their contents first into two
longitudinal trunks, one of which runs along the dorsal, and the other
along the ventral side of the intestine. The ventral trunk opens
anteriorly into the dorsal, which then runs along within the dorsal
mesentery direct to the genital gland. It is here continued into a
spacious lacunar system, which develops in the wall of the gland in
such a way as to split into two diverging lamella?, an outer and an
inner, the latter carrying the germinal epithelium. The dorsal vessel
further gives off a small lateral branch near the point at which the
stone canal enters the circular canal. According to recent researches,
neither a circular canal of the blood vascular system, nor radial
vessels, nor tentacle vessels are to be found. In the Actinopoda (Fig.
371) the blood vascular system is more completely developed. Here,
again, the blood from the lacunar network of the intestinal wall (which
lies on the inner side of the muscle layer) is collected by two vascular
trunks which accompany the intestine along its whole length, i.e. to
the hind-gut ; one of these trunks is the dorsal or mesenterial, and
the other the ventral or anti-mesenterial. They both open anteriorly,
immediately behind the circular canal of the water vascular system,
into a circular vessel which surrounds the oesophagus, from which five
radial vessels run in the radii. Each radial blood vessel lies between
the radial nerve trunk on the outer and the radial water vessel on
the inner side (Fig. 352, p. 409). It gives off lateral branches to
the oral tentacles, the ambulacra! feet, and the papillae. The wall
of the genital glands is everywhere richly " vascularised," either by a
lacunar network or by a more simple splitting of its wall of connective

Km. 371.— Organisa-
tion of Holothuria
tubulosa. Tim blood
vascular system is
marked black. 1, Oral
tentacles ; 2, stone
canals ; 3, water vas-
cular ring ; 4, Polian
vesicle ; 5, gonads ; r,,
longitudinal musclrs ;
7, anterior section of
intestine ; S, ventral
intestinal vessel ; !>,
radial water vessel ; 10,
vascular anastomoses :
11, dorsal intestinal
vessel ; 12, filaments
and strands (of tin'
nature of both muscle
and connective tissue)
which attach the cloaca
to the body wall ; 13,
cloaca ; 14, cloacal
aperture (anus) ; 15,
middle section of the
intestine ; 10, posterior
section of the same ;

17, right branch ial tree ;

18, rete mirabile ; I'.'i,
radial canal of the
water vascular system ;

20, left branchial tree ;

21, tentacle ampulla
(after Milne Edwards
and Carus).

13

452 COMPARATIVE ANATOMY CHAP.

tissue. The blood vascular system of the genital gland may derive its
blood in three mutually exclusive ways : (1) by means of a special
genital vessel from the blood vascular ring, (2) by means of a special
genital vessel from the dorsal intestinal vessel, (3) direct from the
latter, with which the genital gland is in contact.

The ventral vessel of the anterior section of the intestine is almost
always connected with that of the middle section by means of a
usually simple, but sometimes complicated anastomosis (Figs. 371, 10,
383, 27, p. 477).

In the Aspidochirotce principally, but also in many Dendrochirotce
and Molpudiidce, the dorsal vessel becomes detached from the intestine
for a considerable distance ; in Holothuria tulmlosa (Fig. 371) this occurs
along part of the anterior, the whole middle, and part of the posterior
section, and it has a free course as the marginal vessel of the rete
mirabile through the body cavity. It, however, remains in connec-
tion with the lacunar network which is developed in the wall of
the intestine by means of a rich plexus of blood lacunae, known as the
rete mirabile. This network forms a much perforated membrane,
one edge of which is attached to the intestine, while the marginal
vessel runs along the other. The blood of the intestinal lacunar
system, again, may collect in a special longitudinal vessel (collateral
vessel, pulmonary vein) before passing over into the rete mirabile
(Fig. 371).

The rete mirabile is often particularly richly developed iu the loop
formed by the anterior and middle intestinal sections. In the rest of
the small intestine, the vessels which enter the plexus from the mai^ginal
vessel first break up into a bundle of very fine lacuna (capillaries) ;
these finest lacunoe then collect to form vessels which enter the collateral
vessel. It may be said that here, between the collateral and the mar-
ginal vessels, numerous small retia mirabilia (Fig. 371, IS) of the
second order are developed, and that they form webs round the
terminal ramifications of the left water lung. It is doubtful
whether they serve for respiration, since they are not developed in
the wall of the water lung itself, but merely loosely invest the latter.

The dorsal vessel (so called because it lies on the anterior intestinal section
close to the dorsal mesentery) does not run in this mesentery, but somewhat to tin-
left of it. In the middle section of the intestine it comes to lie on the right, and in
the posterior section again on the left side of the mesentery.

At certain points in the course of the vessel, blood glands may be developed.
The spongy, alveolar structure of the vessel then becomes more marked, and many
cells are found embedded in the strands, filaments, membranes, etc., of connective
tissue which traverse the vessel ; these are the formative cells of the blood corpuscles.
At such points the lumina of the vessels (i.e. of the separate lacuna?, which together
form the vessel) are much reduced.

2. Echinoidea. — The arrangement of the blood vascular system
here agrees to a great extent with that in the Holothurioidca. A rich
plexus of vacuoles is developed in the connective tissue layer of the

viri ECHINODERMATA— NERVOUS SYSTEM 453

intestinal wall (with the exception of a larger or smaller portion of
the hind-gut) ; from this plexus the blood collects into two longitudinal
vessels, one outer and dorsal and the other inner and ventral. These
longitudinal vessels do not lie in but on the wall of the intestine, in
its mesenteries. The ventral or inner vessel, along the part where
the accessory intestine is developed, passes on to this latter. Both
vessels open into a blood vascular ring which encircles the oesophagus,
in close contact with the water vascular ring. The position of these
two rings and their relation to one another has already been sufficiently
described (p. 424). From the blood vascular ring five radial vessels
run to the radii, within the bands of connective tissue which separate
the pseudohsemal canals from the radial trunks of the water vascular
system (Fig. 353, p. 410). In Echinoids provided with a masticator}'
framework, the proximal portions of these radial blood vessels descend
in the axis of the lantern along the edges of the five single pyramids
forming the lantern, which are turned towards the oesophagus and to
the nerve ring (Fig. 358, p. 419), before passing through the auricles
to be produced as mentioned into the radii. They are said not to be
in open communication with the circular vessel, but to be divided from
it by a septum. The radial blood vessels during their course give off
lateral branches which run to the bases of the tube-feet.

A lacunar plexus is also developed in the axial organ, immediately
below the surface ; this is either in direct communication with the
blood vascular ring which encircles the oesophagus, or else draws its
blood from the dorsal intestinal vessel. The vacuolar plexus of the
axial sinus is further continued into the wall of the apical circular sinus
of the body cavity, and thence on to the wall of the genital glands.

3 and 4. Asteroidea and Ophiuroidea. — It is doubtful whether a
blood vascular system occurs in these two classes. This system is
certainly wanting in the intestinal wall. The vessels which have
hitherto been regarded as blood vascular, ring, and radial blood vessels
(running between the nerve trunks and the pseudohsemal vessels) agree
in structure with the axial organ, and are said to be direct continua-
tions of this organ. As such they fall under another heading.

5. Crinoidea. — The blood vascular system is here well developed,
and in all its parts shows in a very marked manner the spongy
structure characteristic of Echinoderms (i.e. it consists of lacunar net-
works or plexuses). One lacunar network covers the axial organ, and
another spreads out over the intestinal wall. Both are in open com-
munication with a lacunar plexus which surrounds the oesophagus, and
which may, at one point, become differentiated into a blood gland
(spongy organ) by the deposition of numerous blood formative cells in
its meshes.

XL The Nervous System.

This system in the Echinodermata is developed in a quite peculiar
manner, unknown in other animals. It is composed of three alto-

454 COMPARATIVE ANATOMY CHAP.

gether independent systems : (1) the superficial oral, (i) the deeper
oral, and (3) the apical system.

(a) The superficial oral nervous system is developed on the oral
side of the body, and is alwa3'S more or less superficial. Its most
important and constant constituents are : (1) a nerve ring encircling
the oesophagus, and (2) radial nerves radiating from this ring, and
corresponding in number with the radii. This system innervates the
integument, the ambulacral appendages, and the intestinal canal. It
occurs in all Echinodermata without exception.

(//) The deeper oral nervous system accompanies the superficial
system on its inner side (i.e. on the side turned towards the body
cavity). In the Ophiuroidea and Asteroidea it is paired in each radius,
i.e. its trunks or ganglia lie on the two sides of the radial nerves of
the superficial system. In the Echinoidea and Hvlothurioideu, on the
contrary, it is unpaired in eacli radius, and consists of a trunk or ridge
in close contact with the radial nerve of the superficial system on its
inner side. In the Ophiuroidea and Asteroidea a more or less complete
ring surrounding the oesophagus seems to be developed in this system,
but this is wanting in the Echinoidea and Holothunoidea. The Crinoids
and those Echinoids which have no masticatory apparatus have no
deeper oral nervous system.

This system innervates the muscles which run in the oral side of
the body Avail ; in the Holothurioidea it perhaps innervates the whole
dermo-muscular tube, but in the Echinoidea probably only the muscles
of the masticatory apparatus.

(<•) The apical nervous system is specially strongly developed in
the Crinoidea. It consists of a nerve envelope, surrounding the
chambered organ, and forming a centre from which five nerves run out
along the axial canals of the brachial skeleton, penetrating as far as
to the distal joints of the pinnulse ((/. Fig. 356, S, p. 413).

The apical nervous s}rstem is also continued into the stalk and the
cirri. It innervates the whole of the musculature which moves the
arms and the cirri.

In the Asteroidea the apical nervous system consists of radially
arranged nerve trunks which meet at the centre of the disc ; there is
one trunk for each arm. These trunks run along the middle line of
the arms, immediately above the body cavity, and innervate the dorsal
muscles of the arms.

In the Oj.Jtiiii-iiidea and Echinoidea a delicate nerve trunk runs
within the wall of the aboral circular sinus throughout the whole
course of the latter ; this is the genital nerve ring.

The Holothurioidea have no aboral system.

A. The Superficial Oral Nervous System.

Until comparatively recently this system was the only nervous
system known in Echinodermata,

vin ECHINODERMATA— NERVOUS SYSTEM 455

In the Asteroidea and Crinoidea it occupies throughout life an
epithelial position, but in all other Echinoderms it sinks and becomes
subepithelial, except at the ends of the radii (in the terminal tentacles),
and in the intestine. Here, throughout life, it is epithelial.

The shifting of the superficial nervous system to a position below
the body epithelium, in Echinoderms, goes hand in hand with the
formation of the epineural canals, as already explained (p. 449).

1. Asteroidea (Fig. 354, p. 411).

The radial nerves here form a thickened longitudinal ridge of the
epithelium along the base of each ambulacral furrow, and the circular
nerve a thickened ridge round the mouth. In these nerve ridges the
(ciliated) epithelial cells represent the nerve cells. At their bases they
are continued into nerve fibres, which run in the longitudinal direction
of the nerve ridges (i.e. of the radial nerves), and together form the
deeper layer of the ridge. From the radial nerves a close plexus of
nerve fibres spreads out below the surface of the outer epithelium over
the whole body ; this is especially close in the ambulacral feet. In
the same way a layer of nerve fibres is found over the whole intestinal
epithelium ; this layer increases in thickness orally until it enters the
circular nerve.

2. Crinoidea (Fig. 356, p. 413).

The above account of the nervous system of the Asteroidea is also
applicable to that of the Crinoidea, if it be remembered that the food
grooves of the arms and of the tegmen calycis of the Crinoids correspond
with the ambulacral furrows of the Asteroidea. The food grooves
branch with the arms, and so do the radial nerve ridges of the super-
ficial oral system. The Crinoids, however, differ from the Asteroids in
that the epithelial nerve plexus is limited to the oral side of the calyx
and of the arms, since no epithelium can be made out on the apical
side of the calyx, on the sides and backs of the arms, and on the stalk
and the cirri of adults.

3. Ophiuroidea (Fig. 372).

The superficial oral nervous system, having here taken up a sub-
epithelial position, appears in the form of distinct nerve trunks and of
nerves radiating from these. The central portion consists of the nerve
trunks of the disc and of the five radial trunks of the arms. These
latter run on the inner side of the row of ventral shields, between
these and the vertebral ossicles. The circular nerve throughout its
whole course is beset with nerve cells on the side turned towards the
oesophagus, and the same is the case with the radial nerves on the
side turned towards the ventral shields. The segmentation of the
arms appears in a marked manner in the radial nerve trunks which

456

COMPARATIVE ANATOMY

CHAP.

run through it. At regular segmental intervals, about on a level with
the consecutive pairs of tentacles, these trunks show distinct swellings,
from which most of the nerves originate. In this way the radial
nerves superficially resemble the ventral ganglionic chain of many
Annulata and Arthropoda.

The central portion of the deeper oral nervous system is so closely
applied to the corresponding portion of the superficial nervous sys-
tem that the two can only be distinguished from one another by means

15

PIG. 372.— Nervous system of an Ophiurid (Ophiothrix fragilis) (after Cuenot). Part of the
disc and the base of an arm. 1, Peripheral brauhial nerve ; 2, tentacle nerve ; 3, nerve to the
muscles between the vertebral ossicles ; 4, radial nerve trunk ; 5, ganglion at the base of a spine ;
6, marginal nerve of the disc ; 7, bursal aperture ; S, nervus lateralis ; '.>, gonad projecting into the
bursa; 10, Polian vesicle ; 11, interradial nerve; 12, nerve of the muscnlus interradialis aboralis
(Simroth); 13, nerve ring; 14, dental nerve; 15, enteroccelic nerve ring; 16, spine; I, first oral
tentacle ; II, second oral tentacle ; III-VIII, tentacles of arms.

of careful microscopic examination, e.g. of transverse sections. In the
following description, however, we shall keep the two systems entirely
distinct from one another.

Nerves of the cesophageal ring. — A large number of nerves arise from the
cesophageal ring, and ramify iii the connective tissue layer of the intestinal wall.
Further, at each of the points where a radial nerve trunk joins the cesophageal ring,
the latter gives off two nerves, which run to the bases of the first pair of oral tube-
feet. Each of them there forms a circular ganglion, almost completely surrounding

viii ECHINODERMATA— NERVOUS SYSTEM 457

the base of the root, and then sends off a branch which runs to the tip of the tube-
foot (Fig. 372, 1).

Nerves of the radial trunks. — Two pairs of nerves arise at regular intervals from
the radial nerve trunks, as well in the free portion of the arm as in that included in
the disc. These are the tentacle nerves and the peripheral nerves.

The tentacle nerve has a short course to the base of its tentacle, and then- forms
a ganglion — the tentacle ganglion — which embraces the base of the tentacle. The
nerve contains along its course nerve cells, which are still more plentiful in the ten-
tacle ganglion. The tentacle nerve and the ganglion are, like the radial nerve trunk,
accompanied by an epinenral canal. From this basal tentacle ganglion, the tentacle
nerve, which remains in the subepithelial layer, ascends to the tip of the tentacle.

The peripheral nerve at first runs in close contact with the tentacle nerve of the
same side as far as to the base of the tentacle ; it then runs further laterally, pene-
trates into the lateral wall of the arm, running through the skeletal mass of the
latter, and breaks up into branches which innervate the ventral, lateral, and dorsal
integument of its own side of the arm.

In those Ophiurids, the sides of whose arms are provided with spines (i.e. in the
majority of cases), peripheral ganglia are formed at the bases of these spines.

The nervous system becomes somewhat more complicated in the portion of the
arm which is included in the disc. Branches of the peripheral nerve ascend apically
at the side of the bursa which is turned to the arm, i.e. on the radial side of the
genital plate described on p. 361, when it meets a lateral nerve (8) running along this
plate. Distally this lateral nerve is continued into an aboral marginal nerve (6), the
branches of which innervate the periphery of the disc. Proximally, the lateral
nerve passes over into an interradial nerve (11), which runs along the outer side of
the interradial muscle.

In some cases (Ophiothrix, Ophiocoma, Ophioglypha), the first pair of tentacle
nerves and their accompanying peripheral nerves are followed by a pair of nerves
which run towards the bursal aperture, subsequently uniting with the lateral
nerves.

In 0})hioglijplia, a delicate pair of nerves which innervate the integument of the
ventral (oral) side of the arm runs out between each of the regularly recurring pairs
of tentacle and peripheral nerves.

4. Echinoidea (Figs. 353 and 358, pp. 410 and 419).

In all Echinoids, even in those which are provided with a
masticatory apparatus, the ossophageal ring remains in close proximity
to the mouth, on the inner side of Aristotle's lantern, and at a con-
siderable distance from the water vascular and blood vascular rings,
which appear to be in a manner lifted up by the lantern. From the
oesophageal ring, especially, in the radii, nerves run to the oesophagus,
gradually breaking up into a plexus, which is still traceable even in
the wall of the first coil of the intestine. From the radial nerve
trunks, as in the OpMuroidea, at intervals which correspond with the
ambulacral feet, nerves for these feet and peripheral nerves are
given off. In the regular Echinoidea and the Clypeastroida, the tube-
feet nerves and peripheral nerves arise together, but in Spcdangoida
they arise separately. The points of origin of these two sets of
nerves from the radial nerve trunk are arranged alternately, in
accordance with the alternate arrangement of the two rows of

458 COMPARATIVE ANATOMY CHAP.

ambulacral plates of an ambulacrum, and with the alternate arrangement
of the ambulacral feet at the two sides of the water vascular trunk.

Each tube-foot nerve, with its peripheral nerve, passes out together
with the tube-foot canal of the Avater vascular system, through the
ambulacral pore, on to the surface of the test. The nerve of the
tube-foot is then continued in the epithelial layer to the tip of the
foot, without forming a ganglion. The peripheral nerve, however,
enters an integumental nerve layer \vhich covers the Avhole body and
its appendages.

The network of nerves or layer of nerve fibres, both in the intes-
tine and in the external integument, lies, in regular Echinoids (Cidi.tr-
oi<I<t, Diadc matoida] and in the Clypcastroida, deep in the epithelium,
but is subepithelial in the Spatangoida.

5. Holothurioidea.

The superficial oral nervous system is here subepithelial and
agrees in all respects Avith that of the Echinoidea. The nerve ring
which surrounds the mouth gives off the nerves to the oral tentacles
and the intestinal canal. The latter innervate also the skin round the
mouth, and ramify richly in the connective tissue layer of the intestine,
especially in its anterior portion. The radial nerve trunks give off
lateral branches to the tube-feet or ambulacral papilla?, and "peri-
pheral nerves " to the integument. The latter break up into a
subepithelial network of fibres.

In the Si/ttaptid(c, each radial nerve trunk, soon after it rises from
the oesophageal ring, gives off a pair of nerves to the auditory
vesicles.

B. The Deeper Oral Nervous System.

1. Asteroidea (Fig. 354, p. 411).

In close contact Avith the inner side of each radial nerve ridge
(which is here epithelial) a subepithelial longitudinal band of nerve
cells and nerve fibres run along each side. A similar band accom-
panies the oesophageal ring, at least in its interradial regions. From
the radial bands of the deeper nervous system, lateral nerve bands
arise at regular intervals which correspond Avith the ambulacral feet.
These lateral bands ascend on the outer side of the radial pseudo-
haemal canals, and soon break up into fibres, Avhich probably innervate
the muscles of the ambulacral skeleton. Nerves Avhich emerge inter-
radially from the deeper nerve ring seem to innervate the interradial
muscles of the oral skeleton.

2. Ophiuroidea (Fig. 355, p. 412).

Here also, tAvo lateral bands, consisting of nerve cells and
longitudinal fibres, lie in close contact Avith the inner side of each

vin EGHINODERMATA— NERVOUS SYSTEM 459

radial nerve trunk. The radial nerve trunk of the superficial system
and the paired nerve bands of the deeper nervous system are separated
merely by a thin structureless membrane. The superficial circular
nerve is similarly accompanied by a band of the deeper nervous
system.

The paired bands become thickened at regular intervals, simul-
taneously with the radial nerve trunks of the superficial system.
Between the consecutive swellings of this latter, however, they are
extremely thin. The deeper nerve ring is much thicker in its inter-
radial than in its radial portions.

In each interradius two nerves are given off" by the deeper nerve
ring ; these break up into several branches and innervate the inter-
radial muscles of the oral skeleton.

The paired radial bands of the deeper system give rise, at points
which regularly alternate with the tentacle and peripheral nerves, to
nerves which, first penetrating the cavity of the pseudohsemal canal,
ascend apically, enter the vertebral ossicles, ramify in them, and inner-
vate the intervertebral muscles. The vertebral nerve of one side
always innervates the dorsal and ventral intervertebral muscles of the
same side of the arm, which counteract the homologous muscles of the
opposite side.

3. Echinoidea.

Only those Echinoids which are provided with a masticatory
apparatus have a deeper nervous system ; this bears out the very
probable assumption that this system innervates the masticatory
muscles. It consists of five lamellae consisting of nerve cells and
fibres, which are in close contact with the radial portions of the super-
ficial cesophageal ring and the points of departure of the radial nerves.
Each lamella gives off a pair of large nerves. These ascend along the
edges of the five jaws, then ramify, and most probably, as already
stated, innervate the jaw muscles.

4. Holothurioidea (Fig. 352, p. 409).

The deeper nervous system is only developed on the radial nerve
trunks, which it covers on their inner sides in the form, in each case,
of a single thin band, consisting of nerve cells and longitudinal fibres.
The nerves which proceed from these radial bands seem to serve
principally for the innervation of the dermo-muscular tube.

C. The Apical or Aboral Nervous System.

The apical nervous system of the Asteroi<l<'i.>, Echinoidea, and
Ophiuroidea has already been sufficiently described at the beginning of
this section, p. 454. That of the Crinoidea, however, requires some
further elucidation.

460 COMPARATIVE ANATOMY CHAP.

The chambered sinus which lies in the centrodorsal is surrounded
by a cup-shaped envelope, which consists of ganglion cells and nerve
fibres. The latter are arranged more or less concentrically around the
sinus. The nerve envelope of the chambered sinus is continued along
its prolongation, the stalk canal, and along the canals of the cirri. All
these canals are in fact surrounded by a nerve sheath whose fibres run
longitudinally. The large apical nerve trunks, which run from the
nerve envelope of the chambered sinus into the radii (enclosed in
the nerve canals), consist of nerve fibres and ganglion cells. A kind
of metamerism is apparent in them, since they become somewhat
thickened in the consecutive ossicles of the arm, and give off nerves at
regular intervals corresponding with these ossicles. The apical nerve
trunks divide with the arms, in whose nerve canals they are enclosed,
and run to the tips of the pinimlse.

Round the chambered sinus, commissures are formed between the
nerve trunks arising from the nerve-envelope of the sinus. The courses
of these commissures are explained by the diagrams of the nerve canals
in which they are enclosed (Figs. 327-330, p. 378).

In Antedon and other forms, in the second costal, Avhere the five
primary nerve trunks fork to make the ten brachial nerves, a
peculiar ehiasma nervorum braehialium is formed (cf. Figs. 327
and 329, p. 378). The two ramifications of the nerve, which by
their crossing form the ehiasma, run one above the other without
any mingling of their fibres. Further, the two brachial nerves form-
ing one pair are connected immediately beyond the ehiasma by means
of a transverse commissure.

In each ossicle of the arm, the apical brachial nerve gives off an
upper (oral) and a lower (apical) pair of nerves. These nerves seem to
be principally sensory. They ramify richly in the calcareous substance
of the joints. The increasingly fine ramifications, which radiate to
the surface of the arm (Fig. 356, p. 413), finally reach special groups
of epithelial cells, which must no doubt be considered as sensory cells.
One principal branch of the oral (upper) pair, however, is said to run
to the musculature, uniting the joints of the arm.

Besides these two pairs of nerves proceeding from the apical nerve
trunks, nerves are said to run out at the level of the joint, from
between the consecutive brachial joints, these having for their special
function the innervation of the brachial musculature.

The apical pinnula nerves, which are given off in alternating order
by the brachial trunks to the pinnula?, arise from double roots.

According to the present state of our knowledge on the subject, the
apical nervous system of the Crinoids is formed by the coelomic endo-
thelium. Even in adult animals, those portions of this system which
envelop the chambered sinus, the stalk canal, and the cirrus canals
show a close connection with their places of formation. The apical
nervous system of the Asteroids retains its endothelial position through-
out life.

vin ECHINODERMATA— NERVOUS SYSTEM 461

D. The third Nervous System of the Crinoidea.

Besides the superficial oral and the apical systems, the Crinoids
have another nervous system. This is developed on the oral side of
the disc and of the arms, and is subepithelial in position. It consists
of the following principal parts: (1) a nerve ring encircling1 the
oesophagus close to the mouth ; and (2) five pairs of braehial
nerves. The two nerves of each braehial pair run down the arm
longitudinally at the sides of the radial water vascular canal (Fig.
356, 4, p. 413). They are found on the branches of the first, second,
and other orders. Their condition at the points where the arms
divide is not yet known.

In addition to the five pairs of braehial nerves, the oesophageal ring-
gives rise, in each interradius, to two nerves which branch freely and
run to the bands and mesenteries which traverse the body cavity,
giving off branches also to the tegmen calycis.

Lateral branches of the paired braehial nerves innervate the mus-
culature of the water vascular and tentacle canals which run in the
arm ; they also ascend in the tentacles to innervate their sensory
papillae.

This third nervous system of the Crinoids is connected with the
apical system by means of branches in the following manner :—

1. The two braehial nerves of each arm send off branches alter-
nately (first from the right nerve, then from the left) towards the
apical side of the arm. Each of these unites with one of the pair of
nerves which descend orally from the apical nerve trunk within the
interior of the braehial joint (Fig. 356, 10, p. 413).

2. Certain lateral branches of the pair of nerves which run out
from the oesophageal ring in each interradius seem to run along the
body wall apicalty, and to unite with lateral branches of the apical
nerve trunks which come from the nerve envelope of the chambered
sinus.

Although recent investigators incline to the opinion that this third nervous
system in the Crinoids has no representative in the other Echinoderms, the question
may here be raised whether it does not correspond with a deeper oral nervous system.
If we imagine the deeper oral nervous system of an Opliiurid or an Asteroid to have
become severed from the superficial system and to have shifted further below the
surface, we have a system showing considerable resemblance to the third system of
the Crinoids. Because no connection between the deeper oral system and the apical
system has yet been demonstrated in other Echinoderms, we have no right, considering
the difficulty of investigation on this point, to conclude that no such connection
exists. It is difficult to believe that there are three completely independent nervous
systems in the Echinoderm body.

462 COMPARATIVE ANATOMY CHAP.

XII. The Sensory Organs.

The sensory organs of the Echinodermata show, as compared with
the great complexity of the rest of their organisation, a very low
degree of differentiation. They are for the most part undifferentiated
integumental (tactile) sensory organs. Specific sensory organs are
only rarely developed ; the oral feelers are considered to be olfactory
and gustatory organs ; the red spots at the tips of the arms of
Echinoids (on the terminal tentacles), and also the shining spots on
the integument of the Dt<nl<'ina1'nl<r and related forms, are thought to
be eyes ; the sphseridia of the Eehinoidea, and Baur's vesicles
(otolith vesicles) in the SynaptidcB and Ehtaipwla are considered to lie
organs of hearing or of orientation.

A. The Ambulaeral Appendages as Sensory Organs.

1. The Terminal Tentacles.

We are undoubtedly justified in assuming that originally the radial
canals of the water vascular system in all Echinoderms ended distnlly
in a freely projecting tentacle or feeler, which, covered with a highly
developed sensory epithelium, functioned as a sensory organ.

Such terminal tentacles are well developed in all Asteroids and
Opliinri>i<lx, and are found in them at the tips of the arms, where, as
we have seen, the radial water vascular trunks end. They are sup-
ported by the terminal plates of the skeleton, and are surrounded by
small spines which, when the slightest stimulus is applied to the
terminal tentacle, bend towards one another protectively over it.

It has long been known that the terminal tentacle of the Asteroids
carries a pigment spot, which has been regarded as an eye.

In the Eehinoidea, the terminal tentacle is already somewhat
reduced. The five tentacles are found on the five ocular plates of the
apical system, and the pore which perforates each ocular plate is the
aperture of the distal end of the water vascular trunk, which runs into
the tentacle.

The five terminal tentacles are still further reduced in the Holo-
thurioidea, where they lie round the anus.

Adult Crinoids have no terminal tentacles. The radial canals end
blindly before reaching the ends of the arms and pinnules.

The terminal tentacles, in contradistinction to all other ambulacra!
appendages (ambulacral feet, ambulacra! tentacles, papillae), may be
regarded as primary appendages. In the youngest stages of the
animal, when the radial water vascular trunks have only just arisen
as outgrowths of the hydrocoelomic vesicle, they lie close to the mouth.
The}- attain the position they occupy in the adult animal by the

VIII

EGHINODERMA TA— SENSORY ORGANS

growth and prolongation of the radial canals, during which process
bulgings arise alternately to right and left, which, pushing the integu-
ment before them, form the ambulacral feet or tentacles. Those tube-
feet are therefore most recent in origin which are nearest to the
terminal tentacles and furthest from the ring canal.

The terminal tentacle is the oldest ambulacral appendage ; it
receives the distal end of the radial canal, and is the only unpaired
ambulaeral appendage of the radius.

Special — 1. Asteroidea. — The tentacle which rests on the terminal plate is
covered by a very deep sensory epithelium, which consists of long, thin, sensory

FIG. 373.— Water vascular system of a very young Asteroid, l, Terminal tentacle ; •_', <.-\<-
spot at its base ; 3, ambulacral feet ; 4, railial canal of the water vascular system ; 5, circular canal ;
(>, mouth.

cells. It carries long cilia, and contains beneath the surface a thick layer of nerve
fibres, which represent the distal end of the radial nerve tract of the arm. At its
base, on the side turned to the mouth, it carries a vivid orange-red pigment spot.

2. Ophiuroidea. — The tentacle is surrounded by the terminal plate as by a ring.
There is no eye. The subepithelial radial nerve enters the terminal tentacle and
ends in its epithelium. (These facts recall an analogous arrangement in the
Annelida, where the ventral cord still remains epithelial in the growing caudal end
after it has become subepithelial in other regions of the body). In the Ettryalcc,
which have much-branched arms, no terminal tentacles have been found.

3. Echinoidea. — The terminal tentacle in the adult is usually reduced to a low
papilla which rises above the pore in the radial plate. In Echinocya/mus pusilln*

464

COMPA PA TI VE A XA TOM Y

CHAP.

alone (Fig. 374) does this papilla project somewhat further. The radial canal tra-
verses the pore and ends blindly under the epithelium of the papilla. The radial nerve
trunks also, and with them the epineural canal, pass through the pore. After reach-

ing the papilla, the fibres, which
had hitherto been subepithelial,
enter the epithelium of the
papilla and the epineural sinus
ends. The pseudoksemal sinus,
on the contrary, accompanies
the radial canal and the radial
nerve trunk only to the point
where these latter enter the
pore.

In a few rare cases the
radials are perforated by two
pores for the passage of two
terminal tentacles

Fi<;. 374.— Section through the terminal tentacle of

and certain Pahecchinoidca :
Melonites multipora, Pahcc-
chinus elcgans).

4. Holcthurioidea. — In
Cueumaria cuoiinis and C.

Echinocyamus pusillus (atu-r Cuenot). i, Body epithelium; Lacazii, the radial canals end
2, test : :-i. epineural canal ; 4, radial nerve trunk ; 5, pseudo- rouml the anus just ag in the
liii-inal canal ; ti, radial canal of the water vascular system ;
7, enilijtlieliiim of the body cavity ; s, genital circular sinus ;
9, terminal tentacle.

Echinoidca, but the last ex-
ternally visible trace of the

terminal tentacle has dis-
appi'ared. The radial canal perforates the body wall, accompanied by the radial
nerve trunk and the epineural canal, and ends blindly close below the surface. The
radial nerve unites with those "nests" of cells which represent the integumentary
epithelium and which were described above, p. 415. The pseudohaemal canal ends at
the point where the radial canal enters the body wall.

In other Holothurioidea (e.g. ffolothnrla impnticns) even the last (intrategu-
mentary) trace of the terminal tentacle is wanting. The Synaptidn have no terminal
tentacle, since the radial canals are altogether wanting in them. The distal ends of
the radial nerve trunks, however, perforate the integument, and in this we may
perhaps see a last trace of the terminal tentacle.

•2. The Ambulacral Feet and Ambulacral Tentacles.

Experiments on the living animal show that all the ambulacral appendages
are very sensitive to external stimuli, especially mechanical and tactile. If the
extended ambulacral foot of an Echinoid be irritated, it contracts, and the neighbour-
ing spines bend over it protectively. The same is the case with all the ambulacral
appendages of all Echinoderms. This is what we should expect from the rich inner-
vation of these appendages, which, in addition to their other functions, must be
regarded as organs of touch. Tactile functions could indeed be safely attributed to
them by merely observing how the long, thin, suckerless feet at the end of the
Asteroid arm, or the feet ou the anterior side of the Spataugoid body, stretch out
tentatively in various directions like the " horns " of a snail.

It cannot yet be demonstrated that all or any of the ambulacral appendages of
the Echinodermata have any other special sensory functions. It has been con-
jectured that the oral tentacles are gustatory, but in the Synaptidce alone have

vin ECHINODERHATA— SENSORY ORGANS 465

specific sensory organs (the so-called sensory buds) been found on the inner side of
the oral feelers (i.e. on the side turned to the mouth).

It has further been conjectured that the highly developed sense of smell of many
Echinoderms is located in ambulacra! appendages, and in the Asteroids it has been
localised especially in the terminal tentacles of the arms, above described. Decisive
experiments, however, have still to be made on this point.

The innervation of the ambulacra! appendages is as follows: The tube -feet
or tentacle nerves are always lateral branches of the radial nerve trunks of
the superficial system, these lateral branches being accompanied by ganglion
cells. When the base of the tube-foot or tentacle is reached the nerves have,
in the Echinoidea, an epithelial course ; but in the Ophiuroidca and Holo-
thurioidca, they are subepithelial even within the foot or tentacle. In the Crinoidea
and Asteroidea, these nerves, like the superficial nervous system, are, as we should
expect, epithelial.

The condition of the nerves within the appendages varies.

In the Asteroidea and Crinoidea, no localised distinct tentacle or tube -foot
nerve can be distinguished, a layer of nerve fibres being developed within the
epithelium of the entire foot. In the Ophiuroidca, Holothurioidea, and Echinoidea,
on the contrary, these nerves are distinct even in the feet (in the first two groups
they are subepithelial, and in the last epithelial), and their ramifications can be
clearly followed.

When the nerve reaches the base of a tentacle in the Ophiuroidea, it forms a large
semilunar or semicircular ganglion encircling the base, before it ascends into the
tentacle (Fig. 372, p. 456).

Round the terminal disc, in the tube-feet of the Echinoidea and Asteroidea, the
(epithelial) nerve tissue becomes thickened to form a nerve ring, from which the
nerve fibres (arranged as a whole radially) run inwards close to the epithelium of the
terminal disc.

The way in which the nerve fibres end in the epithelium of the ambulacral
appendages is still a disputed question. According to one view, the nerve fibres are
connected with filiform sensory cells, which have been brought to light by means of
maceration. According to another view, no such cells are present, the nerve fibres
and epithelial cells being merely in contact.

Special. — Any special terminal apparatus of the sensory nervous system has
rarely been observed in the ambulacral appendages. We have, however :—

(a) The sensory buds on the oral tentacles of the Synaptidcc, already mentioned
above. These are arranged in two longitudinal rows along the inner side of each
tentacle. They are conical or papilla-like projections of the tentacle wall, with a
pit-like invagination at the tip. A nerve from below the surface of the cutis enters
the base of the pit, which consists of strongly ciliated sensory cells. It has already
been pointed out that olfactory or gustatory functions have been attributed to these
sensory buds.

(b) The tentacles of the Crinoidea carry scattered sensory papilla; in the form of
distinct slender projections. Each papilla is composed of the fine processes of the
circle of sensory cells, which form its base. It is traversed by a shiny axial (muscle)
fibre, and carries at its freely projecting end three delicate, thin, immobile sensory
hairs (Fig. 356, 14, p. 413).

(c) Similar sensory papillae are found on the tentacles of the Ophiuroid form
Opkiactis virens.

(d) In certain species of the Ophiuroid genus Ophiothrix, the tentacles are
covered with circular rows of conical sensory papilla'. Each papilla seems to
consist of a bundle of long sensory cells, and carries freely projecting sensory hairs
(Fig. 375).

VOL. II 2 H

466

COMPA RA TIVE ANA TO MY

CHAP.

(c) Sensory papillae, sensory hairs, etc., have also been observed on the various
sensory feet of the Echinoidea.

[On the polymorphism of the ambulacral appendages, see the section on the
Water Vascular System, D, p. 431.

•2

B. Nerve Endings in the Integument.

A close plexus of nerve fibres is developed within the body epithelium of the
Echinoidea and Asteroidea. This plexus is more strongly developed, i.e. closer and
thicker at points which are specially sensitive to external stimuli, such as the

fascioles of the Echinoidea, round
•tf-ffc the bases of the pedicellariw, and on

the gills (the so-called papulre) of
the Asteroidea, and at the bases of
the spines of the EcMnoidea.

[For the sensory prominences on
the pedicellarise, see p. 399.]

In the Crinoidea and OpJii-
uroidea, sensory nerves continually
ramifying more and more finely,
run through the (calcified) cutis to
the surface of the body. The
manner in which these nerve fibres
terminate is unknown. Investiga-
tion on this point is the more
difficult as the epithelium appears
to be hardly distinguishable from
the cutis.

At the edges of the food grooves
(on the arms and the oral disc) of
the Crinoidea, alternating with the
trilobed tentacles, groups consisting
of five to six sensory cells with
delicate immobile hairs, occur.

Among the Holothurioidea, a
system of nerve fibres ramifying in
the cutis has been described in
' 'iii-iniiarin. From these branches
run to the nests of epithelial cells
sunk below the surface, which were
mentioned in connection with the
integument, p. 415. A similar
arrangement has been found in
other Adinopoda.

In the Paradinopoda (f>'ijnaji/i>.
Anapta) numerous scattered sensory or tactile papilhe are found on the integument,
which, at such points, bulges out to form prominences. At the centre of such a
prominence a group of sensory cells forms the tactile papilla. A distinct nerve runs
from each papilla to a large tactile ganglion lying in the cutis beneath it. The
epithelial cells surrounding the papilla are differentiated into glandular cells

FIG. 375.— Half of a transverse section through
an ambulacral tentacle of Ophiothrix fragilis (com-
bined from figures by Hamann). 1, Body epithelium ;
2, sensory papillse ; 3, cuticular connective tissue ;
4, nerves to the sensory papillae ; 5, longitudinal mus-
culature ; 6, epithelium of the tentacle canal (7) ;
8, ti-ntacle nerve.

VIII

ECHINODERMA TA —SENSOR Y ORGA NS

46'

C. Auditory Organs, Organs for Orientation.

Two types of organs for hearing or for orientation have been
observed in Echinoderms : (1) the auditory vesicles (Baur's vesicles,
otoeysts) of certain Holothnrioidea ; and (2) the sphseridia of the
EchiTioidea, which have already been described.

1. Auditory vesicles are found only in the Holothurioidea, and
among these in the Paractinopoda (Synaptidce), and among the Adino-
poda in the Ehmpoihi.

They have been best observed in the Synaptida' (Fig. 376). In
this family, five pairs of these vesicles occur in the cutis of the body
wall, near the tentacles, at the points where the five radial nerve
trunks emerge from the calcareous ring. On the outer side of each

FIG. 37ii.— Section through the two auditory vesicles of a radius of Synapta (after Cuenot).
], Epineural sinus ; 2, epithelial wall of the auditory vesicle; 3, otoliths ; 4, nervus acusticus :
5, longitudinal muscles ; 0, pseudohaemal canal ; 7, radial nerve trunk.

nerve trunk lies a pair of otoeysts. Each otocyst consists of a vesicle
filled with fluid, with a wall of (ciliated) plate epithelium. Numerous
otoliths are found vibrating in the fluid. These otoliths are vesicular
cells, the cavity of each being filled by a hard mass (phosphate of
lime). The nerves to the otoeysts (nervi acustici) come from the
radial nerve trunk.

The auditory vesicles of the Elasipoda occur in great numbers ;
there may be fourteen to one hundred or even more. Not infrequently
their manner of distribution is bilaterally symmetrical. For instance,
in Elpidia glacialis, six of the fourteen auditory vesicles occur on the
two lateral radii of the trivium, and one on each of the two radii of
the bivium. In this case no vesicles occur in the ventral median
radius,

2. The sphseridia of the

which are regarded as

468

COMPARATIVE ANATOMY

CHAP.

transformed spines, have already been described in the section on the
skeletal system (pp. 392-3). According to recent researches, there is
only a loose connection between the sphseridium and the shell tubercle,
fibres of connective tissue, and not muscle fibres, uniting them. When
an Echinoid is in the natural position, the sphseridia, which are
developed only on the oral side, hang down perpendicularly in their
niches or chambers, owing to the weight of the dense calcareous mass
which forms their rounded ends. They are thus able, by pressing on
the nerve cushion at their base, to orientate an animal as to its position
in space. Echinoids which are laid on the back, quickly turn them-

selves over again.

D. Eyes.

1. The eye-spots of Asteroids have already been mentioned
(p. 463) in connection with the terminal tentacles. A vivid red eye-

FIG. 377.— Section through the optic cushion at the base of the terminal tentacle of an
Asteroid. 1, Cuticle of the optic cup ; -2, pigment cells ; 3, cuticle of the tentacle epithelium (4) ;
5, nerve layer below the surface of the same ; 6, cutis of connective tissue ; 7, epithelium of the
tentacle canal.

spot is found at the base of each of these tentacles, on the side turned
to the mouth. On closer examination each eye-spot is found to break
up into a large number of single eyes, shaped like cups or hollow
cones. The tips of these conical cups are directed inward towards
the highly developed layer of nerve fibres below the surface of the
tentacle epithelium, while their cavities open outward (Fig. 377). The
wall of each optic cup is formed of pigment cells (with interspersed
unpigmented retinal cells). The cuticle of the tentacle epithelium

VIII

ECHINODERMA T A— SENSOR Y ORGA NS

469

also is continued into each cup. The portions of the cuticle which
belong to the different cells forming the wall of the cup are distinct
from one another, and have been described as rods.

Living Asteroids carry the arm with its tip directed upwards, so
as to enable the eye-spot to function.

2. In Diadema setosum (Eucckinoidea diadematoida), an animal
highly sensitive to light, the skin, which is as black as velvet, is

FIG. 378.— Part of a compound eye of Diadema setosum (after P. and F. Sarasin).

p, Pigment cups.

ornamented with numberless shiny blue spots ; these are gradually
lost on the oral side.

Each of these blue spots, when its surface is examined with the
microscope, breaks up into a number of pentagonal or hexagonal
portions. The number of these varies, according to the size of the
spot, sometimes being many hundreds. Each is a refractive body,
which stands in a cup of black pigment (Fig. 378). The blue colour

. 3

FIG. 370.— Section through the eye of Diadema setosum, diagrammatic (after P. and F. Sara-
sin). 1, Layer of ganglion cells; -2, "cornea"; 3, refractive body; 4, pigment cups; 5, nerve
layer, or plexus ; 6, fibres of connective tissue ; 7, collection of pigment below the nerve layer.

of the spots, which are regarded as compound eyes, is due to inter-
ference.

A section made through such an eye (Fig. 379) reveals : (1) that
the body epithelium, which is covered by a ciliated cuticle, spreads,
much thinned, over the whole eye (cornea) ; (2) that each "refractive
body " consists of a number of vesicular cells (modified epithelial
cells) ; (3) that a pigment cup (consisting of cells which are often
branched and star-like) surrounds the basal portion of each refractive
body ; (4) that the whole eye, with its numerous pigment cups, rests

470 COMPARATIVE ANATOMY CHAP.

directly upon a nerve layer provided with ganglion cells, which, at
the edge of the eye, passes into the usual layer of nerve fibres, found
in all Echinoids below the surface of the body epithelium.

Similar spots are found in other Diadematidce, and species of the related genus
Astropyga.

In the Synaptidcc (8. vittatd), at the base of each tentacle, two pigment spots
occur. It has now been proved that these " eyes" are sensory organs, but a detailed
account of them has not yet been given.

XIII. The Body Musculature.

The special development of the body musculature of the Echino-
derms is directly connected with the peculiar development of the
skeleton.

.Regarding the musculature and the skeleton alone, the Echino-
derms may be divided into three groups.

Holothurioidea. — The skeleton here consists merely of isolated,
and usually microscopically small, calcareous bodies. The presence of
these in the leathery integument does not prevent changes of form in
the tubular body. The body musculature is developed as a dermo-
museular tube. By means of co-ordinated contractions of the
circular and longitudinal musculature of which this tube is com-
posed, the animal is able to make slow, worm-like movements.

Asteroidea, Ophiuroidea, and Crinoidea. — The body in these
classes is drawn out in the direction of the radii in the form of
arms, which are supported by a jointed skeleton. The dermo-
museular tube breaks up into separate muscles, which connect the
joints of the skeleton together. In the Asteroidea alone, a dermo-
muscular tube persists side by side with these isolated muscles.

Eehinoidea. — -The skeleton of the Echinoidea (with but few
exceptions) is a rigid test or capsule. A body musculature would
here be useless, and is therefore wanting.

An exception to this rule is found among the Euecliinoidea in the
Streptosomata, in which the plates of the more or less flexible test
imbricate. It has now been proved that in the Echinothvaridce, five
pairs of muscle lamellae run in meridians from the oral to the apical
side on the inner surface of the test. The contraction of these
muscle lamellae causes a depression of the test.

The musculature of the Echinodermata consists, as a rule, of smooth muscle fibres.
A longitudinally striated fibril of contractile substance lies on one side of the undiffer-
entiated protoplasm of the formative cell which contains the nucleus. Transversely
striated muscle fibres are of less frequent occurrence, but are found in the adductor
muscles of the seizing pedicellariae in the Echinoidea (p. 398), and in the muscles of
the rotating anal spines of Centrostcphanus longispinus.

The musculature of the various organs or systems of organs will be described in
the sections dealing with those organs.

vin ECHINODERMATA— BODY MUSCULATURE 471

A. Holothurioidea.

The dermomuscular tube consists everywhere of an outer circular
muscle layer, and of five radial longitudinal muscles (Figs. 371 and
383, pp. 451 and 477).

The circular musculature lies immediately within the cuticle. It
is usually interrupted in the five radii, and thus consists of five longi-
tudinal interradial strips or bands of muscle, the fibres of which run
transversely. In the Pamd'mopoda (Syna-ptidw) alone, where there are
no radial canals of the water vascular system, the fibres run uninter-
ruptedly round the body.

The longitudinal musculature consists of five strong muscles or
pairs of muscles traversing the body longitudinally along the radii.
These muscles thus cover, on the side of the body cavity, the radial
organs enumerated on p. 409. Anteriorly (at the oral pole) they are
inserted into the radial pieces of the calcareous ring, posteriorly (at
the apical pole) they end near the anus.

In the Dendrochirotce, the longitudinal musculature is differentiated
in a peculiar manner. At the middle of the body, or somewhat in
front of it, the fibres of each of the five longitudinal muscles divide
into two bundles. One of these bundles is continued simply as a
longitudinal muscle along the body wall, while the other freely tra-
verses the body cavity, and is attached anteriorly to a radial piece of
the calcareous ring (Fig. 349, p. 404). The retractor muscles of the
oral region have in this way been derived by the splitting up of the
originally simple longitudinal muscles, and this specialisation became
more marked as the oral tentacles of the Dendrochirotce became more
and more highly developed, and required increasing protection.
Species are to be found in which the separation and branching off' of
retractors from the longitudinal muscles has not yet been perfected.

Apart from the Dendrochirotce, retractors occur only in the genus
Molpadia, and in species of the genera Chirodota and Synapta.

B. Eehinoidea.

The longitudinal muscles of the Echinothuridce (Asthenosoma) have
the shape of semilunar leaves, the convex sides of which are directed
outwards, and are attached to the inner surface of the test ; the
concave edges, on the other hand, face the axis of the test (Fig. 380).
They are inserted into the test at the boundaries between the
ambulacra and the interambulacra, at the lateral edges of the ambu-
lacral plates.

In each muscle lamella, the fibrous bands radiate fan-like (Fig. 380) from a
"centrum tendineum " on the inner edge of the lamella. The uppermost fibres are
attached to the radials, the lowermost to the outer side of the auriculae. Anasto-
moses between the fibres are not infrequent.

472

COMPARATIVE ANATOMY

CHAP.

The five pairs of longitudinal muscles or pairs of muscle lamella? in the Ecliino-
thuridie invite comparison with the five longitudinal muscles or pairs of muscles in
the Holothurioidea. No true homology can, however, be proved with certainty,

m

SiSSss:-.'/'// i.ii\"-v\ N • ^ 11 '

•; /,/',. /.Ili'ls?-^. aS&flftWlM

r

Fig. 380. — Test of Asthenosoma, broken open so as to show the longitudinal muscles. 1, In-
terambulacral plates ; -2, ambulacra! plates; 3, radial canals; 4, centrum tendineum ; 5, muscle
bands ; 6, ambulacra! apophysis (auricula) ; 7, opening muscle of the teeth ; 8, retractors of the
masticatory apparatus.

since no direct relation between the calcareous ring of the Holothurian, and any
definite portions of the Echinoid skeleton (such as the auricules, or the pieces of
the masticatory apparatus) can be established.

C. Asteroidea.

On the apical side of the arms and of the disc, a dermomuscular
layer has been observed, which consists of external transverse, and
internal radial fibres, and runs lengthwise in the arm. This does not
appear to spread (as a muscle layer) to the oral side of the body, where
the ambulacra! skeleton is developed. It may perhaps, however, have
become differentiated here into the special musculature of the ambu-
lacral skeleton.

This latter is developed as follows : —

Ten muscles occur in each skeletal segment.

1. On each side two vertical muscles (or bands), one distal and the other
proximal, connect the adambulacral with the ambulacra! plate (cf. Fig. 309, p. 351).

2. On each side an upper longitudinal muscle connects every two consecutive
ainbulacral plates on their apical side (that turned to the body cavity). The func-
tion of these muscles is to bend the arm upward (Fig. 381, 2 and 7).

3. On each side a lower longitudinal muscle connects every two consecutive
adambulacral pieces ; this muscle counteracts the upper longitudinal muscle.

VIII

ECHINODERMATA—BODY MUSCULA TURE

473

4. An upper transverse muscle connects the two opposite ambulacral plates of
one and the same skeletal segment, on their apical side (that turned to the body
cavity). These muscles, by their contraction, widen the ambulacral furrow (Fig.
382, 3 and 6).

5. A lower transverse muscle connects the two ambulacral plates of a segment
on the lower side, which is turned to the furrow. These muscles, by their contrac-
tion, narrow the furrow.

The musculature of the oral skeleton (Fig. 381) is arranged as follows :—

11

FIG. 381.— Oral skeleton and basal part of the brachial skeleton of Pentaceros turritus,

with the musculature of these parts (after Viguier). From within. 1, First adambulacral plates ;
I-IV, first to fourth ambulacral plates ; or, orals ; 2, dorsal longitudinal muscles ; 3, dorsal trans-
verse muscles (for opening the ambulacral furrow) ; 4, interbrachial pillars ; 5, muscle apophyses
of the first adambulacral plates ; 6, facets of the ambulacral plates for the attachment of the dorsal
transverse muscles ; 7, ditto for the attachment of the dorsal longitudinal muscles ; 8, transverse
muscles between the first adambulacral plates (teeth); 0, dorsal transverse muscles between the first
pair of ambulacral plates ; 10, dorsoventral muscle ; 11, stone canal ; 12, crossed ligament ; 13, ab-
ductor dentium ; 14, adductor dentium ; a, aperture for the first ambulacral feet.

1. A single or double radial muscle connects the distal ends of the first adam-
bulacral plates (teeth, 1 in Fig. 310, p. 352) of one and the same radius (Fig. 381, 13),
and opens these plates.

2. A muscle, which connects the distal ends of the first two adambulacral plates
of two neighbouring radii, and is therefore interradial. These interradial muscles,
by contracting, close the pairs of teeth (Fig. 381, 14). This closing action
is assisted by a transverse muscle, which joins the opposing edges of each pair of
teeth (8).

3. The first pair of ambulacral plates of a radius, like all succeeding pairs, are
connected together by means of a dorsal transverse muscle (Fig. 381, 9).

474 COMPARATIVE ANATOMY CHAP, vin

4. Five pairs of dorsoventral muscles connect the first five pairs of ambulacra!
plates with the dorsal wall of the disc. By the contraction of these, the apical wall
of the disc is approximated to its oral wall (Fig. 381, 10).

D. Ophiuroidea.

A dermomuscular tube is altogether wanting. The muscles which
move the brachial skeleton (the intervertebral muscles) have already
been described, p. 357.

The musculature of the oral skeleton (cf. Figs. 314 and 386, pp. 359 and 486).

1. A musculus interradialis externus connects transversely the opposite distal
surfaces of the oral-angle plates of neighbouring arms. This is the most powerful
of the muscles.

2 and 3. The two oral-angle plates of one and the same arm are connected
by an upper and a lower transverse muscle (musculus radialis superior et inferior),
and approximated by means of their contraction.

The three muscles just described form an outer circle, which is followed, orally,
by a second inner circle, consisting of the following muscles : —

4. A musculus interradialis internus inferior connects transversely the oral
ends of the oral-angle plates of the different arms.

5. The innermost muscles of the oral skeleton consist of fibres which radiate out-
wards. They run as five interradially placed pairs of muscles, from the oral-angle
plates to the teeth (in Ojihiadis only to the upper teeth), for whose movement they
serve. These are the musculi interradiales intern! superiores.

E. Crinoidea.

A dermomuscular tube is wanting. The musculature which moves
the jointed skeleton has already been described, p. 376.

XIV. The Alimentary Canal.

A. General Review.

The alimentary canal, which runs through the body cavity, being-
attached or suspended to the body wall in various ways by means of
mesenteries or mesenterial filaments, commences with the mouth and
ends with the anus.

The absence of the anus in the Oj>hinr<>iilai and in the Asteroid
family Astropedinidce cannot be regarded as an original condition.

In no ease does the alimentary canal run as a straight tube from
mouth to anus, although, in many Synaptidw, its condition is almost
as simple. As a rule, the secreting and resorbing surface of the canal
is increased in one of two Avays :—

1. The alimentary canal, between the mouth and anus, becomes
increasingly lengthened, and thus necessarily forms loops, and has a
winding" course (Holothurioidea, Echinoidea, Crinoidea).

R
Hblothuria

md

os

anus •'

C

•fldinoinetra

'-/,

anus

"• anus
Fio. 382.— Diagram of the course of the alimentary canal in various Echinoderms. The

body is viewed from the oral side, md, ins, Mesenteries ; tig, food-grooves or ainbulacral furrows
of the tegmen calycis ; m, madreporite ; x, commencement of the backward coil of the intestine ;
ii first, i'o second (backward) coil of the intestine ; s, siphon, accessory intestine ; 1 and 2 (in F)
the two points at which the accessory intestine enters the principal intestine ; sj, free portion of
the accessory intestine ; s2) portion of the same in contact with the principal intestine ; coec, in-
testinal poecum. [Combined from several sources.]

476 COMPARATIVE ANATOMY CHAP, vin

2. The alimentary canal runs direct, without coiling, from mouth
to anus, but has sae-like widenings (Ophiuroidea), and further, in the
Asteroids, sends off branched outgrowths into the arms.

