DOLOGY

-

TEXT-BOOK

OF

COMPABATIVE ANATOMY

BY

DR. ARNOLD LANG

u

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

WITH

PREFACE TO THE ENGLISH TRANSLATION BY

PROFESSOR DR. ERNST HAECKEL, F.R.S.

DIRECTOR OF THE ZOOLOGICAL INSTITUTE IN JENA

TRANSLATED INTO ENGLISH BY

HENRY M. BERNARD, M.A. CANTAB.

AND

MATILDA BERNARD

PART I.

MACMILLAN AND CO.

AND NEW YORK
1891

A II rights reserved

\M

TRANSLATORS' PREFACE

THIS translation of the first volume of Professor Lang's Lehrbuch der
Fergleichende Anatomic may be considered as a second edition of the
original work. Professor Lang kindly placed at our disposal his
notes, collected for the purposes of emendation and expansion, and
they have been duly incorporated in the text.

We did not think it necessary to increase the size of the book by
the addition of footnotes, either for the addition of new material
where it seemed to us that the account might with advantage be fuller,
or for calling attention to the opinions of other writers when these
did not happen to agree with that expressed in the text. In one
case a recent discovery (that of Jawarovski of antennae in the embryo
of Trochosa) lent such sudden and unexpected support to the mor-
phology of the Arachnoidea set forth in the text that attention was
called to it in a footnote.

In carrying out our work as translators we have kept as close to
the original as the requirements of our own language would allow.
How far we have succeeded must be decided by those who know by
experience the difficulty of breaking down the long and weighty
German sentences, and of rearranging their contents into readable
English. We have endeavoured to steer a middle course between
the too free use of periphrases which suits the German style, and the
use of single technical terms which better suits the more concise
English sentence. The former style has, we think, some advantages
over the latter, so long as the multiplicity of words does not obscure
the meaning. There can be no doubt that it appeals more directly to
the understanding. The object of such a text-book as this is to

vi COMPARATIVE ANATOMY

enable an ordinarily intelligent reader to obtain a clear grasp of the
facts and of their relation to one another; a knowledge of the
technical terms is of secondary importance, and is easily acquired.

We trust that this translation of the work of our friend and (in
the case of one of us) former teacher, Professor Lang, will prove as
useful to earnest students of biology in the English-speaking world
as it has proved to ourselves.

H. & M. B.

JENA, 1891.

PEEFACE BY PROFESSOR ERNST HAECKEL

THE morphology of the animal body, which is the subject of the
following excellent text-book, has been, during the last forty years,
i.e. since the commencement of my academic career, the favourite
object of my scientific labours. During these four decades no other
science has undergone such profound and striking transformation.

In the years 1852-1856, when I attended in Berlin and Wiirzburg
the masterly anatomical and zoological lectures of Johannes Miiller,
Albert Kolliker, and Franz Leydig, the study of comparative anatomy
and of comparative histology had been brought into great prominence
through Miiller's classical researches in the former, and Leydig's in
the latter field.

Oscar Schmidt's text-book of comparative anatomy which resulted
from the transcription of. Johannes Miiller's summer lectures, and
passed through eight editions, gives evidence of the stage then reached
by that science. Upon this work the present text-book of Professor
Arnold Lang is founded, as its ninth edition ; the present state arid
needs of the science, however, compelled the new editor to rewrite
the whole book on a much larger scale.

Johannes Miiller, the great master in the domains of comparative
anatomy and physiology, died in 1858, after having held sway as the
leader in these sciences for more than quarter of a century. This
very year 1858 saw the simultaneous publication by Charles Darwin
and Alfred Wallace of their preliminary sketches of the Theory of
Selection. The year after appeared the Origin of Species, which at one
stroke ushered in, by means of this theory, a new epoch for the
biological sciences.

viii COMPARATIVE ANATOMY

The theory of the mutual action and reaction of inheritance and
adaptation in the struggle for existence clearly explained the forces
at work in the production of biological phenomena. The facts them-
selves had already been set out in their wonderful array, and com-
parative anatomy had even arranged them with profound philosophical
judgment, but no mechanical explanation of them was forthcoming.

Thomas Huxley in England, and Carl Gegenbaur in Germany, by
means of their well-known text-books of comparative anatomy, were
the first to succeed in revealing in detail the important transformation
which this mechanical explanation of morphological phenomena, by
means of the new theory of descent, had brought about in the
biological sciences.

It was my happiness during the first twelve years of my occupancy
of a University chair in Jena, i.e. from 1861 to 1873, to have
Gegenbaur for my colleague and friend. My own attempts to give
the theory of descent its widest application in those sciences which
are comprised under the term biology owe muck to this stimulating
intercourse, and are embodied in my works : Die G-enerelle M&rplwlogu
(1866), Die Naturliche Schopfungsgeschichte (1868, 8th ed. 1889), and
Die Anthropogenic (1874, 4th ed. 1891).

Oscar Schmidt's and Gegenbaur's text-books and my above-named
works all issued from the University in the small Thuringian town of
Jena, and from the same source has now appeared this text-book of
Professor Arnold Lang, formerly my distinguished pupil, and after-
wards and till quite recently my helpful colleague. Professor
Lang has here successfully carried out the very difficult task
of selecting the most important results from the bewildering mass
of new material afforded by the extensive researches of the last
decades, and of combining them with great judgment. Besides this
he has, more than any former writer, utilised the comparative
history of development in explaining the structure of the animal
body, and has endeavoured always to give the phylogenetic
significance of ontogenetic facts. Lastly, he has, by the clear
systematic reviews of the various classes and orders which precede
the anatomical account of each race, further facilitated the phylo-
genetic comprehension of complicated morphological problems, his

PREFACE BY PROFESSOR ERNST HAECKEL ix

wisely chosen and carefully executed illustrations assisting materially
in this result.

It is therefore with great pleasure that I commend this book to
the English student, in accordance with the wish expressed by my
friend Professor Lang and his translators Mr. and Mrs. Bernard, and
in the hope that the English translation will promote to as great an
extent as the German original the wider study and better com-
prehension of animal morphology, and will attract new students to
this noble science.

ERNST HAECKEL.

JENA, 1891.

AUTHOR'S PREFACE TO PART I1

WHEN requested by the publishers to undertake a new edition of
Dr. Oscar Schmidt's Text-look of Comparative Anatomy, I found that this
could only be done in one of two ways. Either the revision must be
limited to trivial alterations, and to a different choice and a greater
number of illustrations, or the book must be entirely re-written. I
chose -the latter course, which the great advance made in zoological
research seemed to render unavoidable. The result is the Text-book of
Comparative Anatomy, the first part of which I now publish.

In compiling the book I have endeavoured to do full justice to
the numerous important results of the research of the last decades. I
have been less anxious to supply a complete and detailed compendium
of Comparative Anatomy than to emphasise those points which it seems
to me are deserving of special attention. The present work in many
respects exceeds the limits till now usually assigned to text-books of
Comparative Anatomy. It contains, separated as far as possible from
the portion devoted to the main subject, the elements of Compara-
tive Embryology, which will perhaps not be unwelcome to many
students. Following Oscar Schmidt's example, I have prefaced the
Comparative Anatomy of the different animal races by short systematic
reviews, which may be of use to the student of systematic zoology.
The book had also to contain what was necesary for the zoological
education of the medical student.

All these parts, not necessarily belonging to the domain of
Comparative Anatomy, and also many theoretical discussions, are
distinguished from the rest by smaller print.

1 The first part consisted of the first four chapters of this volume.

xii COMPARATIVE ANATOMY

I have taken my own course as regards the arrangement and
method of treating the material. I must naturally leave it to my
fellow-zoologists to decide how far I have succeeded.

I have given special care to the illustration of the book. It
contains a large number of illustrations now for the first time accessible
to the majority of students ; these I have, for the most part, myself
adapted to the requirements of the text. I have to thank the
practised hand of my pupil, the young scientific artist, Mr. Sokolofsky
of Hamburg, for some particularly good illustrations.

I owe my best thanks to my honoured publisher, Mr. Gustav
Fischer, for his obliging courtesy.

JENA, 1888.

CONTENTS

CHAPTER I
THE CELL

PAGE

INTRODUCTION ........ 1

PROTOZOA

Systematic Review ........ 4

1. PROTOPLASM ........ 13

2. ADAPTATIONS FOR LOCOMOTION . . . . . .14

3. MEMBRANES, SHELLS, SKELETAL FORMATIONS . . . .15

4. ADAPTATIONS FOR INGESTION OF FOOD . . . . .16

5. ADAPTATIONS FOR EXCRETION . . . . . . 17

6. TRICHOCYSTS ....... 17

7. STIGMATA (red eye-spots) . . . . . . .18

8. NUCLEI . . . . . . . . .18

9. REPRODUCTION . . . . . . . .18

Literature ......... 22

EGG CELLS, SPERM CELLS, FERTILISATION, SEXUAL REPRODUCTION OF THE

METAZOA ...... 23

THE ANIMAL EGG ........ 25

THE EGG YOLK ........ 26

1. TYPES OF TELOLECITHAL EGGS ..... 27

2. TYPES OF CENTROLECITHAL EGGS .... 28
THE EGG ENVELOPES ....... 28

MALE REPRODUCTIVE CELLS, OR SPERMATOZOA .... 30

MATURATION OF THE EGG . . . . . . 31

FERTILISATION . . . . . . . 32

Literature ......... 34

TISSUE CELLS AND CELL TISSUE ...... 34

1. EPITHELIAL TISSUE . 36

xiv COMPARATIVE ANATOMY

2. CONNECTIVE TISSUE .

3. NEURO-MUSCULAR AND MUSCLE TISSUE

4. NERVE TISSUE ....
Literature

CHAPTEK II

METAZOA

INTRODUCTION .....

Zoopliyta or Ccelenterata

Gastraeadse
Review ........

Literature .........

Porifera
Systematic Review ........

Literature .........

Cnidaria
Systematic Review ........

1. GENERAL .........

2. THE BODY EPITHELIUM . . . . . . .

3. THE GASTRO-CANAL SYSTEM

4. MUSCULATURE . . . . . . . .

5. TENTACLES OF THE CNIDARIA, MARGINAL LOBES OF THE SCYPHOMEDUS.E

6. THE NERVOUS SYSTEM .......

7. THE SENSORY ORGANS .......

8. SUPPORTING ORGANS, PROTECTIVE ORGANS, SKELETON .

9. FUNNEL CAVITIES (Septal Funnels); SUBGENITAL CAVITIES, SUBGENITAL
CHAMBER . . . . . ...

10. THE SEXUAL ORGANS . . . . .

11. THE STRATIFICATION OF THE CNIDARIA BODY ....

12. REPRODUCTION ........

13. ORGANISATION OF THE SIPHONOPHORA . . . . .

14. LIFE-HISTORY OF THE CNIDARIA, ALTERNATION OF GENERATIONS
Literature . . . . . . • .

The fundamental law of Biogenesis — Egg segmentation and the development
of the two primary germinal layers of the Metazoa (gastrulation) — The
ontogeny of the Cnidaria . . . . . .

Literature

CONTENTS xv

CHAPTER III
PLATODES

PAGE

Systematic Review . . . . . . . .133

1. GENEKAL REMARKS . . . . . . .135

2. THE BODY FORM ........ 137

3. THE OUTER BODY EPITHELIUM . . . . .138

4. THE Gr ASTRO-CANAL SYSTEM . . . . . .138

5. SUPPORTING ORGANS, PASSIVE ORGANS OF LOCOMOTION . . 143

6. THE MUSCULATURE ....... 144

7. ORGANS OF ADHESION ....... 145

S. THE NERVOUS SYSTEM ....... 146

9. THE SENSORY ORGANS . . . . . . .149

10. THE BODY PARENCHYMA ..... .151

11. THE EXCRETORY OR WATER- VASCULAR SYSTEM . . . .152

12. THE SEXUAL ORGANS ....... 154

13. ASEXUAL REPRODUCTION AND ITS ORIGIN— THE ORGANISATION OF THE
CESTODA ......... 163

14. ONTOGENY OF THE POLYCLADA ...... 167

15. THE LIFE-HISTORY OF THE TREMATODA . . . . .168

16. THE LIFE-HISTORY OF THE CESTODA ..... 171
THE INFLUENCE OF PARASITISM ON THE STRUCTURE AND DEVELOPMENT OF

ANIMALS ......... 172

STROBILATION AND SEGMENTATION . . . . . .175

Literature . . . . . . . . .176

CHAPTER IV
VERMES

Systematic Review . . . . . . . .178

1. FORM OF BODY AND OUTER ORGANISATION .... 186

2. THE INTEGUMENT ....... 191

3. THE DERMO-MUSCULAR TUBE ...... 193

4. THE PROBOSCIS OF THE NEMERTINA AND ACANTHOCEPHALA . . 197

5. THE INTESTINAL CANAL ....... 200

A. THE FORE-GUT ....... 201

B. THE MID-GUT ....... 205

C. THE HIND-GUT AND THE ANUS 209

xvi COMPARATIVE ANATOMY

6. THE BODY CAVITY, THE MUSCULATURE WHICH PASSES TRANSVERSELY

THROUGH IT, THE DISSEPIMENTS AND MESENTERIES .

7. THE NERVOUS SYSTEM ....

8. SENSORY ORGANS ........

A. ORGANS OF TOUCH . . .

B. EYES

C. OLFACTORY ORGANS (Ciliated Organs) ....

D. ORGANS OF TASTE (Cup-shaped Organs) ....

E. LATERAL ORGANS . . . . .

F. AUDITORY ORGANS .......

G. THE LATERAL EYES OF POLYOPHTHALMUS

9. EXCRETORY ORGANS— NEPHRIDIA .....

10. RESPIRATORY ORGANS .......

11. BLOOD-VASCULAR SYSTEM . . . .

12. GENITAL ORGANS ........

13. PARTHENOGENESIS . . . . . . .

14. ASEXUAL REPRODUCTION BY GEMMATION AND FISSION .

15. STOCK FORMATION .......

16. ONTOGENY OF THE WORMS ......

Literature

CHAPTER V

ARTHROPODA— ARTICULATA

First Division— Crustacea
Systematic Review .....

1. OUTER ORGANISATION ....

A. THE BODY . .

B. THE EXTREMITIES ....

C. THE RESPIRATORY ORGANS— GILLS

2. THE INTEGUMENT ....

3. THE MUSCULATURE . .

4. THE ENTERIC CANAL . .

A. THE FORE-GUT ....

B. THE MID-GUT ....

C. THE HIND-GUT . . .

5. THE NERVOUS SYSTEM ....

6. THE SENSORY ORGANS ....

A. EYES .....

B. OTHER SENSORY ORGANS .

CONTENTS xvii

PAGE

7. THE BLOOD-VASCULAR SYSTEM AND THE BODY CAVITY . . . 358

8. EXCRETORY ORGANS (Antennal Glands, Shell Glands) . . 368

9. THE CONNECTIVE TISSUE .... .370

10. THE SEXUAL ORGANS . . .370

11. SEXUAL DIMORPHISM .... . 376

12. HERMAPHRODITISM IN THE CRUSTACEA . . . 381

13. PARTHENOGENESIS— CYCLIC REPRODUCTION . . . 382

14. ONTOGENY ........ 382

15. PHYLOGENY ........ 406

Literature . . . . • • • • . 411

FIRST APPENDAGE TO THE CRUSTACEA

1. Trilobites . . . . . . . . . .414

2. Gigantostraca ... .415

3. Hemiaspidse .... .416

4. Xiphosura ... .417
Literature .... .... 421

SECOND APPENDAGE TO THE CRUSTACEA

Pantapoda ... .422

Literature . ..... 425

CHAPTER VI
SECOND DIVISION OF THE AETHEOPODA

Tracheata ......... 426

CLASS I. Protracheata ....... 427

Literature ......... 437

CLASS II. Antennata . . . . . . . .438

Systematic Review . . . . . . . .438

1. OUTER ORGANISATION ....... 443

A. THE BODY ........ 443

B. THE LIMBS ........ 445

2. THE INTEGUMENT AND GLANDS . . . . . . 458

3. THE MUSCULATURE ....... 460

4. THE ENTERIC CANAL ....... 460

5. THE NERVOUS SYSTEM ....... 464

6. THE SENSORY ORGANS . . . . . . . 469

A. EYES ......... 469

B. AUDITORY ORGANS ....... 471

C. OLFACTORY AND GUSTATORY ORGANS . . . .474

xviii COMPARATIVE ANATOMY

7. THE CIRCULATORY SYSTEM ....

8. FAT BODIES — LUMINOUS BODIES

9. THE RESPIRATORY ORGANS ....

10. SOUND-PRODUCING APPARATUS ....

11. SEXUAL ORGANS . . . .. .

12. DIMORPHISM — POLYMORPHISM . ~.-;. . .

13. DEVELOPMENT AND LIFE-HISTORY ;.. . ..

14. PHYLOGENY . . . . .
Literature . . . . . • %

CLASS III. Arachnoidea sive Chelicerota
Systematic Review . , . . . ,

1. OUTER ORGANISATION

A. THE BODY ......

B. THE EXTREMITIES . . . • ..

2. THE NERVOUS SYSTEM .....

3. THE EYES. . . . .'.

4. GLANDS OPENING ON THE OUTER INTEGUMENT .

5. THE INTESTINAL CANAL .....

6. THE BLOOD-VASCULAR SYSTEM ....

7. THE RESPIRATORY ORGANS ....

8. THE SEXUAL ORGANS .....

9. ONTOGENY ......

10. PHYLOGENY . .

Literature . . . . '. .

APPENDAGE TO THE ARTHROPODA

Tardigrada . ....

Literature

OF THE

UNIVERSITY

CHAPTEK I

THE CELL

Unicellular animals — The cell as the starting-point in the development of the
higher animals (egg and sperm cells) — The cells which compose the bodies
of those animals (tissue cells and cell tissue).

INTRODUCTION.

THE starting-point of all organic life and of all organic structure is the
cell. The simplest organisms, the lowest animals, are cells. Every
higher animal begins individual existence as a cell, and every higher
organism appears to be composed of cells which have arisen by
multiplication from one cell.

The cell is an organic individual of the first order.

In all cases except that of the lowest organisms the descendants of
one cell unite to form communities or states, which then rank as
individuals of a higher order.

Every higher organism, every bird, every fish, and so on, is such a
cell community. In such a community, the closely united cells share
in the common work. Some undertake one function, some another,
for which they are specially adapted.

Every cell consists of two essential parts : (1) the Protoplasm, and
(2) the Nucleus. The latter may be considered as a special differentia-
tion (specially developed portion) of the protoplasm. Chemically con-
sidered protoplasm is a carbon compound as yet not fully understood ;
it is related to albumen, and is, in life, a stable combination, though
subject to variation within very narrow limits. It is viscid and
capable of swelling. The nucleus lies within the protoplasm as a
chemically and physically differentiated part of it. It is an essential
portion of the cell, in whose multiplication it plays an important part.
According to some observers, if the nucleus is removed from a cell the
latter perishes. If, on the other hand, a nucleus is introduced into an
unnucleated piece of protoplasm, certain characteristic phenomena of
life otherwise absent appear in it.

VOL. I. B

2 COMPARATIVE ANATOMY CHAP, i

There are very simple creatures, small masses of protoplasm, in
which no nucleus has been found. Should the absence of nuclei in these
creatures be established they would rank lower in the scale than cells.
To such organisms the name " Cytods " is applied, and Haeckel has
united them, as the simplest of all organisms, under the group of
Monera.

A frequent though not necessary portion of the cell is the cell
integument or cell membrane, a product of secretion, serving for
protection or support. Such a membrane can also arise by- the
hardening and modification of the peripheral layers of the protoplasm
itself.

A single cell (a unicellular organism, an egg cell) is in itself from
the first capable of all those activities and functions which are contained
in the conception of Life. These phenomena of life, though they may
not as yet be physically and chemically explained, are certainly not to
be referred to the working of any special vital force peculiar to
organisms. There is also no special fundamental substance, no life
substance, which can be found in organisms, and with which a special
vital force is connected. We have to do here with the same forces
and the same substances that we meet with elsewhere in nature.

The life of the cell shows itself in the simplest cases in —

1. Motion. — Protoplasm is contractile. The finest visible portions
can change their relative positions. The cell can change its form and
its position in space.

2. Irritability. — The cell responds to external stimuli by such
movements.

3. Metabolism.- — By means of its life -activity some of the cell
substance is used up, decomposed. What has become useless is
excreted (excretion). By means of the ingestion of food foreign
substances are introduced. These are digestible if, when assimilated
by chemical action, they can be changed into ingredients of protoplasm
(digestion, assimilation). If, owing to their chemical properties, such
a conversion is impossible, they are indigestible, and are expelled out
of the body.

4. Growth. — By nourishment more protoplasm can be produced
than was formerly present. The cell in consequence increases in size —
it grows.

5. Reproduction. — It may be assumed that the size of an indi-
vidual cell is limited. If it exceeds these limits of individual size it
divides into two cells (reproduction by means of division). Each
of the two portions has the same physical and chemical properties as
the mother cell (simplest form of inheritance). The daughter cell by
growth attains the size of the mother cell.

As the cell is the starting point both in the animal and in the vegetable
kingdoms, it can easily be understood that no sharp line of demarcation
between the lower forms in these two kingdoms can be established.
Haeckel has therefore set up an intermediate kingdom, that of the

FIG. 1. — Amoeba polypodia in the successive stages of division. The light spot is the contractile
vacuole, the dark the nucleus (after F. E. Schulze).

FIG. 2.— A, Quadrula symmetrica, after F. E. Schulze. B, Hyalosphenia lata, after F. E.
Schulze. C, Arcella vulgaris, after Hertwig and Lesser. D, Difflugia pyriformis (after Wal-
lich) completed.

4 COMPARATIVE ANATOMY CHAP, i

Protista, consisting of the simplest organisms. But there is also no
sharp line of demarcation between the Protista on the one side and
animals and plants on the other. Some Protista are, chiefly by their
method of nutrition, more nearly related to plants, others to animals.
The latter are called Protozoa, in contradistinction to all other animals,
which are classed as Metazoa.

THE FIKST RACE OR PHYLUM OF THE ANIMAL KINGDOM.

PROTOZOA.
Systematic Review.

CLASS I. Monera.

Simplest organisms. Small masses of protoplasm of varied changing form, in
which till now no nuclei have been demonstrated. Locomotion and ingestion of
food by means of blunt (amoeboid) or long and fine processes (pseudopodia). Repro-
duction by fission and gemmation. All Monera live in water. Protanueba,
Myxodiction, Protomyxa.

CLASS II. Sarcodina.

Unicellular organisms, with nucleus or nuclei. Locomotion and ingestion of food
by means of filose non-vibrating processes of varying length (pseudopodia). Repro-
duction by fission or gemmation.

Sub-Class I. Amcebina.

Naked or shelled Sarcodina of changing shape. Locomotion and ingestion of
food by means of streaming of the body and the formation of processes mostly short
and lobate. Contractile vacuoles generally present. Amoeba (Fig. 1), Arcella
(Fig. 2, C), Difflugia (Fig. 2, Z>), Quadrula (Fig. 2, A), Hyalosphenia (Fig. 2, B}.

Sub-Class II. Rhizopoda.

Sarcodina whose protoplasm secretes a very variously-shaped chitinous, generally
calcareous, shell, which is at first uniaxial. Locomotion and ingestion of food by
means of pseudopodia, which frequently fuse with one another, often in a reticular
manner. Contractile vacuoles generally absent.

A. Imperforata.

Shells of one chamber or more, not perforated by fine pores, but having one or
two larger apertures through which the protoplasm and the pseudopodia pass out.
Miliola (Fig. 3, C), Lituola, Gromia (Fig. 3, A}.

i

B. Perforata.

Shells of one chamber or more, perforated by fine pores for the passage of the
pseudopodia. Globigerina, Rotalia ( Fig. 3, B}.

..•' / '4 -• *•;, c •;.,••• .,:-. .•-•;• •

Ji"fijjj
1 'l/msm:

A1\M^S

Ar^rv-^

m/i m 1 1 ( , ' -

//I \ l\\
// jy v »\

// /m i \

in

•

FIG. 3.— A, Gromia oviformis after M. S. Schultze. £, Rotalia Freyeri, after M. S. Schultze.
C, Miliola (after R. Hertwig) the nuclei in the chambers.

COMPARATIVE ANATOMY

CHAP.

Sub-Class III. Heliozoa.

Globular Sarcodina, naked or clothed with a siliceous skeleton, with fine more
or less stiff pseudopodia radiating on all sides. Contractile vacuoles generally

present in varying numbers. Actinophrys
(Fig. 4), Actinosphcerium, Acanthocystis, Clath-
rulina.

Sub-Class IV. Radiolaria.

Body divided by an originally spherical or
egg-shaped capsular membrane into an outer,
and an inner, nucleated portion (outer and
inner capsules). The outer capsule consists of
protoplasm (without nucleus) and of a gela-
tinous envelope (calymna). The protoplasm
forms a layer round the inner capsule and a
network round the calymna, these two being
connected by means of protoplasmic threads.

FIG. 4.-Actinophrys sol, after Ore- Fine flexible pseudopodia radiate in all dii-ections
nacher. p, Pseudopodia ; n, nucleus ; a,

axial filaments of the pseudopodia. from the surface of the calymna. Skeletons

of extraordinarily various shapes, of silica or of

chitin-like organic substance (acanthin), are seldom wanting. The extra- and intra-
capsular protoplasm are connected through various openings in the cell-membrane.
This division of the marine Sarcodina is wonderfully rich
in forms and shapes. Without contractile vacuoles. Uni-
cellular algse (yellow cells) live symbiotically with the
Radiolaria. The family of the Polycyttaria among the
Spumellaria is distinguished by the formation of colonies.

l**y/.;f%| A. Pomlosa.

£| fi/ Central capsule spherical, without a principal aperture,

°* ~d ^%£ • with innumerable fine pores.

e/r'-*C^SlPa ' *• Spumellaria. — Nucleus central, dividing late in the

& life of the individual. Skeleton siliceous or wanting, never

•/•.«; penetrating into the intra-capsular protoplasm. Thalassi-

Cti - - ,- ffdia collet, Collozoum, Sphcerozoum, Thalassoplancta (Fig. 5),

<- CollospJicera, Dictyastrum.

II. Acantharia. — Nucleus eccentric, dividing early.

/i.Tl

Fio. 5.— Thalassoplancta
brevispicula, part of a sec-
tion, after Haeckel. km,
Capsular membrane ; ip,
intra-capsular ; ep, extra-
capsular protoplasm ; n,
nucleus; nl, nucleoli; ot,
oil-drops ; ca, alveolar ca-

tKurfaS'
s, spicules.

Fio. 6.— Phractaspis prototypus (after Haeckel), skeleton.

Skeleton of acanthin, radiating from the centre of the
central capsule. Acanthometra, Phractaspis (Fig. 6).

PROTOZOA

B. Osculosa.

Central capsule egg-shaped, with a principal aperture at the basal pole of the
chief axis. Skeleton siliceous, always extra- t

capsular. Nucleus dividing late.

III. Nassellaria.— Capsular membrane simple,
a porous area at the oral pole of the chief axis.
Nassella, Cortina (Fig. 7), Cornutdla.

IV. Phaeodaria. — Capsular membrane double ;
at the oral pole of the principal axis an osculum
closed by a radially striped lid, with a central
opening produced in the shape of a chimney. A
collection of pigment bodies (phseodium) in the
calymna. Aulospli<x,ra, Aulactinium (Fig. 8),
Cannopilus, Challengeria.

CLASS III. Flagellata (Mastigophora).

Organisms which are unicellular or united into
simple cell colonies ; properly standing on the
boundary line between the animal and vegetable
kingdoms, since some groups are directly con-
nected morphologically and physiologically with
the lowest plants, others, chiefly by their manner
of taking nourishment (ingestion of solid food)
with animals. Furnished during the principal ring

^1^

1,2,3,4, principal rays ; n, nu-

part of life with one or more vibratile flagella, cleus ; dt, oil-drops ; pc, podoconus.

FIG. 8.— Aulactinium actinastrum, after Haeckel. n, Nucleus ; c, calymna
km, capsular membrane ; op, operculum ; ph, pliaeodium.

8

COMPARATIVE ANATOMY

CHAP.

serving for locomotion, and often also for capturing food. With contractile vacuoles.
Multiplication by fission, or formation of spores, or gemmation, often after previous
copulation of the reproducing individuals.

Order 1. Flagellata s. str.

During active life armed exclusively with flagella (without collar or cilia). Monas,
Euglena, Chilomonas (Fig. 9), Uudorina, Pandorina, Stephanosphcera, Volwx (Fig.
21, p. 21).

1

FIG. 9. —Chilomonas Paramaecium,
after Biitschli. s, Oral aperture ; cv, con-
tractile vacuole ; n, nucleus.

FIG. 10.— Protospongia Haeckelii, after Kent.

Order 2. Choanaflagellata.

Flagella at their basal portion surrounded by a funnel-shaped collar. Phalan-
sterium, Salpingoeca, Protospongia (Fig. 10).

Order 3. Cystoflagellata.

The protoplasm shows a reticulated structure similar to that of vegetable cells.
Noctiluca (Fig. 11), Leptodiscus.

Order 4. Dinoflagellata (Cilioflagellata).

Shelled forms ; besides the freely projecting flagellum there is a second, peculiarly
placed, in a special groove running across the body, giving the appearance of vibrat-
ing cilia which deceived earlier observers. Peridinium, Ceratium (Fig. 12).

CLASS IV. Gregarina.

Parasitic Protozoa of elongated form. Invariably one nucleus. Without pseudo-
podia, without cilia, without contractile vacuoles, without special differentiation of
the protoplasm, with outer cell integument. Multiplication by spore-formation, with
previous copulation or conjugation.

PROTOZOA

Order 1. Monocystidse.
Body simple. Monocystis (in the Earthworm), Urospora (Fig. 13, B).

Order 2. Polycystidae.
Body divided by a partition wall into anterior (protomerit) and posterior (deuto-

f

FIG. 11.— Noctiluca Miliaris (after Biitschli), some-
what altered, tig, Band-like flagellum ; /, flagellum ;
TO, oral aperture ; n, nucleus ; 6 and c, spores of
Noctiluca.

FIG. 12.— Ceratium Tripus (after
Biitschli) somewhat modified.

merit) divisions. The protomerit often again divides, the anterior division (epi-
merit) being furnished with hooks, etc., for adhesion ; this part is lost in time.
Nucleus in deutomerit. Actinocephalus, Stylorhynchus (Fig. * o

13, A\ Clepsidrina.

CLASS V. Infusoria (Ciliata).

Unicellular Protozoa, rarely united in simple colonies,
with cilia or cilia-like processes for locomotion and alimenta-
tion. Generally with contractile vacuoles, oral and anal
apertures. "With double nucleus : a variously formed large
macronudeus and a small micronudeus (erroneously called
nucleolus). Reproduction by fission, conjugation frequent.

Order 1. Holotricha.

The whole surface is equally covered with fine cilia, often
arranged in rows. Paramcecium (Fig. 20, p. 17), Trachelius.

Order 2. Heterotricha.

Possess, besides the clothing of cilia which spreads
equally over the whole surface, a distinctly developed zone
of bristle- or stylet-shaped cilia near the mouth. Spiro-
stomum, Stentor (Fig. 15), Freia, Balantidium.

Order 3. Hypotricha.

Dorsal and ventral surfaces sharply distinguished. Ven-

Fio. 13.— A, Stylo-
rhynchus longicollis,
after Aimd Schneider.
ep, Epimerit ; pm, proto-
merit ; dm, deutomerit.
B, Urospora Ssenuridis.
Conjugation of individ-
uals, after Kolliker.

tral surface ciliated. Chilodon, Euplotes, Stylonychia (Fig. 14), Oxytridia, Urostyla.

10

COMPARATIVE ANATOMY

CHAP.

FIG. 14.— Stylonychia mytilus, after Stein
(from Claus's Zoology), seen from the ventral'
surface. Wz, Adoral ciliated zone ; C, contractile
vacuole ; N, nucleus ; N", micronucleus ; A , anus.

FIG. 15. — Stentor Roeselii, after Stein
(Claus's Zoology). 0, Oral opening with oeso^
phagus ; PV, contractile vacuole N, nucleus.

FIG. 16.— Vorticella microstoma, after Stein (from Claus's Text-book of Zoology), a, Dividing
longitudinally ; N, nucleus ; b, after complete division one part severs itself after having formed
a ring of cilia behind ; w, oral ciliary organ ; c, vorticellse in conjugation ; t, the adhering bud-like
individuals.

PROTOZOA
Order 4. Peritricha.

11

Body globular or cylindrical, only partially ciliated, either near the mouth in a
spiral, or in a belt. Forticella (Fig. 16), Carchesium, Epistylis, Trichodina, Strom-
bidium, Tintinnus, Ophrydium.

FIG. 17.— Podophrya gemmipara, after R. Hertwig (from Claus's Zoology) ; a, with protruded
suctorial tentacles and capturing processes, with two contractile vacuoles; &, with buds into
which processes of the branched nucleus N enter ; c, one of the buds broken loose.

CLASS VI. Suctoria (Acineta).

Ciliated only in swarm-spore stage. With suctorial tentacles, by means of which
they penetrate the bodies of Infusoria (principally) and suck in their protoplasm.
Reproduction by gemmation. Acineta, Podoplirya (Fig. 17), Dendrocometes.

CLASS VII. Catallacta.

Single genus and species : Mago-
sphcera planula (Fig. 18), found
swimming freely in the sea on the
coast of Norway. A globular colony
of pear-shaped cells, the stems of
which meet at the centre of the
sphere ; the outer surface of the
cells provided with cilia.

Reproduction: the colony dis-
solves into single cells, which sink
to the bottom, become first amoeboid,
then encysted, and make a new
colony within the capsule by means
of successive division ; this new
colony frees itself later from the
capsule.

FIG. 18.— Magosphaera planula, after Haeckel.

12

COMPARATIVE ANATOMY

CHAP.