The wall of the intestine, as a rule, in the Echinoderm, consists of
the following layers : (1) a deep inner epithelium, with numerous
glandular cells; (2) a layer of connective tissue ; (3) a muscle layer ;
(4) an outer epithelium, the endothelium of the body cavity. A
system of blood lacunae (absorbing canals) is developed in the layer
of connective tissue in the Holothurioidea, Echinoidea, and Crinoidea.

B. Holothurioidea (Figs. 371, p. 451, and 383).

Course of the alimentary canal. — The mouth lies at the oral
pole (i.e. at the anterior end of the body), the anus at the apical pole.
For the exceptions to this rule, especially Rhopalodina, in which the
mouth lies close to the anus, cf. section III , p. 408.

The alimentary canal is, as a rule, considerably longer than the
body (on an average three times as long), and therefore has a looped
or winding course. From the mouth, it first runs backward towards
the anus (first or anterior section), it then bends for the first time,
and runs forward again (second or middle section) ; lastly, it bends
again near the anterior end of the body, and runs backward once
more, this time to the anus (third or posterior section).

In making these bends, the alimentary canal forms a spiral round
the principal (longitudinal) axis of the body ; this is very clearly
seen by following the line of attachment of its mesenteries to the
body wall.

The anterior section is attached to the median dorsal line interradially. From
this, at the first bend, the mesentery passes across the left dorsal radius into the left
dorsal interradius. The whole of the central section is attached in this interradius.
At the second bend, the mesentery passes over the left ventral radius and interradius,
and over the middle ventral radius, into the right ventral interradius. The third
or posterior section is attached in this latter interradius (Figs. 350, p. 407, and
383).

If a Holothurian is placed upright, with the oral pole upward and
the apical downward, and if we project the loops of the alimentary
canal on to a horizontal plane, or, if we simply view the intestinal
loops of a Holothurian from the oral pole, we see that the digestive
tract runs from left to right, i.e. in the direction of the hands of a
clock. In other Echinoderms with coiled intestine, the coils also run
in this direction.

It was mentioned above that the alimentary canal of many Parac-
tinopoda (Syna/ptidce) is almost straight. This is, however, not an
original condition, as is seen from the following facts : (1) the older
larva and the quite young Synapta have a bent alimentary canal ; (2)
the intestinal mesentery is inserted in the body wall exactly in the

Fig. 383.— Organisation of an Aspidochirotan Holothurian. In the dorsal interradius, the
body wall is cut through on the left, near the dorsal mesentery, and is spread out (after Ludwig, in
Leuckart's Tvfelwcrl-). 1, Genital aperture ; 2, radial plates ; 3, interradial plates of the calcareous
ring ; 4, genital duct ; 5, dorsal or anterior mesentery of the intestine ; 6, stone canals with their
inner madreporites ; 7, dorsal intestinal vessel ; 8, gonads ; 9, anterior section of the alimentary
canal ; 10, ventral intestinal vessel ; 11, posterior section of intestine ; 12, longitudinal muscles ; 13,
posterior edge of the dorsal mesentery ; 14, right branchial tree (water lung) ; 15, circular muscula-
ture of the body wall ; 16, longitudinal muscles ; 17, cut edge of the body wall ; IS, longitudinal
muscles ; 19, partly muscular filaments running from the wall of the cloaca to the body wall ; 20, cloa-
cal aperture ; 21, cloaca ; 22, Cuvierian organs ; 23, left branchial tree ; 24 and 25, longitudinal muscles ;
26, middle section of the alimentary canal ; 27, vascular anastomosis ; 28, fore-gut ; 29, Polian
vesicle ; 30, blood vascular ring ; 31, water vascular ring ; 32, commencement of the radial vessel.

478 COMPARATIVE ANATOMY CHAP.

manner above described, and thus runs in a spiral ; (3) in most cases,
close examination reveals that even in the adult the canal is coiled in
a spiral, although drawn out to a great length, and that the first and
second bendings can still be distinguished as slight curves.

If the typical alimentary canal of an Adinopod were to be
shortened until it was of almost the same length as the body, the con-
dition found in the Synaptidce would arise.

The divisions or sections of the alimentary canal. — In the
intestine of the Holothurioidea, consecutive sections have been
distinguished, but these are never very marked microscopically.
Throughout its whole course, the canal retains its tubular shape.
The different sections are distinguished by their sizes and by the
thickness of their walls, by their colour, their vascularisation, and
especially by their histological structure. The boundaries of the con-
secutive sections are usually externally indicated by circular constric-
tions of varying distinctness ; these constrictions not infrequently
correspond with circular folds projecting into the canal.

The mouth. — Around the mouth, the circular musculature becomes thickened into
a small sphincter muscle.

The more strongly the oral tentacles are developed, the more marked is the
capacity for invaginating the mouth with its tentacles, and with a larger or smaller
portion of the anterior end of the body, into the body cavity. In the Dendrochirotcr,
in which the tentacles are strongly developed, the invaginable portion of the anterior
end of the body is called the proboscis. It is not infrequently distinguished by the
different colouring and constitution of its integument. In all cases, in imagination,
the chief part is played by the retractor muscles (<•/'. p. 471). At the posterior
boundary of the proboscidal region five (interradial or radial) calcareous valves are
occasionally developed ; these, -when the proboscis is invaginated, close the aper-
ture (e.g. Psolus, Figs. 227 and 228, p. 287).

The oesophagus reaches from the mouth to the circular canal of the water
vascular system, or even further. It is attached to the water vascular ring, the cal-
careous ring, the radial canals of the water vascular system, etc., by means of bands
which run out radially, traversing the perioesophageal sinus (see Fig. 365, p. 428).
These bands are chiefly of the nature of connective tissue, but also contain muscle
fibres. The oesophagus, with the complex of surrounding organs, is sometimes
called the pharyngeal bulb.

The cesophagus is followed by a shorter portion known as the stomach intestine,
and this again by the longest part of the digestive tract, the small intestine. This
last forms the larger posterior portion of the first section of the intestine, the whole
of the second section, and by far the greater portion of the third and last section.

The last part of the alimentary canal, the cloaca or rectum, is distinguished by
special thickness, and is attached by radially arranged strands and filaments to
the neighbouring body wall. These strands consist of connective tissue and muscle
fibres.

Into the cloaca or rectum open the water lungs and the Cuvierian organs,
where these are present. These will be described, pp. 487, 488. In some Elpidn<!" .
the anterior part of the cloaca bulges out to form a large cnecum, which projects more or
less far into the body cavity, sometimes reaching almost to the middle of the body.
Since the Elp'uliidic possesses no water lungs, there is some justification for the

viii ECHINODERMATA— ALIMENTARY CANAL 47!J

suggestion recently made that their cloacal ctecum may function as a rudimentary
organ of this kind.

The inner surface of the alimentary canal often shows a longitudinal fold.
Transverse intestinal folds, arranged in longitudinal rows, have been observed in
the small intestine of the Aspidochirotcc.

Finer structure of the alimentary canal. — The wall of the digestive tract
consists of the following layers, which may vary greatly in thickness and special
structure in the different sections : (1) An inner intestinal epithelium ; (2) an inner
layer of connective tissue with the blood lacunte ; (3) a muscle layer (generally con-
sisting of an inner layer of longitudinal and an outer layer of circular fibres, hut in
some Syuaptidce and Aspidoclurotce this order is reversed) ; (4) an outer layer of con-
nective tissue (often very thin) ; (5) the ciliated endothelium of the body cavity.

The inner intestinal epithelium is ciliated, especially in the small intestine.
Over most of the digestive tract it is found as a very deep epithelium covered by a
fine cuticle, its cells being pallisade or thread cells. Glandular cells of various
sorts are specially numerous in the epithelium of the stomach. The epithelium of
the cloaca resembles the outer body epithelium. Into it open the processes of long
subepithelial glands, which are unicellular and tubular.

The anus can be closed by means of a sphincter muscle. Calcareous plates,
papilhe, etc., may occur round it.

^Hl C. Echinoidea.

For the position of the mouth and the anus, cf. p. 338.

In all adult Echinoidea, the length of the tubular intestine is
greater than that of a straight line from the mouth to the anus, so
that the course of the alimentary canal is necessarily coiled.

The simplest arrangement is found in the Glypeastroida (Fig. 382,
E, p. 475). The direction of the intestinal coils will here be described
in the same way as in the Holothurioidea, the viscera being viewed
from the oral side. After traversing the masticatory apparatus, the
alimentary canal turns to the right (following the direction of the
hands of a clock), and makes rather more than a complete coil round
the principal axis of the body. So far, the course exactly resembles
that of the intestine in the Holothurioidea. In the Clypeastroida,
however, the canal now bends back upon itself, and runs direct back
to the anus, which in this division of the Echinoidea lies orally in the
posterior unpaired interradius.

In the endoeyelie or regular Eehinoidea, the arrangement is not
simpler, but still more complicated. After leaving the masticatory
apparatus, the alimentary canal ascends towards the apical system,
then bends round and follows the direction of the hands of a clock
(attached to the inner surface of the test by the mesentery) till it has
run about once round the principal axis. It then bends back upon
itself, coiling in the reverse direction backwards to the apical anus
(Fig. 382, D, p. 475).

The intestine of the exocyclic Spatanijn'ulni resembles in its course
that of the endoeyelie Echinoidea with one difference, caused by the
facts that the mouth has shifted anteriorly along the oral surface, and

480 COMPARATIVE ANATOMY CHAP.

the anus, out of the apical system into the posterior interradius. The
mouth therefore draAvs the commencement of the first intestinal coil
(Avhich runs in the direction of the hands of a clock) forward, Avhile
the anus draws back (i.e. posteriorly) the end of the spiral Avhich runs,
as above described, backAvards (Fig. 382, F, p. 475).

It is Avorthy of note that, in quite young Spatangoidea (Hemiaster
cavernosus, 2 mm. long), the intestine, which appears to end blindly,
ascends direct from the oral to the apical pole. At a rather later
stage the mouth is still central, Avhile the apical end of the alimentary
canal has already someAA'hat shifted, and opens through the anus out-
side of the apical system. At this stage (Avhen the length of the
animal is from two to three mm.) the intestine runs up from mouth
to anus in one single coil, as a spiral in the direction of the hands
of a clock. The complicated arrangement in the adult is thus
secondary, and is no doubt due to the fact that the canal increases in
length more than does the interval between mouth and anus.

Finer structure of the intestinal wall. — This agrees essentially with the struc-
ture described in connection Avith the Holothurioidea. No distinct sections can be
made out in the alimentary canal. That part of it which runs through the mastica-
tory apparatus is often called the pharynx. Its lumen in section is five-rayed, the
layer of connective tissue forming five longitudinal ridges which bulge in the
epithelium. It is connected, in a manner which cannot here be further described,
bv means of five pairs of longitudinal bands of connective tissue, with the surround-
ing masticatory apparatus.

The name oesophagus is generally given to the portion of the digestive tract
which follows the pharynx (and, in the Spatangoidea, to the whole of the first por-
tion of the intestine) as far as the point where, in regular Eehinoids, there is a sac-
like widening, and in the SjMttangoidca a large caecum. In regular Echinoidea, it
includes that part of the intestine which ascends from the lantern towards the
apical system, together with the first portion of the first spiral. In the Spatangoidea
it runs back from the mouth and then bends forward, forming the commencement
of the first intestinal spiral.

The cesophagus is followed by the first intestinal spiral, which runs in the
direction of the hands of a clock. It commences with a slight sac-like swelling
(regular Eehinoids) or a large caecum (Spatangoidea}. In this part of the alimen-
tary canal a rich system of blood lacunae is developed in the connective tissue layer
on the inner side of the otherwise Aveak musculature.

In the second or reverse spiral this lacunar network is wanting. This spiral is
distinguished, more especially in the regular Echinoidea, by its peculiar colouring,
being yellow, whereas the first spiral appears brown.

In regular Echinoidea, the two intestinal spirals have an elegantly undulating
course, regularly ascending and descending.

The second spiral passes without any sharp boundary into the rectum, which, in
the Spatangoidea, runs back from the middle of the body. At its commencement, in
Echinocardium (flavcsccns) and Schizastcr, it has a small diverticulum.

The alimentary canal of the Spatangoidea, which is distended with sand and
mud, is thicker and its walls are firmer than in the regular Echinoidea, whose in-
testine usually contains, besides mud, a large number of unicellular algae. There is a
corresponding ditference in the mesenteries. In the regular Echinoidea the intes-
tine is attached by means of mesenteries practically only to the test, and these

vin EGHINODERMATA— ALIMENTARY CANAL 481

mesenteries are broken through in such a manner as to form elegant arcades. In the
Spalangoidea, however, the different coils of the canal are further connected together
inter sc by mesenteries which are not perforated.

Unicellular glands of various kinds are found, chieHy in the epithelium of the
first section of the intestine. In the Spatangoidea, in the commencement of the first
spiral, there are multicellular flask-shaped glands ; these lie in the connective tissue
layer, the neck-like duct alone opening into the lumen of the alimentary canal.

The accessory intestine (siphon), which occurs in nearly all
Echinoidea, deserves special attention. Near the commencement of
the first spiral, the siphon branches off from the main intestine as a
narrow tube, which again enters the intestine at the end of that spiral,
to which it thus belongs. The siphon always runs along the inner
side of the main intestine (that turned to the principal axis of the
body). In regular Echinoids, it follows the main intestine in its
course ; in the Clypeastroida, on the contrary, its course is somewhat
shortened. In the Spatangoida, the first part of its course is shorter
than that of the main intestine, while the rest follows the coils of the
latter.

The Cidaroida (Dorocidaris paplllata) have no distinct siphon, but
it is very probable that this organ is here represented by a longitu-
dinal furrow bordered by two folds, which furrow is either not yet, or
no longer, shut off' from the lumen of the intestine. This furrow
occurs in the same region of the intestine as the siphon, and also on
the axial side of the canal.

In the Spatangoid genera Brissus, Brissopsis, and Schizaster a
second siphon has been discovered, running parallel to the intestine.

The structure of the siphon resembles, in essential points, that of
the main intestine. It has been conjectured that it, like the accessory
intestine of certain worms, subserves intestinal respiration.

D. Crinoidea.

In this class the alimentary canal is tubular. It descends from
the mouth into the calyx, coiling in the direction of the hands of a
clock (when the body is viewed from the oral side). From the base
of the calyx it again ascends, continuing the same curve, towards the
tegmen calycis, and here enters the anal cone in the anal interradius ;
it then runs through the anal cone, opening outward at its tip through
the anus.

During its course through the calyx, the intestine makes one
complete coil round the principal axis (Fig. 382, B, p. 475). The
alimentary canal of Adinometra affords a striking exception to this
rule, forming, in the same direction as in other Crinoids, as man}r as
four coils (Fig. 382, C). It may be remembered that Adinometra is
also distinguished from all other Crinoids by the eccentric position of
the mouth in the tegmen calycis.

The section of the intestine which lies at the bottom of the calyx

VOL. II 2 I

482

COMPARATIVE ANATOMY

CHAP.

Fin. 384.— Radial-interradial section in the direction of the principal axis through the
calyx of Pentacrinus decorus (after P. H. Carpenter), left half radial, right interradial, dia-
grammatic. In many respects the figure is obsolete and incorrect. The most noteworthy points
are emphasised by special type in the following lettering, so, Subambulacral plates ; ?•«, radial
nerve trunk (hatched) ; /v, radial pseudohivmal canal (black) ; ra, radial canal of the water vascular
system (dotted); cl], brachial crelom ; pi, spongy organ of the blood vascular system; an, circular
canal of the water vascular system ; n, mouth ; an (to the right), nerve ring ; av, blood vascular
ring? pseudolni-mal ring?; mi (to the left), anus; re, rectum; in, plates of the interambulacral
areas ; jm, calyx pores ; cl, the coelom, traversed by a spongy network ; it, intestinal coils
(cut through); gp, axial organ; in.ii, muscles; im, nerve commissure between the axial trunks;
IHI, basal ; n; cirrus canal ; en, cirrus nerve ; ch, chambered sinus, with the centre of the apical
nervous system; >; radial plate; r}, <•._,, first and second costal; iii, distichal ; .••//, syzygial
suture ; gv, genital canal ; ai-n, apical nerve trunk of the arm.

via ECHINODERMATA— ALIMENTARY CANAL 483

is occasionally somewhat widened, and is then called the stomach.
In Ilhizocrinus and Batliycrinus there are, on the external side of the
digestive tract, interradially placed outgrowths. Similar outgrowths
occur in great numbers on the inner side of the tract in Antedon (the
side facing the axis of the calyx). Such a diverticulum, when espe-
cially large and branched, has been called a hepatic caecum, but this
name must not be accepted in any strict sense.

The finer structure of the intestine agrees in essentials with that
in other Echinoderms. The intestinal epithelium is everywhere ciliated
except in part of the rectum. The musculature is weakly developed
or altogether wanting, except near the mouth and in the rectum,
where sphincters are formed. The anal tube or cone consists of the
body wall externally and of the wall of the rectum internally.
Between these two the reduced body cavity is traversed by radially
placed strands of connective tissue.

E. Asteroidea (Fig. 385).

That part of the oral area which is left free by the skeleton is
covered by a soft oral integument, in the middle of which lies the
mouth. This organ can be opened by muscles which run out from it
radially in the oral integument ; it can be closed by circular muscle
fibres, which run round it in the oral integument and in the
cesophagus.

The mouth leads into an oesophagus, which ascends perpendicu-
larly, widens rapidly, and passes over without any sharp boundary
into the stomachal sac.

In Ecliinastcr sepositns, the cesophagus has around it ten outgrowths, whose
walls are very much folded, and whose (inner) epithelium is richly supplied with
glands.

The membranous stomachal sac of the Asteroids is very spacious,
filling the whole disc. Its wall is irregularly folded, and provided
with outgrowths ; it is connected with the wall of the disc by means
of mesenterial strands, partly of connective tissue, partly of muscle.

In the upper (apical) portion of the stomachal sac, five pairs of
brachial diverticula open ; these are the radial cseca, or hepatic
appendages, which stretch more or less far into the arms, according
to family, genus, or species. There is one pair in each arm. These
diverticula of the stomachal sac (which, ontogenetically, develop com-
paratively late) have the following general structure. Each diver-
ticulum consists of a median common tube, which runs in the longi-
tudinal direction of the arm, giving off lateral tubes alternately to
right and left. Each lateral tube receives from each side the openings
of closely crowded glandular lobes, so that the secreting surface is
very large.

484

COM PARA TIVE ANA TOM Y

CHAP.

In the Ecliinasteridce and Asterinidce, the common tube swells into
a large sac.

At the point where the stomachal sac narrows to form the short

FIG. 385. — Alimentary canal and genital organs of an Asteroid, diagrammatic. 1, Braclual
diverticula of tin.- stomach ; -, gonails ; IS, base of the gonacl, wliicli corresponils with its api-rtun- ;
4, stomachal sac ; 5, anus ; 6, rectal diverticula ; 7, apical circular sinus and trunk ; S, one of the
ten radial sinuses and trunks nmiiini; from this latter to the gonads; 9, stone canal in the axial
sinus; 10, madreporite.

rectum, i.e. high up in the apex of the disc, it is once more provided
with diverticula. These rectal diverticula, whose number, arrange-

vni ECHINODERMATA— RESPIRATORY ORGANS 485

ment, and size are subject not only to specific and generic, but often
also to individual variations, are, as a rule, much smaller than the
brachial diverticula of the stomach, with which they otherwise agree
in structure.

The anus is wanting only in the AstropectinidcB. Elsewhere it
lies somewhat eccentrically (never exactly at the apical pole) in the
interradius which follows the madreporitic interradius (in the direction
of the hands of the clock), when the disc is seen from the apical
side.

Finer structure of the intestine. — The intestinal epithelium is ciliated. Glandular
cells, as goblet cells, mucus cells, and granular cells, are everywhere found together
with epithelial cells. The last named appear to secrete especially the digestive
ferment ; they preponderate at the commencement and terminal part of the canal,
and are particularly numerous in the brachial and rectal diverticula. The muscle
layer is well developed in the a-sophagus, the rectum and the rectal diverticnla, less
strong in the stomach, and wanting in the brachial diverticula.

The manner in which the brachial diverticula are suspended to the apical brachial
wall has already been described (p. 440).

The Asteroidea are carnivorous, feeding on other marine animals, especially
Bivalves and Gastropods. When feeding, they evaginate the greater part of the
stomach out of the oral aperture, enveloping their prey with it. The secretion of
the mucus cells yielded during the process appears to be poisonous and to have a
decomposing action. The animals attacked quickly die, and are passed on to the
part of the stomach still remaining within the disc, where they undergo the digesting
action of the secretion yielded by the granular glands (Kb'rnerdrusen).

The evagination of the stomach is brought about by the musculature of the disc,
and its withdrawal by the (partially) muscular mesenterial bands which attach it to
the body wall.

The anal aperture certainly does not serve for the ejection of all the faecal masses.
It is impossible that large masses, such as the shells of Bivalves and Gastropods,
which are found in the stomachs of Asteroids, can be ejected through such a narrow
aperture ; they are no doubt passed out again through the mouth.

F. Ophiuroidea (Figs. 386 and 388, p. 494).

The condition of the alimentary canal in this class is simpler than
in any of the others. The somewhat spacious buccal cavity which is
surrounded by the oral skeleton leads into the digestive sac which
fills the body cavity of the disc, in so far as it is not occupied by the
bursse. An anus is wanting. Special intestinal appendages in any
Avay corresponding with the brachial or rectal diverticula of the
Asteroids are wanting.

XV. Respiratory Organs.

There are no respiratory organs which are homologous throughout
the whole phylum of the Echinodermata. Portions of the body
belonging to very different organs and systems of organs are function-

486

COM PA RA TI VE ANA TOMY

CHAP.

ally modified for the purpose of respiration. All these respiratory
organs (except the respiratory trees of the Holothurioidea) are described

FIG. 380. — Radial - interradial section
through the disc and the base of the arm
of an Ophiurid, in the direction of the prin-
cipal axis. Left half interradial, right half
radial (after Ludwig). m //, //< //x, mil a, Muscles
of the oral skeleton ; an (black), nerve ring ;
r</»._,+«(7j, oral-angle plates; or, oral = ventral
wall of the disc; aav, apical circular sinus
with ring-like strand (cf. Fig. 390). vP, Polian
vesicle ; lii, niesenterial filaments between the
stomachal sac and the wall of the disc ;
d, ccelom ; ulw, apical wall of the disc ; m,
radial canal of the water vascular system ;
rjih, radial psi-udohivmal canal ; IT, continua-
tion of the axial organ in the arm(?), radial
blood vessel(?); rn, radial nerve trunk ot'tlin
oral system: bsj-bsg, first to sixth ventral
shield ; teoi and teo->, first and second oral
tentacles ; /</, torus angnlaris ; L>, teeth ;
o, buccal cavity ; la, entrance to the stomachal
sac; f'yi/i, peripharyngeal sinus; «»ii, peri
stomal plates ; nn, water vascular ring ;
it, stomachal sac ; «;/(•_> — amg, second to sixth
brachial vertebral ossicles ; spt, septum be-
tween the body cavity and the peripharyngeal
sinus.

in connection with the systems of organs to which they belong,
review of these orerans will be found below.

vin ECHINODERMATA— RESPIRATORY ORGANS 487

A. The (inner) Respiratory Trees of the Holothurioidea (Figs. 371

and 383, pp. 451 and 477).

These organs, which are known as water lungs or respiratory
trees, occur as two tree-like delicate-walled branched canals or tubes,
which lie to the right and left in the body cavity, their principal
trunks opening posteriorly into the anterior part of the cloaca. They
open either separately or through a common terminal portion. The
last branches of the respiratory trees end in vesicular widenings,
similar " ampullae " being found also along the branches themselves.
When well developed, the respiratory trees reach far forward into the
body cavity, being attached at various points by muscle fibres or
filaments of connective tissue to adjacent organs, i.e. to the body wall,
the alimentary canal, the pharynx, and the mesenteries. In many
Aspidorhirotce the left respiratory tree is associated with the rete mira-
bile of the blood vascular system in the way described on p. 452.
The delicate wall of the organ consists of an inner ciliated epithelium,
a thin layer of connective tissue, a muscle layer (in which an inner
layer of longitudinal fibres and an outer layer of circular fibres can
be more or less distinctly made out), and, finally, of the ciliated
endothelium of the body cavity.

There can be no doubt that the respiratory trees actually function
as respiratory organs. At regular intervals water flows into them
from the cloaca, and is from time to time expelled through the anus,
discoloured by the admixture of fecal masses.

Respiratory trees are wanting in all Paractinopoda (Synaptidce),
the Pelagothuriida'; and the Elasipoda, unless, in the last-named family,
the diverticulum of the rectum described in the section on the
alimentary canal, p. 478, represents a rudimentary lung.

B. Review of the Respiratory Organs of the Eehinodermata.

(<i) Holothurioidea Aetinopoda (excluding Elasipoda and Pclago-
thuriidce).

1. The respiratory trees, which open into the cloaca.

2. The oral tentacles, and to some extent the delicate-walled

ambulacral tentacles as well.

(//) Holothurioidea, Paraetinopoda and Pelagothuriidse.

The whole of the body wall and the oral tentacles. Respiration
is promoted by the circulation of the body fluid, kept up by means of
the ciliated urns.

(c) Echinoidea.

1. The external gills, as outgrowths of the peripharyngeal sinus,

p. 442.

2. The ambulacral feet, especially those on the apical surface of

488 COMPARATIVE ANATOMY CHAP.

the body, and more particularly the branchial tentacles on

the petaloids, rf. p. 433.
3. The accessory intestine, in which, at least in regular Echinoids,

a streaming of water takes place which does not interfere

with the digestive processes going on in the principal

intestine, cf. p. 481.
('/) Asteroidea.

1. The papulae, rf. p. 439.

2. The ambulaeral feet.
(e) Ophiuroidea.

1. The bursse (respiratory and genital chambers).

2. The ambulaeral tentacles.
(/) Crinoidea.

1. The ambulaeral tentacles.

2. The anal tube (anal cone), which alternately takes in and

gives out water.

XVI. The Cuvierian Organs of the Holothurioidea (Fig. 383, 22,

p. 477).

In certain Holothurioidea, peculiar accessor}^ structures are found
connected with the terminal portion of the respiratory trees ; these
are known as the Cuvierian organs. They occur chiefly in the A$pi-
dochirotce (especially in the genera Holothuria and Mulleria) • in other
Holothurioidea they only occur in isolated cases (Molpadia chilensis,
Cucumaria frondosa). The Cuvierian organs are usually very numerous,
even as many as a hundred occurring in some of the species provided
with them. Although, as already mentioned, they are usually found
in the terminal portion of the respiratory trees, they may shift higher
up the principal trunk, and may even pass over on to the branches.

It is not improbable that they represent morphologically trans-
formed branches of these trees, the structure of their walls agreeing
in general plan with that of these latter.

Two kinds of Cuvierian organs are distinguished — (1) glandular,
and (2) non-glandular.

The Cuvierian organs of the glandular kind are long tubes, the
very narrow axial canals of which open into the terminal section of
the respiratory tree. Each of these axial canals has a spiral course,
and is lined by a unilaminar epithelium. This epithelium is followed
externally by a very thick layer of connective tissue, which projects
into the axial canal in the form of a spiral fold. The next layer
consists (1) of isolated circular muscle fibres, and (2) of external
longitudinal muscle fibres gathered into small bundles. Outside of

O O

the muscle layer there is another layer of connective tissue, which,
on the side of the body cavity, is covered by a peculiarly developed
glandular layer ; this no doubt represents the modified endotheliuru
of the body cavity.

viii ECHINODERMATA—SACCULI OF GRINOIDX 489

In this glandulai1 layer the cells can no longer be recognised except by their
nuclei, no boundaries being distinguishable. The layer is closely packed with
secreted granules. Wandering cells and calcareous corpuscles are found in the con-
nective tissue wall.

The animal, when irritated, vehemently ejects its Cuvierian organs through the
cloaca. (The susceptibility to irritation which leads to such ejection varies greatly
in different forms.) In this process the tubes are not turned inside out, but are
thrown out complete, just as they are in the body cavity, probably through a rent
in the cloacal wall. When these tubes are thus thrown out, water is almost certainly
pressed out of the respiratory trees into their axial canals. The discharged Cuvierian
organs are remarkable (1) for their extreme viscidity; (2) for their extraordinary
extensibility. They can be drawn out to more than thirtj' times their ordinary
length. Their viscidity is no doubt produced by the secreted granules of the
glandular layer. In consequence of these peculiarities, the discharged Cuvierian
organs are weapons of defence ; they remain attached to the body of an enemy, and
impede its movements. They may also be weapons of attack, the prey being
caught and held fast till it dies, when its decomposing remains serve for food.

The non-glandular Cuvierian organs are either tubular, like the glandular, or
branched. They are mostly beset with stalked vesicles. The smooth endothelium
of the body cavity which covers them shows no glandular development of any sort.
The part played by these non-glandular and consequently non-viscid organs is
entirely problematical.

XVII. Excretion.

Special excretory organs are altogether wanting throughout the
Echinodermata. It is probable that fluid excrement is osmotieally
given off, together with the carbonic acid, at the respiratory surfaces
of the body. Further, colom-ed and occasionally crystalline corpuscles,
which are met with in very different parts of the body, chiefly in
the connective tissue layers in most Echinodenns, have been regarded
as products of excretion. They appear to remain in the places of
their formation, this conclusion being arrived at from the fact that
they are present in far greater quantities in old than in young animals.
They are also found within the wandering cells, and it might be
worth investigation whether these wandering cells, which force their
way into the body- and the intestinal-epithelium, do not play some
part in excretion.

XVIII. The Saeeuli of the Crinoidea.

These are peculiar organs which, in certain Crinoids, occur in great
numbers below the integument, principally at the edge of the food
grooves of the pinnulie, the arms, and the disc, less frequently else-
where (intestinal wall, mesenteries). They are globular sacs lying
close below the surface, but having no outer aperture, and are closely
packed with strongly refractive spherules, which, during life, are
colourless, but turn red after death. Close examination shows that

490

COMPARATIVE ANATOMY

CHAP.

these spherules are enclosed, at least at first, in cells, each of which
has a nucleus lying in the base, which is turned away from the

surface. These are regarded as
connective tissue cells.

According to recent ontogenetic research,
eacli individual spherule is the metamor-
phosed product of a mesenchyme cell, and
the sacculi on their first appearance are said
to be nests of such cells.

Sacculi are specially numerous in all
species of Antedon, but are also found in
Eittlioerin us, Pronnichocrinus, Pentaerin ?/*,
Khizocrinus, B/itJnjcrhnis. They are want-
ing in Acfinometi'd; Thaumatocrimts, and
Fi.;. 387. -Diagram of a sacculus. i, Sup,',-- H<>t»pus. Their significance has not been
ticial layer of integument passing over the discovered. They have been regarded by
sacculns ; -2, granular masses within special various authors as calcareous glands, ex-
cells ; 5, the nuclei of these cells ; 4, nuclei t of unice]lular algffi

of the surrounding cutis (3). " . , .

and shine glands, but, according to the

most recent opinion, they are proteid corpuscles, deposited in the connective tissue
cells as reserve stuff', to be used as occasion requires, for the regeneration of broken -
off arms or of the viscera.

In other Echinodenns the contents of wandering cells (especially of those cells
which are massed together below the surface of the Holothurian integument) have
also been claimed as reserves of nutrition.

XIX. Genital Organs.

A. General Morphology.

With rare exceptions, which will be dealt with separately, the
sexes are separate in Echinoderms.

The genital organs are throughout distinguished by great sim-
plicity, as evidenced by: —

1. The entire absence of every kind of eopulatory organ.
The sexual products are ejected from the body, and fertilisation takes
place in the water (except in cases of care of the brood to be men-
tioned later).

2. The entire absence of accessory glands, of widenings or
outgrowths of the ducts, and of complicated adaptations for the
nourishment of the ripening sexual products.

The genital organs consist of variously shaped tubes, within which
the spermatozoa or eggs ripen, and from which they are discharged
through simple efferent ducts.

These gonadial tubes lie in any part of the body cavity ; in the
most complicated cases their wall consists, from without inward, of
(1) the endothelium of the body cavity: (2) a thin muscle layer;

vni EGHINODERMATA— GENITAL ORGAXtf 491

(3) a layer of connective tissue ; and (4) the inner germinal epithelium.
The muscle layer is often wanting.

According to the morphology of the genital organs the Echino-
derins fall into two groups.

In the larger principal group, containing the Echinoidea, Aste-
roidea, Ophiuroidea, and Crinoidea, there are several gonads, each with
a duct and an aperture ; they follow, in their arrangement, the radial
structure of the body, showing at the same time close relations to the
axial organ (or to the wall of the axial sinus). The axial organ has
been compared to a trunk, of which the gonads, as direct prolonga-
tions, are the fruitful branches, on which the sexual products ripen,
as fruit.

The direct organic connection between the axial organ and the
gonads persists throughout life in the Asteroidea, Ophiuroidea, and
perhaps also in the Crinoidea : in the Echinoidea it is only ontogenetic,
and ceases in the adult.

The second group is formed by the Holothurioidea, in which there
is neither axial organ nor axial sinus. The genital organs consist of
a single tuft of gonadial tubes. This tuft lies in the body cavity in
the median (dorsal) interradius, and sends off a duct which runs
forward in the dorsal mesentery, and opens outward in the anterior
region of the body, often very near the mouth.

There is, as a rule, no difference in the microscopic structure
and external appearance of the male and female genital organs in
Echinoderms. In some cases, however, at the time of maturity, they
may be distinguished by their different colouring.

Secondary sexual characters have been noticed only in very rare
cases.

B. Holothurioidea (Figs. 371 and 383, pp. 451 and 477).

In all Holothurioidea, the gonads are developed as a single tuft of
genital tubes, lying in the dorsal interradius. All the tubes of the
tuft converge towards one point, and open into the base of the gonad.
which lies in the dorsal mesentery, and is often somewhat widened
for the reception of the sexual products.

The gonadial tubes are either simple or branched ; in number and size they vary
within pretty wide limits according to the species and the stage of maturity attained.
They may exceed the body in length. They usually hang from the base of the
gonad, on the two sides of the mesentery, into the body cavity, but there are cases
in which they are developed only on one side — the left — of the mesentery (in
Jfofothuria, Miillcria, Labidodemas among the Aspidochirotce, and in many
Elpidiidcc}. The base of the gonad lies in the anterior half of the body, often very
near its anterior end (especially in Kynnptida: and MufjH'diidir, but also in many
Aspidochirota: and Dendrochirotn-).

From the base of the gonad the genital duet runs more or less

492 COMPARATIVE ANATOMY CHAP.

far forward in the dorsal mesentery, to pass through the body wall
at some point of the anterior half of the body in the dorsal median
line, and to open outward through the usually single genital aperture.

The distance of this aperture from the extreme anterior end of the body, how-
ever, varies very greatly. It is greatest in the Elasipoda, and becomes smaller in
the Aspidochirotcc. In the Molpadiidcc and Xynaptidce the genital aperture lies
immediately behind the circle of tentacles, and in the DendrocMrotce it shifts into
that circle, even reaching its inner side. It is found behind the middle of the body
only in Psychropotes longicauda.

The genital aperture is usually inconspicuous. Occasionally it is found on the
tip of a genital papilla ; in species of the genera TJiyone and Cucumuria, this is the
case only in males, a slight sexual dimorphism thus arising.

The occurrence of several genital apertures (2, 4, 8, lt> in certain Elasipoda) is
(juite exceptional. They always belong to one and the same gonad, and to one
genital duct. This latter in such cases, before emerging, divides dichotomously into
as many branches as there are apertures.

C. Asteroidea (Fig. 385, p. 484).

The genital organs are here developed as five pairs of bundles of
gonadial tubes, or as five pairs of rows of such bundles. These pro-
ject freely into the body cavity ; their bases are attached to the apical
(dorsal) body wall, generally at the apical edge of the supramarginal
plates, or on a level with this edge. Exactly over the point of
attachment, i.e. over the base of each gonadial tuft, the efferent duct
traverses the body wall (between two neighbouring skeletal plates) to
open outward at the surface through one, less frequently through
several, genital apertures. These apertures are quite small, and are
often only clearly visible at the season of sexual maturity, when the
genital products are ejected.

The bases of all the gonadial bundles are still connected with the
axial organ even in the adult (<•/. p. 445 on the axial organ and the
axial sinus). The axial organ is continued along the inner apical
(dorsal) body wall (that turned to the ccelom) as a pentagonal strand
running- round the apical pole and the anus, which agrees in
structure with the organ itself. At each of the five interradially
placed corners of the ring it sends off a pair of strands which run
peripherally. There are thus in all flve pairs of strands radiating
from the ring ; these run to the bases of the five pairs of gonadial
tufts, and where these are in rows, from tuft to tuft of each row,
connecting their bases.

Just as the axial organ is surrounded by the axial sinus, so are
all its derivatives surrounded by a coelomic sinus, a direct prolonga-
tion of the former.

The aboral ring-like strand lies in a ring-sinus, attached to its
wall by a suspensory band. This sinus is also continued along the
five pairs of strands which radiate from the ring-like strand ; when it

vni ECHINODERMATA— GENITAL ORGANS 493

reaches the bases of the gonads it is further continued along all the
individual tubes to their tips. The gonadial tubes thus have a double
wall — first, their own wall, which is a continuation of that of the out-
growths from the axial organ ; and secondly, an outer wall, which is
a continuation of the wall of the axial sinus. Between these two
walls lies the narrow coelomic sinus, which is in open communication
by means of the sinuses of the genital strands, with the ring sinus,
and through this latter with the axial sinus.

The relations existing between the gonads, the axial organ, and
the system of sinuses, is clearly elucidated by the ontogeny of the
Asteroidea, which shows that in quite young animals the axial organ
grows out apically, and first forms the ring strand. Out of this the
genital strands bud, and from these latter again the bundles of
gonadial tubes arise, which are at first solid outgrowths, and only
become hollow secondarily. During this whole process the growing
axial strand, which finally produces the rudiments of the gonads,
continually carries the axial sinus along with it, so that the ring-like
strand, the genital strand, and the genital tubes are entirely sur-
rounded by a sinus, which constantly remains in open communication
with the axial sinus.

At those points of the genital strands from which the gonadial
bundles bud, i.e. at the future bases of the gonads, the duct which
perforates the body wall is formed from within at a later stage.

The form of the individual gonadial bundles requires little notice. The genital
tubes of which each bundle is composed are usually not long, sometimes they
resemble short sacs and are vesicular, they are rarely branched.

Of much greater interest are the number and arrangement of these bundles.

In the simplest cases, five pairs of gonadial bundles are present ; this is the case,
as far as examination of the various species on this point has taken place, in the
following families : the Aster inidcc, Sol aster idee, EcMnasteridoe, Linckiida:, Astcri-
id«\ In these, the five pairs either lie in the disc, one bundle at each side of each
interradius, or have shifted into the bases of the arms (Echinastcrida; Liiickiidir}.
More than five pairs of gonads are found in the families of the Astropcctinida',
PI iitticcmtida', and Gyituwufcriidn-. They either lie in the disc in rows at the sides
of the interradii, or the five pairs of rows stretch into the arms. This last arrange-
ment is found in the most extreme form in Luliliu, where, on each side of each arm,
a row of nine runs to near its tip, one or two pairs occurring on each skeletal
segment.

In all cases, each bundle has its separate genital aperture.

As a rule, each bundle has only one aperture, but it sometimes happens (Asterias,
Solaster) that the duct which traverses the apical body wall branches, and opens
through several genital pores lying near one another.

Aster ina gibbona is an exception to the rule that the genital apertures lie on the
apical side of the disc or arms. The apertures here lie on the oral side, a peculiarity
connected with the fact that these Asteroids do not simply eject their eggs into the
water, but attach them in combs or plates to stones, etc.

It must, finally, be noted that the aboral ring sinus is not always simple, but
may break up into a circular network of sinuses (e.ij. Echinastcr scpositus}.

494

COMPARA TIVE ANA TOMY

CHAP.

D. Ophiuroidea.

In structure and development the genital organs in this class
strongly resemble those of the A&teroidea. The gonad is connected
with the axial organ by means of an aboral ring-like strand, and both
the gonads and this strand are surrounded by ccelomic sinuses, which
communicate with the axial sinus.

The only important difference in the genital organs of the two
classes is caused by the fact that, in the Ophiuroidea, the gonads do not
open outward directly, but by means of five pairs of large sac-like

FIG. 388.— Stomach and bursae of a young Ophioglypha albida, in its natural position in
tlit- disc, the dorsal wall of which is removed. 1, Bursie ; 2, cavity of the disc ; 3, iuterradial ; 4.
radial bulgings of the digestive sac (after Ludwig).

invaginations of the body wall into the coelom of the disc, these
sacs themselves communicating with the exterior through five pairs
of slit-like apertures lying at the sides of the bases of the arms on the
lower (oral) side of the disc. These sac-like invaginations of the
body wall are the bursse or bursal pockets, their outer slit-like
apertures being known as the bursal (genital) apertures, which have
already been mentioned (Figs. 245, 246, and 314, pp. 300, 301,
and 359).

1. The Bursse (Figs. 388 and 389).

These are large sacs within walls, which fill up the body cavity of
the disc round the digestive sac. Their walls are attached to that of

VIII

ECHINODERMA TA— GENITAL ORGANS

495

the digestive sac and the apical body wall by means of strands of
connective tissue, and consist of the same layers as the body wall,
though in the bursse these layers are much thinner. Calcareous
corpuscles may be either present or wanting in the connective tissue.
The inner epithelium of the bursse is in some parts strongly ciliated.

The outer apertures of the bursae lie at the sides of the proximal
portions of the arms, which are included in the disc. Each bursa has,
as a rule, one aperture, but in 2

the genus Ophiura (formerly
Ophioderimi) there are two aper-
tures on each side of the base of
an arm, one distal and the other
proximal. Both these apertures,
however, lead into one and the
same bursa, and the double aper-
ture (in Ophiura) can be deduced
from the ordinary single aperture
by assuming that the margins of
the latter fuse at about the
middle of their length.

The gonads are attached to
the wall of the bursa, on the
side turned to the body cavity

(Figs. 391 and 392).

FIG. 380. —Part of a preparation of Ophio-
The glypha similar to that in Fig. 388, after removal

sexual products pass into the of the stomach and the gonads (after Ludwlg).
\ Ofthe two bursee, that on the left has been removed,

bursa, and are ejected thence ], Dorsal shields of the arm; 2, dorsal wall of the

through the aperture. This is, disc: 3- bursa with its tjp (4>; 5< penstome;

• i £ ,1 f 6. vertebral ossicle in the base of the arm ; 7, genital

however, Only One of the fimC- plate ; 8, row of buwal plates or scales.

tions of the bursa, and, in most

Ophiuroidea, as it appears, not the principal function.

The bursse have a more important function as respiratory organs.
Sea-water can enter them, and through their thin walls exchange of
gases can take place between it and the body fluid. It would be in-
teresting if it could be proved that, as in the mouth and oesophagus
of the Corals, the sea-water enters through one (more or less proximal)
part of the bursal aperture, and flows out again through another (more
distal) part. The proximal aperture of each bursa in Ophiura is per-
haps an inhalent, and the distal an exhalent aperture.

In certain Ophiuroidea (e.g. Ainpkium sguainata, magellanica, Ophi/i-
cantha vivipara, marsupialis, Ophiogh/pha hexactis, Opkiomt/xa vivipara, etc.),
the bursre serve as brood chambers. The eggs pass through their
whole development in them, until all the organs of the young Ophiurid
are formed.

2. The Genital Apparatus (Figs. 390-393).

The most interesting point in connection with the genital apparatus
is the peculiar course of the apical ring sinus with the ring-like strand

496

COMPARATIVE ANATOMY

CHAP.

it contains. Fig. 390, which represents the ring sinus in horizontal pro-
jection, illustrates its course in five outwardly directed radial and five

FIG. 390.— Course of the aboral circular sinus, with the ring-like strand contained in it in
the Ophiuroidea (diagram after Ludwig). l.Gonails; 2, axial sinus with axial organ ; 3, mouth;
4, circular sinus with ring-like strand, on the side of the bursal wall turned to the interradius ; 5,
interradial portion of the ring sinus and strand, bent downwards orally (Fig. 3S6, left oar) ; 6, bursal
aperture ; 7, radial (apical) region of the ring sinus (Fig. 3S6, right uar); 8, lateral branches of
the same on the bursal wall turned to the radius.

inwardly, if. orally directed interradial curves. In this undulating-
course the ring sinus descends on the inner wall of the disc alternately

from the apical to the oral side,
and then again ascends to the
apical side, the radial curves lying
apically and the interradial (those
near the bursse) orally.

This peculiar course is no doubt con-
nected (1) with the orally directed course
of the axial sinus, the axial organ, and

the stone canal which opens outward
FIG. 391.— Bursa of Ophioglypha, seen from ., ,. ,

the side turned towards the intei-radius (diagram Orally through its madreporite (Fig.
after Ludwig). 1, The tip of the bursa, lying on 361> 6> P- 422). For the ring sinus
the dorsal side of the digestive sac ; 2, the gonads is the continuation of the axial sinus,

sessile on the bursal wall ; 3, distal portion of a
bursa (that turned to the periphery of the disc) ;
5, proximal portion of the same (that turned to
the centre of the disc) ; 4, the rows of bursal
scales along the aperture.

and the ring-like strand is the continua-
tion of the axial strand. It is now im-
possible to determine whether the axial
sinus and the axial organ, in bending
orally, drew the ring sinus interradially

in the oral direction (in the first place this could of course only apply to the madre-
poritic interradius), or whether, on the contrary, the ring sinus, shifting orally, drew

VIII

ECHINODERMA T A— GENITAL ORGANS

49:

the axial sinus, etc. with it downwards ; i.e., it is impossible to decide which organ
took the lead in shifting. (2) As the gonads which bud from the ring-like strand
open into the bursa?, which latter, however, open outward orally, it is to some
extent explicable why the ring-
like strand descends interradi-
ally to the bursre.

The whole problem is still
further complicated by the ques-
tions : (1) What was the original
function of the bursse ? (2) Is
the ventral position of the burste
the primitive position ? (3) Is
the opening of the gonads into
the bursre a recent specialisation
in the Ophiuroidea ?

The curved-in portion
of the ring - like strand
(with the sinus enclosing
it) runs along that side of
each bursa which is turned
to the interradius. It,
however, gives off a branch to the wall which is turned towards
the radius (of the arm), this branch running along this wall from
its periphery to its proximal part. Both walls of the bursa, therefore,

Fn;. 3<.t-2.— Transverse section through the disc of an
Ophiurid (Ophioglypha) at the base of an aria (after
Ludwig). 1, Dorsal wall of the disc ; 2, bulging of the
digestive sac ; 3, bursa ; 4, gonad on the bursal wall ; 5, base
of the arm ; (5, ventral wall of the disc ; V, bursal aperture ;
S, genital plate ; 9, bursal scale.

FIG. 393.— Section through an ovary of an Ophiurid (Ophioglypha lacertosa) (after Cue'not).

1, Muscle trunk, cut through transversely ; 2, nerve ring ; 3, bursal wall ; 4, aperture of the ovary
into the bursa ; 5, wall of the genital sinus ; 6, genital sinus ; 7, the endothelium of the genital
sinus, which covers the gonadial wall ; 8, cavity of the gonad ; 9, eggs in a more mature condition
than the rest ; 10, ring-like strand in the aboral ring sinus (11).

the abradial wall, i.e. that turned to the interradius, and the adradial
wall, i.e. that turned to the arm, have a genital strand. The abradial
genital strand of each bursa is merely a part of the apical ring
strand, while the adradial is a lateral branch of that strand. These

VOL. II 2 K

498 COMPARATIVE ANATOMY CHAP.

five pairs of adradial genital strands recall the five pairs of genital
strands of the Asteroidea.

The gonadial tubes are sessile upon the genital strands of the
bursa?, and project freely into the body cavity of the disc. These
gonadial tubes are either single pear-shaped tubes, great numbers of
which are arranged in rows along the genital strands, or they are
collected into tufts, and then there is one tuft on the adradial and one
on the abradial wall of the bursa.

In the former case (e.g. Opliioglyplta, Ophiomyxa, Ophiocoma) the two rows of
genital tubes (the adradial and the abradial) stand rather low down on the wall of
the bursa, near its aperture, almost parallel with the edges of the latter. Each
genital tube has its special aperture into the bursa.

In the second case, the points of insertion of the two tufts of gonads within the
ventral region of the bursal wall, i.e. the position of the bases of the gonads,
seems to vary greatly, and each tuft appears to have only one aperture into the
bursa (Opliiopholis. Ophiothrix).

It is still an open question whether the genital apertures are constant in adult
Ophiurids, or whether they only break through into the bursal cavity at the time of
maturity.

The gonadial tufts arise as originally solid buds from the genital strands, and,
while forming, bulge out the wall of the sinus which contains the strand ; the tubes
are thus here also surrounded by a genital sinus, which communicates with the ring
sinus, and through it with the axial sinus (Fig. 393).

The ring-like strand is attached by a thick band to the wall of the ring sinus.
It is solid, and consists of two kinds of cells: (1) cells which entirely resemble
those of the axial organ, of which the ring-like strand is a prolongation ; (2)
enclosed in these cells, there is a strand of cells proved to be genital germ cells
(rachis genitalis). The cells of the former kind progressively decrease in number,
and those of the second kind increase in number the nearer the ring-like strand
approaches the gonads. The former are not even continued into the gonadial
tubes, while the latter kind yield the germinal cell material of the gonads. It is
very probable that, after the sexual products have been ejected, a new formation
of germinal cell material takes place, by some kind of forward movement, from the
rachis genitalis.

The development of the genital system from the axial organ and the axial sinus
proceeds in the same manner as in the Asteroidea.

OphiacMs virens, a form distinguished by reproduction by means of
fission, and by the peculiar arrangement of the appendages of the
water vascular system, is the only Ophiurid in which the burste are
altogether wanting. The gonads open direct outward on the oral side.

E. Echinoidea (Figs. 358 and 370, pp. 419 and 443).

Although the genital system of the Echinoidea appears to resemble
in its development that of the Asteroidea and the Ophiuroidea (a point
on which, however, further research is desirable), marked deviation
takes place in the adults.

The gonads, at least in regular Echinoids, are five in number, and

VIII

ECHINODERMA T A— GENITAL ORGANS

499

lie in the apical region of the body cavity, in the interambulacra. The
five genital ducts ascend towards the apex, there perforate a coelomic
circular sinus which surrounds the rectum, pass through the genital
pores of the basals, and then open outward, sometimes at the tips of
projecting papillae.

The gonads. — These, in a mature condition, are large acinose
organs, which are suspended to the inner wall of the test by an
exactly interradial principal
suspensor, and by various
other bands of connective
tissue. They are not sur-
rounded by a genital sinus.

The number of gonads
was originally five. Five
are found in all the regular
Echinoids (Cidaroida and
Diadematoida), and also in
many Clypeastroida. In the
Spatangoida, the Holedi/poida,
and many Clypeastroida, the
number is reduced, the pos-

1 . FIG. 304.— Cystechmus vesica A. Ag. Apical portion

tenor Unpaired gonad With of the test from within, with the three gonads. 1, Au-

the

genital

pore
-, i • ii_ £ 1-4, ri''lit posterior gonad;

to it bein the first to dis-

belonin terior ambulacrum ; 2, left anterior ; 3, left posterior ;

circular sinus (after A.

Agassiz).

appear. In the Spatangoida,

the reduction may go still further, the right anterior, and in a few
-cases the left anterior as well, disappearing (Fig. 394).

Further details on this point are to be found in the section on the skeletal system
•(cf. pp. 321-324 on the genital pores). It is there shown that these pores are by no
means necessarily limited to the basals.

The genital apertures. — The genital papilla?, on the tips of which the genital
apertures lie, are specially well developed in the Spatangoida.

The ring sinus encircles the anus with the periproctal sinuses, the stone canal,
and the axial sinus. Its wall is formed on the one side by the test, and on the
•other by a circular lamella of connective tissue which is covered on both surfaces by
endothelium, on the apical surface by that of the ring sinus, and on the oral by that
•of the general body cavity.

The lower wall of the apical ring sinus is broken through in Dorocidaris, so
that the circular sinus is here in open communication with the general body
•cavity.

In all other cases, the ring sinus is entirely closed on all sides in adult
.Echinoids.

In adults, there is no trace of a ring-like strand enclosed in the

ring sinus,
is thus lost.

The connection between the axial

organ and the gonads

500

COMPARATIVE ANATOMY

CHAP.

F. Crinoidea (Fig. 395).

In the Crinoids, a genital strand runs through the arms, branching
with them, and entering into their last ramifications — the pinnulse.
While this genital strand, which is to be found even below the food
grooves of the tegraen calycis, remains as a rule infertile in the calyx

and in the arms, in the
pinnules the germinal
cells which it contains
give rise to the genital
cells. The genital strand
in the pinnulse becomes
a gonadial tube.

On the position of the
genital strand cf. p. 414
and Fig. 356. It runs
between the three brachial
sinuses of the body cavity
(between the dorsal canal

and the two ventral
canals), below the food
grooves of the arms.

It is contained in a
special ecelomie sinus
ovanai (like the ring-like strand
and the genital strands
of the Astcroidea and the
Ophiuroidea), to the wall
of which it is attached by
filaments of connective
tissue.

The coslomic sinus is continued on to the gonadial tubes of the
pinnulse and there becomes the genital sinus.

The genital strand is at first solid, but at a later stage becomes a
hollow genital tube. This genital tube widens in the pinnules into
the gonadial tube, which, in mature pinnules, is filled either with eggs
or spermatozoa, these having their origin in the cells of the wall of the
gonadial tubes (the germinal epithelium).

In a transverse section of the genital tubes of the arms, the wall
appears thickened at one part. This thickening is the section of a
ridge whose cells seem to correspond with those of the germinal epi-
thelium of the gonadial tubes.

It is very probable also, that after the ejection of the sexual pro-
ducts from the pinnules in Crinoids, the new formations of these pro-
ducts proceed from the germinal cells, which are pushed forward out
of the ridge of the genital tube into the pinnules.

FIG. 3i>5. —Transverse section through an
pinnule of a Crinoid, diagrammatic. 1, nerve trunk of the
apical nervous system in the joint of the pinnula ; 2, genital
sinus ; 3, germinal epithelium of the gonadial rachis (genital
strand or tube) ; 4 and 0, sinuses of the brachial coelom ;
5, deeper longitudinal nerves of the pinnula ; 6, tentacle canal ;
7, radial canal of the water vascular system ; S, nerve ridge of
the superficial oral system ; 10, sacculus, see p. 490.

via ECHINODERMATA— GENITAL ORGANS 501

That the cells of the genital ridge (and, indeed, originally all the cells of the
genital strand) are germinal cells is further proved by the fact that, in exceptional
cases, gonads may develop in the arms also, and even under the ambulacral furrows
of the calyx (e.g. in individuals of the species Antedon and Actinomctm, and in one
species not specified).

The gonadial tubes are sometimes long, sometimes egg-shaped. They run
through a larger or smaller number of joints of the pinnule. At the time of
maturity they swell and often bulge out the wall of the pinnule in such a way as
to show at a glance which pinnules contain ripe sexual products.

The manner in which the ripe products are ejected from the pinnules is not yet
satisfactorily explained. There seem to be no constant genital apertures in the
adult. It appears that the ejection takes places through two merely temporary
apertures (one on each lateral wall of the pinnule).