The Protozoa are unicellular organisms, or simple colonies of
similar unicellular organisms. The typical character of a unicellular
organism often appears disturbed by the presence of more than one
nucleus, the original simple nucleus by successive division separating
into several or even many nuclei. These divisions of the nucleus are
in some cases connected with reproduction, as its first stage, in others
the rest of the cell remains altogether unaffected.

Although the Protista are unicellular organisms they show a
remarkable variety in form, and in some cases a great complexity of

FIG. 19.— Amoeba polypodia. In the successive stages of division. The light spot is the
contractile vacuole, the dark spot the nucleus (after F. E. Schulze).

structure. Modifications may arise specially adapted to the most
varied functions of life; these, however, unlike the modifications in
the Metazoa, are always in one and the same cell. Nowhere in the
organic world does the cell reach so high a degree of morphological
differentiation as in certain Protozoa. In the lower Protozoa, on the
other hand, the cell in its simplest form shows itself capable of all
essential life processes. An Amoeba (Fig. 19) may consist of a small
mass of uniformly granulated protoplasm containing a nucleus.
Locomotion is caused by the streaming forward of the protoplasm at
some points and the consequent formation of processes of varying

! PROTOZOA 13

shape (amoeboid). At other points processes already formed are
withdrawn. The mass of the plasm flows towards the newly formed
processes, and locomotion thus appears as an irregular streaming.
In this way small foreign bodies are taken into the Amoeba body ;
if they can be assimilated they are digested, if not they are left
behind as the Amoeba moves forward. The Amoeba grows as its
nourishment increases. It multiplies by the nucleus becoming con-
stricted in the shape of a dumb-bell and finally dividing into two nuclei.
After complete division of the nucleus the plasm of the body also
becomes constricted and falls into two parts, each with its nucleus.
In this way each daughter Amoeba is, except in the matter of size, like
the mother. Reproduction by fission.

I. Protoplasm.

The protoplasm of many Protozoa (of certain Monera, of the
Pihizopoda, a few Amoeba, and most Flagellata) is tolerably homogeneous,
i.e. uniformly granular. In most cases, however, there is a differentia-
tion into an outer and an inner layer ; the former firmer, hyaline or
more often finely granular, and generally more contractile (cortical
layer, ectoplasm, eetosare) ; the latter more liquid and granular
(medulla, endoplasm, endosare). In some Heliozoa the endosarc is
the more homogeneous, the eetosare granular. There is generally no
sharply defined barrier between the two layers. In the Radiolaria
the protoplasm is divided by a membrane (capsular membrane) into
two parts, the extra- and intra-capsular protoplasm, which, how-
ever, communicate with one another by means of various perforations
of the membrane, and thus do not correspond with the ecto- and
endo-plasm above mentioned. In a similar way, in the RUzopoda with
calcareous shells, some of the protoplasm surrounds the shell.

The capsular membrane possesses either numerous fine pores (Spumellaria,
Acantharia), or one single round aperture (osculum), with a porous cover or
operculum (Nasellaria), or, in addition to two or more apertures, one principal
aperture closed by a radially striated cover produced externally in the shape of a
tube (Phoeodaria}.

Both ectoplasm and endoplasm are distinguished, in the case of
most Protozoa, by special structural modifications and differentiations.
The ectoplasm supplies the adaptations for locomotion and alimenta-
tion— pseudopodia, cilia, flagella, suctorial tentacles which also serve
as feelers, oral and anal apertures. It frequently forms on its surface
a cell integument (cuticle) which may form the substratum of a
great variety of shell structures. The ectoplasm generally also supplies
the material for the various skeletons met with in many forms. The
contractile vacuoles and the stinging capsules (trichocysts), where such
are found, almost always lie in it. In some cases it gives rise to
special contractile portions (Infusoria, Vorticella).

14 COMPARATIVE ANATOMY CHAP.

The digestion of solid food takes place in the endoplasm. In it
lie the nucleus or nuclei. It often contains non-contractile vacuoles,
food vacuoles, products of excretion (crystals), fat drops, oil drops, etc.,
gas-bubbles, pigment granules. The endoplasm occasionally shows slow
streamings (circulation in Infusoria).

In the Heliozoa, the protoplasm becomes, by means of the ap-
pearance of numerous non-contractile vacuoles, spongy, alveolar. In
the Cystoflagellata we find a central plasmic portion from which the
protoplasm radiates to the surface in a network forming numerous
vacuoles. This arrangement of the plasm resembles that in the plant
cells. Granular movement can be seen in the cords and strands of the
protoplasm ; between them lies cell-sap.

In many true Flagellata which take in no solid food but feed in
the manner of plants the protoplasm contains small pigment bodies
(chlorophyl or similar pigment), organs of assimilation which form
amylum. Chlorophyl (proper to the animal body and formed by it)
is also to be found in Infusoria (in a Forticella) in a dissolved form.

Besides these there are unicellular algae, which live symbiotically
with many forms of Protozoa in the same way as Algae and Fungi live
together to form lichens (yellow cells, and pigment bodies of the
phseodium (?), of the Radiolaria, chlorophyl bodies of many Infusoria).

II. Adaptations for Locomotion.

The locomotion of the Amoeba (and many Monera) by means of blunt
processes of varying shape has been described above. In the Rhizopoda,
Heliozoa, and Radiolaria, there are long filose processes of the exoplasm
(where such a differentiation exists) which radiate from the body on
all sides — the so-called pseudopodia. These processes serve, however,
more for taking in food and as a hydrostatic apparatus than as organs
of active locomotion. There are two principal sorts of pseudopodia —
myxopodia and axopodia. The former are not stiff, they are pro-
trusible and retractile, can fuse with neighbouring pseudopodia into a
network, and, chiefly in Rhizopoda, can collect into small masses by
flowing together outside the body at the points where they meet with
food. Such myxopodia are characteristic of the Rhizopoda and most
Radwlaria. The axopodia, which are found in the Heliozoa and in
Acantharia among Radiolaria, are, on the contrary, more or less stiff,
and not inclined to reticulate and fuse. In their axes there generally
runs a stiff axial filament, a sort of elastic organ of support formed of
organic substance. These axial filaments run towards the central
point of the body — to the boundary of the endoplasm (Actinosphcerium),
or to the nucleus near the centre (Adinophrys), or they meet actually
in the centre (Acantharia). All pseudopodia show more or less swift
granular streamings.

In the Rhizopoda with calcareous shells, part of the protoplasm spreads itself
in a layer over the latter, and from this layer the pseudopodia radiate. In the

i PROTOZOA 15

Radiolaria, whose bodies are surrounded by a thick gelatinous envelope (calymna),
filled with vacuoles and alveoles, the case is more complicated. The extra-capsular
protoplasm forms a layer on the exterior of the capsular membrane (sarcomatrix)
and, further, a network on the surface of the calymna (sarcodyctium). From the
latter, which is connected with the former by intra-calymnary protoplasmic cords
and strands, the extra-calymnary pseudopodia radiate.

Flagella cannot be sharply distinguished from pseudopodia. They
are processes of the exoplasm (where such exists) which, in the
Flagellata and in early stages of the life of many other Protista, appear
at special points of the body in small numbers (one or two, rarely
more).

Undulating membranes have also been observed in a few Flagellata.
Cilia are characteristic of the Infusoria and the young stage of the
Suctoria. These are fine vibratile processes of the ectoplasm, which
vary in length, strength, and shape ; they are arranged in different
characteristic ways in each division, either spreading over the whole
body or restricted to certain regions, and specially forming spirals about
the mouth, or belts.

In the Cystoflagellata (Noctiluca), besides the ordinary flagellum at
the base of the oral aperture, there is a large band-like flagellum
which moves slowly ; this is a protoplasmic outgrowth of complicated
structure (Fig. 11, p. 9, bg).

In the Gregarina special organs of locomotion are wanting; the
ectoplasm here appears peculiarly contractile, just as in the Infusoria,
where it is often differentiated into parallel contractile and non-
contractile strips (furrows and ribs). Another differentiation of
the ectoplasm is the so-called stalk muscle of the Forticella (Fig. 16,
p. 10), which in contracting rolls itself up spirally. Here belong also
the myophrises of the Acanthometridce — filamentous processes which can
contract suddenly, but not repeatedly, and which are arranged on the
sarcodyctium in a circle round each skeletal spine. It is supposed
that they perform hydrostatic functions.

In the Suctoria are found variable processes, mostly terminated by
a knob, and used as suctorial tubes, which are closely connected with
pseudopodia. The contents of the body of the penetrated Infusorian
or Alg flows through the suctorial tube into the body of the Acinetan.

III. Membranes, Shells, Skeletal Formations.

These are extraordinarily numerous. Many Protista, Amoeba, and
Flagellata are naked. In simple cases the protoplasm secretes at the
surface a chitinous membrane (Gromia), which may be composed of
small plates (Arcella). Occasionally small foreign particles are united,
by a binding medium supplied by the body, into a sort of case
(Difflugia). A fine cuticle is found in most Infusoria • in some cases
this may harden into shells or carapaces. A cuticle (cell-integu-
ment) is further found in Gregarina and many Flagellata, and can

16 COMPARATIVE ANATOMY CHAP.

develop into a shell by becoming firmer and severing itself from the
protoplasm. The structure of the shell of the Diiwflagellata is com-
plicated ; in this case alone the shell consists of cellulose (the substance
of the membrane of plant cells). Gelatinous envelopes are also widely
dispersed ; they are found in Heliozoa, all Radiolaria (Calymna) and
many Infusoria, especially in the attached forms. In the last case they
are often plastered over with small foreign particles. In the marine
Rhizopoda there arise, by secondary impregnation of an originally
chitinous membrane with carbonate of lime, and by a further deposit
of the same, variously shaped calcareous shells. These have either
one chamber (Monothalamia) or become many-chambered (Polytha-
lamia), with varied arrangement of the chambers. The shells either
possess one large aperture for the emission of the protoplasm (Imper-
forata), or they are perforated by numerous fine pores (Perforata).

In Heliozoa pieces of skeleton come into existence by the impreg-
nation of an organic substratum with silica; these generally lie
loosely on the surface of the soft body. In Clathrulina, however, a
globular, much fenestrated, siliceous skeleton is formed.

The skeletal formations of the Radiolaria, which are rarely
absent, show a wonderful diversity of structure. They consist of
fenestrated spheres, several of which, connected by means of rods,
are often contained one within another, or of regularly arranged
radial spicules, or of bivalve shells, etc. etc. We have to distinguish
two altogether different skeletal forms. One consists of silica and
never penetrates into the central capsule; the other consists of
an organic substance akin to chitin (aeanthin), and is always cen-
trogen, i.e. it radiates from the middle of the central capsule
(Acantharia).

In one principal division of the Gregarina (Polydstidce) the extended
body is divided by a partition into an anterior part (protomerit) and
a posterior part (deutomerit). Another partition may cut off from
the protomerit a front portion serving for adhesion and temporarily
provided with hooks, etc. (epimerit). The partitions are, like the
cell integument itself, prpducts of the ectoplasm.

In contrast to the envelopes and skeletons above mentioned, we
have the cysts or capsules. The formation of capsules and cysts
(eneystation) takes place in the most varied Protozoa for protection
against dessication and putrefaction, after acquisition of food to admit
of undisturbed digestion, in hibernation, etc.; very often reproduction
(by fission, gemmation, or sporulation) takes place after the formation
of, and under the protection of, a cyst.

IV. Adaptations for Ingestion of Food.

The organs of locomotion (amo3boid processes, pseudopodia,
suctorial tubules, flagella, cilia, etc.) generally also serve for the
seizing and sucking in of food, and for the formation of currents which

PROTOZOA

17

bring it within reach. There are no special points of the body for
the taking in of food in the Monera, Sarcodina, Orega/rina, Suctoria, and
those Flagellata which feed after the manner of plants. In other
Flagellata and in the Infusoria, there is at one special part of the body
(in the Flagellata at the base of the chief flagellum) a depression of
the ectoplasm (mouth and oesophagus), through which solid food is
passed into the endoplasm. In the In-
fusoria there are, closely connected with
the mouth, cilia specially arranged, in
circles or spirals, which whirl into it the
minute nutritive particles. A certain part
of the body may also temporarily function
as mouth. In this case the aperture is
only visible at the moment of feeding.
An anal aperture or anal spot can also
be found for the evacuation of undigested
remnants of food.

V. Adaptations for Excretion.

The so-called contractile vaeuoles
may, with doubtful accuracy, be con-
sidered as adaptations for excretion, i.e.
for removing from the body the useless ^«- ^.-Paramecium aureiia. M,

. & . , . J , Medulla ; C, cortical layer ; n, macro-

products Of metabolism. These Vaeuoles nucleus ; n', micronucleus ; CM, cuticle ;

are found in most Amoeba, Heliozoa, ci> cilia • f> trichocysts ; v, food vacu-
Flagdhia (excepting Cystoflagellata), and ^L^^',%^^

Infusoria, but are wanting in the Rhizo- ation of vacuole (after Ray Lankester).

poda, Badiolaria, Gregarina, and Cysto-

flagellata. They vary greatly in number. Where there is only one
it generally has a fixed position. There is no sharp distinction be-
tween contractile and non-contractile vaeuoles. The first are vaeuoles
without walls, filled with liquid, which in cases where a differentiation
into ecto- and endo-plasm exists generally lie in the former. They
gradually expand and then contract more quickly, often suddenly.
Before and during contraction they move towards the surface and
empty out their contents through an aperture only visible at the
moment of evacuation. They again arise as a small drop, or as several
drops which unite later. Perhaps they also serve the purpose of
keeping up an exchange of water (evacuation of water taken in by the
mouth), and thus of respiration.

VI. Triehoeysts.

Small vesicles containing threads, which can be shot out rapidly
like the nematocysts of the Ccelenterata, are sometimes found in the
ectoplasm of the Infusoria and in one of the Flagellata (Fig. 20 t).
VOL. i o

18 COMPARATIVE ANATOMY CHAP.

VII. Stigmata (red eye-spots).

Stigmata are found, generally singly, in many coloured Flagellata.
It is highly doubtful if these are organs sensitive to light.

VIII. Nuclei.

These seem to be absent in the Monera. In all the other Protozoa
they are found, either singly (many Amoeba, some Heliozoa, all Gregarina,
and most Flagellata and Infusoria) or in numbers. They lie in the
endoplasm — in the Eadiolaria in the central capsule, in the Gregarina
in the deutomerit — and are either vesicular, with membrane, sap, and
one or more nucleoli, occasionally with a sort of nuclear framework,
or homogeneous. They vary greatly in shape. The nuclear processes
in most Infusoria present complications. We here find a double
nucleus, viz. a large macronueleus, and, lying more or less close to
this, a small mieronueleus (formerly erroneously called nucleolus).
The mieronueleus plays an important part in conjugation; the
macronueleus, on the contrary, during this process falls to pieces and
degenerates in a peculiar manner. When conjugation begins, the
mieronueleus divides twice, i.e. into four parts. Three of these four
parts disappear, while the fourth again divides into two nuclei — the
migratory nucleus and the stationary nucleus. The most important
process during conjugation is the mutual exchange of the migratory
nuclei of the two conjugating individuals A and B. The migratory
nucleus of A passes over into B and fuses with the stationary nucleus
of B, while the migratory nucleus of B passes over into A to fuse with
the stationary nucleus of A. A new macronueleus and a new micro-
nucleus arise, in the individuals which separate after conjugation, out
of the nucleus which results from the fusing of the migratory and
stationary nuclei.

The division of the nucleus is either a direct division (constriction,
dumb-bell stage, separation into halves), or it resembles the indirect
nuclear division found among Metazoa, which will be described later.

IX. Reproduction.

The phenomena of reproduction among the Protozoa deserve more
detailed investigation, as we find included under this head a tolerably
complete series of intermediate stages between the simplest reproduc-
tion by fission and sexual reproduction.

Reproduction by simple binary fission (cross, longitudinal, and
diagonal fission) takes place in nearly all divisions of the Protozoa.
It is especially characteristic of the Monera, many Rhizopoda, many
Flagellata, and all Infusoria. It is, however, not observed in the
Gregarina.

Reproduction by budding or gemmation is, in the simplest cases,

i . PROTOZOA 19

distinguished from the above in that one part (the bud) is smaller
than the other (the mother). The small size of the bud in most cases
makes possible the production of numerous buds on the surface of the
mother. This sort of reproduction is often found together with repro-
duction by fission in Ehizopoda, Heliozoa, Radiolaria, a few Gregarina,
Noctiluca, and Suctoria.

In many forms reproduction by fission and gemmation are probably
preceded by a conjugation (temporary connection or fusing) or
copulation (permanent fusing) of two individuals.

In many Protozoa, belonging to the most varied forms, the
individuals which are produced by fission or gemmation do not
separate entirely from each other, but remain more or less closely
connected, and so form colonies of unicellular organisms.

These colonies are of the greatest importance, as they represent a
lower stage of the cell colonies of the Metazoa, and in many cases
reproduce in a manner which strongly reminds us of the sexual
reproduction of Metazoa and plants (see below).

As an example of reproduction by gemmation we choose the Noctiluca, where
it occurs, probably after previous copulation of two individual Noctiluca, side by
side with simple reproduction by fission. The typical process is briefly as follows.
Gemmation occurs only in such individuals as have, when copulation has ended, lost
their organs of locomotion and mouths, and are thus simple globular bodies, on
the walls of which the chief mass of protoplasm (central plasm), with the nucleus,
is still to be found in its original place. The central plasm at this spot bulges out
somewhat, its nucleus divides by a kind of indirect fission, and the prominence at
the same time separates into two by a furrow. The division of the protoplasm is,
both here and in the following stages, merely superficial, since its deeper part remains
undivided. By continuous fission, 4, 8, 16, 32, 64, and up to 512 -nuclei arise, and
the same number of superficial prominences of protoplasm. Then each prominence
with its nucleus separates below the surface also from its neighbours and forms a
separate bud, on which a flagellurii and a peculiar process are developed ; this bud
leaves the mother animal as a spore (Fig. 11, b, c, p. 9). The further develop-
ment of these spores into young Noctiluca, has not yet been investigated. The whole
process of the formation of buds is very similar to the discoidal furrowing of the
meroblastie egg in the Metazoa, of which we shall speak later.

Eeproduction by continuous fission and spore-formation is very
common among the Protozoa. In the latter case the whole body falls
to pieces, or else the greater part of it simultaneously dissolves into a
usually very large number of nucleated portions, i.e. spores. Both
these methods of reproduction are generally accomplished in resting
encysted individuals, and often after previous copulation or conjugation ;
this is especially the case with Gregarina and Flagellata. They may
occur together with ordinary reproduction by fission. The spores are
generally capable of free movement, occasionally amoeboid, or as
swarm spores they are furnished with a flagellum or several flagella
(Flagellata, Radiolaria, some Heliozoa, and Ehizopoda). Occasionally
(Gregarina) the spores themselves redivide, and only the portions thus
arising grow into adult animals.

20 COMPARATIVE ANATOMY CHAP.

Reproduction of Colony-forming Protozoa.

Freely swimming and stationary attached colonies alike come into
existence by incomplete fission and gemmation. Among the colonial
Radiolaria (the Polycyttaria among the Spumellaria) separate colonies
can mingle with each other ; colonies also can multiply by fission.

The ordinary method of reproduction of colonial Flagellata and
Radiolaria is the production of swarm spores furnished with flagella,
which takes place by simultaneous, or more often by successive, division
of the body into numerous portions.

In the Radiolaria, the contents of the central capsule alone take
part in the formation of spores ; and this process is preceded by the
early or later division of the originally simple nucleus. In the case
of Radiolaria which do not form colonies, every spore becomes
a Radiolarian. In the colonial species, however, two sorts of spores
are developed alternately — (1) isospores, representing the usual
Radiolarian spores, and (2) anisospores, of which again there are two
kinds, smaller mierospores and larger maerospores. The isospores
develop direct into young Radiolaria ; the anisospores most probably
do so only after the copulation of a micro- and a macro-spore. The
young Radiolarian, by repeated fission and gemmation (formation of
so called extracapsular bodies), produces a colony. The macro- and
micro-spores are either formed in one and the same individual, or in
different individuals of the colony. We have here in all probability
a regular alternation of generations, one of which reproduces by means
of isospores, the other by copulating anisospores.

The reproduction of the colony-forming Flagellata is particularly
important and interesting. In the simplest cases every individual
of the colony falls by successive fission into a certain number
of portions which sever themselves from the mother - colony,
forming daughter -colonies. In other cases (Pand&rina), after a
number of generations reproducing themselves as above, a generation
arises whose individuals also divide; but the resulting portions
(gametes) do not remain united, they separate. These gametes
copulate in pairs, the individuals of each pair often differing in size.
The product of copulation (zygote), after a resting stage of some
duration, again produces a colony by continuous incomplete fission.
The reproduction of the Eudorina is distinguished from that of the
Pandorina by the formation of two sharply contrasted sorts of gametes,
male and female. The sexual generation which produces gametes,
and which follows a series of generations reproducing in the usual
asexual manner, is either male or female. In the female colony there
are certain individuals (ovoid gametes) distinguished by unusual
size; in the male the individuals divide into groups (plates) of 32-64
spermoid gametes, each of which has two flagella. The plates sever
themselves and swim about freely. If such a structure comes in

PROTOZOA

21

contact with a female colony, it remains attached to it and breaks up
into single male gametes. Volvox (which leads almost directly by its
method of reproduction to the higher plants and animals) is closely
connected with JEudorina.

In Volwx (Fig. 21) the colony appears on a higher scale of de-
velopment, as a division of labour takes place among the different
individuals. Only some individuals are capable of reproduction ; in the
asexual generation these individuals are the parthenogonidia (every
parthenogonidium produces, by continuous incomplete fission, a colony
which separates from the mother colony); in the sexual generation
they are gametes. Unlike the Endorina, Volwx produces, in one and
the same colony, male and female gametes. The female gametes are
simple individuals of the colony, only distinguished by their greater

, s

J

FIG. 21.— Volvox globator.— Sexual, hermaphrodite colony, after Cienkovsky and Biitschli,
combined and somewhat diagrammatic. S, Male gametes (spermatozoa) ; 0, female gametes (eggs).

growth. The male gametes, on the contrary, arise in masses, by the
successive fission of certain individuals (cells) of the colony. From one
such mother-cell as many as 128 male gametes may proceed. The
male gametes separate, move about by means of their flagella, and
copulate with the female gametes. By successive division (into 2, 4,
8, 16, and so on) of the stationary zygotes a colony, i.e. a young Volvox,
arises.

We have here before us a true alternation of generations,
asexually reproducing generations alternating with sexually reproducing
generations. The sexual reproduction corresponds with the method of
reproduction of the Metazoa and higher plants. The hermaphrodite
Volvox (Flagellate colony) corresponds with a very simple hermaphro-
dite metazoon. The female gametes represent the eggs, the male
gametes the spermatozoa; the copulation of the female and male
gametes corresponds with a simple form of fertilisation of the egg by
means of spermatozoa. The zygote represents the fertilised egg. The
formation of a new Volvox colony by successive fission of the zygote

22 COMPARATIVE ANATOMY CHAP.

answers to the repeated division of the egg-cell, described in the
Metazoa as furrowing or segmentation.

In the colonial Infusoria also the copulation of larger with smaller
individuals has been observed (Fig. 16, p. 10, c).

Review of the most important Literature.
General works on Zoology, Comparative Anatomy, Ontogeny, etc.

G. Cuvier. Lemons d'anatomie compare, publics par DumAril et Duvernoy. 5 vols.

Paris, 1799-1805. Id. Second edition, 8 vols. Paris, 1835-46.
J. F. Meckel. System der vergleichenden Anatomic. 6 Bande. Halle, 1821-1833.
C. E. von Bar. Ueber Entwickelungsgeschichtc der Thiere. 2 Bande. Konigsberg,

1828 bis 1837.

J. Miiller. Handbuch der Physiologic des Menschen. 2 Bde. 4 Aufl. Coblenz, 1844.
R. Leuckart. Ueber die Morphologie und die Verwandtschaftsverhaltnisse der

wirbellosen Thiere. Braunschweig, 1848.
v. Siebold and Stannius. Handbuch der Zootomie. Berlin, 1854. (Vertebrates

incomplete. )
F. Leydig. Vom Bauc des thierischen Kdrpers. I. Band. 1. Halfte. Tubingen,

1864. (With atlas.)
The same. Lehrbuch der Histologie des Menschen und der Thiere. Frankfurt,.

1857.
C. Vogt. Zoologische Brief e. 2 Bande. 1859.

E. Haeckel. Generelle Morphologie der Organismen. 2 Bde. Berlin, 1866.
The same. Natilrliche Schopfungsgeschichte. New edition, 1890.

The same. Anthropogenic. Leipzig, 1876.

T. H. Huxley. A Manual of the Anatomy of Invertebrated Animals. 1877.

The same. A Manual of the Anatomy of Vertebrated Animals. 1871.

C. Gegenbaur. Grundzuge der vergleichenden Anatomic. Leipzig, 1870.

The same. Grundriss der vergleichenden Anatomie. Leipzig, 1878. Translated by

F. J. Bell, revised by E. Ray Lankester, as Elements of Comparative Anatomy.

London, 1878.

F. M. Balfour. Comparative Embryology. 2 vols. London, 1880-81.

C. Glaus. Lehrbuch der Zoologie. 3 Auflage. Marburg u. Leipzig, 1885.

H. Ludwig. J. Leunis Synopsis der Thierkunde. 2 Bande. Hannover, 1883-1886.

C. Vogt u. E. Jung. Lehrbuch der praktischen ^vergleichenden Anatomie. I. Band.

1888.

V. Cams. Geschichte der Zoologie. Munchen, 1872.
Ch. Darwin's Works.
Milne-Edwards. Lemons sur la physiologic et V anatomic compare'e de I'homme et des

animaux.
Bronn's Klassen und Ordnungen des Thierreichs, etc. etc.

Protozoa.

E. Haeckel. Studien uber Moneren und andere Protisten. Leipzig, 1870.

The same. Die Radiolarien. Eine Monographic. Bd. I. II. u. III. Berlin, 1862,

1887, 1888.
The same. Report on the Radiolaria collected by H. M.S. "Challenger." London,

1887.
M. Schulze. Ueber den Organismus der Polythalamien. Leipzig, 1854.

i EGG CELLS, SPERM CELLS 23

R. Hertwig. Der Organismus der Radiolarien. Jena, 1879.

Ch. G. Ehrenberg. Die Infusionsthiere als vollJcommene Organismen. Leipzig, 1838.

Claparede et Lachmann. Etudes sur Us Infusoires et Us Ehizopodes. .Geneve, 1858-

1859.

Fr. Stein. Der Organismus der Infusionsthiere. I. -III. Leipzig, 1859, 1867, 1879.
0. Biitschli. Ueber die Conjugation der Infusorien, in: Studien uber die ersten

Entwickelungswrgange, etc. Frankfurt, 1876.
Saville Kent. A Manual of the Infusoria. London, 1880-1882.
Dujardin. Histoire naturelle des Infusoires, in : Suites a Buff on. Paris, 1841.
Carpenter. Introduction to the Study of the Foraminifera. London Ray Soc. , 1862.
J. Miiller. Ueber die Thalassicolen, Polycystinen und Acanthometren. Abhandl. der

Berliner Akadcmie, 1858.
K. Brandt. Die koloniebildenden Radiolarien (Sphaerozoeen) des Golfes von Neapel.

Berlin, 1885.
0. Biitschli. "Protozoa," newly edited, in Bronn's Klassen und Ordnungen des

Thierreichs. I. Band. Not yet completed.

Balbiani. Various treatises in the Journal de TAnatomie et de la Physiologic. I. -IV.
Various important treatises of Haeckel, Hertwig, Schneider, Engelmann, Maupas,

F. E. Schulze, Greef, Bergh, Czienkovsky, Jikeli, etc.

Egg Cells, Sperm Cells, Fertilisation, Sexual Reproduction
of the Metazoa.

Whereas Protozoa are either simple cells or colonies of similar
cells, all other animals, or Metazoa, appear as complicated communities,
the individual cells of which are no longer similar. Division of
labour arises among the cells; every cell (or group of cells) in
the community having to fulfil only one special function is con-
stituted in correspondence with this function (polymorphism of the
cells of a Metazoan colony). However wonderfully complicated such
a cell community may be, it always develops (except in cases of
asexual reproduction) by means of the continuous division of one
single cell, the fertilised egg. This is the product of the fusing of
a female reproductive cell with a male reproductive cell, i.e. it is
the result of fertilisation. Reproduction by means of such sexually
differentiated reproductive cells is called sexual reproduction. In
all forms of Metazoa (with a few not quite certainly established
exceptions) sexual reproduction occurs at least at times, and con-
stitutes an essential characteristic by which Metazoa are distinguished
from Protozoa. It is true that we found in the latter the beginnings
of sexual reproduction. As among the Protozoa a series of phenomena
lead up to the sexual reproduction of the Volvox colony, so the
latter is directly connected with the simplest form of the sexual
reproduction of the Metazoa.

In sexual reproduction we have to bear in mind two distinct points :

1. The fusing of the cells, or more accurately of two cell

nuclei, which here takes place — a phenomenon which is analogous

to the processes of copulation and conjugation in the Protozoa. The

24 COMPARATIVE ANATOMY CHAP.

origin and meaning of this phenomenon are not certainly ascertained.
Many see in them a strengthening of the product of the fusion, i.e. the
young new individual.

2. The different sizes, or the the sexual dimorphism of the
fusing reproductive cells. This is to be explained by the principle
of the division of labour. The reproductive cells have a double object
to fulfil : (1) such a cell must mingle with another (fertilisation) ; and
(2) must, after mingling, form a new organism like that of its parents.
To secure the first object free locomotion is useful, so that the repro-
ductive cells may seek each other and meet ; and further, in certain
circumstances a power of resistance to external influences is needed.
To fulfil the second object the cell must be of a considerable size, and
contain, if possible, nutritive material which can be used during
development. Both these objects cannot be fulfilled by each of the
reproductive cells without disadvantage. Here division of labour
steps in. Some cells fulfil the first object ; they move about with great
ease ; they are resistent, and moreover very small (the smallest cells
of the organism). Their smallness has a further advantage ; they
are produced in greater numbers, and can easily penetrate the second
sort of reproductive cell. These are called male reproductive cells,
sperm cells, sperm filaments, or spermatozoa.

Other cells fulfil the second object. They are large and often
filled with much reserve material (the largest cells of the organism).
They substitute size and mass for free locomotion. These are called
female reproductive cells, or egg cells.

Male and female reproductive cells are either formed in one and
the same metazoan individual (hermaphroditism), or in two different
individuals, male and female (gonoehorism, separation of the sexes).
The latter appears among the Metazoa generally as the rule, the
former as the exception. The causes which determined the separation
of the sexes are most probably quite similar to those which brought
about fertilisation in the animal kingdom. If one remembers that even
among hermaphrodite animals a copulation of two individuals often
takes place, or adaptations are present which prevent the fertilisation
of the eggs of an animal by the spermatozoa of the same individual
(self-fertilisation), the recent opinion that all Metazoa were originally
sexually separated, and that hermaphroditism has developed secondarily
from the male or female condition gains in probability.

The utility of cross-fertilisation places in a new light the utility of
a sexual differentiation of the reproductive cells into freely movable
spermatozoa and massive eggs. So as still further to secure cross-
fertilisation we find copulation in very many Metazoa. These animals
possess special copulatory organs, by means of which the spermatozoa
from the male body are carried into the sexual organ of the female,
and thus into the neighbourhood of the egg.

THE EGG 25

The Animal Egg.

The mature egg, capable of fertilisation, is everywhere in the
animal kingdom a simple cell, and shows the typical structure of such
a cell. It consists of protoplasm, called yolk, and of a nucleus which
is named the germinal vesicle in an unfertilised egg. The egg
is either naked or surrounded by one or more membranes and
envelopes. These are of very different nature according to their
origin. They are either secreted by the egg cell itself, and then
answer, as real primary yolk or egg integuments, to the membranes
of ordinary cells, or they are in various ways supplied by the
surrounding tissues of the mother body, and are laid round the egg
externally. In this case they may be considered as secondary or
accessory egg envelopes. The eggs arise in special organs of the
Metazoan body, called ovaria or germaria. These, in the simplest
cases, are masses of cells, some of which by stronger growth become
egg cells.

The processes of the formation, growth, and ripening of the egg in
the animal kingdom are as various as is the structure of the ovary
itself. It is especially the necessity for the abundant nourishment of
the eggs which determines the most manifold modifications. The eggs
are developed either from a mass of protoplasm with nuclei scattered in
it, or from an assemblage of similar little cells clearly defined one from
the other. In the first case the nuclei, in the second the cells, multiply
by fission. After this multiplication has lasted for some time the proto-
plasm round the nuclei in the first case separates off and gives rise
here also to independent cells. All these young cells are young egg
germs, and capable of growing and ripening into eggs ; but only in the
rarest cases do all the cells become eggs — a large number of germ cells
have almost always another fate.

The nourishment of the egg, speaking generally, is secured in the
following ways. In animals without a blood -vascular system and
body cavity the ovaries lie on the wall of the intestine, or of a
gastro- vascular system, which proceeds from the main intestine
(Coelent&rata, Platodes). In many animals the eggs are developed on
the wall of the body cavity and nourished by the body fluids,
into which they sooner or later pass and mature while suspended
in them (many higher worms). In the case of the greater number
of the higher animals the nourishment of the egg is secured by a
rich provision of blood-vessels in the ovaries. In those cases
where only some of the egg germs are developed into eggs,
the remainder often serve as nourishment for these or contribute
to their nutrition. Amoeboid moving egg cells can feed upon the
neighbouring egg germs after the manner of Amcebce ; or the surround-
ing egg germs store up food, which they give over to the growing
egg cell, either by emptying their contents into it (Cephalopoda)

26

COMPARATIVE ANATOMY

CHAP.

or by transfusion (follicle formations). In Insects the egg germs
can develop in the ovaries alternately into nutritive cells and egg
cells. In most Platodes the case is very complicated ; the germ
cells of an original germarium here fall into two more or less
distinct groups. The germ cells of one group (germarium) become
egg cells ; those of the other group (vitellarium) become nutritive or
yolk cells filled with nutritive yolk. In other Platodes only some
of the originally numerous germaria retain their primitive function,
while all the others are changed into vitellaria to supply nutritive
yolk. The eggs either absorb the nutritive yolk of the yolk cells

before fertilisation, or else many yolk cells
are stored in an egg capsule together with
a few fertilised egg cells, and are used up
during development.