Round the mouth, finally, there are five genital strands with the
sinuses in which they lie, running from the periphery, i.e. from the
bases of the arms below the food grooves of the tegmen calycis. It is
not certainly known what becomes of these genital strands ; according
to some accounts, they are continued round the mouth into the strands
of the axial organ. They are said also to develop ontogenetically as
outgrowths of that organ (<•/. p. 446).

If the axial organ of the Crinoids is homologous with that of the Ophiuroids,
Asteroids, and Echinoids (which homology cannot be considered as certainly estab-
lished), then we should have the same relations subsisting between the axial organ
and the genital organ in the Crinoids as in the other groups above mentioned. But
in the Crinoids the genital strands, which only become fruitful as gonadial tubes in the
pinnule, are oral outgrowths of the axial organ, whereas in other Echinoderms (apart
from the Holotlmrioidca, which are quite isolated) they are apical outgrowths.

G. Origin of the Sexual Products.

The first origin of the sexual products has been accurately described
for the Ophiurid Amphiuni squamata. They, and the cells of the axial
organ, arise out of one and the same rudiment, which consists of the
endothelial cells of the eoalom. The Echinoderms would thus, as
far as the origin of the sexual products is concerned, agree with the
Annulata, the Mollusca, and the Vertebrata.

The specific cells of the axial organ seem incapable of becoming

germinal cells.

H. Hermaphroditism in Eehinoderms.

Hermaphroditism is an altogether exceptional phenomenon in
Echinoderms, and is only of frequent occurrence in one order of the
Holothurians, the Paradinopoda (Synaptidce). Apart from this order,
it is only certainly established in one Asteroid (Asterina gibbosa) and

one Ophiurid (Ampliiura squamata).

(«•) Paractinopoda. — All species of the genera Xynapta and Anapta, examined
with regard to this point, and a few species of the genus Chirodota, are herma-
phrodite.

502 COMPARATIVE ANATOMY CHAI-.

Eggs as well as spermatozoa are produced in the gonadial tubes, but the two
products do not ripen simultaneously (Synajrta inhccrcns). The spermatozoa only
ripen after the ejection of the eggs.

(&) Asterina gibbosa. — Here also the eggs and the spermatozoa are formed in the
same genital organs, again not being simultaneously produced. The young animals
are males, the adults females.

(f) Amphiura squamata. — The simple pear-shaped gonads are very few in number.
On the average, the adradial and the abradial walls of a bursa have only one gonad
each sessile on it. The adradial gonads are testes, the abradial ovaries. Only a few
eggs in the ovary and a small number of spermatozoa in the testes ripen at one time.
These two kinds of sexual products here also, as it appears, do not ripen simul-
taneously in one and the same animal. The eggs are developed in the bursre.

I. Care of the Brood and Sexual Dimorphism.

Little by little, somewhat numerous cases of care of the brood have become known
among the Holotlmrioidea, Echinoidcu, Asteroidca, and Ophiuroidea. These are not
infrequently connected with a more or less pronounced sexual dimorphism, the
adaptations for care of the brood being found only in the female.

The eggs of an Echinoderm in which the brood is cared for are, as far as investiga-
tion on this subject has gone, distinguished by remarkable size, and by a rich pro-
vision of nutritive yolk, from those which are ejected into the water, and are destined
to develop into free-swimming larva?.

The following review makes no claim to be exhaustive.

(o) Holothurioidea. — In Psolus ephippifer (cf. Fig. 228, p. 2S7) large plates are
found on the back of the female, raised from the dorsal integument by means of
stalks. Between the stalks a brood chamber, roofed over by the contiguous plates,
arises ; in this the fertilised eggs which emerge through the dorsal genital aperture
pass through their development.

In Cucumaria crocea, the developing eggs are retained in a dorsal trough, which
arises by the swelling up and bulging outward of the body wall in the two dorsal
radii.

Another kind of care of the brood is found in Cucumaria. Icevigata and C. minute.
Two sacs here project froin the body wall into the body cavity ; these are brood
pouches, which shelter the developing brood. The sacs are probably mere imagina-
tions of the body wall ; their outer apertures, however, have been discovered only in
G. miniita. The two sacs belong to the two ventral interradial areas ; in C. In ri/jatu
they lie near the middle of the body, in C. mi /tufa anteriorly.

In PhyUophonis urna and Chirodota rotifcra, the body cavity serves as a brood
chamber. It is, however, unknown how the fertilised eggs pass in nnd the young
Holothurioidea out of it.

In other Echinoderms, as might be anticipated, we find the spines occasionally
acting as protections for the brood.

(b) Echinoidea. — In a few Cidaroida (e.g. C. canaJiculata, C. //////•/,/•. C. mcm-
branipora) the eggs are retained on the apical area of the test, and here develop,
protected by the spines, which bend together over them. The same is the case in
many Spatangoida, but the members of this order have become still more specialised
for this function. In certain forms either some or all of the petaloids (cf. p. 347)
sink in deeply, and thus give rise to brood chambers (marsupia) into which the eggs
pass from the genital aperture. The brood developing in such a marsupiuni is pro-
tected by the bending together of the larger spines which border it. In the Schizastcr
figured on p. 294, the anterior unpaired petaloid ; in Hemiastcr cavcrnosus, in which
this arrangement is best known, the paired petaloids are the most deeply sunk. As

VIII

ECHINODERMATA—CARE OF BROOD

503

this is only the case in the female, \ve here have a striking sexual dimorphism.
Similar adaptations for the care of the brood seem to occur in Moira, Anochanus,
etc.

(c) Asteroidea. — Among the Asteroidea, the Pterasterina: (Pteraster, Hymenaster)
are very remarkable for the care of the brood. The whole of the apical body wall
carries large peculiarly shaped paxillse or calcareous pillars, from the free ends of
which radiate, like the spokes of a wheel, a varying number of calcareous rods (<•/'.
p. 391). All these calcareous stars of the paxillfe are connected by an integument
in such a way that, between this in-
tegument (supradorsal membrane) and
the dorsal wall of the body beneath it,
a large brood chamber is formed. This
chamber communicates with the ex-
terior at many points : (1) through a
large aperture at the apical pole
(osculum) usually surrounded by five
valves of considerable size (Fig. 396) ;
(2) through numerous contractile pores
or spiracles in the membrane which
covers the brood cavity ; (3) through
regular segmentally recurring apertures
at the sides of the arms. These aper-
tures can be closed by means of small
spines or scales. These "segmental"
apertures are regarded by the present
writer as ventilating apertures, as they
appear to serve the purpose of keeping
up an active circulation of water in the
brood cavity.

The sexual arrangements in the

Ptcrasterince. are unfortunately still un- surrounded by five valves (after Sladen).
known. All specimens as yet desciibed

show the brood membrane. Possibly they are all females, and the males are still
unknown, or the Pterasterina' may be hermaphrodite. Or, again, there may be a
far-reaching dimorphism, which has led to the males being described as a separate
species.

Leptoptychaster kerguelenensis, an Astropectinid, shows us the care of the brood,
seen in the Pterasterina:, to a certain extent in statu nascendi. The eggs which
emerge from the genital aperture pass into the interstices between the stalks of the
still separate paxillte, and there pass through the first stages of their development.
At a later stage, also, as young Asteroids, they remain for some time on the body of
the mother.

In Asterias spirabilis, similar arrangements are found, but the embryo is con-
nected by means of a stalk to the body wall of the mother.

Other Asteroids (e.g. species of Echinaster and Asterias} protect the brood which
collects on the oral side ; it develops under the shelter of the arms, which simply
bend round over it, so forming a temporary brood chamber.

Ophiuroidea. — In the description of the bursse, p. 495, it was mentioned that, in
many Ophiurids, these function as brood chambers, and the best-known cases were
Driven.

FIG. 3!iii.— Hymenaster pellucidus, Wyv. Thom-
son, from the apical side. The osculum is seen,

504 COMPARATIVE ANATOMY CHAP.

XX. Capacity for Regeneration and Asexual Reproduction by means of

Fission and Gemmation.

The capacity for regeneration is, as a rule, highly developed in Echinoderms.
Defects in the body wall are in this way easily and quickly repaired in all Echino-
derms. The Echinoidca, even, in which this capacity is in other ways slightly
developed, easily repair smaller or greater defects in the body epithelium which covers
the test. For example, in Dorocidaris papillata, portions of the test over which
the epithelium has been damaged or destroyed are cast off, and as soon as a fresh
integument has formed a new test can undoubtedly be produced below it.

The capacity for regeneration may increase to an extraordinary degree within the
different divisions, and the sensitiveness to external stimulus increases in proportion,
till a stage is reached when voluntary amputation by means of muscular contraction
takes place in response to external stimuli.

The Crinoids completely regenerate lost viscera, and it even appears as if such
loss is not altogether involuntary, in certain species and under certain conditions
voluntary amputation taking place. This, however, is not certain.

Crinoids easily regenerate brokeu-off portions of arms or whole arms ; several or
indeed all the arms may, under favourable circumstances, be regenerated. The arms
break off easily at their bases ; it even appears as if Antedon, in response to injurious
stimuli, voluntarily throws off its arms.

The regeneration of the portions of arms (bitten off, possibly by enemies) or of
whole arms takes place very easily in many Asteroidea and Ophiuroidca. The
frequency with which Asteroids and Ophiurids with regenerated arms or arm tips are
met with demonstrates both the frequency of mutilation and the great utility of
regeneration.

Species of Asteroids in which the disc with the other arms are regenerated at the
base of broken-off arms are less common. Such regenerations give rise to the well-
known "comet" form of Asteroids (Fig. 397, B). Regeneration of the whole body
from one arm never occurs in Ophiurids. It has been suggested that the difference
between Asteroids and Ophiurids in this respect is connected with the fact that, in
Asteroids, intestinal diverticula project into the arms, and that the genital products
are often developed in them, which is never the case in Ophiurids.

Animals in which half the disc is retained can regenerate the rest of the body
both among the Ophiuroidea and the Asteroidea.

Defects both great and small in the disc are repaired.

In Linckift inuJtifora, an Asteroid distinguished by an extraordinary capacitv for
regeneration, cases have been known in which, after the animal has lost the greater
part of an arm, two new tips have been formed by the wounded surface, and in one
case regeneration led to the formation of a complete new Asteroid at such a
point. This latter case is illustrated in outline in Fig. 397 C. The new animal
consists of two discs with their arms, connected by the regenerating stump of the
arm.

Holothurioidea. — Here also the capacity for generation seems to be very great.
Not only are tentacles and integumental defects repaired, but the ejected viscera
(intestine, respiratory trees, and even the calcareous ring, the water vascular ring,
and the gonads) can be regenerated. In Synapta, after the body has been completely
cut to pieces, its anterior portion can regenerate the whole. In a Cucumaria, the
two separate halves can grow into complete animals.

Increase in the capacity for regeneration is accompanied by increased irritability.
Many HoJothurioidea, especially Aspidochir<it« , when strongly stimulated, contract

VIII

EOHINODERMA TA—REGENERA TION, ETC.

505

so violently that the intestine is torn out (usually behind the calcareous ring), and
together with the right respiratory tree is ejected through a rent in the cloacal
wall.

In certain Holotkurioidca, the integument, when strongly irritated, easily dis-
solves into slime. A Stichopus was observed to come entirely out of its skin, i.e.
the whole integument dissolved into slime, so that only the dermomuscular tube

FIG. 397.— A, Ophidiaster diplax. A specimen with the arms (3, 4, 5) in the act of being re-
generated ; and two (1 and 2) being constricted off (after Haeckel). B, Linckia (Ophidiaster) multi-
fora, a "comet" form. One arm is in the act of regenerating the disc and the other four arms
(after Haeckel). C, The case given in the text of a specimen of Linckia multifora (after P. and F.
Sarasin).

enclosing the viscera remained. That regeneration follows such a phenomenon has
not yet been established by observation.

The Synaptidcv react on stimuli (often quite slight) by falling to pieces, the
circular musculature being at certain points so strongly contracted that the parts
thus constricted break off.

It will no doubt be proved in time that all these manifestations of increased
irritability which coincide with increased capacity for regeneration are of use to the
animal.

Asexual reproduction (schizogony). In certain Echinodenns, the strongly
developed capacity for regeneration has had as a consequence asexual reproduction.
It is, indeed, not certain that, in the cases to be quoted below, the division into

506 COMPARATIVE ANATOMY CHAP.

parts is purely voluntary (i.e. results from causes entirely within the animal itself)
and not to some extent due to external stimuli, however slight. In any case, the
final result of the regeneration which follows is the same — the multiplication of
individuals.

Fission of the body into two halves of approximately equal size with subsequent
regeneration has been observed in Ophiuroidea, Asteroidca, and Holothurioidcu.
In the two former classes the plane of fission passes through the middle of the disc
(through the mouth and digestive sac), in the Holothurioidea (Cucumaria') it passes
transversely through the tubular body, dividing it into an anterior (oral) and a
posterior (apical) half.

In the Ophiuroidea, reproduction by means of fission has been observed in the
following genera : Ophiactis (Mulleri, Sarigiiy. vircns), Ophiocnida (scxradia), Ophio-
coma (pumila, Valencia}}, Ophiotlteln (isidicoJa, dividua).

Among the Astcroidea, schizogony is specially characteristic of many species of the
genus Asterias (acutispina, atlantica, calamaria, microdiscus, tenuispina], and is also
found in Aster ina Wega, CribreUn scxradiata, SticluMcr aJbulus.

Another kind of asexual reproduction seems to be very common in the family of
the Linckiidce, In these Asteroids, the arms become constricted off at their bases,
after which not only does the disc regenerate the arms which have been cast off, but
each individual arm regenerates the disc and the other arms ("comet " forms of
Asteroids, Fig. 397 A, B).

Asexual reproduction does not, as a rule, appear to take place simultaneously with
sexual reproduction, but there are exceptions to this rule.

XXI. Ontogeny.

In all Echinoderms, except those few forms in which care of the brood occurs, the
fertilised eggs develop into free-swimming, bilaterally symmetrical larvae, which are
transformed into the radially built Echinoderm after passing through an often very
complicated metamorphosis.

The larvse of the different classes of Echinoderms will first be compared ex-
clusively according to their external characteristics.

A. The various Larval Forms of the Echinodermata.

We shall first construct a hypothetical larval form, and then deduce the various
larval forms from it (Fig. 398, A).

The body of the larva is egg-shaped, and concave on the ventral side. In the
base of the concavity lies the larval mouth. Near one of the poles (i.e. near the
posterior end), but still on the ventral side, there is a second aperture (proceeding
from the blastopore of the gastrula larva) ; this is the larval anus. A ciliated band
which runs back upon itself surrounds the mouth along the edge of the ventral con-
cavity ; posteriorly it runs over the ventral side in front of the anus, and is the
circumoral ciliated ring. The aperture of the mouth and its immediate surround-
ing are also ciliated (adoral ciliated band).

1. Holothurioidea. — The Holothurid larva known as Auricularia (Fig. 398, A)
differs but little from the hypothetical form. The ventral concavity becomes more
complicated, lengthening on each side posteriorly and anteriorly, while a posterior
median portion with the anus forms a ventral prominence, the anal area, and a
median portion in front of the mouth forms another prominence, the preoral area.
The ciliated band which runs longitudinally along the ventral depression assumes

VIII

ECHINODERMATA— ONTOGENY

so:

in consequence a more complicated form, and takes a winding course. This descrip-
tion will be elucidated by the figures.

Here, as in all other Echinoderms, the ciliated rings are mere remains of the
cilia which covered the whole body in an earlier stage, i.e. in the gastrula.

2. Asteroidea (Fig. 398, B). — The Asteroid larvae are known as Bipinnarise and
Brachiolarise. The chief distinction between them and the Auricularia is the
preoral ciliated ring. This is a circle on the preoral area, and tcithtu the circumoral
ciliated ring, from which it is altogether distinct.

The comparison of a Bipinnaria with an Auricularia led to the conjecture that
the preoral ciliated ring of the former corresponds with a preoral portion of the
common circumoral ciliated ring of the latter, which has become distinct and has
closed to form a ring. Direct observation of the ontogenetic development of the
ciliated ring of the Asteroid larva has entirely confirmed this conjecture.

B

FIG. 398.— A, B, C, Auricularia, Bipinnaria, and Tornaria (Enteropneustan larva), from the
right side, diagrammatic. 1, Pretroclial area; 2, oral area; '.',, postural area; 4, anal area; I, pre-
oral ; II, circumoral : III, anal or principal ciliated ring ; 5, neural plate ; <w, month, an, anus.

The Bipinnaria passes through an Auricularia stage. The general ciliatiou of
the body, belonging to an early stage, disappears first from the ventral side, which
becomes depressed, then from the dorsal side, in such a way as to leave a band
running back on itself at the edge of the ventral depression ; this corresponds entirely
with the course of the circumoral ciliated band in the Auricularia. In the frontal
region, however, where the two lateral strips of the circumoral band approach each
other in the median line, a ciliated island is for a time retained connecting them
(Asterias rubens). The covering of cilia thus forms an X-like cross on the frontal
region. By the disappearance of the ciliation from the centre of the X, the preoral
section of the ciliated ring is separated from the rest, and forms the distinct preoral
ring enclosed within the circumoral ring.

The process in Asterias vulgaris seems to take a somewhat different course, but
has the same final result. On the frontal region, where, in A. rubens, an isolated
ciliated area remained to form a connection between two portions of the circumoral
ciliated band, this connection arises only secondarily by the approximation of the
two portions in the middle line. The further process of separation of the preoral
ring from the rest, which latter then represents the secondary circumoral ciliated
ring, agrees with that in A. rubciix.

The ventral depression (in which the mouth lies) which, in the Auricularia, runs
forward to the right and left of the preoral portion of the circumoral ciliated ring,
is now able, after the latter has become constricted off as a ring, entirely to surround

.508

COMPARATIVE ANATOMY

CHAP.

that portion ; it forms a moat round the preoral area, which becomes raised up like a
shield.

An adoral ciliated ring, closely encircling the mouth and extending some way
into the buecal cavity, is also present.

The body is produced into longer or shorter processes or arms, in the regions of the
preoral and circumoral ciliated rings. An anterior unpaired frontal process, belong-
ing to the ciliated ring, is distinguished by its constant occurrence and its greater
length.

In some species, the ciliated band disappears on this frontal process, which, on

20-

FIG. 399.— Older Auricularia, seen diagonally from the lower and left side (after Semon). 1,
Circumoral ciliated ring ; 2, liydropore ; 3, liydrocoel ; 4, adoral ciliated ring ; 5, median or stomach
intestine ; i>, nerve band ; 7, hind-gut ; 8, left enteroccel ; 9, calcareous wheel ; 10, fore-gut, oeso-
phagus ; 11, right enteroccel.

the other hand, divides into three branches, apparently covered with protuberances
at their tips. Such larvas are known as Brachiolarise.

There are, further, Asteroids whose larvae do not at all resemble the Bipinnnrian
and Bracliiolarian larva?, or else show only a superficial resemblance to them ; if.
below the account of the larva of Astcrina yibbosa (p. 525).

3. Ophiuroidea. — The Ophiurid larva is known as the Pluteus, and can he
just as easily deduced from the hypothetical larval form of the Echinoderms, sketched
above, as the Auricularia and the Biplnnana. The gastrula stage is followed by
the bilateral stage with depressed ventral surface, in the centre of which lies the
larval mouth. A circumoral ciliated band is retained, running along the edge of this
ventral depression. This band always remains single. While the preoral area (the
larva being viewed from the ventral side) remains very small or is even indistinguish-
able, the anal area appears very large. The body is produced into processes or arms,
which may become very long, and are supported by calcareous rods. These pro-

VIII

ECHINODERMA TA—ONTOGEN Y

509

cesses are of two kinds. One kind, which belong to the region of the circumoral
ciliated ring, are paired, and diverge in a forward direction. Two arms are
distinguished by their constant occurrence and special length ; these belong to the
posterior and lateral region of the circumoral ciliated band. Opposed to these paired
processes of the circumoral ciliated band pointing anteriorly is an unpaired, posterior,
postanal process projecting backwards from the posterior end of the anal area ; its
tip may carry a cap of cilia.

In Ophiuroidea in which care of the brood occurs, the typical larval forms are not
developed.

4. Echinoidea (Figs. 400 and 401). — The larva of Echinoidea agrees to a great
extent with that of the Ophiuroidea, and is, like it, known as the Pluteus. The only

anus

FIG. 400.— Larva of an Echinid (Pluteus) from
the ventral side. 1, Ciliated "epaulettes " ; out, an-
terior ; post, posterior; de.c, right ; sin, left.

FIG. 401.— Spatangid larva (Pluteus)
from the ventral side. 1, The three proci-ssr~
of the anal. area.

important difference is that the two lateral arms which, in the Ophiurids, are the
most constant and the longest, seem to be altogether wanting in the Echinoidea.

The Pluteus of Echinus has eight arms or processes, and at the bases of each of
the four posterior arms a ciliated " epaulette" (Fig. 400).

The larva? of Arbacia and Spatanyus (Fig. 401) have no ciliated "epaulettes,"
but Arbacia has two and Spatancjus three long posterior processes of the anal area,
which, like all the other processes, are supported by calcareous rods.

Echinoids in which care of the brood occurs develop direct without meta-
morphosis.

5. Crinoidea (Fig. 402). — The free-swimming larva of Antalo/i is long and egg-
shaped. At the frontal pole, the thickened but somewhat depressed ectoderm (the
neural pit or plate) carries a tuft of long flagella. The larva, in swimming, has the
frontal pole, which corresponds with the anterior end of other Echinoderms, directed
forwards.

510

COMPARATIVE ANATOMY

CHAP.

ant

The body is surrounded by five ciliated rings, distinct from one another ; these
cannot be ontogenetically derived from one continuous ciliated ring.
The most anterior ring is interrupted on the ventral side.

The second ring runs somewhat diagonally from above downwards and forwards,
the third is directed downwards and backwards, so that there is a large interval

between the second and third rings ventrally.

In this region, the ventral side sinks in to
form a large ciliated vestibular depression.

A smaller depression on the ventral side
between the first and second ciliated ring is
known as the adhesive pit. The larva attaches
itself at this point, by means of a special
secretion yielded by the glandular cells of the
depression.

To the left, between the third and fourth
rings, there is a small aperture, the primary
madreporite (water pore).

The intestine lies as an entirely closed sac in
the posterior part of the larva. The free-swim-
ming larva has neither larval mouth nor larval
anus. The definitive mouth breaks through the
floor of the vestibular depression later.

The whole anterior part of the larva, as far
FIG. 402.— Free -swimming larva of as the third ciliated ring, becomes the stalk, and
Antedon, from the right lower side (after the posterior part the calyx of the attached

T>IIU-PT\ T IT" mi,** tt,,n ,>;i ,*,-, *-,,,i „:,,,-.„ . i -ti —

larva.

The free-swimming Crinoid larva cannot with
certainty be derived from the same hypothetical
form as other Echinoderm larvse. The difficulty

consists in the varying number and arrangement of the ciliated rings, which most
recall the condition in the Holothurid larva (pupa), to be described later. The
vestibular depression of the Antedon larva may, however, be compared without
forcing to the ventral depression of the other Echinoderm larva. A thickening of
the ectoderm, comparable with the neural plate of the Antedon larva, also occurs, as
we shall sec in other Echinoderm larvae.

Bury). /- V, The five ciliated rings ; 1, the
neural tuft ; 2, the adhesive pit ; 3, the
vestibular depression ; d, dorsal ; !>,
ventral.

B. Ontogeny of the Holothurioidea.

The segmentation of the ovum is total and equal, and leads to the formation of a
cceloblastula, whose unilaminar cell wall usually consists on one side of somewhat
larger cells. By invagination of this part of the blastula wall, a ccelogastrula is
formed. The invaginated part, i.e. the archenteron, is a blindly ending tube, with
narrow lumen (archenteric cavity), which is far from filling the segmentation cavity.
This latter is filled with an albuminous, fluid or semifluid, mass, the gelatinous
nucleus.

The ectoderm and the endoderm are ciliated.

During the process of invagination (occasionally even during the blastula stage)
cells arise by division from the ectoderm, but more especially from the endoderm,
which, as mesenchyme cells, wander into the enclosed jelly-like substance, multiply
by division and, in ever-increasing numbers, occupy the blastocn-1. From them is pro-
duced all the connective tissue of the Holothurian body. The calcareous corpuscles
arise exclusively in the mesenchyme.

VIII

ECHINODERMATA^ONTOGENY

The blind end of the lengthening archenteron bends to that side, which becomes
the dorsal side of the larva as it rapidly grows bilaterally symmetrical (Fig. 403, A),
and at the same time it moves somewhat to the left side. As the hydro-enteroccel
vesicle, it soon becomes entirely constricted off from the rest of the archenteron,
which opens outward through the blastopore (Fig. 403, B, C, D, 4, 5).

This constriction from the archenteron, the hydro-enteroccelomic vesicle, is of
the greatest importance, because out of its wall arises the whole musculature of the
body and all the internal epithelia, i>. the ccelomic and water vascular epithelia.

The hydro-enterocu'lomic vesicle increases in length, alongside of the intestine,
in the direction of the blastopore, and again divides into two vesicles by means of a

Fin. 403. — Formation of the larval mouth
and the hydro - enterocoalomic vesicle in the
gastrula larva of Synapta digitata, diagram-
matic (after Selenka). .1. Gastrula, the archen-
teron bent towards the dorsal side ; P>, archen-
teron, opening outward through the hydropore ;

C, hydro-enteroccel, constricted from the intestine;

D, intestine, opening outward through the larval
mouth ventrally. 1, Segmentation cavity, blasto-
cojl ; 2, archenteron ; 3, blastopore ; 4, hydropore ;
5, hydro-enterocoel ; 6, intestine ; 7, oesophagus ;
8, mouth; ant, anterior; post, posterior; r, ven-
tral ; d, dorsal.

Fin. 404.— Auricularia with the left half of
the ectoderm removed, from the left (after
Ziegler's model). The organs lying in the
segmentation cavity (11) are seen. 1, Cut edge
of the ectoderm ; 2, mouth ; 3, oesophagus ;
4, mesenchyme cells ; 5, mid-gut or stomach
intestine ; 6, anus ; 7, hind-gut ; 8, left entero-
ceel, still connected with the hydroccel 10, the
latter showing slight indications of the first
radial outgrowths ; 9, dorsal pore or hydro-
pore ; 11, blastoctt'1, segmentation cavity.

transverse constriction. The anterior vesicle (that further from the blastopore) is
the hydroccelomic vesicle, which at once sends oft' a canal to the dorsal side, which
opens outwards through a pore on the left of the middle line. The canal is the
primary stone canal, and the pore the primary madreporite. The hydroccelomic
vesicle is the rudiment of all the rest of the water vascular system, and in the first
place, of course, of the circular canal (Fig. 404).

The first appearance of the various structures just described does not occur in the
same order in all species of Holothurioidea examined on this point. The hydro-
enterocu'lomic vesicle may become connected with the exterior through a stone canal
before it has divided, or even (a unique condition found in Synapta digitata) before

512

COMPABA TIVE ANA TOMY

CHAP. VIII

it has itself separated from the archenteron (Fig. 403). In the last case, it can be
established that the archenteron which begins with the blastopore opens outward for
a time through a second aperture, the primary madreporite.

After the hydro-enterocalomic vesicle has become constricted from the archen-
teron, the intestine grows further, its blind end bending to the ventral side (that lying
opposite to the water pore), which commences to become depressed and to sink in.

The blind end of the intestine soon becomes applied to the, ectoderm of the
depressed ventral side of the larva, about half way down the body, or a little in

CCJi

FIG. 405.— Young Auricularia of
Synapta, from the ventral side (after
Semon). 1, Circumoral ciliated band ;
2, entero - hydrocrel ; 3, calcareous
wheel ; 4, adoral ciliated ring ; os,
mouth; mi, anus; u, mid -gut or
stomach intestine ; (5, larval nerve
band.

FIG. 40(3.— Older Auricularia, seen diagonally from the
left lower side (after Semon). 1, Circumoral ciliated band ;
2, hydropore ; 3, hydroccel ; 4, adoral ciliated band ; 5, mill-
gut or stomach -intestine; 0, nerve band; 7, hind -gut;
S, left enteroccel ; 9, calcareous wheel ; 10, fore-gut, oeso-
phagus; 11, right enteroccel.

front of the middle point. Where the two touch one another, an aperture, the mouth,
breaks through.

The median portion of the intestine (the mid-gut) swells up and becomes distinct
both from the fore-gut and from the hind-gut.

In the meantime, the larva has undergone a change of shape through which it
reaches the Auricularia, stage, the depression of the ventral side being the most
important part of this change. The general ciliation has disappeared ; of the
complete covering of cilia, only the circumoral ring and the adoral band are
retained, and the region immediately around the mouth has become depressed to
form the oral vestibule (Figs. 404-407).

The transformation of the Auricularia into the barrel-shaped larva (Figs. 408-

413). The Auricularia does not change direct into a young Holothurid, but passes

through an intermediate stage, which was formerly known as the pupal stage,
because during it no nourishment is taken.

FIG. 407.— Older Auricu-
laria (after Semon). an,
Anus; os, mouth; 1, out-
growths (primary and
secondary) of the primitive
horse - shoe - shaped hydro-
coal ; 2, stone canal ; 3, left,
4, right enterocoelomic sac,
which have become closely
applied to the mid-gut.

PIG. 408.— Auricularia,
in which the circumoral
ciliated ring is beginning
to break up into lengths
(after Semon). The horse-
shoe - shaped hydrocoel is
growing round the intestine.
The first pieces of the cal-
careous ring (1) have ap-
peared.

VOL. II

514

COMPARATIVE ANATOMY

CHAP.

The Auricularia assumes the shape of a barrel. The circumoral ciliated ring
atrophies in sixteen places, which are indicated in the diagram (Fig. 413). The
sixteen lengths of the ring which remain continue to grow and join, as indicated

FIG. 400.— Old Auricularia.

Transition to the barrel -shaped
pupa, the whole body decreasing
considerably in size. 1, The nerve
bands, in the act of forming the
nerve ring ; 2, primary tentacle.

FIG. 410.- Intermediate stage be-
tween Auricularia and the barrel-
shaped pupa of Synapta (at in
Semon). l-l", the rudiments of the
five ciliated ring's. 1, The oral funnel;
2, the primary ; 3, the secondary out-
growths of the water vascular ring:
4, pieces of the cah-aivous ring; 5,
ecekmiic vesicle ; 0, water vascular ring.

by dotted lines (Fig. 413), to form five ciliated rings entirely encircling the barrel-
shaped body. The centre of the former oral area becomes surrounded by four lengths
of the ciliated band which join together to make a square. The part of the oral

Km. 411.— Young barrel-shaped
larva (pupa) (after Semon). 1, Oral
funnel ; 2, tentacles ; 3, pieces of the
calcareous ring ; 4, Polian vesicle ;
5, left cceloni ; 0, hind-gut ; 7, audi-
tory vesicles ; 8, secondary out-
growths of the hydroccel ring.

Km. 412.— Barrel-shaped larva,
with the tentacles (1) beginning to
project from the opening oral funnel
(after Semon). 2, Water vessels of
the body— secondary outgrowths of
the circular canal ; 3, the rapidly
swelling enteroccel.

area enclosed by this ring sinks below the surface, and thus increases the size of the
oral vestibule. The ciliated square itself becomes depressed and forms the oral
shield. The spacious oral vestibule becomes cut off from the exterior, with the
exception of a very narrow aperture, and shifts quite to the front, so that now the

vm

ECHINODERMA T A— ONTOGENY

515

three apertures, that of the oral vestibule, the mouth itself lying in its floor, and the
anus, lie almost in the axis of the barrel-shaped body.

In the older Auricularia stages and during the transformation into the barrel -
shaped larva important internal processes take place.

Calcareous corpuscles appear early (even in the younger Auricularia} in the
mesenchyme. In the best known Auricularia, that of Synapta diijitata, these
bodies appear in the form of wheels in the two
posterior tips of the larva (cf. Fig. 404, p. 511,
and following).

The hydroccelomic vesicle assumes the form
of a horse-shoe with the curve towards the dorsal
side. On the convex side of this horse-shoe-
shaped vesicle five outgrowths appear. The two
arms of the horse-shoe then close round the fore-
gut, growing towards each other round it until,
finally, they meet and fuse (probably in the
right half of the body). The horse-shoe-shaped
hydroccel is now the closed water vascular ring
surrounding the fore-gut. It continues, as before,
to communicate with the exterior through the
primary stone canal and the dorsal water pore.

The five outgrowths of this hydroccel ring
now become more distinct. They are originally
directed forwards, but very soon, with further
growth, bend backward, and, as the rudiments
of the radial canals of the water vascular
system, grow further back under the body wall,
in the five radii. The rudiments of the tentacle
canals appear very early on the rudiments of
the radial canals as orally directed lateral out-
growths.

of the Auricularia unite to form the rive
complete rings (I -I').

Fir,. 413.— Diagram illustrating the
rise of the five ciliated rings of the
Holothurian pupa from the pieces of the
ciliated bands 1-7 and l'-7' of the Auri-
culuria (after Ludwig). The pieces of
the ciliated band are marked by broad
black lines, the interruptions being left
clear. 8, The preoral ; 9, the postoral
The above account of the first processes of intermediate piece of the ciliated ring ;
differentiation in the hydroccel vesicle are those os, mouth. The dotted lines give the
found in Cucumaria Planci, the ontogeny of direction in which the pieces of the ring
which has recently been carefully investigated.
In other Holothurioidea, at least in Synapta
digiLata, according to former authors, the ontogenetic processes differed essentially
from these. The first five outgrowths of the hydroccel in Synapta develop exclusively
into the tentacle canals, and only after the appearance of these and alternately
with them, five other outgrowths form the rudiments of the radial canals.

This and certain other discoveries led to the conclusion that the radial canals
in the Holothurians arise interradially and only shift into the radii secondarily,
hence it was inferred that the tentacle canals of the Holothurioidea were homologous
with the radial canals of other Echinoderms, and that the radial canals of the Holo-
thurioidea are not represented in other classes. The above discoveries in the larva
of Cucuitiaria Planci dispose of this suggestion, which must always have appeared
improbable to comparative anatomists.

It is a very noteworthy fact that, in Synapta, the radial canals appear onto-
genetically, whereas they are wanting in the adult.

The Polian vesicle also arises as an outgrowth of the circular canal ; in Cucu-
maria Planci, it forms at the point where it lies in the adult, i.e. in the left dorsal
interradius.

The tube-feet arise as outgrowths of the radial canals, which push the ectoderm

516

COMPARATIVE ANATOMY

CHAP.

out in front of them (Fig. 415). The first two tube-feet in Cucumaria Planet arise
simultaneously at the posterior end of the body. Both these feet belong to the
medioventral radial canal.

The differentiation of the enteroccel vesicle. — After the hydro-enteroccel
vesicle has divided into the hydroccel and the enteroccel vesicles, the latter grows
backwards longitudinally, the growing posterior end pushing its way gradually over
the intestine, on to its right side (i.e. into the right-hand portion of the segmenta-
tion cavity). The anterior part of the enteroccel vesicle now lies to the left, the

posterior to the right, near the intestine (Fig.
404). These parts become completely separated
by a constriction which stretches transversely over
the intestine, into a left and a right enteroccel
vesicle.

Each of these vesicles becomes applied to the
intestine, and, increasing in size, takes the shape
of a hollow disc resembling a watch glass.

The nervous system of the larva. — On each
side of the Aurictihirin larva, on the ventral side,
in the oral area, there is a ciliated ectodermal
ridge, beneath the surface of which ganglion cells
lie and longitudinal nerve fibres run. The ridge
consists of two limbs, meeting in an obtuse angle
which is open towards the mouth. From the ends
and angles of these ridges nerve fibres go otf to
the circumoral ciliated band.

Formation of the tentacles. — The tentacle
canals, whether as lateral outgrowths of the radial

Km. 414.— Young Synapta (Pentac-
tula) (after Semon). 1, Oral tentacles ;
2, auditory vesicles ; 3, pieces of the
calcareous ring ; 4, water vascular
ring : 5, Poliaii vesicle ; t>, radial

LlUg . i», J lc > ") J .. -I

vessels of the water vascular system ; canals or direct outgrowths of the water vascular
7, hind - gut ; 8, calcareous wheels ; system, grow towards the oral vestibule, and press

'.i, mid-gut ; 10, madreporite ; 11, stone
canal.

out the ectodermal wall of the latter before them.
The ectodermal covering thus afforded them is
derived from the oral shield, i.e. indirectly from parts of the original circumoral
ciliated ring of the Auricularia larva. The tentacles (five of which form first)
remain hidden in the oral vestibule during the pupal stage.

Transformation of the barrel-shaped larva into the young Holothurian (Figs.
414 and 415).— The eternal changes are as follows. The ciliated rings atrophy.
The tentacles project freely from the expanding and widely opening vestibule, and
increase in numbe'r. In the Actitiopocla tube-feet are formed in all the five radii.

It is an important fact that, in the comparatively simple transformations of the
Holothurian, not only does the whole of the larval epithelium pass into the body
epithelium of the adult, but none of the larval organs are eliminated.

The following description applies to Cucumaria Phmci.

The tentacles.— It is somewhat remarkable that the five tentacles first formed do
not each arise from a single radial canal. Two of the five tentacles, on the contrary,
receive their canals from the medioventral, and two others from the left dorsal
radial canal. The fifth tentacle belongs to the right dorsal radial canal. Other
tentacles appear only at a very late stage, two more being added first, a sixth and a
seventh. These belong to the two lateral ventral radii, which up to this time have
been without tentacles.

The stone canal. —An anterior outgrowth forms on the primary stone canal ; the
epithelium of this outgrowth flattens, giving rise to the madreporitic vesicle. On
the wall of this vesicle the mesenchyme forms an incoinplete shell, perforated like a
lattice.

vin ECHINODERMA TA— ONTOGENY 5 1 7

The water pore lying on the right side of the mesentery (now formed) disappears

19

FIG. 415.— Longitudinal section of a larva of Cucumaria doliolum (after Selenka). 1, Frontal
prominence with enclosed jelly-like substance ; 2, tentacle vessels ; 3, (5, 15, 14, 13, radial
vessels ; 5, stone canal ; 8, madreporite ; 4, Polian vesicle cut oft' ; 7 and 12, coelom, enterocoel
vesicle ; 0, anus ; 10 and 11, the tirst two tube-feet ; 16, circular canal of the water vascular
system ; 17, oesophagus ; 18, mouth ; 19, niesenchyme cells ; II, III, IV, V, ciliated rings.

later, and still later the madreporitic vesicle opens into the body cavity, and thus
becomes the secondary inner madreporite.

518 COMPARATIVE ANATOMY CHAP.

The radial canals. — The five radial canals do not develop with equal rapidity,
nor indeed do the radial nerves and the radial longitudinal muscles. The medio-
ventral organs (radial canal, radial nerve, longitudinal muscle) in all cases arise first,
then follow the organs of the two dorsal radii, and only at a later stage, those of
the two lateral ventral radii.

The tube-feet. — In agreement with the order of appearance just described, the
two first tube-feet, already mentioned above, belong to the ventral radius (Fig.
415). The two next in order also belong to the medioventral radial canal, and,
according to the rule which applies to all the newly developing tube-feet, arise in
front of those already present. The fifth tube-foot belongs to the left dorsal radial
canal. (The correspondence of this order with that of the rudiments of the tentacles
should be noted.)

According to observations which have been made, it appears that
when, in a Holothurian, the tube-feet are scattered, this arrangement
is, ontogenetically, secondary. In the same way animals which have
several rows of tube-feet in each radius have only two rows in a young
stage, or a zigzag row of alternating feet.

The nervous system.- — The first part of the nervous system to
appear is the oral circular nerve, and this arises as an eetodermal
circular ridge on the floor of the oral vestibule in the larva. It
sends out five band-like processes in the direction of the rudiments of
the radial canals ; these are the rudiments of the radial nerves.

In that the rudiments of the circular nerve and the radial nerves
become subepithelial, there arises between them and the body
epithelium which closes over them a narrow space ; this is the
epineural canal.

The rudiments of the five radial nerves grow backwards together
with those of the radial canals.

In Cucumaria- Planci there seems to be no larval nervous system.
In Syimpta, on the contrary, the larval nervous system yields the
rudiments of the definitive system. The two lateral nerve ridges of the
Auricularia larva, when the oral vestibule of the barrel-shaped larva is
formed, shift into it. Their free ends then become connected from
the two sides to form a ring encircling the mouth, which is the rudi-
ment of the nerve ring.

The intestine shows, at an early stage, the coils characteristic of the adult.

The first portions of the calcareous ring to appear are the five radial pieces : these
arise on the radial canal and, like all the calcareous structures, are produced by the
mesenchyme. The medioventral calcareous piece is from the first the largest.

The enteroccel. — The right and the left enterocoel vesicles grow round the intes-
tine. At the point where they meet ventrally they open into one another.
Dorsally they press the mesenchyme cells together in a vertical lamina. In this
way the dorsal (anterior) mesentery arises. The middle and the posterior mesen-
teries probably arise in consequence of the two enteroccel vesicles twisting round the
intestine posteriorly.

The visceral wall of the enteroco.4, which becomes applied to the intestine,
presses the mesenchyme cells, which have greatly increased in number, against the
endodermal intestinal wall, till they form a layer, which becomes the connective

vin ECHINODERM AT A— ONTOGENY 519

tissue layer of the intestine. In a similar way, the parietal wall of the enterocrel
presses the peripheral mesenchyme cells to form a layer below the ectodermal body
epithelium ; this layer forms the cutis of the body wall.

Whilst these processes have been going on (i.e. while the enteroccel vesicles are
increasing in size till they surround the intestine, the parietal wall becoming applied
to the body wall, and the visceral wall to the intestinal wall of the young animal,
and the two vesicles opening into one another ventrally), the narrow slit-like cavities
of the watch -glass -shaped enterocoel vesicles in the larva have become the large
spacious body cavity.

The visceral wall of the enteroccel yields the intestinal musculature and the
endothelial covering of the intestine ; the parietal wall of the enteroccel gives rise
to the longitudinal and circular musculature of the body wall and its endothelial
covering. Since the musculature of all parts of the water vascular system is
formed from the epithelial wall of that system, it follows that the whole of the
musculature of the Holothurian body is of epithelial origin.

The blood lacunar system arises in the form of cavities in the connective tissue
(mesenchymatous) layer of the different organs (integument, intestine).

The rudiments of the genital organs, of the respiratory trees, nnd of the Cuvierian
organs are unknown.

Not all the Holothurids pass through a distinctly marked Auricularia stage.
The gastrula larva of Cucumaria Planci, for example, passes direct into the stage of
the barrel-shaped larva. Nevertheless, " when the oral depression of the Cucumaria
begins to form, it is beset along its edge with wreath-like ectodermal protuberances
(ciliated protuberances) which, taken together, may be compared with the ciliated
ring of the Auricularia larva." One trace of the Auricularia stage seems thus to
be retained.

The preoral region in the Cucumaria larva rises up sharply as the frontal
prominence, and in its gelatinous nucleus an excessive growth and multiplication of
the mesenchyme cells early takes place. This is resorbed at a later stage as
nourishment for the formation or further development of the different organs.

The plane of symmetry of the young Holothurid does not coincide with that of
the larva, but deviates from the latter, anteriorly to the left, and posteriorly to the
right. The longitudinal axis of the young Holothurian deviates anteriorly ventrally,
and posteriorly dorsally, from that of the larva.

C. Ontogeny of the Echinoidea.

The following description is an epitome of the results gained by various investi-
gators in the examination of different Echinoids.

Segmentation is total, and, in a peculiar manner, unequal. The inequality
between the blastomeres, however, soon almost entirely disappears. A spherical,
egg-shaped, or (Echin ocyam us pusillus) long elliptical cceloblastula arises, with a
unilaminar wall, whiSh becomes covered with long flagellate hairs (one on each cell).

The wall of the blastula becomes thickened at the vegetative pole. At this
thickened part, the blastoderm cells divide actively, so that the wall becomes bi- or
trilaminar. The deeper cells pass in succession into the segmentation cavity,
become amu-boid, and are the first mesenchyme cells. At this same point the
blastula wall sinks in to form the archenteron. The blastula becomes a gastrula.
During this process of invagination, the wandering of mesenchyme cells out of the
wall of the archenteron into the blastoctel continues.

In Echinocyamus (and other Echinoids?) the cells of the blastula wall at the
animal pole (the point opposite the later blastopore) are longer than the rest even at

520

COMPABA TI VE ANA TOM Y

CHAP.

the blastula stage, and their flagella are less movable. This differentiated part
(larval sensory organ ? neural plate ?) can also be recognised in the subsequent
stage.

First Pluteus stage. — The gastrula becomes concave on the ventral side ; on the
opposite (dorsal) side it becomes convex. The larva is now bilaterally symmetrical.
The blastopore at first indicates the posterior end ; then it shifts somewhat on to
the ventral side on to a mound-like bulging of the body (the anal area), which lies
posteriorly to the ventral depression. The anterior edge of this anal area becomes
drawn out into two anteriorly diverging processes, the two posterior ventral arms

ant

PIG. 416.— EcMnocyamus pusillus, gas-
trula, forty hours after fertilisation (after
Theel). 1, Blastocwl ; 2, frontal thickening
of the ectoderm ; 3, niesenchyme cells ;
4, formation of these wandering cells at the
base of the archenteron ; 5, the first two cal-
careous spicules ; 6, archenterou ; 7, primitive
mouth, blastopore.

tie i'

post

Fie;. 417.— Echinocyamus pusillus, young
Pluteus, about forty-eight hours after fertil-
isation (after The'el), from the ventral side.
1, Rudiment of the larval mouth ; 2, the first
arms ; 3, rudiment of the hydro-enteroccel
at the base of the archenteron ; 4, larval
skeleton ; 4j, dorsal branches of the same :
5, archenteron ; 0, primitive mouth, blasto-
pore.

(Fig. 417). The circumoral ciliated ring, which is continued on to the arms,
becomes raised above the general ciliated covering of the body.

During the first larval stage the following important internal changes take pla<-f.
The two first lateral calcareous spicules develop in the mesenchyme and send sup-
porting rods into the only two arms present at this stage, the posterior ventral arms.
(The first rudiments of these two spicules can be made out even in the gastrula
larva, Fig. 416.)

Formation of the hydro-enteroccel. — The anterior blind end of the archenteron
has, on each side, a small outgrowth, which lengthens posteriorly. The archenteron
becomes constricted immediately behind these outgrowths, which finally become
separated from it in the form of a horse-shoe-shaped vesii-le, with two limits directed
posteriorly and applied to the archenteron. This hydro-enteroccel vesicle at once
divides into its two lateral limbs, forming two lateral hydro-enteroco-l vesicles. In

VIII

EGHINODERMA TA—ONTOGEN Y

521

— rfors

certain species this division seems to take place as early as the constriction of the
hydro- en terocoel from the archenteron.

From the very commencement of the formation of the hydro-enteroca-l, the three
sections of the intestinal canal, the hind-gut, the mid-gut, and the oesophagus, begin
to be marked off by constrictions, which become more distinct. All three, especially
the middle section (the stomach intestine), begin to widen out like vesicles. The
blastopore having shifted, as larval anus, far from the vegetative pole ventrally to
the centre of the anal area, the hind-gut is drawn up and forms an angle with the
mid-gut (Fig. 418). After the hydro-enteroccel has been constricted off, the new
blind end of the intestine bends ventrally
and soon meets a small invagination of the
ectoderm of the depressed ventral surface.
These break through into one another, and
the larval mouth thus enters into open
communication with the lumen of the
fore-gut.

During the latter part of the first larval
stage, contractile fibres appear in the wall
of the cesophagus, and enable this section
of the intestine to contract strongly.

Of the two hydrocoel vesicles lying at
the sides of the posterior portion of the
fore-gut, that on the left opens outward
by a water pore at the centre of the dorsal
surface.

Second larval stage. — The two posterior
dorsal arms grow out. They are supported
by the rod-like processes of two new cal-
careous bodies, which have appeared in the
mesenchyme. On the left side, in the
angle between the posterior dorsal and the
posterior ventral arms, an ectodermal in-
vagination appears (Fig. 419, 3) ; this
sinks into the blastoccel in the shape of a

flask. This invagination plays an important part in the transformation of the larva
into the young Echinodenn.

Third stage (fully-grown Pluteus larva). — The two anterior dorsal and the two
anterior ventral arms continue to grow (cf. Figs. 400 and 401, p. 509). On the dorsal
side, a fifth unpaired calcareous spicule arises, and, in the immediate neighbourhood
of the water pore, sends off processes, two of which enter the anterior dorsal arms
and support them. The body has shortened, and its posterior region has become
rounded off.

Further differentiation of the hydro-enteroccel. — We resume the description of
this organ from the stage in which it consisted of two lateral vesicles applied to the
intestine. Each vesicle now becomes divided by a constriction into an anterior
and a posterior vesicle. The two anterior vesicles lie at the sides of the posterior
portion of the esophagus, the two posterior at the sides of the stomach intestine.
The left anterior vesicle opens outward through a water pore ; the other three are in
no way connected with the future water vascular system ; they are distinguished as
the right anterior, right posterior, and left posterior enterocoal vesicles. Some-
what later, three vesicles are seen on the left side (Fig. 419, 2, 4, 5). The anterior
and middle vesicles are in communication, whereas the posterior is distinct, and
becomes applied to the middle vesicle, assuming a crescent shape. The left

post

FIG. 418. — Echinocyamus pusillus, the
same larva from the left side (after Theel).
1, The first larval arms; 2, rudiment of the
larval mouth ; 3, ectoderm ; 4, hydro-enterocoel
rudiment ; 5, archenteron ; 0, blastopore ;
7, larval skeleton.

522

COMPARATIVE ANATOMY

CHAP.

anterior (enteroccel) vesicle is the one which opens outward through the water
pore ; it, however, does not become the hydroccel, but the middle vesicle (which is
in communication with it) now represents the rudiment of the hydroccel. It is
probable that this is produced by constriction from the left anterior hydro-enterocosl
vesicle.

On the left side, the development of the hydro-enterocoel is now as follows.
The water pore leads into the posterior end of a left anterior enteroccel vesicle,
which, again, communicates, by means of a constricted portion, with the hydroccel

ant

lex

FIG. 419.— Dorsal aspect of an Echinoid
Pluteus, to illustrate the relations of the hydro-
enterocoel (after Bury). «. nt, Anterior ; post, pos-
terior ; sin, left ; dex, right ; 1, larval oesophagus ;
'2, left anterior enteroccel ; 3, ectodernial in-
vagination ; 4, rudiment of the hydrocoel ; 5, left
posterior enteroccel vesicle ; (i, stomach intes-
tine ; 7, right posterior enterocoel ; 8, hydro-
pore; 9, unpaired dorsal skeletal piece; 10, right
anterior enteroccel. The arms are not fully
represented.

FIG. 420.— Rudiment of the Echinoid
in the Pluteus larva of Echinocyamus
pusillus (at'ler Theel). The Pluteus is seen
from the dorsal side, and only the left side
is completely drawn. 1, Arms of the
Pluteus ; 2, aperture of invagination of the
sac (3), whose floor will form the oral body
wall of the Echinoid ; 4, outgrowths of the
hydrocoel, which push this wall before them
and form the first ambulacral tentacles ;
skeletal rods of the Pluteus ; t>, hydrocoel ;
7, hydropore ; 9, portion of a skeletal piece
lying in the neighbourhood of the hydm-
pore, which will probably become the
madreporite ; 8, stomach-intestine.

vesicle. This latter is embraced posteriorly by the horse-shoe-shaped left posterior
enterocic-1. The stone canal does not arise out of the water pore, but out of the
connecting piece between the left anterior enteroccel vesicle and the hydroccel
vesicle, which becomes drawn out into a canal. The left anterior enteroccel
seems to become the madreporitic ampulla.

(The above description of the differentiation of the hydro-enteroccel must not be
considered as fully established. The observations are not quite continuous, and
do not all agree.)

VIII

EGHINODERMA TA— ONTOGENY

523

Transformation of the Pluteus larva into the young Echinoid. — This meta-
morphosis is far from having been satisfactorily described, its investigation being
exceedingly difficult.

An important part in the shaping of the Echinoid body is played by the above-
mentioned flask-like invagination of the ectoderm on the left side. The thicken-
ing floor of this invagination grows towards the hydroca'l, and becomes externally
applied to this latter as the "Echinoid disc" (Fig. 420). The thin lateral walls of
the capacious flask, which is still connected by its neck with the larval ectoderm,
are known as the amnion.

The hydroccel vesicle assumes the horse -shoe -shape, and at the same time puts
forth five outgrowths which push before them the Echinoid disc, i.e. the floor of

ant

ii

Fio. 422.— Lateral view of a very
young Echinoid (Echinocyamus
pusillus), forty -live days old (after
Theel). The first tube-feet and spines
of the Echinoid are seen, and, attached
to its back, the remains of the calcar-
eous rod of a larval arm.

Fio. 4-21.— Dorsal aspect of a larva of Echi-
nocyamus pusillus, about forty-five days old
(after The'el). 1, The larval anus, with their
calcareous rods ; 2, unpaired calcareous rod,
taking part in surrounding the dorsal pore. (3) ;
4, spines ; 5 and fi, primary tentacles of the
young Echinoid ; a tit, anterior ; post, posterior ;
sin, left ; dex, right.

the flask-like invagination. Five hollow tubes thus now project into the cavity of
the flask, which continually becomes more and more spacious ; these are the five
primary tentacles, which receive their covering from the spreading Echinoid disc.
This Echinoid disc forms the oral wall (that is, no doubt, only the epithelium and
the nerves ?) of the young Echinoid, while the apical wall is formed direct from the
larval dorsal ectoderm of the Pluteus.

The fate of the amnion is differently described for different forms. Sometimes
it is said to pass over into the young Echinoid, the amnion sac opening and spread-
ing out, and yielding the circular integumental region between the apical and oral
surfaces of the body. At other times, again, the amnion sac is said to remain closed

524

COMPARATIVE ANATOMY

CHAP.

and the amnion, together with part of the larval integument, are lost when the
larva changes into the young Eehinoid.

The larval arms disappear, and their spicules are for the most part absorbed.
As a rule, one or other of the arms of the Pluteus still adheres to the quite young
Eehinoid (Fig. 422).

The intestine, at least the whole stomach, the spreading enterocu/1, and the
growing hydroccel are taken over into the young Eehinoid ; the latter, however,

has, at first, neither mouth nor
anus. In Echinoids, therefore,
the larval mouth and anus do not
pass direct into the corresponding
organs of the adult.

Formation of the mouth and
the definitive oesophagus. — Ac-
cording to one account, the oeso-
phagus only grows out from the
intestine after the horse -shoe-
shaped hydroccel has closed ; it
then passes through the water
vascular ring and opens outward
at the centre of the Eehinoid disc
through the definitive mouth.

The pedicellarise arise very
early. They are even occasionally
seen on the dorsal side of an older
Pluteus larva.