The nucleus or the germinal vesicle
(vesieula germinativa) of the animal egg
is remarkable for its relatively great size.
It consists of an outer layer, in most cases
differentiated into a distinct membrane,
surrounding the light - coloured clear
nuclear fluid. In this lie one or more
solid nueleoli or germinal spots (maculae
germinativse), which are often connected
with each other and with the nuclear
membrane by a network of fine threads.
In many eggs the germinal vesicle lies,
throughout, in the centre of the egg • in others it does so at least in
the very early stages.

FIG. 22. — Ovarian egg of an
Echinoderm, after O. Hertwig. In

the middle the germinal vesicle with
the nuclear framework and the germ-
inal spot.

The Egg Yolk.

Investigations made by means of the improved optical appliances
for research have lately shown that protoplasm itself (in Protozoa,
egg cells, and tissue cells) exhibits a fine structure. It consists of
very small firmer particles, arranged in the finest network of threads,
which form the spongioplasm, and, lying between them, clear homo-
geneous more fluid portions, forming the hyaloplasm. Protoplasm
thus constituted only in the rarest cases forms the whole of the yolk.
In most cases reserve nourishment in the form of fat or oil drops,
small plates and spheres, is found in the protoplasm, these being used
as food by the developing egg. These constitute, in contradistinction
to the formative yolk — i.e. the actual living active protoplasm of the
egg — an inert lifeless constituent, only serving as nutriment, the
deutoplasm or nutritive yolk. The quantity and arrangement of the
deutoplasm in the egg is of great importance, since this determines
the course of its first segmentation.

It rarely happens that there is no deutoplasm in an egg. Less

i THE EGG 27

rarely we find only a very small quantity equally distributed through-
out the protoplasm (holoblastie alecithal eggs, Fig. 23, C). Such
eggs are found among animals which can at a very early stage of
development find their own food, or else among those whose embryos
are developed within, and nourished by, the mother body.

In most cases the egg contains a considerable quantity of nutritive
yolk. Two different types of eggs are distinguished, according to the
arrangement and position of this yolk.

I. Types of Teloleeithal Eggs.

A. In the simplest case the comparatively small quantity of nutri-
tive yolk is imbedded in the formative yolk, principally in one hemi-

D

'"">-_

"* x,

FIG. 23.— Structure of different eggs. A and B, Holoblastie telolecithal eggs. C, Holo-
blastie alecithal egg. D, Centrolecitlial egg (of a spider). E, Meroblastic telolecithal egg.
dp, Deutoplasm ; n, nucleus, or germinal vesicle.

sphere. This hemisphere is called the vegetative ; the other, which
contains the germinal vesicle, the animal hemisphere. Corresponding
animal and vegetative poles are also distinguished.

B. The quantity of the nutritive yolk increases so much that the
formative yolk is reduced to a smaller or larger segment at the animal
pole, in which lies the germinal vesicle. Besides this a thin layer of
protoplasm spreads all round the egg as a rind. In the remaining
portion of the egg the formative yolk is so much displaced by the
development of the nutritive yolk, that it remains merely as a cement-
ing substance in the interstices between the elements of the latter
(holoblastie teloleeithal eggs, Fig. 23, A).

28 COMPARATIVE ANATOMY CHAP.

C. The quantity of formative yolk is, in comparison with the
enormously developed nutritive yolk, so small that it is only a small
mass enclosing the germinal vesicle at the animal pole, and a very
thin layer round the whole egg. In by far the greater part of the
egg the formative yolk is quite supplanted by the nutritive yolk
(meroblastie teloleeithal eggs, Fig. 23, E).

II. Types of Centroleeithal Eggs.

The formative yolk is arranged in a regular layer round the whole
egg, and besides this in a mass containing the germinal vesicle at the
centre of the egg. The region between the centre and the circum-
ference is occupied by nutritive yolk, either

A. In largely preponderating quantity (holoblastie eentroleeithal
eggs of many Crustaceans), or

B. Almost, or quite exclusively (meroblastie eentroleeithal eggs of
the Tracheata and many Crustaceans, Fig. 23, D).

The Egg Envelopes.

These are divided into (1) primary envelopes — yolk membranes,
egg membranes ; and (2) secondary accessory envelopes.

I. The yolk membrane is secreted by the yolk itself. It can be
formed at different stages of the ripening of the egg, and shows great
diversity of structure. Occasionally it is double. It is often pene-
trated by numerous pores (zona radiata). Not infrequently there is
a special opening, the mieropyle. Both the pores and the micropyle
are connected with the nutrition of the egg, and serve in many cases
for the passage of the spermatozoa.

II. The secondary envelopes are also very various. They are
alike in one point — none of them are formed by the egg itself.

a. The chorion is a membrane which is very often secreted by
the cells of the ovary which surround the egg (follicle cells). It lies
between the egg and the follicle cells.

b. Other accessory envelopes are sometimes found, which are
only added to the egg later by means of special glands on its way
through the passages leading out of the ovaries. Such are the egg
capsules, albuminous and gelatinous envelopes, calcareous shells, etc.
Either only one of these egg envelopes is formed, or else two or more
are simultaneously produced.

As an example of the development and ripening of an egg, we choose first the
egg-formation of the mussel (Fig. 24, A, B, C]. The eggs here arise out of definite
cells of the germinal epithelium, which grow more strongly and soon project beyond
the epithelium, with which, however, they remain for a time connected by a long
stalk-like process. Through this stalk, in all probability, the nourishment of the
egg by the epithelium takes place. Yolk granules, continually increasing in number,
appear in the protoplasm of the egg. The nucleus becomes vesicular. The egg
secretes on its surface a yolk integument, which is broken through at the point

THE EGG

29

where the egg joins the germinal epithelium, so that when the egg severs itself an
opening, the micropyle, is left.

The marine planaria offer a further example of a simple egg- formation (Fig. 24, D}.
These possess very numerous ovaries, whose nourishment is provided for by their

kb ksch

FIG. 24.— A, -B, C, Three ovarian eggs of a mussel in different stages of development, after
Fleming, m, Micropyle ; dh, yolk membrane. D, Ovary of a marine planarian with eggs in
different stages of development ; da, branch of the intestine or gastro-canal ; U, germinal layer ;
e, advanced egg.

position between the branches of the intestine and close to its walls. In each ovary
there is a small germinal layer (kl) formed by a mass of small cells with nuclei and
little protoplasm. Some of these cells grow and
become eggs, numerous yolk granules forming in
them, and their nuclei changing into the charac-
teristic germinal vesicle. Other cells remain
small ; they take up a position between the
strongly growing egg cells, and so form a frame-
work in the ovary, which is continued into the
oviduct.

As a type of a perfect meroblastic telolecithal
egg with complicated envelope-formation we have
the bird's egg (Figs. 25 and 26). The egg is
fertilised within the mother body, and has
already begun to develop when it is laid. The
different parts which can then be recognised have
very different meanings and origins. In the
interior of the egg we recognise the yellow yolk
sphere, the well-known yellow of the egg (Fig. 25). This is formed in the ovary,
and represents the real egg. In the ovary it is merely a simple meroblastic egg cell,
consisting of the following parts :

FIG. 25.— Ovarian egg cell of a Fowl.
ksch, Formative yolk ; kb, germinal
vesicle ; wd, white yolk ; gd, yellow yolk ;
dh, yolk membrane. (After O. Hertwig.)

30

COMPARATIVE ANATOMY

CHAP.

1. An outer yolk membrane, secreted by the yolk itself.

2. The formative yolk, a small whitish mass in which lies the germinal vesicle
at one pole of the egg, viz. the animal pole.

3. The yellow nutritive yolk, which represents the principal mass of the egg,
and which appears in concentric layers.

4. The white nutritive yolk, which forms a thin outer layer round the yellow
yolk beneath the yolk integument and the formative yolk, and which also sinks into
the middle of the yellow yolk from the animal pole in the form of a thick strand with
a swollen rounded end.

When the egg thus formed passes out of the ovary into the oviduct, the walls of
the latter secrete around it the last envelopes, which are :

1. The albumen with the chalaza, i.e. somewhat denser spirally twisted

c7i.L

FIG. 26.— Diagrammatic longitudinal section of anewly-laid Hen's egg, after Allen Thomson
and O. Hertwig. b.l Formative yolk; w.y. white yolk; y.y. yellow yolk; w. albumen; ch.l.
chalaza ; a.ch. air-chamber ; i.s.m, inner ; s.w. outer layer of the shell membrane ; s. shell, v.t. yolk
membrane ; x. a somewhat fluid layer of albumen surrounding the yolk.

strands of albumen, which go from the yolk membrane out towards the two ends of
the egg.

2. The double shell integument, which surrounds the albumen on all sides.
The two membranes which form this integument separate, and between them a cavity
filled with air arises at the blunt end of the egg — the air chamber.

3. The porous calcareous shell. These three parts, therefore, represent the
accessory integuments of the egg.

Male Reproductive Cells, OP Spermatozoa (Fig. 27).

These belong to the smallest cells formed in the animal body. Each
spermatozoon is a simple cell, as is shown by its development which is
generally complicated, and is very difficult to observe. A very common
form of the spermatozoon is the so-called pin-shaped. Such a sperma-
tozoon consists of a small knot, the head, representing the remains of a

MATURATION OF THE EGG

31

cell nucleus, and a mobile filamentous appendage, the tail, which is of

protoplasmic nature, and is much like the flagellum of the Flagellata.

Besides the tail there may be accessory nagella. Between the head and

the tail a special intermediate portion is sometimes interposed. Other

forms of spermatozoa are occasionally

found ; round, pear-shaped, etc., either /^ X* "^-w^^'^Nn I

stationary or moving like Amcebcv. The '

spermatozoa arise in the testis from a

germinal layer or epithelium, as do

eggs. After repeated division of the

original formative elements, cells are

produced which are equivalent to egg

germs, and which may be distinguished

as sperm germs. Whereas, however,

the egg germs become eggs direct by

means of growth and maturation, the

sperm germs are still further divided FIG. 27.— various forms of sperma-

and produce spermatozoa. We have J0203" £ Of» Mwnmai ; a, of a Turbei-

. lanan, with two accessory flagella ; c. d,

already seen a phenomenon similar, and e, of Nevada; /, of a crustacean;

though not in all points parallel, in 9, of a Salamander (with undulating mem-

i

An Ordinary Cell Of the Colony brane);^ The commonest pin-shapedform.

there becomes by growth a large egg, or by division a mass of small
spermatozoa.

Maturation of the Egg.

The ejection of the directive or polar bodies is the 'last stage in
the maturation of the egg which precedes fertilisation. The germinal
vesicle moves towards the surface of the egg (towards the animal pole
in the case of eggs differentiated into poles), and here undergoes

kb
pn
nu

... kf
kb

FIG. 28.— Disorganisation of the germinal vesicle and formation of the nuclear spindle in
eggs of Asterias glacialis, after O. Hertwig. x, Prominence of protoplasm ; Itf, germinal spot
which divides into two distinct substances, pn and nu ; fc&, germinal vesicle ; sp, nuclear spindle.

considerable changes. It becomes partially disorganised (Fig. 28).
Out of part of its contents is formed that spindle-shaped figure
(Fig. 28, B) which is characteristic of indirect nuclear division (see
below, pp. 35,36). The one half of the spindle enters a small mass of
protoplasm which projects from the surface of the egg. This prominence

32

COMPARATIVE ANATOMY

CHAP.

tHen becomes completely constricted off from the egg (Fig. 29) as the
first polar body. In a similar manner a second polar body is formed.
The formation of a polar body thus appears like a process of gemma-
tion, or a sort of cell division, in which one cell — the daughter cell—
the polar body, is very much smaller than the other — the egg.

ir

VI

FIG. 29.— Formation of the polar bodies in Asterias glacialis, after O. Hertwig. sp, Nuclear
spindle ; rk\ first, rfc2, second polar body ; efc, female pronucleus.

The half of the nuclear spindle which remains in the egg after the
formation of the second polar body changes into an egg nucleus very
different from the original germinal vesicle, especially in its size, being
much smaller. This germ is known as the female pronucleus.

In many divisions — e.g. many insects, — there are species in which, alternating with
the sexual generations, which multiply by means of fertilised eggs, there occur other
generations, which reproduce by means of parthenogenetic eggs, i.e. by means of
such eggs as develop without fertilisation. In these eggs, according to some
observers, only one polar body is formed. Many hypotheses have been brought
forward as to the meaning jof all these various phenomena of maturation. I can
here only refer to the works of Biitschli, Balfour, Minot, Sabatier, van Beneden,
Weismann.

Processes similar to those of the expulsion of the polar bodies from
the egg have been observed in the formation and ripening of the
spermatozoa. The nucleus of the ripe spermatozoon is called the male
pronueleus.

When the polar bodies have been expelled the egg is capable of
fertilisation.

Fertilisation.

This process takes the following course ; out of numerous sperma-
tozoa pressing towards the egg there is only one, normally, which
fertilises it. This is the one which first touches the egg at a definite
point as it appears, viz. at the animal pole, in eggs differentiated into

FERTILISATION

33

poles, near the polar bodies. Here, at the touch of the head of the
spermatozoon, a prominence of the outer protoplasmic layer is formed —
the receptive prominence — into which the spermatozoon penetrates.
Gradually it presses further into the egg, its tail seeming to fuse with

C

FIG. 30.— Fertilisation of eggs of Asterias glacialis, after Fol (from O. Hertwig's Lehrbuch
der Entu-icUungsgeschichte). One of the spermatozoa which have entered the mucilaginous
envelope comes in contact with the receptive prominence. In C the yolk membrane is formed.

the protoplasm of the egg ; the head (the remains of the original
nucleus) increases somewhat in size. As male pronucleus, it moves
forward to meet the female pronucleus. Finally they fuse and form
one single nucleus, the so-called segmentation nucleus. The egg is
fertilised.

It seems tolerably certain that where the egg envelopes have
a micropyle, the spermatozoon enters through it. When the first
spermatozoon has penetrated the egg, the micropyle is closed by a
fresh secretion from the yolk, so that no more spermatozoa can enter.

FIGS. 31 and 32.— Fertilised eggs of a Sea-urchin, after O. Hertwig. Male (sk) and female (el)
pronuclei moving towards each other.

In other cases, at the moment when the first spermatozoon presses
in, a membrane begins to rise from the yolk, which makes the entrance
of other spermatozoa impossible. There are adaptations also with the
object of preventing the entrance into the egg of more than one
spermatozoon.

VOL. I D

34 COMPARATIVE ANATOMY CHAP.

Abnormally, two or more spermatozoa may enter an egg. In sue** a case several
male pronuclei may fuse with the female. Compare on this point the works of Fol and
Hertwig. It is not improbable that twin- and triple-formations may be produced
by such over - fertilisation. The development after over -fertilisation in all cases
deviates characteristically from the normal course.

The most essential morphological characteristic of fertilisation is
the fusing of two sexually differentiated cell nuclei, the male and the
female pronuclei. In the conjugation of the Protozoa (Paramcecium)
also we have to do (cf. p. 18) with a fusing of two nuclei (the stationary
and the migratory nucleus). Fertilisation in this latter case, however,
is mutual, and we cannot distinguish the conjugating cells as male I
and female.

Fertilisation is either internal, i.e. takes place within the mother body, or
external, i.e. spermatozoa and eggs are expelled from the parental bodies and meetj
each other outside in water. In the first case at least the reproducing animals]
possess special organs of copulation.

Various theories about the nature of fertilisation have been recently brought I
forward, especially by Biitschli, Balfour, Sabatier, van Beneden, Hertwig, Weismann, j
Geddes and Thomson, and others.

Literature.
Comprehensive Works.

Besides Balfour's Comparative Embryology, consult especially :
O. Hertwig. Lehrbuch der Entwickelungsgeschichte des Menschen und derm

Wirbelthiere. 3d edition. Jena, 1890.
W. Waldeyer. Eierstock und Ei. Leipzig, 1870.

The same. Ban und Entwickelung der Samenfdden. Anat. Anzeiger. Jena, 1887 J
where complete bibliography is given.

Tissue Cells and Cell Tissue.

We have till now considered (1) unicellular organisms, and (2) the]
egg- and sperm-cells, which fuse to form the starting point in thej
individual development of all the higher, i.e. multicellular, animals, j
We will now briefly deal with the manner in which the Metazoanj
body is composed of cells, and consider the various cells of which!
it consists. The observation of the cells of the animal body and of
their complexes, the tissues, is the object of the science of tissues ori
Histology. How different cells and complicated tissues arise out oq
simple indifferent cells, is the subject of Histogeny.

As our first principle we can state that all cells and tissues of the
adult animal body arise by means of repeated division from the]
fertilised egg -cell. These phenomena of division are the same as
those with which we became acquainted among the unicellular j
Protozoa as widely spread asexual reproductive processes. Whereas,!
however, in most Protozoa the products of division separate, and, like j
the mother cell, lead an independent life, in the Metazoa the descend-!

TISSUE CELLS AND CELL TISSUE

35

mmm

ants produced by repeated fission of the fertilised egg -cell remain
bound together in space. Similar cases were found among the
Protozoa ; we called them cases of colony formation. While, however,
there, all the cells of the colony remained alike and each maintained
itself quite like a Protozoan individual, the cell communities of the
Metazoa by dividing among the
individual cells the various duties
of life, so that some cells are ex-
clusively adapted for the per-
formance of one function, some
for the performance of another,
raise themselves into stable and
well-ordered states, the citizens
of which (the cells) are dependent
upon one another and can no ^p
longer exist alone.

The division of the egg-cell r^
and its descendants occurs under
peculiar inner conditions, which
chiefly concern the nucleus. Direct
nuclear division during cell divi-
sion is distinguished from indirect /
or karyokinetie nuclear division. /
The first and, as it appears, the I
rarer agrees in essentials with that w ;
already figured in the Amceba (p. ™
12, Fig. 19). The second shows
various modifications. The follow-
ing course may be taken as typical ("
(Fig. 33, A-H). {:

[Among the constituents of \
the cell nucleus are to be dis- /
tinguished the aehromatin, —

that part which does not stain at
all or only very slightly when
treated with colouring solutions,
viz. the nuclear fluid, and a part of division, with indirect division of the nucleus

,1 ... ; ,1 T.I (diagram!

WE*

Fig. 33.— A-H, Consecutive stages of cell-

,-, ... f . ,! n-i (diagrammatic).

the constituents of the fibrous

net-work ; — and ehromatin, which freely imbibes colouring matter, viz.

the nucleoli and other granules of the fibrous network.]

1. At the beginning of cell division, there appear near the
nucleus two opposite attraction centres, round which the portions of
protoplasm group themselves in a radiate manner (formation of the
amphiasters). The ehromatin of the nucleus arranges itself as a
tangle of fibres (Fig. 33, B).

2. The nuclear membrane becomes indistinct ; the tangled ehromatin
falls into several loops (Fig. 33, C).

36 COMPARATIVE ANATOMY CHAP.

3. These loops arrange themselves in an equatorial plane between
the two attraction centres, in such a way that their free ends are
directed outwards and their angles inwards. Fine achromatine fibres
run to the attraction centres (Fig. 33, D).

4. The chromatin loops split lengthwise, so that their number is
doubled (Fig. 33, E\

5. The one half of the chromatin loop which has thus arisen moves
towards one attraction centre, the other towards the opposite centre.
The halves thus move away from each other ; fibres of achromatin
stretch between them (Fig. 33, F).

6. The chromatin loops of each side have moved quite near their
attraction centres. Their order now becomes irregular, and they
again unite into a single tangle, round which again a nuclear
membrane can be recognised (resting stage, Fig. 33, G and H).

During the last stages, at the surface of the cell in a plane between
the two attraction centres, a circular furrow appears, which becomes
deeper and deeper, and finally, when the two new nuclei have reformed,
divides the cell into two halves, each with a new nucleus.

The cells which arise in the animal body by repeated division of
the egg develop in various ways, but always in such a way that the
greater number of them remain bound together in a special manner,
forming the so-called tissues. Four chief sorts of such tissues can be
distinguished :

1. Surface or epithelial tissue.

2. Connective tissue.

3. Nerve tissue.

4. Muscle tissue.

I. Epithelial Tissue.

This is the simplest form of tissue, and, as Comparative Histology
and Histogeny teach, the most primitive complex into which cells can]
combine. It may therefore be correctly described as primitive tissue,]
from which all other tissues are derived. Even among the Protozoa!
epithelium-like combinations of cells occur, as, e.g., in Pblvox, where the!
individuals (cells) of a colony are placed side by side in a layer which, !
like a spherical mantle, encloses a central cavity. Such a form is, in'
many Metazoa, the immediate result of the first division of the egg ; *
it is here called the Blastula. The epithelial character consists in the"
regular juxtaposition of cells into superficially extended membranes. \
These cover the outer and inner surfaces of the animal. The cells of |
an epithelium lie either in a single layer side by side (unilaminar
epithelium)^ or several layers lie one above another (multilaminar
epithelium). Different sorts of epithelium are always distinguished j
according to the differences in the individual cells which form them.
Thus we speak of tesselated epithelium when the cells are flat, of |
columnar epithelium when the cells are cylindrical, and so on. The!

EPITHELIAL TISSUE

37

epithelial cells generally secrete externally layers of varying thickness,
which often harden and possess different chemical and physical pro-
perties ; these form a cuticle /?
over the epithelium. ££ &

In the simplest cases
every epithelial cell pos- ;*•$

sesses nearly all the qualities ;-*£. 'S':

of a Protozoon. Division
of labour, however, soon
steps in, so that the epi-
thelial cells belonging to
different areas undertake
different functions. In a
simple condition, which we
meet with in certain low
Metazoa, e.g. the Hydra, and
which is also passed through
by many higher Metazoa in
the course of their develop-
ment, the wall of the pouch-
like body consists of two
contiguous epithelial layers,
an outer (eetodermal epi-
thelium) and an inner
(endodermal epithelium),
the latter lining the central
cavity of the pouch (arch-
enteron). The two layers
pass into each other at
the opening of the pouch
(mouth). This stage is
known as the Gastrula. A
certain not very sharply
defined division 'of labour
occurs between the cells of
the inner epithelium and
those of the outer epi-
thelium.

OUter epithelium appear ' Hum of intestine with amoeboid processes ; i, multilaminar
tO be Specially Suited, in ePithelium ; ^ body epithelium of a marine planarian,
, . , , with pigment cells, rod cells, and sub -epithelial glands.

correspondence with their

position, to carry on the relations between the animal body and the
outer world, and further to promote locomotion and movement ; those
of the inner epithelium are adapted more for the ingestion and
digestion of the food taken into the gastric cavity. In correspond-
ence with these functions, differences in the structure and in the
size of the cells of the two layers appear. The higher we rise in

FIG. 34.— Various forms of epithelium, a, Ciliated
epithelium ; &, columnar epithelium in profile ; d, surface
view ; c, tesselated epithelium ; e, the same from the sur-

rpi v f ,1 face; /, epithelium of collar cells, with flagella (from the

' endoderm of a sponge) ; g, flagellate epithelium ; h, epithe-

38 COMPARATIVE ANATOMY CHAP.

the animal kingdom the more sharply defined and the more thorough
is the division of labour. The epithelial areas no longer consist of
similar cells, but the cells of the same epithelium, having different
functions to perform, are differently constructed.

The comparison of the adaptations by which different cells
(especially epithelial cells) of the Metazoa seem to be suited for definite
functions, with the adaptations we have described among the Protozoa,
is very instructive. In the latter, in the most complicated cases,
different portions of one and the same cell appear specially adapted
for the performance of definite functions. Among the Metazoa, how-
ever, in a number of tissue cells one of the many adaptations of a
complicated unicellular Protozoon becomes the principal adaptation
suppressing all or most of the others, or totally obliterating them;
in this way cells suited for the performance of one special function
arise. We find among the Protozoa — as adaptations for movement,
for taking in food, and for respiration — cilia, flagella, and so on.
Epithelial cells of Metazoa very often assume such adaptations as the
most prominent characteristic of the cells. Either all cells of an
epithelium are covered with cilia — then we speak of ciliated
epithelium — or only groups of cells or single cells are thus covered.
The cilia of a cell may be replaced by a flagellum, e.g. in the
endoderm of the sponges. Then the epithelial cells so provided are
called flagellate cells, which often strongly remind one of the
Choanoflagellata.

These cilia and flagella serve among the Metazoa, as among the
Protozoa, for very various purposes.

1. They cause locomotion in Metazoa of small size which live in
water (cilia of the body epithelium of a few fihabdoccelidce, ciliated
plates of the Ctenophora, cilia of the wheel organ of some Rotifera\
and especially in the free swimming larval stages of many Metazoa
(general ciliation of many larvse, ciliated rings and bands of the larvae
of Platodes, Vermes, Molluscs, and Echinoderms).

2. They serve for whirling food within reach, as they surround
the oral aperture, e.g. Rotatoria.

3. They continually cause new digestible material to pass over the
digesting epithelium ; in many cases they at the same time cause in
the enteric cavity a constant renovation of water and of the food
suspended in it. Intestinal respiration. Ciliated epithelium of the
intestinal canal.

4. They constantly bring fresh oxygenated water into contact with
the epithelium. Respiration. Ciliated epithelium of the gills.

5. If they carry on these activities at certain parts of the
epithelium where sensory nerves come to the surface, they subserve
special functions of sensation. Ciliated grooves. Olfactory pits.

Very many Protozoa use pseudopodia and amoeboid processes for
locomotion and ingestion of food, and also perhaps for respiration. In
many Metazoa the taking of nutritive particles into the intestinal

i EPITHELIAL TISSUE 39

epithelial cells by means of the protrusion of pseudopodia-like or
amoeboid processes is an important function of these cells, which serve
exclusively for purposes of nutrition. In many lower Metazoa the
amoeboid character of one or of all the cells of the intestinal epithelium
is so marked that they sever themselves from the complex of epithelial
cells and float about independently in the intestinal cavity.

Formations similar to the contractile vaeuoles of many Protozoa
are found in the so-called excretory cells of the Turbellaria. In their
protoplasm, and in that of their processes, small drops (products of
metabolism) collect, and these may mingle to form one large drop
(vacuole). The drop is emptied into the lumen of the cell (see Fig.
109, p. 152), and thence expelled by means of the excretory ducts.

In certain Protozoa the ectoplasm gives rise to stinging cells.
In most Coelenterata the production of such stinging capsules, the
so-called nematoeysts, is always the chief function of very many
ectodermal, cells (cnidoblasts) which do no other work for the
organism.

In the Protozoa a membrane or shell for the protection of the
unicellular organism is provided by the secretion of a resistent external
envelope ; so also in the Metazoa the epithelial cells, and naturally
the cells of the outer body epithelium especially, provide a protection
or covering for the body by the formation of outer eutieular
membranes. Such eutieular formations arise by the mingling of the
secretions of different cells, or the transformed products of the proto-
plasm, to form a layer. They may vary very much chemically and
physically, and stand in just the same relation to the epithelial cells
which produce them as do the glandular secretions to the gland cells.
To these eutieular formations belong the chitinous integuments of the
Annulata and Arthropoda, which sometimes by calcareous deposit become
carapaces as hard as stone ; these integuments are secreted by the
epithelium (hypodermis) which underlies them.

The Cuticle is often penetrated by fine perpendicular pores, which
probably owe their origin to the fact that, as the cuticle increases in
thickness, the cell protoplasm remains connected with the eutieular
surface by fine processes. A fine cuticle is also to be found in
ciliated epithelium, in which case the cilia pass through the pores to
the exterior.

There are, however, among the Metazoa a series of cells with
special adaptations and functions which are wanting in the Protozoa.
The aggregation of a large number of cells which can adapt themselves
to the most various activities, presents much more favourable conditions
for a far-reaching division of labour than are offered in the case of
unicellular organisms. In the first place, we have in the epithelium
the most various glandular cells — cells which are distinguished by
peculiarities of form and structure, and possess a protoplasm capable
of transforming the nutritive substances provided by the body into
different sorts of secretions, or of assimilating from the body super-

40

COMPARATIVE ANATOMY

CHAP.

fluous or waste material, which they then remove out of the body.
Dermal glands secrete mucus or other substances.

When glandular cells remain single they form unicellular glands ;
if many combine for the same purpose they are called multieellular
glands.

In the simplest cases, these glands are epithelial cells, chiefly
distinguished by their size, and often pear'- shaped. The nucleus

lies at the basal end of the cell, i.e.
the end away from the free surface
of the epithelium. The secretion
collects in the cell and is pushed
forward towards its surface. Where
a cuticle is developed it is often
broken through above the gland cell,
forming a pore for the discharge of
the secretion (Fig. 35, A).%

Unicellular glands often partially
leave the epithelium, the greater part
of the gland projecting into the
underlying tissue. They then only
penetrate between the epithelial cells
by means of thin processes, the
efferent ducts (Fig. 35, B).

A greater or smaller area of
epithelium often consists entirely or
principally of glandular cells (Fig.
35, (7). Such areas generally sink into
the epithelium, so that a glandular
epithelial sae arises, into whose cavity the secretion is emptied (Fig.
35, D, E). Such sacs again may form many branches (Fig. 35, G).
The glandular portion is then frequently limited to the blind and
often lobate ends of the sacs, while the rest serves as an efferent duct.
Glandular cells again can sink under the epithelium of these glandular
sacs, and only remain connected with it by their efferent ducts (Fig.
35, F).

It is only natural that certain cells of the epithelium which
envelopes the body should undertake the function of carrying on the
relations between the body and the outer world, i.e. of receiving im-
pressions and imparting them to the body. Such epithelial cells are
called sensory cells. They can either occur singly or form, in com-
bination, a sensory epithelium. Originally these cells are very slightly
differentiated, very slightly qualified for the reception of varied
impressions. But here also, as the organism grows more perfect,
division of labour steps in. Certain sensory cells appear qualified for
the perception of definite sorts of impressions only. They enter into
combination with certain tissue elements and form specific sensory
organs, which give rise to perceptions of either touch or sound, sight,

FIG. 35. — A-G, Various forms of glands.

r CONNECTIVE TISSUE 41

smell, or taste. The sensory cells will engage our attention again
later.

Epithelial cells may become pigment cells by the deposition in
their protoplasm of pigment material ; these pigment cells are an
important element in the external colouring of the animal. Deposited
in certain sensory cells, pigment serves for the absorption of light and
heat rays, and so contributes to the sensations of light, colour, and
warmth.

Certain epithelial cells often perform the function of supporting
their companions. They form, by mingling with each other, a tissue
with meshes, the interstitial tissue, which is related to the other
epithelial cells as the mortar in a wall is related to the bricks.

Since the protoplasm of neighbouring epithelial tissue mingles,
and since in epithelial cells it is the nuclei only and not their proto-
plasm that divide, masses of protoplasm with nuclei scattered through-
out them may arise. In such masses no cell boundaries can be recog-
nised. They are called syneytia.

The epithelial cells do not always lie close to each other; they
are sometimes separated by clear intercellular spaces for transmission
of fluid, or by intercellular substances ; but in such cases they remain
connected by means of very fine protoplasmic processes which stretch
across these intercellular spaces.

II. Connective Tissue.

Under this name is comprised a long series of tissue forms, which
may have very different ^

origins, structures, and func- n.--.| V ••"•'"[" iVi'Ai \ i • •.
tions. They originate either iM
directly or indirectly from
the epithelium. Their essen-
tial office is to bind together
different portions of the
body and different organs,
or to serve as support for
these by their possession of
a certain degree of firmness.
We divide the connective
tissues, according to an im-
portant difference in their :J||
origin, into two principal PPI |

groups. ef \ g

I. GelatinOUS tiSSUe (Fig. FIG. S6.— Gelatinous tissue of a Scyphomedusa (diagram-
36) takes its Origin direct matic).e, Epithelium ; jjr, Jelly; fy, epithelial cell passing in-
from an epithelium. To *» ^jdly; 60, branched cells in the jelly; e/, elastic fibres.

this belong, e.g., the gelatinous tissue of Medusae and the Ctenophwa.
Between the inner epithelium which lines the intestine and the

42 COMPARATIVE ANATOMY CHAP.

outer which covers the surface of the body a homogeneous jelly, con-
taining a large proportion of water, is secreted by the epithelium.
Cells leave the epithelium and enter the jelly, where they assume a
different form. Sometimes they become spindle shaped, sometimes
much branched cells connected together by their branches, sometimes
elastic fibres. Occasionally such cells show amoeboid movement.
They can even become contractile muscle-cells.

II. The Connective tissue proper does not take its rise direct
from the epithelium. In early stages of development of the animal
variously sized groups of cells sink out of the epithelium below the
surface, multiply by fission, and so produce the actual formative cells
of the connective tissue. This fills the interstices between the organs
and other tissues, or forms pillars, strands, plates, and variously
shaped supporting masses. It often forms membranes round other
organs and tissues, or lines cavities. Such a superficial extension may
even assume the character of an epithelium.

Two chief types of connective tissue proper are to be distinguished.

A. The cells of the connective tissue lie close together and form
no intercellular or connecting substance.

Vesicular connective tissue (Fig. 37). Vacuoles filled with fluid
occur in the cells, which, growing in size, cause vesicular swelling.

PIG. 37.— A, Younger, B, older vesicular connective tissue of a Platode. v, Vacuoles.

The protoplasm is then often limited to a thin layer surrounding the
vacuole, and this can mingle with the neighbouring cells. Small
aggregations of protoplasm may still be found massed round the nuclei.

Vesicular connective tissue passes into retieular connective
tissue, when the fluid -filled spaces of neighbouring cells unite with
one another. The connective tissue then takes the character of a
spongy network containing imbedded nuclei, and the intracellular fluid
which has flowed together becomes in a certain sense intercellular
(retieular connective tissue of many Platodes). In fatty tissue (Fig.
38) smaller or larger fat drops appear in the protoplasm of the cells.
In pigment tissue colouring matter is deposited.