The water pore becomes the

Fi.;. 4-23.-Echinocyamus pusillus, young Eehinoid, madreporite, and the unpaired
about forty-five days old, from the oral side (after The"el). spicule which, in the older Pint- US,
"nt. Anterior unpaired ambulacrum; post, posterior un- arose in its immediate neighbour-
!"!!'*! *"!ei™M"!a"T: h*?*'165 ;>.^ilies;>3;sl3l(uvl" hood, changing into a lattice-like

plate, becomes the madreporitic
basal. Four other plates, which
arise over the right enteroccel

of the larva, become the other basals. In their centre the dorso-central is soon
distinguishable. On the oral side, in the peripheral part of the original Eehinoid
disc, where the primary tube-feet developed, the first ambulacral and interambu-
lacral plates appear, with the rudiments of the spines and the sphseridia, both
of which form independently over the plates (Fig. 423). In the future oral area,
which is surrounded by a circle of ambulacral and interambulacral plates, thirty
small calcareous centres form, three in each radius and three in each interradius ;
these are the rudiments of the plates of the masticatory apparatus. The middle
calcareous plates of the interradii become the teeth.

Little or nothing is known of the ultimate fates of the other twenty-five pieces, or
of the enteroccel, of the hydroccel (e.g. the order of appearance of the tube-feet), or as
to the appearance of the nervous system, the origin of the radial plates, etc.

idia in their niches ; 4, pieces of the masticatory apparatus ;
5, teeth ; 6, oral integument, the mouth is not yet formed ;
7, radial skeletal plates; S, interradial skeletal plates.

D. Ontogeny of the Asteroidea.

Segmentation is total, and leads to the formation of a cceloblastula, through the
invagination of which a ccelogastrula arises. The formation of the mesenchyme
takes place in the manner already described for the Holothurioidca and the Echin-

VIII

ECHINODEEMA TA — ONTOGENY

525

oidca, and commences either in the blastula stage or not until the gastrula stage.
In the former case, that part of the blastoderm which becomes invaginatecl to pro-
duce the archenteron yields the mesenchyme, and, after invagination has taken
place, continues to produce it. In the second case, also, the cndoderm is the place
of formation of the mesenchyme cells, which wander into the blastoccel. Such
observations, however, seem to point to the fact that, although most mesenchyme
cells arise from the endoderm, the ectoderm also takes part in their formation.

In the older gastrula stage of Aster las vulgaris, the ectoderm seems to be
thickened at the (aboral) pole opposite to the blastopore. This may be the rudi-
ment of the neural plate.

As a further illustration of the development of the Asteroidea, we shall utilise
the observations made on Asterina gibbosa, in which form, however, a typical

ant

post

FIG. 424.— Asterina gibbosa,
gastrula four days old ; ap-
proximately horizontal longi-
tudinal section, from tlie ventral
side (after Ludwig). ant, An-
terior ; post, posterior ; dex, right ;
sin, left ; 1, segmentation cavity ;
•2 and 3, right and left coelomic
outgrowths of the archenteron ;
4, blastopore.

yio.-'f

Km. 425.— Asterina gib-
bosa, larva at the end of the
fourth day, horizontal longi-
tudinal section seen from the
ventral side (after Ludwig).
The enterocoel outgrowth.-,
have grown in length. 2, right
enteroccel outgrowth ; 3, left
or hydro - enterocoel (.nit-
growth ; 5, intestine; 6, an-'
teri.>r unpaired coelom. The
crelom is still in open com-
munication with the intestine.

FIG. 42ti.— Asterina gib-
bosa, larva at the com-
mencement of the fifth
day, horizontal longitudi-
nal section (after Ludwig).
The enteroccel has become
constricted off from the
intestine. Lettering as
before.

Bipinnaria larva does not attain development. In the course of the description
observations made on other Asteroids will be referred to.

In the ovoid gastrula of Asterina, the blastopore does not lie altogether at the
posterior pole, but is shifted somewhat on one side, which in the further develop-
ment of the animal becomes distinguishable as the ventral side. Two sections can be
made out in the archenteron, a short cylindrical commencement (posterior section),
and a vesicular blind terminal part (anterior section). This description applies to
the gastrula on the second day.

Third day. — The rudiment of the hydro- enteroccel vesicle. — The anterior
vesicular section of the archenteron, which represents the rudiment of the hydro-
enteroccel vesicle, bulges out posteriorly on each side, while its wall becomes
thinner (Fig. 424). The two bulgings grow out longitudinally backwards, at the
sides of the posterior part of the archenteron, and become the two hydro-enteroccel

526

COMPARATIVE ANATOMY

CHAP.

vesicles, which continue to grow backward in proportion as the posterior part of the
archenteron, the larval intestine, grows anteriorly (Fig. 425).

Fourth day. — The whole hydro-enteroccel becomes constricted off from the
larval intestine, and is now found as a large vesicle occupying the anterior part of
the larval body, and continued posteriorly in the two long hydro-enteroccel vesicles,
the one on the left being longer than the one on the right (Fig. 426).

An imagination of the ectoderm, somewhat anteriorly to the middle of the
ventral side, represents the rudiment of the larval mouth and oesophagus, and.
towards the end of the fourth day, breaks through into the larval intestine.

Anteriorly a cushion-like thickening of the body appears encircling a depression.
ant

unt

post

FIG. 4i8.— Larva of As-
terina gibbosa four days
Old, just hatched, from
the ventral side (after
Ludwig). 1, Larval organ ;
2, blastopore. Here, ami
in the following figures,
a at = anterior ; post = pos-
terior.

FIG. 42!"". — Asterina
gibbosa, larva six days
old, from the left side
(after Ludwig). r, Ven-
tral side ; /-I, dorsal
side ; 1, larval organ.

FIG. 427. — Asterina gibbosa,
larva five days old, horizontal
longitudinal section seen from the
ventral side. First rudiment of the
hydroccel outgrowth (7) on the left
of the hydro-enteroccel vesicle (3).
The two enteroccel vesicles have
opened into one another posteriorly
at 8.

This circular cushion, the rudiment of the larval organ, slants from above anteriorly
to below posteriorly (Figs. 428-430).

At the end of the fourth day the embryo leaves the egg-envelope, and swims
abont freelv by means of the cilia covering its entire surface.

Fifth day. — The two hydro-enteroccel vesicles grow round the larval intestine
above and below. — Where they meet below, somewhat to the left of the median line,
they form a ventral mesentery, which, however, rapidly disappears, the two vesicles
opening into one another at this point. Above the intestine a dorsal mesentery,
lying to the right of the median line, arises in a similar manner, and persists.

The left hydro-enterocoel vesicle bulges out laterally somewhat behind its middle
point. This bulging is the rudiment of the hydroccel (Fig. 427).

At this stage, therefore, the hydro-enteroccel consists of the following sections in
widely open communication with one another.

1. Anterior unpaired enteroccel (6 in Figs. 426, 427), lying in the larval organ.

2. Right enteroccel vesicle (2 in the tigs.), anteriorly in wide open communica-
tion with No. 1, and ventrally in open communication with—

VIII

ECHINODERMA TA—ONTOGEN Y

r>27

3. The left enteroccel vesicle (3 in the figs.). This vesicle, again, lias an out-
growth (7) on the left, which is—

4. The hydroccel vesicle.

Simultaneously with the formation of the hydroccel rudiment, that of the water
pore appears dorsally, somewhat to the left of the median line, as an invagination
of the ectoderm, which grows towards the left enterocrel, and breaks through
into it.

Sixth and seventh day. — The outer form of the larva has been considerably
modified on the fifth day. The larval organ has increased in size, and its sloping
circular ridge projects considerably beyond the surface of the larval body.

The rudiment of the hydroccel has grown out further backward, but is still
anteriorly in open communication with the left enterooiel. Five outgrowths (Nos.
I-V in Fig. 435) now appear at its posterior edge ; these are the rudiments of the
five radial vessels. The water pore (dorsal pore, madreporite) still leads into the

post

FIG. 430. — The same
specimen of Asterina gib-
bosa viewed from the
left and. from the ventral
side. 1, Larval organ with
its dorsal and vi-ntnil
lobes ; 2, larval month.

^1

FIG. 431.— Asterina gibbosa, larva at the
beginning of the eighth day, from the left
side; the larval nnjan is very largely de-
veloped (after Ludwig).

left enterocoel. "A channel develops on that wall of the hydroccel which faces the
interior of the body, which soon closes to form a canal." This canal runs towards
the point where the dorsal pore opens into the left enteroccel. One end of this
canal remains in open communication with the hydroccel, while the other enters
the enteroccel quite near the aperture of the dorsal pore. This canal is the stone
canal of the future Asteroid. The dorsal pore of the larva does not thus lead
direct into the stone canal, but enters it through the left enteroccel (Fig. 436).
Only at a later stage does the dorsal pore come into direct connection with the stone
canal.

Formation of the hydro-enteroccel in other Asteroids. — In the larva of Astfrias
nil/iaris also, the entero-hydroccel arises in the form of two lateral diverticula of
the blind and somewhat swollen archenteron, whose wall has become thinner. The
two diverticula soon become constricted from the archenteron, and become distinct
vesicles. Each sends off an outgrowth towards the dorsal surface, a growth of the
ectoderm running in towards it. The two meet and fuse, become hollow, and form
the stone canal with the water pore. Thus in the young Bipinnaria larva of
Asterias vulgaris, the bilateral symmetry is so marked that a right as well as a
left stone canal attains development (Fig. 432). The right pore, however, soon
disappears, and the right canal somewhat later.

528

COMPARATIVE ANATOMY

CHAP.

The two lateral mesoderm vesicles lengthen and fuse in front of and above the
mouth, and, further, surround the intestine. On the left vesicle (hydro-enteroccel
vesicle) a transverse constriction appears, which finally divides it into two vesicles,
an anterior, which at its posterior end opens outward through the stone canal and
water pore, and a posterior (Fig. 433).

Further development of the hydroccel in Asterina gibbosa. — After the seventh
day the five outgrowths of the hydroccel (Figs. 437-440) become trilobate, and
later have five lobes. The unpaired terminal lobe of each outgrowth is the rudiment
of the terminal tentacle, the paired lobes are the rudiments of the- first two pairs

ant

ant

post

FIG. 432. — Larva 01
Asterias vulgaris, about
four days old, from the
dorsal side (after Field).
1, Circumoral ciliated
band ; -2, n.outh ; 3, right
and left hydro-enteroccel
vesicles, with their hydro-
pores (4) ; 5, cesophagus; ti,
mesenchymatous muscle
fibres ; 7, stomach intes-
tine ; 8, anus. The month
and the anus lie on the
side turned away from
the reader.

FIG. 433. —Dorsal aspect of a
Bipinnaria larva to illustrate the
developmentofthehydro-enteroccel

(nttiT Bury). 1, Larval cesophagus;
2, left anterior enteroccel ; 3, liyilm-
pore ; 4, rudiment of the hydrtifu-l ;
5, stomach intestine ; 0, terminals ;
7, left posterior enteroccel vesicle ;
S, dorsal mesentery ; 9, right posterior
enteroccel vesicle ; 10, madreporite ;

11, blood vesicle, pulsating vesicle ;

12, right anterior enteroccel.

dea

FIG. 434.— Asterina gib-
bosa, larva six days old,

horizontal longitudinal sec-
tion from the ventral side
(after Ludwig). The hydro-
ccel (7) has become con-
stricted posteriorly from the
left enteroccel. An out-
growth of the intestine (S) is
the first indication of the
future oesophagus of the
Asteroid.

of tube-feet. Each new pair of feet arises between the terminal tentacle and the
foot last formed.

The five outgrowths of the hydroca-1 become outwardly visible, bulging out the
body. On the left side of the seven-days-old larva there are thus visible five flat
protuberances arranged in a convex arch directed upward and backward ; these
protuberances become more marked on the eighth day, and are then divided either
into three or five lobes each (Figs. 438-440). These are the first indications of the
young Asteroid, the rudiments of its ambulacra! arms.

The rudiment of the definitive cesophagus appears in the form of a bulging of
the left side of the archenteron, that facing the hyd-roccel. This arises in the region
which corresponds with the anterior part of the gastrula intestine (Fig. 434, 8), and
has nothing to do with the larval oesophagus. This latter degenerates on the
eighth or ninth day, and the larval anus also disappears.

VIII

ECHINODERMA T A— ONTOGENY

529

The larval organ attains its greatest development on the eighth or ninth day, it
diminishes in size later,' and finally is altogether resorbed, without giving origin to
any organ of the young Asteroid. Its wall consists of three layers : (1) an outer
ciliated larval epithelium, (2) the inner epithelium of the unpaired section of the
enteroccel which fills up the whole cavity of the larval organ, and (3) between these
two layers, one of mesenchyme cells differentiated into muscle fibres. The larva uses
the organ for locomotion and for temporary attachment.

Soon after the ambulacral (oral) rudiments of the arms have appeared on the left
side in the form of the five protuberances (1-5, Fig. 437) mentioned above, five
mesenchyme thickenings form, which also bulge out the ectoderm, and represent
the antiambulacral (apical, dorsal) arm rudiments (I-Y). Three of these are

FIG. 435.— Asterina gibbosa, larva six
days old, seen from the left (after Ludwig).
I-V, The live primary bulgings of the
hydroccel (7) ; 3, the left enteroccel, opening
outward through the hydropore or dorsal
pore (11) on the dorsal side; 5, intestine;
6, anterior enteroccel, enterocoel of the
larval organ ; 9, larval mouth ; 10, mesen-
tery ; 11, dorsal pore ; 12, ectoderm of the
larval organ.

FIG. 436.— Asterina gibbosa, larva eight
days Old, seen from the dorsal and somewhat
from the left side, optical longitudinal section
(after Ludwig). II, V, second and fifth primary
bulgings of the hydroccel ; 3, left enteroccel ;
5, intestine ; 7, hydroccel ; 11, dorsal pore ;
13, oesophagus of the Asteroid ; 14, rudiment of
the stone canal.

found on the left ventral side, and two somewhat to the left of the median line, on
the dorsal side of the larva. The five stand in a curved row, the curve opening
anteriorly, the plane in which they lie making an angle with that of the ambulacral
arm rudiments.

These two sets of arm rudiments then shift towards one another, until their
planes are nearly parallel.

Appearance of the skeletal plates. — As early as the time when the hydroccel
bulgings begin to become trilobate, a small calcareous body appears in the mesen-
chyme on each side of each primary bulging, on the proximal side of the lateral
secondary bulgings (i.e. of the rudiments of the first tube-feet). These are the
rudiments of the first five pairs of ambulacral plates. When a second pair of lateral
lobes appear distally to the first pair on each hydrocoel bulging, a second pair of
VOL. II 2 M

530

COMPARATIVE ANATOMY

CHAP.

calcareous bodies form between this and the first pair ; these are the rudiments of a
second pair of ambulacral plates, and so on.

FIG. 437.— Asterina gibbosa, larva at the end
of the eighth day, from the left side (after
Ludwlg). 1-5, The ambulacral rudiments of the
arms over the primary hydroccel bulgings; I-V, the
antiambulacral rudiments of the arms.

FIG. 438.— Asterina gibbosa, larva ten
days old, seen from the left and some-
what from the ventral side (after Ludwig).
The ambulacral arm rudiments (1-5) now
have five lobes, ft, Terminal lobe, terminal
tentacle.

As early as the seventh day, the rudiments of the apical skeletal plates, eleven
in number, appear. These eleven rudiments lie superficially below the ectoderm of

ax

Fin. 430.— Asterina gibbosa, young
Asteroid, with much reduced lateral
organ (lu) at the end of the tenth day,
from the loft side (after Ludwig).
The first rudiments of the ambulacral
skeleton have appeared (five pairs of
ambulacral plates). The mouth of
the Asteroid has not yet formed.

Fni. 440.- Asterina gibbosa, young Asteroid
eleven days old, horizontal section immediately
below the oral surface (after Ludwig). 1-5, The five
five-lobed outgrowths of the hydroccel ring (an) which
has not yet closed ; ax, the two outgrowths at the two
ends of the horse-shoe-shaped hydroccel, which by
growing out towards one another and opening out
into one another close the hydrocoel ; lo, interradius of
the larval organ ; m, interradius of the madreporite.

the apical region. Five of them appear in the mesenchyme of the five apical brachial
rudiments, and become the terminals of the Asteroid arms, always remaining at the

VIII

EGHINODERMA TA—ON TOGENY

531

tips of the growing arms (Fig. 441, /j - /.,). Five others appear within the anteriorly
open curve made by the five terminals, and alternate with these latter ; these are
the primary interradials (basals) of the dorsal surface of the Asteroid disc (ba^ - &«.,).
One of these (ba^ always lies on the right near the dorsal pore, and, growing round
this pore later, becomes the madreporitic plate. The eleventh plate lies in the
centre of the two curves just mentioned, and is the rudiment of the central plate (tv).

The basals and the central appear on the right side of the larva over the right
enteroccel. The relation of the terminals to the enteroccel has not yet been
certainly ascertained. It has
been proved that in the Bipiii-
iini'iii, they appear even before
the rudiment of the five hydro-
ccel outgrowths, above the left
enteroccel.

Metamorphosis of the larva
into the young Asteroid.—
This is throughout a continuous
process. Only two parts of the
larva are not taken over by
the young Asteroid, viz. the
larval organ and the larval
oesophagus, which are gradually
resorbed. The. anus of the
Asteroid does not indeed de-
velop out of that of the larva,
but at the same point.

The last remains of the
larval organ are found on the
ventral side of the young
Asteroid lying excentrically
in that interradius in which
the hydroccel closes to form
the water vascular ring ; view-

ba3

FIG. 441.— Asterina gibbosa, young Asteroid ten days
old, dorsal view (after Ludwig). I-V, The antiambulacral
arm rudiments ; Z, interradius of the larval organ (Jo) ;min-

terradius of the madreporite (mp) ; MI - to3, the hve basals ;
^^ the five tenninals; C(,, ceiltral.

ing the body apically this interradius follows the madreporitic interradius on the
right (cf. Figs. 440, 441 ; the arrows indicate these interradii).

The mouth and oesophagus of the young Asteroid arise by the outgrowth from
the left side of the archenteron, mentioned above, reaching the body wall and finally
breaking outward through it (thirteenth or fourteenth day). The oesophagus is then
grown round by the hydroccel, which closes to form the water vascular ring. Only
shortly before this takes place does the hydroccel become entirely constricted off
from the enterocoel, and the dorsal pore comes into direct connection with the
stone canal.

The intestine widens into a sac, five radially placed outgrowths appearing in it,
directed towards the rudiments of the arms. At the point where the larval anus
formerly lay, in the interradius between the first and second apical brachial rudi-
ments, the definitive anus breaks through.

The two curves formed by the five apical (antiambulacral) and the five oral
(ambulacral) arm rudiments approach one another more and more, the zone of the
body wall which separates them (and which is, with regard to the animal, equatorial)
becoming continually narrower. Finally, the edges of the apical and those of the
oral rudiments touch to form the young Asteroid. During this process the arm
rudiments unite in the following peculiar manner: Number 1 unites with II.
•2 with III, 3 with IV, 4 with V, and 5 with I.

532 COMPARATIVE ANATOMY CHAP.

In the meantime new pairs of outwardly projecting lateral lobes (rudiments of
tube-feet) have appeared on the hydroccel bulgings, which have developed into the
radial water vascular trunks ; these new growths always appear distally to those
already formed and proximally to the median terminal lobe (terminal tentacle).

The nervous system develops as an epithelial circular cushion in the oral area,
even before the mouth has broken through its centre.

The skeleton receives the addition of fifteen new plates on the apical side outside
of the basals, five being radial and five pairs interradial.

In each interradius orally (on the thirteenth day) a plate forms between the
separate proximal pairs of ambulacral plates. These five plates are the rudiments
of the orals (odontophores).

At the sides of the ambulaeral plates the adambulacral plates appear. The
remaining pairs of ambulacral and adambulacral plates arise in the same order as
the pairs of tube-feet, always proximally to the terminals of the arms, and distally
to those already formed.

The five first and the five second pairs of ambulacral plates unite with the five
first pairs of adambulacral plates to form the oral skeleton.

The five radial outgrowths of the intestine cprickly grow into the arms, forking,
and thus producing the ten brachial diverticula of the digestive sac. Five pairs of
small interradial outgrowths on the water vascular ring represent the rudiments of
Tiedemann's bodies. None of the tube-feet at first have suckers. The formation
of the nerve ring is followed by that of the radial nerve ridges, which, like the
former, are epithelial in position, persisting as such even in the adult Asteroid. The
continuous ciliated covering of the larva is at no time interrupted, but passes
direct into the ciliated covering of the Asteroid.

We shall not enter upon the accounts given of the rise and development of the
blood vascular system, since there is nothing more problematical in the anatomy
of the adult Asteroids than this system.

Where, among Asteroids, a typical Hi-pinna ria larva is developed, the meta-
morphosis which produces the young Asteroid seems to resemble in essentials that
of Asterina. The rudiment of the young Asteroid is found in the posterior part of
the larva which contains the swollen mid-gut. At first, as in Astrrina, this
rudiment is double, i.e. it consists of an oral rudiment, arising close to the hydrocoal,
and an apical rudiment, the two uniting round the intestine. The larger anterior
portion of the larval body, together with the ciliated rings of the Bipinnaria, are,
like the larval organ of Asterina, gradually resorbed.

E. Ontogeny of the Ophiuroidea.

According to the present state of our knowledge the development of the Ophiur-
oidea does not appear to differ so greatly from that of the Asteroidea, in spite of
the difference in shape of the larva?, as to need detailed description. We shall
therefore limit ourselves to a few points.

Development of the hydro - enteroccel.— The first rudiment of the hydro-
enteroccel has not been observed with as much certainty as could be desired. In
the quite young P/nffus larva an enteroccel vesicle lies at each side of the oesophagus.
Somewhat later the larva possesses, besides this pair of vesicles, another pair of
enteroccel vesicles at the sides of the stomach-intestine, these latter having, as it
appears, been constricted off from the former. The left anterior vesicle at this
stage enters into communication with the exterior through the dorsal pore (water
pore). On the left side there now arises, between the anterior and the posterior
enteroccel vesicles, apparently by constriction from the latter, a third new vesicle,

VIII

EGHINODERMA TA—ONTOGEN Y

53*

the hydroccel vesicle (Fig. 442). This at once becomes entirely distinct, and
lengthens out anteriorly below the left anterior enteroocel vesicle. At its outer left
edge it then produces five outgrowths, the rudiments of the radial portions of the
water vascular system. Between the fourth and fifth outgrowths (reckoning from
before backward) a dorsally directed diverticulum further grows out from the
hydrocoel vesicle, which, after
a very short course, comes in
contact with the left anterior
enteroccel vesicle, and opens
into it immediately below the
aperture of the water pore.
This diverticulum is the rudi-
ment of the stone canal. Its
connection with the dorsal pore
(madreporite) is thus secondary,
and is brought about by means
of the left anterior enteroccel,
which no doubt becomes the
ampulla.

The long hydrocoal vesicle,
with its five outgrowths, then
clasps the larval oesophagus
like a halter, and grows round
it ; this larval oesophagus
apparently becomes the defini-
tive oesophagus, while no
definitive anus replaces that
of the larva.

First appearance of the
plates of the skeleton. — Soon
after the formation of the stone

post

FIG. 44-2. ^Dorsal aspect of a young Ophiuroid Pluteus,
to illustrate the hydro - entered tl (after Bury). 1, Larval
(esophagus ; 2, left anterior enterocael ; 3, hydropore ;
4, hydroccel ; 5, left posterior enterociel vesicle ; 6, stomach-
intestine ; 7, right posterior enterocoel vesicle ; 8, right

canal, ten skeletal plates appear

on the Plvtevs larva, five on

the left and five on the right anterior enterocoel vesicle.

side, i.e. above the left and

right posterior ccelomic vesicles. The five on the right side are the radials of the

apical system ; the five on the left are the terminals. In the middle of the right

side the rudiment of the central plate then appears, and on the left side, immediately

iu front of the water pore, another plate appears, which is the fifth oral, the one

which becomes the madreporitic plate. The madreporite thus belongs ontogenetically

to the oral system of plates. The other parts of the skeleton form only after the

metamorphosis.

F. Ontogeny of the Crinoidea.
The Ontogeny of Antedon alone has been investigated.

1. Embryonic Development.

Here also a ccelogastrula is formed by the invagination of a cceloblastula. The
transverse slit-like blastopore indicates the posterior end of the larva, which at a
later stage becomes bilaterally symmetrical. The segmentation cavity is filled by
an albuminiferous gelatinous mass (gelatinous nucleus).

COMPARATIVE AX ATOMY

CHAP.

After the process of imagination has begun, the formation of mesenchyme also
commences, proceeding from the blind end of the archenteron, which here is bi-
laminar. The cells of the layer which is turned to the segmentation cavity wander
into that cavity, i.e. into the gelatinous nucleus which fills it, and become mesen-
chyme cells (Fig. 443). The formation of mesenchyme proceeds actively during
the whole process of invagination along the whole archenteron. but chiefly at its
base. Here, indeed, the formation of mesenchyme is observed long after important
processes of separation and differentiation have been accomplished in other parts
of the archenteron.

The formation of mesenchyme takes place here more actively than in any other
Echinoderm in which it has been observed, so that the large segmentation cavity
soon appears to be crowded with mesenchyme cells.

The ectoderm becomes covered with cilia.

The blastopore closes completely in the course of the second stage of develop-

FIG. 443.— A. Horizontal longitudinal section through an embryo (gastmla) of Antedon
twenty-six hours old ; B. the same of one forty-eight hours old, in which the archenteron, which
has become constricted oft', is divided into two sections (af'tf r Seeliger). 1. Ectoderm ; 2, mesen-
chyme cells ; 3, place of formation of the mesenchyme cells at the base of the archenteron ; 4,
blastopore; 5, enrlodenn ; 6, archenteric cavity: 7. mesentero-hydrocoel vesicle; S, entni
vehicle.

mem. The archenteron then lies as a closed vesicle in the posterior region of
the segmentation cavity.

An important process soon takes place. The archenteric vesicle or archenteron
becomes constricted by a circular furrow (Fig. 443 IV. This constriction leads to
a complete division of the archenteron into an anterior and a posterior vesicle.
The anterior is somewhat larger than the posterior ; the formation of the mesen-
chyme continues actively on its wall (Fig. 444).

From the anterior vesicle are derived the intestine and the hydroccel ; from the
posterior the ccelom with the chambered sinus, etc. We here have a remarkable
difference between Anii'dmi and other Echinoderms, in which latter, as above
described, the anterior blind end of the archenteron always yields the ccelom.

The anterior vesicle is in close proximity to the ectoderm, on the ventral side.

The posterior vesicle becomes a transversely placed tube, whereas the anterior
is produced into a horn, both dorsally and ventrally. These two horns clasp the
posterior vesicle from its anterior side. The larva is now distinctly bilaterally
symmetrical (Fig. 444).

The next changes to occur arc the following :—

VIII

ECHINODERMA TA—ONTOGEN Y

535

The two horns of the anterior vesicle grow out towards one another round the
posterior vesicle until they touch, and so form a hollow ring surrounding the
posterior vesicle, but not closed posteriorly.

The posterior vesicle (enteroccel vesicle) assumes the shape of a dumb-bell, its
two lateral parts swelling up, while the transverse connecting piece becomes
narrower. It is this connecting piece which
becomes encircled bjr the anterior vesicle
(Fig. 444).

The ectoderm thickens on the ventral side.

The germ, which till now is approximately
spherical, begins to lengthen from before back-
ward (in the direction of the principal axis).

The anterior vesicle forms a large ventrally
directed outgrowth, the first rudiment of the
hydroccel (Fig. 445, 3). A small outgrowth
of its anterior wall is the rudiment of a sinus,
which has been called by some the parietal
cavity, and by others the anterior enteroccel (2).
The circular anterior vesicle itself becomes the
intestine (5, 7).

In the posterior (enteroccel) vesicle the two
lateral swellings increase in size, while ;the
connecting piece becomes thinner and thinner,
and finally, at a later stage, entirely disappears.
The enteroccel vesicle is thus divided into a
right and a left enteroccel sac.

During the next period, which more or less
corresponds with the fourth day of develop-
ment, the embryo increases somewhat more in
length. Anteriorly, in the frontal region, i.e.

FIG. 444. — Horizontal longitudinal
section through an embryo of Antedon.
fifty - seven hours old (after Seeliger).
1, Point at which the neural plate becomes
differentiated ; 2, ectoderm ; 3, mesen-
cliyme ; 4, place of formation of the
niescnohynie ; 5, rudiment of the intes-
tine ; 6, rudiment of the left ccelom ;
7, ventral outgrowth of the mesentero-
hydrocoel vesicle ; 8, rudiment of the right
9, transverse duct, connecting the

at the end of the embryo diametrically opposite C(Eiom

to the point where the now vanished blastopore two rudiments of the

lay, a ciliated tuft forms. Ciliated rings

appear in the arrangement characteristic of the free-swimming larva (cf. Fig. 402,

p. 510).

The ectoderm which carries the neural tuft thickens (neural plate), becomes
multilaminar, and at the same time appears to be slightly depressed (Figs. 446 and
447). The deep cells become ganglion cells, and nerve fibrillse also appear below
the surface, closely applied to the neural plate and formed by the ectoderm ; these
are the rudiments of the larval nervous system.

Ventrally from the neural plate, close behind it in the median line, a pit-like
depression forms ; this is the adhesive pit, so called because, at a later stage, the
free-swimming larva attaches itself by means of it.

Another depression, which rapidly deepens and increases in circumference, lies
in the thickened ventral ectoderm, and is the rudiment of the vestibule, whose signi-
ficance will be explained later.

The two ccelom sacs have become completely detached, the connecting piece
having disappeared. That on the right spreads chiefly dorsally forward into the
segmentation cavity and then over the intestine, even crossing the median line on to
the left side. The left c«elom sac, however, spreads chiefly backward and grows
round the intestine posteriorly like a cap, until it touches the posterior wall of the
right sac. Dorsally it touches the latter somewhat to the left of the median line,
and a mesentery is thus formed which runs dorsally somewhat to the left of the

536

COMPARATIVE ANATOMY

CHAP.

median line, but on the posterior side shifts the more to the right the nearer it
approaches the ventral side. This is the principal mesentery. Ventrally the two
ccelom sacs still remain far apart.

In the anterior vesicle, the hydrocoel rudiment, together with the rudiment of
the parietal sinus, becomes separated from the rudiment of the intestine. After
this separation has taken place, the hydroccel vesicle still remains for a short time
in open communication with the parietal sinus.

The hydroccel vesicle lies close below the thickened ventral ectoderm, shifted
somewhat far to the left.

The rudiment of the parietal sinus becomes a transverse tube.
The rudiment of the intestine changes shape. It is no longer, as before, an
incomplete hollow ring, placed vertically, through which the connecting piece of

the two crtlom vesicles passed (Fig. 445).
The connecting piece having degenerated,
the lumen of the tubular ring has room
to expand from before backward, until
the hollow ring becomes a vesicle.

Towards the close of embryonic
development, in the fifth stage, the first
rudiment of the calcareous skeleton
appears. In an embryo one hundred
hours old, the rudiments of the following
plates were found : 5 orals, 5 basals, 3-5
infrabasals, and about 11 segments of
the stalk.

The five orals have a superficial
position at the posterior part of the

JlOft

PIG. 445.— Posterior end of an embryo of
Antedon sixty hours old, seen from the right
side (after Seeliger). 1, The outline of the right
ccelom sac ; 2, rudiment of the parietal sinus ;

3, rudiment of the hydroecel ; 4, mesenchyme ; embryo, making a horse - shoe - shaped
5, ventral, and 7, dorsal process of the mesentero- arch which i(j anteriorly and down-

hydrocoel vesicle ; 6, connecting duct between the

right and the left coelom sacs. , wards- The left end of the arch reaches

further forward than the right.

As a rule (with the exception of the first oral, which indicates the end of the
left side of the arch) the five orals lie round the left ea4om sac.

The five basals have exactly the same arrangement as the five orals, merely lying
somewhat further forward than the latter. All of them, except the first basal, lie
above the right ccelomic vesicle.

The 3-5 infrabasals again, which are still extremely small, lie in front of the
basals, but further below the surface of the embryo.

In the anterior half of the embryo the infrabasals are followed by the row of
stalk plates. This row forms an arch which is concave towards the ventral surface,
so that the most anterior, the pedal, plate, lies near the floor of the vestibular
invagination.

The newly-arising skeletal plates of the stalk appear at the posterior end of the
row, generally (but not exclusively) immediately in front of the future centrodorsal,
between it and the last formed most posterior stalk plate.

Up to this time the embryo has lain enclosed in the egg-membrane, on the pin-
nulse of the mother, but it is now ready to be hatched. Even at this stage its
organisation leads to the conjecture that the calyx will be produced from the larger
posterior half, which alone contains the internal rudiments, while the stalk of the
future attached larva will be produced from the anterior half.

vii r

EGHINODERMA TA— ONTOGENY

537

2. The Free-swimming Larva (Figs. 402 (p. 510), 446, 447, 448).

The external form of the free-swimming larva has already been described on
p. 510.

The duration of this manner of life differs greatly in individuals of the same
brood, varying from a few hours to several days.

Ectoderm. — For the ciliated bands, see above, p. 510.

In the intermediate zones, which are free from cilia, a fine cuticle becomes
differentiated from the ectodermal epithelium, the cells of which later begin to
secrete a homogeneous substance which separates cell from cell, so that the epi-
thelium conies to resemble a connective tissue.

The neural plate, the neural tuft, and with them the larval nervous system attain
their highest development at this stage, but undergo complete degeneration in the
next. The ganglion cells below the neural plate
increase in number, and the layer of nerve fibres
spreads out over the whole anterior end of the larva.
Fine nerve trunks run to the ciliated rings, and two
specially strong ventral nerve trunks run back at
the sides of the vestibular invagination, their an-
terior parts being beset with isolated ganglion cells.

The adhesive pit becomes larger and deeper,
and towards the end of this period loses its ciliation
and assumes a glandular character.

The vestibular invagination spreads over the
greater part of the ventral side. It closes and
becomes a tube, the lateral edges of the invagination
growing towards one another and fusing in the
median line. This process takes place from behind
forwards, and is not fully completed during this
period, a small aperture being retained anteriorly.
The ciliation of the vestibule disappears.

The intestine alters its shape. It spreads out
somewhat, assuming first the form of a hollow plate,
with the concavity directed ventrally and the con-
vexity dorsally. In the ventral concavity lies the
hydrocu?! vesicle. At a later stage the intestine
again becomes rounded and vesicular.

FIG. 44(1. -Free-swimming larva
of Antedon, fifty-two hours old,

from the left side (after Seeliger).
I-V, The five ciliated rings ; 1, the
parietal sinus ; 2, the vestibule,
already closed posteriorly ; 3, the
hydrocoel ; 4, the hydropore ; 5, left
enterocoel vesicle ; 6, intestine ;
7, right enteroccel.

The two enteroccel sacs continue to change their
positions and to spread out. The right vesicle pro-
duces anteriorly five tubular outgrowths, which become grouped round the principal
axis. These five tubes arise, widened like funnels, from the right ccelom, then
narrow anteriorly, and, losing their lumina, run out as strands. They are the
rudiments of the chambered sinus.

The skeletal pieces of the stalk are at this time horse-shoe-shaped, and tend to
enclose the five tubes of the chambered sinus. When they become complete rings
the chambered sinus passes through them.

The hydrocoel vesicle becomes completely constricted from the parietal sinus,
flattens in the dorsoventral direction, and at once assumes the horse-shoe shape.
The gape of the horse-shoe points at first backwards and to the left, and finally to
the left and forwards. Five ventrally directed outgrowths appear on it, out of each
of which, at a later stage, three tentacle vessels arise.

538

COMPARATIVE ANATOMY

CHAP.

The rudiment of the primary stone canal appears at the blind end of the left
limb of the hydroccel, as a dorsal process, running to the left.

The tubular parietal sinus, which is now completely isolated from the hydroccel,

Fiu. 447.— Median longitudinal section
through a free-swimming larva of An-
tedon, twenty -eight hours old, in the
act of becoming attached (after Seeliger).
I-V, The ciliated rings ; 1, neural plate with
nerve fibres (2), and ganglion cells (3) ;

4, gelatinous nucleus, the mesenchyme
cells which crowd it are not represented :

5, the tubes of the chambered organ ; 6, in-
testine ; 7, right coelom ; 8, vestibule ;
9, parietal sinus ; 10, right enteroccel ;
11, hydroccel ; 12, adhesive pit ; 13, left
enteroccel.

on.

FIG. 448.— Free-swimming larva of
Antedon, forty -eight hours after
being hatched, from the left si. IP.
specially to illustrate the rise of the
skeletal plates. 1-V, The ciliated rin^s :
liai-ba-,, the live basals ; »/•]-'"'.•;, the live
orals, those lying on the right side re-
presented as discs ; 1, vestibule ; 2, in-
testinal vesicle ; 3, right enterocoel ;

4, calcareous joints of the stalk :

5, pedal plate.

has shifted to a position in front of and above the latter. Its posterior end grows
out till it touches the ectoderm immediately in front of the fourth ciliated ring
ventrally and to the left, and finally breaks out through the hydropore at this point.

3. Attachment of the Larva and its Transformation into the Stalked Form

(Figs. 449-453).

Attachment takes place by means of the adhesive pit, which yields a sticky
secretion ; and since this pit lies ventrally at the anterior part of the body, the
attached larva has at first a position parallel to the surface to which it is fastened,
and the vestibule lies immediately above that surface. The body, however, soon
becomes erect, and the adhesive pit takes up a terminal position.

Very soon after attachment the ciliated rings disappear, and so does the

VIII

ECHINODERMA TA— ONTOGENY

539

neural tuft; the neural plate flattens out, and the larval nervous system
completely degenerates.

The ectoderm cells continue to yield an intermediate substance. Many of them
sink below the surface, the consequence being that the distinction between the
body epithelium and the mesenchymatous cutis is entirely obliterated.

The vestibule becomes completely constricted off, the last remains of the aperture
of invagination closing. At the same time it shifts entirely to the posterior end
of the larva, the end which now freely projects, twisting through an angle of 90°, so
that its thickened epithelial floor, which before lay parallel to the principal axis, now

r-3

d

ba.

FIG. 440.— Young attached larva of An-
tedon, forty-eight hours old, from the left
side (after Seeliger). The vestibule 1ms be-
come entirely constricted otr, but the distinc-
tion between the calyx and the stalk is not yet
pronounced. Imi-bn^, Basals ; orror3, orals ;
•Hi, infrabasals ; 1, pedal plate ; 2, parietal
canal ; 3, hydrocoel outgrowths ; 4, vestibule ;
5, intestinal vesicle ; G, left cuelom sac ; 7, right
cttlom sac ; 8, calcareous joints of the stalk.

"bets

01*2

post

Fi<;. 4"iii. — Young attached larva of
Antedon, forty -eight hours after being
hatched, from the left side (after Seeliger).
co, Stalk; ca, calyx; bai-ba3, basals; nr]-'"':;-
orals ; il>, infrabasals of the. left side ; 1, pedal
plate ; 2, hydropore ; 3, left coelom ; 4, right
crelom ; 5, joints of the stalk.

lies at right angles to it. The larval body becomes club-shaped, the anterior body
forming the handle of the club. The vestibule, which continues to increase in size,
comprises the entire posterior part of the club (the calyx) ; it becomes pentagonal.
and imprints the same shape upon the whole posterior part of the bod}7, and thus
first determines the radiate structure (Figs. 450, 451, and 452).

The anterior end of the larva becomes the apical end of the stalk, the pos-
terior end becoming the oral side of the calyx of the attached Pentacrinus-like
larva.

The hydrocoel undergoes the same twisting and shifting as the vestibule, beneath

540

COMPARATIVE ANATOMY

CHAP.

whose floor it lies after as before the process. It has passed from the horse-shoe
shape to the circular, but the hydroccel ring still remains unclosed for a long time
at the point where the gape of the horse-shoe formerly was. Its five outgrowths

ant

FIG. 4'.!.— Stalked larva of Antedon,
eighty-four hours old, with twenty-live
tentacles, from the right side (after
Seeliger). Calcareous plates not repre-
sented. 1, Right coelom sac ; 2, stomach
intestine ; 3, left coelom sac ; 4, saccnli ;
"'. M'stihule, still closed; 6, the fifteen
primary tentacles; 7, the five pairs uf
secondary interradial tentacles ; s, oeso-
phagus ; '.I, hind-gut ; 10, axial organ ;
11, fibrous strands in the stalk, continua-
tions of the axial sinus.

FIG. 4.V_'.— Transverse section
through the region of the left
or oral coelom of an attached
larva of Antedon, 108 hours
old (after Seeliger). I-V, The
five radii ; 1, left oral coelom ;
2, oesophagus ; 3, stone canal ;
4, parietal canal.

I-'H;. 453.— Diagrammatic
transverse section through the
region of the aboral ccelom in
an Antedon larva. 108 hours old
(after Seeliger). I-V, The five
radii; hf'i-?«ig, the five basals ;
1, right or aboral coelom ; -2, hind-
gut ; 3, axial organ; 4, parietal
sinus ; 5, oesophagus.

push up the ectoderm of the floor of the vestibule into the vestibular cavity ; they
soon appear to be trilobate, so that in all 5 x 3 tentacles are present, ten more being
added to them, which arise in pairs at the bases of the primary outgrowths.

viii ECHINODERMATA— ONTOGENY f>41

The stone canal breaks through into the parietal sinus. — The point at which
this occurs does not, however, correspond with the point at which the hydroccel and
the parietal sinus originally were in open communication.

The parietal sinus also takes part in the shifting just mentioned. In the free-
swimming larva it lay in front of the hydroca-1. This position it retains, while
shifting backward (towards the oral end) together with the hydroccel. It thus
approaches its external pore. Compared with the other growing organs, it remains
small and stationary.

The hydrocoel is thus connected by the stone canal with the parietal sinus, and
this latter is in open communication with the exterior through the hydropore.

The intestine. — An extraordinary process goes on in the intestinal vesicle.
Numerous cells become detached from its wall, and wander into its cavity, which
they finally completely fill. They fuse for the most part into a large yolk-like mass,
which is entirely resorbed at a later stage as nutritive material.

The floor of the vestibule deepens at the centre, and is produced anteriorly like a
funnel towards the intestinal vesicle. This funnel, which passes through the hydro-
ccel ring, becomes the oesophagus, and joins a posterior process of the intestinal
vesicle which grows out to meet it.

The intestinal vesicle divides into a stomachal section to the left of the larva
and a narrower portion running dorsoventrally on the right side. This latter part,
the hind-gut, rises with a broad base out of the former and ends blindly. The blind
end of the hind-gut then grows over to the left ventrally.

The ccelom. — The two ccelom sacs shift and spread out in a peculiar manner.
The left sac shifts quite posteriorly, and becomes the oral ccelom, which grows round
the oesophagus on all sides from above downward, thus assuming the shape of a
hollow horse-shoe which clasps the oesophagus, the stone canal, and the parietal
canal (counting these in order from within outward). The gape of this horse-shoe
is directed ventrally to the left, and since the tips of its two arms grow towards one
another, a short longitudinal accessory mesentery arises. The right ccelom vesicle
passes through changes of form, expansion, and shifting which are difficult to describe
briefly, and becomes the aboral or apical ccelom. In consequence of the shiftin^s of
these vesicles the longitudinal principal mesentery which separated the originally
right from the originally left ccelom vesicle becomes a transverse mesentery,
separating the oral from the apical ccelom, and surrounding the oesophagus like a
diaphragm. Near the right (now aboral) ccelom also, a longitudinal, somewhat
diagonal accessory mesentery develops, which runs somewhat to the right of the
ventral median line. The walls of the apical (originally right) ccelom are continued
anteriorly into the walls of the five tubes which form the chambered sinus, but at
the point where they pass into one another they are so pressed together that no
open communication exists between the two sinuses.

The axial organ (genital stolon) of the calyx arises as a thickening in the left
epithelial wall of the longitudinal accessory mesentery of the aboral ccelom, at its
most anterior (apical) end, where the chambered organ commences. As the genital
strands of the arms and pinnulae most probably arise as outgrowths of the axial
organ, it might thus be proved that in the Crinoids also the genital cells are derived
in the last instance from the endothelium of the body cavity. The cushion-like
thickening increases in length, becoming partly constricted from the mesentery ;
posteriorly, it reaches to the oral ccelom ; anteriorly, the axial organ passes into the
stalk up the centre, between the five tubes of the chambered sinus.

The formation of trabeculse begins in the aboral ccelom. Single endothelial
cells lengthen and project like pillars into the ccelomic cavity. A similar process
takes place in the hydroccel.

The skeleton. — When the vestibule shifts to the posterior end of the larva the

542 COMPARATIVE ANATOMY CHAP.

c

skeletal plates are also shifted. The horse-shoes formed by the tive orals and ^
five basals close, and form two flat rings or circles. The circle of the orals shifts
backward and on to the roof of the vestibule in such a way that the five plates together
form a pyramid, the truncated tip of which forms the centre of the vestibular roof, at
the extreme posterior end of the larva. The orals have thus shifted away from the
region of the left (oral) ccelom.

The circle of basals which lies in front of (apically to) that of the orals forms a
pyramid in the body wall of the calyx, around the aboral coelom, the truncated end
of this pyramid lying at the commencement of the stalk, or at the anterior (apical)
end of the calyx. The orals and the basals together form a pentagonal double
pyramid, truncated at both ends. At the truncated end of the basal pyramid, round
the uppermost (most posterior) joints of the stalk, lie the four or five small infra-
basals. The number of joints in the stalk increases, and the anterior body of the
larva becomes, as the stalk, more and more distinctly demarcated from, the posterior
body or calyx, which has now become five-rayed.

The orals and basals alternate with the primary outgrowths of the hydroccel.
i.e. they are interradially placed. If we indicate that primary outgrowth of the
hydroccel which, when the larva is viewed from the oral side, comes next in the
direction followed by the hands of the clock, to the hydropore (which lies ventrally
to the right) as No. I, and those which follow in this direction a.s II, III, IV, and
V, and if, again, we indicate that oral or basal which lies in the interradius between
radii I and V as the first, and those which follow in this direction as orals (or basals)
2 to 5, it can be proved that the hydropore, in the older stages of the attached larva,
is enclosed by the basal part of the first oral plate. In these stages it is also seen
that the infrabasals fuse to form a single plate, the centrodorsal, at the centre of
which there is an aperture for the passage of the chambered organ.

Daring the first developmental periods which follow the attachment of the larva,
the sacculi appear. Five of these first arise exactly radially at the bases of the
middle tentacle of each group of tentacles, on the outer side of the circular canal.
These sacculi can be ontogenetically derived from groups of mesenchyme cells.

4. The Stalked Larva after the Vestibule has been Perforated.

(From five days to the sixth week after hatching, Fig. 454.)

The calyx becomes more and more distinct from the stalk.

The roof of the vestibule becomes ever thinner at its centre, an aperture finally
forming. Radial incisions run from this central aperture towards the peripheral base
of the roof, so that this latter becomes divided into five interradial lobes or valves,
each of which contains an oral plate. This pyramid of valves can open and shut.
The vestibule has opened outward.

The five tentacles meanwhile lengthen and receive their papilla?. They usually
project outwards from between the five oral valves.

The definitive nervous system (which is oral and superficial) rises quite inde-
pendently of the larval nervous system, which entirely disappears. The first ap-
pearance of the nerve ring was observed very late, long after the perforation of the
vestibule. The ectoderm of the oral disc, i.e. the peripheral portion of the original
vestibular floor (the central part having sunk in to form the oesophagus), becomes
thickened in a ring which is bordered by tentacles, and here becomes multilaminar.
The cells of the deeper layer yield the nerve tissue.

Neither the rudiments of the deeper oral nervous system nor those of the apical
system have as yet been certainly observed.

V11I

ECHINODERMA TA— ONTOGENY

543

ant

Alimentary canal. — The mouth from the very first does not lie exactly at the
centre of the oral disc, but somewhat eccentrically in the interradial area bordered
by the first and fifth radii.

The stomach becomes a capacious sac, and the yolk-like mass contained in it is
gradually absorbed. The hind-gut arises out of it (in interradius III-V), being
broad at the base, and then thinning into a
tube which has the following course. Viewing
the larva from the oral pole, the hind-gut runs
(in the direction of the hands of the clock) in
the horizontal mesentery, near the body wall,
through the interradial space IV-V, then runs
across radius V, and immediately after opens
outwards in the interradial space V-I, through
the anus which has in the meantime broken
through the calyx laterally. This is the same
interradius in which the hydropore lies, on
the original ventral side of the bilaterally
symmetrical larva. The ectoderm takes no
part in the formation of the anus.

In the ccelom sacs profound changes are
going ou, which may be briefly summarised
as follows.

(a) The chambered sinus gives up all
connection with the original right, now the
aboral ccelom sac.

(b) The mesenteries (both the principal
and the longitudinal accessory mesentery) are
completely resorbed, and as a consequence
the right and left cosloms unite to form one
large body cavity.

(c) The trabeculse (of endothelial origin)
become very strongly developed, and traverse
the body cavity in all directions as a network.

(/I) The axial organ becomes differentiated
as an independent solid cell strand, lengthens
till it reaches the tegmen calycis (oral disc),
and at a later stage becomes hollow.

(c) In the parietal sinus, which comes to
lie quite in the body cavity, two sections
become more and more distinct ; the one
vesicular, and the other a narrow canal-like

Fio. 4.34.— Calyx of a decalcified larva
of Antedon, five weeks old, with extended
tentacles, from the left and lower side (after
Seeliger). I-V, The live radially placed
primary sacculi ; 1, axial organ ; 2, right
(aboral) coelom ; 3, principal mesentery,
between the right (aboral) and the left (oral)
coelom; 4, hind -gut which follows the
stomach ; 5, oral ca'lom ; 6, hydroeael ring ;

section opening outwards through the hydro- 7, two of the ten secondary tentacles ;

pore. The former, into which the primary 8, tentacle papilla- ; !>, primary tentacles

stone canal enters, loses its independent ("nly seven of tllB total "u"^er, fifteen, are

,,-,,. i ii j.1 • IT i-i represented); 10, oral lobes ; 11, stone canal ;

endothehum, and the thin wall which

' 12, hydropore; 13, oesophagus ; 14, stomach ;

separates it from the ccelom also probably 15> continuation of the chambered organ in

disappears, so that it ceases to exist as a the stalk (Hi).

separate cavity. The stone canal now opens

into the general body cavity, which is thus in communication with the exterior

through the narrower section of the original parietal sinus, and through the

hydropore in the anal interradius.

In the hydroccel, the water vascular ring completely closes. The whole of the
musculature of the hydroccel is formed by the hydroccelomic epithelium itself. The

544 COMPARATIVE ANATOMY CHAP.

trabeculce within the system of canals increase in number. In the tentacles, the
following changes have taken place. Formerly the twenty-five tentacles were
arranged in five radial groups of five each. The five tentacle canals of a group were
connected by a common tentacle canal rising with the circular canal. Now all the
five tentacle canals of a group rise separately from the water vascular ring.

Further, during the period of the attachment of the stalked larva, four new stone
canals appear, and four new calyx pores form in the other interradii. These and
all that arise later cannot, of course, form in the same way as the primary calyx
pore.

The stage, the development of which lias just been described, has been called
the Cystid stage, owing to the absence of arms, to there being no clear division
of the calyx into dorsal cup and tegmen calycis, and to the occurrence of the
rudiment of the genital organs as an axial organ, whereas later the genital glands lie
in the arms and especially in the pinnules.

In opposition to the above view it may be remarked :

1. That neither the want of arms nor the absence of division of the calyx into a
dorsal cup and an ambulacra! disc is characteristic of the Cystidea.

2. That the skeletal system of the attached larva of Comatula is altogether
radiate, consisting of the three circles, the radials, the basals, and the infrabasals.
On the other hand, the irregular arrangement of the skeletal plates is, as a rule,
characteristic of the Cystidea. Those Cyst ids which most resemble the larva of
Comatula in the number and radial arrangement of the skeletal plates are also those
which are, of all Cystids, the most nearly related to the Grinoidea.

3. The hydrocoel of the stalked Comatula larva consists simply of the water
vascular ring and a circle of tentacles, which receive their canal direct from the water
vascular ring. In the Cystidea, radial canals must have run out from the water
vascular ring, below the food grooves of the ambulacra, giving off tentacle canals to
right and left, and also probably penetrating into the arms.

4. The appearance of the first rudiments of the genital glands in the body merely
proves that the definitive position in the arms is secondary, and this applies to all
Echinoderms.

The position of the anus, indeed, agrees in both.

5. Last Stage of the Attached Stalked Larva— Pentacrinus Stage.
(Of. Fig. 326, p. 375.)

This stage is distinguished by the rise of the arms, which begin to grow out in
the radii between the circle of the orals and that of the basals. Each rudiment of
an arm is, from the very first, supported on its apical side by a newly arising skeletal
plate. These plates are the five radials of the dorsal cup. Distally from each
radial on the growing arm, two new skeletal plates follow one another, the first and
second costals. The growing rudiment of the arm then forks, the distiehals form,
and so on.

During the formation of the arms the five middle and strictly radially arranged
tentacle canals of the five tentacle groups become the radial vessels, which fork with
the arms. Fresh investigation of this point is, however, much needed.

The interval between the oral pyramid and the bases of the arms increases, and
the tegmen calycis thus comes into existence. The pyramid of five oral valves in
the middle of the latter does not grow further, and the valves with their skeletal
plates finally disappear. Round the anus, which comes to lie in the tegmen calycis,
an anal plate develops temporarily.

At this stage the resemblance of the attached and stalked larva of Antedon

vin ECHINODERMATA— PHYLOGENY 545

to the Inadunata, especially to the so-called Larviformia (<•/. pp. 303, 328, etc. ),
is so striking as to be at once recognisable.

The calyx, with the arms, sooner or later breaks away from the stalk, and
can either move by using the arms as paddles or catch on to objects by means of its
cirri. When it breaks loose from the stalk, some of the uppermost whorl joints on
which cirri have formed remain connected with it ; these fuse with one another, and
with the centrodorsal. The basals, again, fuse to form a rosette, which is soon over-
grown on all sides by the large apical centrodorsal plate.

XXII. Phylogeny.

No other phylum of the animal kingdom is so sharply marked off from all others
as the Echiuoderms. Their organisation is in all points strange ; even the radiate
.structure is strange, in so far as it is, unlike that of many Ccelcntcrata, only a mask
which hides a complicated and hitherto inexplicable asymmetry. We are not in a
position to compare an adult Echinoderm with the adult representative of any other
phylum from a phylogenetic standpoint.

The difficulties which meet us in attempting to reconstruct the phylogenesis of
the Echinodermata are still further increased by the fact that the typical charac-
teristic Echinoderm larva cannot at any stage of its development be compared with
the adult or larval form of any other animal. An exception to this statement may,
however, perhaps, be made in favour of the Enter ojmeusta, which will be described in
the next chapter.

If, taking the gastrsea theory as a foundation, we assume for the Metazoa a
common bilaminar racial form, it seems, in view of the above-mentioned difficulties,
that the racial form of the Echinodermata must have branched off extraordinarily
early, perhaps at a stage corresponding phylogenetically with the gastrula. By such
an assumption, the Echinoderms and their larva; would be removed from the sphere
of comparative anatomy and comparative embryology, except in so far as such com-
parative enquiry were limited to the Echinoderm phylum itself.

It appears to us, however, that attempts to approximate the Echinodermata to
Metazoa standing higher than the CVelenterata should not be abandoned. Recent
anatomical and ontogenetic researches have brought to light facts which open up new
prospects. We may mention the demonstration of a neural plate and of a larval
nervous system, the attempts to demonstrate that there are two pairs of enteroccel
vesicles, the proof that the first rudiments of the gonads proceed from the endothe-
lium of the cculom, the suggestion that the stone canal or the hydropore should be
regarded as a nephridial canal, etc.

All this, of course, does not justify us in closely comparing the Echinoderm larva
with other definite forms, adult or larval, belonging to Metazoan classes higher than
the Ccelenterata, except perhaps the Enteropneusta. But these discoveries and new
views tend to make the Echinoderm body appear somewhat less strange, since we
find in its organisation important points in which it is fundamentally in agreement
with the so-called Triploblastica.

It cannot be doubted, and has never been doubted, that the Echinodermata form
a distinct, naturally marked out phylum of the animal kingdom, or, in the language
of Phylogeuy, that all Echinoderms have had a common racial form.

AVithin the phylum of the Echinodermata, further, the classes are again quite dis-
tinct and naturally marked off from one another. Among known Echinoderms,
there are no intermediate forms between the Pdmatozoa, the Holotkurioulea, the
EcMnoidea, the Astcroidca, and the Opkiuroidoa. Every known Echinoderm can at
once be recognised as either an Echinoid, an Asteroid, a Holothurid, etc. The Cystidea
VOL. II 2 N

546 COMPARATIVE ANATOMY CHAP.

alone, perhaps, form an exception to this rule, "showing decided resemblance to
the Criiwids on the one hand, with an occasional possible approach to the opposite
extreme, i.e. the Holotliurioidea, on the other. It is, however, very difficult to judge
of the Cystidcii; since conclusions as to the inner organisation drawn exclusively
from the structure of the skeleton cannot be regarded as altogether trustworthy.

It appears to us that there is not the least justification for deducing the different
Echinoderm classes in any definite way from one another, nor can we at all accept
the recently urged view that the Holotliurioidea stand nearest to the racial form.
On the contrary, the morphology of the genital organs leads us to believe that the
Holot Imriiiidea are distinct from all other Echinoderms, with the possible exception
of the Cyst idea.

If we review the whole morphology of the Echinodermata, our phylogenetic specu-
lations are, first of all, influenced by the fundamental fact that the radiate, but at
the same time asymmetrical Echinoderm proceeds ontogenetically from a bi-
laterally symmetrical larva, the so-called Dipleurula.

The Dipleurula Larva.

This larva is regarded from two opposite points of view. (1) The bilateral
structure is thought to have been secondarily acquired, within the different groups
of the Echinoderms, in adaptation to the free-swimming manner of life. (2) The
bilateral structure of the larva has been inherited from the common racial form
of the Echinodermata, or from the larva of such a form. The first view is now
generally abandoned. The manner of life might indeed have called forth external
bilateral symmetry of form, but certainly not the marked bilateral symmetry of
structure of the internal organs.

If we now try to sketch a hypothetical phylogenetic stage based upon a com-
parison of the various Dipleurula lame of the Echinodermata, the following is the
result : The body was freely movable, ovoid, and bilaterally symmetrical ; the mouth
lay anteriorly on the ventral side, the anus at the posterior end, or posteriorly on
the ventral side. In the frontal region there was a nerve centre below the surface
of the ectodermal epithelium which was differentiated into a sensory organ (neural
plate). Running back from the nerve centre along the ventral side, below the surface
of the body epithelium, were two nerve trunks beset with ganglion cells. The intes-
tine was divided into the ectodermal(?) (esophagus, the wider endodermal mid-gut,
and the hind-gut, which was also endodermal. At the sides of the intestine were
two pairs of cielomic vesicles, the anterior pairs at the sides of the tesophagus, the
posterior at the sides of the mid- and hind-guts. The two anterior cu-lomic vesicles
(or their posterior portions) were connected with the exterior by a canal laterally or
dorsally (cf. the interesting temporary occurrence of a hydropore on the right side in
Asteroids, especially in Aster ias i-n/i/arix, p. 527). The genital products developed
out of the endothelium of the cd-lom.

Such an organisation has nothing strange about it. It has almost as much claim
to be classed with the Venues as Sayitttt has, for which latter classification, however,
not much can be said. It is further possible that the racial form possessed special
organs for locomotion, respiration, etc., about which nothing can now be positively
affirmed, since they have in all cases disappeared from the ontogeny of the Echinoderms.
It is not at all likely that the ciliated rings have any phylogenetic significance.

vin ECHINODERMATA—PHYLOGENY 547

Metamorphosis of the Dipleurula Larva.

The larva develops, through metamorphosis, into the young Echino-
derm, which, under its radiate mask, is asymmetrical. The radiate
structure is amalgamated with an asymmetrical structure.

Here again, we are not altogether without light as to the phylo-
genetic significance of this process. We agree with the majority of
modern authors in believing that a radiate structure of body arises
as a consequence of an attached manner of life. We are, therefore,
justified in assuming that the radiate Echinoderm arose from a free-
moving racial form, in adaptation to a newly-acquired, attached
manner of life.

All Eehinoderms must, therefore, once ' have been attached
animals.

If now, we wish to ascertain in what special manner attachment
took place, we unavoidably turn for an answer to this special question
to the Crinoids. These are the only Eehinoderms which, in all prob-
ability, never again gave up the attached manner of life. That the
only Crinoid about whose ontogeny we know anything, Antedon, is a
form which has actually once more become free, i.e. has, as a secondary
specialisation, given up the attached manner of life, detracts in no way
from the arguments based upon its ontogeny.

All other Eehinoderms whose ontogeny we can investigate have
long since given up the attached manner of life, and, with the excep-
tion of certain (analogous) cases among the Asteroids (e.g. Asterina),
do not an}- longer pass through an attached larval stage. Hence the
methods of development of other Eehinoderms, even when simpler
than that of the Crinoids, must in comparison with the latter be regarded,
phylogeneticully, with some suspicion.

From the developmental history of Antedon then, we learn that the attachment
of the Dipleurula larva of this animal took place by means of the ventral side of
the anterior end of the body. In a similar way the Dipleurula larva of Aster iiia
attaches itself by means of the larval organ which develops anteriorly.

Authors who have recently attacked this problem assume that attachment took
place on the right side ; making this assumption in order to explain the asymmetry
which follows. To us also this assumption appears necessary, but it should be
specially stated that the attachment took place on the right anteriorly.

When this assumption is made we must further ask : What were the changes
which the attached manner of life induced ?

It is difficult, with the embryological material we have at present, to obtain an
adequate idea of the resulting processes, and only a very tentative explanation can
be given.

Judging by analogy from the modifications which, in other parts of the animal
kingdom, result from an attached manner of life, it may be assumed that the arrange-
ments for conducting food were the first to become adapted to the new condition of
existence. The mouth left its unfavourable position and wandered along the ventral
side, first to the left, i.e. to the side which was now uppermost (being opposite to the.
point of attachment). In this shifting the cesophagus pushed the median and

548 COMPARATIVE ANATOMY CHAP.

ventral -wall of the left anterior ccelom in front of it, embedded itself to a certain
extent from without in the ccelomic vesicle, so that this vesicle surrounded it in the
shape of a horse-shoe. Round the mouth, the body wall (and with it the left anterior,
ccelomic vesicle which lay here) grew out into, five tentacles which, as in so many
attached animals, served for bringing in food, for the sense of touch, and for respira-
tion. (Compare the tentacles and the horse-shoe-shaped tentacle -carrier of the
Bnjozoa Cephalotliscns, etc.) Thus, the left anterior ccelom, which from the very
first was, like the right, connected with the exterior by means of a canal, produced
the primary horse-shoe-shaped hydroccel with the primary tentacles and the hydro-
pore (stone canal). In this way the first impulse towards the development of the
radiate structure was given. The horse-shoe finally closed to form the circular canal.

The right anterior side of the body, which was used for attachment, could be
produced into a stalk, as is the case in most Pelmatozoa. (The larval organ of
Aster ina may be a modified reminiscence of such a stalk.) The right anterior
ccelomic vesicle, which lay m this region, now serving for attachment, lost its efferent
aperture, atrophied, or became a cavity of the stalk (chambered sinus and its con-
tinuation in the Crinoids(1), ccelom of the larval organ in Aftterina(V).

The body now developed principally in the oral and tentacular region (on the left
anterior side). The posterior portion of the body with the anus near its end was
originally like a lateral outgrowth or shield on the body, which gradually subsided
and disturbed less and less the radiate appearance.

According to this view, the greater development of the anal interradius which is
found in many Crinoids, especially in palaeozoic forms, may possibly be an original
condition, in connection with which we have the occurrence of special anal plates in
the anal interradius. The anus also may originally have lain outside of the circle of
tentacles, a supposition which harmonises with its position in the Cystids and in
the ontogeny of A ittedon.

Concurrently with these changes, the left posterior ciulom, which lay nearer than
the right to the mouth, which had shifted to the left, now upper, side, grew round
the oesophagus, and forming a vertical mesentery, became the oral crelom. The
right Cftlomic vesicle, however, spread out chiefly in the lower (originally the right)
region of the body, and became (also forming a vertical mesentery) the apical ctelom.
The mesentery dividing the two (oral and apical) sections of the crelom would natur-
ally be horizontal (transverse).

In the vertical mesentery, the rudiments of the gonad (the axial organ) arose
as a ridge-like thickening and growth of the endothelium on one side ; in mature
animals, this opened outward through a genital duct and aperture in the region be-
tween the mouth and the anus.

This phyletic stage, deduced as a result of the attached manner of life, may be
called the Pentactsea.

For the protection of the body, calcareous plates developed in the mesenchyme
below the integument, at first, perhaps quite irregularly.

From the Hypothetical (unknown) Pentactaea Stage to the known

Echinoderm.

Most Echinoderms gave up the attached manner of life at a later stage. The
known case of Antcdon, in which an animal in the highest degree adapted for the
attached life resumed the free life, is specially welcome and useful in this connec-
tion.

The ancestors of the Holothurioidea were probably the first to renounce the
attached manner of life, although not as early as the Pentactea stage.

vin ECHINODERMATA—PHYLOGENY 549

The organisation of the Pentactaea would only become completely adapted to the
attached manner of life, when the number of tentacles and the surface for taking in
nourishment increased. Such increase might take place in various ways ; we have
many examples among attached animals belonging to other divisions of the animal
kingdom.

The Pentactaea may have become perfected in one direction as follows :—

The interval between the bases of the primary tentacles and the mouth increased,
while the (apical) interval between the primary tentacles and the attached pole
remained the same or decreased. By the shifting of the circle of tentacles away
from the mouth, the basal piece of each tentacle canal would be drawn out radially
below the oral body wall, and would become a radial canal, from which new lateral
tentacle canals would bud out alternately to right and left, always proximally to
the shifting primary (now terminal) tentacle, the body wall projecting in the form
of tentacles. There thus arose in each radius a double row of tentacles, which, to
speak exactly, stood in the corners of a zigzag line. Nutritive particles, descending,
would be sent on to the mouth between these rows of tentacles. This adaptation
became perfected when the floors between these rows of tentacles sank in the form of
furrows — the food grooves or ambulacral furrows ; these furrows then became pro-
vided with means of transport, in this case with cilia. The epithelial cells lining
the furrows gradually became sensory cells, and epithelial nerve ridges, — the radial
nerves, — would arise, which would meet round the mouth as the nerve ring.

This rise of the radial nervous system of the Echinodermata as a natural
development of the ambulacral furrows, and of the palisades of sensory tentacles border-
ing these furrows on each side, may be without difficulty followed out in detail,
although there is no space for such an attempt in this volume.

At this stage are found the armless Cystids with their carapace of plates.
The genital organ is enclosed in the body, and opens outward through a single
aperture (the " third " aperture of the Cystideo).

It was, perhaps, at a similar phyletic stage that the ancestors of the Holotlmri-
oidca gave up the attached manner of life. For locomotion, they used the tentacles,
arranged in five meridional double rows, the body elongated, and the anus shifted
to the apical end, which became free. Food was taken in directly at the mouth ; to
assist in alimentation, the oesophagus became modified into a pharynx, and the ten-
tacles lying close to the mouth became specialised. The food grooves lost their
function, leaving, however, behind them the epithelial nerve ridges, which continued
to innervate the tube-feet. The food grooves could now close over to form a tube
for the protection of the nerve ridges ; these latter became the subepithelial radial
nerves, and the lumina of the closed tubes the epineural canals. According to this
view, therefore (and this applies also to the Opliiuroidca and Evliinoidca), the sub-
epithelial radial nerves, with their epineural canals, are the original food grooves
closed over to form tubes. They gave up their original function as " food grooves "
in proportion as, with the adoption of a free manner of life, food was taken in direct
at the mouth.

It is a question of subsidiary importance, in the derivation of the Holothurid
body, which still possesses the single genital organ and the single aperture, whether
the condition of its skeleton is to be regarded as original, or whether it has not rather
been derived from the carapace of plates of a Cystid-like animal by means of the
multiplication of the skeletal pieces, their loose arrangement, and their decrease
in size.

The longitudinal and circular musculature may be new, but may just as well have
been inherited from an attached ancestral stage, in which they could be functional;
just as are the longitudinal and circular muscles of Actinia.

After the foregoing description it is obvious that the Paradinopoda (Synaptidce)

550 COMPARATIVE ANATOMY CHAP.

cannot be regarded as primitive forms of Holothurioidca. They are, on the contrary,
highly specialised forms, which, in adaptation to the limicolous life, have lost the
tube-feet and the radial canals, which, however, still occur ontogenetically.

We have thus seen how the Pentaetsea might become modified in the direction
of certain Cystids and of the Holotlmrioidea.

By remaining attached, the Pentacta?a might develop in another direction.

The body carried by the stalk might remain small, but become drawn out into
processes or arms in the directions in which the primary tentacles travel from the
mouth, the ends of these arms being always marked by the possession of the primary
tentacles. Secondary tentacles then rose out of the radial vessels which ran along
the arms (the tentacle canals of the primary tentacles) in the way above described ;
the food grooves and their nerve ridges also formed in the same way. A still more
complete adaptation to the attached manner of life was attained by the branching of
the arms and the formation of pinnula?. In this way the surface for capturing food
was continually increased.

The direction of adaptation here indicated might be called the Crinoid direction,
the Crinoids having, in fact, gone furthest in this direction.

The development of the crown of arms on a body which remained small had
necessary consequences. The body (calyx, disc) and the stalk (should this latter
develop) would have to gain the necessary stability for carrying the growing arms.
This was provided for by the formation of the more or less firm carapace of plates.
The stalk attained firmness by the development of joints ; the calyx, by that of the
dorsal cup, and here all the facts seem to indicate that in the racial form of
the Crinoids, the dorsal cup had a definite composition, viz. five infrabasals, five
basals, five radials arranged in the typical manner, and the anals. For the
protection of the mouth, five orals were added, forming together a pyramid which
could be opened and closed. For the support of the arms, and in connection with
the developing capacity for unfolding and closing the crown of arms, the jointed
brachial skeleton formed.

As the arms grew out from the small body, the ccelom was produced into them,
and processes of the single rudiment of the gonad (the axial organ) spread in one
way or another into them. They became fertile more or less far from the calyx (or
disc) and yielded the gonadial bundles, each of which opened outwards through one
or more special apertures.

In this point also, the Crinoids are the most extreme forms.

The Echinoidca, Ophiuroidea, and Asteroidca appear also to belong as lateral
branches to this Crinoid development.

First, and probably very early, the Echinoids seem to have branched off. They
became free, used their tentacles for locomotion, and took in food direct through the
mouth, the food grooves, with the nerve ridges, becoming the subepithelial radial
nerves. The arms were again incorporated into the enlarging calyx, or test ; in that
the apical skeleton of the arms degenerated, and thus brought the (ambulacral) ends
of the arms close up to the continually decreasing apical capsule. As this latter
was free, the anus could shift into its centre.

We have come to hold this view of the derivation of the Echinoidea from attached
ancestral forms with arms, a view which, as far as we know, has never before been
published, chiefly for the following reasons :—

The Echinoidca, possess five pairs of gonads, which are at first connected with the
axial organ by means of an aboral circular strand.

This important distinction from the Holothurioidca, with which the Echinoidca
are usually compared in other points, can only be explained by the assumption that
the Echinoidea, originally possessed arms which contained the fertile outgrowths of
the central genital rudiment. That the gonads now lie interradially, presents no

vin ECHINODERMATA—PHYLOGENY 551

difficulty. They could easily have been in the short thick arms, while the terminal
portions of their efferent ducts opened outward intcrradially.

The Ophiuroidea branched oft' from the series leading to the Crinoids by the
readoption of the free manner of life later than the Echinoids. They used the
free arms for locomotion, and took food direct into the month. The tentacles never
became ambulatory tube-feet, bnt only retained their respiratory functions. The
food grooves closed to form tubes, becoming subepithelial radial nerves with their
epineural canals. For further protection, the radial longitudinal rows of ventral
shields became arranged over the closing food grooves. The use of the arms almost
exclusively as locomotory organs determined their slender form, which makes them
appear as mere appendages of the body, and it further led to the return of the
gonads into the disc.

The Asteroidea were the last to branch off from the series of attached Echino-
derms with arms, by the adoption of a free manner of life. They used their tube-
feet first for locomotion, and secondly for seizing and holding fast prey, which was
enveloped en masse direct by the evaginated oral portion of the intestine and drawn
through the mouth into the stomach. The food grqoves now no longer serve as
such, but are retained as deepened arabulacral furrows, from the bases of which the
closely crowded tnbe-feet rise, and into which they can withdraw. Over the
tentacles, withdrawn within the furrow, the spines which border the furrow can
bend protectively together. Deep in the base of the furrow, the radial nerve ridge
is found still in its epithelial position.

From the standpoint, of the foregoing it is to be expected that the ontogeny of
the Holothurioidea, which earliest gave up the attached life, should show the least
trace of the phyletic stage of the attached animal, and that by the avoidance of those
complicated rearrangements which attachment caused, their development should be
much simplified. The facts agree with this expectation, and also with the expecta-
tion that an attached stage is most likely to be ontogenetic'ally repeated in the
Asteroids (cf. the ontogeny of Asferi/ia, with its temporary fixation by means of the
anterior part of the body, the larval organ).

In the foregoing attempt to trace the phylogeny of the Echinoclerms, we have
avoided going into details, and we have also avoided all reference to many
important points, such as the hydropore and the stone canal, the hydrocujl and the
left anterior enteroccel, etc. etc. These can only be elucidated by renewed research,
which must be both extensive and thorough. In applying our views to explain
special points in Echinoderm morphology, it must be acknowledged that in the
majority of cases it does not suffice for a full explanation, and that it cannot indeed
at present be reconciled with many ontogenetical and anatomical facts. The recent
researches in Echinoderm morphology and the attempts at phylogenetic explanations,
which are continually suggesting new points of view, justify us, however, in hoping
that, little by little, many of these interesting and important problems will receive
a satisfactory solution.

Review of the most important Literature.

Comprehensive Works. Text Books. Treatises of Wider Scope. Re-
searches extending over some or all of the Classes.

A. Agassiz. Palceontological and embryological development. Address before the
American Association for the Advancement of Science. Boston Meet imj. Cam-
bridge, 1880. Also Ann. and Mag. Nat. Hist. (5). Vol. VI. 1880.

552 COMPARATIVE ANATOMY CHAP.

E. Baudelot. Contributions fi I'histoire du systeme nerveux des Echinodenncs. Arch.

deZool. Exp. (1). Vol.1. 1872.
P. H. Carpenter. Several important treatises especially on the SkeJdul System, from

1870-1890, in various Journals, principally in Quart. Journ. Microsc. Science,

and in the Ann. and Mag. Nat. Hist.
L. Cuenot. Etudes sur le sang, son r6le et sa formation dans laserie animal e. Part

II. Invertebrds. Archives Zoologic expe'rimcntalc (2). Tome IX. 1891.
- Etudes morphologiques sur les Echinodermcs. Arch. Biologic (van Benedcn}.

Tome XI. 1891.

Herb. E. Durham. On wandering cells in Echinodcrms, etc., more especially

regard to excretory functions. Quart. Journ. Mic. Sci. Vol. XXXIII. 1892.
Jobs. Frenzel. Beitrage zur vcrglcichend.cn Physiologic und Histologie der Ver-

dauung. I. Mitth. Der Darmkanal der Echinodermen. Arch. /. Anat. v.

Physiol. Physiol. Abth. 1892.
Greeff. Ueber den Bau der Echinodermen. Sitzungsbcr. d. Gcsellsch. f. Naturiciss.

zit Marburg. 5 Parts. 1871, 1872, 1876, 1879.
0. Hamann. Die wandernden Urkcimzellen und Hire Rcifungsstdtten bei den Echino-

dermen. Zeitschr. /. wiss. Zool. 46 Bd. 1887.
M. M. Hartog. The true nature of the madreporic system of Echinodcrmata, with

remarks on Ncphridia. Ann. and Mag. Nat. Hist. (5). Vol. XX. 1887.
Herapath. On tlie pedicellarice of the Echinodermata. Quart. Journ. Microsc.

Science. Vol. V. 1865.

Carl Jickeli. Ucber den Ban der Echinodermen. Zool. An~. 7 Jahry. 1884.
A. Kowalevsky. Zur Kcnntnissder Excretionsorganc. Biol. Ccntralblatt. 9 Bd. 1889.
H. Ludwig. Morphologische Studien an Echinodermen. Leipzig, 1877-1882.

Preprints from the Zeitschr. f. iviss. Zool.

- Bronn's Klassen und Onlninnj, -n des Thierrcichs. 2 Bd., 3 Abtli. Echino-
dermen. Holotlmrioidea. Asteroidea in progress.

Job. Miiller. f'Hicr ili~n Bau der Echinodermen. Berlin, 1854. Abhandl. ' Akad.

IFissenscli. Berlin, 1853.
M. Neumayr. Morphologische Studien uber fossile Echinodermen. Sitznngsbcr.

Akad. Wissensch. JFicn. mathcm.-naturw. Cl. 84 Bd. 1 Abth. 1881.

- Die Stiimme des Thierreichcs, etc. 1888.

E. Perrier. Reclwrches sur les pedicellaires et les ambulacrcs des Asteries et
des Oursins. AnnaJes des Sciences natur. (5). Tomes XII. and XIII.
1869-1870.

- Echinodermcs. Part I. Stellerides. Expcd. du Travailleur et du Talisman.
Paris. 1895.

G. J. Romanes and J. E. Ewart. Observations on the locomotor system of Echino-

di-ntiata. Proceed. Roy. Soc. London. Vol. XXXII. 1881. Philos. Trans-

actions London. 1881. Part III.
C. F. and P. B. Sarasin. Uebcr die Anatomic der Echinothurideii und die Phylo-

genie der Echinodermen. Ergcbnisse nat. Forschungen Ceylon. 1 Bd. 1888.
R. Semon. Die Homologicn innerhalb des Echinodxrmenstammes. Morph. Jaltr-

buch, 15 Bd. 1889.
— Die Entwickelung der Sijnaptu diijituta und die Stainmcsgcschichfc der Echino-

dermen. Jcnaische Zeitschr. f. Naturwiss. 22 Bd. 1888.
W. P. Sladen. On Hie homologies of the primary larval plates in the test of Brachiate

Echinodcrms. Quart. Journ. Microsc. Science (2). Vol. XXIV. 1884.
Friedr. Tiedemann. Anatomic der Rohrcnholothuric, des pomeranzenfarbigen

Seesterns und des Stein- Seeigels. Landslmt, 1816.
K. A. Zittel. Handbuch der Paliiontologie. 1 Bd. Miinchen and Leipzig, 1876-

1880.

vni ECHINODERMATA— LITERATURE 553

Principal Systematic Works and Treatises specially dealing with the
Morphology of the Skeletal System. Pedicellarise.

A. Agassiz. -Revision of the Echini (Illustrated Catalogue of the Museum of Com-
in/ratice Zoology at Harvard College. No. VII.). Cambridge, Mass., 1872-1874.

- North American Starfishes (Mem. of the Museum of Comp. Zool. Vol. V. No. I).
Cambridge, Mass., 1877.

- Report on the Echinoidca. Report on the scient. results of the voyage oj
H.M.S. "Challenger," Zool. Vol. III. Part IX. 1881.

- Report on the Echini (Results of dredging by the "Blake" XXIV. Part L).
Memoirs of the Museum of Comparative Zoology at Harvard College. Vol. X.
No. I. Cambridge, Mass., 1883.

F. A. Bather. British fossil Crinoids. A scries in progress in Ann. and Mag. Nat.

Hist, beginning (6). Vol. V. 1890.

- The Crinoidea of Gotland. Part I. The Crinoidea inadunata. Kongl. Svenska
Vctenskaps-Akademiens Handlingar. Bandet 25. Stockholm, 1893.

J. F. Brandt. Prodromus description is animalium ab H. Mcrtcnsio in orbis terrarum
circumnai'igatione observatorum. Fasc. 1. Petropoli, 1835.

Leopold von Buch. Ueber Cystideen. Abhandl. Bcrl. Akad. 1844.

P. H. Carpenter. Report on the Crinoidea. I. The Stalked Crinoids. Voyage of
the " Challenger." Vol. XI. PartXXXlI. London, 1884. II. The Comatulce.
Ibid. Vol. XXVI. Part LX. 1888.

- A Series of important works, chiefly on the Morphology of the Skeletal System,
appearing from 1870-1890 -in English journals, especially Quart. Journ. of
Microsc. Science, and Annals and Magazine of Natural History.

G. Cotteau. Echinides. PaUontologic franqaisc. Vols. VII., IX., X. Paris, 1862-

1879.

D. C. Danielssen and Joh. Koren. Holothurioidea. The Norwegian North. Atlantic

Expedition. 1876-1878. VI. Cbristiania, 1882.

E. Desor. Synopsis des Echinides fossiles. Paris and Wiesbaden, 1858.

F. Dujardin. Rechcrchcs sur la Comatule de la Mcditerrante. L'lnstitnt. Tome

III. 1835.

P. M. Duncan. A revision of the genera and great groups of the Echinoidea.
Journ. Linn. Soc. London. Vol. XXIII. 1889.

P. M. Duncan and W. Percy Sladen. A memoir on the Echinodcrmata of the arctic
sea to the west of Greenland. London, 1881.

R. Etheridge jun. and P. H. Carpenter. Catalogue of the Blastoidea in the Geo-
logical Department of the British Museum (Natural History), with an account of
the morphology and systematic position of the group and a revision of the genera
and species. London, 1886.

J. W. Fewkes. On the serial relationship of the ambulacral and adambulacral Cal-
careous Plates of the Starfishes. Proceed. Boston Sor. JV. H. Vol. XXIV.
1890. Primary spines of Echinoderms. Ibid.

- On the development of the Calcareous Plates of Asterias. Bull. Mus. Harvard
Coll. Vol. XVII. 1888.

— On the development of the Calcareous Plates of Amphiura. Bull. Mus. Comp.
Zool. Vol. XIII. No. IV. 1887.

Alex. Foettinger. Sur la structure des pedicellaircs gcmmiformes d'Echinides.
Arch, de Biol. Vol. II. 1881.

E. Forbes. A history of British Starfishes and other animals of the class Echino-
dcrmata. London, 1841.

E. Fraas. Die Asterien des wcissen Jura von Schivabcn und Franken, mit Untcr-

554 COMPARATIVE ANATOMY CHAP.

suchungen uber die Structur der Echinodcrmcu und das Kalkgcriist dcr Aslcrieii.

Palaeontogr. 32 Bel. 1886.
Wilhelm Giesbrecht. Derfcinerc Ban dcr Seciyelzdlme. Morph. Jahrbuch. 6 Bd.

1880.
J. E. Gray. Synopsis of the species of Starfislics in the British Museum. London,

1866.

G. Hambach. Contributions to the anatomy of the genus Pcntremites, with descrip-
tion of new species. Trans, of the Acad. of Science of St. Louis. Vol. IV .

No. I. 1880.
Clem. Hartlaub. Beitrag zur Kcnntniss dcr Comatulidcnfauna dcs Indisclicn

Archipcls. Nova Acta Acad. COBS. Leop.-Carol. Germ. Nat. Cur. 58 Bd.

No. 1. 1891.

C. Heller. Die Zo&phyten und Echinodermcn dcs Adriatischen Mecres. "Wien, 1868.
0. Jaekel. Beit-rage zur Kenntniss dcr Palccozoischen Crinoidcn Dcutschlands. Pal.

Abh. New Series. III. 1895.
K. Lampert. Die Seewalzen. Rciscn im Archipcl dcr Philippinen von Dr. >'.

Semper. 2 Th. Wiss. Rcsultatc. 4 Bd. 3 Abth. Wiesbaden, 1885.
A. Ljungman. Ophiuridea viventia hue usque cognita. Stockholm, 1867.
P. de Loriol. Echinologic hclvttiquc. I., II., III. Geneva, 1868-1875.

- Monographic dcs Crinoidcs fossilcs dc la Suissc. Geneva, 1877-1879.

— Paleontologie francaisc. Terrain jurassiquc. Tome XI. Crinoides. Part I.
1882-1884. Part II. 1884-1889.

S. Loven. £tudcs snr ?es £chinoidtes. K. Svensk. Vet.-Akad. Handl. 11 Bd.
(1873-75). Stockholm, 1874.

- On Pan rtalcs ia, a genus of Echinoidca. K. Svensk. Vet. -Akad. Handl . 19 Bd.
1884.

Echinolorjica. Bihang till K. Svensk. Vet.-Akad. Handl. XVIII. Afd. 4.

1892.
Chr. Fr. Liitken. Additamenta ad historiam Ophiuridurum. Copenhagen, 1858-

1869.
Hubert Ludwig. Trichastcr clegans. Zcitschr. f. wisscnsch. Zool. 31 Bd. 1878.

- Zur Kcnntniss dcr Gatlung Brisinga. Ibid.

- Das Mundskclct dcr Asteri.cn und Ophiuren. Ibid. 32 Bd. 1879.

- Ueber den primaren Stcinkanal der Crinoideen nebst verglcichcnd-anatomischen
Bcmerkunyen. Ibid. 34 Bd. 1880.

- Zur Entirickflunysgeschichtc dcs Ophiurenskclctcs. Ibid. 36 Bd. 1882.

— Qphiopteron elegans, cine neue, wahrschcinlich schwimmeiule Ophinridenform.
Ibid. 47 Bd. 1888.

- Ankyrodcrma musculus (Hiss. ), eiiic Molpadiide des Mittclmccres, ncbst Bcincr-
kungcn zur Phylogenie und Systcmatik der Holothuricn. Ibid. 51 Bd. 1891.

- Holothv roidea of the "Albatross" Expedition. Mem. Mas. Harvard. XVII.
3. 1894.

Th. Lyman. Ophiuridce and Astrophytidce. Ulustr. Catalogue of the Museum of
Comp. Zool. Harvard College. I. Cambridge, Mass., 1865.

- Report on the Ophiuridea. Report on the scientific results of the voyage of
H.M.S. "Challenger." Vol. V. Part XIV. London, 1882.

Meyer. Uebcr die Laterne dcs Aristotcles. Arch. f. Anat. u. Physiol. 1849.

J. S. Miller. ./ natural history of the Crinoidca or lily-shaped animals. Bristol,

1821.
Job. Miiller. Ueber den Bau des Pentacrinns caput Mcdusce. Abhandl. d. Akad. d.

irixxensch. Berlin, 1841.

- Ueber die Gattnini Cmnatiila Lam. und ihrc Arten. Abhandl. d. Akad. <l.
Wisseiisch. Berlin, 1847.

vin ECHINODERMATA LITERATURE 555

Job. Miiller and Fr. H. Troschel. Si/stem der Aster iff en. Braunschweig, 1842.
Edm. Perrier. Observations sur Us relations qui existent cntrc hs dispositions cles

pores arnbulacraires a I'ext&rieur ct a finterieur du test dcs Echinides reguliers.

Nouv. Arch. Museum. Tome V. 1869.
Edm. Perrier. Becherches sur les p&liccllaires ct les ambulacres dcs Asteries ct des

Oursins. Annales des Sciences natur. (5). Vols. XII. and XIII. 1869-1870.

- Revision de la collection de StclUridcs du Museum d'histoire ntiti< relic de Paris.
Paris, 1875-1876.

- Memoire sur les etoiles de mer, recucillics dans la mcr des Antilles ct Ic golfc if it
Mcj'ique. Paris, 1884.

H. Prouho. Recherclies sur le Dorocidaris papillata et quclques autrcs Echinides de
la mMitcrranee. Arch, de Zool. cxper. (2). Tome V. 1887-1888.

F. A. Quenstedt. Pctrefactenkunde Deutschlands. 3 Bd. Echiniden. Leipzig,

1872-1875.

Ferd. Romer. Monographic der fossilen Crinoidcenfamilic der Blastoidecn. Arch,
f. NaturgcschicJitc. 1851.

G. O. Sars. Memoire pour servir d la connaissance des Crinoides vivants. Chris-

tiania, 1868.

- Researches on the structure, etc., of the genus Brisinga. Christiania, 1875.

M. Sars. Bid-rag til Kundskaben om Middclhavets Littoral-Fauna. Christiania,

1857.
— Ovcrsigt of Norges Echinodcrmer. Christiania, 1861.

E. Selenka. Beitriige zur Anatomic und Systcmatik der Holothuricn. Zcitschr. f.
wissensch. Zool. 17 and 18 Bd. 1867, 1868.

C. Semper. Reisen im Archipel der Philippinen. \ Bd. Holothurien. Leipzig,
1868.

W. Percy Sladen. On a remarkable form of Pedicellaria, and the functions per-
formed thereby, etc. Annals and Mag. of Nat. History (5). Vol. VI. 1880.

- On the homologies of the primary larval plates in the test of Brachiate Echino-
derms. Quart. Journ. Microsc. Science (2). Vol. XXIV. 1884.

- Report upon the Asteroidea collected by H.M.ti. "Challenger." Vol. XXX.
1889.

Hj. Theel. Report on the Jfolot.hurioidca collected during the voyage of the
"Challenger." Parti. Vol. IV. 1882. Part II. Vol. XIV. 1886.

- On the formation and rcsorption of the skeleton in the Echinoderms. Ofv. Ak.
Forh. 1894.

Volborth. Ucber Achradocystitcs und Gystoblastus, zwei neue Crinoiden-Gattungen.

Mem. Acad. Imp. Sc. St. Petersbourg. 1870. Tome XVI. No. II.
C. Viguier. Anatomic comparec du squelctte dcs Stclle rides. Arch. Zool. expert HI.

Tome VII. 1879.
C. Wachsmuth and F. Springer. Revision of the Palwocrinoidca. Proceed. Acad.

Nat, Sc. of Philadelphia. 1879, 1881, 1885.

- (1) Discovery of the ventral structure of Taxocrinus and Haplocrinus, and
consequent modifications in the classification of the Crinoidca, Proceed. Acad.
Nat. Science Philadelphia. 1889.

(2) C rotalocrinus : its structure and zoological position. Ibid.

- Tli e perisomic plates of the Crinoids. Pro. Acad. Nat. Sc. Philadelphia. 1890.
T. Wright. Monograph of the British fossil Echiuodcrmata of the bolithic forma-
tion. London, 1857-1880.

- Monograph of the British fossil Echinodermata of the cretaceous formation.
London, 1864-1882.

K. A. Zittel. Handbuch der Paldontologie. 1 Bd. 1876-1880.

556 COMPARATIVE ANATOMY CHAP.

Anatomical Monographs.

A. Agassiz. Revision of the Echini. Mus. Compai: Anatomy Harvard Coll. Vol.
VII. 1872.

- North American Starfishes (Mem. of the Mus. of Comp. Zool. Vol. V. No.
I.). Cambridge, Mass., 1877.

Report on the Echinoidea collected by H.M.S. "Challenger." Vol. III.

London, 1881. (Also contains anatomical details.)

Nic. Christo Apostolides. Anatomic et developpement des Ophiurcs. Arch. Zool.

cfptr. genir. Vol. X. 1882.
Alb. Baur. Beitriige zur Naturgcschichte der Synapta digitata. Ada. Acad. Cccs.

Leop. -Carol. Nat. Curios. 1864.
P. H. Carpenter. Report- upon the Crinoidea collected during the voyage of H.M.S.

" Challenger " during the years 1873-1876. Part 1. General Morphology, with

descriptions of the stalked Crinoids. Vol. XL 1884. Part 2. The Comatulce.

Ibid. Vol. XXVI. 1888. (The second part is almost exclusively systematic

and descriptive.)
William B. Carpenter. Researclies on the stri/ft are, physiology, and development of

Antcdon rosaceus. Philos. Transactions. Vol. CLVI. 1866. Addendum.

Pro. R. Soc. Vol. XXIV. 1876.
L. Cuenot. Contribution a t etude anatomique dcs Asttrides. Arch. Zool. cxperim.

(•2). Tome V. Supplementary (1887-1890).

Etudes anatomiqucs et morphologiqucs sur Ics Ophiures. Arch. Zool. cy.ptr.

(2). Tome VI. 1888.

D. C. Danielssen and J. Koren. HoJothurioidca. The Norwegian North Atlantic
Expedition. 1876-1878. Christiania, 1882.

Fredericq. Contributions a I'etude dcs Echinides. Arch, dc Zool. exp&rim. Tome
V. 1876.

0. Hamann. Bi'ltr/iijc -ur Histoloyie der Echinodermen. 1. Heft. Die Holothurien.
Jena, 1884. 2. Heft. Die Asteriden. Jena, 1885. 3. Heft. Anatomic n ml
Histologie der Ecldniden und Spatangiden. Jena, 1887. 4. Heft. Anatomic
und Histologie der Ophiuren und Crinoidcn. Jena, 1889. (Reprints from the
Jc /in incite n Zeitxdir. f. N«ti.irn-!sx.)

- Bcitritge ;///• Histologie. der Echinodermen. 1. Die Holothurien (Pedata) und
das Nervensy stem der Asteriden. Zeitschr. f. wiss. Zool. 39 Bd. 1883.

£. Herouard. Rechcrchcs sur les Holuthurics dcs cotes dc France. Arch. ZuoL

exper. (2). Tome VII. 1889.
C. F. Heusinger. .///<//'//;/ /«.•//,« Untersvchvnydcr Comatula meditcrranca. Zeitschr.

f. orrjirniscJic Physik. Tome III. 1828.
C. K. Hoffmann. Zur Anatomic der Eehincn und Spatangcn. Niederl. Arch.

Zool. 1 Bd. 1871.

— Z nr Anatomic der Asteriden. Niederl. Arch. f. Zool. 2 Bd. 1874.
G. F. Jaeger. Dc Holothuriis. Diss. inaug. Zurich, 1833.
Et. Jourdan. llci-lta-clics sur Vkistologie des Holothunes. Ann. Mus. H. N.

Marseilles. Tome I. 1883.
R. Kohler. Rechcrchcs sur les Echinides des cotes de Provence. Ann. Mits. H. N.

Marseilles. Tome I. 1883.
W. Lange. Bcitrmj zur Anatomic und Histologie der Astericn und Ophiuren.

Morph. Jahrbuch. 2 Bd. 1876.

Leydig. Anatomische Notizcn iiber Synapta digitata. Mailer's Arch. 1852.
Loven. E-tudes sur les Echinnidcs. Svensk. Vetensk.-Akad. 11 Bd. 1874.
H. Ludwig. Be it rage ~IT Kcnntniss der Holotlmricu. Arbeit. Zool. lust. Wurzbunj.

2 Bd. 1874.

viii EGHINODERMA TA —LITER A T URE 557

H. Ludwig. Bcitrdgc zur Anatomic dcr Crinnidccn. Leipzig, 1877. Zeitschr. f.
u-iss. Zool. 26 Bd. 1877. 28 Bd. 1877.

- Zur Anatomic dcs Rhizocrinus lofotensis. Ibid. 29 Bd. 1877.

- Ucbcr Rhopalodina lagcniformis. Ibid. 1877.

- Beitrdge zur Anatomic dcr Astcriden. Ibid. 30 Bd. 1878.

- Ucbcr Asthenosoma varium Grube und iiber ein neues Organ bci den Cidariden.
Ibid. 34 Bd. 1880. Berichtigung im Zool. Anz. 3 Jahrg.

- Ueber den prividren Stcinkanal dcr Crinoidcn ncbst vergl.-anatomischcn
Bcmerkungen ubcr die Ecltin<,il< rmcn iiberhaupt. Ibid. 34 Bd. 1880.

- Neue Bcitrdgc zur Anatomic dcr Ophinren. Ibid. 34 Bd. 1880.

- Nochmals die Ehophalodi.ua lagcniformis. Ibid. 48 Bd. 1889.

- Die Secwalzen. Bronn's Klasscn und Ordnungcn dcs Thier-Rcichs. 2 Bd. 3
Abth. JSchiiiodermata. Leipzig, 1889-1892.

Hub. Ludwig and Ph. Barthels. Bcitrdgc zur Anatomic dcr Holotlmrien. Zeitschr.

f. u-iss. Zool. 54 Bd. 1892.
T. Lymau. Report on the Ophiuridea. Voyage of H.M.ti. " Challenger," Zool.

Vol. XV. Part 14. London, 1882.

J. Miiller. Ucbcr den Bau von Pcntacrinus caput Medusa'. Abhandl. d. Akad. d.
JVissensch. Berlin, 1841.

- Ueber Synapta digitata und iiber die Erzeugung voii Schnecken in Holothurien.
Berlin, 1852.

E. Perrier. Memoirc sur I 'organisation el le devcloppcment de la Comatule de la
Meditcrranee (Antcdon rosacea Linck). Nouv. Archin's Mus. Paris, 1886-
1892.

- Rcclicrclies sur V anatomic et la regeneration des bras de la Comatula rosacea.
Arch. Zool. expcr. Tome II. 1872.

H. Prouho. Recherchcs sur le Dorocidaris papillata et. quclqvcs aufres iZch in ides de

la Mtditcrranee. Arch, dc Zool. exper. (2). Tome V. 1887.
A. de Quatrefages. Memo-ire sur le Synaptc dc Duvernoy. Ann. Sc. natur. (2).

Tome XIV. 1842.
C. Semper. Kurzc anatomischc Bcmerkungen iiber Comatula. Arbeit. Zool. Inst.

Wiirzbiirg. 1 Bd. 1874.
C. F. and P. B. Sarasin. Ucbcr die Anatomic dcr Echinothuriden und die Phylo-

gcnic dcr Echinodcrmen. Ergebnisse Nat. Forschungcn Ceylon. 1 Bd. 1888.
G. 0. Sars. Researches on the structure etc. of the genus Brisinga. Christiania,

1875.

— Memoirc pour scrvir d la connaissancc des Crinoides vivants. Christiania, 1868.
Selenka. Beitrdgc zur Anatomic und Systematik dcr Holothurien. Zeitschr. f.

u-iss. Zool. Two papers in Vols. 17 and 18. 1867-1868.
R. Semon. Bcitrdgc zur Naturgeschichte der Synaptidcn dcs Mittelmcers. Mitth.

Zool. Stat. in Neapcl. 7 Bd. 1887.

Von Siebold. Zur Anatomic der Kecstcrne. Miiller s Archiv. 1866.
H. Simroth. Anatomic und Schizogonic dcr Ophiactis virens. Zeitschr. f. wiss.

Zool. Two papers in Vols. 27 and 28. 1877-
Reinhold Teuscher. Bcitrdge zur Anatomic dcr Echinodcrmen. I. Comatula medi-

ti'i-ranea. Jcnaischc Zeitschr. 10 Bd. 1876.
Friedr. Tiedemann. Anatomic der Rohrenholothurie, des pomeranzcnfarbigcn

Sccstcrns und dcs Stein- Seeigcls. Landshut, 1816.
Hj. Theel. Report on the Holothurioidca. Voyage of H.M.ti. " Cliallenger. "

Zoology. Part I. Vol. VI. Part XIII. 1882. Part II. Vol. XIV. Part

XXXIX. 1885.
J. V. Thompson. Sur le Pcntacrinus europaeiis, I'e'tat dc jeuncsse du genre Comatula.

L'lnstitut. 1835.

558 COMPARATIVE ANATOMY CHAP.

G. Valentin. An<it<nn!c du genre Echinus. Monographics d' Echinodermes, par

L. Agassiz. Neuchatel, 1841.
Alex. Weinberg. Die Morphologic dcr febcmlcn Crinoiden mil Bczichung auf die

form Antedon rosacca Linck. Naturhistoriker. 5 Jahrg. 1883.

Works dealing with Single Organs or Systems of Organs.

H. Ayers. On the structure and function of the Sphceridia (if the Echinoidca. Quart.

Journ, Microsc. Science (2). Vol. XXVI. 1886.
E. W. M'Bride. The development of the, genital organs, ovoid gland, axial ami aboral

sinuses in Amphiura squamata, together i/'lf/t so me remarks on Lud trig's Harmc/I

System in this Ophiurid. Quart. Journ. Microsc. Science. Vol. XXXIV. 1893.
The development of the dorsal organs, genital rachis and genital organs in Asterina

gibbosa. Zool. Anz. 16 Jahrg. 1893.
E. Baudelot. Contributions a I'histoire du systeme nerveux des Echinodermes. Arch.

dc Zool. exper. 1872.
W. B. Carpenter. On the nervous system of Crinoiden. Proceed. Eoy. Soc. London.

Vol. XXXVII. 1884.
L. Cuenot. Etudes sur le sang, son role ct sa formation dans la seric animtili .

Part II. Invertebres. Arch. Zool. exp&r. (2). Tome IX. 1891.

E. Haeckel. Ucber die Augen und Nervcn dcr Seesterne. Zeitschr. f. wiss. Zool.

10 Bd. 1860.
Otto Hamann. Beitriigc zur Histologie dcr Ecldnodermen. 1. Die Holothurien

(Pedata] und das Ncrvensij stem dcr Asteriden. Zeitschr. f. wiss. Zool. 39 Bd.

1883. 2. Mitth. 1. Nervensystem der pedaten Holothurien. 2. Cumersche

Organe. 3. Nervensystem und Sinnesorgane der Apoden. Ibid.
R. Kb'hler. ll<:ch<*rch<:s sur Tappareil circulatoire dcs Ophiurcs. Ann. Sc. nat. (7).

Tome II. 1887.

F. Leipoldt. Das angcbliche Excrctionsorgan der Seeigel, untersucht an Sphocrechinus

granularis unil Dorocidaris papillata. Zeitschr. f. iviss. Zool. 55 Bd. 1893.
Theodore Lyman. The stomach and genital organs of Astrophytida'. Bull. Mus.

Co/up. Zool. Harvard College. Vol. VIII. No. 6. Cambridge, Mass., 1891.
A. M. Marshall. On the- nervous system of Antcdon rosaceus. Quart. Journ.

Microsc. Science. Vol. XXIV. 1884.
C. Mettenheimer. Ucber die Gesiehtsorgam des violcttcn Seesterns. Mailer's Arch.,

1872.
E. A. Minchin. Notes on the Cuvierian organs of Holothuria nigra. Ann. and

M«g. Nat. History (6). Vol. X. 1892.
J. Niemiec. Rcchcrchcs sur les xentouscs dans le regne animal. Recueil Zool. Suisse.

Tome II. Encore un mot, etc. Ibid. 1885.
Owsjanikoff. Ucber das Nervensystem der Seesterne. Bull. Acad. St. Petcrsbounj.

1870. Tome XV.
Ed. Perrier. Rccherchcs sur les pediccllaircs ct les ambulacres des Asteries et des

Oursins. Ann. dcs Sciences natur. (5). 12 and 13 Bd. 1869-1870.
Recherches sur Vappareil circulatoire dcs Oursins. Arch, de Zool. exper.

Tome IV. 1875.

G. J. Romanes and J. C. Ewart. Observations on the locomotor system of Echino-

dermata. Philos. Transact. London. Part III. 1881.
A. Russo. Ricerche citologiclie sugli dementi seminali delle Ophiurece (spermato-

gcncsi-oogencsi) Morfologia dell' appareccMo riprodutore. Internat. Monatschr.

f. Anat. und. Phys. 8 Bd. 8 Heft.
C. F. and P. B. Sarasin. Die Augen und das Integument dcr Diadematiden.

vin EGHINODERMATA— LITERATURE 559

Ergebnissc naturwiss. Forschungcn auf Ceylon in d. Jo hint 1884-1886. 1 Bd.

1887.
Rich. Semon. Das Nervensystem dcr Holothurien. Jcnaische Zeitschr. f. Natur-

in'xii. 16 Bd. 1883.

H. S. Wilson. The nervous system of tlic Astcridce. Transact. Linncan Society.
Vol. XXIII. 1860.

Ontogeny.

A. Agassiz. North American Starfishes, 1864. Mem. of the Museum of Comp. Zool.
Harvard College. Vol. V. 1877.

- Revision of the Echini. Illustr. Catalogue of the Museum of Comp. Zool.
Harvard College. 1872-1874.

Nic. Christo Apostolides. Anatomic ft dtveloppement des Ophiures. Arch. Zool.

c.'-m'r. Tome X. 1882.
J. Barrois. RechercJies sur Ic develop/if ment dc la Comatule (C. mcditerranca}.

Rei'iicil Zool. Suisse. Tome IV. 1888.
E. W. M'Bride. The development of the genital organs, ovoid gland, axial ami aboral

sinuses in Amphiura squamata, together with some remarks on Ludwig's Hu'iiinf

System in this Ophiurid. Quart. Journ. Micros?. Science. Vol. XXXIV.

Part II. 1893.
— The development of the dorsal organ, genital rachis and genital organs in

Asterina gibbosa. Zool. An^. 16 Jahrg. 1893.
H. Bury. The early stages in the development of Antedon rosacca. Philos. Transac-

tions. Vol. CLXXIX. 1888.

- Studies in the embryology of the Echinoderms. Quart. Journ. Microsc. >SV.
Vol. XXIX. 1889.

- The Metamorphosis of Echinoderms. Ibid. Vol. XXXVIII. 1895.

P. H. Carpenter. On some points in the anatomy of larval Comatula. Quart. Journ.
Microsc. Sc. (2). Vol. XXIV. 1884.

- Notes on Echinoderm morphology. No. 11. On the development of th<

plates in Amphiura squamata. Quart. Journ. Microsc. 8c. Vol. XXVIII.

1888.
William B. Carpenter. Researches on the structure, physiology, and development of

Antedon rosaccus. Philos. Transact. Vol. CLVI. 1866. Addendum. Pro.

Roy. Soc. Vol. XXIV. 1876.
J. W. Fewkes. On the development of calcareous plates of Amphiura. Bull, of the

Museum of Comp. Zool. of Harvard College. Vol. XIII. 1887.
A. Fleischmann. Die Entwickelung des Eies von Echinocardium cordatum. Zeitschr.

f. u-iss. Zool. 46 Bd. 1888.
Geo. W. Field. The larva of Astcrias vulgaris. Quart. Journ. of Microsc. 8c. Vol.

XXXIV. Part 2. 1892.
Alexander Goette. Vergleichcndc Entivickclungsgeschichte dcr Comatula mediter-

ram'a. Arch. f. mikr. Anatomic. Tome XII. 1876.
E. Korshelt. Zur Bildung des mittlcren Keimblatts bet den Echinodermen. Zool.

Jafirb. Abth. Morph. 3 Bd. 1889.
Kowalevsky. Leitrage ~ur Eiitu-ii-ki'Iinujxi/eschichfe dcr Holothuricn. Me'in. dc

I'Acad. Imper. de St. P&crsbourg. (7.) Tome XI. 1867.
H. Ludwig. Zur Entwickclungsgcschichtc des Ophiurenskclctcs. Zeitschr. f. iviss.

Zool. 36 Bd. 1881.

- Entwickelungsgeschichte dcr Asterina gibbosa Forbes. Zeitschr. f. wiss. Zool.
37 Bd. Also in Morph. Studien an Echinodermen. 2 Bd. 1882.

- Zur Entivickelungscjcschichte der Holothuricn. Sitzber. K. Prcuss. Acad.
d. Wiss. 1891. X. XXXII.

560 COMPARATIVE ANATOMY CHAP, vm

E. W. M'Bride. The organogcuy of Asteriiia gibbosa. Pro. R. Sue. L1V. 1894.
E. Metschnikoff. Studicn iibcr die Enttcickelungder J-Ji-//im>i/i n/ic/i u/i<l Ncmrrtinen.
Mem. Acad. St. Petersburg. Tome XIV. (No. 8). 1869.

- Eiitwickclung von Comatula. Bull. Acad. St. Petersbourg. XV. (Col. 508.)
1871.

- Vcrgleichcnd-embryologischc Studicn. 5. Ueber die Bilding dcr IVandcrzellen
bri Astericn und Echiniden. Zeitschr. f. iviss. Zool. 42 Bd. 1885.

Job. Miiller. Classical Treatises on the larval forms of Echiiwdcrms, and their meta-
morphoses. Abhandlungen d. Konigl. Akad. d. Wissensch. zu Berlin. 1848,
1849, 1850, 1852, 1853, 1855.

Edm. Perrier. Memoirc sur I' organisation ct la cteveloppement de la Comatule de la
Mtditcrrane'e. Nouv. Arch, du Mus. Hist. nat. Paris. 1886-1892.

A. Russo. Si'i-i'i-al treatises in Neapolitan journals. 1891-1892.

Oswald Seeliger. Studicn zur Entwickclungsgescliichte der Crinoiden (Antedon
rosacea}. Zool. Jahrb. v. Spengel. Abth. f. Anat. u. Out. 6 ]Jd. Jena, 1892.

Emil Selenka. Die Krimblutter der Echinodermen. Studicn iib. d. Entwickcl-
ungsgcsch. d. Thicre. 2 Heft. 1883.

- Zur Entwickelung der Holothurien. Ein Beitrag zur Keimblattertheorie.
Zntsckr. f. iciss. Zool. 27 Bd. 1876.

R. Semon. Die Eniwickelung <l<r Synapta digitdta und die StammcsgcschicTde der

Echinodermen. Jenaisclie Zeitschr. f. Naturwiss. 22 Bd. 1888.
— Zur Morphologic der bilatcralen Wimpcrschniire dcr Echinodermenlarven.
Jena. Zeitschr. f. Naturwiss. 25 Bd. 1891.

C. Wyville Thomson. On the embryology of the Echinodcrmata. Nat. Hisiunj
Review. 1863-1864.

- On the cmbryogcny of Antedon rosaceus. Philosoph. Transact, of the Hoy. Soc.
London. Vol. CLIII. 1S65.

Hjalmar Theel. On the development of Echinocyamus pusillus (0. F. Midler).
Nova Ada Reg. Soc. Sci. Scr. 3. Vol. XV. (No. 6). Upsala, 1892.

Phylogeny.

Besides the above-named treatises and works of Al. Agassiz. P. H. Carpenter,
Cuenot, Haeckel, Ludwig, Neumayr, Perrier. Sarasin. Seeliger, Semon,
compare especially 0. Biitschli, Versueh dcr Ableitung des Echinoderms aus cine,-
bilatcralen Urform. Zeitschr. f. wiss. Zool. 53 Bd. Suppl. 1892.

CHAPTER IX
ENTEROPNEUSTA l

BILATERALLY symmetrical, long, vermiform animals with soft skin.
The body is divided into (1) a preoral proboscis, (2) a short "collar,"
(3) a long trunk. The wide mouth lies at the boundary between the
proboscis and the collar. The anus is terminal. The long intestine
may be divided into four sections. The mouth leads direct into (1)
the buccal cavity, which runs through the collar, and sends off a
diverticulum anteriorly into the proboscis. The buccal cavity leads
into (2) a branchial intestine, which is in open communication with
the exterior through numerous pairs of consecutive gill-pouches. The
branchial intestine passes, through an intermediate portion, into (3)
the hepatic intestine, which is often provided with two longitudinal
rows of caeca. Lastly, the hepatic intestine runs into (4) an efferent
intestine, which opens outwards through the anus.