B. The formative cells of connective tissue form externally a
substance, the intercellular substance, in which they come to lie em-

CONNECTIVE TISSUE

43

bedded. This substance is either secreted by the protoplasm of the cells,

or it is produced by the metamorphosis of the outer protoplasmic layers.

However much developed the intercellular substance is, according

to recent research it is probably nearly always penetrated by very fine

FIG. 38.— Fatty tissue, after Ranvier (from Claus's
Text-book of Zoology). F, Fat-cells ; B, connective tissue
fibrils.

•%

FIG. 39.— Chorda tissue, after Leydig.

processes of the connective tissue cells, which are thus maintained in
organic union. Various hollow spaces are also frequently found in it.

Cellular vesicular connective tissue arises from the vesicular con-
nective tissue above described by the cells secreting an external mem-
brane or cuticle which connects
them together, e.g. Chorda
dorsalis (Fig. 39).

There are kinds of cellular
connective tissue in which the
cells do not take the vesicular
form, but remain compact and
mostly round. The inter-
cellular substance is incon-
siderable in comparison with
the cells. If it increases in
mass, the tissue passes into the
next form.

In fibrous connective tis-
sue the intercellular substance
is considerable. When boiled
it yields glue. It is differen-
tiated into fibres, which often
Unite in bundles, showing FlG- 40.-Reticular connective tissue, after

. , Gegenbauer.

the most varied arrangements.

They often run parallel, often cross each other, or branch and anasto-
mose with each other. The cells of the connective tissue seldom

44

COMPARATIVE ANATOMY

CHAP.

remain round ; they generally assume a long spindle shape or a branched
form. In the last case there again arises a sort of retieular connect-
ive tissue. The processes of the cells themselves are often differenti-
ated into fibres, which associate themselves with those formed from
the intercellular substance. Fibrous connective tissue may have a
loose texture or a firm texture, as in tendinous tissue (Fig. 41). The
fibres swell on treatment with acids and alkalis. If they are elastic
and retieular, and if they do not swell under the influence of acids
and alkalis, we have elastic connective tissue.

Ill

FIG. 41.— Tendinous tissue, from the
longitudinal section of a sinew, after
Gegenbauer.

FIG. 42.— Hyaline cartilage. To the right above
are depicted the protoplasmic threads which connect
the cartilage cells.

Cells of connective tissue may become pigment cells by the deposi-
tion of colouring material.

Cartilaginous tissue is a very good instance of the structure and
rise of real connective tissue. The cells generally remain round.
They secrete membranes, which continually become thicker and
stronger by their own growth, and which finally mingle with those of
the neighbouring cells and form a tolerably firm intercellular substance,
the cartilaginous substance, which in boiling yields chondrine. The
cartilage cells continue to divide ; the daughter cells again surround
themselves with membranes. Sometimes the membranes of various
generations can still be distinguished — the less easily, of course, the
older they are (Fig. 42).

If the cartilaginous substance is homogeneous and structureless we
have hyaline cartilage ; if it is fibrous we speak of fibrous cartilage.
In cartilaginous tissue also the intercellular substance seems penetrated
by exceedingly fine processes of the cartilage cells, which apparently

CONNECTIVE TISSUE

were already present when the first membrane was secreted. Calcified
cartilage is formed by the appearance of calcareous deposits in the
intercellular substance.

Cartilaginous tissue, on account of its firmness, serves as support-
ing tissue in vertebrate and in some
invertebrate animals.

Bone tissue forms, par excellence,
the supporting tissue of vertebrates.
The intercellular or bone substance
becomes as hard as stone by a com-
bination of lime-salts with some ground
substance, which yields glue on being
boiled, and does not dissolve under
treatment with acids. In it are scattered
the cell elements (bone cells); they are
much branched, and connected by their processes ; they are arranged in
parallel layers, often concentrically round the cavity (Fig. 44). Bone
tissue arises out of indifferent connective tissue cells, which are arranged
in strands or flat expanses, and which function as formative cells
of the bone tissue, osteoblasts (Fig. 45, a). They produce on one side
bone substance, often in the direction of cartilaginous masses, which
they supplant, a,t the same time forming processes which remain im-

43. -Fibrous cartilage, after ciaus.

Fio. 44.— Bone cells, after Gegenbauer.

FIG. 45.— Bone tissue, a, Osteoblasts
b, bone cells (after Gegenbauer).

bedded in the bone substance. New masses of bone substance being
constantly formed from the osteoblasts, some of the latter come to lie
in the bone substance, and become bone cells.

Dentine is nearly related to bone tissue. Here the formative cells
(odontoblasts) do not enter into the dentine which they have: secreted ;
they all remain at its base, but send into it numerous finely branched
processes (fibres), which run parallel to each other in as many little

46 COMPARATIVE ANATOMY CHAP.

channels of the bone substance. The fibres are connected by fine
anastomoses.

It is clear from the above that bone tissue shows much similarity
in its origin with gelatinous tissue, and dentine with epithelial cuticular
formations.

The blood cells and lymph cells which float in the blood, the
lymph or the coelomic fluid of animals, at first rise out of connective
tissue cells. The blood has even been described as fluid connective
tissue, the blood fluid representing the intercellular substance, the
blood corpuscles the connective tissue cells. Lymph corpuscles often
show amoeboid movement, and are capable of taking in solid materials
(e.g. products of excretion, food material, products of suppuration, and
foreign bodies).

III. Neuromuseular and Muscle Tissue.

The elements of both muscle and nerve tissue originally come from

the epithelial cells. Both tissues appear simultaneously in the animal

kingdom, and are connected in their origin. In its simplest form

neuro-museulap tissue is met with among the lower Codenterata (e.g.

>3MH«iH93Sgi?ss&aaH£^'' Hydra}. Here, in the outer epithelium,

, /\ are found cells which form processes

• ' ./ inward, and produce a layer of fibres

, •' > -{ - ] close under the epithelium. These

fibres are contractile, and represent
muscle processes of the epithelial cells.
FIG. 46. -Neuro-muscular ceils of the Tne latter, which contain the nucleus,
Hydra, after Kieinenberg. q, Muscle pro- take part with their companions in
cesses of the same. limiting the. surf ace of the body. These

cells undertake the relations with the outer world, being suited to
receive external impressions. The stimulus created by such impres-
sions is carried through the protoplasm of the cell to the muscle pro-
cesses, which contract in consequence. The cells, with their processes,
are accordingly called neuro- muscular cells. In them the chief
functions are of nerve and muscle tissue localised, in the most general
undifferentiated manner, in different parts of one and the same cell.
In consequence of the intimate connection of the neuro-muscular cells
with one another, a locally created stimulus of one or more cells is
communicated to the neighbouring cells, and thus to their muscle
processes. The principle of the division of labour here again supplies
the key for understanding the further differentiation of neuro-muscular
tissue. One portion of the neuro-muscular cells undertakes principally
the function of contractility, and the greater part of its protoplasm
becomes differentiated into contractile substance ; another portion of
the neuro-muscular cells, while remaining in close organic connection
with the first, performs the function of receiving external impressions,

MUSCLE TISSUE

changing them into sensations and carrying on the stimulus to the
muscular elements.

In accordance with this, muscle tissue in its simplest form appears
as a system of epithelial muscle cells (Fig. 47, a). They still lie in the
epithelium, but no longer take part in limiting the body surface.
They form inwards muscle processes which run under the epithelium.
They are distinguished from neuro-muscular cells by the fact that their
chief function is contractility, and that the body of the cell no longer
acts as a neural portion. This cell body then appears merely a remainder
of the original formative substance of the fibres ; it lies below the
upper epithelial cells, wedged in between them. In this way a neuro-

FIG. 47.— Muscle fibres, a, Epithelial muscle cell (ez), with fibre ; 6, sub-epithelial muscle fibre,
with attached protoplasmic body, both of Cnidaria ; c, longitudinal section of muscle fibre of a
Nematode ; q, transverse section of the same ; e, of an Hirudo ; e\, the same in transverse section ;
d, dorso-ventral muscle fibre of a marine Planarian ; /, the same of an Hirudo ; g, branched muscle
fibre from the jelly of a Ctenophore ; p, protoplasm, or, in e and e1? medulla ; cs, contractile
substance ; n, nucleus.

muscular cell becomes an epithelial muscle fibre, with a protoplasmic
body containing a nucleus attached to it. From this to sub-epithelial
muscle fibres (Fig. 47, b) is a short step. These no longer lie in, but
under the epithelium. They consist of a contractile fibre, which
carries along its whole length, on the side bordering on the epithelium,
a thin layer of protoplasm.

In other cases, e.g. in the Ctenophora, epithelial cells can leave their
complex and enter into the jelly secreted below the epithelium. Here
they often grow into fibres branched at each end (Fig. 47, g), whose
protoplasm changes into a contractile substance. Such fibres are
known as mesenehymatous muscle cells, as opposed to epithelial
muscle cells.

48 COMPARATIVE ANATOMY CHAP.

The above differentiations of muscle tissue are seen within the
division of the Cnidaria. Various forms of muscle elements can here
be found in one and the same animal.

Even if it is not impossible that muscle fibres respond to certain
direct stimuli by contraction, it is generally characteristic of them
that they respond only to stimuli communicated to them by the nerve
elements.

A general division into two principal types of muscle fibres may
be made : (1) smooth, and (2) transversely striated.

Smooth muscle fibres are almost always simple muscle cells.
One single cell forms a fibre. The contractile substance is either
formed by the cell on one side, so that the cell with the nucleus appear
as an attached body (Fig. 47, c, Cj), or the contractile substance becomes
differentiated on the whole surface of the formative cell, and then
tubular muscle fibres arise (Fig. 47, e, e^. In these we distinguish the
outer contractile eortiele layer from the central medulla which con-
tains the nucleus ; this medulla fills the axis of the muscle fibre, and
represents the more or less unchanged remnant of the protoplasm of
the formative cell. The contractile substance of the smooth muscle
fibres often appears longitudinally striated, and under the influence of
suitable reagents separates into the fine long fibrillaB which are the
cause of this longitudinal striation. The smooth muscle fibres are
often branched at one or both ends, especially in those cases in which
they have an isolated course (Fig. 47, d, /). The above described muscle
fibres of the Ccdenterata are of this sort.

Striated muscles. — These are considered as physiologically the
most efficient muscle elements. Even among the Coelenterata, the con-
tractile fibres of the epithelial muscular cells appear transversely
striated. The muscle processes of several cells unite to form a group
of striated fibres. Striated fibres form the chief mass of the muscula-
ture in the Arthropoda and Vertebrate. They arise out of muscle-
forming cells, the greater portion of one such cell becoming
differentiated into transversely striated fibre. The remainder of the
protoplasm, with the nucleus, often persists on the surface of the fibre.
Perhaps it is this protoplasmic layer which produces the Sareolemma,
the membrane which envelops the muscle fibre. In nearly all cases,
however, the number of nuclei increases as the muscle fibre grows and
differentiates further ; the fibre is thus to be considered as a many-
celled structure proceeding from one cell by incomplete division.
Often, however, several cells lying one behind the other share in
the formation of a muscle fibre. Striated muscle fibre appears not
only transversely but longitudinally striated. The transverse striation
comes from a regular alternation of singly refractive with doubly
refractive elements.

There are, at the present time, many different opinions about the finer structure
of the striated muscle fibre. According to the last view of Van Gehuchten, sup-
ported by investigations of the transversely striated muscle fibres in the Arthropoda,

MUSCLE TISSUE

49

the muscle fibre, like the protoplasm of the cells in general, consists of a network
of close and very fine fibres, the spongioplasm, and of an intermediate homogeneous
fluid substance, the hyaloplasm. The former is contractile and elastic, the latter is
purely passive. In muscle fibres the spongioplasm is regularly arranged (Fig. 49, A).
It consists of parallej. filaments, which run longitudinally and are bound by transverse
filaments at regular intervals in a plane at right angles to the longitudinal axis.
A transverse section lying in such a plane would form a plate with reticularly
arranged filaments and hyaloplasm between them (Fig. 49, B}. There are no

FIG. 48.— Transversely striated
muscle fibre, after Gegenoauer.

FIG. 49.— Transversely striated muscle fibre of
an Arthropod, after Van Gehuchten. A, Lateral
view. B, In transverse section.

transverse connections between the longitudinal filaments, except at these regular
intervals, so that on a transverse section of the fibre made at any other part of it
only the transverse section of the longitudinal filaments would be seen. It is clear
how, by such an arrangement of the finer portions, the transversely and longitudin-
ally striated appearance of the fibre could be produced. There are weighty objec-
tions, however, to these views of Van Gehuchten.

The muscle fibres contract in response to stimuli transmitted to
them by the motor nerve fibres. They therefore always stand in con-
nection with the ends of such fibres in a manner which cannot here
be further described.

Muscle fibres joined together by connective tissue unite to form
bundles, bands, or tubes. These again can be united in layers or in
thick muscle strands.

Muscle fibres, smooth as well as striated, arise, even in many higher
Metazoa, out of epithelium ; in many of these, however, the muscle-
forming cells (Myoblasts) are descendants of cells which at an early
stage of their development sank below the level of the epithelium to
which they belonged. Both kinds of formation may occur in the same
animal.

VOL.

IB*4^v
Of THE Tr\

UNIVERSITY)

.llu £

50 COMPARATIVE ANATOMY CHAP.

IV. Nerve Tissue.

As muscle elements pure and simple may be imagined to arise out
of neuro-muscular cells by the gradual differentiation of most of the
protoplasm of one part of the cell into contractile substance, the
muscular function thus being brought to the front at the expense of
the other possible functions, so nerve elements may be produced by
the suppression of the contractile part of the cell and the further
differentiation of the neural portion. We can perhaps imagine that
the simple sensory cells of the body epithelium of the lower Metazoa
arose in this way, always presupposing that they remained in connection

.si

FIG. 50.— Piece of a muscle lamella of the septum of an Actinian (Anthea cereus), with nerve
plexus, after O. and R. Hertwig. m, Muscle fibres ; sz, sensory cell, with sensory hair ; gz, ganglion
cell.

with the neighbouring contractile elements, either by simple contact
or by processes. In this way a stimulus received by the sensory cells
could be transmitted to the muscle cells. The sensory cells are
epithelial cells, which are generally distinguished by a delicate, usually
immobile, sensory hair, projecting outwards. Division of labour could
then go further. Single sensory or nerve cells of the most undiffer-
entiated sort, found in contact with other sensory cells and muscle cells,
could give up their connection with the surface of the body and sink
into the lower part of the epithelium, thus playing the part of inter-
mediaries between sensory and muscle cells (Fig. 50), and transmit the
stimulus received by the former to the latter. Such cells are found in
many Coelenterata. They are here already known as gang-lion cells.
They possess processes by which they are connected with each other
and with the sensory cells and muscle fibres. They represent the|

NERVE TISSUE

51

central elements of the nervous system; in them the impressions
which have come from the sensory cells become sensations which can
be transformed into will impulses ; from them proceed the stimuli
which cause the muscle fibres to contract.

The nervous system becomes complicated as low down as among
the Ccelenterata, but more especially in animals of a higher grade.
Between the central parts, i.e. the ganglion cells on the one side, and

FIG. 52.— Ganglion cell
from a human spinal
FIG. 51.— Ganglion cell from the anterior cornu of the ganglion (after Gegen-

human spinal cord (after Gegenbauer). p. Pigment ; n. bauer). n, Nuclei of the

nerve. neurilemma.

the sensory and muscle cells on the other, nerve cells are interposed,
which, being stretched like fibres, become nerve fibres. These under-
take exclusively the function of the transmission of sense impressions
from the sensory cells to the ganglion cells, and the transmission of
stimuli from the ganglion cells to the muscle fibres.

Besides this, the glanglion cells no longer appear scattered,
arranged in a plexus, but they unite into masses, which are defined
as the central organs of the nervous system, e.g. the brain. The
nerve fibres also unite into nerves. There are two sorts of nerve

52 COMPARATIVE ANATOMY CHAP.

fibres : (1) Sensory nerve fibres, which transmit sensory impressions
from the peripheral sensory cells to the central organ; (2) Motor
nerve fibres, which transmit stimuli from the central organ to the
muscles. The majority of sensory cells also do not remain in their
undifferentiated condition. Division of labour steps in here also.
Some cells seem specially suited for the reception of light and colour
sensations, others for those of sound, others again for sensations of
smell and taste. Tactile cells still remain in the most undifferentiated
condition. Many sensory cells which are qualified to receive one and
the same class of stimuli become, by the addition of accessory tissues,
combined into complicated sensory apparati — the specific sensory
organs : the organs of sight, hearing, smell, taste, and touch.

The ganglion cells possess one or more processes (unipolar, bi-polar,
multi-polar ganglion cells), one or more of which pass over into the
nerve fibres, while others only serve perhaps for the nourishment of
the cells. The process of a unipolar ganglion cell sooner or later
divides into at least two branches, one motor and the other sensory.
The ganglion cells in peripheral ganglion centres are often enclosed
in envelopes of connective tissue.

Nerve fibres may have many branches, and are often finely
striated in a longitudinal direction. They are either naked — in some
of the lowest Metazoa — or enveloped in a sheath, the neurilemma,
which is supplied by the surrounding connective tissue. When many
nerve fibres form one nerve, the single fibres of this nerve are, 'among
all the higher animals, kept apart from one another by this sheath ;
the whole then in transverse section produces the effect of a spongy
tissue, in whose larger or smaller meshes the transverse sections of
the nerve fibres lie. I (I The neurilemma is not generally continued on
•S to the ganglion cells.

A further distinction between two sorts of nerves has been made,
especially among the Vertebrata : (1) nerves without medulla, which
remain simple ; and (2) nerves containing medulla, in whose fibres
two parts are found — an outer oleagenous tubular medullary sheath, and
a fibre surrounded by this — the axis-cylinder. On entering a ganglion
cell, the latter alone penetrates its process — it alone represents the path
of transmission. Both these sorts of nerve fibre are enclosed in neuri-
lemma sheaths.

In many of the lower animals the nervous system remains for the
most part in its original place of formation, i.e. in the body epithelium.
In the higher animals the nervous system remains in connection with
the body epithelium through the sensory apparatus. This helps us to
understand why, in the embryonic development of the highest animals,
the nervous system is always produced by the outer epithelial layer.

Literature.

Th. Schwann. MiTcroskopische Untersuchungen iiber die Uebereinstimmung in der
Structur und dem Wachsthum der Thiere und Pfianzen. Berlin, 1839.

i LITERATURE 53

A. Kolliker. MikrosJcopische Anatomie oder Geivebelehre des Menschen. 1850-54.

3 Parts.

The same. Handbuch der Gewebelehre des Menschen. 5th edition. Leipzig, 1867.
F. Leydig. Lehrbuch der Histologie des Menschen und der Thiere. Frankfurt, 1857.

(A standard work. )

The same. Vom Ban des thierischen Korpers. Tubingen, 1864. Vol. I. with atlas.
C. Frommann. Zur Lehre von der Structur der Zellen. Jenaische Zeitschrift fur

Naturwiss. Bd. IX. 1875.

W. Flemming. Zellsubstanz, Kern, Zelltheilung. Leipzig, 1882.
H. Frey. Handbuch der Histologie' und Histochemie des Menschen. 3d edition.

Leipzig, 1870.
L. Ranvier. Traite" technique d' Histologie. A German translation of the same by

Nicati and v. Wyss. Leipzig, 1877.
H. Fol. Lehrbuch der vergleichenden mikrosTcopischen Anatomie. Leipzig, 1884. Till

now only the first part has appeared, treating of technical microscopy.
Ph. Stohr. Lehrbuch der Histologie. 2d edition. Jena, 1888.

CHAPTEE II

Introduction to the study of the Metazoa — Ccelenterata or Zoophyta as the
lowest Metazoa constructed essentially of two cell layers — Yolk segmenta-
tion and formation of the two primitive germinal layers of the Metazoa
(Gastrsea theory).

INTRODUCTION.

IN contradistinction to Protista dr Protozoa we have real Animals or
Metazoa. The bodies of the former consist of one single cell or of
several similar cells (with the exception of F'olwx), each of which, how-
ever, is competent to perform all vital functions (cell colony) ; the
bodies of Metazoa, on the contrary, always consist of a number of cells
which are not all similar, but have divided among them the different
forms of vital activity (cell community). The division of labour may
be more or less complete, and according to it the degree of. morpho-
logical complexity and of physiological perfection is determined.
There are animals which are morphologically (according to structure)
and physiologically (according to their vital activities) only a little
raised above the Protozoan colony, e.g. the Hydra.

The bodies of these animals consist of only slightly different
sorts of cells : digesting cells, neuro- muscular cells, stinging cells,
and formative cells of eggs and spermatozoa. All these kinds of cells
are, however, indispensable to the existence of the Hydra body ; not
one of them can be removed from the body without endangering
its existence. The whole body is nevertheless physiologically an
individual, but, as opposed to the cell, an individual of a second,
i.e. a higher order — a person. Most animals remain at this stage of
individuality. A Medusa, a Worm, a Crustacean, or a Mammal is
such an individual of the second order. In many animal divisions,
however, individuals of the second order multiply by fission or gemma-
tion. The new individuals thus arising remain united, and together form
individuals of the third order — an animal stock. The single indi-
viduals which collectively form such a stock may remain similar, and
they then are related to the stock in just the same way as the cell-
individuals of a Protozoan colony are related to the colony ; or division

CHAP, ii INTRODUCTION TO STUDY OF METAZOA 55

of labour again steps in, resulting in variety of development in body
form and structure of the persons forming the stock (polymorphism).
Then such a stock is also physiologically again an individual of the
third order. The single persons become equivalent to instruments of
this complex individual, and bear the same relation to it as the various
cell elements of a single individual, e.g. a Hydra, bear to it. As instances
of animal stocks without division of labour among the persons we have
most Corals ; and of stocks with far-reaching division of labour and
polymorphism the Siphonophora.

Even in the lowest Metazoa the cell elements are not found
scattered in the body without any special arrangement. On the con-
trary, we find even among the simplest Ccdenterata that they are
arranged in two epithelium-like layers, which are closely contiguous and
form the wall of the body, which is pouch-shaped and provided with an
opening. In keeping with the physiological activities of the various
cells, the stinging cells and the neuro-muscular cells form the outer
layer, while the digesting cells form the inner layer, which is turned
towards the pouch cavity, i.e. the gastric cavity. The reproductive
cells lie protected in the deeper portions of the outer layer. These
two layers, which occur in the development of all Metazoa, are called
the Ectoderm and the Endoderm.

Either similar or dissimilar cells or tissues, therefore, may combine
to form cell or tissue complexes. Such complexes are called organs
when the cells or tissues combining to form them perform in common
one or more functions. The endoderm of the Hydra is a primitive
organ, all the cells of which undertake the digestion. The tentacles
of the Hydra are slightly more complicated organs; they serve as
organs of touch, as weapons, and as organs for seizing food. For these
purposes they contain neuro-muscular and stinging cells. For the
nutrition of the tentacles canals lined with endoc[erm enter them from
the gastric cavity. In this way the most various elements are drawn
into the service of one or of several functions after which the organs
are named. We therefore speak of the sensory organs and the organs
of movement, respiration, etc.

Several organs of the same sort with similar functions may occur
in the same body (this is especially the case among the higher animals);
these are then portions of an organic system — muscular system,
vascular system, nervous system, etc.

The observation of the construction of the animal body out of cells,
tissues, organs, etc., is the object of Anatomy, microscopic and mac-
roscopic. These become Comparative Anatomy when the structure
of animal bodies is considered comparatively. Comparative anatomy,
again, Js the indispensable aid of zoology in one of its chief tasks — the
discovery of the natural relations of affinity among animals and of the
conjectural course of development of the animal world in the earth's
history. Comparative anatomy seeks to define the relations of affinity
between the different portions which combine to form the animal body.

56 COMPARATIVE ANATOMY CHAP.

It deals naturally, not only with the bodies of animals now living, but
also with the bodies of animals of past geological periods in so far as
these are attainable in a fossilised condition ; not only with the bodies
of adult animals, but with all the consecutive stages of development of
such animals. For an animal form is characterised not only by its
structure in a fully grown and sexually mature stage, but by its struc-
ture in all the previous consecutive stages of its development. Com-
parative anatomy only considers organs, as we have already said,
according to their structure and their connections of affinity, not ac-
cording to their physiological activities. The relationship of two
organs rests upon their descent from the same organ of a common
racial form. The proof of this relationship establishes the Homology
of the organs in different animals. Thus the anterior or posterior
extremities of the Amphibia, Eeptiles, Birds, and Mammals are homo-
logous to each other and to the pectoral or ventral fins of the Fish,
because these organs have a common origin. The limbs of the Ferte-
brata and those of the Arthropoda are not homologous, but only ana-
logous, because they cannot be referred to a definite organ in a
common racial form. They were first formed independently within
each of these groups, and the superficial similarity which they exhibit
is only the result of their adaptation to the" same function.

Zoological research has further proved that in the process of time
organs can undertake functions quite different from those which they
originally performed (principle of the change of functions). The air-
bladder of fishes, for example, is principally a hydrostatic apparatus
used by these animals for rising or sinking in the water. At the same
time, in certain fishes the air-bladder may also undertake the secondary
function of a respiratory apparatus. This secondary function becomes
in the higher Vertebrata the chief function ; the lung rises out of the
air bladder, and the original function is quite lost. The so-called
rudimentary organs are of great importance in comparative anatomy ;
these are degenerated organs which are not in the condition to perform
any useful function for the organism. They are remains of originally
well-developed and functionally important organs, retained by inherit-
ance, but in the act of disappearing. Thus the human processus vermi-
formis is a small remainder of an intestinal caecum which is greatly
developed in certain Mammalia of a lower order and energetically
takes part in the work of digestion.

How is the rise of the lowest, simplest Metazoa to be imagined ?
This question is answered by various theories ; one of these, the
Gastrsea theory, has been very generally accepted. This theory rests
upon two series of facts :

1. In the development of very many Metazoa there arises, by
repeated division of the egg cell, a hollow group of similar cells,
which in its structure shows a general correspondence with a
Protozoan colony (Polvox, Magosphcera). The cells of this group

ii INTRODUCTION TO STUDY OF METAZOA 57

(blastula) arrange themselves in all Metazoa in a double layer, and in
many cases they do this in the simplest manner, i.e. by the wall of the
hollow sphere sinking in at one spot. The sunken portion lines the
non-invaginated portion of the hollow sphere, and so we have a
pouch with a double wall (ectoderm and endoderm). The inner
layer, the endoderm, surrounds a cavity, the areh-enteron, which
opens outward by an aperture, the primitive mouth or blastopore.
The outer layer, or ectoderm, everywhere supplies the outer integu-
ment and the nervous system of the animal; the inner, the enteric
epithelium and the glands which proceed from it. This germinal form
is called Gastrula.

2. The body of one of the lowest Metazoa, e.g. one of the simplest
Ccelenterata, throughout life consists of two layers, which in all essentials
correspond with the two germinal layers of the Gastrula. The outer
layer, the ectoderm, represents the outer integument ; the inner, the
endoderm, the epithelial wall of the intestinal cavity. The latter
surrounds the intestinal cavity, which opens at one point, the mouth.

From these two series of facts the following conclusions may be
drawn : —

1. All Metazoa are descended from one common ancestral form,
which possessed essentially the structure of one of the lower Ccelenterata.
This hypothetical ancestor, the Gastrcea, is met with in all Metazoa
as a transitionary stage in their development — as a Gastrula.

2. The Gastrcea itself arose in a similar way from a Protozoan
colony in the shape of a hollow sphere by the formation and the
gradual deepening of a depression, just as in the individual develop-
ment of many animals the gastrula arises by invagination out of a
hollow group of cells, the product of the segmentation of the egg.

The three chief Divisions of the Metazoa.

A. The body consists essentially of two layers — the ectoderm and the endoderm.
There is no middle layer as a rule, and where such does occur, its close relation to
either the ectoderm or the endoderm, or to both, is clear. Intestine with one external
aperture — the mouth. A body cavity between the intestine and the integument is
wanting ; so also are blood-vessels and excretory organs. A nervous system is either
wanting, or, where it occurs, is little centralised.

Comprises : The Second Race or Phylum of the Animal Kingdom— Zoophyta or
Ccelenterata.

B. With well-developed mesoderm sharply distinguished from the ectoderm and
endoderm. Gastric cavity with a single aperture opening externally (mouth). Body
cavity and blood-vascular system wanting. Excretory organs (water- vascular system)
present. Nervous system centralised.

Comprises : The Third Race or Phylum of the Animal Kingdom— Platodes.

C. With well developed mesoderm, sharply distinguished from the ectoderm and
the endoderm. Intestine with two external apertures (oral and anal). Generally
with a body cavity in mesoderm. Blood-vascular and excretory systems usually
present. Nervous system centralised.

58 COMPARATIVE ANATOMY CHAP.

Comprises all the remaining races of the animal kingdom, — viz. the fourth,
Vermes ; the fifth, Arthropoda ; the sixth, Mollusca ; the seventh, Echinodermata ;
the eighth, Tunicata ; and the ninth, Vertebrata.

THE SECOND KACE OE PHYLUM OF THE ANIMAL KINGDOM.

ZOOPHYTA OR CCELENTERATA.
Systematic Review.

CLASS I. Gastrseadse. Without pores in the body wall and without tentacles.
,, II. Porifera or Sponges. With pores in the body wall, without tentacles.
,, III. Cnidaria or Stinging Animals. Without pores in the body wall, with

tentacles.

Of the three chief classes of the Ccelenterata, the Gastrceadce show essentially the
structure of a Gastrula, while the other two groups contain animals more highly
differentiated, which, developing in entirely different directions, cannot be com-
prehended in one description.

CLASS I.

The Gastrseadse.
Systematic Review.

A. The Physemaria, Haliphysema, Gastrophysema.
B.. The Dicyemidse. Dicyema (Fig. 53).
C. The Orthonectidse. Rhopalura, (Fig. 54).
Appendage : Trichoplax adhcerens (Fig. 55).

The Gastrceadce are animals whose structure essentially corresponds
with that of the Gastrula. In some forms the organisation is com-
plicated ; in others, no doubt in adaptation to the parasitic manner of
life, somewhat simplified. The Physemaria are bi-laminar tubes attached
to the sea bottom by that portion of their bodies which is opposite to
the aperture. The ectoderm consists of fused cells (syncytium) ; the
endoderm of collar cells, each with a flagellum. The sexual products
are developed in the endoderm. Foreign bodies are contained in the
ectoderm. Were the body wall of the Physemaria perforated by pores
they would have to be considered as the simplest sponges.

The bodies of the Dicyemidce (Fig. 53) and Orthonectidce (Fig. 54)
which are parasitic in Cephalopoda, Echinodermata, and Turbellaria, also
consist of two principal layers; the ciliated ectoderm forms an un-
broken layer of not very numerous cells round the inner solid layer,
which is generally considered to be endoderm, and this layer consists
either (Orthonectidce) of a mass of cells, or of one single multinuclear
axial cell (Dicyemidce). The oral opening and gastral cavity have here
disappeared in the same way as in the Cestoda. The body of the
Orthonectidce is outwardly ringed, and between ectoderm and endoderm
has a layer of ectodermal muscular fibres. In the Orthonectidce
spermatozoa and eggs are produced in the endoderm, but in different

59

disi. . .lar individuals. In the Dicyemidce no spermatozoa have yet
been discovered, but many egg-like germs, which apparently without
fertilisation develop as eggs within the axial cell.

Th'-i course of development is as follows. The egg or the unicellular
germ divides into two unequal portions. The larger segmentation cell
which is thus produced (macromere) remains at first undivided, while
the smaller (micromere) divides repeatedly. The descendants of the
latter grow round the larger cell, finally completely surrounding it,

FIG. 53.— Young Dicyema, after Whitman.
e, Ectoderm ; en, endoderm cell, with nucleus
(n) ; em, embryo.

FIG. 54.— Rhopalura Giardii, (j> , after
Julin.

and form the ectoderm, and, in the Orthonectidce, the muscular fibres as
well. The large cell remains undivided in the Dicyemidce and be-
comes the axial cell, while in the Orthonectidce it yields by division the
group of endodermal cells.

Appendage : Triehoplax adhserens (Fig. 55). — This is a remark-
able animal discovered in the Graz marine aquarium, which presents
the appearance of a thin flat ciliated body like an Amoeba, irregular
and varying in shape. It is composed of three layers — the lowest,
which adheres to the surface on which the animal rests, consists of
cylindrical cells, the uppermost of tesselated epithelium. The layer

60 COMPARATIVE ANATOMY CHAP.

between these consists of branched and partly anastomosing cells,
which lie in a hyaline ground substance. The cells of the lowest layer
possess processes, which pass into the processes of the cells of the
middle layer without sharp distinction. As long as we Lave no know-

FIG. 55.— Part of a vertical section through the body of Trichoplax adhaerens, after
F. E. Schulze.

ledge of the reproduction and development of this animal, judgment
as to its morphology must be suspended.

[Note. — Of. p. 175. Where the author suggests that the Dicyemidce and Ortho-
nectidce, on account of their similarity to the sporocysts, are degenerated Trematoda.
-Tr.]

Literature.

E. Haeckel. Biologische Studien. Heft II. 1877.

Ed. van Beneden. Recherches sur Us Dicyemides. Bull. Academic Belgique.

Bruxelles, 1876.
C. 0. Whitman. A Contribution to the Embryology, Life-History, and Classification

of the Dicyemids. Mittheil. aus d. zool. Station zu Neapel. T. IV. 1882.
A. Giard. Les Orthonectides. Journal de I' Anatomic et de la Physiologic. T. XV.

1879.

E. Metschnikoff. Untersuchungen ilber Orthonectiden. Zeitschr. f. wiss. Zoologie.

Bd. 35. 1881.

Julin. Contribution & Vhistoirc des Mesozoaires. Archives de Biologie. T. III.
1882.