There is an unpaired proboscidal coelom, which opens outward at
the base of the proboscis either by a single pore on the left side, or by
two symmetrical pores. The coelom of the collar is paired, and there
are two pores at its posterior end. The coelom of the trunk is also
paired. The integumental and intestinal musculature are derived
from the coelomic walls. The nervous system consists of a layer of
nerve fibres in the integument, this layer being thickened to form a
mediodorsal nerve cord provided with ganglion cells, and a similar
medioventral cord, both extending the whole length of the trunk.
The dorsal cord sinks, in the collar, below the integument, forming
the collar cord. A capillary network is found within all the limiting
or basal membranes of the body. A large contractile vessel lies in
the dorsal middle line of the collar and trunk, and a similar vessel in
the ventral middle line of the trunk. In the former, the blood flows

1 In this chapter the author, relying largely upon Spengel's monograph on
Balanoglossus, rejects the proposed affinity between the Eiiteropneusta and the ancestors
of the Vertebrata. As this affinity has been very widely accepted in England and
America, the student should consult M'Bride's " Review of Spengel's Monograph "
(Quart. Journ. Mien*. Sci., vol. xxxvi., 1894), which is written from this latter point of
view. — TR.

VOL. II 2 O

562

GOMPARA TIVE ANA TOMY

CHAP.

B

from behind forward, in the latter, from before backward. A pulsating
heart vesicle, which serves to propel the blood, though not belonging

to the vascular system,
is found in the proboscis,
above the intestinal
caecum.

The sexes are sepa-
rate in the Entero-
pneusta. The gonads are
tubes or sacs which lie
in longitudinal rows in
the anterior region of
the trunk (in and behind
the posterior branchial
region) and open through
dorsal genital apertures.
There are no copulatory
organs. Reproduction
is sexual. Development
is with metamorphosis
(in which case the larva
is known as the Tornaria)
or with abbreviated
metamorphosis.

Marine, inhabiting
sand or mud. Four
genera : Ptychodera,
Glandieeps, Sehizoear-
dium, Balanoglossus.

I. Outer Organisation

(Fig. 455, A, B).

In the longworm-like
body, three principal
regions can be distin-
guished, corresponding
with three consecutive
sections of the coelom :

FIG. 45"..— A, Ptychodera minuta, from the dorsal side ; , ,

B, Balanoglossus Kowalevskii. After drawings by Peters these are the prODOS-

aud Minot, in Spengel's inonoyrai.1i. 1, Proboscis; 2, collar ; eidal region, the Collar

3, branchiogenital region ; 4, hepatic region ; 5, genital folds ; res-jon ancl fne trunk
(i, anus ; 7, branchial pores.

region. Compared
with the long trunk, the first two regions are short.

A. The proboscis is in the shape of an acorn (hence its German
name " Eichel "). It can be distended and contracted, and, in the
limicolous manner of life, is the chief or°;an for burrowing;.

O o

ix ENTEROPNEUSTA—BODY EPITHELIUM 563

B. The proboscis is joined to the second division — the collar, by
a short stalk or neck.

This region forms a somewhat projecting ridge anteriorly round
the neck of the proboscis, its anterior wall surrounding the neck like
a stand-up collar. The neck is only joined to the collar dorsally, for
between the neck of the proboscis and the ventral wall of the collar,
gapes the wide unarmed mouth. This leads into the buccal cavity,
which runs through the collar. The proboscis is thus a preoral
division of the body. The collar is marked off from the trunk by a
circular furrow of varying depth, over which it sometimes bulges out
posteriorly. Further, immediately in front of its posterior boundary,
there is a circular furrow round the collar itself.

C. In the long trunk, in the genera Ptychodera and ScJmocardium,
three regions can be distinguished : the branehio - genital, the
hepatic, and the abdominal regions.

1. The anterior part of the branehio -genital region, which
follows the collar, is distinguished by the branchial pores, and the
posterior part by the gonadial apertures. The gonads, however, may
stretch for some distance anteriorly into the branchial, and posteriorly
into the hepatic regions. The branchial pores are found on the
dorsal side, arranged in two longitudinal rows, or, when they are
small and circular, in two more or less deep longitudinal furrows
converging posteriorly. The pores may take the form of transverse
slits.

The genus Ptt/clwdem is distinguished by a longitudinal fold or
ridge on each side for the reception of the gonads. These two
genital folds (Fig. 455, A) can, when well developed, bend towards
one another over the back, and so form, on the dorsal side of the
branchial region, a branchial vestibule.

2. The hepatic region is distinct only in Ptychodera (Fig. 455, A)
and Schizocardium. In these genera it is distinguished by two
longitudinal rows of projecting brown or green liver-eseea. Even in
cases in which these creca do not appear to be arranged in two
longitudinal rows, it can be shown that their insertions on the body
form two such rows, but, there not being sufficient room for their
swollen ends one behind the other, many of them are pushed out
irregularly to the right or left. A mediodorsal strip of the hepatic
region always remains uncovered.

3. The cylindrical, delicate walled abdominal region tapers off
posteriorly, as a rule gradually, to the terminal anus.

II. The Body Epithelium.

The body is everywhere covered by a thick ciliated epithelium,
in Avhich, apart from the nerve-elements, undifferentiated epithelial
cells and gland cells can be distinguished. The latter are always

564 COMPARATIVE ANATOMY CHAP.

epithelial in position. The finer structure of the epithelium cannot be
dealt with here.

III. The Nervous System (Figs. 456 and 457).

The facts of fundamental importance will be treated of first.

The whole nervous system, with the single exception of a part
situated in the collar, lies in the body epithelium.

Below the surface of the whole of the body epithelium, there is
an uninterrupted layer of nerve fibres, a close continuous nerve
plexus.

The principal or main portions of the nervous system are
merely local thickenings of this network ; and are the two longi-
tudinal nerve cords, one mediodorsal and the other medioventral,
which run throughout the whole length of the trunk.

At the boundary between the collar and the trunk, the
network of nerve fibres thickens into a nerve ring, which forms a
more specialised connection between the dorsal and ventral nerve
cords.

The dorsal cord is produced anteriorly as far as to the base of
the proboscis, where it divides into two diverging branches, which
encircle that base. This circular thickening of the epithelial nerve
plexus, however, is not sharply circumscribed.

The collar portion of the dorsal cord leaves its epithelial position,
and runs longitudinall}7 through the ccelom of the collar above the
buccal cavity.

Special. — Besides smaller ganglion cells, others of remarkable size, so-called
giant ganglion cells, sometimes occur in the thickenings of the nerve plexus, especi-
ally in the collar region. Near the nerve cords and fibres the body epithelium has
few or no glands, and above the cords it is thickened. Further, all along these
cords, i.e. in the dorsal and ventral middle lines, and especially in the latter, the
body wall appears to have sunk in.

The cord in the collar, which may best be regarded as the central portion of the
nervous system, forms the so-called dorsal nerve cord of the collar. This lies in
the median line in the coelom of the collar above the pharynx, and consists of several
parts. There are two perihaemal tubes, which clasp between them the collar
portion of the dorsal blood vessel ; the nerve substance itself lies upon these, or
in a channel formed by them. The nerve cord of the collar is a thick, almost
cylindrical band, the dorsal part of which consists of cells, which, however, are not
nerve cells but may be glandular ; the ventral part, that turned to the intestine,
consists of nerve tissue, and is a direct prolongation of the dorsal nerve cord of the
trunk. Anteriorly, i.e. at the anterior end of the collar, this band divides, one
portion running into the nerve tissue of the base of the proboscis, and the other
into the epithelial nerve tissue of the circular collar ridge which surrounds this base.

In the genus Pti/fhuili-rri, the nerve cord of the collar is connected with the
epithelium of the dorsal body wall in the median plane by means of a varving
number of epithelial tubes, the so-called roots of the nerve cord. Of these, only

ix ENTEROPNEUSTA— ALIMENTARY CANAL 565

one or more of the most anterior are really hollow, containing an axial canal.
The posterior roots are solid strands of epithelium. No outer aperture of the axial
canal has been observed, although the root tissue passes directly into the body
epithelium, and the limiting membrane, elsewhere found below the epithelium, is
interrupted at the point where they join. On the peripheral, outer side of the
roots, the plexus of nerve fibres of the integument is continued into the nerve tissue
of the collar cord. The roots themselves are connected only with the dorsal layer
of cells of that cord.

The collar cord contains cavities : these are either numerous small medullary
cavities, arranged more or less exactly in two longitudinal rows, or they form one
continuous central cavity, an axial canal (Ptyehodcra) which either (in one species)
opens outward at the anterior and posterior boundaries of the collar region, or (in all
other species in which this point has been investigated) ends blindly at these points.

The axial canals of the roots of the collar cord (which run in the dorsal mesentery)
are in communication either with the axial canal of that cord or with its medullary
cavities.

IV. Sensory Organs.

Even the most recent careful investigations have not been able to
demonstrate with certainty the existence of specific sensory organs.
Undifferentiated sensory cells may be scattered over the whole of the
integument. At the posterior, and especially at the postero-ventral
part of the proboscis, and, further, at the anterior edge of the collar,
the constitution of the body epithelium is such as to make it highly
probable that it is a sensory epithelium. In one species alone,
Balanoglossus canadensis, in the postero-ventral wall of the proboscis,
there is a deep epidermal pit, which is the only structure that can,
with any probability, be claimed as a localised sensory organ.

On the sensory organs of the free-swimming larva, (/. the section
on Ontogeny, p. 586.

V. The Alimentary Canal.

The alimentary canal runs as a large and usually straight epi-
thelial tube through the body, from the wide oral aperture, at the
anterior and ventral end of the collar, to the terminal anus. It is, as a
rule, attached to the body wall by both a dorsal and a ventral mesentery
traversing the body cavity. It is developed in ways peculiar to the
different regions of the body. Especially noteworthy is the fact that,
in the branchial region, it forms a branchial intestine, communicating
by means of two longitudinal rows of branchial canals (gill-slits) with
the exterior.

A. The mouth is followed by the spacious bueeal cavity, which
traverses the collar region, and is provided with a thick epithelial
wall.

B. The roof of the buccal cavity grows out to form a diverticulum
directed anteriorly ; this runs through the neck of the proboscis,

566

COMPARATIVE ANATOMY

CHAP.

reaching as far as the base of that organ. This is the proboseidal
divertieulum of the buccal cavity (Fig. 456), and is preoral. Its
epithelial wall is a continuation of the epithelial wall of that cavity.

Special. — In the proboscidal divertieulum. a narrower posterior neck and an
anterior head or body can usually be distinguished. In section, the neck appears

•11

FIG. 45i5.— Ptychodera minuta. The proboseidal, collar, and anterior branchial regions, cut
through the middle line, and seen from the cut surface, diagrammatic (after Spengel). 1, Probos-
eidal divertieulum of the buccal cavity ; 2, ventral septum of the proboscis ; 3, skeleton of the
proboscis ; 4, buccal cavity ; 5, ventral vessel of the collar ; 6, ventral nerve cord of the trunk ; 7,
oesophagus ; 8, ridge forming the boundary between the oesophagus and the branchial intestine ; 9,
ventral blood vessel of the trunk ; 10, dorsal blood vessel of the trunk ; 11, branchial intestine,
with the gill-slits; ll>, dorsal nerve cord of the trunk; 13, dorsal blood vessel of the collar; 14,
roots of the collar cord ; 15, collar cord ; 16, proboscis pore ; 17, coelom of the collar traversed
by muscle fibres ; 18, anterior wall of the collar ; 19, nerve layer at the base of the proboscis ;
20, central blood vascular cavity of the proboscis ; 21, heart vesicle ; 22, proboseidal glomerulus ;
23, proboseidal epithelium; 24, a part of the longitudinal musculature traversing the proboseidal
cavity.

semilunar, with the concavity directed downwards. In ScJtizomrcUu.ni and Glandi-
ccps, the head is continued anteriorly into a narrow blind canal, the vermiform
process, which runs through the proboseidal cavity almost axially.

In Balanoylossus c<i/itnt/'/tsis, the neck is wanting, the head of the divertieulum
consequently forming a constricted vesicle. In other species, the continuity of the
lumen may be interrupted here and there.

ENTEROPNE USTA —ALIMENTA R Y CA NA L

567

Iii certain sections, the tissue of the proboscidal diverticulum, especially that of
its head, has a vesicular appearance which recalls that of the noto-chord of Verte-
brates. The proboscidal diverticulum of the Enteropneusta has even been called the
chorda, and has been directly homologised with this structure in Vertebrates. More
recent researches have, however, shown that the tissue of the diverticulum is an epi-
thelium in direct continuation with the intestinal epithelium of the buccal cavity.
This epithelium seems to consist of thread-like, cells, which at one point swell up to
form vesicles or vacuoles containing a fluid clear as water. The vacuoles of adjacent
cells have not room enough to lie in the same level of the epithelium. They have
to make room for one another, and thus come to lie in very different levels of the
thickened epithelium. This makes the proboscidal diverticulum appear to have a
vesicular structure, especially in tangential sections.

C. The branchial intestine. — At the posterior end of the collar,
the buccal cavity passes on into the branchial intestine, which lies in the
anterior portion of the branchio-
genital region of the body.

Here, as already mentioned,
the intestine is in communication
with the exterior through two
rows of pouch-like canals, which
will be described more in detail
later.

The numerous pouches of each
row follow one an-
other closely, and mutually flatten
one another, so that their lumina
become slit -like, and lie trans-
versely in the body at right angles
to its longitudinal axis ((/. Figs.
457, 458, 459).

5-

longitudinal

From each of these flat,
vertical, and transverse gill-
pouches, an inner aperture, the

Fin. 457.— Portion of the branchial region of
an Enteropneustan, cut down through the
middle line, and seen from the cut surface,
diagrammatic. 1, Dorsal vessel ; 2, body wall ;
3, cavity of the trunk ; 4, branchial pore ; 5, con-
tinuation of the trunk crelom in the tongue-shaped
process ; 6, branchial septum ; 7, lowest tip of the
gill-Slit, leads into the alimentary branchial tongue ; 8, wall of the alimentary canal ;
canal, and an outer aperture, the 9, ventral vessel ; 10, dorsal mesentery ; ll, ventral
branchial pore, leads to the -sentery^ 12, gill-slits ; 13, branchial tongue

exterior.

The inner aperture, the gill-slit, is as long as the gill-pouch
itself, and would have the shape of a very long O were it not com-
plicated by the formation of the tongue. The intestinal wall projects
from the upper end of the gill-slit in the form of a hollow process
down into the slit, changing the 0 into a very long vertical TJ
(Fig. 457, 12). This hollow process is the tongue. Its cavity is in
open communication with the ccelom of the trunk. It either hangs
freely down into the gill-slit (Balanoglossus, Glandiceps), or is attached
to the wall of the gill-pouch by means of rods or buds, the so-called
synaptieulse which run across the limbs of the U-shaped gill-slit
transversely, making the latter fenestrated.

568

COMPARATIVE ANATOMY

CHAP.

The partitions between the consecutive gill -pouches are called
septa, and the edges of these which are turned to the intestine are
the septal edges. If a lateral wall of the branchial intestine be
viewed from the intestinal cavity, the septal edges and branchial

1011

FIG. 458.— Ptychodera minuta, transverse section through the branchial region, .somewhat
diagrammatic (after Spengel). 1, Dorsal nerve cord ; -2, dorsal blood vessel ; 3, branchial furrow ;
4, body epithelium ; 5, gonad ; 6, longitudinal muscle layer of the integument ; 7, ventral blood
vessel ; 8, ventral nerve cord ; 0, crelom of the trunk ; 10, genital pore ; 11, branchial pore ; 12,
branchial tongue ; 13, dividing ridges ; 14, branchial septum ; 15, cavity of the branchial intestine ;
16, oesophagus.

tongues are seen regularly alternating. The septa, like the tongues,
are hollow, their cavities communicating with the coelom of the trunk.
But whereas the septal edges are continued both dorsally and ventrally
into the wall of the intestine, the edges of the tongues turned towards
the intestine, the so-called backs of the tongues, are, of course, only
in connection with the intestinal wall dorsally.

IX

ENTEEOPNE UST A— ALIMENTARY CANAL

569

The epithelial walls of the gill- pouches and of the tongues are
ciliated.

The depth (measured dorsoventrally) of the area occupied by the
gill-slits on the lateral wall of the branchial intestine varies greatly.

In all cases the gill-slits leave only a narrow strip of the intestinal
wall in the dorsal median line ; this strip is the epibranehial streak.
Ventrally they never extend so far towards the median line. They
either leave a narrow strip of the intestinal wall, the hypo-branchial
streak, which is at any rate wider than the epibranehial streak (Schizo-
cardiuni), or they only extend a very short way on to the ventral wall
(Glandiceps), or again they only reach about half way down the lateral
wall (Balanoglossux}. In the last case the hypobranchial streak

FIG. 459.— Vertical longitudinal section
through the anterior part of a row of
gills, and through a collar pore of Schizo-
cardium brasiliense (alter Spengel). rplt
Anterior aperture of the collar canal (into
the cfelom of the collar) ; cp», posterior aper-
ture (collar-pore) of the same (into the first
gill-pouch); bpi-bpe, first to sixth branchial
pores (outer apertures of the gills) ; 6si-6s5,
first to fifth gill-pouches ; b/j, life, fourth and
fifth gill-slits (apertures of the gill-pouches
into the branchial intestine) ; dmn, dorsoven-
tral musculature ; cc, ccKlom of the collar ;
)•(•, ccelom of the trunk ; c, blood vessels ;
l<; continuation of the ccelom of the trunk
into the branchial tongues ; I, branchial
tongues ; s, branchial septa ; szi, first anterior
septal bar or prong; fe, tongue bars or
prongs ; crs, septum dividing the trunk
from the collar.

V

sz

occupies the ventral or nutritive half of the branchial intestine, which
is thus more or less distinct from the dorsal or respiratory half, into
which the gill-slits open. The distinction between these two halves is
still more marked in Ptychodera (Fig. 458, 15, 16), inasmuch as they are
here separated by longitudinal ridge-like projections of the intestinal
wall, which run on each side along the boundary between the two
(13). The two ridges growing towards one another may even touch,
in which case open communication between the branchial intestine
above and the oesophagus below ceases.

The form of the outer apertures of the gill-pouches, the branchial pores, has
been described above. The furrows in which they lie correspond, in Balanoglossus,
Glandiceps and Sclii~ocardium, with the submedian line, which is indicated by the

570 COMPARATIVE ANATOMY CHAP.

interruption of the longitudinal musculature. In Ptyckodera, the branchial pores
lie mediad of this line.

The gills are, as a rule, paired, but in species of Ptychodcra, those belonging to
one side may be shifted in front of those of the other side by as much as half the
breadth of a gill.

At the posterior end of the branchial region, even in adult animals, new gills are
continually being formed.

In Ptycliochra elavigera, each gill -pouch has a long ventrally directed diver-
ticnlum.

The cavity of the branchial tongue is lined with endothelium, and traversed in
various directions by fibres, some of which, no doubt, are muscular.

The efferent section of the gill-pouches is provided with a musculature, which
cannot here be described. The pores also may be provided with an encircling sphincter
musculature of their own.

In Jlalanoglossus Kowalerskii, the posterior edge of the collar is continued back-
ward as two outgrowths, which cover the most anterior branchial pores. These
outgrowths have been called the opercula, and the small space they enclose, the
atrium.

For the blood vessels and the skeleton of the branchial intestine, see below,
pp. 584 and 580.

I). The afferent intestine, which follows the branchial intestine,
runs through the posterior gill-less part of the branchio-genital region,
and at its posterior end passes over into the hepatic or stomach intestine.
In some forms the afferent intestine is distinguished by the fact that
it sends off dorsally to right and left short canals, which open outward
on the dorsal surface. These efferent canals are known as the unpaired
intestinal pores, because they are for the most part unpaired.

Special. — These unpaired intestinal pores are found in Schizocardium Irasilicnse,
f^inidiccps Hacksii, and Gl. talaboti. In Scld. brasiliensc the openings are irregular,
either paired or unpaired. Twenty-nine in all have been observed, thirteen on the
left and sixteen on the right side, and among them seven pairs. The afferent section
of the alimentary canal is, in this species, distinguished by a -strong circular
musculature. In Gl. Hacksii, nine unpaired pores were observed in the young
animals examined, the most anterior being on the right, and the rest on the left.
In Gl. talaboti, all the pores in this region are unpaired ; in the animals examined
they are arranged in nine groups at irregular distances from one another. The
efferent canals of each group probably open into a common ampulla, which, on its
part, opens outward through a single aperture.

E. The hepatic or stomachal region of the intestine is, in all

Enteropneusta, distinguished by the fact that its epithelium is ciliated
and contains numerous globules of a secretion, usually green in colour.
This section of the intestine seems to have a musculature of its own
only in Schizocardium brasiliense, in which it is developed as a fine
layer of longitudinal fibres. The hepatic intestine is no doubt the
part of the alimentary canal of the greatest importance for digestion ;
the network of vascular capillaries, which will be described later, is
specially strongly developed in its walls.

The hepatic intestine appears as a specialised section of the

ix ENTEBOPNEUBTA—GCELOMIG SACS 571

digestive tract only in those species of Ptycliodem and Scliizocardium
in which it gives off on each side dorsally a row of finger- or wedge-
shaped outgrowths, which push out the body Avail in such a way
as to form the above-mentioned liver-caeca. The aperture of each
caBcum into the alimentary canal is a narrow transverse slit. Food
never passes into the liver-caeca. The capillary network is exceedingly
close in their walls, and the intestinal epithelium of the caeca is, as a
rule, much folded.

In Glandiceps Hachii, an accessory intestine occurs in the hepatic
region : this is a straight canal, ca. 6 mm. long, which branches off
from the median dorsal surface of the intestine proper about the
middle of the region, and again enters it at the posterior end of the

same region.

In Scliizocardium brasiliense, Glandiceps Hacksii, Balanoglossus Kowa-
levskii, and E. Merschkovskii (but not in Fti/chodem and not in E. Knpfferi,
and B. canadcnsis) paired intestinal pores, leading outward dorsally, are
found in the most anterior hepatic region, or in the region immediately
in front of it, intercalated between it and the afferent intestine. Schi.
brasiliense has one pair, Gl. Hacksii three pairs, and Balanoglossus Kowa-
levskii four to six pairs of such pores. They emerge mediad of the
siibmedian line, and may be provided with cilia and with sphincter
muscles.

F. The hepatic intestine is followed by the efferent section,
which gradually passes into the narrower rectum, this in its turn
opening outward through the anus. Where, in this section, a proper
musculature is found, it is very weakly developed.

VI. The Ccelomie Sacs and the Body Musculature.

We here use the expression coelomie sacs rather than ecelomie
cavities, the former implying that they have walls of their own.

Five ecelomie sacs occur in the body of an Enteropneustan, these
being divided among the principal regions of the body as follows :—

The proboscis contains one unpaired ecelomie sac.
The collar contains two paired ecelomie sacs.
The trunk contains two paired ecelomie sacs.

The coelomie sacs fill up almost the whole of the space between
the intestinal epithelium and the body epithelium, i.e. the seg-
mentation cavity or blastoecel of the larva, with the exception of
a system of spaces, serving as the blood vascular system, which will
be described later.

In each ccelomic sac there can be distinguished, at the least, a
visceral wall in contact externally with the intestinal epithelium, and
a parietal wall, in contact internally with the body epithelium.

Where the ecelomie sacs are paired, i.e. in the collar and the

572 COMPARATIVE ANATOMY CHAP.

trunk, the two lateral sacs come in contact with one another above
the intestine to form a bilaminar dorsal mesentery, and below the
intestine to form a bilaminar ventral mesentery.

In the adult animal these mesenteries are nowhere retained in
their full extent.

Each coelomic sac has an anterior and a posterior wall. The
posterior wall of the collar sac becomes applied to the anterior wall
of the trunk sac and thus forms a bilaminar transverse and vertical
septum, separating the eoelom of the collar from that of the
trunk.

The walls of the coelomic sacs, in the larva, are epithelial.
Throughout the greater part of these sacs, however, the epithelial
cells become transformed into muscle fibres to form the musculature of
the body and of the intestine. And this takes place to such an
extent that over large areas no endothelial lining to the body cavity
is any longer demonstrable.

Connective tissue is also produced by the Avails of the coelomic
sacs.

The musculature of the Enteropneusta consists exclusively of
smooth fibres.

Lymph cells (probably amoeboid) float in the fluid of the body
cavities : these are presumably produced by the peritoneal endo-
thelium.

A. The Ccelom of the Proboscis.

The proboscis ecelom is, as above mentioned, unpaired. The
parietal wall lies under the proboscidal epithelium, the visceral wall
envelops not only the proboscidal diverticulum of the buccal cavity,
but a complex of other organs as well, which lie posteriorly in the
base of the proboscis ; these basal organs, to a certain extent, bulge
out the ccelomic wall, like the finger of a glove, from behind forward,
into the proboscis cavity.

This cavity has three outgrowths directed backward towards the
neck, one ventral and two, a right and a left, dorsal. The left out-
growth is produced backward into a canal lined with ciliated epithe-
lium, which opens outwards through the proboscis pore. This pore
lies dorsally to, and on the left side of, the neck, at a greater or
less distance from the median line.

In a few forms (constantly in Balanoglossus Kupfferi and B. ama-
densis, and occasionally in Ptychodera minuta and B. KowalevsMi) a
second proboscis pore occurs, through which the right dorsal outgrowth
of the proboscis eoelom opens outward. This secondary proboscis
pore rises much later ontogenetically than the primary.

It has been conjectured that water is taken in through these pores
for the purpose of swelling the proboscis. There is no justification
for ascribing to them any excretory function.

The visceral wall of the proboscidal coelomic sac and, in general.

ix ENTEROPNEUSTA—GCELOMIG SACS 573

the walls of the posterior outgrowths, retain an epithelial character,
while the parietal wall develops muscle and connective tissue. This
parietal wall consists of the following parts : —

1. Close under the basal or limiting membrane of the body
epithelium, there is an outermost layer of circular muscle fibres.

2. This latter is followed by a massive layer of longitudinal
muscles, filling up the greater part of the proboscis. The very
complicated course of the longitudinal muscle fibres cannot here be
described in detail. They are stretched like the strings of an
instrument between two points of the proboscidal wall, one behind the
other, so that they cross one another in every direction.

3. Dorsoventral muscle fibres form a dorsoventral muscle
septum exactly in the median plane of the proboscis. This is,
however, not developed through the whole length of the cavity,
but reaches only as far forward as the proboscidal diverticulum of
the intestine or its vermiform process. This muscle septum thus has a
free anterior edge. The fibres of the septum, which descend from the
median line, when they reach the basal organs, diverge to right and
left, clasping these organs between them, then again uniting beneath
them, form the ventral portion of the septum. This ventral portion
is distinctly a double muscle lamella. The two constituent lamellfe
are separated by a structureless limiting lamella, which is a continua-
tion of the limiting membrane of the ventral proboscidal epithelium.
The circular muscle layer of the proboscis passes through the limiting
membrane of the ventral septum in bundles.

The ventral septum is interrupted at its most posterior part, so
that it has a free posterior edge as well as a free anterior edge.

That portion of the proboscidal cavity which is free from muscle
fibres, is to a great extent filled with connective tissue, in Avhich
irregular spaces are found as remains of the cavity. A space free
from connective tissue and varying in size is retained round the basal

organs.

B. The Ccelomie Sacs and the Musculature of the Collar.

The collar region of the body contains not only its own two
coelomic sacs, but outgrowths or processes of the trunk ccelom as
well ; these latter have been called peripharyngeal or pel-internal
cavities, and will be described with the trunk ccalom. The two
lateral coelomic sacs of the collar are, in adults, nowhere completely
separated from one another by mesenteries. The median ventral
mesentery is retained for a short distance in the posterior region of the
collar. The dorsal mesentery extends further forward, but never as
far as to the anterior end of the collar. In Balanoglossus Kupfferi, both
the mesenteries are altogether wanting.

The divisions of the collar coelom are complicated by the appear-
ance of folds in the inner or visceral wall. The two lamella of

574 COMPARATIVE ANATOMY CHAP.

these folds lie close to one another, being only separated by a limiting
membrane containing vessels. According to the courses of these
vascular folds which project into the collar coelom, the Enteropneusta
can be divided into two groups.

Group 1 — Balanoglossus, Glandiceps, Schizocardium. — A fold commences at the
posterior end of the co-lorn on each side, near the ventral median line, and ascends
diagonally in a curve anteriorly to the neck of the proboscis.

Group 2 — Ptychodera. — A medio-ventral vascular fold runs anteriorly from the
posterior end of the collar region, dividing at <i short distance from the anterior
end of that region into two folds, which ascend perpendicularly and encircle the
buccal cavity.

The Avails of the collar coelom are for the most part developed as
muscles.

1. The parietal wall consists first of an outer layer of longitudinal muscle
fibres. These commence, it is true, posteriorly in the visceral wall, then run slant-
ingly forwards and outwards towards the integument, traversing the coelom. Only
in the anterior collar region do they run close under the integument to the anterior
end of the collar. Only in this anterior region of the collar also is a circular
muscle layer developed on the inner side of the longitudinal musculature.

2. The visceral wall contains first an inner longitudinal musculature, according
to the arrangement of which the Enteropneusta may be divided into two groups.

Group 1 — Schizocardium, Glandiceps, Balanoglossus. — A bundle of longitudinal
fibres rises on each side anteriorly, from the proboscidal skeleton, which will be
described later. These muscles spread out fan-like towards the septum between the
collar and the trunk. The fibres composing this fan slope more and more the
nearer the ventral middle line they become attached to that septum. Anteriorly, the
two bundles of fibres surround the efferent vessel of the collar.

Group 2 — Ptychodera. — The numerous bundles of longitudinal fibres run
parallel to one another. Only a few of them, viz. those lying nearest the median
line run as far as the neck, enclosing the efferent collar vessel and becoming
attached to the skeleton of the proboscis. None of the rest reach the anterior
end of the collar region, but become attached to the posterior wall of the vascular
fold mentioned above as encircling the buccal cavity.

The visceral wall of the collar coelom further consists of a transverse musculature.
This also is differently arranged in the groups mentioned above, its distribution
being determined by the courses of the vascular folds.

In «roup 1 (Schiwcardium, Glandiceps, and Balanoglossus} the transverse fibres
run out laterally from the dorsal point of attachment of the ventral mesentery on
each side and then upwards, to become attached to the vascular fold above described
which runs from behind and below, anteriorly and upward.

Where the ventral mesentery is wanting, the transverse fibres run without
interruption from the right to the left vascular fold, surrounding the buccal cavity
ventrally. The transverse fibres are quite short posteriorly and limited to the
ventral side, for here the vascular folds are but commencing to diverge. Anteriorly,
as the folds gradually meet over the buccal cavity, the fibres become longer and
longer ; they finally form, near the insertion of the neck, circular loops almost
completely surrounding the buccal cavity, these loops being interrupted for only a
short distance dorsally.

In group 2 (Ptychodera} this transverse musculature is altogether wanting

ix ENTEROPNEUSTA—CCELOMIG SACS 575

throughout the larger part of the collar, viz. in all that part which lies behind the
circular vascular fold. In front of this fold, however, it is as highly developed as
in the same area in group 1. The muscles arise dorsally at each side of the pro-
boscidal skeleton, and form loops encircling the buccal cavity.

3. The anterior wall of the collar ccelom. — A strong bundle of muscle fibres
arises to right and left of the neck of the proboscis ; these radiate from the anterior
wall of the collar ccelom to the edge of the collar. Both ventrally and dorsally
the marginal fibres of these two radiating bundles usually pass beyond the median
line, cross one another, and intermingle at these points.

Besides the musculature hitherto described of the parietal visceral and anterior
walls of the collar ccelom, there are further isolated radial muscle fibres, which
connect the outer with the inner wall, and also with the anterior wall. These
fibres together form a system of crossing fibres, some running slantingly from the
outer wall inward and forward, and others inward and backward.

The cavity of the collar is filled with connective tissue, which penetrates every-
where between the muscles, leaving only certain spaces free. Such a free space is
as a rule found to the right and left in the posterior part of the collar. The collar
ccelom here is continued on each side into a canal lined with a ciliated cylin-
drical epithelium ; each canal opens through a pore (collar-pore), not on the
external surface of the body, but on the anterior wall of the first gill-pouch, near
the branchial pore. The canal projects from this pore freely forward into the
collar ccelom, and, on its outer surface, that turned to this ccelom, is covered with
plate epithelium.

As to the function of these two collar pores, we can say no more than what
was said above about that of the proboscis pore.

In Balanoglossus Kupffcri a cushion-like thickening of the epithelium is found
on each side of the body on the septum dividing the collar from the trunk. This
thickening occurs both on the anterior surface, that turned to the collar civlom,
and on the posterior surface, that turned to the trunk ccelom. These thickenings
probably function as lymph glands.

C. The Ccelomic Sacs and the Musculature of the Trunk.

The Perihsemal and Peripharyngeal Cavities of the Collar Region.

The trunk ccelom is uninterrupted throughout its whole length.
Its composition out of two lateral ccelomic sacs can still be recognised
in the adult animal, the ventral mesentery being entirely, the dorsal
partially, retained.

The ccelomic sacs of the trunk send outgrowths anteriorly into
the cavity of the collar, which push before them the wall of that
cavity ; these are the perihsemal and peripharyngeal cavities
(Fig. 460).

The perihsemal cavities are two dorsal prolongations of the trunk
ccelom, which traverse the collar region and the neck of the
proboscis as far as the proboscidal skeleton. They run below the
collar cord and above the buccal cavity. In the median line the two
cavities are separated by a structureless partition, a limiting mem-
brane, in which the dorsal vessel runs. The perihsemal spaces are
almost entirely filled by longitudinal muscle fibres formed by their

576

COMPARATIVE ANATOMY

CHAP.

dorsal walls. These are the immediate anterior continuations of the
dorsal longitudinal musculature of the trunk. In Pti/rhodera, a single
weak layer of longitudinal muscle also develops on the ventral walls
of the perihsemal cavities. In Schizocardiv/m and Glandiceps, on the
contrary, a transverse musculature is here developed. In the genus
Balanoglossus, both longitudinal and transverse muscles are wanting
in the ventral wall.

Besides the muscles which have been mentioned, there are fibres
which traverse the perihcnemal cavity transversely, chiefly in a dorso-
ventral direction.

The peripharyngeal cavities are also anterior continuations of

FIG. 460.— Ptychodera minuta, diagram of tlie collar
coelom and of the anterior region of the trunk, in an
almost median longitudinal section (after Spengel), some-
what modified. 1, Anterior wall of the collar; 2, collar
ccelom ; 3, peripharyngeal cavity ; 4, perihiemal canal ;
5, buccal cavity ; 6, septum dividing the collar from the
trunk ; 7, trunk ccelom ; 8, oesophagus.

Fio. 401.— Ptychodera minuta,
transverse section of the body through
the genital region, diagram to illustrate
the arrangement of the coslom. d, Dor-
sal ; i', ventral ; 1, dorsal mesentery ;
2, dorsal accessory chambers of the
trunk ccelom ; 3, lateral mesenteries ;
4, body epithelium ; o, parietal wall ;

6, visceral wall of the trunk coelom ;

7, intestinal epithelium ; S, intestinal
cavity ; !>, principal chamber of the
trunk coelom ; 10, ventral mesentery.

the trunk coelom, which push in between the buccal cavity (pharynx)
on the one side and the collar coelom on the other, and, in Ptychodera,
surround the buccal cavity. Anteriorly, they end dorsally at the
point where the proboscidal diverticulum of the intestine arises, and
laterally, at the points of attachment of the vascular folds. The
inner wall of the peripharyngeal cavities consists of a layer of circular
muscle fibres which surrounds the buccal cavity, closely applied to
its epithelium, and only separated from it by a limiting membrane.

In Schizocardiwm, the two peripharyngeal cavities are less exten-
sive, they lie at the sides of the buccal cavity, without surrounding
it either ventrally or dorsally. Each peripharyngeal cavity forms a

ix ENTEEOPNEUSTA—GCELOMIG SACS 577

triangle, whose sides are constituted as follows. The first (posterior)
side corresponds with its origin out of the trunk ccelom ; the second
(dorsal) with the lateral edge of the perihaemal cavity ; the third
(anterior and lower) with the line of insertion of the vascular fold.
There is a corresponding limitation and circumscription of the trans-
verse muscle fibres on the lateral walls of the buccal cavity. Never-
theless a closed muscular envelope is formed: (1) dorsally, by means
of the transverse musculature on the lower walls of the perihsemal
cavities ; (2) ventrally, by the transverse musculature of the collar
coelom.

Peripharyngeal cavities are found, not only in Ptyckodcra and Schizucardium,
but in Balanoglossui Kowalcvskii, in the same form as in Schizocardium, but pro-
vided with longitudinal instead of transverse muscles, not belonging, however, to
the inner or visceral wall, but to the outer parietal wall which is in contact with
the collar ccelom.

In Ptychodcra, a further complication occurs in the divisions of the trunk
ccelom. In the anterior hepatic and the branchiogenital regions, an accessory or
lateral mesentery (Fig. 461) occurs on each side dorsally in the submediaii line, in
addition to the two principal (median) mesenteries. This accessory mesentery runs
from the intestine to the integument, dividing the ccelom at this point into four
chambers, two large, ventral, principal chambers, and two small dorsal accessory
chambers. The accessory chambers open posteriorly into the principal chambers,
the accessory mesentery disappearing ; anteriorly, they narrow and end in the
branchial region. They are here no longer in contact with the intestine, but only
with the integument, the accessory mesenteries here shifting their visceral edges
of attachment on to the integument, in a manner which cannot here be described
more in detail.

By far the greater part of the walls of the trunk ecelom are
taken up in the formation of musculature. The parietal wall most
especially becomes differentiated into a powerful dermo-museular tube
which gradually diminishes in strength posteriorly.

The most important and constant part of this dermo-muscular
tube is the longitudinal musculature.

The longitudinal musculature, which is specially strongly developed on the
ventral side of the body, in the genital folds (where these are developed), and on
the dorsal side in the branchial region, is interrupted in the dorsal and ventral
median lines by the median mesenteries. A similar interruption takes place in
the branchiogenital region in the submedian lines, in which the gonads, and, in the
genera Balanoylossus, Glandiccps, and Sc/iizocardium the gills also, open.

By these four lines of interruption, the longitudinal musculature is divided
into two dorsal and two ventrolateral areas. (B. canadcnsis has two streaks on
each side free from muscle, and gonads open in both. )

Each longitudinal fibre runs in a curve between two points, one behind the
other, on the limiting membrane of the body epithelium. Eacli fibre thus crosses
numberless others.

In Ptychodera, in addition to these, an outer circular muscle
layer is also differentiated from the parietal wall of the trunk coelom ;
the fibres of this layer pass through the mesenteries.

VOL. II 2 1>

578 COMPARATIVE ANATOMY CHAP.

A true circular muscle layer is nowhere else developed. Such a
layer is, however, functionally replaced by pseudo-circular muscle
fibres which run on the inner side of the longitudinal musculature,
but which in reality do not form a closed ring.

In Scliizocardium, the bundles of these pseudo-circular muscle fibres run on
each side from the dorsal edge of the mediodorsal mesentery to the dorsal edge of
the ventral mesentery. Similar bundles arise near the ventral edge of the ventral
mesentery, and break up into fibres on the lateral walls of the body, ascending
along the inner side of the longitudinal musculature, traversing it, and becoming
attached to the limiting membrane of the body. In Glandiceps, this latter system
is repeated (but of course reversed) on the dorsal side.

Balanoglossus has neither the outer true, nor the inner pseudo-, circular mus-
culature.

Radial muscle fibres connect the limiting or basal membrane of
the body epithelium with the limiting membrane of the intestine
throughout the whole coelom. In the genital folds, these fibres are
stretched between opposite points of the integument. In the region
of the lateral mesenteries, similar fibres stretch between these and the
integument.

VII. The "Heart Vesicle" (Figs. 456, 21, p. 566; 464, 11, p. 583).

This is one of the names l suggested for a small closed sac which
lies upon the proboscidal divertieulum of the intestine in the basal
part of the proboscis. Its ventral wall bends down somewhat over
the divertieulum to right and left, and it is separated from the latter
by a small blood sinus. Posteriorly, towards the neck of the pro-
boscis, the "heart vesicle" is drawn out to a small tip, which is
traversed by fibres, most probably muscular, chiefly in a transverse
direction, while the rest of the vesicle contains a fluid as clear as
water. The median part of the posterior and dorsal wall is in con-
tact with the body epithelium of the neck of the proboscis.

The ventral wall is formed of a single layer of transverse muscle
fibres and pear-shaped cells, while the rest of the wall is represented
by a plate epithelium. The existence of a closed circular muscula-
ture has not yet been demonstrated.

The " heart vesicle " in Xchiwdrilit/in (and to a lesser degree in Glandiceps also)
is ] iroduced anteriorly into two large symmetrically arranged tips, the auricles.
From the posterior tip of the " heart vesicle " two bundles of muscle fibres arise which
pass anteriorly into these auricles, each one giving off fibres, one after the other, to
the wall of the auricle it enters.

It must be emphasised that the "heart vesicle" does not belong to the blood
vascular system, and does not communicate with it, but is merely in contact with
part of that system. If, therefore, tin- vesicle propels the blood, this can only

1 Morgan suggests "Proboscis vesicle." — TK.

is ENTEROPNEUSTA— LIMITING MEMBRANES 579

occur in a way similar to that seen in the lower Crustacea, where the contractions
of the intestines are able to set the body fluid in motion.

The " heart vesicle " appears, according to recent researches, to be of ectodermal
origin, and thus cannot be considered as a ccelomic vesicle.

VIII. The Limiting- Membranes, the Proboseidal Skeleton, and

the Branchial Skeleton.

Throughout the whole body of the Enteropneusta, the walls of
organs which are in contact are separated from one another by
structureless limiting membranes, which are to be regarded as
secretions of these walls. These limiting membranes must be thought
of as composed for the most part of two adhering laminae. The blood
vessels lie within the limiting1 membranes ; they represent a system
of spaces between the two laminae.

In secreting the limiting membranes, the histological character of
the secreting walls is of no consequence. A muscle wall can secrete
a limiting membrane just as well as an epithelial wall.

After the foregoing description the reader will be able to under-
stand without further assistance the occurrence and arrangement of
limiting membranes in the body. He will, for example, know that a
limiting membrane exists everywhere below the body epithelium,
secreted by that epithelium on the one hand, and by the parietal
wall of the coelom on the other.

A similar limiting membrane must also occur between the visceral
wall of the ccelomic sacs and the intestinal epithelium, as also between
the anterior and posterior walls composing the septa, which separate
collar from trunk, peripharyngeal cavities from collar coelom, etc. etc.

At certain points, especially in the proboscis and on the branchial
intestine, the limiting membrane becomes thickened, and forms the
proboseidal and branchial skeletons.

A. The proboseidal skeleton consists of a median body and two
limbs diverging backward. The body of the proboseidal skeleton lies
in the neck of the proboscis, between the neck of the proboseidal
diverticulum of the buccal cavity above, and the ventral body
epithelium of the neck of the proboscis below. The limbs diverge to
right and left into the collar region, clasping from above the entrance
to the buccal cavity, and in close contact with its epithelium.

The proboseidal skeleton is further strengthened laterally by the
ehondroid tissue which becomes attached to it. The ground sub-
stance of this tissue is identical with the substance of the proboseidal
skeleton, and of the limiting membranes generally. It is secreted by
the anterior wall of the collar coelom, and by the posterior wall of
the proboseidal coelom, or by the latter alone ; but cell processes of
these walls remain in the secreted ground substance, and these processes
may break up into cell groups or nests, which give sections of the

580

COMPARATIVE ANATOMY

CHAP.

chonclroid tissue a certain similarity to cartilage. This chondroid
tissue is most developed, forming a mass thicker than the pro-

CUxt

FIG. 402.— Gill slits and branchial skeleton of an Enteropneustan. The six hindermost
gills seen from the intestine, the three posterior in the act of forming, diagrammatic. The black
parts represent the U-shaped gill slits ; the dotted parts, the skeletal forks. 1, Branchial tongue ;
2, branchial septum ; 3, anterior prong ; 4, median or septal prong ; 5, posterior lingual prong of
a three-pronged skeletal fork.

boscidal skeleton, which always remains at its centre, in the genera

Schizocardium and Glandiceps.

B. The Branchial skeleton (Figs. 462 and 463). (Of. here pp.

567 and 568 on the gill slits, the branchial septa, and the branchial

tongues.)

The branchial skeleton here, again, consists of local thickenings of

the limiting membrane, which separates
the epithelium of the branchial intes-
tine from the visceral wall of the trunk
coelom of the branchio-genital region.
These thickenings are in the form of
upright three -pronged skeletal forks,
which are arranged on each side, in a
single longitudinal row, throughout the
whole length of the branchial region.
The number of forks corresponds with
that of the gills. The free ends of the
prong are turned dowmvards, and the
connecting piece upwards. The three
prongs of a fork are arranged as
follows. The middle prong lies in a
branchial septum, under the surface
FIG. 403.— The three anterior forks of of the septal edge, which is turned
the branchial skeleton in Baianogiossus towar(]s the cavity of the branchial in-

Kowalevskii (after Spengel). The most

anterior (I) has only U-o prongs. 1, A pos- tBStme. ThlS Septal prOIlg forks at

terior lingual prong ; 2, a septal prong (in jfcg f ree lower Clld, giving off a short

its origin double); 3, an anterior lingual anterior and & posterior branch.

The anterior prong of a fork lies

on the posterior wall of the branchial tongue, immediately in front of
the septum ; the posterior prong in the anterior wall of the branchial

u

ix ENTEROPXEUSTA— BLOOD VASCULAR SYSTEM 581

tongue, immediately behind the septum. Each fork thus has a
median septal prong, and an anterior and a posterior lingual prong.
Each tongue has two prongs, one anterior and the other posterior,
but these belong to two different forks. Each septum has only one
prong. A very minute examination shows, however, that each septal
prong consists of two fused prongs. Two-pronged forks must, there-
fore, be the ultimate elements of the branchial skeleton. Each fork
would lie with one prong in a tongue and the other in a septum.
The two septal prongs belonging to two consecutive two-pronged
forks, are, however, in every case fused together.

The most anterior skeletal fork, and it alone, has two prongs.

In the formation of the lingual prongs, the branchial epithelium
(belonging to the intestine) and the mesodermal, inner wall of the
lingual cavity (belonging to the visceral wall of the trunk coelom)
take part, but the septal prongs are secreted exclusively by the
branchial epithelium of the septal edge.

IX. The Blood Vascular System.

The blood vascular system consists of spaces in the limiting' mem-
" branes of the body. The tAvo lamella? of the limiting membranes
simply remain apart at certain points, thus forming the walls of
the vessels. An endothelium-like covering of the inner side of the
separating lamellae has only been found in Ptychodera, and. in isolated
parts in Schizocardium and Glandiceps. Nothing of the sort has been
observed in Balanoglossus. In Pt//chodera, isolated blood cells float in
the colourless blood fluid.

The lacunar blood vessels of the Enteropneusta do not arise by
the separation of the formerly contiguous lamellae of a limiting
membrane. They are, rather, persistent portions of the larval
segmentation cavity or blastoecel. The organs of the larva lie in
a spacious blastocoel, which narrows and disappears in proportion as
the organs (especially the ccelomic sacs) increase in size, these
swelling up in such a way that their walls come in contact with one
another and with the body and intestinal epithelium. Certain
cavities, however, persist, which afterwards form the blood vascular
system. The blood cells and endothelial cells of the vascular system
are, in all cases, of mesenchymatous origin.

The arrangement of the vascular system may be roughly described
as follows.

There is a capillary network in all the limiting membranes of
the body, especially in that of the integument and of the intestine.
This network is in connection with larger vessels, i.e. (1) with a dorsal
vessel which, in the dorsal mesentery, runs through the trunk and
collar and communicates with the blood vessels of the proboscis, and
(2) with a ventral vessel which, running in the ventral mesentery of

582 COMPARATIVE ANATOMY CHAP.

the trunk, receives blood from the proboscis through two lateral vessels
or vascular plexuses within the two vascular folds of the collar, these
vessels or plexuses usually uniting in the ventral median line at the
posterior end of the collar region. The dorsal and the ventral
vessels of the trunk have muscular walls, which, however, do not
properly belong to them, but are borrowed from the apposed walls of
the mesenterial portions of the coelomic sacs. In the proboscis, the
blood vascular system, by increase of its surface towards the probosci-
dal ccelom, to right and left of the basal complex of organs, gives
rise to the so-called proboseidal gill or glomerulus.

Special. — The finer details cannot be entered upon.

1. Vessels of the trunk. — While the dorsal mesentery is retained in the
abdominal part of the body (as already noted, the ventral mesentery persists
throughout the whole trunk) it may disappear in the anterior trunk region with the
exception of the part which contains the dorsal longitudinal vessel. The muscles
of the vascular trunks are transverse or circular muscles, 'and, in part at least,
continuations on to the mesenteries of the circular musculature of the body. The
ventral vascular trunk of B. Kowalevskii is provided, not with a transverse, but
with a longitudinal musculature. The musculature (which is yielded by the
mesenterial endothelium) always lies on the side of the limiting membrane
away from the lumen of the vessel, and facing the body cavity.

2. The dorsal vessel of the collar is the direct continuation of the dorsal vessel
of the trunk. It runs between the two perihremal cavities, from whose walls it
borrows its musculature. Passing out again from between these cavities, it loses
its musculature and opens, in the proboscis, into a blood sinus, the basal sinus of
the proboscis. This is a space left between various heterogeneous organs, — the
proboscis pore, the diverticulum of the intestine, the posterior tip of the "heart
vesicle," the epithelium of the neck of the proboscis.

This basal sinus communicates, on the one side, with the capillary network in
the wall of the proboscis, and on the other, through a narrow slit, with the central
blood sinus of the proboscis which lies in front of it.

3. The central blood sinus of the proboscis (Fig. 464, 9) is a space in that limit-
ing membrane which separates the "heart vesicle" (dorsally) from the proboseidal
diverticulum of the buccal cavity (ventrally). It has no musculature of its own.
This, however, is supplied by the ventral transverse musculature of the "heart
vesicle" which lies above it. In Scliizucu rilinni, the central blood sinus is con-
tinued in a peculiar manner, which cannot here be further described, on to the
two "auricles" of the "heart vesicle."

4. The proboseidal glomerulus (Fig. 464, 10) consists of two lateral principal
portions and a dorsal connecting piece. Each of the principal portions has the
form of a unilaniinar honeycomb, with deep cells. The base of the comb is formed
by the right or left lateral walls of the basal complex of proboseidal organs, i.e. of
the " heart vesicle " and the proboseidal diverticulum of the intestine. The apertures
of the single "cells," however, are turned towards the proboseidal ccelom. The
walls of the cells are formed by folds of the visceral endothelium of the proboseidal
ccelom. They are hollow, and the cavities are blood sinuses, which open into a
common cleft-like sinus in the base of the comb. This latter, again, communicates,
by means of a slit-like transverse aperture, between the "heart vesicle" and the
proboseidal diverticulum, with the central sinus of the proboscis (Fig. 464, 9).
Posteriorly, each lateral principal part of the glomerulus becomes simpler and

IX

ENTEROPNEUSTA— BLOOD VASCULAR SYSTEM

583

simpler, and its sinus, which anteriorly is so complicated, becomes on eacli side a
simple vessel in the limiting membrane between the proboscidal intestine and the
visceral ccelomic endothelium.

These two vessels are the efferent proboscidal vessels.

Blood reaches the central sinus of the proboscis in the following ways : (1)
from the dorsal vessel of the trunk and collar through the basal blood sinus,
(2) out of the integumental vascular network of the proboscis through vessels or a
vascular plexus, which ascends in the limiting membrane of the ventral septum,
and lastly (3) out of this vascular network through a vessel, which descends along
the free edge of the " heart vesicle."

The "heart vesicle" propels the blood by means of its ventral wall, which lies
upon the central sinus. Its function is considered to be that of driving the blood

FIG. 464.— Diagrammatic transverse section through the proboscis of an Enteropneustan.
1, Dorsal proboscidal septum ; 2, proboscidal epithelium ; 3, blood lacunae of the integument ; 4,
circular musculature ; 5, longitudinal musculature; 6, ventral proboscidal septum; 7, proboscidal
divert iculum of the bnccal cavity ; S, proboscidal ccelom ; 9, central blood sinus of the proboscis ;
10, proboscidal glomerulus ; 11, " heart vesicle."

through the narrow passages of the glomerulus, and through the efferent proboscidal
vessels finally into the ventral vessel of the trunk.

If the proboscis pore takes in water for swelling the proboscis, it appears
pretty certain that the glomerulus (formerly called the proboscidal gill) must,
irrespective of other unknown functions, also serve for respiration.

5. The efferent proboscidal, and collar, vessels. — From their origin out of the
two posterior tips of the lateral portions of the glomerulus, these vessels turn
ventrally, and, running very close to one another, traverse the chondroid tissue of
the neck. In the anterior and upper area of the collar region, they enter the two
vascular folds, in whose limiting membranes they run, breaking up into more or
less rich plexuses. Their courses then, naturally, correspond with those of the
vascular folds, which take their name from the vessels within them. In Schizo-
cardium, fialanoylossus, and Glandiccjjs, where the two vascular folds descend
slantingly to the ventral median line of the posterior end of the collar, they neces-

584 COMPARATIVE ANATOMY CHAP.

sarily also have slanting systems of blood vessels, which open into the anterior
end of the ventral vessel of the trunk.

In Ptychodera, on the contrary, where the vascular folds descend perpendicularly
in the anterior part of the collar region, and form a ring around the buccal cavity,
which is then continued as a medioventral fold to the posterior end of the collar,
the vessels which run in these folds naturally also fonn a similar ring, which
passes into a medioventral collar vessel continuous with the ventral vessel of the
trunk.

The vessels of the collar are distinguished from the principal vessels of the
trunk by the fact that they possess no musculature.

A circular space, running in the limiting membrane of the septum separating
the 'collar from the trunk, is in open communication with the ventral vascular
trunk.

6. The vascular capillary networks of the integument and of the intestine are
everywhere in communication with the two principal vessels. In the collar, a
connection is formed between the integumental and the intestinal plexuses by
plexuses running in the mesenteries. "Where peripharyngeal cavities are found
(Ptychodera, Schizocardium) the integumental plexus lies in the peripheral walls of
the cavities, viz. those turned to the coelom of the collar. Of all the sections of
the intestine, the hepatic is most distinguished by the closeness of its capillary
plexus and its rich supply of blood.

In some species of Ptijchodcra, dendriform, blindly-ending, vascular caeca project
from the dorsal side of the collar cord and sometimes also from the dorsal septum.

7. Lateral vessels — Ptychodera. — Two lateral vessels, provided with muscular
walls, run through the branchiogenital and hepatic regions. They originate
anteriorly out of the vascular network of the integument, run backward in the
submedian line, and enter the lateral mesenteries, in whose limiting membranes
they run. At the posterior ends of these mesenteries, at the boundary between
the branchiogenital and the hepatic regions, they pass over on to the intestine,
being continued in two vessels running along close below the liver-caeca of the
intestine, and finally open into the intestinal capillary network. The anterior
portions of the lateral vessels, which might be called the genital vessels, are
connected with the capillary network of the gonadial walls. Similar lateral
vessels also occur in ScMzocardium.

In JBalanoglossus and Glandiceps there are, usually in the hepatic region, two
lateral vascular trunks of the intestine, which open anteriorly and posteriorly into
its capillary network. Their musculature consists, in Glandiceps, of circular
fibres ; in Balanoglossus, of longitudinal fibres. These perhaps correspond with
the posterior or intestinal portion of the lateral vessels of Ptychodcra.