F. E. Schulze. Ueber Trichoplax adhaerens. Zoolog. Anzeiger. Bd. VI. 1883.

S. 92-97.

CLASS II. Porifera or Sponges.
Systematic Review.

Sub-Class I. Calcaria. — Skeleton composed of spicules of carbonate of lime,
always present. According to the structure of the soft body, Ascones, Sycones, and
Leucones. Olynthus, Ascandra, Sycandra, Leucandra.

Order 1. Calcispongige.

Sub-Class II. Non-Calcarea. — Skeleton seldom wanting, but never of calcareous
spicules, rather of siliceous spicules or spongin fibres. According to the structure of
the soft body, Leucones.

PORIFEEA
Order 2. Hexactinellidse.

61

Siliceous needles, isolated or bound together by masses of silica into a continuous
firm framework, tri-axial. Flagellate chambers cylindrical, placed radially, similar
to the radial tiiles of Sycandra. Most forms fossil. Living: Euplectella, Hyalonema.

FIG. 56.— Skeleton of a horn sponge attached to a stone, o, Oscula.

Order 3. Spiculispongiae.

Skeleton consisting of independent siliceous spicules of various kinds, rarely
wanting. The spicules are often bound together into bundles by an organic sub-
stance, or form firmly connected skeletons interlocking by means of knotty out-
growths ; they are, however, never cemented together by siliceous masses. Gfeodia,
Plakina, Chondrosia, Oscar ella, and Halisarca (without skeleton), Tethya, Tiiberella,
Suberites.

Order 4. Halichondrina.

Skeleton composed of siliceous spicules, chiefly uniaxial, cemented together by a
more or less horny substance (spongin). Halichondria, Reniera, Spongilla (in fresh
water) Myxilla, Clathria.

62 COMPARATIVE ANATOMY CHAP.

Order 5. Ceraspongia (Fig. 56).

Skeleton consists of horn fibres (spongin). Proper spicules wanting. Fragments
of foreign spicules, sand, etc., are often used for strengthening. Spongelia,
Euspongia officinalis (bath sponge), Aplysina.

The form of the body in the Porifem is so wonderfully varied that
no general description of it is possible, on account of their great variety
in shape. Sponges or sponge stocks can be knob-like, pear-shaped,
crust-like, funnel-shaped, cylindrical, or spherical. Many are irregularly
branched. Some have a radiate structure. All are attached, or have
part of their bodies buried in mud.

With the exception of the Spongillidce, all sponges live in the sea.
In many sponges the external forms to be met with, even in one
and the same species, varies to an extraordinary degree. The same
individual even, in different parts of its body, may show differences
of texture and structure, and variations in the composition of the
skeleton.

The inner structure of sponges is not less varied. As an example
let us take Olynthus (Fig. 57). This sponge is vasiform and rather
thin walled ; it is attached by its blind end, while
the opposite free end is broken through by an
opening (oseulum). The body wall is perforated
by pores which can open and shut. The water
streams through the pores into the body cavity,
which may be compared with the gastric cavity of
the gastrula, and flows out through the oseulum.
The wall, as far as is at present known, consists
of two layers : (1) an .outer layer formed of a
tolerably homogeneous fundamental substance, in
which are imbedded cells and calcareous needles ;
(2) an inner epithelium of collar cells. Perhaps
here also, outside the layer which contains the
skeleton, there is a thin tesselated epithelium, in
which case the body would consist of three layers
— an outer ectodermal layer, an inner endodermal epithelium, and an
intermediate mesodermal layer of connective tissue. Those sponges,
which are essentially of the same degree of organisation as the Olynthus,
are called Ascones.

A higher degree of organisation is attained when the body wall
becomes thicker and cylindrical tubes or pouches arranged close
together penetrate into the thickened wall radially round the central
cavity (Fig. 58). The outer surface of the sponge is then often raised
in numerous cones over these radial tubes. The radial tubes are lined
with collar epithelium, while the epithelium of the central or gastral
cavity is changed into a pavement epithelium. The outer pores
in this case lead first into the radial tubes, from these into the central
cavity, and thence through the oseulum to the exterior. Sycones.

II

PORIFERA

63

In most sponges, however, the " canal system " is more complicated.
The collar epithelium is limited to numerous so-called "ciliated
chambers," which are sac-like, and generally lie
scattered in the much thickened mesoderm of
the body wall (Fig. 59, gk). The pores of the
outer surfaces of the body lead into much-
branched canals of varying width, which are
lined with tesselated epithelium; these, as
afferent canals, enter the 'ciliated chambers.
Other canals of varying width, which often
unite into larger canals, lead out of the
chambers as efferent canals into the variously
shaped central cavity, which again opens out-
ward by means of an osculum. Such forms
are known as Leucones.

The movement of the flagella of the collar
epithelium maintains a constant stream of water
through the canal system of the sponges. The
water enters by the pores, passes through the
canal system, and flows out again through the
osculum.

The canal system may vary extraordinarily in details.
Its structure and arrangement are of importance in
classification.

The coalescence of the afferent canals often causes a
system of large lacunae and cavities lying quite near the
surface, the sub-dermal spaces ; into these the pores open
either directly or through canals, the water passing on
from the sub-dermal spaces by special canals into the
ciliated chambers.

The canal system of the sponges may be greatly de-
veloped in comparison with the solid matter of the
middle layer (mesoderm), or the solid tissue may pre-
ponderate. In the first case the sponge has a loose, in
the second a firm, texture.

p1G. 53.— Sycandra ciliata,
Haeckel, after Vosmaer.
Longitudinal section through

the body wal1 in tlie upper

The mesoderm of the sponges is represented
by a middle layer of connective tissue, chiefly
gelatinous, with cells imbedded in it. The
latter are either spindle-shaped or star-shaped, occasionally vacuolated.
Some of them often contain colouring matter (pigment cells) ; others
can move like Amwbce (migratory cells). Long spindle-shaped and
finely-branched cells occasionally lie concentrically at the commence-
ment of the afferent canals, and no doubt serve as contractile cells
for closing the pores.

The mesodermal connective tissue is, in sponges, the place of for-
mation of the very varied skeletal structures. These consist either
of carbonate of lime, or of silica, or of horn known as spongin.

64

COMPARATIVE ANATOMY

CHAP.

Skeletons of silica and of horn fibres are found combined. The sili-
ceous or calcareous skeletons consist of small bodies of extraordinarily
different shapes, the so-called spicules, most probably formed in the

FIG. 59.— Part of a section through Halisarca lobularis, after F. E. Schulze. ec, "Ectodermal
pavement epithelium ; gh, gastral cavity ; m, mesoderm ; p, pores ; gk, ciliated chambers ; zk, affer-
ent canals ; o, eggs in different stages of segmentation.

cells. There are uniradiate, triradiate, quadriradiate, sexiradiate, multi-
radiate forms, stars, spheres, etc. The skeleton of a sponge may con-
sist of only one sort of
spicule, or two or more
sorts may occur together.

The single spicules lie
either loosely near each
other, or are cemented
together into coherent
frameworks. The same is
the case with horn fibres.
The ordinary bath sponge
is only a framework of such
horn fibres ; it is merely
the skeleton of a marine
animal (Fig. 56).
A nervous system is not yet with certainty proved to exist in the
Porifem.

Reproduction is either asexual or sexual.

Asexual reproduction takes place by external or internal budding
or gemmation.

External gemmation. — A sponge may put out buds at various

FIG. 60.— Various forms of skeletal spiculae
from Sponges.

POEIFERA

65

poin ! oi bh< body surface. These, without detaching themselves,
grow \d can in their turn form buds. Sponge colonies thus

ouds may at various points grow together, the colony itself

arise.

may again iave the appearance of a plexus or framework. The holes
and interspaces of such a colony may then again assume the character
of a canal system (pseudo-canals). These must, however, according
to their origin, be sharply distinguished from the real canal system
which runs through the walls of every sponge individual Separate
sponge individuals may also fuse and form colonies. The number of
oscula generally corresponds with the number of individuals which
form the colony (Fig. 56, o).

In the so-called internal gemmation groups of cells called
gemmulse detach themselves from the sponge body, and after a
period of rest develop into complete sponges. Observers differ as to
the finer processes which take place during the development of these
gemmulse.

Sexual reproduction. — Sponges are either hermaphrodite or
disecious ; in the former case the eggs and spermatozoa are not pro-
duced at the same time in the same individual or colony ; they are
protandrously hermaphrodite. The eggs and spermatozoa seem to
develop from mesoderm cells.

Development. — The course of development of the sponge from the fertilised egg,
which often begins within the mother body, seems, to judge from the as yet insuffi-
cient and often contradictory observations which have been made, to be so varied

FIG. 61.— Sections of three stages of development of Oscarella lobularis, after K. Heider.
A, Gastrula which has attached itself. B, Rudiments of mesoderm and canal system. C, Forma-
tion of the osculum and ciliated chambers, e, Ectoderm ; en, endoderm ; m, mesoderm ; o, osculum ;
p, pores ; ivk, ciliated chambers.

that it is hardly possible to form a generally applicable scheme. We select the
newly investigated development of Oscarella (Halisarca] lobularis (Fig. 61).

By means of repeated egg division a freely swimming larva arises, the Blastula.

This is a hollow sphere, whose wall consists of one single layer of flagellate cells.

The blastula by invagination becomes a gastmla. This attaches itself by the

gastrula mouth or blastopore, and the aperture gradually narrows and finally closes!

VOL. I F

66 COMPARATIVE ANATOMY CHAP.

Gelatinous substance is secreted between the ectoderm and endoderm, and into
this cells migrate most probably from the endoderm. Thus the connective tissue
mesoderm arises. At the same time radial invaginations of the endoderm vvhich lines
the arch-enteron are formed in the mesoderm, and grow towards the ectoderm. These
invaginations are rounded off and become the ciliated chambers (wL). Their com-
munication with the gastric cavity becomes narrowed. The ciliated chambers become
connected with the surface, either by the formation of pores through the external
membrane (in the case of chambers lying superficially), or by the formation of short
invaginations of the ectoderm, which finally reach the ciliated chambers. The
osculum arises at the aboral pole, by the lengthening and breaking through of the
body cavity. Sycone stage.

According to these observations, the epithelium of the ciliated chambers, the
efferent canals, and the central cavity (gastric cavity) is of endodermal origin ; the
tesselated epithelium on the surface of the body and the epithelium of the afferent
canals (at least partly) of ectodermal origin.

According to other observers, in fresh water sponges the ectoderm is thrown off
by the larva, and the whole adult sponge is derived from the endoderm.

The observations of several investigators agree in establishing the fact that the
gastrula of the sponge attaches itself by the edges of the blastopore. The osculum
of the sponge therefore represents neither the blastopore of the gastrula nor the
mouth of the Ccelenterata. The Porifera thus appear as a laterally developed group
of the lower Metazoa, which do not admit of direct comparison with other Coslenterata,
but are only distantly related to them.

Literature.

O. Schmidt. Die Spongien des Adriatischen Meeres. Leipzig, 1862. Drei Supple-

mente, 1864 bis 1868.

The same. Grundzuge eincr Spongienfauna des atlantischen Gebietes. Leipzig, 1870.
Vfl^aeckel. Die Kalkschwamme. 3 Bde. Berlin, 1872.

0. Schmidt. Die Spongienfauna des mexikanischen Meerbusens und des caraibischen

Meeres. Jena, 1880.

F. E. Schulze. Untersuchungen uber den Bau und die Entwickelung der Spongien, in

Zeitschrift f. w. Zoologie. Bd. 25-35. 1876-1881.

G. C. Vosmaer. Porifera. In Bronn's Klassen und Ordnungen des Thierreiehs.

Leipzig, 1882.
N. Polejaeff. Report on the Calcarea in Chall. Exped. Rep. vol. VIII., part XXIV.

London, 1883.
F. E. Schulze. Report on the Hexadinellidce. Chall. Exped. Rep. vol. XXL, 'part

LIII. London, 1887.
Compare also the older works and treatises of Grant, Lieberkiihn, and Bowerbank,

and newer investigations of Zittel, Barrois, Keller, Heider, Marshall, Lenden-

feld, Gotte, etc.

CLASS III.— Cnidaria.

Systematic Review.

Sub-Class 1. Hydrozoa.

Prototype : Hydropolyp or Hydrula. In all Hydrozoa an ectodermal oeso-
phagus is wanting ; the mouth leads direct into the endodermal gastric cavity. Gastral
filaments are wanting. The sexual products mostly arise from the ectoderm. The
sexes are generally separate.

II

CNIDARIA

67

Order 1. Hydridse (Fresh-water polyps).

Single individuals or small stocks without envelopes consisting of a few similar
individuals. Reproduction asexual by gemmation, and sexual. Hydra develop direct
from the egg. Hermaphrodite. Hydra, in fresh water.

Order 2. Hydromedusse.

Hydroid colonies, which are at least dimorphic, since, besides the sterile nutritive
polyps, there arise by gemmation sexual persons, which either detach themselves as
Craspedote Mcdusce and swim about freely, or remain united with the colony as
medusoid gonophores. In one series of Hydromedusce the attached Hydroid form is

FIG. 62.— Bougainvillea ramosa (after Allman), with budding Medusa, h, Nutritive polyps ;
mk, Medusa buds ; m, detached young Medusa (Margelis ramosa).

suppressed, as the Craspedote Medusa develops direct from the fertilised egg into
another Medusa. The systematic relationship of single forms is naturally determined
both by the Hydroid and by the Medusa forms. As the whole life-history and
development of only a minority of the many species is available, and in many species
only the Hydroid form is known, in others only the Medusa form, a natural system
of the Hydromedusce is still a desideratum.

Hydroid form. Medusa form.

Sub-Order 1. Hydrocorallia.

Hydroid stocks, with calcified peri- Wanting,

derm skeleton. The sexual products are
produced in gonophores. Stylaster,
Millepora.

68

COMPARATIVE ANATOMY

CHAP.

Hydroid form.

Sub-Order 2.
Tubularia.

Small hydroid colonies, naked or
covered with a chitinous envelope (peri-
derm). The chitinous envelope never
widens into a cup (theca) round the polyp
head. In many forms the Medusce are
reduced to gonophores, which do not de-
tach themselves.

Medusa form.

Anthomedusae.

Craspedote Medusce, without marginal
vesicles and otoliths, with ocelli at the
bases of the tentacles. Gonades in the
outer wall of the gastric peduncle ; 4,
seldom 6 or 8, radial canals.

v Syncoryne Sarsii.
Podocoryne earned.
Eudendrium ramosum.
Bougainmllea ramosa (Fig. 62).
Stauridium cladonema.
Cordylophora lacustris \
(in fresh water). J
Tubularia larynx.
Unknown.

Examples.

Sarsia tubulosa.
Dysmoi*phosa carnea.

Lizusa octocilia.

Margclis ramosa (Fig. 62, m).

Cladonema radiatum.

Wanting.

Wanting.
Ctenaria ctenophora.

Sub-Order 3.

Campanaria.

Small hydroid stocks with chitinous
periderm, which widens round the polyp-
head into a theca, into which the head
with the tentacles can be withdrawn.
The Medusa buds or sessile gonophores
generally arise united into groups in
special modified polyps devoid of ten-
tacles and mouth (gonangia).

Leptomedusse.

" Craspedote Medusce, some without,
some with, marginal vesicles, the latter
developed from the base of the velum
with ectodermal otolith cells. Ocelli at
the tentacle bases sometimes present,
sometimes wanting. Gonades always in
the course of the radial canals. Number
of radial canals various, often very great "
(Haeckel).

Examples.

Campanularia geniculata.
Unknown.

Campanulina tenuis.

Unknown.

Unknown.
Laomedia Caliculata.

The Plumularia and the Sertularia are
generally placed near the Campanaria.
These are elegantly branched Hydroid
stocks. In the first, the cups (thecrc)
which contain the nutritive polyps are in
a single row, in the second in two rows
on opposite sides of the stem. The sexual
products form bud-like outgrowths (gono-
phores), which generally arise in groups
on special modified polyps devoid of

Obelia geniculata.
f Eucope campanulata.
1 (Fig. 65, p. 74.)
Phialidium variabile.
GastroUasta Raffadii.
Aequorea Forskalea.
Wanting.

n CNIDARIA 69

Hydroid form. Medusa form.

mouth and tentacles, and they are sur-
rounded by a chitinous periderm. It is
as little known in this case as in that
of the Hydrocorallia, whether these
gonophores are degenerate Medusae which
remain sessile, or simple sexually differ-
entiated Hydropolyp buds.

Sub-Order 4.
Wanting. Traehomedusse.

Craspedote Medusae with auditory or-
gans (tentaculocysts), with endodermal
otolith cells, which sometimes stand
freely on the margin of the umbrella,
sometimes enclosed in auditory capsules,
Ocelli generally wanting. Gonades
always in the course of the radial canals.
Number of radial canals 4, 6, or 8, never
more ; between them often blind centri-
petal canals. Direct development with
metamorphosis. Olindias Miilleri, Rho-
palonema velatum, Aglantha digitalis,
Geryonia probosddalis, Carmarina
hastata.

Sub- Order 5.
Wanting. Narcomedusse.

Craspedote Mcduscc with auditory or-
gans, always standing freely on the
margin of the umbrella, with endodermal
otolith cells. Ocelli mostly wanting.
Tentacles at some distance from the
margin of the umbrella, inserted on the
exumbrella, bound to the umbrella mar-
gin by clasps (peronia), Avhich divide it
into a number of collar lobes. Gonades
on the gastric peduncle, often spreading
from it peripherally in radial gastric
pouches. Radial canals sometimes
wanting, sometimes spreading out in the
shape of flat gastric pouches. Circum-
ferential canal sometimes obliterated.
Number of tentacles, lobes, and pouches
variable ; seldom 4, generally 8 or more,
up to 32. Development usually direct,
with metamorphosis. Cunina, Pcgan-
tha, ^Egincta, ^Eginopsis, Solmaris.

Order 3. Siphonophora.

Polymorphic, freely swimming Hydrozoa stocks, whose individuals or persons,
Craspedote Medusce, are modified for special functions.

70 COMPARATIVE ANATOMY CHAP.

Sub-Order 1. Siphonanthe.

The heteromorpliic persons bud on a variously formed stem, which may be com-
pared with the gastric peduncle of a Medusa.

Family 1. Calyconecta. — Without pneumatophore and feeler (taster) ; with one
or more swimming bells at the upper end of the stem. The remaining hetero-
morphic persons arranged in groups (cormidia), which can detach themselves
from the stem as Eudoxice and Ersoece. Praya (Fig. 85, p. Ill), Diphyes,
Abyla, Hippopodius.

Family 2. Physonecta. — With pneumatophores, without aurophore, with several
swimming bells, and with feelers. Apolemia, Agalma, Anthemodes, Hali-
stemma, Physophora, Forskalia.

Family 3. Auronecta. — With one large pneumatophore, under which stands a
circle of swimming bells, and in the dorsal middle line of the latter a large
medusoid air bell (aurophore), which secretes gas, and may be considered as a
modified swimming bell. Stem shortened and thickened. Without feeler (?).
Stephalia (Fig. 84, p. 110), Auralia, Rhodalia.

Family 4. Cystonecta. — With large pneumatophore, without aurophore.
Swimming bells and covering pieces wanting. Hhyzophysa, Physalia. Stem
under the pneumatophore very much shortened, and flattened into a disc.

Sub-Order II. Disconanthe.

The heteromorphic individuals bud from the under side of a disc, which contains
a many-chambered pneumatophore, and may be compared with a Medusa umbrella.
The margin of the disc carries a ring of numerous tentacles. In the middle of the
subumbrella stands the central gastric peduncle as chief siphon.

Family 5. Disconecta. — Discatia, Porpita, Porpalia (Fig. 87, p. 114), Velella.

Sub-Class II. Scyphozoa.

Prototype : The Scyphopolyp or the Scyphula. In the Scyphozoa we find an
ectodermal oesophagus. Gastral or mesenterial filaments are present in all cases on
the septa or gastric ridges. The sexual products arise out of the endoderm. The
sexes are generally found in separate individuals.

Order 1. Anthozoa (Corals).

Attached individuals or colonies. The body remains essentially on the samel
grade as the Scyphula. The ectodermal oesophagus sinks in the form of a tube into;
the spacious gastric cavity, and round it the latter is divided by septa into a variable \
number of separate pouches. The free internal edges of these septa run through i
the gastric cavity to the aboral end of the body.

A. Octocorallia.

Sub-Order 1. Alcyonaria. — Generally with 8 septa and 8 pinnate tentacles. Polyp
colonies of very various shapes. Skeletal forms very varied. Alcyonium,
Pennatula, Kophobelerrwwn (Fig. 63), Gorgonia, Isis, Tubipora.

B. Tetracorallia.

Sub-Order 2. Rugosa. — Number of septa great, a multiple of 4. With calcareous
skeleton. Fossil Paleozoic forms.

CN ID ARIA

71

C. Hexacorallia.

Sub-Order 3. Antipatharia (horn corals).— With 6 or 24 simple tentacles. Colonies

with horny axial skeleton. Antipathes (6 tentacles, only 2 developed septa),

Gerardia (24 tentacles and septa).
Sub-order 4. Madreporaria (stone corals).

— Mostly colonies, more rarely in-
dividuals with strongly developed

calcareous skeleton. 6n simple ten-
tacles, and septa present in great

numbers and variously arrranged.

Madrepora, Astroides, Fungia, Astrcea,

Mceandrina, Cladocora, Caryophyllia,

Flabellum.
Sub -Order 5. Actinaria (flesh corals). —

Mostly individuals with 611 tentacles,

and septa generally in considerable

numbers and varied arrangement.

Without skeleton. Cerianthus, Zoan-

thus, Actinia, Anemonia, Adamsia,

Edivardsia.

Order 2. Scyphomedusse (Acraspeda).

Mostly free-swimming individuals of bell-

or disc-shape, in which the mesodermal

supporting lamella is developed into

a large gelatinous mass. The ecto-

dermal oesophagus lies within a gastric

peduncle which hangs down from the

middle of the subumbrella. The 4

radial pouches of the Scyphula de-
generate in the higher forms. The

exumbrellar and subumbrellar walls of

the peripheral portion of the gastric

cavity grow together in such a way

that only a variously developed system

of radial gastro-canals remains. With

tufts of gastral filaments. A true velum

is wanting ; in its place, marginal

lobes, containing processes of the

gastro-canal system are present.
A. Medusce with deeply vaulted um-
brella.— The 4 radial gastric pouches and
the septa which separate them more or
less clearly retained.
Sub -Order 1. Stauromedusse. — 4 septa

retained (Lucernaria) or reduced to 4

pillars ( Tessera), 4 or 8 gonades in the subumbrellar wall of the 4 gastric

pouches, without sensory bodies (rhopalia). Lucernaria (attached, with 8

marginal lobes, each carrying a tuft of stinging batteries). Tessera (free, with-
out distinct marginal lobes, with 8 tentacles).
Sub-Order 2. Peromedusse. — The 4 septa reduced to 4 pillars, in consequence of

which the 4 gastric pouches are united into a circumferential sinus. With 4

;

FIG. 63.— Kophobelemnon Leuckartii.

72 COMPARATIVE AXATOMY CHAP.

inter-radial rhopalia. S or 16 marginal lobes, 4 or 12 tentacles. Pa'icvlj.w.
Pcriphylla,

Sub-Order 3. Cubomedusse (Charybdeidse). 4 septa retained. 4 pair of gonades
on the septa, freely protruding into the gastric pouches. With 4 perradial
rhopalia, which contain tentaculocysts with endodermal otolith sacs, and are
provided with one or more eyes. 4 interradial tentacles or tufts of tentacles.
Mostly with velarium. Charybdca, Chirodropus,

B. Mxlusic with flat, disc-like umbrella. — The 4 primary gastric pouches of
Scyphula degenerated by disappearance of the septa, instead of which there are S.
16. 32, or more radial canals of varying width, and often branched or anastomosing,
as survivals of the gastric cavity on the growing together of the subumbrellar and
exumbrellar intestinal walls. The 4 interradial gastric ridges or taenioles are
retained as remains of the septa, and these carry the phacelli, or tufts of gastral
filaments. Development either direct with metamorphosis, or with alternation of
generations. In the last case an attached Scypliula arises out of the gastrula and
develops into a young attached Medusa (Scyphistoma). This multiplies in most
cases by a sort of repeated fission or gemmation (strobilation). The constricted
M>:duso. (JEpJiyra} changes through metamorphosis into the adult form.

Sub-Order 4. Discomedusse.

Family 1. Cannostomse.— With simple month tube, without oral arms, with s<juare
month, and short solid tentacles. Nausitlwe (Fig. 67, p. 77).

Family 2. Semostomse.— With 4 long Hag-like oral arms and cross-shaped mouth.
With long hollow tentacles. Pclayia noctiluca. Cyanca, Aurdia aurita.

Family 3. Rhizostomse. — Mouth grown together. With numerous small suckers
<m the 8 long root -like oral arms ; without tentacles. Cassiopcca, Ptlcnuc
(PJiizostoma}, Cutylorldza., Cmml»;ssa, Cannorliiza (Fig. 70. p. 85).

Sub-Class III. Ctenophora.

Cnidaria with sensory bodies at the aboral pole ; with 8 meridional rows of
ciliary or ctenophoral plates, with ectodermal (esophagus. Without gastral filaments.
Development direct, without alternation of generations. Mesodermal jelly strongly
developed, with muscle, nerve, and connective tissue elements. Hermaphrodite.

Order 1. Tentaculatae. — With 2 capturing filaments with one row of simple
branches ; the filaments can be withdrawn into sacs situated in the lateral
perradii. Gastro-canals ending blindly.

Family 1. Cydippidw. —Body globular or egg-shaped. Hormiphora (Fig. 68,

p. 79).
Family 2. Lobatae. — Body compressed in the lateral plane, with 2 oral lobes

in the median plane. Encltdris.

Family 3. Cestidse. — Body band-like, compressed in the lateral plane, with-
out oral lobes. CV.s^/s.

Ordf-r 2. Nuda. — Without capturing filaments. Mouth wide, oesophagus very
spacious. Castro-canals much branched and anastomosing.
Family 4. Beroidae.— Bcruc.

The Hydrozott and Scyphozoa probably descend from attached forms, which are
only distinguished from gastrula attached by the abond pole, either, on the one hand
(•HydrozfM\ by their possession of circumoral tentacles, or, on the other (Scyphozoa],
by the formation of an ectodermal rcsophagus. Indications of this descent can still
be seen in the fix-ely swimming Hyfli-(i'm<:dns«: and Sc'iijilionKdii.ttf.e in their radial

CNIDARIA— GENERAL

73

structure and in the absence of all organs at the aboral side of the body (exumbrella),
i.e. at that part of the body by which the stationary forms are attached ; and further
in the fact that very many Hydromedusce and many Scyphomcdusce actually pass
through such a stage of attachment in the course of their development. As for the
Ctcnophora, their descent from attached forms, and, generally, their position among
the Cnidaria, seems very doubtful. The presence of complicated sensory bodies at
the aboral pole, and the constant occurrence of 8 rows of rowing plates with corre-
spondingly arranged gastro-canals, and the marked departure from the strictly radial
type, point to the fact that in their case the adoption of the Swimming manner of
life dates very far back.

si

I. General.

We shall better understand the varied organisation of the Cnidaria
if we keep clearly in mind that they can in all cases be referred to
one of the three following forms :
(1) the Hydropolyp or Hydrula, (2)
the Scyphopolyp or Scyphula, (3) the
Ctenophora.

1. The simplest form (from which
all others can be deduced) is that of
the Hydropolyp or Hydrula.

A Hydropolyp (Fig. 64, A) (type :
Hydra) is a pouch- shaped gastrula
attached by its aboral pole, and
possesses, round its mouth, hollow
tentacles as evaginations of the body
wall. The gastric cavity is continued
into the tentacles. Between endo-
derm and ectoderm there is a struc-
tureless supporting membrane (si).

A more highly developed form,
the Craspedote Medusa (Fig. 64, B]
may proceed from the Hydropolyp
by adaptation to a free -swimming
manner of life. The aboral portion

of the Hydropolyp body (from the FIG. 64.-A, Diagram of a Hydropolyp
attached pole to the tentacles) (longitudinal section); B, of a Craspedote

spreads out like an umbrella or bell, £££ *\£5ttJS£ ilSfti

and becomes the exumbrella of the tween ectoderm and endodenn ; rk, radial

Medusa. The oral portion of the canal; gl> vascular lamella or cathammai

i i / <• .-, , , plate ; v, velum ; rik, circumferential canal.

body (from the tentacles to the

mouth) also widens out, deepens, and becomes the subumbrella. We
thus have a convexo-concave body, on whose circular margin the
tentacles are radially arranged (Fig. 65, A). The mouth lies in the
middle of the concave side, and generally on the summit of a pro-
jection (gastric peduncle, gastric tube).

The supporting membrane of the Hydropolyp thickens very much,

3'

74

COMPARATIVE ANATOMY

CHAP.

and becomes the gelatinous dise of the Medusa, an elastic passive
organ for locomotion and support. The partial fusing of the inner
wall of the oral portion of the body with that of the aboral portion
considerably reduces the gastric cavity lying between them, which
originally spread throughout the whole extent of the disc. There
remain only :

1. the cavity of the gastric peduncle into which the mouth leads.

2. A central stomach above the oral peduncle. These two parts
form the main intestine.

3. A peripheral canal at the edge of the disc (the circumferential
canal) which is continued into the tentacles.

4. Radially arranged connecting canals between the central

stomach and the circumferential
canal. These canals become
nutritive gastro- canals, which
also serve the purpose of blood-
vessels (which are wanting) and
convey food from the central
stomach to the organs at the
disc's edge. These two portions
form the peripheral intestine.

In some Craspedote Medusce,
besides the radial canals and
the circumferential canal, the
endoderm from the centre to
the circumference persists, its
two layers being pressed to-
gether by the strongly de-
veloped oral or subumbrellar
and aboral or exumbrellar jelly,
thus forming the so-called
vascular lamellae or eatham-
mal plates, the layers of which,
separating in radial strips, form
the radial canals.

The radial canals in simple

FlG.65.-Eucope campanula, partly after Medm? *™ 4 ln number, and
Haeckel A, From the surface. B, Section in the ar6 placed CrOSS-WlSC. Ihe
direction a-b-c in Fig. A. a-b, Perradius ; 6-c, ad- radii in which they lie are

-M ji „ ,«j" /T?' CK. A i\
Called PSrradll (Fig. 65, A, a-b).

In Order to define the position
Of Other tentacles, Canals, and

organs, the radii exactly half
way between the perradii have been called interradii.

Half way between the 4 perradii and the 4 interradii lie the 8
adradii (b-c) ; half way between the 8 former and the 8 latter lie the
16 subradii.

radius ; t. tentacle ; sb, marginal vesicle ; g. gonades ;
mr, gastric peduncle; r, radial canals ; if velum •
ri, circumferential canal ; ex, exumbrella ; su, sub-
umbrella; ga, jelly; ty, tentacular vessel ; b-6, main

ii CNIDARIA— GENERAL 75

From the margin of the disc a thin muscular membrane projects
into the subumbrellar cavity like a diaphragm ; it is known as the
Velum (v\ and is one of the principal locomotory organs of the
Craspedote Medusae.

There are many other organs at the margin of the disc, the princi-
pal being :

1. A double nerve ring.

2. Marginal vesicles.

3. Eye spots.

4. A ring of nematoeysts.

We shall speak of these organs later on.

The derivation of the Medusa form from the Hydropolyp is the
more probable as Medusce belonging to many divisions, in their in-
dividual development, proceed from Hydropolyps by lateral budding.
Such Medusce are then specially developed buds which have the
function of forming the sexual products and of scattering them abroad.

2d Form : the Scyplwpolyp or the Scyphula (type : Lucernaria or
a simple Coral individual) is the original form of the Scypliozoa, as
the Hydropolyp is of the Hydrozoa. It is distinguished from the
Hydropolyps in that at the oral pole the region about the mouth sinks
in to form an cesophageal tube (Fig. 99, p. 130). The epithelium
which lines it is thus ectodermal. Around this tube the gastric cavity
falls into 4, 6, or 8 compartments separated by walls ; these partition
Avails, whose free edges project into the gastric cavity as septa or
tsenioles, bearing either mesenterial thickenings or gastral filaments,
are continued along the body wall even to the aboral end of the body.
All Corals remain essentially at the stage of the Scyphopolyps (Fig.
66 ; Fig. 82, p. 107). Their typical form is that of the cylinder
or reversed truncated cone. The two ends of the body are almost
circular ; by the aboral pedal disc the animal attaches itself, and
hollow tentacles stand round the oral disc (especially at the edge).
In the middle of the oral disc lies the generally elongated slit-like
external mouth.

The external mouth, in keeping with the Scyphopolyp plan, leads,
not directly into the gastric cavity, which is lined with endoderm, but
into a tube which has arisen by an invagination from without, and is
therefore lined with ectoderm. This tube (cesophageal tube, stomo-
daeum) is open at both ends. One opening, the external mouth, leads
to the exterior; the other (the enteric aperture) leads into the spacious
gastric cavity.

The body wall consists of ectoderm and endoderm, between which
a solid middle layer is interposed, of which we shall speak later. The
endoderm and the middle layer raise themselves from the body walls,
forming vertical ridges which project into the gastric cavity through-
out the whole length of the body ; these ridges are the septa. In the
neighbourhood of the oesophageal tube these septa project far enough

76 COMPARATIVE ANATOMY CHAP.

to fuse with its walls, so that the gastric cavity here appears divided
into chambers or compartments arranged radially round the resophageal
tube (Fig. 66, to the left).

In the remainder of the body (Fig. 66, to the right) the septa pro-
ject their free inner edges into the gastric cavity, which thus appears
divided into a central portion and into radial niches separated by the
septa (analogous to the gastro-canals of the other Cnidaria). These
niches, in the neighbourhood of the cesophageal tube, are continued
direct into the radial compartments between the tube and the body

r

FIG. 66.— Diagrammatic transverse section of a Coral individual ; to the left on the level of
oesophagus, to the right on the level of the gastric cavity, a-b, Direction of the plane of sym-
metry.

wall, and these compartments again are continued into the axial cavities
of the tentacles, which are lined with endoderm (cf. Fig. 82, p. 107).
The axial cavities of the tentacles occasionally open externally at the
tip by a pore.

The number of tentacles represents, generally, the number of the
septa. The Odocomllia have 8 septa, and 8 tentacles placed so as
to alternate with them. The Tetmcorallia generally have a large
number of septa which are always a multiple of 4. The Hexacorallia
possess 6 or 6 ft partition walls and tentacles, arranged in a definite
order which cannot here be described. We can only say, quite gener-
ally, that the oldest septa project farthest, and the youngest septa least
far, towards the axis.

Most Corals have not a strictly radial structure ; on the contrary
we often find, anatomically and ontogenetically, a bilateral symmetry
in the arrangement of the parts of the body. The slit-like shape of
the mouth even is a departure from radial structure. A plane through
the chief axis of the body in the longitudinal direction of the mouth
(Fig. 66, a-b) is, in fact, a median plane — the only plane which divides
the body into two exactly similar halves.

More exact ontogenetic investigation has shown that (e.g. in the
Hexacorallia) two septa lying opposite each other to the right and left
of this plane are first formed ; these septa incompletely divide the
gastric cavity into two portions of unequal size. Two new septa are
then formed symmetrically in the larger — say the anterior — division ;

CNIDARIA— GENERAL

77

then two symmetrical septa in the posterior division. The rise of the
other septa does not by any means occur in all Anthozoa after one
and the same simple plan. (In most Actinaria the other septa arise in
pairs, each pair in the space between two older septa.) We repeat
what we said above, that the septa project the farther towards the
chief axis of the Corals the older they are. The tentacles also arise
symmetrically 'with reference to the median plane over the interseptal
spaces of the gastric cavity.

The mesenterial thickenings, or mesenterial filaments, are in-
serted in all Corals at the free inner edges of the septa.

sf.

Sk.

FIG. 67.— NausithoS. pr, Perradii ; ir, interradii ; ar, adradii ; sr, subradii ; rl, marginal lobes ;
t, tentacle ; gf, gastral filaments ; m, circular muscle of the subumbrella ; sk, sensory bodies
(rhopalia) ; g, sexual glands (goriades) ; in the middle the cross-shaped mouth.

As the free- swimming Craspedote Medusa can be referred back to
the Hydropolyps or Hydrula, so the free-swimming Acraspede or Scypho-
medusa (Figs. 67 and 70, p. 85) can be referred back to the Scypho-
polyp or Scyphula. The lower Acraspeda, with deeply vaulted, often
cup-shaped body, are only slightly distinguished from the Scyplmla,
and we find among them forms still attached (e.g. the Lucernaria).

The 4 radial gastric pouches separated by septa are still present,
or they flow together to form a great circumferential sinus, the septa
dwindling into 4 small points of connection between the sub-
umbrellar and exumbrellar gastric walls, between which the circum-
ferential sinus remains in wide-open communication with the central

78 COMPARATIVE ANATOMY CHAP.

stomach. The 4 septa which carry phaeelli or tufts of gastral fila-
ments are continued as gastric ridges or taenioles on the exumbrellar
gastric wall to the aboral pole of the body. In the higher Acraspeda,
the Discomedusce, the exumbrellar and subumbrellar walls of the
peripheral intestine (circumferential sinus) coalesce, so that here again
there arises a cathammal plate, in which variously -shaped radial
canals and radial pouches remain as survivals of the circumferential
sinus ; at such points the lamellae of the cathammal plate separate,
leaving between them spaces, the lumens of these canals. The deriva-
tion of the Discomedusce from a Scyphula form is further justified by
the fact that in many of them the Scyphula appears as an attached
early stage (Fig. 99, p. 130).

The third form of the Cnidaria is the Ctenophora (Fig. 68). Its
body is ovate, with 2 dissimilar poles; its principal axis, which
connects the two poles, coincides with the long axis of the oval. At
one pole of the chief axis (the oral) lies the mouth. The opposite
pole here, as in other Cnidaria, is called the aboral pole.

The oral aperture leads into a spacious cavity lying in the chief
axis, which has its rise, ontogenetically, through an invagination from
the exterior, and is lined, like the oesophageal tube of the Scyphozoa,
with ectoderm. We call this cavity the cesophageal cavity (" stomach "
of authors) (s).

In form the cesophageal cavity is neither round nor radial, but
very much flattened ; in a transverse section its lumen appears like a
slit. In this we find the first departure from the radial body struc-
ture of the Medusa. A plane running in the direction of the flattened
cesophageal tube, and in which the chief axis lies, is called the median
plane (c-d).

The oesophageal tube leads through another opening into a smaller
cavity lying above it and lined with endoderm — the stomach (m). The
stomach is elongated at right angles to the chief axis and the median
plane, and thus, when the animal is viewed from either the oral or
the aboral pole, forms a cross with the oesophagus. A plane running
through the chief axis in the direction of the stomach stands at right
angles to the median plane, and is called the lateral plane (<?-/).

The median and lateral planes thus cross each other at right
angles in the chief axis, just like the cross axes of the Medusa, each
axis consisting of two opposite perradii. While, however, in the
radially constructed Medusa the cross axes are quite similar, and the
planes which run through them in the chief axis divide the body into
four entirely similar quarters, the two cross axes in the Ctenophora are
not alike, and the lateral and median planes divide the body into four
quarters, of which only the two which are diametrically opposite are
similar. Either of the two planes by itself, however, cuts the body
into two similar halves.

At the aboral pole of the Ctenophora, as opposed to all other Ccelen- '

GNIDARIA— GENERAL

79

terata, there is a complicated sensory organ, the sensory body, which,
according to its structure, is best described as an auditory organ, but
perhaps serves for regulating the position of the body in the water.

From near the aboral to near the oral pole, there run along the
surface of the body in 8 meridians 8 rows of swimming plates, the so-

PIG. 68.— Hormiphora plumosa, after Chun. A and B, From the side : A, seen in the direction
c-d in Fig. C ; B, seen in the direction e-f in Fig. C. C, As viewed from the aboral or sensory
pole, u-b, Chief axis ; c-d, direction of the median plane ; e-f, direction of the lateral plane ; a,
oral ; b, aboral pole ; m, stomach ; s, oesophagus ; sg, oesophageal vessels ; tg, tentacular vessels ;
ag, aboral vessel ; r, ribs ; ts, tentacle sheath ; pp, polar plates. D, Portion of a transverse
section through the oesophagus (s) ; tb, tentacle base ; siv, oesophageal papillae ; ga, branches of
the tentacular vessel.

called ribs (r). If we use the terms applied to the Medusa, these ribs
are adradial.

Between oesophagus and stomach on the one side and the outer
integument on the other, there is a jelly which is strongly developed
in most Ctenophora, and in which various tissue elements are imbedded.

Gastro-canals, lined with endoderm, similar to the radial canals of

80 COMPARATIVE ANATOMY CHAP.

the Medusae (or the radial gastric pouches of the Corals}, branch out
from the stomach in various directions through the jelly, reaching
almost to the surface of the body. We can distinguish four kinds of
these canals.

A. Four canals which to a certain extent rise interradially out of
the stomach. Each of these canals branches dichotomously, and 8
adradial canals thus arise and run to the ribs, where they enter as
many adradial meridian or rib vessels, which run under the ribs.

B. Two canals which run along the broad side of the oesophagus
perradially and in the lateral plane (i.e. in the lateral perradii)
towards the oral pole, where they end blindly; these are the ceso-
phageal vessels (" gastric vessels " of authors) (sg).

C. Two canals which arise perradially and in the lateral plane
(i.e. in the lateral perradii), and run to the walls of the tentacle sacs
(tentacular vessels) (tg).

D. An unpaired canal (ag) running in the chief axis of the body
towards the aboral pole, and dividing, under the sensory body into
2 branches, which lie in the median perradii. Each of these branches
again divides into 2, and so 4 interradial small branches arise ;
2 of these, which are diametrically opposite, generally open exter-
nally at the sensory pole, while the others, at right angles to the latter,
end blindly. Less frequently all 4 branches open externally. The
sensory body lies between these 4 branches of the aboral vessel
(" funnel vessels " of authors).

In many Ctenophora there is a further development of two solid
pinnate tentacles, into which the gastro-canal system is not continued.
In quite young animals the tentacles lie near the aboral pole. At a
later stage, however, they move towards the oral pole.

The tentacles are inserted at the base of sac-like depressions of the
outer integument, the so-called tentacle sheaths (ts) into which they
can be withdrawn. They lie in the lateral perradii. In many Cteno-
phora the body is round in transverse section ; in others it is com-
pressed either in the lateral or the median plane. The lateral com-
pression is so great in the Cestidce that the body appears drawn out
into a long ribbon in the median plane.

If we now compare the Ctenophoral forms with other Cnidarian forms, we are at
once struck by important distinctions between them. The Ctenophora, depart in a
peculiar manner, though in a direction different from that of the Corals, from the
purely radiate fundamental form, the two cross axes being unequal. They agree
with the Scyphozoa in the possession of an ectodermal oesophagus, but otherwise
differ decidedly from them.

According to the opinion of some observers, the Ctenophora must be derived from
Craspedote Medusae. The transition from the one group to the other is made evident
by an interesting Cladonemid, Ctenaria Ctenophora. In this animal the umbrella
is much vaulted, the subumbrella much deepened ; 4 radial canals rise out of the
stomach and bifurcate into 8 adradial canals. There are only two pinnate, per-
radially placed tentacles. We need only say that the subumbrellar cavity of Ctenaria

II

CNIDARIA—BODY EPITHELIUM

81

can be compared with the cesophageal cavity of the Ctenophora. The further homo-
logies (or analogies ?) in the gastro-canal system and in the tentacles then follow of
themselves.

II. The Body Epithelium.

It is not possible to carry out a sharp histological distinction be-
tween the ectodermal epithelium which clothes the whole exterior of
the Cnidarian body and the endodermal gastro-canal epithelium. We
Tiere find the ectoderm and the endoderm, histologically, still in a
rather undifferentiated condition ; this is seen most clearly from the
fact that the endoderm can in some groups supply nerve and muscle
elements and stinging cells. These are tissue elements, which among
the higher animals arise almost exclusively out of and in the ectoderm.

The close connection of the body epithelium of the Cnidaria with
the nervous and muscular systems is characteristic. The cell-elements
of these systems often take part with the other ectoderm cells in limit-
ing the outer surface of the body; or they lie wedged in between
them somewhat below the surface. Finally, we find them in many
Cnidaria close under the body epithelium, but often still outside of the
supporting membrane which divides the ectoderm from the endoderm.

The body epithelium is either naked, or may be covered with cilia
or flagella over greater or smaller ex-
panses. The swimming or rowing
plates, which are arranged in eight
meridional rows in the Ctenophora, arise
out of cilia cemented together.

In the body epithelium of all
Cnidaria, except the Ctenophora, the
stinging cells are found as a very
characteristic element (Fig. 69, a, b).
These stinging cells contain a stinging-
capsule, with a spirally coiled filament,
often bearing barbed hooks ; when the
skin is irritated this filament is evagin-
ated and shot forward, and has the
effect of a sting.

The stinging capsules or nematocysts are
microscopic adhesive organs, and are to the
Cnidaria at the same time weapons of defence
and of offence. They are particularly numerous

in exposed parts of the body, and in organs

droid, Cordilophora, after F. E. Schulze.
which are used in seizing prey-round the c> Seizing or adhesive cells Qf a cteno.

louth, on the tentacles, at the margin of the phore, after Chun,
disc in the Medusae. On the tentacles of

Hydro- and Scypko-medusce, and especially on the capturing filaments of the
Siphonophora, they are assembled in masses, and form "stinging knobs " or " stinging
batteries."

VOL. I r,

FIG. 69.— a, li, Stinging cells of a Hy-

82 COMPARATIVE ANATOMY CHAP.

In the body epithelium of the Ctenophora we find peculiar adhesive
cells (Fig. 69, c) with uneven and sticky surfaces. Their bases are
prolonged into spirally coiled contractile filaments.

Besides the nerve cells, the sensory, muscle, stinging, flagellate,
ciliated, and covering cells, various glandular cells and pigment cells
are found in the body epithelium of many Cnidaria, and are especially
numerous in the Ctenophom.

III. The Gastro-eanal System.

The gastro-canal system in its general arrangement has already
been mentioned in the general review. It is the most characteristic
system of organs of the Cnidaria, and in some groups reaches a very
high degree of complication. This complication stands in direct rela-
tion to the complication of other portions of the body, which fact is at
once comprehensible when we learn that in the Cnidaria the gastro-
canal system undertakes not only the digestion, but also the circulation.
The more massive the body and the more numerous and complicated
its organs, the more necessary is it that the nutrition of these organs
should be provided for by gastro-canals or vessels.

From such a standpoint the complicated arrangement of the gastro-canal system
of the Medusae and Ctenophora is at once comprehensible. In the Medusce the
margin of the umbrella is distinguished by the possession of numerous organs
(tentacles, velum, auditory vesicles, stinging cells, nerve ring, eye spots, and sensory
bodies). The subumbrellar side is strongly muscular, while the exumbrella is devoid
of organs. The jelly being so largely developed as a passive organ for motion and
support, the special gastro-canals (radial vessels) must run near the subumbrella to
convey food to the organs on the margin of the disc. This purpose is also served by
the circumferential canal, into which they enter. The relations of the gastro-canals
to the other organs of the body are just as clear in the case of those Ctenophora
which have a massive gelatinous tissue between the ectoderm and endoderm. The
most important organs of the body, apart from the sexual organs, are in this group :
the sensory body of the aboral pole, the 8 ribs, and the 2 tentacles ; answering to
these, we find an aboral vessel, also 8 vessels running from the stomach to the ribs,
and which enter 8 meridional vessels, and, further, 2 vessels which run to the base
of the tentacles.

According to the ontogenetic origin of the gastro-canal system, two
principal types can be distinguished in the Cnidaria. In one type,
which is found among the Hydrozoa, the whole gastro-canal system
rises out of the endodermal enteron of the larva. In the second type,
which is characteristic of the Scyphozoa and Ctenophora, the gastro-
canal system consists of an ectodermal and an endodermal portion.
The former, which we call stomodaeum in embryos, larvae, and
generally in young transition stages, and oesophagus in adult animals,
arises at the oral pole by a depression of the ectoderm into the body.
This oesophagus is represented in the Scyphomedusce by the inner
lining of the oral or gastric peduncle as far as the point of insertion of

ii CNIDARIA— GASTRO-CANAL SYSTEM 83

the gastral filaments, in Corals by the tube which leads into the
gastric cavity, and in the Ctenophora by the beginning of the gastro-canal
system, hitherto unsuitably named the stomach. The mouth of the
Hydromeduscv thus leads direct into the endodermal gastric cavity ;
whereas the mouth of the Scyphomedusce, Anthozoa, and Ctenophora leads
first into the ectodermal oesophagus, and from that through the enteric
aperture, which corresponds with the mouth of the Hydromedusce, into
the endodermal gastro-canal system.

In all animals, from the Ccelenterata upwards, there is an ectodermal oesophagus.

The endodermal gastric cavity is met with in its simplest form in
the Hydroida as a simple pouch adhering closely to the inner surface
of the ectoderm, from which hollow or solid processes extend into the
axes of the tentacles. In the long, or principal axis of the body,
thickenings of the gastric wall, gastric ridges, which are occasionally
4 in number, regularly arranged crosswise, are commonly found.

We have already described the form of the gastric cavity of the
Scypliula in the introduction ; and we also there described the general
arrangement of the gastro-canal system in the Anthozoa. We add
here that the septa which separate the gastric pouches from each
other round the oesophagus are occasionally broken through near the
oral disc by an aperture, so that a sort of circumferential canal arises.
The hollow processes of the gastro-canal system in the tentacle axes
sometimes penetrate to the exterior at the tips of the tentacles through
pores. The free edges of the septa, which are turned towards the
principal axis of the Coral individual, are thickened, and are prolonged
as mesenterial filaments, freely projecting into the gastric cavity.
The epithelium of these filaments contains numerous gland cells, and
sometimes stinging cells as well. Some of these filaments, the so-
called aeontia, are particularly long and vermiform, and can be
quickly shot out of the body, either through the mouth, or through
special pores in the body wall. These are found in the Actinia.

How the gastro-canal system of the Craspedote Medusae (Hydromedusce)
can be traced back to the gastric cavity of the Hydrula form, and the
gastro-canal system of the Acraspeda (Scyphomedusce) to the gastric
cavity of the Scyphula form, has already been shown in the general
review.

The radial canals in the Craspedote Medusce are comparatively
seldom limited to the number 4. In most forms their number is
greater, and in a few (^Eguortdce) they are very numerous (over 100)
and run radially from the central stomach to the margin of the disc,
and here enter the circumferential canal. There are also in a few
Craspedota, near and between the radial canals, centripetal canals,
which start from the circumferential canal and run a certain distance
towards, but do not reach, the central stomach. The radial canals
may be branched, and these branches may either end blindly or enter
the circumferential canal. The central stomach may be divided, the

84 COMPARATIVE ANATOMY CHAP.

divisions being superimposed one on another in the principal axis.
The lowest division is the opal stomach, continued in the oral or
gastric peduncle, which hangs down from the middle of the sub-
umbrella. The gastric peduncle, at whose free end lies the mouth, may
be very variously developed, from a short wide tube to a long tubular
structure protruding far beyond the subumbrellar cavity. The mouth
is either simple, square, or cross-shaped, or produced into 4 points
or lobes, and sometimes supplied with oral tentacles, or with variously
shaped papillae. The edge of the mouth is generally well armed with
nemato-cysts.

The gastro-canal system of the Acraspeda or Scyphomedusce
(Fig. 67, p. 77 ; Fig. 70) shows, in the arrangement of its single
sections, a still more varied structure than that of the Craspedota. In
some groups it is wonderfully complicated, and sometimes shows great
similarity with the gastro-canal system of certain Craspedota ; in such
cases, when we further think of the great similarity in body form, it is
difficult to believe that the Acraspeda and Craspedota are two sharply
divided branches of the Cnidaria.

Let us first consider the oral tube or oral peduncle, which, unlike
that of the Craspedota, contains the ectodermal oasophagus. The oral
tube is usually short, and has either a simple square or cross-shaped
aperture, or, as in most large Acraspeda, is produced into 4 long
strong oral arms. These 4 perradial oral arms become, by means
of a bisection so deep as to reach their bases, the 8 oral arms of the
PJiizostomce, which are distinguished by the following peculiar organisa-
tion.

Each oral arm becomes deeply furrowed on the side turned to the
chief axis, forming a channel in the longitudinal direction. This
inner channel corresponds with an externally projecting mid-rib.
The channel becomes deeper, and the curled edges of the oral arm
which border it unite over it and grow together, so that it now
becomes a closed canal. Such concrescence is completed along the
whole length of the oral arm to its base, and also spreads to the edges
of the oral aperture lying at the base of the arms, and the canal thus
becomes completely closed. The concrescence of the curled edges of
the arm, however, takes place in such a way that numerous small open
canals remain (suctorial mouths) (Fig. 70 D, sin) • these lead from the
exterior into the central canal of the arm. This again leads into the
closed oBsophagus. In all Rliizostomce the original oral aperture is thus
closed, and is replaced by the numerous suctorial mouths on the hollow
oral arms.

In the endodermal gastro-canal system, here as among the
Craspedota, we can distinguish a central or main intestine from the
peripheral intestine (Kranzdarm). The main intestine can separate
into two divisions, one lying above the other, the lower of which
•always communicates with the peripheral intestine.

On the wall of the main intestine of all Acraspeda (as opposed to

CNIDARIA— GASTIiO- CANAL SYSTEM

85

the Craspedota), there are 4 interradial or 8 adradial mobile gastral
filaments, or, usually, tufts (phacelli) of gastral filaments, whose bases
define the boundary between the ossophagus and the endodermal main
intestine. The peripheral intestine is very variously developed. In

FIG. 70.— Cannorhiza connexa, after Haeckel. A, Seen from the subumbrella ; B, from the
side ; C, from the subumbrella, after removal of the oral arms and buccal stomach by cutting
through the oral pillars. D, Section in the direction of the principal axis and the line abc in Fig. A.
ci-Tj, Near an interradius ; &-c, perradius ; og, subgenital ostia ; ma, oral arm ; ps, perradial, is, inter-
radial sensory body ; g, gonades ; mp, ap, oral pillars, arm pillars ; pfk, pillar canals ; pk, perradial ;
ik, interradial gastro-canal ; 6-6, principal axis ; sgp, subgenital portions ; bm, buccal stomach ; ak,
arm canal ; sm, suctorial mouths ; gc, central stomach.

the lower Acraspeda, which also remind us of the Scyphula in their
attachment by an aboral stalk (Lucernaria) and their cup-like shape,
the peripheral intestine consists of 4 wide pouches divided by narrow
septa, which communicate with the chief intestine and also open into

86 COMPARATIVE ANATOMY CHAP.

a circumferential canal at the margin of the disc by means of breaches
through these septa. If the septa are reduced to 4 small points of
concrescence between the exumbrellar and subumbrellar walls of the
peripheral intestine, directly on the circumference of the main intestine,
the 4 pouches coalesce to form a spacious circular sinus, which com-
mands the whole edge of the disc, and communicates with the principal
gastric cavity between the points of concrescence. In the higher
Acraspeda this circumferential sinus becomes divided, by the partial
concrescence of its exumbrellar and subumbrellar walls, into 8, 16, 32,
or more radial chambers or radial canals, which in many forms, by
anastomosing or branching, form a very ornamental net- work of canals
running towards the edge of the disc (Fig. 70, 0).

Excretory pores of the gastro-eanal system. — In various Medusw,
in Acmspeda as well as in Craspedota, small apertures have been observed
at the margin of the disc ; these often lie on the points of papillae, and
by means of them communication between the peripheral gastro-canal
system and the outer world is established.

The gastro-eanal system of the Ctenophora (Fig. 68, p. 79), in
its general arrangement, has already been delineated. We add here
that the meridional vessels in lobate Ctenophora, and also in the Cestidce
and Beroidce, communicate with each other and with the cesophageal
vessels at the oral portion of the body, and that in the Beroidce they
also send out numerous branching and anastomosing processes, some
of which enter the jelly, or join to make a peripheral net-work. The
nourishment of the often much -developed oral lobes of the lobate
GtenopJwra is provided for by the meridional vessels, which traverse the
oral lobes in various arabesque-like patterns.

Histological. — Each epithelial cell of the gastro-canal system very commonly
carries one single flagellum ; these cells are thus flagellate cells. Among the epithelial
cells there are gland cells, stinging cells, cells with various contents as products of
metabolism, epithelial muscle cells, etc. Veiy often the gastric epithelial cells send
out amceboid or pseudopodia-like processes on that side of them which is turned to
the lumen of the gastro-canal system, and by the help of these they take into their
cell bodies small particles of food in the manner of the Rhizopoda (intracellular
ingestion of food).

IV. Musculature.

In the Hydroida and Siphonoplivra we find, in the first place, a
system of longitudinal fibres which run, buried in the epithelium, from
the oral to the aboral pole, and in the tentacles. These fibres, which
correspond to the processes of the ectodermal neuro- muscular or
epithelial muscular cells, serve for contracting the body and the
tentacles. In these forms again, and especially in the Siplionopliora,
there is a system of circular fibres which run under the endodermal
epithelium as processes of the endodermal epithelial muscular cells.
By the contraction of these fibres the body and the tentacles are

n CNIDARIA— MUSCULATURE 87

extended. In the medusoid swimming bells of the Siphonophora, just
as in the Craspedote Medusce, a layer of striated ectodermal circular
muscle fibres is developed in the subumbrella and in the velum.

We find the two systems of muscles in the Medusce also, the longi-
tudinal and the circular, though the latter is here ectodermal. The
longitudinal muscle fibres are generally smooth, the circular muscle
fibres usually striated. The muscle fibres are mostly processes from
the epithelial muscle cells ; but there are also muscle bands and strands
which quite detach themselves from the epithelium, and run in the
gelatinous connective tissue as independent mesodermal muscles. The
exumbrella is poor in muscles, sometimes even having none at all. In
the remainder of the body, the longitudinal as well as the circular
musculature falls into three systems : (1) into a system spreading over
the gastric or oral peduncle ; (2) a system spreading over the sub-
umbrella from the base of the oral peduncle to the margin of the disc ;
(3) a system developed at the margin of the disc itself (musculature
of the tentacles, the velum, and the velarium). In correspondence
with this, the three systems of the longitudinal musculature are as
follows : —

A. The longitudinal musculature of the oral peduncle (serving for
its contraction and retraction).

B. The radial muscles, which run radially from the base of the oral
peduncle towards the margin of the disc.

C. The longitudinal muscles of the tentacles and marginal lobes.

The three systems of the circular musculature are as follows : —

A. The circular musculature of the oral peduncle.

B. The circular musculature of the subumbrella, developed in the
Craspedota over its whole extent, but in the Acraspeda generally form-
ing a narrower but very strong peripheral muscle (Fig. 67, m, p. 77)
near the circumference of the subumbrella.

C. The strong circular musculature of the Craspedote velum, and
the circular musculature of the Acraspede marginal lobes and velarium.

In Scyphostoma and many lower Acraspeda, especially in attached
forms, e.g. Lucernaria, there are 4 (seldom 8) interradial septal
or funnel muscles (peduncular muscles) which, starting from near the
oral disc, or that portion of the subumbrella which borders on the oral
peduncle, run through the body as far as the aboral attached apex of
the exumbrella. They lie in the 4 septa which separate the 4
gastric pouches on their axial sides, and then proceed upwards in the
prolongations of the septa, i.e. in the 4 gastric ridges or ta3nioles.
According to recent research, they arise in the ectodermal cells of a
solid prolongation of the 4 interradial septal funnels, which grow
towards the aboral pole ; we shall speak of these septal funnels later.

The endodermal musculature of the Corals, in contradistinction to
that of all the other Cnidaria, is at least as much if not more developed

88 COMPARATIVE ANATOMY CHAP.

than the ectodermal. The musculature is most highly developed in
those Actinia which have no skeleton. It 'shows in its arrangement
much similarity with that of the Scyphomedusce, e.g. the Lucenmria.
We have an ectodermal longitudinal muscle system and an endoder-
mal circular muscle system (leaving out of sight for the present the
fact that single portions of the musculature detach themselves from
the epithelial matrix and become mesodermal).

I. The ectodermal longitudinal muscular system forms (a) the
longitudinal muscles of the tentacles and (b) the radial muscles of the
oral disc. This system is wanting in the oesophagus, in the outer wall
of the body, and in the pedal disc. Only in some forms, which possess
no septal longitudinal muscles (Cerianthus), there are in the body wall
strong ectodermal longitudinal muscles which serve as retractors of the
body.

II. The endodermal circular muscular system extends all over the
surfaces of the body, and forms a layer of circular fibres in the body
wall, an inner circular muscular layer in the tentacles, a layer of con-
centric fibres on the oral disc, and a circular muscular layer round the
oesophagus.

The musculature of the septa in Corals deserves special atten-
tion ; it consists of a system of longitudinal and of a system of
transverse muscles. The longitudinal muscles run along the septa from
the pedal to the oral disc, and serve as retractors of the body. They
are mostly so strongly developed that they form longitudinal thicken-
ings on each septum in the space between it and the next septum (Fig.
66, p. 76). The transverse musculature is less strongly developed. It
is attached on one side to the body wall, on the other to the pedal disc,
the oral disc, and the oesophageal tube. The transverse muscles lie on
one surface of each septum, the longitudinal on the other side. Their
relative positions on the different septa varies in different divisions,
and is always very characteristic. There is generally only one plane,
which can be made to divide the body in such a way that the arrange-
ment of the muscles on the septa on each side of it is exactly similar.
This median plane runs in the principal axis in the direction of the
flattened oesophagus or the slit-like oral aperture.

In most Actinia with numerous septa of various ages and sizes,
septa of equal sizes always go in pairs. The longitudinal muscular
thickenings of such a pair of septa are turned towards each other.
The longitudinal muscles are therefore turned towards the space
between the two septa of such a pair — the so-called intraseptal space —
and the transverse muscle layers are turned towards the space between
this pair and the next on each side — the interseptal spaces.

In those Alcyonaria which have 8 partition walls, the muscular
thickenings of the 4 septa which lie on one side of the median plane
are all directed towards one side (Fig. 66).

There are, besides these, other types of muscle and septal arrange-
ment in the Corals.

ii CNIDARIA— MUSCULATURE 89

The longitudinal muscles of the septa show a great analogy with
the septal or peduncle muscles of the lower Acraspeda. The former
appear, however, according to present knowledge, to be endodermal
muscles.

The muscular elements of Corals are either epithelial muscle cells
(endodermal musculature), or sub-epithelial muscle cells (ectodermal
musculature), or mesodermal muscles (in some species at different parts
of the body).

The following applies to the musculature of the Medusae, Siphonophora, and Corals.
If it is much developed in one part, the muscle lamella lays itself, for the purpose of
superficial increase, in parallel folds like the leaves of a book (Fig. 71, C]. These folds
may again fold themselves in a more or less complicated manner, so as to have a
feathered appearance in transverse section.

A c

•mt,

FIG. 71.— Diagrammatic representation of the various arrangements of the ectodermal
Cnidarian musculature in transverse sections through the body wall, e, Ectoderm ; en, endo-
derm ; m, muscle lamella ; mz, cell bodies of the muscle fibres ; sm, supporting membrane, jelly.

The mesodermal supporting substance or supporting membrane takes part in the
folding of the contiguous muscle lamellae by itself running within the folds in the
form of lamellae.

When muscle folds completely detach themselves from their matrix, the epi-
thelium, and when the free edges of each fold coalesce, there arise out of these sub-
epithelial muscle folds mesodermal muscle tubes (D}t which are then surrounded on
all sides by the supporting substance.

In the musculature of the Ctenophora, we in the first place dis-
tinguish an ectodermal and a mesodermal portion, between which,
however, no very sharp boundary can be made. The ectodermal muscu-
lature consists of elongated sub-epithelial fibres on the boundary be-
tween the epithelium and the jelly; these may have very various
courses. Sometimes we can more or less clearly distinguish a system
of longitudinal from a system of circular fibres. The ectodermal
musculature is also continued on the oasophagus. The mesodermal
musculature, which lies in the jelly, is more strongly developed than
the ectodermal musculature ; its fibres, which have an isolated course
and are elegantly branched at each end (Fig. 47, g, p. 47), run in
various directions, though usually radially, being stretched between
the different parts of the gastro-canal system and the outer integument.
The contraction of the tentacles and their lateral filaments is brought

90 COMPARATIVE ANATOMY CHAP.

about by strands of longitudinal muscle fibrillse which run in their
solid axes and which may be partially striated. Longitudinal and
circular muscle fibres have also been observed in the walls of the gastro-
canals.

The Velum of the Cmspedote Medusa (Fig. 64, B, v, p. 73 ; Fig. 65,
p. 74 ; Fig. 72, #, p. 95) is a thin membrane which projects from the
margin of the disc like a diaphragm into the subumbrellar cavity.
The gastro-canal system is never continued into this membrane, which
consists of the following portions : —

1. A continuation of the epithelium of the exumbrella.

2. A continuation of the epithelium of the subumbrella. These two
epithelia coalesce at the free inner edge of the velum. Under the former
lies a thin supporting lamella, the continuation of the disc jelly ; under
the latter, a layer of ectodermal circular muscle fibres (m), a continuation
of the circular musculature of the subumbrella.

V. Tentacles of the Cnidaria, Marginal Lobes of the
Seyphomedusae.

All Cnidaria (with the exception of the Rhizostomce among the
Scyphomedusce and the Amalthceidce among the Craspedota) possess
tentacles arranged in a circle round the mouth, at a greater or less
distance from it. These tentacles are evaginations of the body wall,
into which (with the perhaps only apparent exception of the Ctenophora)
hollow or solid processes of the endodermal gastro-canal system pene-
trate. The tentacles are pre-eminently organs for catching food, and
at the same time sensory organs of touch. We shall see further on
that some of them are partially transformed in the Craspedote and
Acraspede Medusce into specific sensory organs.

The structure of the tentacles, their number, their division and
arrangement on the body, and their relation to the gastro-canal system,
offer in the various divisions many modifications of great importance
in classification.

Form of the tentacle. — The tentacles are, speaking generally,
cylindrical filaments. In the Hydroids they are usually simply fila-
mentous, less frequently knobbed at the free end, and still less fre-
quently branched (Cladocoryne). Among the Craspedote Medusce also
we generally meet with filamentous tentacles. The family of the
Cladonemidce alone is distinguished by tentacles which are dichoto-
mously branched, or feathered on one side (provided with collateral
filaments), and often knobbed. The tentacles of the Siplionopliora and
the Ctenophora are also feathered on one side. The tentacles of the
Acraspeda are simple. Among the Antlwzoa, the Alcyonaria possess
tentacles feathered in two rows, but all other divisions have simple
filamentous or vermiform tentacles.

Number and arrangement of the tentacles. — Among the Hydroida

ii CNIDARIA— TENTACLES, MARGINAL LOBES 91

the tentacles, in varying number, are either arranged in a circle at a
short distance from the mouth (e.g. Hydra), or in two circles (e.g.
Tubularia), or they are dispersed over the body of the individual,
though not on its stalk (e.g. Coryne). In the Scyphostoma the tentacles
(about 24 in number) are inserted at the edge of the oral disc of
the cup-shaped body. In the Craspedote and Acraspede Medusae the
tentacles are invariably found on the margin of the disc. In the
simplest case we find in the former 4 perradial, and in the latter
4 perradial and 4 interradial tentacles. In most Medusae the
number of tentacles increases in a regular manner — secondary, tertiary,
etc., being inserted between the primary. Only in a few Craspedota
the number is reduced to two tentacles, opposite one another, and less
frequently to one single tentacle (in the sub- family of the Euphysidce
among the Codonidce, and in the tentacle-bearing "persons" of one
principal division of the Siphonophora (the Siphonantha). Complete loss
of all tentacles is a distinguishing mark of the Amalthceidce among
Craspedota and the Ehizostomce among Acraspeda. In a few Medusse,
especially among the Narcomedusce, the points of insertion of the
tentacles move up from the edge of the disc a greater or less distance
on to the exumbrella.