8. The branchial vessels. — These vessels have been best investigated in Ptycho-
i/i'rii . A branchial capillary network is found in the limiting membranes both in
the branchial tongues and the branchial septa, i.e. in the limiting membrane
which separates the epithelium of the branchial intestine from the visceral layer of
the mesoderm which lies outside it. Into this plexus, vessels having a definite con-
stant course and of large size enter : (1) a vessel along the back of each tongue, (2)
a vessel along the inner side of each lingual prong, i.e. on the side turned to the
lingual cavity, (3) a vessel along that edge of each septal prong which is turned to
tin- body wall. These last-named vessels run ventrally into the capillary network
of the lower, nutritive part of the branchial intestine (i.e. of the oesophagus),
and must be considered as efferent vessels of the branchial septa.

The branchial capillary network receives its blood from afferent branchial
vessels, which originate out of the dorsal vessel and (in Ptychodera clariycra} have
the following arrangement : Each afferent branchial vessel, soon after its origin

ix ENTEROPNEmTA— GONADS 585

out of the dorsal vessel, divides into two, one running into a tongue and the other
into the branchial septum next in front. The lingual vessel divides again into
two branches, which are continued into the two above-mentioned vessels of the
lingual prongs (" tongue-bars ").

It is not known in what way the blood is again carried out of the branchial
tongues.

X. The Gonads.

The sexes are separate in the Enteropneusta. The gonads are
simple or branched sacs of various shapes which project into the
body cavity of the trunk, towards which, however, they are completely
closed. They form on each side a conspicuous longitudinal row in
the genital region of the trunk, which, however, is not sharply de-
marcated from the branchial region in front of it nor from the
hepatic region behind it. At the posterior end of each row of gonads,
a constant formation of new gonads takes place.

The gonadial sacs open outward through simple efferent ducts
and genital pores, which always lie dorsally in the submedian line
close to, but on the outer side of, the branchial pores (Fig. 458).

These gonads, which open laterally to the branchial pores, form
the row of principal gonads, and their pores are the primary
principal pores.

A certain agreement in the number of the gonadial pores with
that of the branchial pores is sometimes found.

The arrangement of the gonads may become complicated.

A. One and the same gonadial sac may open outward through
accessory pores, which lie either medianly, or laterally, to the
principal pore.

Such accessory pores are found in Schizocardium brasiliense and
Glandiceps talaboti, in the latter in great numbers.

B. Accessory gonads may occur in addition to the principal
gonads, opening outward through secondary genital pores.

In Balanoglossus Kupfferi, such accessory gonads form a complete
row running parallel with the principal row, along its median side.
The same is the case in Glandiceps talaboti, although here the accessory
row is not quite complete. In Balanoglossus canadensis, both principal
and accessory gonads occur, there being several rows of each. The
pores of all the gonads lie in the submedian lines, which are free
from muscle, and are in this case widened into broad streaks.

When accessory gonads occiir in species of Pti/chodera (e.g. Pt.
aurantiaca, bahamensis, erythrcea) their pores always lie laterally to the
principal pores.

Structure of the gonads. — The gonads consist (1) of a wall turned to the
coelomic cavity and belonging to it, constructed of a tesselated epithelium and
fine muscle fibres, and (2) of a massive inner germinal layer, consisting of germinal
cells and covering or follicle cells ; this layer is continued into the epithelium of
the efferent ducts.

586 COMPARATIVE ANATOMY CHAP.

Between these two layers lies (3) a limiting membrane, in which a rich capillary
network may be developed, or else the membrane is divided into its two lamella
by a continuous slit-like blood sinus.

The origin of the gonads is not yet certainly known. They were formerly held to
be derived from the ectoderm, but the most recent researches seem to show that
they arise as local accumulations of the mesenchyme cells which occupy the
blastoccel. In any case the connection of the gonads with the body epithelium
by means of the ducts is secondary. They originally lie isolated between this
epithelium and the parietal layer of the ccelom.

XI. Ontogeny.

The development of the Enteropiieusta is sometimes connected with metamor-
phosis, a pelagic larva, the Tornaria larva, being developed. This larva in many
respects recalls the Bipinnaria larva of the Asteroids, and
was at first considered to be an Echinoderm larva. In other
cases development is abbreviated, and is indeed almost direct,
for though a free larva develops from the fertilised egg, it
lives at the bottom of the sea, and shows no signs of many of
the most important characters of Tornaria.

A. Structure and Metamorphosis of the Tornaria larva.

t The egg segmentation and gastrula are unknown.

Fm. -k;.-,.— Very young 1. Outer organisation. — The youngest larval stages
specimen of Tornaria observed are almost egg-shaped (Fig. 465). At the anterior

^fto^pengel).*11!, A|! lwle there is a Pair of brown eye-spots, at the posterior the
ical plate with t'yi-s ; anal aperture, and in the middle of the ventral side the
•2. in-eoral area; 3, pre- long transverse mouth. The thin transparent integument is
oral ciliated ring; thickened only in the region of two ciliated rings, which
4, oesophagus; 5 month; bord in & maimer soQn b described somewhat

t>, stomach intestine ;

7, anus; s, hind-nut; deel)enecl oral area, at whose centre the mouth lies. The
0, postoral ciliated rin^c ; ciliated rings are strictly bilaterally symmetrical. A preoral
10, postoral area ; ii, ciliated ring runs from the anterior ventral edge of the oral
proboscis pore; 12, pro- area fonvards and llpwards on each side to the frontal region,
boscidal coelom ; 13, , , .. . .

muscle to apical plate. wnere the eyes lle> and marks oft a preoral area. A second
ciliated ring runs back on each side, almost longitudinally,

from the frontal region, then bends round on to the ventral side, and here,
behind the mouth, passes into the corresponding ciliated band of the other side
of the body.

This postoral ciliated ring forms the dorsal and posterior boundary to the oral
area, and marks off a postoral area, which comprises the dorsal and posterior (anal)
regions of the larval body. The preoral and postoral ciliated rings unite for a very
short distance at the apical pole. The oral area enclosed within these two rings has
the form of a transverse ventral saddle, drawn out on each side towards the apex.

The next remarkable change which is externally visible is the appearance of a
ciliated ring at right angles to the principal axis. This surrounds the posterior
part of the postoral area, and is the principal ciliated ring (Fig. 466, 9). The
po.storal area is by it divided into an anterior and a posterior region. The posterior
region is the anal area, with the anus at its centre. In the anterior region, a
dorsal area can be distinguished from a ventral zone. Behind the principal ciliated
ring, a second weaker ciliated anal ring may appear (Fig. 466, 8).

IX

ENTEROPNEUSTA— ONTOGENY

587

During the further transformations which take place in the larva, the anal area
which is bordered by the greatly developed principal ciliated ring remains almost
unaltered, while the oral area, pushing out before it the preoral and the postoral
rings, sends symmetrical extensions (Fig. 466) into the pre- and postoral regions,
as follows : —

From the anterior and lateral tips of the oral region which stretch to near the
apical pole, two extensions, one on each side, run posteriorly and ventrally into the
preoral area (13), two more posteriorly and dorsally into that region of the postoral
area which was above distinguished as the dorsal area (2).

In this way the larva, seen from the apical pole, has temporarily a four-rayed
appearance.

From the lateral and dorsal regions of the oral area, two extensions invade a
postoral area dorsally (4). From the posterior lateral edges, two inconspicuous

c

2

PIG. 466. — A, B, C, Tornaria Mulleri (?). A, From the ventral side ; B, from the dorsal side ; C,
in profile (after Spengel). 1, Apical plate, with the eyes and apical tuft ; 2, anterior dorsal lobes of
the oral area ; 3, " heart vesicle " ; 4, posterior lorsal lobe of the oral area ; 5, collar coelom ; li, trunk
coelom ; 7, anus ; S, secondary anal ciliated ring ; 9, principal anal ciliated ring ; 10, postoral area ;
11, proboscis pore ; 12, proboscidal coelom (water sac) ; 13, anterior ventral lobe of the oral area,
14, oral area; 15, ventral "saddle"; 10, ventral zone of the postoral area; 17, anal area; 18,
cesophagus ; 19, stomach-intestine.

extensions may spread posteriorly. The ventral zone, however, bulges forwards
ventrally towards the oral area (15).

These changes bring about the peculiar indented course of the preoral and
postoral ciliated rings, shown in the figures.

The ciliated rings may even become still more folded. Such folding reaches
the highest degree in Tornaria Grcnachcri, hardly 1 cm. long, in which the
anterior ventral and the anterior dorsal extensions of the oral area bulge out at
the cilia-carrying edge to form numerous long, narrow, freely projecting accessory
lobes, resembling tentacles.

In the frontal region, on the apical eye-bearing plate, which here becomes
differentiated, a tuft of delicate immobile cilia develops early.

The larva swims in such a way that the anterior or apical pole is directed
upward and the anal pole downward.

The metamorphosis of the Tornaria larva into the young Enteropneustan is
accompanied by the following external processes :—

588 COMPARATIVE ANATOMY CHAP.

The body lengthens and its preoral section becomes produced into the proboscis,
at whose tip the eyes are still long visible, until they degenerate together with the
apical plate and tuft.

The preoral and postoral ciliated rings degenerate, but the whole body becomes
covered with cilia.

The principal ciliated ring (Fig. 467) persists for some time, eventually, i.e.
when the anal area has increased in length, surrounding the body about half
way between the mouth and anus. A circular furrow between it and the base
of the proboscis is the first indication of the posterior boundary of the collar
region.

The whole ectoderm of the oral area degenerates during metamorphosis, the
body epithelium proceeding exclusively from the ectoderm of the preoral, postoral,
and anal areas, which increases in thickness. This phenomenon, together with the
lengthening of the larva, causes a very marked diminution in the transverse section
of the body, an essential accompaniment of which is the approximation of the
larval integument to the parietal walls of the ccelomic vesicles.

2. Anatomical. — The apical plate of the Tornarin larva, which completely
degenerates later, consists of a dorsal and a ventral half. Below the surface of the
specially thick dorsal half there is a layer of nerve fibres.

The centre of the apical plate is formed by a small group of long narrow
sensory cells carrying delicate immobile cilia (sensory hairs).

The two eyes which are embedded in the apical plate are optic pits, whose
floors are formed by cells which are pigmented at their bases (retinal cells ?). The
optic pit is filled with a clear substance, which is a continuation of the cuticle
covering the apical plate, and is called the lens. The apertures of the two pits
diverge anteriorly and laterally. Round them in the apical plate theie is a
deeper layer of elements, which are considered to be ganglion cells.

The alimentary canal. — Even the youngest larvae observed showed the division
of the alimentary canal into oesophagus, mid-gut or stomach, and hind-gut, which
is characteristic of all the larval stages.

The oesophagus ascends vertically. It is a flattened tube provided with a
circular musculature, its thickened dorsal and ventral walls being ciliated.

The stomach is a large egg-shaped sac, whose axis lies horizontally, and into
whose anterior pole the oesophagus enters. The epithelial cells of the sac are at
first low, but they lengthen at a later stage. The originally thin-skinned stomach
develops in this way a thick wall, which is probably non-ciliated except at two
points. A ciliated cushion is found ventrally, at the entrance to the stomach, and
the efferent aperture is surrounded by long hairs, which perhaps act as a fish-trap
apparatus.

The hind-gut, in the youngest larvse, is an almost cylindrical tube with thin
walls. At a later stage its anterior part becomes swelled up, so that the hind-gut
as a whole becomes funnel-shaped or conical. As, however, the aperture com-
municating with the stomach remains small, the wall of the hind-gut is applied
over a considerable area to the posterior wall of the stomach. The aperture lies at
the centre of this area. Immediately in front of the anal aperture there is a circle
of cells provided with cilia.

Formation of the gills. — The formation of the first pair of gills takes place
shortly before or after metamorphosis. Two lateral cteca arise at the posterior end
of the oesophagus, grow out towards the integument, and break out laterally and
dorsally through the external branchial pores. The cesophageal apertures of these
branchial diverticula are at first round, later they become U-shaped, by the lateral
dorsal intestinal wall growing out into the diverticulum. This outgrowth is the
rudiment of the branchial tongue.

ix ENTEROPNEUSTA— ONTOGENY 589

In Tornaria Agassizi, it was observed that the pores of the second pair of gills
arise earlier than those of the first.

In other cases, the new pairs of gills always form behind the last produced.
During the process, the growing oesophagus, the posterior part of which is now the
branchial intestine, passes from a vertical to a horizontal position.

Even in the adult animal, new gill-pouches are constantly being formed at the
posterior end of the branchial region. The course of development is always the
same as that of the first gills in the larva.

The collar canals, with their pores, develop shortly before the rudiments of the
second pair of gills arise, most probably as outgrowths of the anterior wall of the
first gill-pouch, near its external pore. These outgrowths run forwards towards the
collar ci.elom.

The first rudiment of the proboscidal diverticulum of the buccal cavity has
been observed in Tornaria Agassizi, as a small bulging of the anterior wall of the
larval oesophagus directed anteriorly, and lying immediately above the mouth.

The proboscidal ccelom. — The rudiment of the proboscidal ccelom has not as
yet been observed with as much certainty as could be desired. According to one
observer, it arises as an outgrowth of the intestine at the boundary between the
oesophagus and the stomach.

In the youngest stages which have been closely observed, the proboscidal ccelom
(the so-called water sac of the larva) is an almost cylindrical tube lined with
tesselated epithelium, the inner half of which is slightly widened. This tube
becomes attached by its inner end to the anterior wall of the oesophagus, near
the point at which the latter opens into the stomach. Here it sends two processes
(the reins) to right and left on to the lateral walls of the oesophagus, on which
it appears to ride. From the oesophagus, the tube, traversing the blastoccel,
ascends almost vertically. Shortly before it reaches the dorsal ectodermal wall, it
is continued into a short internally ciliated tube, the epithelium of which suddenly
changes its character. This short tube, lined with cylindrical epithelium, is the
rudiment of the proboscidal canal, and the pore to which it leads is the proboscis
pore.

The proboscidal coelom is further attached to the apical plate by means of a
strand in the manner illustrated in Fig. 466 ; the strand consists of contractile
fibres surrounded by a nucleated envelope, a continuation of the wall of the water
sac. The contractile fibres of this strand are continued on to the tips or reins of
the water sac, which grasp the oesophagus between them.

The further fate of the water sac, briefly stated, is as follows. It swells up,
and soon changes from a tube into a vesicle, the greater part of the wall of which,
during metamorphosis, becomes applied, as parietal wall of the proboscidal ccelom,
to the body epithelium of the anterior or proboscidal section of the body, the
muscular apical strand becoming shorter and shorter, and finally altogether
disappearing.

The manner in Avhich the epithelial walls of the water sac become differentiated
into the proboscidal musculature cannot here be described.

" Heart vesicle." — Authorities differ as to the first rudiments of this organ. In
the youngest stage investigated by the most recent observers, it is a small cellular
structure with an internal cavity lying so close to the ectoderm that it may be
either ectodermal or mesenchymatous. It appears to the right, in front of and
near the proboscis pore. The body becomes a hollow vesicle, leaves the ecto-
derm, sinks below the surface, and becomes applied to the right side of the water
sac. Transverse muscle fibres develop on its ventral wall. The water sac then forms
posterior outgrowths, which grow round the "heart vesicle'' on its right dorsal
and ventral sides. The two dorsal posterior and the ventral posterior sections

590

COMPARATIVE ANATOMY

CHAP.

of the proboscidal ccelom thus come into existence. The " heart vesicle " on its
lower side, however, always remains separated from the dorsal wall of the ventral
posterior outgrowth of the water sac by a space, into which the proboscidal
diverticulum of the buccal cavity grows from behind, but in such a way that
between it (the diverticulum) and the superimposed "heart vesicle" a space
remains, which appears to become filled with blood at an early stage. This is the
central blood sinus of the proboscis.

The ccelomic sacs of the collar and trunk. — These two pairs of coelomic sacs
appear, in Tornaria, to have a common rudiment in the following way. The
edge formed by the anterior wall of the hind-gut, bending round backwards into its
lateral walls, is produced anteriorly to right and left as hollow sacs, or in other cases

as a pair of solid bilaminar plates. These become
applied to the stomach, but are on the other
hand separated by a large space from the ecto-
derm. These two sacs or plates become con-
stricted from their matrix, the hind -gut, and
grow round the stomach dorsally and ventrally.
In each, apparently, an anterior portion becomes
constricted off. This anterior pair of sacs or
plates is the rudiment of the collar coelomic
sacs, the posterior, which only secondarily
extend backwards along the sides of the hind
gut, is the rudiment of the trunk coelomic sacs
(Fig. 467). These two cceloms are therefore
enteroccels. Where the first rudiments of the
cceloms are solid bilaminar plates, a space soon
arises in them by the separation of the two
laminae. These small spaces, whether present
from the first, or formed later, begin to increase
in size at the end of the larval period. The
two pairs of ccelomic rudiments become vesicular.
FIG. 4G7.-Collar and trunk of an The Quter wall becomes applied to the body
Enteropneustan (Tornaria Krohni)

immediately after metamorphosis, from

epithelium as the transverse section of the

the ventral'side (after Spengel). i, Pro- growing larva decreases during metamorphosis
boscis ; 2, collar ; 3, trunk ; 4, collar in the way already described, and forms the
ccelom ; 5, septum between the collar dermomuscular tube. The inner wall lies upon

and the trunk ; ti trunk coelom ; 7, veil- ,,-,,- i j-i • 11

the intestine, and represents the visceral layer
tral mesentery ; 8, principal ciliated L

ring; 9, wall of the mid-gut; 10, wall of the ccelomic sac. The dorsal and ventral
of the hind-gut ; 11, anus. mesenteries are formed where the right and

left trunk coelomic sacs and the right and left

collar ccelomic sacs, in surrounding the intestine, come in contact dorsally and
ventrally in the median plane.

These processes, of course, go hand in hand with a progressive reduction of the
blastoccel, which contains a number (small at first, but increasing later) of
mesenchyme cells of unknown origin. The remains of the segmentation cavity
represent the blood vascular system.

Nervous system. — Shortly before the conclusion of metamorphosis, the two
longitudinal nerve trunks arise as local differentations of the body epithelium,
below the surface of which a layer of nerve fibres forms. The collar cord, also, at
first lies superficially in the integument, and is nothing more than the collar
portion of the dorsal epithelial longitudinal cord. This part only sinks below the
surface at a later stage. According to recent observers, the process recalls thr
sinking in and constriction of the neural tube in Vertebrates.

ix ENTEROPNEUSTA— PHYLOGENY 591

Gonads. — The development of the gonads has already been sufficiently described
above, p. 586.

B. The almost Direct Development of the Balanoglossus Kowalevskii.

We can only select a few of the principal points in this development for
description. Segmentation is total and equal, and leads to the formation of a
cceloblastula, out of which, by invagination, a ccelogastrula is produced. This
latter becomes covered with cilia, and a ciliated ring forms round the. blastopore,
which diminishes in size and finally closes ; this ring corresponds with the principal
ciliated ring of the Tornaria larva. At this stage the larva leaves the egg to live
at the bottom of the sea, without showing any trace of the form, or of the charac-
teristic ciliated rings, of Tornaria.

The differentiations which take place in the archenteron are important. Its
anterior part becomes constricted off as a semilunar vesicle lying transversely. This
takes up the whole of the most anterior part of the blastocLel and becomes the
proboscidal ccelom, which thus, according to these observations, is an enteroccel.
Two pairs of lateral outgrowths become constricted otf from the rest of the archen-
teron, the anterior being the rudiment of the collar ccelom, and the posterior that
lit' the trunk ccelom.

The blastociel is small from the first.

The mouth is said to arise by the simple breaking through of the intestine out-
wards, and the anus in a similar way, in the place of the original blastopore. Thus
the whole of the intestinal wall is of an endodermal origin.

The developmental processes in B. Kowalevskii cannot here be further described :
we refer the reader to the account of the formation of the organs in the Tornaria
larva, given above.

XII. Phylogeny.

The systematic position of the Enteropneustan must still, or rather again, be
considered as altogether uncertain. In any case, the Eiiteropneusta are not closely
related to any single large division of the animal kingdom. Special affinities with
the Chordata, the Echinodermata, and the Nernertines have been long suggested,
and in quite recent times also with Cephalodiscus and Rhabdopleura.

A. The relation of the Enteropneusta to the Chordata has been maintained
on the following grounds :—

1. The Chordata and the Enteropneusta show a very far-reaching and extra-
ordinary agreement in their gills. This agreement holds good even in details
(branchial tongues, branchial skeleton, synapticulse) if the gills of Amphioxus are
taken for comparison.

2. The proboscidal diverticulum of the Enteropneusta is, in structure and origin,
comparable with the chorda of the Chordata.

3. The proboscidal skeleton of the Enteropneusta corresponds with the sheath
of the chorda.

4. The body cavities in the two groups are of enteroceelomic origin ; the pro-
boscidal ccelom corresponds with the anterior unpaired mesoderm vesicle of
Amphioxus.

i>. The collar cord of Balanoglossus corresponds with the dorsal cord of the
Chordata, and arises in the same way as the neural tube of Vertebrates, by sinking
in and covering over.

The most recent researches have, however, yielded resxilts unfavourable to this
assumed homology.

592 COMPARATIVE ANATOMY CHAP.

1. The great similarity of the Enteropneustan gills to those of Amphioxus iu
finer structure remains, but the detailed comparison of point with point makes a
real homology doubtful, and seems to oblige us to consider the resemblance as at
the most a very remarkable case of convergence.

Further, the following considerations must also be taken into account. The
gills of Amphioxus arise ontogenetically as segmental structures, while those of
the Enteropneusta, although standing in a longitudinal row, belong to the unseg-
mented trunk.

The gills of the Chordata receive their blood from the ventral vascular trunk,
those of the Enteropneusta from the dorsal trunk.

Should it be proved that the larval oesophagus of the Enteropneusta proceeds
from an ectodermal invagination, and is a stomodseum, then the gills of the
Enteropneusta would lie in an ectodermal intestinal region, in contradistinction
to those of the Vertebrata, which belong to an endodermal intestinal region.

2. The proboscidal diverticulum is a preoral outgrowth of the wall of the buccal
cavity, and is lined with epithelium. It does not thus agree with the tissue of
the noto-chord. It is further very questionable whether the buccal cavity,
and, with it, the diverticulum are endodermal formations. The proboscidal
diverticulum lies below the continuation of the dorsal blood vascular trunk (below
the central blood sinus of the proboscis) ; the chorda of Vertebrates, on the con-
trary, lies above the dorsal blood vascular trunk (aorta). K"o homology between
the two is possible.

3. The proboscidal skeleton, as a thickened limiting membrane, could at the
most be compared only with the inner cuticular sheath of the chorda.

4. The body cavity is an enteroccel, in many different divisions of the animal
kingdom (either constantly or exceptionally) other than the Enteropneusta and the
Chordata. The ccelomic vesicles (mesoderm vesicles, primitive vertebrae) of the
Vertebrata show a segmental arrangement, corresponding with the metamerism of
the other organs, while in the Enteropneusta no such arrangement exists.

5. The collar cord of the Enteropueusta is only the anterior continuation of the
dorsal nerve cord of the trunk. It does not sink below the surface until all its
parts are formed. The corresponding ventral nerve cord of the Enteropneusta does
not exist anywhere in the Chordata.

The following further points must be emphasised.

The gonads of Amphioxus arise segmentally from the endothelium of the body
cavity, while the rows of gonads in the Enteropneusta lie in an unsegmeuted
region. The first origin of the gonads of the Enteropneusta is, indeed, unknown,
but their rudiments are found in the blastoccel very early. The manner in which
the genital products are ejected iu the two groups is altogether different.

In the Chordata, the blood in the dorsal vessel flows from before backward, in
the ventral from behind forward ; the reverse is the case in the Enteropneusta.

A comparison of the two collar pores in the Enteropneusta with the most
anterior pair of nephridia in Amphioxus could only be of value were the develop-
ment of these organs known. In all probability the former are of ectodermal and
the latter of mesodermal origin.

There is nothing we know of in the Chordata comparable with the Tornaria
larva.

These considerations render any relationship between the Enteropneusta and
the Chordata, at least at present, highly improbable.

B. The relationship between the Enteropneusta and the Nemertina is so very
problematical that it cannot here be discussed.

C. Eolation of the Enteropneusta with the Annelida. — The attempt to bring
the Enteropneusta and the Annelida into even a distant genetic relationship is

ix ENTEROPNEUSTA— PHYLOGENY 593

supported chiefly upon a comparison of larval forms. The, following characteristics
of the Trochophoran and the Tornarian larvte have been pointed out.

The neural plate, the apical sensory organs, and the muscles which become
attached to the neural plate correspond in the two. The divisions of the intestine
also agree, that is, if the fore-gut of the Tornaria is an ectodernial stomodaeum,
and the hind-gut a proctodreum. The two pairs of ccelornic sacs are comparable
with the two anterior pairs of mesoderm vesicles.

The comparison of the ciliated bands presents difficulties. Three ciliated rings,
a preoral (surrounding the apical area with the neural plate), a postoral, and a
preanal, are typically ascribed to Trochophora, the last of which is supposed to
correspond with the principal ciliated ring of Tornaria, while the preoral and
the postoral rings of Trochophora are wanting in Tornaria. And it is argued
that the absence of these rings in Tornaria has led to the specialisation of the
preanal as the principal ciliated ring.

On the other hand the preoral ciliated ring of Tornaria cannot be compared
with the preoral ring of Trochophora, because it does not surround the apical
plate. This latter lies, on the contrary, outside the frontal area, at the opposite
end of the oral region to the mouth ; the preoral ring passes in front of it.

Compared with the principal ciliated ring, the pre- and postoral rings are,
perhaps, of small morphological significance, since they are wanting in the non-
pelagic larva of Balanoglossus Kowalcvskii, whereas the principal ring occurs
in it.

In a comparison of the Tornaria with the Trochophora larva, the great import-
ance of the following differences must not be overlooked.

1. Trochophora possesses typically one pair of primitive kidneys, which are
wanting in Tornaria.

2. Tornaria has a preoral ccelom, which is absent in the T roclioplioru, larva.

If now we turn to the organisation of the adult animals for light as to this
question of the relationship between the Annelida and the Enteropneusta, we find
immediately that insurmountable difficulties stand in the way of any close com-
parison. Only in the blood vascular system is any fundamental agreement found.
The blood in the dorsal and ventral longitudinal vessels has the same course in
both groups. But, on the other hand, a comparison of the nervous system of the
Enteropneusta with that of the Annelida encounters difficulties similar to those
found in comparing it with that of the Vertebrata. The gills in the two groups
are altogether heterogeneous. The typical Aunelidan kidneys are wanting in the
Enteropneusta, for the collar pores, which are probably of ectoblastic origin, can
hardly be regarded as a pair of uephridia.

Thus the relationship of the Enteropneusta to the Annelida appears at the
best to be extremely distant.

D. The relation of the Enteropneusta to the Echinodermata. — This relation-
ship is claimed on the ground of the agreement existing between the larval forms.
The similarity of the Tornaria, especially with the Bipinnaria larva of the
Astcroidca, is, indeed, so striking that the first observer of Tornaria expected
for certain that it would develop into an Echiuoderm.

A closer comparison yields the following results :

1. If we place the Tornaria and the Bipinnaria similarly with regard to the
position of the mouth and the anus, a striking agreement in the conformation
of the regions of the body and in the ciliated rings bordering them is observed
(Fig. 468). In both we find a separate preoral ciliated ring in the same position,
bordering a preoral area. In both we can distinguish, lying behind this, a
deepened oral area with the mouth in its ventral centre. The postoral ciliated
ring of the Tornaria corresponds with the large circumoral ring of the Bipinnaria.
VOL. II 2 Q

594

COMPARATIVE ANATOMY

CHAP.

The region of the body lying behind this occupies in both a large part of the dorsal
posterior side of the body ; in it lies the anus.

2. Whereas, in Tornaria. in the postoral area of the body, a large ciliated
ring bounds an anal area, such a ring is \vanting in Bipinnaria.

3. The apical plate with the two eyes and the tuft, which is so sharply marked
in Tornaria, is wanting in the developed £ijiininn-ia. Too great a significance
must not now, however, be attributed to this fact, since something like a neural
plate (an ectodermal thickening with long cilia) has been observed in the quite
young larvse of Asteroids and Echinoids, and a neural plate with, a layer of nerve
fibres, ganglion cells, and ciliated tuft, although without eyes, has been demon-
strated in the apical region of the Antedon larva.

4. The intestine of Tornaria shows the same divisions as are found in that
of the Echinoderm larvse, viz. : cesophagus, stomach - intestine, and hind -gut.
Whether, however, these three sections correspond with one another in the two

8

HI

FIG. 408.— A, B, C, Auricularia, Bipinnaria, and Tornaria (Enteropneustan larva), from the
right side, diagrammatic. 1, Preoral area ; -2, oral area ; 3, postoral area ; 4, anal area ; I, preoral :
II, circumoral ; III, anal or principal ciliated ring ; 5, neural plate ; o.s, mouth ; on, anus.

groups must remain uncertain so long as the origin of the cesophagus and hind-
gut is not definitely ascertained. In this matter the Echinoderms are in the same
position as the Enteropueusta.

The cesophagus, in the Echinoderms, is sometimes described as an ectodermal
stomodreum, sometimes as a section of the archenteron. The latter, according to
the most recent investigations, is the case, e.g., in the Holothurioidea, and the
former in Antedon and others. In this case the cesophagus, even within the
Echinodermata, is not an homologous structure ! In the case of the Enteropueustn,
in the interest of other views, doubt has been thrown upon the statement that the
larval cesophagus is a part of the archenteron.

The hind-gut in the Echinodermata is, by all authorities, considered to be
endodermal. The same was affirmed of the hind-gut of the Enteropueusta, but
this has recently been strongly doubted.

5. The condition of the cceloni in the two larval forms would show great agree-
ment if two pairs of ccelomic sacs can really be attributed typically to the Echino-
derm larva, a point which recent research makes more and more probable, and if
also the endodermal origin of the ccelomic vesicles of the Euteropneusta could be
proved.

It would then be evident that the two anterior ccelomic sacs of the Echinoderm
lame (the left of which is the hydroccel) correspond with the two ccelomic sacs of

ix ENTEROPNEUSTA— LITERATURE 595

the collar of the Enteropneusta, and the two posterior sacs of the former with the
two trunk coelomic sacs of the latter. The two anterior sacs are in communication
with the exterior through the collar pores ; this communication (hydropore, water
pore) in the Echinodermata is usually limited to the left anterior sac, i.e. to the
left hydrocoel vesicle, but occasionally in Asteroids — a matter of great importance
— appears on the right side as well.

From these considerations, it seems that the prospect of establishing a funda-
mental agreement in structure between the Enteropneustan larva and that of the
Echinodermata is very hopeful. This relationship between the Enteropneusta and
the Echinodermata seems to rest upon more solid ground than do any of the others
which have been attributed to either of these two groups.

At the same time any attempt to compare adult Echinoderms with adult
Enteropneusta is at present completely futile. The Echiuoderms and Entero-
pneusta could, as far as we can see, only be genetically connected through some
common racial form far back in their phylogeny — a form which corresponded with
the Tornarian and the Dipleurulan larva.

Further, before we can feel any certainty on these questions of affinity, new and
more exact ontogenetic researches must be made. The origin of the proboscidal
coelomic vesicle of the Enteropneusta has to be established, as has also that of the
"heart or proboscis vesicle." Attention must be directed to the question as to
whether a preoral section of the body corresponding with the proboscis of the
Enteropneusta is present (if only as a rudiment) in the Echinoderm larva, as for
instance in the preoral section of the body in the Antedon larva (?), or in the larval
organs of Asterina and other Asteroids (?). With reference to the " heart vesicle "
we are reminded of the statement that a "pulsating vesicle" occurs, apparently
not of enterocoelomic origin, in Echinoderm larvse. This has to be confirmed.

The Relationship of the Enteropneusta to Cephalodiscus and Rhabdopleura
will be considered in the Appendage to this chapter.

Literature.

Alex. Agassiz. The history of Balanoglossus and Tornaria. Mem. Amer. Acad.

u f Arts and Sc. Vol. IX. 1873.
W. Bateson. The early stages in the development of Balanoglossus. Quart. Jonrn.

Microsc. Se. (N.S.). Vol. XXIV. 1884.
The later stages in the development of Balanoglossus Kowalevskii, with, a

suggestion on the affinities of the Enteropneusta. Quart. Journ. Microsc. Sc.

(AT.S.). Vol. XXV., Suppl. 1885.

- Continued account of the later stages in the development of Balanoglossus
Kincalcvskii and on the morphology of the Enteropneusta. Quart. Journ.
Microsc. Sc. (N.S.). Vol. XXVI. 1886.

- The ancestry of the Chordata. Quart. Journ. Microsc. Sc. (N.S.). Vol.
XXVI. 1886.

Gilbert C. Bourne. On a Tornaria found in British Seas. Journ. Mar. Biol.

Assoc. (2). Vol. I. 1889.
R. Kohler. Rechcrches anatomiqucs sur une nouvelle espece de Balanoglossus.

Bull. Soc. Sc. Nancy (2). Tome VIII. 1886. An almost identical work in

Internal. Monatsschr. Anat. Hist. 3 Bd. 1886.
A. Kowalevsky. Anatomic dcs Balanoglossus dclle Chiajc. Mem. Acad. I/nji.

Sc. St. Petersboury (7). Tome X. 1867.
A. Krohn. Beobachtungcn iiber Echinodcrmenlarven. Arch. f. Anat., Pliysiol.

u. wisscnsch. Mai. 1854.

596

COMPARATIVE ANATOMY

CHAP.

A. F. Marion. Etudes zoologiques sur deux especes d'Enteropneustes. Arch. Zool.

g£ner. et exper. (2). Tome IV. 1886.
E. Metschnikoff. Untersuchungen uber die Metamorphose cinigcr Secthiere. 1.

Ucbcr Tornaria. Zeitschr. f. wiss. Zool. 20 Bd. 1870.
T.H.Morgan. Grou-th and metamorphosis of Tornaria. Journ. Morph. V. 1892.

The Development of Balanoglossus. Journ. Morph. Vol. IX. 1894.

Job. Miiller. Ucbcr die Larvcn und die Metamorphose dcr Eehinodcrmtn. Part 2.

Akad. d. Wissensch. 1848. Berlin, 1850.
Wladimir Schimkewitsch. The Fauna of the White Sea : Balanoglossus Mcresch-

korskii Wagner. St. Petersburg, 1889. (In Russian.)
J. W. Spengel. Die Enteropneusten. Fauna and Flora des Golfes von Neapcl.

18 Monoyraphic. Berlin, 1893. The most important recent work.
W. F. R. Weldon. Preliminary note on a Balanoglossus larva from the Bahamas.

Proceed. Roy. Soc. London. Vol. XLII. 1887.
R. v. Willemoes-Suhm. Biologische Beobachtungcn uber niedere Meeresthiere. 4.

Ucbcr Balanoylossus Kupfferi aus dem Oeresund. Zeitschr. f. wissensch. Zool.

21 Bd. 1871.
A. Willey. Amphioxus and the ancestry of the Vertebrates. 1894.

Appendage to the Enteropneusta.

Cephalodiseus and Rhabdopleura.
I. Cephalodiseus (Figs. 469-471).

The body is about 1 mm. long, almost bean-shaped, bilaterally

symmetrical ; it is rounded
posteriorly and anteriorly
flat, with a slight backward
slope. The most important
organs which can be distin-
guished externally are found
in this anterior sloping sur-
face, while in the whole of
the rest of the body only
one organ appears, viz. a
cylindrical stalk or pedicle,
which rises from the ventral
side of the rounded posterior
end of the body.

In that part of the body
which projects most anteri-
orly, i.e. the anterior end
of the dorsal side, lies the
anus, while somewhat be-
FIO. 4titi.— Cephalodiseus dodecaiophus, from the hind the anterior end of

ventral side (after M'Intosh). 1, Tentacles ; 2, buccal the ventral side, the HlOUth,

shield = proboscis ; 3, mouth ; 4, buds ; 5, pedicle ; 6, trunk.

and between
slope mentioned above, the inter -stomatal region.

the two the
The median

IX

GEPHALODISGUS

597

line between the mouth and anus is . the inter- stomatal middle
line.

In the inter-stomatal region or in its immediate vicinity lie the
following parts of the body : (1) the preoral buccal shield with its
two pores; (2), the central nervous system; (3), the twelve feathered
tentacles ; (4), the apertures of the two oviducts ; (5), the postoral

15

Fio. 470.— Median section through Cephalodiscus dodecalophus somewhat near the median
plane (after Harmer). 2, Nervous system ; 3, anterior paired ccelom (collar coelom) ; 5, fold of the
anterior trunk region ; 8, paired trunk ccelom; 9, pharynx ; 10, oesophagus ; 11, stomach-intestine ; 12,
hind-gut ; 13, buccal cavity ; 14, pedicle ; 15, ovary ; 16, anus ; 17, oviduct ; 18, buccal shield ; 19,
coelom of the same = proboscidal ccelom ; 20, one of the proboscis pores ; 21, anterior diverticulum
(proboscidal diverticulum) of the pharynx.

lamella; (6), the two postoral pores of the body cavity; (7), the
two gill-slits (Figs. 470, 471).

The buccal shield, which is comparable with the proboscis of the
Enteropneusta, is a plate, which projects downwards from the inter-
stomatal region immediately in front of the mouth by means of a
rather short stalk, in such a way that its free surface, the epithelium
of which is enormously thickened, faces ventrally (Fig. 470, 18).

The central nervous system lies in the hypodermis, almost at

598

COMPA EA TIVE ANA TOMY

CHAP.

the centre of the inter-stomatal region. If we describe the mouth
as lying posteriorly and ventrally to the stalk of the buccal shield,
the central nervous system lies exactly opposite, that is, posteriorly
and dorsally to the same structure.

Twelve tentacles rise dorsally at the base of the stalk of the
buccal shield, six to the right and six to the left of the central nervous

system. The nervous system
extends into the dorsal hypo-
dermis of these tentacles (Fig.
471). The tentacles are large,
are feathered on both sides like
the feathers of birds, and are
knobbed at their free ends.

Two pores, placed symmet-
rically to the inter-stomatal
median line, break through
the most anterior part of the
central nervous system. There
is thus open communication
between the body cavity of the
buccal shield (the proboscis)and
the exterior (proboscis pores).
Between the central nervous
system and the anus there is on
each side an aperture. These
two apertures belong to the
oviducts.

Outside of the inter-
stomatal region, but in its
immediate proximity, the
following parts are found.

Immediately behind the
mouth, covered by the oral
disc, a thin lamella hangs ven-
trally and laterally down from
the body like an apron ; this
is the postoral lamella (Fig.
471, 5). In the posterior

FIG. 471.— Horizontal section through Cephalo-
discus (after Harmer). 1, Tentacles ; 2, nervous
system ; 3, anterior paired body cavity (collar coelom) ;

4, collar pore ; 5, folds of the anterior trunk region, the

postoral lamella ; 6, branchial pores ; 7, mesentery ; angle formed by this lamella

5, paired trunk coelom; 9, pharynx; 10, oesophagus: -TV ,v i i • f j
11, stomach : 12, hind gut. Wlth the body> a Pore 1S found

on each side (collar pore).

These pores lead into the paired middle body cavity, of which we shall
speak later. Immediately behind these pores, and like them over-
arched by the lateral folds of the postoral lamella, two branchial pores
lead from without into the pharyngeal section of the intestinal canal.
Musculature. — From near the mouth, longitudinal muscle fibres
run back along the ventral side, and enter the pedicle. Muscles are

ix CEPHALODISCUS 599

also found in the stalk of the buccal shield, which radiate out from
the stalk into the shield.

Body cavity. — In a young bud, the body appears to be divided
into three sections (anterior, middle, and posterior) by two circular
furrows. Each of these three sections possesses a separate body
cavity. The most anterior section, out of which the buccal shield
proceeds, has an unpaired body cavity, into which a short intestinal
diverticulum enters from the middle section. The body cavity in
both the middle and posterior sections is paired, the two lateral halves
being separated by mesenteries. The boundary between the middle
and posterior sections, the latter of which becomes the greater part of
the body cavity in the adult, becomes more or less indistinct. The
body cavity of the middle section is retained in the adult in the
postoral lamella and in the region of the central nervous system and
of the feathered tentacles, into which it is continued. The body
cavity of the posterior section in the adult contains the whole of the
alimentary canal and the ovaries, these organs almost entirely filling it.
It is continued into the pedicle.

The alimentary canal forms a loop in the body, with a ventral
section which runs backwards and into which the mouth leads, and a
dorsal section running forwards and opening anteriorly through the
anus. The mouth leads first into the " pharynx," which com-
municates with the exterior by means of the two gill-slits mentioned
above. A thin divertieulum runs out anteriorly from the pharynx
below the nervous system into the stalk of the oral disc. The
pharynx is followed first by an oesophagus and then by a very
spacious stomach or stomach intestine, which occupies by far the
greater part of the body cavity. At the point where the pedicle joins
the body, the stomach passes over into a narrower section of the
intestine, which, immediately behind the stomach, ascends, and then,
bending forward, runs along the dorsal side as the hind-gut to the
anus.

Genital organs. — Male genital organs have not been observed.
The female organs consist of two ovaries, lying in the anterior
part of the body ; they are continued into two strongly pigmented
oviducts, which open outward through the apertures already men-
tioned (Fig. 470, 17).

Reproduction. — Besides multiplying sexually by means of eggs,
Cephalodiscus also reproduces itself asexually by gemmation. The
buds always form on the pedicle, near to its free end. Almost all
adult individuals have from 1 to 3 buds.

Many individuals of Cephalodiscus live together in a ramifying
and anastomosing system of tubes secreted by themselves, these tubes
having occasional apertures. The animals, throughout life, when
not disturbed, remain in the immediate vicinity of these apertures,
through which they protrude their unfolded crowns of tentacles.

600 COMPARATIVE ANATOMY CHAP.

C. dodecalophus, the only known representative of the genus, was
found in the Magellan Straits at a depth of 245 fathoms.

Systematic position. — Cephalodiscus shows in the following
points a remarkable agreement with the Enteropneusta.

1. The body falls into three sections (distinct even in the young
bud), one preoral and two postoral. The preoral section, the so-
called buccal shield, corresponds with the proboscis, the middle section
with the collar, and the larger posterior section and the pedicle with
the trunk of the Enteropneusta.

2. These sections correspond with special sections of the coelom,
an unpaired coelom in the buccal shield, and two pairs of coeloms in
the body proper. AVe recognise here the unpaired proboscidal coelom
and the paired collar and trunk coeloms of the Enteropneusta.

3. The pores of the coelom of the buccal shield correspond with
the proboscis pores of the Enteropneusta proper, which are also often
two in number.

4. The pores of the pair of coeloms in the anterior body correspond
with the collar-pores.

5. Cephalodiscus and the Enteropneusta have gill-slits, the former
having one and the latter many pairs.

6. The anterior diverticulum of the buccal cavity corresponds
with the proboscidal diverticulum of the Enteropneusta.

7. The central nervous system corresponds with the collar cord
(which, however, in this case is not sunk below the skin) of the
Enteropneusta and with its immediate continuation on to the base of
the proboscis.

The differences existing between Cephalodiscus and the Entero-
pneusta may well be attributed, at least in part, to the tubicolous, half-
sedentary manner of life of the former. These are : (1) the anterior
position of the anus, and the consequent looped course of the
alimentary canal ; (2) the general crowding together of the most
important external organs (apart from the pedicle or stalk) at the
most anterior part of the body ; (3) the presence of a tentacular
crown, consisting of twelve feathered tentacles ; (4) the presence of
the pedicle or stalk ; x (5) the occurrence of asexual reproduction by
means of gemmation ; (6) the small number of gill-slits and genital
organs ; (7) the form of the body in general and especially that of the
proboscis ; (8) the absence of a blood vascular system.

II. Rhabdopleura.

This form, which was formerly classed with the Bryozoa? is no
doubt somewhat nearly related to Cephalodiscus, but is further
removed than the latter from the Enteropneusta.

1 An apparently homologous structure is, however, figured by Bateson on a young
Balanoglossus Ktiii'uli-cxIciL

a Iii the first volume of this book, indeed, Rhabdopleura appears among the Bryozoa.

IX

RHABDOPLEURA

601

This animal forms colonies by gemmation. Each individual
consists of a body and a contractile stalk, both of which are
enclosed in a horny tube. This tube is supine at first but rises
erect later. It is secreted in successive rings by the buccal shield.
The tentacles can be protruded through the aperture of the tube,

FIG. 472.— Rhabdopleura Normani Allm.,

individual, from the right side (after Lankester).
1, Buccal shield ; 2, feathered tentacles ; 3, region
of the "collar pore " ; 4, anus ; 5, trunk ; 0, stalk
or pedicle.

the body being withdrawn again into the tube by means of the stalk.
All these tubes are lateral branches of a creeping basal tube which
spreads out and branches on the substratum, and appears to be divided
into chambers by septa.

The stalks of the individuals enter this radical tube, and are
continued in it as thin strands covered with cuticle, which in travers-
ing it perforate the septa.

The axis of each individual stalk is formed by a strand of tissue

602

COMPARATIVE ANATOMY

CHAP. IX

ff

somewhat of the consistency of cartilage. Similar skeletal pieces

support the tentacles and their branches.

Some insight into the chief anatomical features of the individuals will

be gained from the figures (Figs. 472, 473). The principal differences

between Rhabdopleura and Cephalo-
fZiscws are: (1) the gill -slits are
wanting in the former ; (2) there
are only two feathered tentacles ;
(3) the proboscis pores, i.e. the
pores of the ccelom of the buccal
shield, are wanting.

The genital organs are still
imperfectly known. In some speci-
mens a testicle tube, which runs
longitudinally and asymmetrically
along one side of the body, and
bulges out the body wall, has been
demonstrated : this tube opens out
near the anus. Rhabdopleura, like
Ceplialodiscus, is a deep-sea form.

Further research is needed be-
fore we can establish the exact
relationship of Ceplialodiscu-s and
Rhabdopleura to the Bryozoa.

Fir,. 473. —Rhabdopleura Normani,
median longitudinal section, diagrammatic
(after Fowler). 1, Tentacle of one side, in-
dicated by dotted lines ; 2, anterior paired
ctelom (collar ccelom) ; 3, anus ; 4, posterior
paired or trunk co;lom ; 5, hind-gut ; 6, mid-
gut ; 7, buccal cavity ; 8, mouth ; 9, anterior
diverticulum of the buccal cavity (proboscidal
diverticulum) ; 10, ccelom of the buccal shield
(proboscidal ccelom) ; 11, buccal shield.

Literature.

G. J. Allmann. On Khabdopleura, a new
form of Polyzoa, from deep-sea dredg-
ing in Shetland. Quart. Jo-urn.
Microsc. Sc. Vol. IX. 1869.
G. Herbert Fowler. The morphology <>f
Rhabdopleura Normani Allm. Fest-
schrift zum Siebenzigsten Geburtstag R.
Lciickan't's. 1892.
Sidney F. Harmer. Appendix to M'lntosh : Report on Cc2)halodiscus dodccaloph u $

M'lntosh. Rep. Voy. of the "Challenger" Zool. Vol. XX. 1887.
E. R. Lankester. A contribution to the anatomy of Rhabdopleura. Quart. Journ.

Mlcrosc. Sc. Vol. XXIV. 1884.

W. C. M'lntosh. Report on Ccphalodiscus dodccalophus M'lntosh, anew type of the
Polyzoa, procured on the voyage of H.M.S. " Challenger." Rep. Voyage
"Challenger," Zool. Vol. XX. 1887.

G. 0. Sars. On Rhabdopleura mirabilis. Quart. Journ. Microsc. Sc. Vol. XIV.
1874.

INDEX

Numbers in Italics give Systematic Position. Numbers in Black Type

refer to Figures

Amph. = Amphiueura
Ast. = Asteroidea
Blast. = Blastoidea
Crin. — Crinoidea
Cyst. = Cystidea
Echiu. = Echinodermata
Echinoid. = Echinoidea

Eut. = Enteropneusta
Gastr. = Gastropoda
Hoi. = Holotnurioidea
Lam. = Lamellibranchia
Moll. = Mollusca
Oph. = Ophiuroidea

cavernosus, apical system, 324
Abranchia, 12
Acanthaster, 299

,, echinaster, 421

Ellisii, 421
Acanthochiton, 165
Acanthodoris, 13
Acanthology, 387

Acanthotrochus /inrabilis, "wheel," 337
Acephala, 14, 177
Acer a, 10
Aceste, 293

Acetabula (Moll.), 117 ; (Echinoid.), 390
Aciculidce, 6
Acmaea ( = Tcctura), 5
Acmaeidce, 5
Acrocidaris, 290
Acrocrinidce, 308
Acrocrinus, 309
Actceon, 110
Actceonidce, 10
Actinocrinidci', 307
Actinocrinus, 307

,, proboscidalis, apical system,

329

,, verneuilianus, 423

Actinocucumis, 338
Actinometm, 313; food grooves, diagr.,

366

strota, 365

Actinopoda, 285; oral region, section, 428
Adambulacral ossicles, 352 ; radii, 316
Adetes, 293
Adradii, 316

13
^Eolis, alimentary canal, 192

,, rujtbranchialis, 12
JErope, 293
^Esthetes, 166
^Etheria, 62
Agaricocrinus, 307
Agassizia, 293
Agassizocrinus, 304
Agelacrinus, 313

,, ciiicinnatensis, 313

Aglossa, 14, 177
Amaiid, 76
Ambitus, 338
Aniblypygus, 345
Ambulacra, 339

Ambulacral brush (Echinoid.), 433; gills
(Echinoid.), 433 ; ossicles, 352 ; radii,
316

.1 in iiwnitidea, 22
Amuion, 523

Amcebocytes (Moll.) 200 ; (Echin.), 415
Amphiaster, 297
Amphidromus, 160
^•1 iiiphineura, 2
Amphipeplea, S

,, leuconensis, 8

Amphisphyra, 46
Amphisternal test, 349
, 300

magellanica, 495
squamata, apical system, 327 :
stone canal, 422

604

COMPARATIVE ANATOMY

161

e, 300

.-,-, 307

Ampullae, madreporic, 420
,, tentacle, 430
100
(Lanistes), Boltemana, shell.

161
(Ceratodes), chiquitensis, 161

, 161
i, 161

purpvrea, 161
(Ceratodes), /w,/
Swainsoni, 161

r, 6

A nanch t/tida , 293
Anapta, 466
Anaspidce, 10

A a" fin", 21
Anatinidce, .'!
Anchylosis (Crin.), 378
Ancitlct, 13
Ancyloceras stage, 68
Ancylus, S
A nkyrodi rum, 287
Annulus, 126
Anochanus, 503

AiU'iti'iiti.i, J? ; circulation, 207; gills,
95 ; larva. 264; section, 221
,, cyijnua, 34

Aiio'niia, lo ; shell, 63
Aiimiiiidce, 14
Antednn, 313 ; embryo and larva, 510,

534-543

Aiiteduit in cisa, '312 ; larva, apical system,
318

,, phalanyium, young stage. 375

,, rosan-nx, 378

,, tuberculosa, 297
Antedonidte, 373

Antheneidce, 296
Antispadix, 116
Aorta, 198, 202
Apetala, 293

e, 310

.t, 3

Aplustridce, 110

AjJ>f*i", 10, 78 ; nervous system, 140
Aplysidla, 181
Aplysiidce, 10

Apophyses (Chiton), 39 ; (Echin.), 350
Aptychi (Ceph.), 71
A fi <-li a i n'des, 293
Arbacia, 290, 291
A i-baciidoe, 290
.1 i-f'i, 15

,, barbata, eye, 175

,, Xoce, 206
Archosociititi-ii/ir. 289
Ari-!i<> ocidaris ( = Echiiwcrinus), 289
ArcMster, 296
Archastt ridce, 296

'i". 295
is, 13
Architcenioglossa, 5
Arcidce, 1-5

Arcus (Echinoid. ), 400
Argeutea (Ceph.), 197
Ai'!/<ni"i'f", .'4- 24 ; gonads, female, 230 ;

nervous system, 148
„ '•'•/", male, 243

Arion, S

„ ater, 9
A /-i"/i//"iita, 8
.1 ristocystis, 313
Aristotle, lantern of, 400-403
Artieulamentum, 39
Ai-ficii'o.ta, 309
Ascocystls, 313
Ascoylossa, 11
Asiphoniata, 44
AsperyilliiiH (Brechties), 21. 20

„ i/ichitfi'iiii'iii. 66

Aspidobranchia, ,}

Aspidochirotce, 285 ; organisation, 477
Aspidodiadema, 290
A *2) idodio.de i a atii >< • , 290
Aspidosoma, ."..<•'
'-, 50

ti.i, pedicellarise, 395
mbens, madreporic plate,

421
„ ,. pedicellariae, 395

A.-:f' , . .. '

"<-i'ti,-i]ii ita, 506
, 506
i'.i, 506
capensis, 421

506
'-, 421
s, 507
spirabilis, 503
., sticmitliii. arm, 396

(Stollasterias), xolseUuta. arm,

396

,, teni'lafiiim, 421
„ vv.lgaris, 507
Asteriidce, 299
Asterina, 297

,, i/Hibosa, circular canal, etc., 425 :

ontogeny, 525-531
TIV.v", 506
Asterinidce, 297

Asler&idea, 295 - 299 ; alimentary canal,
etc., 484; arm, section, 411; optic cushion.
section, 468 ; oral skeleton, 352 ; pedi-
cellariw, 395 ; stone canal, section. 421 :
water vascular system, 463

Asterofisis, 297

Asthenosmna, 290

v. re us, spine, 390 : test, 472 ;
viscera, 443

Axtrochde, 301

Astroclon, 301

IXDEX

605

Ast r<>c ni<lii-. 301
Astrocrinidce, 315
Axfrocrintis, 315

„ Bfitiiiei, 315

Astrogomphits, 301
Astroijnnium, 296

A.-fmi-n'ctcn (i.umiitiacus, branchial skele-
ton, 351

A*ti\'pectinidce, 296
Astrojphytidce, 301
ton, 301

„ Li nckl, 301

a, 301

!/!'.; -90

Astroschema, 301
Astrotoma, 301
Atelecrinus, 313
Atelestocrinus, 304
Atlanta, 90, 109

,, Penmii, 7
At!i/itti<lce, 6
Atys, 46

Auricula, 350, 402
Auricularia, 506, 507, 508, 511, 512,

513, 514

,, of Synapta, 512, 514

An ricid idee, 8
Avicula, 17
.1 ri',- id idee, 17

Axial organ ( = ovoid gland), 437, 445, 446
Azyyobranchia, 5

BACULITES stage, 68
Buhi.nocrinus, 313
Bcdanoglossus, 562

„ canadensis, 565

„ Eoimh'cskii, 562 ; branchial

skeleton, 580

„ Knp/eri, 571

„ Merschkovskii, 571

Sarrandeocrinidce, 309
Barrandeocrinus, 309
Barrel-shaped larvse (Echiu.), 514
Bari/criu/'g, 304
Basals, 318

Basibrachial cartilage (Ceph.), 126
Basipterygial cartilage (Ceph.), 127
Basommotophom, S
Batluibiaxtfr, ,'.'"/'
Bathycrinidce, 304
tit'tlii/cri/tii* Aldrichianv.s, axial canals,

378

Baflti/di.'rt*, 13
Batocrinus, 307

,, pyrifi-'i-mis, 307

Baur's vesicles, 467
Belem nites, 24 ', shell section, 69
Belemnitidce, 24
Belem nocrin us, 304
Belemnoteuthis, 24
Better ophant idee, 5
Belosepia, 24

,, shell section, 69

; 29S

c--i, 2S5
!>' r< il< in. 13
Bipinnaria, larva, 507 ; dorsal aspect, 528

Hiri/lm, 14

Bivium, 325, 347, 407

Blastoidea, 314

Bin uiti-'i'i", 109

Bojauus, organ of, 215, 221

B ''//(/(/', 13

]',< 'fit i-iiiciilarix, 289

Bothriocida ro ida, 289

Botyocrin «.«, 304

BowrgueticrinidcB, 310

Bouryuftin icrinus, 311

Brachiolaria larva, 508

Branchiopneusta, 78

Brechites (Aspergillum), 21. 20

Brisinga, 299

Brisimjidce, 299

Brissopsis, 293

Brissits, 293

Buccal shields (Oph.), 336 ; plates (Echin-

oid.), 344
Bi'-CciniJ.cE, 6
Bucciiiiti/t, 160
Buliminus, 8
Bi'.lii/udidce, 8, 9
Bvlimus, 8

,, oblongus, 75

,, percersus, 160
Bidla, 10

,, hydatis, nervous system, 140
Bullidcc, 10

Btdh>idea (Cephalaspidce), 110
Bursffi (Oph.), 494
Byssus, 112-115, 114

yVS, 310

Calamus, 69
Calcareous cells, 190
< 'alct iii'i'inus, 304
, 296

, 309
ddlocystis, 332
Callum, 64
<_'</! i/ nine, 293
Cahjptnea, 108
Calyptrceidce, 6
Calyx, 319
<_'<tn<t:riifti, 306

i, 310

, 6
, 18
, 6

Cardiacea, 18
Can! in^t,',: 293

<_'<iriliii!<r, 18

Cardita, 17
Curilit i.kc, 17
Cardit'/n, 18

„ edule, nervous system, 144 ; tuber-
culatum, 18

606

COMPARATIVE ANATOMY

Carina (Echinoid.), 401
C't.rinaria, 90, 109
Caroliu, 63
Carjjiici-iiinx, 307
Caryocrinus, -il-J

,, ur/iuti's, apical capsule, 332
Cassidaria tyrrhcna, 163
Cassidiidce, 6
Cassidiiioiiii-ii, 293
Cassidulus, 293

„ pacijicus, perisome, 348
Cassis sulcosa, 6

s, 304
. ''93
Caudina, 281
Curnl inia, 11

„ tfidcittattt, 91
Ciirnl liiiidce, 11; diagram, 80
Centrodorsal, 375
Centrostephanus longispiniis, pedicellariap,

397

Cfphalaspidte (Bulloidea), 10
Cephalic cartilage (Gastr. ), 126 ; cone, 104 ;

disc, 103
Cephalodiscus dodecalophus, 596 ; sections,

597, 598

Cfi>lnd.iiph<ifi.i, 3, 177
Ci'iihulupoda, 21-25; eye, development,

171 ; embryo, 116 ; gonad, female, 230 :

male, 231 ; ink-bag, 197 ; heart, 199 :

retinal cells, 173
Cerata, 98
Ceratodes (see Ampullaria), 161

iis, 29n ; spine of, 389
i-niiiilii-nlata, 502
hystrix, peristome, 344
membranipora, 502
rii.; 502

i, oral area, 345
,, friljufiiidcs, surface of test, 389
Cidaroida, .'!>»; apophyses, 350

Cin ,!,'//<'. 7.".