In the Corals the number of tentacles inserted at the circumference
of the oral aperture represents, in a general way, the number of gastric
pouches separated by septa. Each tentacle lies above a gastric pouch,
which is produced into it in the form of an axial canal. In the Alcyo-
naria we have 8, in the HexaGorallia 6 or Qn, in the Tetracorallia 4 or
4% such corresponding tentacles.

The majority of the CtenopJiora (Tentaculata) possess two solid
tentacles or retractile filaments, feathered on one side, lying in the
lateral plane ; these can be withdrawn into special sacs or sheaths, and
arise in the neighbourhood of the aboral pole ; later, however, they
generally move towards the oral pole.

Structure of the tentacles. — The tentacles of the Hydromedusce
and Scyphomedusce consist: (1) of an ectodermal epithelium, generally
with stinging knobs or batteries ; (2) of a layer of ectodermal longi-
tudinal muscle fibres lying directly under this ; (3) of a structureless
elastic supporting membrane ; and (4) of an axis of endodermal cells.
This axis is hollow in most forms, and the cavity is in open communi-
cation with the gastro-canal system of the body ; or it is solid, and
then consists chiefly of a single row of disc-like cells, which are super-
imposed like the coins in a roll of sovereigns (e.g. in Obelia, in many
Trachomedusce, Narcomedusce, and the related Hydroids, also in Stauro-
medusce and Cannostomce). This axial pillar must serve as an elastic
organ of support. The hollow tentacles are mostly far more mobile
and more strongly contractile than the solid. The tentacles of Corals
are also hollow ; their structure differs considerably from that of the
Hydro- and Scyplw-medusce. The ectoderm and endoderm in Corals are
separated by a supporting substance which contains cells of connective

92 COMPARATIVE ANATOMY CHAP.

tissue. Under the outer epithelium lies an ectodermal muscle layer,
and under the inner epithelium a similar endodermal layer. In some
cases muscles may also run within the supporting substance.

The tentacles of the Ctenophora are solid. They are generally
provided with seizing or adhesive cells. Their axes are usually occupied
by strongly developed longitudinal muscle-fibres. These fibres arise,
as it appears, at an early stage out of special " mesodermal " elements,
i.e. a group of cells which sever themselves from the primitive endo-
derm of the young gastrula larva. We ought therefore, perhaps, to
compare the solid axis of the Ctenoplwmn tentacle with the solid endo-
dermal axis of the tentacles of many Medusce.

The marginal lobes of the Seyphomedusse (Fig. 67, rl, p. 77 ;
Fig. 70). The marginal lobes of the Scyphomedusce or Acraspeda are
just as characteristic of them as is the velum of the Hydromedusce or
Craspedota. As a real velum is wanting in all Scyphomedusce, so are
marginal lobes wanting in all Hydromedusce. Like the tentacles,
together with which they are found, the marginal lobes are processes
of the body wall at the edge of the umbrella, into which the prolonga-
tions of the gastro-canal system extend. Unlike the tentacles, they
are broad and flat, serving as rowing organs, with muscles on the con-
cave subumbrellar side. In the simplest cases there are 8 adradial
marginal lobes, mostly, however, there are 16 subradial lobes, and
their numbers are often still further increased. In the Cubomedusce
and in many Rhizostomce the lobes grow together to form a circular
rim of varying width, the so-called velarium which, however, can
always be easily distinguished from the true velum of the Craspedota
by its supply of gastro- canals. The clusters of tentacles of the
Lucernaria are secondary outgrowths of the marginal lobes, between
which rudiments of the primary tentacles can sometimes be found.

VI. The Nervous System.

This, which appears in the Cnidaria for the first time in the
animal kingdom as an independent system, is marked by a rather
diffuse arrangement (want of definite centralisation), and by the close
relation which it bears to the body epithelium during the whole life.
In the Hydra the outwardly directed cell body of the so-called neuro-
muscular cells probably plays the part of an undifferentiated sensory
nerve cell. But as early as the Craspedote Medusce we find an inde-
pendently developed nervous system, close under the epithelium, outside
the supporting membrane or jelly; it forms a plexus of bipolar and
multipolar ganglion cells with connecting fibrillse. This plexus is con-
nected by fibrillse on the one hand with the epithelial sensory cells
(tactile, auditory, visual, olfactory cells), on the other hand with
muscle fibres. In correspondence with this, the nervous tissue is
particularly strongly developed on the margin of the umbrella, which

ii ( CNIDARIA— NERVOUS SYSTEM 93

is so rich in sensory and locomotor organs. At this part in the
Craspedota a central apparatus, in the form of a double nerve ring, is
developed (Fig. 72, A and B, nrl} nr2) p. 95). One of these nerve rings,
the upper one, is placed on the exumbrellar, the other on the subum-
brellar margin, both close to the base of insertion of the velum. The
first innervates chiefly the sensory organs of the umbrella margin, the
latter the musculature of the velum ; the two are connected by fibrillse
running through the supporting membrane. A peripheral nervous
system is present in the form of a plexus of ganglion cells and fibres
connected with the lower nerve ring, especially on the subumbrella ;
it supplies the musculature of this part with nerves. The aboral
exumbrella, not only in the Craspedota, but also in the Acraspeda, is
devoid of a nervous system, as also of sensory organs and musculature ;
this is explicable by the ontogenetic and phylogenetic derivation of
these animals from a form attached by the aboral pole.

In the Siphonophora also a plexus of ganglion cells at various parts
of the body has been demonstrated.

In the Scyphomedusce or Acraspeda also a plexus of ganglion cells
is developed in the subumbrella. It is connected with considerable
central accumulations of nerve tissue on the subumbrellar margin.
These are developed in the Cubomedusce as 8 (4 perradial and 4 inter-
radial) ganglia, connected together by a circular nerve, from which
sensory nerves proceed to the rhopalia (sensory bodies) and to the
tentacles, and motor nerves to the musculature. In the Discomedusce
there are also 8 ganglia at the bases of the 8 sensory bodies, but here
there is no connecting nerve ring.

In the Corals a plexus of ganglion cells lying deep in the body
epithelium has also been demonstrated. This isl particularly strongly
developed on the oral disc, and at the base of the tentacles.

In the nervous system of the Ctenophora we can, according to recent
research, which still needs extension, distinguish the following parts :
(1) a diffuse, ectodermal, ganglionic plexus under the body epithelium,
which is spread over the whole surface of the body, and may be con-
tinued on to the wall of the oesophagus ; (2) fine, branching, nucleated
strands, which penetrate the jelly, and are connected with the muscles
of the jelly by lateral branches — a connection between these strands
and the ectoderm is not yet proved ; (3) 8 meridional nerve strands,
running under the 8 ribs, and ciliated bands (see description of
sensory body), — the rib-nerves. It is a remarkable fact that it has till
now been impossible to prove a connection between these three parts
of the conjectural nervous system and the sensory body at the ab-
oral pole.

A fact which deserves special mention is that, in the Medusce
and Siphonophora, and especially in the Actinia, a more or less
extended plexus of ganglion cells may lie under the endodermal
epithelium.

COMPARATIVE ANATOMY

CHAP.

VII. The Sensory Organs.

We meet with special sensory organs for the first time in the
animal kingdom among the Cnidaria. The development and distribu-
tion of these organs is directly related to the manner of life. In
attached forms (Hydroids, Corals) there are only organs of touch (the
tentacles) ; in free-swimming forms (Acraspeda, Craspedota, Siphonophora,
Ctenophora} organs of smell, hearing, and sight may be added.

• The organs of touch are primarily the tentacles of which we have
already spoken. The sense of touch is due to special tactile eel
belonging to the body epithelium ; these are provided with a projecting
tactile hair, which is either flexible, mobile, or stiff. The basal ends
of such cells are continued into nervous processes which are connected
with the nervous system. Tactile cells are to be found, not only in
the tentacles of the Cnidaria, but in great numbers on the margin of
the disc of the Medusae, and especially in the Ctenophora, scattered over
the whole free surface of the body.

As organs of smell, or perhaps rather taste, we have :

1. Small elub-shaped papillae, which in certain Leptomedusce are
found generally in great numbers at the edge of the umbrella between
the tentacles, being attached to the margin by thin stalks. They
contain a narrow blind canal, lined with thick cylindrical endodermal
epithelium, which comes from the circumferential canal.

2. Pit-like depressions on the sensory bodies or rhopalia of the
Acraspeda, lined with a sensory epithelium which is much folded and
provided with long flagellate hairs.

Auditory organs (perhaps also organs for regulating the position
of the body in the water) are found in Craspedote and Acraspede Medusce
and in the Ctenophora.

We can distinguish three types of auditory organs : (1) auditory
vesicles, or marginal vesicles with ectodermal otoliths ; (2) tentaeulo-
eysts, or auditory tentacles ; and (3) the so-called sensory body of the
Ctenophora.

I. The auditory or marginal vesicles are found in the division of
the Leptomedusce (Vesiculatce). These are, in the simplest cases, open
pit-like depressions of the subumbrellar epithelium near the base of
insertion of the velum. Within these auditory pits are one or more
otoliths, which have come from ectodermal cells, while the cells which
form the base of the pit bear auditory hairs, on which the otoliths
rest (e.g. Mitrocoma). Auditory vesicles rise out of these auditory pits
by the closing of the pit, which moves towards the exumbrellar side of
the base of insertion of the velum and here forms an externally
rounded protuberance (Fig. 72, A). Auditory pits and auditory vesicles
receive their nerves from the subumbrellar nerve ring. In the simplest
cases we find 8 adradial auditory vesicles, but the number is often much
greater, and mounts up to many hundreds.

CNIDARIA— SENSORY ORGANS

95

ec

nr,

II. Tentaeuloeysts OP auditory tentacles are, among the Medusce,
the most widely spread organs of
hearing. They are short transformed
tentacles in which the solid endo-
dermal axis in the Craspedota (Tracho-
and Narco-medusce, Fig. 72, B, C\ or
the peripheral distal end cells of the
hollow tentacle-canal (Acraspeda, Fig.
73), form one or more otoliths, which
are here, therefore, endodermal. The
ectoderm of the tentaculocysts of the
Craspedota and often also a sensory
cushion round the base of the tenta-
culocyst carry long stiff auditory
hairs. In many TracliomeduscK the
sensory cushion rises round the base
of the tentaculocyst into a circular
wall (Fig. 72, C\ which may even
completely close over the tentaculo-
cyst. Thus closed vesicular auditory
organs again arise, but these have
quite a different origin and morpho-
logical signification from the true
marginal vesicles of the Leptomedusce.

Between the inner wall of the
auditory vesicle and the tentaculocyst
which it encloses, the auditory hairs
are stretched like chords. The tenta-
euloeysts always receive their nerves
from the exumbrellar nerve ring. In
the simplest cases we find 4 interradial
tentaculocysts, but their number is
often considerably greater.

The sensory bodies or rhopalia
of the Aeraspeda (Fig. 73) are com-
pound sensory organs, of which the
auditory organ is the principal factor.
These are, at the same time, meta-
morphosed tentacles into which a
hollow process of the gastro- canal
system penetrates. The endodermal
cells at the peripheral blind end of
these processes produce an otolith or
a mass of otoliths. The outer epi-
thelium of the auditory body carries
the auditory hairs. In it, also, on
one side lie the eye or eyes ; close to it also lie the olfactory pits. The

FIG. 72. — A. Auditory vesicles of
-ffiquorea Forskalea. B, Tentaculocysts
of Cunina lativentris. C, Tentaculocysts
of Rhopalonema velatum. A and B, Trans-
verse sections of the margin of the disc ; C,
section of the margin of the disc, after
Hertwig. ec, Ectoderm ; er, endoderm of
the circumferential canal ; nri, upper ; nr%,
lower nerve ring ; r, circumferential canal ;
sm, supporting membrane; o, otolith; hh,
auditory hairs ; hg, auditory cells ; Me, tenta-
culocysts ; bl, auditory vesicles; g, jelly;
m, muscle lamellae ; en, endoderm cells of
the tentaculocysts. In A and B the velum
is bent centrifugally outwards.

96

COMPARATIVE ANATOMY

CHAP.

whole epithelium round this compound sensory organ, which is pro-
tected in special indentations in the disc margin by lobes, is a sensory
epithelium with a thick nerve plexus. There are either 4 or 8, less
frequently 12, 16, or even more rhopalia on the margin of the
Acraspede umbrella.

III. The sensory body at the aboral pole, so characteristic of the
Ctenophom, is a compound sensory organ of a very peculiar kind, which,
according to its structure, is an auditory organ, or rather, perhaps, an
organ for regulating the movement of the swimming plates.

The sensory body is constructed as follows. It consists at first of
a shallow pit- like depression between the 4 branches of the aboral

ee

FIG. 73.— Sensory bodies of Nausithoe, after Hertwig (optical transverse section of the margin of
the disc), sf, Sensory fold of the margin of the disc ; ec, ectoderm ; en, endoderm ; ga, gastro-
canal ; Tik, auditory body ; o, otolith ; se, sensory epithelium ; oc, eye ; I, lens ; g, jelly.

so-called funnel vesse1. The ciliated body epithelium which forms
the base of this pit tLickens considerably (Fig. 75, se). Its elements
are thread-like cells. In this " sensory cushion " are found deposits
of pigment, which perhaps represent simple organs of sight. Round
the edge of the pit there rises a membrane which unites above in the
shape of a bell, forming a sort of roof to the pit, which is thus trans-
formed into a vesicle. The membrane is composed of the long cilia
at the edge of the pit cemented together. It is broken through by
slits at 6 places, and through these the sea water can freely reach
the interior of the vesicle. Two of these slits, which are opposite
each other, belong to the median plane. The other 4 lie interradi-
ally. On the sensory cushion within the vesicle rise 4 S-shaped
radially-placed springs ; these likewise consist of fused cilia, and their
free upper ends enter a globular mass of otoliths, which they support.
From the 4 springs 4 rows of cilia run out through the 4

CNIDARIA— SENSORY ORGANS

97

interradial slits in the bell -shaped roof. They divide into 8
•adradial ciliated bands, which run along the aboral surface of the
Ctenophoran body towards the upper end of the rows of swimming
plates. Closely bordering on the sensory body are two ciliated
epithelial bands, the so-called pole plates (Fig. 74, pp), whose edge is
thickened. They lie in the median plane of the body ; at the point
where they come in contact with the sensory body are found the two
above-mentioned median slits through the bell covering that body.

Organs of sight. — Simple organs of sight occur principally as
pigment spots in such Leptomedusce as possess no marginal vesicles

FIG. 74.— Aboral pole of the body of
Callianira, after Hertwig. ws, Ciliated
bands ; /, springs carrying the mass of
otoliths (o) ; sk, sensory body ; pp, pole
plates ; to, openings of the 4 branches of
the aboral vessel or gastro-canal.

FIG. 75.— Halves of two sections through the
sensory body of Callianira, after Hertwig. A
passes through one of the 4 springs (/) which
carry the otolith mass (o) ; d, roof of the sensory
pit ; se, sensory epithelium of the sensory body ;
p, pigment.

fpcellata), and in Anthomedusce. They consist of pigment cells and
sensory cells, to which, in some cases, a cuticular thickening is added
as lens. Such a lens is less often wanting in tike visual organs of the
Acraspeda (Fig. 73, oc, I), whose structure is occasionally considerably
complicated ; they form part of the marginal bodies. In the Cubo-
medusce a vitreous body and a retina are developed between the lens
and the pigment cup. In Charylde 5 there are compound eyes : each
marginal body bears 2 large unpaired and 4 small paired eyes.
Eye spots, sometimes with lenses, are found at the tips of the feelers
in some Siphonophora. In the Ctenophora there are, as has already
been mentioned, pigment spots in the sensory cushion of the sensory
body.

VIII. Supporting Organs, Protective Organs, Skeleton.

The formations belonging to this category, which are very hetero-
geneous, can at once be divided into two principal groups, according to
their ectodermal or mesodermal origin.

VOL. I H

98

COMPARATIVE ANATOMY

CHAP.

1. Eetodermal supporting and protective organs. — These are
found in attached Cnidaria, and especially in those which form animal
stocks by asexual reproduction (incomplete fission and gemmation).
We can easily understand why such stocks, which in their natural
state imitate trees, bushes, grass, feathers, crusts, leaves, etc., need
special adaptations for holding the parts upright in the water, to
support and at the same time to protect them. We also see why
such supporting organs are of no use, or of very little use to
attached Cnidaria which do not form colonies, and why they are often
wanting, or only slightly developed in such forms. (Examples : Hydra,
the attached Scyphomedusce, and the Actinia among Corals.)

The ectodermal supporting formations are simplest in the Hydroids.
Here the body epithelium generally secretes a chitinous cuticle (peri-

Fio. 76.— Diagrammatic representation of the structure of a Stone Coral (Madreporarian),
after v. Koch. Only the lower aboral portion of the body is taken into consideration, fp, Foot-
plate ; ap, exotheca ; mp, theca ; ss, sklerosepta ; hs, sarcosepta. White parts = calcareous skeleton.
Streaked parts = ectoderm. Black parts =mesoderm. Dotted parts = endoderm.

derm) which surrounds the body like a tube. This tube surrounds
either only the stem, the branches of the stock, and the stalks of the
individuals, or, further, it widens out round the individuals into cups
into which they can be withdrawn. In the division of the Hydro-
corallia the periderm becomes calcified, and forms a framework of
many tubes, arranged in a complicated manner, and reticulately bound
together.

The calcareous skeletons of the Stone-corals (Hexacorallia, Madre-
poraria), and the horny skeletons of the Alcyonidce (Octacorallia) belong
to the order of skeletons secreted by the ectoderm.

The origin of the calcareous skeleton of the Stone-corals (Fig. 76),
and its relation to the soft parts of the body, are as follows : —

The young Coral, still devoid of skeleton, having attached itself
by the aboral end of its body, secretes from the ectoderm of its pedal

II

CNIDARIA— SKELETON 99

disc a foot-plate formed of globular calcareous grains, and thus con-
nects the ectoderm of the foot with the surface to which the body is
attached. Then from this foot-plate gradually arise, by calcareous
secretion from the ectoderm, radially arranged perpendicular ridges,
the star-ridges or sklerosepta. These are naturally covered on all
sides with ectoderm, and they raise the pedal disc, in as many folds
as there are ridges, into the gastric cavity. In the same way a cal-
careous tube, the theea (Mauerblatt), arises, partly by the coalescence
of the ends of the star-ridges, and partly perhaps also by the rising of
a circular wall out of the pedal disc ; this again raises the basal body
wall in folds into the gastric cavity, just as do the star-ridges. By
the formation of this calcareous tube the gastric cavity is divided
into a peripheral portion, lying outside the theca, and a central
portion, the two being in open communication above its free upper
edge.

In the axis of the Coral a calcareous pillar often rises from the
pedal disc and projects into the gastric cavity ; this is the eolumella.
The star-ridges may fuse with this eolumella, they may also stand
out above the before-mentioned tube as ribs. By a further calcareous
secretion from the ectoderm round the base of the body wall, the
exotheea arises ; this is lined by ectoderm, and forms an outer circular
calcareous wall of varying height above the pedal disc.

The peripheral ends of the star -ridges can also unite with the
exotheea, though of course only by breaking through the body wall ;
indeed the theca may entirely or partially coalesce with the exotheea,
displacing the intermediate soft portions.

Most Madreporaria, by incomplete fission or gemmation, form
variously -shaped Coral colonies, in each individual of which the
skeletal arrangement just described is repeated. Complications may
arise by the complete or partial fusing, or else the complete disappear-
ance of the thecse of the various individuals, etc.

The star-ridges or sklerosepta never correspond in position with
the ordinary septa or sareosepta, but on the contrary alternate with
them, so that a skleroseptum always lies between 2 sareosepta, and
a sarcoseptum between 2 sklerosepta. In consequence of this the
sklerosepta imitate the sareosepta in number and arrangement.

The skeletons of the Madreporaria are either massive and close
(M. aporosa), or they are perforated by small cavities (M. porifera). The
calcareous skeletons of the fossil Ruyosa had probably an origin similar
to that of the Hexacorallia.

The second kind of ectodermal Coral skeletons, the horn skeletons,
which are found in many Alcyonaria and in the Antipatharia, are
usually hollow axial skeletons ; they run through the bodies of these
colonial Corals, and thus take the shape of their often elegantly
branched stocks. It seems at first paradoxical that these axial skele-
tons should be ectodermal. So as to explain this fact, we shall briefly
describe the formation of the horn skeleton of Gerardia. The stocks

100 COMPARATIVE ANATOMY CHAI

of this Coral form a sort of crust over foreign bodies, preferring th<
axial skeleton of other dead Alcyonaria. The surface by which they
adhere to these bodies, and which is of course ectodermal, puts forth
externally, and thus between itself and the surface to which it adheres,
a lamella of horn, which, together with the foreign body (axial skeleton
of another Alcyonid), forms the axial skeleton of the whole stock.
"Now, however, it not infrequently happens that the Gerardia colony
tries later to spread out further than is allowed by the surface it rests
on, and then growths which bear young polyps appear on its branches,
and into these a new formation of horn enters, attached to the
original horny secretion; this new formation has a similar origin
with the first, but encloses no foreign body " (v. Koch). In the other
Alcyonaria which have a horny axial skeleton, the formation of the
skeleton is similar (Fig. 82, B, p. 107), but that part of the axial
skeleton which is attached to a foreign substance is very much reduced,
whereas the free part rising from it is considerably developed, and
forms the generally much-branched principal mass of the skeleton.
Horny axial skeletons are thus always lined with an ectodermal axial
epithelium. In the axial skeleton of the Alcyonaria , lime may be
found in larger or smaller quantities, and sometimes preponderates.
In his horny joints alternate with calcareous. The mesoderm of the
soft part of the Coral stocks which covers the axial skeleton often
contains calcareous spicules. In such cases (e.g. Gorgonia) we have j
an ectodermal horny axial skeleton and a more peripheral mesodermal
skeleton formed of calcareous spicules.

2. Mesodermal supporting organs. — The first of such organs
which we shall consider is the thin structureless membrane, which, |
throughout the whole Hydroid body, separates the ectoderm from the j
endoderm. In the Craspedote Medusae this membrane thickens into the j
more or less strongly developed structureless elastic disc jelly ; it is
retained as a thin membrane only in the tentacles, and generally also
in the oesophageal tube. In the Scyphomedusce the mesodermal sup- \
porting jelly begins to be more highly differentiated, connective tissue ;
cells appearing in it, and fibres, which are either processes of such cells
or differentiations of the intercellular substance (Fig. 36, p. 41). In
the same way we find, in the Corals, a hyaline mesodermal layer
throughout which cells are scattered. The membrane is everywhere
reduced to a thinner layer in the tentacles. In many Corals (most
Alcyonidce) the skeletal elements are found in this layer. They con-
sist of variously shaped calcareous spicules, which arise in special cells
and are found in varying numbers. In the mesoderm of AlcyoniumJ\
and in the peripheral portions (the rind) of the stock of the precious*
Corals of commerce and other forms, they are isolated. Occasionally,
however, a new calcareous substance is deposited between the calcareous
spicules, cementing these into a firm mass, and leading to the formation-;)
of axial skeletons such as that of the precious Coral of commerce.

The greatest differentiation, histologically, within the Cnidaria is

ii CNIDARIA— SUBGENITAL CAVITIES 101

shown by the jelly which functions as supporting tissue in the Cteno-
phora, containing, as it does, nerve, muscle, and connective tissue
elements. All these elements are usually represented by more or less
branched fibres.

As to the origin of the various mesodermal supporting formations, we have the
following remarks to make. The homogeneous supporting membrane of the Hydroids,
and the corresponding jelly substance of the Medusce, Corals, and Ctenophora is a
product of secretion deposited between the ectoderm and endoderm. -Whether both
layers take part in this secretion or only one, and if so which, is a question difficult
to decide. The cell elements which lie in the jelly, on the contrary, seem almost
exclusively to arise by the immigration of cells from the ectoderm. In Corals
the ectoderm soon becomes multi-laminar. The cells of the deeper layers become
mesodermal connective tissue cells by the rise between them of homogeneous sub-
stance. Many observers, therefore, consider the Coral mesoderm as only a more
deeply lying portion of the ectoderm.

IX. Funnel Cavities (Septal Funnels) ; Subgenital Cavities,
Subgenital Chamber.

These formations, met with in many Acmspeda (analogous forma-
tions are very rarely found in Craspedota) are in the lower forms
represented by 4 interradial funnel-shaped depressions of the sub-
umbrella round the oesophagus (Fig. 99, H, /, st, st', p. 130). They
project more or less far into the gastric cavity, within the septa which
separate the 4 gastric pouches. In Discomedusce they become 4 some-
what flat interradial subgenital cavities. Their roof is close to the
subumbrellar wall of the central gastric cavity in which the genital
organs develop. The membrane which separates the two cavities thus
becomes the genital membrane. In two families of the Rhizostomce,
(the Versundce and Crambessidce), the 4 sub-genital cavities unite in
the centre of the umbrella to form a spacious sub-genital chamber
(porticus subgenitalis, Fig. 70 D, sgp, p. 85), which opens outward by
four interradial apertures through the subumbrella into the umbrella
cavity. The dorsal roof of the chamber is formed by the gastro-genital
membrane, which separates it from the gastric cavity above it. The
Subgenital chamber separates the central gastric cavity from the
oesophagus. The two remain connected only by the 4 canals which
run perradially (pfk).

X. The Sexual Organs.

The Cnidaria are either sexually separate, like the Hydrozoa and the
Scyphozoa, (with a few exceptions, e.g. Hydra and a few Cladonemidce,
Cerianthus, Chrysaora), or hermaphrodite like the Ctenophora.

In the colonial Cnidaria we find male and female individuals either
in the same stock (monoecious), and this is the rule ; or on different
stocks (dioecious).

102

COMPARATIVE ANATOMY

CHAP.

Testes and ovaries are, taken as a whole, very simply constructed ;
they are vesicles or spheres, with numerous egg and sperm cells at
different stages of development.

In Hydra both sorts of sexual products lie in the deeper part of
the body epithelium. In the other colonial Hydroids they are met

with in specially shaped so-
called medusoid individuals, of
which more will be said below,
either being formed primarily
in these individuals, or reaching
such a position secondarily from
the stem.

In all Medusae the sexual
glands or gonades show, by
their position, a close relation
to the nutritive gastro- canal
system. In the Craspedota (Fig.
77), they lie in varying numbers
either on the wall of the oral
tube (Narcomedusce and Antho-\
medusce), or on the radial canals j
(LeptomeduscesLiidTrachomedusce).
Where there are 4 radial canals
there are 4 gonades, and where j
there are 8 radial canals, 8 i
gonades. With increase in the j
number of radial canals there
may also be increase in the i
number of gonades. In the j
Acraspeda 4 (less frequently 8)
globular or band -like gonades j
are usually developed ; these
are sometimes folded, or curled, -i
or clustered, and occasionally !
of considerable size ; they lie
in the subumbrellar wall of the j
gastro-canal system, sometimes nearer the circumference, at other times
nearer the central gastric cavity. In the Pelagidce and Cyanide?, the
gonades hang down as 4 gastro-genital sacs from the subumbrella into
the umbrella cavity ; in the Rhizostomw and Aurelidce, on the contrary,
they lie on the upper surface of the subgenital cavities or the sub-
genital porticus (Fig. 70, p. 85). In the Craspedota the ripe genital I
products pass directly out into the umbrella cavity by the bursting of
the gonade ; in the Acraspeda they pass inward into the cavity of the
gastro-canal system, and reach the exterior thence through the mouth, j

The sexual organs of Corals lie in the septa, near the free edges
which project into the gastral cavity.

FIG. 77. — Eucope campanulata, partly after
Haeckel. A, From the surface. B, Section in the
direction a-b-c of Fig. A. a-&,Perradius ; b-c, adradius;
t, tentacle ; sb, marginal vesicle ; g, gonades ; mr,
gastric peduncle ; r, radial canals ; v, velum ; ri,
circumferential canal; ex, exumbrella; su, sub-
umbrella; ga, jelly; tg, tentacular vessel; b-b,
principal axis.

ii CNIDARIA— STRATIFICATION OF BODY 103

In the hermaphrodite Ctenophora we meet with them on or in the
walls of the meridional vessels, in such a way that on the one wall of
the meridional vessel the male elements lie, and on the opposite wall
the female elements. These sexual glands are so arranged that in
each of the 8 regions of the body, separated by the meridians, there
are either 2 testes or 2 ovaries. The ripe sexual products fall into
the meridional vessels, and through the gastro-canal system reach the
stomach and O3sophagus and pass out through the mouth.

In the Cnidaria no special ways of transmission for the sexual
products, and no special copulatory organs, are developed.

Concerning the origin of the sexual products we may say, generally, that in very
many Hydrozoa they are developed out of the ectoderm, but in the Scyphozoa out of
the endoderm. Observers are not yet agreed about the origin of the sexual products
in the Ctenophora.

Since in the one form, the Hydroids, the sexual products come from the ectoderm,
and in a related form from the endoderm, too great significance should not be
attached to the place of their origin.

XI. The " Stratification " of the Cnidarian Body.

In the lowest Cnidaria the body during life consists of two layers
of epithelium separated by a supporting lamella ; these two layers are
similar to the two epithelial germinal layers of the gastrula larva.
The musculature is formed by processes of the ordinary epithelial cells.
Only the sexual products arise and continue to lie imbedded in the
epithelium.

As the complication of the organism increases, there is a tendency
for certain tissues and organs to detach themselves from the epithelium
and to take up a position beneath it. This tendency is shown by the
various tissues and organs approximately in the following order : —

1. The sexual organs, which in the lowest Cnidaria are already
subepithelial, and in the higher Cnidaria come to lie altogether or
partly in the jelly.

2. Connective tissue elements, which immigrate into the gelatinous
supporting membrane.

3. The musculature, whose elements first arrange themselves into
a subepithelial muscle layer, and then also move (partially at any rate)
into the jelly.

4. The tendency to take up a position deep in the body affects the
nervous tissue far less. In consequence of the inseparable connection
of the nervous system, on the one hand with the sensory organs, which
in accordance with their functions must remain at the surface, and on
the other hand with the musculature which tends to sink below it, this
system takes up an intermediate position.

We observe, then, in the Cnidaria the progressive development of
an intermediate layer between the outer body epithelium and the
inner intestinal epithelium, this intermediate layer being formed of

104

COMPARATIVE ANATOMY

CHAP.

heterogeneous elements of independent origin, connective tissue, muscu-
lature, nervous system, and sexual organs. The outer epithelium gives
rise principally to the connective tissue and the nervous system, while
the musculature and the sexual organs may be produced either by the
outer or the inner epithelium.

It is evident that the development of such an intermediate layer,
which we call mesoderm, is the necessary preliminary of a higher
organological differentiation of the body.

XII. Reproduction.

Asexual reproduction by fission and gemmation — Stock formation — Division
of labour and polymorphism.

Asexual reproduction is very common among the Cnidaria side by
side with sexual reproduction. Among the Ctenophora alone it has not
been observed. In Hydra we find asexual reproduction by gemmation

FIG. 78.— Bougainvillea ramosa (after Allman), with budding Medusae, h, Nutritive polyps 5
mk, medusa buds ; m, detached young Medusae (Margelis ramosa).

side by side with sexual reproduction in adult animals. Buds are formed
by hollow outgrowths of the body wall. These buds grow, and at the
distal end a breach is formed — the oral aperture, round which the
tentacles arise by means of new outgrowths. Such buds can detach

II

CNIDARIA—REPROD UCTION

105

themselves from the mother body, or they may in small numbers
remain united with it for some time. In the last case small Hydra
colonies composed of similar individuals arise.

In the same way elegant and richly branched colonies arise in
most Hydroids (Fig. 78). The individuals of such stocks are, how-
ever, generally not similar, but, as a consequence of more or less
division of labour, Dimorphism or Polymorphism takes place. We
distinguish : (1) sterile nutritive persons, which remain on the level of
the Hydroid, and undertake the feeding of the stock, the gastric cavities
of all the individuals of the stock being in communication with one
another ; (2) sexual persons, which undertake the duty of ripening the
sexual products, and also of planting
them out and dispersing them, so that
the young brood of Hydroids proceeding
from the fertilised egg may attach them-
selves in new places and form new stocks.
The sexual persons which are destined
for a free-swimming life, and which are
buds of the Hydroid stock, attain a struc-
ture corresponding with this manner
of life, they become young Craspedote
Medusce, which detach themselves from
the stock, swim away, and — often after
longer or shorter metamorphoses — ripen
the sexual products. That the Craspe-
dote Medusa is only a metamorphosed
Hydroid suited for a free -swimming
manner of life has already been pointed
out. In the accompanying illustrations
(Fig. 79, A-E), we can see how a bud

t TT j -j ij i -.L FlG- 79-— A> B> c> A E> Diagram-

Of a Hydroid Stock develops into a matic representations of the formation

Craspedote Medusa.

The development of free-swimming
sexual persons has also the further
advantage that it makes cross-fertilisa-
tion possible.

In many Hydroids, however, there develop on the stock by bud-
ding sexual persons whose structure approaches more or less nearly to
that of the Medusa, but does not reach it. Such medusoid sexual
persons or gonophores (Fig. 80) do not detach themselves as free-
swimming Medusce, but develop the sexual products while remaining
connected with the stock. It is not possible to define the various sorts
of medusoid buds as stages in the development of the Medusa form, as
we cannot see what advantage could be gained by the greater or less
development of the Medusa form in these attached sexual persons.
They should rather be regarded as so many stages of degeneration,
due to the fact that the sexual buds no longer detach themselves from

of a Craspedote Medusa by budding
from a Hydroid. Black portions =
gastric cavities, en, Endoderm ; e, ecto-
derm ; TO?*, gastric peduncle ; v, velum ;
ra, radial canal ; r, circumferential canal.