,.'./

is, 54 ; shell of, 127
Cirrobi-n iii-liiii, 1 .'
Chcetaster, ^98
Ckcetoderma, .',. 88
(.'kii tndi-rnKitid.ee, 3
I '/m linli-riilfl.tuW., 3

(_.'/,,> f mi, nuiilo:, 3
Cho.ina, 18
f'lin inn era, 51
Chambered sinus, 446
n:, 18

ij;, 6

C'hiasma uervorum brachialium (Criu.),

377, 460

Chiastoneury, 135, 136, 137
L'hi/'odota, 288

„ riilifi'ra, 402

ctenidium, 85, 87 ; heart. 199 :

eye, section, 167 ; ovary, 227 ;

uephridial and genital systems,

217

cujctaitii.f, 165
Icevis, 88, 165
Pallasii, 82

Pulii, 165 ; development, 249
i-iiliii-i/ndus, 129

sicidus, 165 ; nervous system, 130
Chitonellus, 3 ; section, 41

< 'h ifniiiiln . : ; gills, 87
Chlamys, 17

t'ltoaitiiiiiplialiis Mnackl, 160
Choristes, 108
Chromatophores, 53
Chromodoris, 13

< 'ii'iidlu, 75
Cladohepatica, 13
C'lausilia, 8
Clavagella, 21
Clavo.gcUiili.i:, Jl
Clavulas (Echinoid.), 391
Cleiocrinus, 309
Cleodora, 30

Clio, 11

„ striata, anatomy, 190
Clioue, 90

, 11

( lionopsis, 90
Clypeaster, 291, 292

rosaceus, apical system, 322

Clypeastroida, 291 ; system of plates, 346
C!>/peus, 293
Ciiemidaster, 298

„ Jl>////, 297

<.',,(/iixfi',-, 315; ambulacrum section, 383

,. bilobatux, 314
Codasteridce, 315

y, 293

„ ellipticu, apical system, 325
Col I yr it idee, 293
Colochirus, 287

,'ii,',',ii /-•, calcareous body, 337

i, 3, 2 ; spine, 40 ; sections, 40, 212 ;

Columella, 123

Columellar muscle, 120-123

Columna (C'rin. ), 373

Columuals, 373

Columnar layer of shell (Moll.), 57

ComutiiUdi.i , 313

Conchyolin, 26, 57

Conidce, 6

Conodypeus, 291

Coralliophila, 183

Coralliophillida, 183

( 'ni-bii-iiffi, li

Corbis, 93
', 19

INDEX

607

Corethraster, 298

Corium (Moll.), 39 ; (Echin.), 414

Corona, 339

Coronaster, 299

Corpus epitheliale, 171

Corylocrinus, 332

Coryptella, 13

Costals (Grin.), 371

Oranchia, 24

Crassatella, 49

Orassatettidce, 17

Crenella, 115

Orepidulidce, 138

Cribrella, 299

,, sexradiata, 506
Crinoidea, 302-316 ; arm, 372 ; arm, sec-
tion, 413 ; ovarial pinnule, section, 500
Crioceras stage, 68
Crista acustica, 169
Cromyocrinus, 304
Crossaste/; 298
Orotalocrinidcc, 308
Crotalocrinus, 308

,, pulcher, 373

,, negosus, arm disc, 372

Oryptoblastus, 315
Oryptochiton, 41
Oryptodon Moseleyi, 50
Cryptoschisma, 315
Cryptozonia, 297
Crystalline stylet, 191
Cteuidium, 84, 85
Ctenodiscus, 296

,, procurator, 296

Ctenopidce, 102
G'ucumaria, 287

crocea, 502

crucifera, cruciform body, 337

doliolum, section, 517

frondosa, 488

Lacazii, 464

Icevigata, 502

longipeda, "stool," 337

minuta, 502

planci, 287
Culeita, 297
Cultellus, 19
Cuprcssocrinus, 304
Cuspidaria, 21
Cuspidaridce, 21
Cuttlefish ( = Cephalofiodu), 21
Cuvierian organs (Hoi.), 488
Cyathocrinidce, 304
Cyathocriuus, 304 > apical system, 329

,, longimanus, 304, 364

Cyclns, 81

,, cornea, development, 262, 263
Gydophoridce, 5
Cydophorus, 100
Cydostomu, 100

,, elegans, nervous system, 139 ;

radula, 182
Cydostomidce, 0

Cylichna, 46

Cymbulia, 11 ; larva, 257
' 'ymAulidce, 11
Cymbuliopsis, 11
Gyphosoma, 290
GyphosomatidcE, 290
Cypraea, 5
(Jypraeidce, 5
Cyre.ua, 17
Cyrenidce, 17
Cyrtoceras group, 68
Cystechinus, 293

,, vesica, gonads, 499

Cystidea, 313
Cystoblastus, 313

,, Leuchtenbergi, 312 ; apical

side, 332

Cystocidaris ( = Echinocystites), 289
C'ystocidaroida, 289
Cystocrinoidea, 313
Cystoid stage, 544
Ci/tkerea (Meretrix), IS
„ chione, shell, 63

DAUDEBARDIA (Helicophanta], 9
,, breoipes, 9

„ /'I/fa, intestine, section,

39 ; nephridia, 220 ;
pallial organs, 76, 77
,, saulcyi, 76

Decadocrinidce, 304
Decadocrinus, 304
Decapoda, 24
Dei ma, 285
Deimatidce, 285
Ddphiiiuli<li.i:, 170

Deltoid plates (Blast.), 331 ; (Criu.), 364
Dendrochirotce, 287
Dendrocrin idee, 304
Dendrocrinus, 304
Dendronotidcc, 13
Dentaliuin, 13, 33 ; shell, 59, 156

,, entalc, 113, 159 ; alimentary

canal, 193 ; larva, 258 ; on-
togeny of, 258
Dermatobranchus, 48
Dexiobranchcea, 11
Dextral twist (Gastr.), 56, 60, 160
Diadema, 290

,, setosi'ni, 392 ; compound eye,

469

Jtiodematidce, 290
Diudematoida, 290

,, apophyses, 350

Dialyneurous nervous system (Gastr.), 138
Diaphragm (Ceph.), 127 ; cartilage, 127
Diaulula, 13
Dibranchia, 24 ', eye development, 171 ;

musculature, 127 ; shells, sections, 69
Diceras, IS
Dichocrinus, 308
Dicyclic base (Criu.), 328
Dicyclica, 304

608

COMPARATIVE ANATOMY

Diffuse liver (Gastr.), 192
Digonopora, 8
Dimerocrinus, 367
Dimyaria, 14, 124 ; shell, 63
Diotocardia, 4, 30
Diplei'/'tda larva, 546
Diplopodia, 290
Discodoris, 13
Discoidea, 291
Distichals (Crin.), 371
Docoglossa, 5
Dolabella, 10
Doliidce, 6
Dolium, 102
DonacidcK, 18
Donax, 18
Dondersia, 3

,, batyidensis, 251

,, flavens, 184
DondersiidiK, 3
Dorididcc cryptobranchiata, 13

,, phanerobranchiata, 13
Doridiidce, 10
Doridium, 46
Doridopsidce, 13
Doriopsis, 214
Doris, respiratory aiid circulatory organs,

98

Dorocidaris papillata, 388
Dorsal axis (Blast.), 331 ; cartilage (Ceph.),

127

Dorycrimis, 307
Dosidicus, 69
Dotonidce, 13

Dreissensia polymorpha, gills, 94
B/'eissensiidce, 17
Dysaster, 293
Dy taster, '296

ECHINASTER, 299

, , sepositus, 483

Echinasteridce, 299
Echinidce, 290
Echinobrissus, 293
Echinocardium, 293

„ ftavescens, 398

Echinocidaris, 290

„ nigra, ambulacrum, 393

„ pustulosa, 291

Echinocumi.s, 293
Echinocorys, ,?9.3
Echinocrepis, 295
Echinocrinus (Archceocidaris), 289
Echinocyamus, 291

,, pusillns, gastrula, 520 ;

Pluteus larva and young,
520-524 ; tentacle, sec-
tion, 464

Ediinocystis (Cystocidaris), 289
Echiuodermata, alimentary canal, 475 ;
representatives of principal divisions,
316
Echinodisctts, 293

Echinodiscus biforis, 424
Ech inoencrinus, 313

„ armatus, 333

Echiiifiuh-fi, 288-295; endocyclic, 322;
exocyclic, 322 ; larva, 509 ; organisa-
tion of regular, 419 ; pedicellarise,
397 ; Pluteus, 520-524 ; radial region,
section, 410
Echinolampas, 293
Echinometra, 290
EchinometridcK, 290
Echinoneidce, 293
Echinoneus, 293
Ediinopsis, 290
Eckinospatayus, 293
JSchinosphcera, 386
Echinothrix, 290
Echinothuria, 290
EchinothuridcK, 290

Echinus, 290 ; apical system, 318 ; masti-
catory apparatus, 401
,, acutus, 398
,, melo, 398
Edrioaster, 387
Elceocrinus, 315
Elasipoda, 285
Eledone, 24

,, moschata, secondary body cavity,

diagram, 214
Eleutlierocrinus, 315

,, Cassedayi, apical side,

331 ; oral side, 383
Kl/ndm, 2S5

,, glacialw, 467
Elpidiidce, 285
Elysiadix, 12

Emarginula, 4 ', shell, 59, 156
Embryonic cone, 257
Enallocrinus, 308
Encope, 293

,, Vulenciennesi, system of plates,

346
Eiicrinus, 304 ; axial canals, 378

,, liliiformis, 304
Eitdocemtidai, 67

Eudogastric coil (Gastr.), 67, 68, 159
Enoploteuthis, 127 ; musculature, 127
Ens-is, 19

Enteropnensta, 561-596 ; branchial region,
567, 580 ; larvae, 586-590, 594 ; pro-
boscis, section, 583
Eiitucula.t Lttdwigii, 246, 247
Kittoconcha, 183

,, mirabUis, 247, 248

Entovalva, 229
Knlini-qxis, 293
K»l is Driii/iiiKindi, 183
Epineural canal, 449
Epipodial lobes (Ceph.), 38
Epipodium, 106
Episteruum, 349
Epistroma, 349
Eretmocriiius, 307

INDEX

609

Erisocrintts, 372
Ery cilia, 17
Erycinidce, 17
Einifiti'i-niili'ii, 296
Eucalyjotocrinidce, 309
Eucalyptocrinus, 309
EucladocriiiHs, 308
Eucystoidea, 313
Eudiocrinus, 313
Euechinoidea, 290

,, diailfiiiiitui/la, 469

Eulamellibranchia, 17 ; gills, 92
En lima, 183
EuMmidcB, 6
Eumargerita, 108
Eupachyc/rinus, 372
Enplocamus, 13
Eiinjahe, 301
Euryale, 301

Euthyueurous condition (Gastr.), 133, 158
Exogastric coil (Gastr.), 159
E.i'i'i/i/i-ii, 62
Extracrinus, 313

FACELLINA, 13

Fakes (Echinoid.), 400

Faorina, 293

Fascicles (sernites), 349

Ferdiiia, 326

Ferment cells, 190

Fibi/laria, 291

Flbulariidce, 291

FilibrancMa, 14 ', gills, 92

Fins (Ceph.), 54 ; (Oph.), 392

Fiona, 13

Firola corona ta, 7

Firoloides, 90

Fissurdla, 4 i ctenidiuni, 85 ; shell, 59,

156

Fissu/rellidae, 4
Fistuland, 64
Fistuluta, 303
Flabellina, 13
Fleming's cells, 162
Floscelle, 347
Foramen basale (Echinoid.), 400; exter-

num (Echinoid.), 400
Forbesiocrinus, 309 ; dorsal cup, 369
Fork pieces (Blast.), 331 ; (Echinoid.), 400,

401
Funnel (Ceph. = siphon), 265 ; ciliated

(Echiu.), 437
Fusidce, 6

GALATEA, 17

Galeomma, 17

Gcderopygus, 322

Galvina, 13

Ganeria, 297

Gasteropteridce, 10

Gasteropteron Meckelii, 10 ; genital organs,

233
Gastrochccna, 65

VOL. II

Gastrocoma, 304

Gastropoda, 3-13 ; hypothetical primitive,
150, 151 ; racial form, 159 ; sections,
47, 110 ; shell, 59, 156
Genicopatagus, ^93
Genital plates (Echinoid.), 321 ; stolon

(Crin.), 446
Gilbertsocrinus ( = 011acrinus), 306

,, tuberculosus, system of

plates, 367

Gills, adaptive, 97 ; anal, 97
Gissocriinifi, 304
Gladius, 69
','/n /i diceps, 562

,, Hacksii, 570

,, talaboti, 570

<il«nilina, 104
Glands —

,, albuminous (Gastr.), 232

anal (Gastr.), 177
,, blood-making (Ceph.), 97
,, byssus (Lam.), 112
,, calcai-eous (Gastr.), 42

digestive, 190-195
,, digitate (Gastr.), 237
,, goblet (Echiu.), 415
,, granular (Echin.), 415
,, hermaphrodite (Gastr.), 235
,, hypobranchial, 101

labial (Gastr.), Ill, 178
Leiblein's (Gastr.), 188
lime (Gastr.), 99

,, lymph ( = ovoid) (Echin.), 445
,, mucous (Gastr.), 42

uephridial (Gastr.), 219
,, nidamental, 233, 241
., oviduct (Ceph.), 241
., ovoid ( = lymph) (Echin.), 437,

444

pedal, 106, 111
pericardial (Keber's organ), 214,

215

,, pigment (Gastr.), 42
,, poison (Gastr.), 188

purple (Gastr. ), 74, 101
prostatic (Gastr.), 233
salivary, 184-187
shell (Gastr.), 237 ; larval, 253
slime (Gastr.), 237
stalk (Echin.), 399
,, sugar (Amph.), 187
Glandular pedicellarise (Echiu. ), 398
(,'fii.ncus, 13
Gleba, 11

GlocJiittiiim, larva, 264 ; paranitici<in, 263
Glomerulus (Ent. ), 582
Glossophora, 177
Glycymeridce, 19
Glycymeris, 19
Glyphocyphus, 290
Glyptaster, 306
Gly piaster idee, 306

<,'! II /itlK-fi II !>.-<, 307

2 R

610

COMPA KA TI VE ANA TOMY

Glyptosplwzrites, 313
G<>it taster, 396
Goniodoris, 13
Goniopygus, 290
Goryonocephalus, 301
Granatoblastidce, 315
Granatocrinus, 315

,, Norwood t, 314

Graphiocrinus, 304
Gryphcca, 62
Gymnasteria, 297

,, carinifera, pedicellarire, 395

Gymnasteriidcc, 297
Gymnosomata (Pteropoda), 11
Gyroceras, group, 68

H^MOCYANINE, 200

Hemoglobin, 200

Hcemolymph, 200

HaliaiJea, 6

Haliotidce, 4

Hallotis, gills, 90, 121 ; shell, 59, 156

Halopsyche, 90

Hamit&s, stage, 68

Haploffiniis, 303

,, mespttifornvis, 303, 334

Harpidce, 6

Hectocotylisation, 118, 242
Helcion, 108
II el taster, 299
Heliasteridae, 299
Hi licarion, 8
Helicidce, 8
Helidnidce, 5

Helicophanta (Daudebardia), 45
11,-Hd.er, 160

Helix, 8 ; anatomy, 236 ; alimentary canal,
186 ; circulatory system, 204 ;
sections, 99, 100, 104, 181 ; shell,
123

,, aspersa, 75
,, Ghiesbreghti, 183
,, pomatia, 9 ; nervous system, 142 ;

shell, 60

,, Waltimi, 258
Hemiaster, 293

,, caver nosus, 480
Ili'iiiir/i/nrtf!, ,''.>'•>
Hem icidaridce, 290
Hemicosmites, 332
Hemieuryale, 300
Hi in ipatagus, 293
Hemipholis, 327

,, cordifera, disc section, 435

Hi 'miji/1' ustes, 293
Hi rmccidce, 12
Hero, 13

Herpetocrinus, 304
Heteroblastus, 315
Heterocrinus, 304
*, :">.'>
iii, 15, 124
a, 6

Hexacrinidcc, SOS

Hexacrinus, 308

Hinge (Lam.), 61

Hipponycidce, 6

Hipiwnyx, 108

Hippopus, 18

Hippurites (Rudistes), 62

Hippuritidce, 18

Histioteuthis, 54

Holaster, 293

,, suljii/'bir/itiifis, apical system,
324

Holectypoida, 290

Holcctypus, 290

,, depressus, apical system, 322

Holohepatica, 12

Holopidce, 304

Holopus Ranyi, 305

Holostomata, 44

Holothuria, 285, 316

,, impatiens, 464

,, Murrayi, "stool," 337

,, /uli/i/uxa, organisation, 451

Holoth wrioidea, 2S5-2SS ; calcareous bodies,
337 ; ciliated rings, 515 ; diagram, sec-
tion, 407 ; oesophagus of deudrochirote,
404 ; oral region of actinopod, 428 ;
organisation of aspidochirotau, 477 ;
radial region, section, 409 ; stone canal,
etc. 418

Homalonyx, 45

Ilni/iocriiiiis, 304

Hood (Ceph.), 37

Hook sacs (Gastr.), 180

l/l/i'/ina, 76

Hybocrinu-s, 304

Hydreiocrinus, 372

Hydrobiida', 6

Hydrocoel, 511

Hydrocoenidce, 5

Hydrospire( Blast.), 382

Hymcnastcr, ;19S

,, ccelafi's, 298

,, nobilis, 298

pelluciditJi, apical side, 503

ITyocrinidce, 304

Hyocrinus Bethellianus, 305. 335

Hyphalaster, 296

Hi/pobrancMcea, 98

Jfypocrinus, 304

Hypoplax, 64

lANTHINA, 108
lanthinidce, 184
Ichthyocrinidce, 309
IcMhyocrin&idce, 308

Irlitl/ i/oci'i mix, ,!ll'.l
/i/, i/ii(, 13

Idiosepion, 24
Ihjaster, 296
flyodcemon, 418
Inadunata, 303

,, Jistidata, 303, 363

INDEX

611

/ in/if imata larvi/ormia, 303, 363
Infrabasals, 318

Ink-bag (Ceph.), 177, 196 ; morphology,
197

ix, 17

, shell, 63
Interambul.icra, 339
Interambulaeral radii, 316
Interdistichals, 363
Interuodes, 374
Interpalmars, 363
Interradii, 316
Intersecuudibrachs, 363
Intertertibrachs, 363

Illl-l'illllS, 304

Irpa, 418
frregulares, 315

Isaster, 293
Isomyaria, 124

JANUS, 13 ; nervous system, 141
Jouannetia, 19

,, Cum'uKjii, 19
Juglandocrinus, 332

KEBER'S organ (red-brown organ), 215
Kelhja, 17
Kentrodoris, 13

Kleinia luzonica, apical system, 342

Kolya, 285

Kollicker's canals (Ceph.), 168

LABIAL palps (Gastr.), 104 ; (Lain.), 105

Labidiaster, 299

Labidodemas, 491

Labrum (Echinoid.), 348

Lacuna, 108

Lcetmoyone, 285

Laganidce, 291

Laganum, 291

„ depressum, apical system, 323
LanifUiiridce, 6
LameUibranchiata, 14-21 ; byssus, 114 ;

diagrams, 50 ; heart, 199 ; shell, 59, 62,

156

Lancet plates (Blast.), 380
Lanistes (see Ampullaria}, 161
Larviformm, 303
Lascea, 17

Lateral organs (Gastr.), 165
Lecythocrinus, 304
Leda, 14
Lepetidce, 5
Lepidocentrus, 289
Lepidomenia, 13

„ hystri.i:, 132

Leptoconchus, 183
Lepton, 17

Leptoptychaster kergudenensis, 503
Leskiidce, 295
Leuconia, 109
Ligament (Lam.), 61
Li/iui, 105

i;, 8
helidna, anatomy, 189

,, J.i-sl/ft-iii, 11

,, retroversa, 160
Limacinida:, 11 ; diagram, 80

/,i/ll<l/Hllltiit/H', 12

Limn.'-, 8 ; vascular system, 204
/. ini/i/ii1, :1 3
Limncea, S

„ ii/ii/xm'cii/a, 100

„ x/i/:/,iti//x, genital organs, 235
Lininccidce, 8
Linckia, 298

,, multifora, 244
Linckiidce, 298
IJnthia, 293
Lipocephala, 101, 177
x, 15

., 108

Littorinidce, n

Li twites, 68
Lobiger, 12
Loligo, 24 ', gonad, male, 231 ; nervous

system, 148
,, riityci'is, 23
Loligopsis, 24 ; shell, section, 69
Lomanotidce, 13
Loven, law of, 341-344
Loveuia, J'J-1

Lucina Pennsylvania, shell, 63
Lucinacea, 50
Li/cinidce, 17
Luidia, 296
Luuula, 341
Lutraria, 51
Lutrariidce, 19
Lyonsia, Norweyica, 51
Lyonsiidce, 21

M Al'TRA, 18

Madrid 01, IS
Macula, 168

,, acustica, 168
Madreporite, 321, 416-423
Magiliis, 183
Malletia, 49
Mullens, 17
Mantle, 26 ; cavity, 26
Miiri/arita Groenlandica, 4
Margaritana ( Unio) Mar gar it if c fan, 17
Marginaster, 297
Marginellidce, 6
Mitriacriiuifi, 307
Marionia, 13
Marseniada:, 225
Marsupia, 347

ccelatus, tegmen calycis,

369
Ma mi ip ites, 304

„ ornatus, apical system, 329

M<ii-ii-xi<t, 65
Megallcesthetes, 166

612

COMPARATIVE ANATOMY

Megistocrinus, 307
Mdampas, 109
Melanidi.1 . <'>
Meleaffri.ii a, 17

,, ni fi.fi/iiritifertt, 17

Melibe, 13
Mellita, 293

,, testmliiifitft, 293
Meloorinidce, 307
Mdocrinus, 307

„ typus, 307

Melonites, 289

,, i/iii/fi/mra, apical system, 340
Mdonitidce, 289
Membrana limitans, 173
Meona ventricosa, apical system, 323
Meretrix (Cytherea), 18
Meridostemal test, 349
Mes-ites, 313 ; ambulacrum, section, 386
Mesoblastus, 315
Mesodesmatidce, IS
Mesoplax, 64
Mesorchium, 230
Metttbhtsttm, 315
Metacrinus, 313

„ angulatus, tegmeii calycis, 365

,, Murray i, 311

Metaplax, 64
Metapodium, 106
Metnxlini, 298
Micnesthetes, 166
Micraster, 293

,, coraiii/niiiniii, apical system, 324
Millericrinus, 310
Milncria, 49
Mimaster, 296
Mithrodia, 2'.t!>
Miim, 101
Mitridce, 6
Modiola, 15
Mndiolaria, 15
Moira, 503

Mollusc, hypothetical primitive, 26
Molpadia, 287

„ rliilfiitiix, 488
Mnl/Hiiliiiln, 387
Monocyclic base (Grin.), 328
Monocydica, 303
t, 8

iii. 14, 124 ; .shell, 64
Monopleurida', IS

Monotocardia, 5 ; diagram, 31 ; gills, 90
Mmifncuta, 93
Muelleria (Moll.), 124
MuMeriacea, 124
Muelleridtz, 124
Mulleria (Echin.), 285
Murex, 101

„ trim fa lux, alimentary canal, 189
Ma rii-iiln , fi
Mi'fehi, 17
Mali I inn , fil

Myarca, IS
Mi/iiiln, 64
M iiiiilif , 10
M I/IK-IIII mo, 51
M//<i/i.xi<l(i , 24
Mi/fiiitr<ii:hHs, 2SS

246

, 15

M i/l /lux rilnlix, 15

M i/.i-iister, 298

NACELLA, 108
Nacreous layer of shell, 57
, 108

i , 107
Natantia, 10

Xntii-ii .liixi'jiliini/, 107 ; swelling of fnut.
119

Xu.tii iiin. <;

.' .'

, ..'.'; eye, 169; diagram, 37;
gonads, 230 ; nervous system,
145, 146 : pallia! complex,
82 ; tentacles, 117

Po/iqiilinx, 22

group, 68
Nectria, 296

Needham's (spermatophoral) pouch, 238
, 303

in, 3

, 3

tri, 78

Nefttunea, contraria, 160

Neritidce, 5
Neritince, 5
Neritopsidce, 5

Notaeum, 10
ri-liux, 10

[in n<-tiii i ix, nervous system, 141

i.', 10

90

/"'", 11
Nuchal plate, 126
Xiti-lt'iililitxtiilii , -115
Nucula, 14- ; ctenidia, 85

„ n nr/i'llx, 14

i;, 14

12

OCTOPODA, 24

Or/,,/,, nliilil , 24

Octopus, 24 ', anatomy, 147 ; nervous
system, 148

,, fii/i/iii-i.x. ,.'.'/• gonad, female, 231 ;

male, 239

Ocular plates (Echinoid.), 321 ; (Ast.) 354
Odoutoblasts, 183
Odontophore, 336
< h</<i/ixiiln , .'4
o//(/'v»/'//'.v. 293
Oliva, 30

INDEX

613

Olividce, 6

Ollacrinus (see Gilbertsocrinus)

Omanastrephes, 24 ; gonad, female, 230 ;

shell, section, 69
Ommatophore, 102

Ommatost replies, 24 ', nervous system, 148
Oncidiella, 46
Oncidiidu; S
Oncidium, 44, 104

celticum, genital organs, 235 ;

larva, 256

Oneirqphanta mutabilis, rod, 237
Onychoteuthis, %4
Operciilum, 31
Ophiagantha, 300

,, marsupialis, 495 ; vivipara,

495
Ophiactis, 300

Miilleri, 506
poa, 300

,, Savigny, 506
„ virens, 498 ; disc, section, 426
Ophiarachna incrassata, vertebral ossicles,

356

Ophi Master Germani, 421
Ophioceramis, 327
Ophiocnida, 300

,. sexradia, 506

Ophiocoma, 300

„ pionila, 506

„ Valencice, 506 ;

Ophiocreas, 301
Ophiodermn ( = Ophiura), 495
Ophiodiaster, 298

„ i/if>lax, regeneration of arms,

505
Oph'm<il //pint, 300 ; bursa, 495, 496 : disc,

section, "497

,, albida, stomach, 494

,, hexactis, 495

,, lacertosa, ovary, section, 497

Ophioglyphidce, 300
Ophiohelus umbella, arm joint, 357
Ophiomastix, 327
Ophiomitra exigua, 327
Ophwrnusivm, 300

„ ml iilum, apical system, 327

Ophiomyxa, 300

„ rirfpara, 495

Ophiomyxidce, 300
Ophionereis, 422
Ophioplocus, 422

Ophiopteron degans, brachial joints, 391
Ophiopya longispinus, oral skeleton, 359
Ophiopyrgus, 327
Ophiothda dividua, 506
„ isidicola, 506
Ophiotholia, 392
Ophiothrix, 300

,, frar/ilis, ambulacral tentacle,

section, 466 ; nervous sys-
tem, 456
Ophiozona, 300

VOL. II

Ophiura, 300

Ophiura; 300

Ophiuroidea, 299-301 ; arm, section, 412 ;
arm, joint, 357 ; disc, section, 486 :
Pluteus, 533 ; nervous system, 456 ;
ring sinus, 496

Opisthobranchia, 10, 33 ; heart, 199

Opisthopneumonia, 76

Oral angle plates, 358
,, lobes (Lam.), 105

Orocystis Helmliackeri, 313

Orqphocrinus, 315

„ stelliformis, 314

Orphniirgus, 418

Orthoceras group, 68

Orthopsis, 299

Oscainius, 10

Osphradium, 84, 162-164

Ostracoteuthis, shell, section, 69

Ostrea, 17

„ edulis, anatomy, 16

Ostreidtt', 17

Oivenia, %4

Oxygynis, 108

Oxynoe, 12

Oxynoidea, 12

PAL^SASTER, 295
Palceasteroidea, 295
Palceechinoidea, 288
Palaxchimis, 289

„ elegans, 289

Pcdceobrissus, 293
Palceocoma, 295
Pahrocrinoidea, 303
Palceopneustes, 293

„ Murrayi, 294

Palceostoma, 295
Palocotropus, 429
Pallial cavity, 26 ; line, 64, 124 ; sinus,

64, 124

Palliata ( = Tectibranchia), 46
Pallium, 26
Palmars (Crin.), 371
Palinipes, 297
Paludina, ctenidium, 85

,, riripara, circulatory system,

203 ; development, 252. 254,

255

1'flllllli lli/ll_l , O

Pancreas (Cepli.), 196
Pandoridce, 21
Pami /ii'li in, JS5
Papulffi, 439
Paractinopoda, 288
Paramenia impexa, 216
,, palifera, 184
Parameniida; 3
Parapodia, 106
I'n ni r<-Iiiiffter, 296
1'in-ii.tidi-nia, 290

Parasira (Tremoctopus) catenukita, 230
Parelpidia, 418

2 R 2

614

COMPARATIVE ANATOMY

Parmophorus (Scutum), 5

Patella, 72 ; nephridia, 218 ; nervous

system, 138 ; section, 188
,, rulyata, 4

Patelluhtf, 5
Patinella, 108
Paxillae, 391, 503
Pecten, 17 ; eye, 174
,, Jacobeus, 16
Pectinibranchia, 5
Pectinidce, 16
Pectinura, 300
Pectunculus, 15 ; eye, 175
Pedicellariae, 393-399
Pedicellaster, 299
PediceHasterid.ee, 299
Pedina, '290
Pedipes, 109

Pelagothuria natatrix, 285, 286
Pelagothuriidce, 285
Pelanechinus, 290
Pelecypoda, 14
Pelmatozoa, 302-315
Peltastes, 290
Peltella palliolum, 9
Peltulti", 10
Pen (Ceph.), 69
Peneagone, 285
Pentaceros, 297

„ turritus, oral skeleton, 473

Pentactrnthhi', 296
Pentar.riitida, 313
Pentacrinus, 313

„ decoms, 377 ; calyx, section,

382

Pentactcea, larva, 548
Pentactula, 516
Pentagonaster, 296
Pentagonasteridu; 296
PentephyUitm, 315

Pentremites, 315, 314 ; ambulacrum,
section, 382 ; apical system, 330 ;
organisation, 380
Pentremitidce, 315
Pentremitidea, 315
Peraclis, 11
Periechocrinus, 307
Periostracum, 58
Periscoechinoidea, 289
Peristome, 339
Perna, 17

„ Epliippinm, shell, 64
Peronella orbicularis, 424
Peronia (Moll.), 45 ; (Ech.), 290
Perradii, 316
Petalodium, 347
Petricula, 52
Petricolidce, 18
Phcenoschisma, 315
PJianerozonia, 296
P/iarus, 51, 115
Philine, 46
Philiiiid/i, 10

Philonexidcc, 24
Philonexis, 24

,, (Octopus) carence, hectocotylisa-

tion, 243
Pholadacea, 19
Pholadidcc, 19
Pholadidea, 19
Ph./iladomyidce, 51
Pholas, 19

„ dactylus, valve, 67
Phormosoma, 290
Phorus exutus, 5
Pliragmocoue, 68
Phyllidiidcc, 13
P/iyllirhoe, genital organs, 238

,, buceplialum, 12

Phyllirh»ida\ 13
PhylldbranchidcK, 12
Phyllodes, 347
Phylloplwrus, 287

„ urna, 502

Physa, 8

„ fontinalis, 8
Physetocrinus, 307
Pinna, 17

Pinnules (Crin.), 371 ; (Blast.), 381
I'ii-ni'i/stis, 336
Pisidium, 17
Pisocrinus, 304
Placobranchidce, 12
PlacopJwra, 2
Placuna, 15

I'lii mi, i'ii!ir, S

Planorbis, S, 58

Plastidogenic organ, 445

Plastron (Ecliiuoid.), 348

Plates, included, 340 ; isolated, 340 ; half,

340 ; primary, 340
Platycrinidce, 307
Platycrinus triacontadactylis, 308
„ fvberosus, tegmeu, 335

Platydoris, 13
Plesiocidaroida, 289
Plesiospatangidce, 293
Pleura?, 182
Pleurobranchcea, 10

„ Meckelii, genital organs,

237

Pleurdbranchia, 47
PlenrobraJichidce, 10
/'/I'l'i-n/irniic/iKs, 18

,, auranttacus, 10

1'1,'11,-n /i'/i,-/i/ie, 13

'iifn, 13

t , 13

in, 5 ; shell, 59, 156
Pleurotomaridce, 5

e, 6
'i, 125

Pluteus, larva, 508, 509, 522
Plutonaster, 296
Pneumoderma, 11, 79
I'n: niii<itlermatid(e, 11

INDEX

615

Podocidaris, 290
Podocyst, 258
Polian vesicles, 416, 423
Polycera, 13
Polyplacophora, 2
Polytremaria, 5 ; shell, 59, 156
Pompholyx solid a, 160
Porcelainous layer of shell, 57
PorceUanaster, 296
Porcellanasteridce, 296
Pore rhombs (Cyst.), 384
Porocrinus, 313
Poromya, 51
Poromyidce, 21
Postpalmars (Grin.), 371
Poteriocrinus, 304
Pourtalesia, 295

„ Je/reysi, 295 ; apical system,

'325

Pourtalesiidce, 295
Prseputium, 235
Prismatic layer of shell, 57
Proboscidifera holostomata, 6

„ siphonostomata, 6

Promachocrinus, 313
Proneomenia, 3

„ Sluiteri, 3 ; nervous system,

132 ; sections, 42

,, vagans, 184

Proneomeniidce, 3
Proostracum, 68
Propodium, 106

Prosobrcmchia, 4 ', giUs> 90 > heart, 199 ;
proboscis, 179 ; sections, 110 ; snout,
section, 181 ; tentacles, 102
Prosoplax, 64
Prostata, 233

Protobranchia, 14 ', gills, 92
Protocrinus, 313

„ oviformis, 312

Prunocystis, 332
Prymnadetes, 293
Prymnodesmia, 293
Psammobia, IS
Psammobiidce, 18
Pseudar chaster, 296
Pseudoconch, 47
Pseudolamell ihranchia, 15
Pseudomelanidce, 6
Pseudomonocyclic Criuoids, 329
Psolus, 287

„ ephippifer, 287
Psychropotes, 285

„ longicauda, 285

Psychropotidce, 285
Pteraster, 298
Pterasteridce, 298
Pterasterince, 503
Pterobranchia, 12
Pterocera, 107
Pteropoda, 10 ; larva, 257
,, gymnosomata, 11
,, t/iecosmnata, 11

Pterotrachceidce, 6
Pterotrachea, 109

,, auditory organ, 168

,, (Firola) coronata, 1

PtycJwdera, 562

„ aurantiaca, 585

,, ba/uimensis, 585

,, clavigera, 570

., erythrcea, 585

„ minuta, 562 ; branchial region,

section, 562 ; head region,

section, 566, 568 ; ccelorn,

diagrams, 576

PulniiiHatd, S, 32 ; eye, 170 ; radula, 183 ;

renal ducts, diagram, 75
Pupa, 8
Pupidce, 8
Purpura, 101
Purpuridce, 6
Pygaster, 291
Pi/gurus, 293
Pi/ram idellidce, 6
Pyriform vessel (Gastr.), 220
Pyritla, 44

,, tuba, 73
Pythonaster, 29S

RACHIAL teeth, 182

Rachiglossa, 6

Radial shields, 362

Radials, 318

Radiiififiifn; 18

Radula, 177, 180-183

Red-brown organ (Keber's organ), 215

Regulares, 315

Reptantia, 10

Requienia, IS

Retaster, 298

Rete miraliile (Hoi.), 452

Retiocrinidce, 306

Retiocrinns, 306

Rhabdopleura, 600-602

,, Normani, 601 ; section,

602

Rhachiglossa, 6
Rhinophore, 48, 103
Rhipidocrinus, 306
Rhipidoglossa, 4
Rhizochilus, 183
Rhizocrinus, 311

,, lofutensis, 335 ; axial canals,

378 ; stone canal, 423

„ Rawsoni, 335

Rlwdocrinidce, 306
Rhodocrinus, 306
Rhodope Veranii, 281
RJwpalodina, 287

„ derivation of, 408

Rimula, 43
Ringiculidce, 46
Rissoicke, 108
Rizzolia, 13
Russia, 24

616

COMPARATIVE ANATOMY

Rostellaria, 102

„ rectirostris, 6

Rostrifera, 6

Rostrum (Cepli.), 68 ; (Gastr.), 102
Rotula, 293

Rotulse (Echinoid.), 400
Rudistes (Hippu rites), 62
Ruff (Gastr.), 107

SACCULI (Grin.), diagram, 490

Salenia, 290 ; apical system, 319

Saleniidce, 290

Saxicava, 19

Scceurgus, 243

Scalariidce, 6

Scaphander, 46

Scaphandridce, 10

Scaphites stage, 68

ScapJwpoda, 13

Sckizaster, 293

„ fa iialiferus, pedicellarise, 397
lacunosus, 294

Schizoblastus, 315

Schizocardium, 562

,, brasiliense, gills, section,

569

Schizogoriy, 505

Schwammerdam's vesicle, 233

Scissurella, 5

Scotoplanes, 418

Scrobicularin jiipi'rafa, 52

Scrobiculariidce, 51

Scurria, 5

Scuta buccalia, 360

Scutella adoratia, 360

Scutella, 293

„ sexforis, 292

Scutellidve, 291

Scutum (Parmorphus), 4

Scyllceidce, 13

Scytaster, 326

Sea urchins (= Echinoid.), 316

Semiproboscidifera, 6

Semites ( = fascicles), 349

Seniper's organ, 166

Sepia, 224 ; alimentary canal, 189 ; body
cavity, 213 ; cephalic cartilage,
126 ; ctenidium, 85 ; diagram, 36 ;
gill, 96 ; gonad, male, 231 ; onto-
geny, 266, 267 ; nervous system,
148 ; renal sacs, 213 ; shell, sec-
tion, 69 ; spermatophore, 242
(irulentu, shell, 70
officincdis, circulatory system, 209 ;
eye, 172 ; genital organs, female,
240 ; genital organs, male, 239 ;
renal sacs, 224
,, Savigniana, 83

Sepiadarium, 24

H:-/iiii/it, 24 ; nervous system, 148

Sepioloidea, 54

Sepioteuthis, 24

Septibranchia, 21 ; gills, 92

Sickles = radii (Echinoid.), 400

Siteiiiri. Xarxii, 21

Si liquid, 115

Siuistral twist (Gastr.), 56, 60, 160

SimipaHiata, shell, 63

Siphon (Ceph.), 37 ; = accessory intestine
(Eehinoid.), 481 ; (Gastr.), 43; (Lam.),
exhaleut, 49 ; (Lam.), inhalent, 49

Si pin i H in ta, 44

Siph o it ode n taliu m , 13

Siphonoplax, 64

Siphonostomata, 44

Siplnmcle (Ceph.), 37

Solaridie, 6

Solasteridce, 298

Solen, 19

Solenidce, 19

Solenocurtus, 19

Solenogastres, 3

Solenogastridce, section, 212

Solenomya, 91

Solenomyidce, 14

Solenojwda, 228

Spadix, 116

Spatagocystis, 295

Spatangidce, 293

Spatangoida, 293 ; ambulacral brush,

section, 433

Spatangoidea, larva, 509
Spatangomorpha, 293
Spatangus, 293

„ jmrpureus, apical system, 324 ;

oral region, section, 441
Speugel's organ, 84
Spermatophore, 241, 242
Sjilin ri'i-liiitiia. .".in

„ gramdaris, pedicellaria?, 397,

398

Sphseridia (Echiiioid.) 392 ; section, 392
fyi/iti /-in in, 17
Sphcrmnis, 336
Spiculum amor is, 237
Spiracles (Blast.), 382
Xpiriila, 24 ', shell, section, 69

„ prototypes, 23
Spirulidce, 24

Spirulirostra, 24 ', shell, section, 69
SpondylidcB, 53
^/iiuii/i/fiift, 62
HI HI i tr/ iobran chcca, 11
Spongylocentrotiis, 290
Star-fish ( = Axtrriiiilini'). 316
Steganobranchia, 12
Hti'/iiiiiiii-riii us. ,;n~
Stdidiocrimis, 307
HI <•!! aster, 296

Stelleridea ( = Asteroidea), 295
Stenoglossa, 6

Sternum, 348

Stewart's organs (Echinoid.), 442

Hticliaster, 298

,, nlbi'lus, 506

IXDEX

61'

s, 285
„ /iijiiiitici/s, rod and supporting

plate, 337

,s'/////i /• /,///«•/•/«•, sections, 245, 246
Stolasterias ( — Asterias -culsdlata), 396
Sttnnat iiiln , ~i
Stone canal, 416-423
Streptoneurous condition (Gastr.), 133,

158
Streptosomata, 290

Strmnbus, 102

Strongylocentrotus lifit.1t/.f, 398
x, 307

regalis, apical border, 368

Stylonvmatophora, 8
S'Ubemargimula, 89

Xit/Hiii/fi/iift'ii, 17
Xiifi-lnritln , ,V
Sytnbathocrinus, 3<>4

Syiiapta, 288 ; young, 516 ; auditory
vesicle, section, 467

„ digitotti, 288 ; larva and young,
511

,, inlnt /r,i.v, calcareous body, 337

„ vittata, 470
Synapticulae, 567

Synaptidce, :.'SS ; ciliated urn, 438
Syuostosis (Grin.), 376
Syzygy (Grin.), 376

T.EXHiGLOSSA, 6

Talarncriinis, 308

Tapes, IS

Tapetum lucidum, 175

„ iiiiiltihrinii-liinl i/*, 310

Tedibranchia (—Pattiata), 10; .section,

110

Tfdum ( = At: iima), 5
Teeth (Ast.), 473 ; (Echin.), 400, 401 ;

pliaryngeal (Gastr.), 180
Tegnieu calycis, 302, 369
Teginentum, 39
Tcli'ion-iitn.i, 307
Tellina, 18
Tellinacea, IS
Tell ini dee, 18
Temnopleuridce, ~90

Terebra, 102
Terebridce, 6

Ti'i-i'il i mi, 65
Ti'inliiiiil'i , 21
Teredo, 21

„ nnvaMfi, 20
260

development, 259,

Terminal plates (Ast.), 354
Testacellu, 8, pallial organs, 77
., Ini/infii/cii, 9, 44, 45

Testacdlidt.1', 8

, 13

, 13

..'.'

Thaumatocrinus, 309

310

Thru fin, .'I
T/iyca, 183

,, cdm-inu-lia, 214-246 ; sections, 244,
246

„ ch if en. vis, 417
Thy san "I i' a Hi in, 2 4
189

princeps, 289 ; 319
Tiedemanu's bodies, 416-424
Titiscania, 4

larva, 507

J //<'.«/-./, 589
GreiKU'lteri, 587
Km/i it i, 586, 590
Mi>r/i',-i, 587
, 46

Torus angnlaris (Oph.), 361
Tosia, 326
Toxiglossa, 6
Toxopneustes, 290 ; masticatory apparatus,

402

,, (li-oefMC/tii.'itxix, apical system,

320 ; system of plates, 341
Tfemi.idtipus, 24

,, riiiliii'i'ns, 196

T i- if/i aster, 301

„ dcijuns, 394
Tricuspi.l body (Lam.), 194
Tridacna, 18
Tridacitida; IS
Trifnrix, 160
Trigoniu, 15
Triijniiiiiln , 15
TrinjMi, 13
Triploblastica, 545
Tripiii'iix/i'*, ..'•'"'
Tritii/iin, 13

, 13

Trivinm, 325. 347, 407
i, 253, 260

ng gills, 90
Troostoblastida . 315
Trotistarriiiiix, 315

Tubercle, 388

Turiiiiii'lliiln', 6
TII i-li! a ill" . 5
Turin, 166
Turgescence, 118
Tui-i'ilitf stage, 68
Turrit dl a, 30

618

COMPA RA TIVE ANA TOMY

', 6

Tylaster, 297
Tylodina, 10

UlNTACRINUS, 309

Umbrella, 10
Umbretlidfv, 10
Uucmi, 182
Unio, 17

„ Margaratlferus
35

,, pinctorum, 17
Unionidce, 17
1'iii induce, 50
Uniophora, 299
Urechinus, 298
Urns, ciliated, 437
Utriculus, 180

!, S

\'nij!n uliifi, 45
Valvaster, 299
Valvatidce, 6
Veliger, larva, 1, 257
Velum, 250
\'<'in-f(icea, IS
Vc nericardia, 17
Ven&ridce, IS
Venus, 18

17,

Vermetus, 108, 248
Vertebra] ossicles (Oph. ), 356
Verticordiii/tt; 51
Vitrinn, S
Volutuht', 6
Vulsella, 17

WACHSMUTH and Springer's rule, diagram,

374
Wandering cells (amcebocytes) (Echin.),

415
Water lung (Hoi.), 487 ; pore (Scaph.),

221 ; pores (Criu.), 377

XENOCRINUS, 306

Xflliiphnridcr, 6

Xylopliaga, 21

YOLDIA, 14
Ypsilotlitiria, 409

ZEUGOSRANCHIA, 4
Zonites, S

,, fi/lyens, apical system, 326
Zoroasteridce, 298
Zygoueury, 137

THE END

Printed l<y K. &; R. CI.AKK. LIMITKD,

BOOKS ON ANATOMY AND PHYSIOLOGY.

ELEMENTS OF THE COMPARATIVE ANATOMY OF VERTE-
BRATES. Adapted from the German of Prof. ROBERT WIEHERSHEIM. By
W. NEWTON PARKER, Professor of Biology in the University College of South
Wales and Monmouthshire. With additions by the Author and Translator.
270 Woodcuts. Medium 8vo. 12s. 6d.

THE STRUCTURE OF MAN. By Prof. R. WIEDERSHEIM. Translated
by H. M. BERNARD and G. B. HOWES. 8vo. 8s. net.

A TEXT-BOOK OF COMPARATIVE ANATOMY. By Prof. ARNOLD
LANG. With Preface to the English Translation by Prof. Dr. ERNST HAECKEL.
Translated into English by HENRY M. BERNARD, M.A. Cantab., and MATILDA
BERNARD. Vol. I. Svo. 17s. net.

LESSONS IN ELEMENTARY ANATOMY. By ST. GEORGE MIVART,
F.R.S., Author of "The Genesis of Species." Fcap. Svo. 6s. 6d.

TEXT-BOOK OF ANATOMY AND PHYSIOLOGY FOR NURSES.

Compiled by DIANA CLIFFORD KIMBER, Graduate of Bellevue Training
School. Svo. 10s. net.

A COURSE OF INSTRUCTION IN ZOOTOMY. Vertebrate. By
T. JEFFERY PARKER, F.R.S., Professor of Biology in the University of
Otago, New Zealand. With Illustrations. Crown Svo. 8s. 6d.

AN INTRODUCTION TO THE OSTEOLOGY OF THE MAMMALIA :

being the Substance of the Course of Lectures delivered at the Royal College
of Surgeons of England in 1870. By Sir WILLIAM HENRY FLOWER, F.R.S.
Illustrated. Third Edition. Revised with the assistance of HANS GADOW,
Ph.D. Crown Svo. 10s. tkl.

THE MYOLOGY OF THE RAVEN (Corvus corax Sinuatus). A Guide to
the Study of the Muscular System in Birds. By R. W. SHUFELDT, of the
Smithsonian Institute, Washington, U.S.A. With Illustrations. Svo.
13s. net.

A TEXT-BOOK OF PHYSIOLOGY. By MICHAEL FOSTER, M.D., F.R.S.,

Professor of Physiology in the University of Cambridge, and Fellow of Trinity
College, Cambridge. Illustrated. Svo.

Part I. Comprising Book I. Blood — The Tissues of Movement, the Vascular
Mechanism. Sixth Edition. 10s. 6d.

Part II. Comprising Book II. The Tissues of Chemical Action, with their
Respective Mechanisms — Nutrition. Sixth Edition. 10s. 6d.

Part III. The Central Nervous System. Sixth Edition. 7s. 6d.

Part IV. Comprising the remainder of Book III. The Senses and some Special
Muscular Mechanisms ; and Book IV. The Tissues and Mechanism of Reproduction.
Fifth Edition. 10s. 6d.

THE CHEMICAL BASIS OF THE ANIMAL BODY. An Appendix to
Foster's "Text-Book of Physiology." By A. SHERIDAN LEA, M.A., D.Sc.,
F.R.S. , University Lecturer in Physiology in the University of Cambridge.
Fifth Edition. Svo. 7s. 6d.

PHYSIOLOGY. By MICHAEL FOSTER, M.D., F.R.S. With Illustrations.
Pott Svo. Is. [Science Primers.

MACMILLAN AND CO., LTD., LONDON.

BOOKS ON PHYSIOLOGY AND EMBRYOLOGY.

A COURSE OF ELEMENTARY PRACTICAL PHYSIOLOGY
AND HISTOLOGY. By Prof. MICHAEL FOSTER, M.D., F.E.S., and
J. N. LANGLEY, F.R.S., Fellow of Trinity College, Cambridge. Sixth
Edition. Crown 8vo. 7s. 6d.

LESSONS IN ELEMENTARY PHYSIOLOGY. By T. H. HUXLEY,
F.R.S. With numerous Illustrations. Fourth Edition. Pott 8vo.
4s. 6d.

QUESTIONS ON HUXLEY'S LESSONS IN ELEMENTARY
PHYSIOLOGY. For the Use of Schools. By THOMAS ALCOCK, M.D.
Fifth Edition. Pott 8vo. Is. 6d.

A TEXT-BOOK OF THE PHYSIOLOGICAL CHEMISTRY OF
THE ANIMAL BODY. ' Including an Account of the Chemical
Changes occurring in Disease. By ARTHUR GAMGEE, M.D., F.R.S.,
Emeritus Professor of Physiology in the Owens College, Victoria
University, Manchester, etc. 8vo. Vol. I. 18s. Vol. II. 18s.

A TREATISE ON COMPARATIVE EMBRYOLOGY. By F. M.

BALFODR, M.A., F.R.S., Fellow and Lecturer of Trinity College, Cam-
bridge. In two Vols. Second Edition. Medium 8vo. Vol. I. 18s.
Vol. II. 21s.

THE ELEMENTS OF EMBRYOLOGY. By Prof. MICHAEL
FOSTER, M.D., F.R.S., and the late F. M. BALFOUR, F.R.S., Professor of
Animal Morphology in the University of Cambridge. Second Edition,
revised. Edited by A. SEDGWICK, M.A., Fellow and Assistant Lecturer
of Trinity College, Cambridge, and W. HEAPE, M.A., late Demonstrator
in the Morphological Laboratory of the University of Cambridge.
Illustrated. Crown 8vo. 10s. 6d.

LESSONS IN ELEMENTARY BIOLOGY. By T. JEFFEKY PARKER,
F.R.S., Professor of Biology in the University of Otago, New Zealand.
Illustrated. Second Edition. Crown 8vo. 10s. 6d.

A COURSE OF PRACTICAL INSTRUCTION IN ELEMENTARY
BIOLOGY. By T. H. HUXLEY, F.R.S., assisted by H. N. MARTIN,
F.R.S., Professor of Biology in the Johns Hopkins University, U.S.A.
Third Edition, revised and extended by G. B. HOWES, Assistant
Professor, Royal College of Science, and D. H. SCOTT, Ph.D. With a
Preface by T. H. HUXLEY, F.R.S. Crown 8vo. 10s. 6cl.

AN ATLAS OF PRACTICAL ELEMENTARY BIOLOGY. By

G. B. HOWES, Assistant Professor, Royal College of Science, Lecturer on
Comparative Anatomy, St. George's Medical School, London. With a
Preface by T. H. HUXLEY, F.R.S. Medium 4to. 1 4s.

MACMILLAN AND CO., LTD, LONDON.

.

,

•

.

rfm

, • ~ . .

.

:

.

.

•

j -
;

'

i

•

..

. i

;

.

I

. •
'

.

•

I

•

H

.
.
,

'
'

H

* ^

mm

.
i

.

. .

-

• • . . ..