106

COMPARATIVE ANATOMY

CHAP.

the stock as free-swimming Medusce. Degeneration can go so far that
the original Medusa organisation becomes quite unrecognisable (Fig.
80, C).

The division of labour among the persons of a stock goes still
further in some Hydroids, and leads to the formation of polymorphic
stocks. Besides the ordinary nutritive and sexual persons, feelers
(tasters) devoid of mouth and tentacles, and thorn-like protective per-
sons (guard polyps) provided with a hard periderm skeleton, may
occur; between these latter the other persons can withdraw.

Reproduction by gemmation and fission is relatively rare in the
Medusce of the Hydrozoa, and in the Craspedote Medusce generally. In
the division of the Anthomedusce gemmation has till now been observed
only in the family of the Sarsiadce. Here numerous buds are formed
either at the edge of the umbrella, or on the very much lengthened

FIG. 80.— A, B, C, Three different types of gonophores from Hydrozoa. e, Ectoderm ; en,
endoderm ; es, outer ectodermal envelope of the gonophore ; u, umbrella ; ra, radial canal ; r, circum-
ferential canal ; t, tentacles ; m, gastric peduncle ; o, mouth ; ov, gonade (ovary) ; v, velum.

gastric peduncle. These buds grow into young Sarsice like the mother
animal, then detach themselves and swim about independently. Here
we have formation of free-swimming Medusa-stocks without division
of labour and without polymorphism of the individuals. The Medusce
attain full development only after their detachment from the mother
animal.

Eeproduction by repeated binary fission has also been observed in
Craspedote Medusce. In some cases (Gastroblasta) peculiar free-swimming
Medusa -stocks arise which have the following structure. A single
Medusa provided with tentacles and marginal vesicles carries on its
subumbrella numerous gastric pouches. The number of these gastric
pouches determines the number of persons in the stock, which are
so far incompletely divided from each other that their discs never
separate.

Asexual reproduction by a sort of fission occurs also in young
forms of the Discomedusce (e.g. Aurelia), i.e. in the young attached stage
known as Scyphistoma and described above. In the simplest case

CNIDARIA— REPRODUCTION

107

(monodise Strobila) the disc of the Scyphistoma (Ephyra) constricts
itself and separates from the peduncle, on which by regeneration a new
disc is afterwards formed. New discs, however,
are mostly formed between the peduncle and
the older discs before the latter detach them-
selves; then we have the typical polydise
Strobila (Fig. 81).

In Corals, reproduction by gemmation and
by incomplete fission is very wide spread. It
is, however, rare in the naked Actinia. It leads
to the formation of those occasionally very
large Coral-stocks whose skeletons are well known
as Eeef or Stone Coral, the Coral of commerce and
other Alcyonaria. The gemmation and stock
formation of the Alcyonaria is the most fully in-
vestigated. At certain points of the mother-polyp
outgrowths, the so-called stolons, make their
appearance and are arranged on the mother-
polyp in ways characteristic of each different
group. They are either simple, or branched in
a reticular manner. By new outgrowths, and
local widenings of the endodermal canals which
they contain, there arise, on these stolons,
young daughter animals, in which the mouth,
oasophageal tube, septa, and tentacles are formed.

FIG. 81.— Polydise Strobila
of Aurelia aurita, after
Haeckel.

In this way arise

B

FIG. 82.— Diagrams illustrative of gemination and stock formation in various Alcyonaria.
A, General diagram. B, Gorgonia. C, Tubipora. D, Alcyonium. Black portions the cavities of the
gastral system, s, GEsophagus ; se, septa ; mf, mesenterial thickenings ; dh, gastric cavity ; sk, axial
skeleton, drawn in layers to illustrate the manner of its origin.

108 COMPARATIVE ANATOMY CHAP.

stocks formed of individuals whose gastric cavities remain connected,
like those of the Reef Coral, by characteristic arrangements of the canal
system. The above illustrations (Fig. 82, A-C) show diagrammatically
the manner of gemmation and the beginning of stock formation in
different Alcyonaria.

In the Alcyonaria also, division of labour takes place, with the
resulting dimorphism or polymorphism of the persons (zooids). Thus,
side by side with the normal persons there are other persons without
tentacles and with septa reduced in number (two) whose chief function
is the taking in of water into the canal system.

XIII. Organisation of the Siphonophora.

It is most suitable for our purpose here to describe the structure
of the Siphonophora, as it can only be explained and understood by
help of the phenomena of asexual reproduction by means of gemma-
tion, of stock formation, and of the division of labour. The Siphono-
phora have actually been long considered by most investigators as
polymorphic animal stocks, although zoologists had not agreed as to
the significance of the separate parts.

The following description corresponds in general with the views
recently put forth by Haeckel.

In the order of the Siphonophora two animal groups have till now
been united, which, apart from the fact that both are Medusa- stocks,
have nothing in common, and, in any case, have quite different origins.
We shall therefore treat of these two groups — (1) the Siphonanthe and
(2) the Diseonanthe — separately.

I. The Siphonanthe. — These may be conceived of as colonies of
Craspedote Medusce by comparing their whole body with a Craspedote
Medusa on whose gastric peduncle numerous young Medusce have
arisen by gemmation, somewhat in the same way as in Sarsia
Siphonophora. While, however, the mother Medusa of the Sarsia is
radially constructed and all the daughter Medusce are like each other
and like the mother, the mother animal of a Siphonanth which is
recognisable in its young or larval stage is a much metamorphosed
Medusa. Its disc is mostly changed into an air vesicle, it possesses
only one tentacle (which also occurs in the Craspedota), and its gastric
peduncle is lengthened out into the generally very long " stem " of
the Siphonanth. The daughter Medusce, budding from the stem, are
neither like one another nor like the mother animal. They divide
between them the general work, and are consequently variously
modified to suit their special functions.

If we now more closely consider the body of a Siphonanth, we
must first bear specially in mind those parts which can be compared
with parts or organs of the mother Medusa of a proliferous Craspedote.

A. The pneumatophore or swim -bladder (Fig. 83) lies at the

GNIDARIA— ORGANISATION OF SIPHONOPHORA

109

upper end of the stem, and represents a metamorphosed Medusa
umbrella. (It is wanting only in the order of Calyconecta, where
the umbrella of the larval mother
Medusa develops into the first pro-
visional swimming -bell and is then
thrown off.) At one point of the
exumbrella an invagination forms at
an early stage for secreting air, the
air vesicle, which expands so much
that it represents by far the largest
portion of the original disc ; it always
remains in open communication with
the exterior by means of the aperture
of the invagination (the pore of the
air vesicle). Around the air vesicle,
in the bell which has been so much
modified and has become more or
less globular, there are 8 (less fre-
quently 4 or 1 6) endodermal chambers
divided by septa ; these open under
the air vesicle into each other and
into the endodermal axial canal of the
stem. These chambers correspond
with the radial canals of the Medusa.
The pneumatophore serves as a hy-
drostatic apparatus, which keeps the
whole Siphonophora colony floating in
the water. The air can be expelled
through the pore of the air vesicle,
and again secreted by the ectodermal
glandular epithelium at its base.

B. Only one of the tentacles is
fully developed. This is moved from
the margin of the disc on to the
subumbrella to the base of the stem,
and is probably usually thrown off at
an early stage.

C. The stem of the Siphonanth,
which is generally long, tubular, and
contractile, more rarely short and
flat, answers to the gastric peduncle
of a Medusa. An aperture (primary

oral aperture) is but rarely found at its lower end. The view that
these three parts together are equivalent to a Medusa is supported by
ontogenetic observation, as is to a certain extent evident from what has
already been said. The gastrula which develops from the fertilised egg
grows into a Siphonanth larva, such a larva possessing at first only

sb, Pneumatophore ; sg, swimming-bell ; ds,
bract ; t, tentacles ; groj, go?, go$, gonophores ;
hy, oral or gastric peduncle (siphon) ; p,
feeler or taster (palpon) ; A-H, various
groups of appendages which are never found
in this way together in any single Siphon-
anth. Black portion = gastric system.

110

COMPARATIVE ANATOMY

CHAP.

these three parts — umbrella, tentacle, and gastric peduncle. This
medusoid larva is bilaterally symmetrical. Its umbrella has a deep
cleft, it possesses only one tentacle, its gastric peduncle is filled with
yolk. The Siphonophora colony arises on the gastric peduncle by
gemmation.

Let us now consider the polymorphic appendages of the Siphonanth
stem, which we compared with the daughter Medusce budding on the
gastric peduncle of Sarsia siphonophora. All these appendages are
arranged on the stem in a line whose position is called ventral. The

au

FIG. 84.— Stephalia corona, after Haeckel. A, Halved longitudinally. B, From life, sb, Swim-
bladder ; au, aurophore ; sg, swimming-bells ; ka, canal system of the stem (chief stomach, st) ; go,
gonophore clusters ; o, aperture (mouth) of the peduncle (chief gastric tube, st) ; hy, gastric tubes
(siphons) ; t, tentacles.

line generally becomes a spiral, because of the spiral twisting of the
stem. Highest up on the stem under the pneumatophore (when one
is present) the so-called swimming-bells or neetophores are inserted ;
these are wanting only in the Cystonecta. The swimming-bells exclu-
sively and alone provide for the locomotion of the whole stock. They
have lost all those Medusa organs which were of no use to or even
hindered the fulfilment of this function, first of all therefore, mouth,
gastric peduncle, and tentacles. The locomotory organ of the Medusa,
the disc or umbrella, however, is all the more strongly developed ; ft
is much vaulted with a strong circular muscular layer in the sub-
umbrella. Its edge projects in the shape of a true velum. At the
base of the velum runs the circumferential canal, into which 4 radial
canals enter. The swimming-bells are so inserted on the stem by their

CNIDARIA— ORGANISATION OF SIPHONOPHORA

111

aboral or apical poles, that the aperture of the bell is turned downwards
and outwards, away from the apex of the stem.

When the swimming-bells contract, and so expel the water down-
wards out of their subumbrellar cavities, the whole stock is propelled
by the recoil in the opposite
direction, i.e. upwards. The swim-
ming-bells are not regularly, but
bilaterally symmetrical, which is
explicable by their insertion on
the stem, and by the position they
have to assume to effect the motion
forward of the whole stock. As
to the number and arrangement of
the swimming-bells, we find either
one or two opposite each other,
or several, or often very many,
arranged in two or many rows in
circles round the stem. The points
of insertion of the bells, however,
always lie in a spiral, which is
sometimes much extended, some-
times much compressed. The
direction in which this spiral is
twisted is the opposite of that
in which the other appendages
are arranged. The radial vessels
of each swimming - bell, which
unite at its apical pole, are in
open communication with the
endodermal axial canal of the
stem.

Beneath the swimming-bells the stem carries the following different
kinds of appendages, which we may also consider as modified Medusae :

A. The gonophores or reproductive persons (Fig. 83, gov go2, go3).
—To these belong exclusively the function of forming the sexual
products. They are either male or female. The typical organisation
of a Craspedote Medusa is still more or less faithfully maintained in
them. They possess a bell-shaped umbrella with velum, circumferen-
tial canal and radial canals, and, further, a gastric peduncle which
projects into the subumbrellar cavity (occasionally with an oral
aperture in addition), and in whose wall, as in the Codonidce among the
Craspedote Medusce, the sexual products arise. The umbrella is here,
probably, a protective apparatus. Occasionally the rudiments of
tentacles are still found on the margin of the disc. Sometimes, how-
ever, the whole Medusa form is considerably degenerated.

B. Sterile persons. — These perform the functions of taking in food,
and digestion, of protection, touch, etc. The Medusa structure in them

FIG. 85.— Praya galea, after tfaeckel.

112 COMPARATIVE ANATOMY CHAP.

is always obscured, and often so much so as to be unrecognisable.
The following different kinds of such sterile persons may be dis-
tinguished :

a. Persons in which the following typical organs of a Medusa may
still be recognised : (1) a variously-shaped protective or braet as
metamorphosed umbrella. It serves as umbrella or shield, and affords
protection, not only to the other parts of the same person, but also
to the neighbouring persons, which can withdraw under it. (2) The
oral or gastric peduncle (siphon), the chief organ for taking in food
and digestion. The siphon is often stalked, and the edge of the
mouth widened into a funnel, or produced into 4 points, or prolonged
like a proboscis. (3) A very contractile tentacle or capturing filament,
which is placed at the base of the gastric tube. The tentacle is
feathered on one side, i.e. it is provided with one row of lateral
branches, whose ends are armed with stinging batteries. Such a
sterile person simultaneously performs the functions of taking in food
and of protection (Fig. 83, A}.

b. Persons distinguished from those just described by the fact that
the contractile hollow siphon has lost its mouth, and so appears
changed into a taster or feeler (palpons). The tentacle at the base of
the feeler becomes an unfeathered, long, and very retractile sensory
filament (Fig. 83, B).

c. Persons in which the umbrella is completely degenerated, and
which consist of nothing but siphon and tentacle (Fig. 83, C).

d. Persons which have retained exclusively the function of pro-
tection, and in whom the umbrella alone, in the form of a bract, has
attained development, while the formation of siphon and capturing
filament has been suppressed (Fig. 83, E).

e. Persons reduced to tasters, without bracts and without sensory
filaments,

C. Special swimming-bells. — Nectophores, agreeing in structure
with the ordinary swimming bells developed at the upper end of the
stem are found in some Siphonanths on other parts of the stem as well.

These various appendages, or heteromorphic persons (A-C), of
which several may be wanting, occur in different and often very
characteristic order and manner of division on the stem. They are, in
the first place, arranged in many Siphonanths in distinct groups,
repeated at regular intervals and separated by internodes of the stem.

The following are the chief modifications which occur in the com-
position of such a group, which is known as a eormidium :

A. The eormidium consists of (1) a gonophore and (2) a sterile
person with bract, siphon, and capturing filament (Fig. 86).

B. To these two persons a third person, a special swimming-bell, is
added.

C. The eormidium consists of (1) one or more gonophores, (2) one
sterile person with siphon and tentacle, but without bract. .

ii CNIDARIA— ORGANISATION OF SIPHONOPHORA 113

D. It consists of (1) one or more gonophores, (2) one siphon with
tentacle but without bract, (3) one or more palpons with tentacle but
without bract.

E. It consists of (1) a group of gonophores, (2) a siphon together
with a tentacle, (3) one or more palpons without tentacle, (4) several
bracts, some of which perhaps belong to the palpons and to the siphon.

Less frequently we find in the cormidium several siphons with
tentacles.

The cormidia described under A and B can detach themselves
from the stem, and only when they are thus free-swimming Eudoxice
(A) or Erscece (B) do they ripen the sexual products in their gono-

\

FIG. 86.— Single Cormidium (Eudoxia) from Praya galea, after Haeckel. ds, Protective or
bract ; r, radial canals of the same ; st, portion of the stem ; Jiy, siphon or gastric peduncle ; t, ten-
tacle ; rik, stinging knobs ; sg, gonophore.

phores. From the fertilised egg a medusoid Siphonophoran larva is
then produced, and from this, by budding, comes the polymorphic
Siphonmth stock.

In many Siphonanths the arrangement of the heteromorphic
persons in special cormidia is either more or less obscured (e.g.
Rliizophysa, several Agalmidce and Forskalidce) or quite suppressed, so
that the persons are irregularly distributed on the stem (Physalia,
Agalmopsis). In this case the persons are generally appendages, in
which, apart from the gonophores standing in groups or clusters, the
medusoid structure is more or less completely degenerated : i.e. siphons
with tentacle, taster with or without sensory filament, or isolated
bracts.

This dispersed arrangement is to be explained in this way : the
parts belonging to a sterile person, such as siphon or taster, bract or
tentacle, become detached, and move away from each other, and stand

VOL. I I

114

COMPARATIVE ANATOMY

CHAP.

separately on the stem. These dislocated portions or organs are able
to multiply independently.

II. The Diseonanthe (Disealia, Porpeta, Porpalia, Velella). —

These have to be interpreted quite differently from the Siphonanthe.
According to the harmonious and convincing teaching of ontogeny and
comparative anatomy, these animals must be considered as Medusce
with marginal tentacles. These Medusce have a gastric tube with
mouth (principal siphon) in the centre of the subumbrella in the
typical way, but also produce secondary siphons or palpons, by gemma-
tion, on the subumbrella (just as in the G-astroblasta) ; out of the
wall of these secondary siphons the Medusa-shaped gonophores

FIG. 87.— Porpalia prunella, after Haeckel. cd, Central gland ; Ik, air chamber ; cp, central
pore of the same ; rk, radial canal ; sp, supporting plate ; eu, exumbrella ; su, subumbrella ; t, ten-
tacles ; g, gonades ; o, mouth ; ms, principal siphon ; gm, accessory siphons.

bud (Fig. 87). All the tentacles belong to the margin of the un-
divided persistent Medusa umbrella. In the umbrella (on which in
Velella a vertical crest, generally placed diagonally, rises) an air
vesicle is developed on the exumbrellar side ; this is often of very com-
plicated structure, many chambered, and originally octoradiate ; the
chambers communicate with the exterior through numerous pores.
The young stages of the Diseonanthe are typical Medusce, with 8 (later
16) tentacles at the disc margin, and with one central gastric peduncle
or siphon. The gonophores detach themselves as free -swimming
medusoid sexual persons, and only ripen the sexual products after their
separation.

The view of the Siphonophoran body here brought forward takes up
a position intermediate between two diametrically opposed theories,

ii CNIDARIA— ALTERNATION OF GENERATIONS 115

each of which has long had its supporters. According to one theory,
the whole Siphonophoran body represents a single Medusa person, and
all its separate appendages — the nectophores, siphons, tasters, tentacles
and gonophores — are nothing but displaced organs of this Medusa,
whose number increases by multiplication. According to the other
theory, the Siphonophoran body is a free-swimming polymorphic Hydroid
stock, and each of the appendages just enumerated, even each tentacle,
is to be considered as a more or less modified person, suited to some
special function, in consequence of an extreme division of labour,
either, therefore, as a metamorphosed Hydroid, or (as e.g. the necto-
phores and gonophores) as a metamorphosed Medusa.

XIV. Life-history of the Cnidaria, Alternation of Generations.

We shall return later to the special ontogeny of the Cnidaria, the
arrangement of the layers of the body, and the development of the
organs. Here we shall restrict ourselves to depicting the general
course of their life-history.

Hydra multiply both asexually by gemmation, and sexually by
means of fertilised eggs. From the latter, by a gradual course of
development, Hydra again arise.

In very many Hydromedusce an attached Hydroid form arises out of
the fertilised egg, out of which, by budding, comes a Hydroid stock,
which is at least dimorphic and often polymorphic. Some buds
become sterile nutritive persons, others sexual persons. The latter
detach themselves from the stock as free -swimming Craspedote Medusce
(Fig. 78, m, p. 104) and form the sexual products. From the
fertilised egg an attached Hydroid may again be produced. We thus
find here, in the cycle of development, two consecutive generations, as
it were, intercalated : (1) the dimorphic or polymorphic Hydroid stock
which reproduces by gemmation ; and (2) the Medusa which arises by
gemmation, detaches itself, swims about, and reproduces itself sexually.
Such an alternation of differently formed generations which multiply
in different ways is called alternation of generations (metagenesis).
It follows from our description that this alternation of generations is
the result of division of labour between the single persons of a
Hydroid stock. Each Medusa is originally equivalent to a nutritive
person, and it owes its structure to adaptation to the special function
of forming the sexual products and of dispersing them by means of
its free locomotion.

We must not, therefore, consider the Hydroid form as a young
stage of the Medusa form. Nutritive polyp and Medusa are sisters.
The one sister develops further than the other and reproduces sexu-
ally, while the latter remains sterile.

There are two other methods of development to be derived from
the alternation of generations of the Hydro -Medusa. There are
Hydroid stocks in which the sexual persons do not detach themselves

116

COMPARATIVE ANATOMY

CHAP.

from the stock, but remain connected with it as medusoid gonophores.
From the fertilised eggs of such Hydroids other Hydroids are produced.
On the other hand there are Hydro-Medusae in whose whole life cycle
no attached Hydroid stock is developed. From the fertilised egg of such
a Craspedote Medina another sexual Medusa is produced, often after a
series of metamorphoses.

In the Discomedusce also a kind of alternation of generations occurs.
The fertilised egg may develop into a young attached Medusa, which
reproduces asexually by axial budding (strobilation, Fig. 81, p. 107)

p - x^-

FIG. 88. — Nausithoe. pr, Perradii ; ir, interradii ; ar, adradii ; sr, subradii ; rl, marginal lobes ;
t, tentacles ; gf, gastral filaments ; m, circular muscle of the subumbrella ; sfc, sensory bodies
(rhopalia) ; g, sexual glands (gonades) ; in the middle the oral cross.

or by lateral budding. The constricted young Medusae (ephyrse),
whose organisation, but for absence of gonades, is essentially the same
as that of Nausithoe (Fig. 88), undergo a more or less complicated
metamorphosis, till they again become adult sexually mature Medusce.
Here, however, the organism which multiplies asexually is really a
young stage of the sexually differentiated Medusa, not a sister as in
the Hydro-Meduscz. The young Scyphistoma does not need to multiply
asexually. It can detach itself from the stem and develop direct into
a Medusa. There are also very many free-swimming Scypho-Medusce
from whose fertilised eggs a new Medusa is produced again direct
without the intervention of an attached stage in which multiplication
is asexual. This direct development is usually accompanied by meta-

ii CNIDARIA— LITERATURE 117

morphosis. Development with alternation of generations and without
it can occur in the same species.

In Corals a free-swimming larva is produced from the fertilised
egg, and this attaches itself and develops into a Coral, which either
remains a single individual or produces a Coral -stock by means of
gemmation and incomplete fission.

The Ctenophora, without exception, develop direct.

From the fertilised egg of the Siphonophora a medusoid organism
arises, which, by budding, yields the polymorphic animal stock. Here
also (Physalia, Disconanthe) the medusoid gonophores can detach them-
selves from the stock and lead a free life as sexual individuals. In
many Siphonanthe groups of persons, the already mentioned Eudoxice
and Erscece, detach themselves from the stock, and swim about freely
as minute new animal stocks. In their gonophores the sexual pro-
ducts are developed. From the fertilised egg a medusoid organism
arises, which by gemmation again becomes the polymorphic animal ]
stock. We thus have here again to do with a sort of alternation of
generations.

Literature.

R. Leuckart. Zoologische Untersuchungen I. Giessen, 1853.

The same. Zur ndhern Kenntniss der Siphonophoren von Nizza. Arch. f.

Naturgeschichte. 1854.

C. Gegenbaur. Beobachtungen uber Siphonophoren. Zeitschr. f. w. Zool. 1853.
The same. Neue Beitrdge zur Kenntniss der Siphonophoren. Nova acta. Tom. 27.

1859.

C. Vogt. Memoires sur les Siphonophores. Memoires de VInstitut genevois. 1854.
A v. Kolliker. Die Siphonophoren oder Schwimmpolypen von Messina. Leipzig,

1853.
Milne-Edwards et J. Haime. Histoire naturelle des Coralliaires. 3 vols. Paris,

1857-60.

Th. Huxley. The Oceanic Hydrozoa. Roy. Society. London, 1859.
De Lacaze-Duthiers. Histoire naturelle du corail. Paris, 1864.
Th. Hincks. Natural History of the British Hydroid Zoophytes. 2 vols. London,

1868.
Kolliker. Antomisch-systematische Beschreibung der Alcyonarien. Die Pennatuliden.

Abhandlungen SenTcenb. nat. Ges. Frankfurt. Bd. 7 and 8. 1872.
C. Glaus. Ueber Halistemma tergestinum. Arbeit. Zool. Institut. Wien. 1878.
G. J. Allman. A Monograph of the Gymnoblastic or Tubularian Hydroids. London,

1871-72. 2 vols.

N. Kleinenberg. Hydra. Leipzig, 1872.
0. and R. Hertwig. Das Nervensystem und die Sinnesm-gane der Medusen. Leipzig,

1878.

The same. Die Actinien. Jen. Zeitschr. f. Natumuiss. Bd. XIII. u. XIV. 1879-80.
R. Hertwig. Ueber den Bau der Ctenophoren. Jena, 1880.
C. Chun. Monographic der Ctenophoren. In: Fauna und Flora des Golfes von

Neapel. Leipzig, 1880.

E. Haeckel. System der Medusen. Jena, 1880, 1881.
A.Andres. Monografia delle Attinie. 1. Theil, In : Fauna und Floi^a des Golfes

von Neapel. 1884.

118 COMPARATIVE ANATOMY CHAP.

G. von Koch. Monographic der Gorgoniden. In : Fauna und Flora des Golfes von
Neapel. 1887.

R. Hertwig. Die Aktinien der Challenger Expedition. Jena, 1882.

E. Haeckel. Report on the Siphonophora of the Challenger Expedition. 1888.

A. Goette. Ueber die EntwicJcelung von Aurelia aurita und Cotylorhiza tuberculata.
1887.

Numerous important works and treatises by Metschiiikoff, F. E. Schulze, Haeckel,
Moseley, L. Agassiz, A. Agassiz, Claus, Weismann, Hamann, Grenadier,
Lacaze-Duthiers, Jourdan, Jikeli, Lendenfeld, Chun, C. Keller, Wilson, Fol,
Semper, Dana, G. von Koch (a series of important treatises on Corals in
Gegenbaur's Morph. Jahrb., vols. iv-ix). A. von Heider, etc.

The fundamental law of Biogenesis. — Egg segmentation and the development of
the two primary germinal layers of the Metazoa (gastrulation)— The ontogeny
of the Cnidaria.

Every Metazoon is, at the commencement of individual existence, a simple cell,
an egg cell ; i.e. its development starts from a point at which the Protozoon remains
during its whole life. Further, by the repeated division of the fertilised egg of every
Metazoon, a germ is produced, whose structure repeats in a general way the structure
of a simple Coslenterate. This germ, which is known as the gastrula, consists of two
cell layers, the ectoderm and the endoderm, which may be compared with the two
layers composing the body of a simple adult Ccelenterate.

Fundamental law of Biogenesis. — The frequently observed parallelism between
the consecutive stages of individual, or ontogenetic development, and the grades of
development presented by the animal kingdom, is thus explained by the theory of
descent. Every animal, in its ontogeny, passes through, in an extraordinarily ab-
breviated and concise manner, the long series of its ancestral forms. " Ontogeny,
or the history of the development of the individual, is a short recapitulation of
the history of the race, or phylogeny." This sentence, in which the fundamental
law of Biogenesis is formulated, contains a generalisation of the fact that every
animal passes on to its descendants by inheritance, not only its organisation at an
adult stage, but also its own course of development.

In the course of time adaptation, i.e. the survival of the fittest in the struggle for
existence, interferes with the action of heredity, so that species do not remain
constant, but change according to circumstances. In the same way, the ontogenetic
process of development also, i.e. the series of consecutive stages of development of a
species, may be subject to such modification that it no longer faithfully recapitulates
the process of development of its ancestors. The repetition of ancestral development
caused by inheritance is called " palingenetic " ; the modifications of ancestral
development caused by adaptation are "caenogenetic."

It is extremely difficult to determine in a concrete case what is palingenetic and
what is csenogenetic. It is at once evident that a purely palingenetic course of
development never occurs anywhere. It is only by the discovery of similarity be-
tween one stage of development in an animal and another adult animal, that we can
decide that that stage has something palingenetic in it. When a comparison of a
stage of development in one species with other species in an adult condition is not
possible, we have no sure means of judging what is palingenetic in it and what
csenogenetic. For instance, we know of no adult animal with which the Echinoderm
larva can be compared ; and we cannot consequently know whether these larvae have
in any way retained the organisation of their primitive Echinoderm ancestors.

Our justification, again, for holding that a stage of development is probably

ii FUNDAMENTAL LAW OS BIOGENESIS 119

palingenetic, when other organisms in adult condition possess essentially the
organisation of that stage of development, stands or falls with the assumption that
in the course of the earth's history some lower animal forms have remained very little
altered by the side of others which have developed into higher organisms, and that
it is not the case that nearly all lower animals living to-day are the degenerate
descendants of ancestors once highly developed. Palaeontology teaches us, in fact,
that during those epochs of the earth's history which are accessible to investigation,
and among those organisms which we know in a fossilised condition, many forms and
groups of forms have remained unaltered for immensely long periods, while other
forms on the contrary have progressed. On the other hand, we cannot doubt that
retrogressions in the development of the organic world are by no means rare
phenomena. The acquisition of a stationary manner of life, for example, and still
more of parasitism, necessitate such retrogressions. In many cases, in consequence of
new conditions, sexual maturity may be shifted to an earlier stage of development,
better suited for competition, and the "adult" form may gradually cease to be
developed. The Axolotl, for instance, generally becomes sexually mature at a stage
with external gills — a so-called larval stage — and reproduces itself at that stage. It
only rarely develops into the "adult" animal. If, in consequence of certain condi-
tions, the development into an adult animal were altogether and always to cease to
take place, we should have before us a case in which an animal, according to the
common conception more lowly organised, descended from one more highly
developed. If we wish, therefore, to compare the developmental stages of an animal
with the final stages of other animal forms, in order to demonstrate their palingenetic
meaning, we must always be' able to bring forward good reasons for believing that
these animal forms are not simplified or degenerated.

If the distinction between palingenesis and csenogenesis is difficult, even when we
can illustrate it by adult animals, the difficulty increases when the comparison
remains purely ontogenetic ; i.e. when we can only compare developmental stages of
one animal with developmental stages of other animals. It is now accepted, that
when two animal groups have similar larval forms, these groups are racially related.
The larva of Balanoglossus agrees in many important points with the larva of the
Echinoderm ; Balanoglossus is therefore considered to be related to the Echinoderms.
It is possible that this view is correct ; but we cannot say that this relation is prob-
able ; for we can bring forward no reason to show that either of these larval forms
has any palingenetic significance.

In dealing with ontogenetic problems we have to take into account another series
of considerations, which are partly a simple result of a consistent carrying out of the
Darwinian principles. We give only the most important.

1. We are more likely to be justified in considering an ontogenetic process of
development palingenetic, when it exhibits from first to last an unbroken series of
self-supporting stages of development, so that we can imagine the larva at each stage
to be an adult sexually mature animal. But we certainly do not find such a method
of development entirely realised anywhere in the animal world.

2. It is an advantage for an animal whose organisation is adapted to its condi-
tions, i.e. which maintains with success the struggle for existence, to reach its adult
form, not only as directly and as quickly as possible, but economically, i.e. without
the development of parts which have become useless.

3. Such a direct and abbreviated development can take place only when the
developing animal is from first to last provided with nourishment by means of which
it can develop. This occurs either by the egg receiving from the mother body
nutritive yolk to help in its development, or by its being nourished, as in viviparous
animals, direct from the mother body. Everything that is connected with such

120 COMPARATIVE ANATOMY CHAP.

nourishment of the developing animal is a secondary addition to the original develop-
ment. We cannot think either of a bird embryo at the time of incubation, or of a
mammalian embryo in its egg envelopes as an independent and self- feeding animal.

4. Direct development may occur together with gradual development or meta-
morphosis in the ontogenetic development of the animal. Of all the original in-
dependent stages of development, often only one or a few are preserved, viz. those
which are specially capable of competition, whilst all the others disappear in the
direct development. Thus the insect egg develops direct into the caterpillar or
larva by the help of the nutritive yolk. The larva at this stage of development,
being suited for competition, procures food for itself independently, grows vigorously,
and then again develops direct (pupal stage) by the help of the stored -up food into
the adult winged insect.

5. The effect of the struggle for existence on the larvse at the various stages of
development, and especially on those which feed independently, is just the same as
on adult animals. It is therefore to be expected that the different manners of life
(attached, free-swimming, parasitic, etc.) during the stages of development should
determine adaptations and modifications in the larval organisation similar to those
in adult animals. Such modifications, however, have a limit not clearly known to
us, just because the stages they affect are stages of development whose purpose is
to produce an adult sexually mature animal.

6. The more important organs or systems of organs are, the earlier do they
begin to form in the larva ; or, in other words, the order of their development in
time is in direct correspondence with their importance to the adult animal.
Whatever holds good for the adult animal also naturally holds good for every
organ and for every adaptation ; its development becomes more and more direct, and
more and more abbreviated ; it adapts itself more and more to the purpose of reaching
as soon as possible the form and arrangement which belong to the adult animal.
Increasingly defined localisation of the developing parts is a necessary consequence
of their earlier commencement.

We see from the above that in phylogenetic investigations it is perhaps still
more difficult to decide the true bearing of ontogenetic than of anatomical facts.
We can only attain to phylogenetic conclusions of a certain degree of probability
when comparative anatomy and comparative ontogeny go hand in hand, when com-
parative anatomy takes into account the developing organs, and when comparative
ontogeny does not leave out of consideration the last stages of development.

Segmentation and gastrulation. — When we come to investigate in what manner
the bi-laminar germ, the gastrula of the Metazoon, arises out of the fertilised egg,
we find what appear to be very different methods of development. Numerous
thorough investigations have proved that this variety of methods is almost exclu-
sively caused by the quantity and distribution of the nutritive yolk in the egg.
If we assign to the influence of the nutritive yolk the importance that is due to it,
we shall be convinced that one single process underlies all these different phenomena.
We must keep well in view, (1) that the nutritive yolk, or deutoplasm, is an
inert, lifeless nutritive material deposited in the egg cell, and (2) that the forma-
tive yolk, or the protoplasm, with the nucleus it encloses, is the only living active
portion.

We have already spoken of the variations in amount and distribution of the
deutoplasm in the egg. We shall now describe the first appearance of segmentation
or furrowing in the different kinds of eggs. The following are the types of eggs
whose segmentation we propose to describe : —

Holoblastic alecithal egg. Holoblastic telolecithal eggs, with varying quantity
of nutritive yolk : Eupomatus (Annelid), Discocelis (Polydad), Bonellia (Ecliiurid],

ii SEGMENTATION AND GASTRULATION 121

Ctenophora. Holoblastic centrolecithal eggs : Geryonia (Craspedote, Medusa}. Mero-
blastic telolecithal eggs : Shark. Meroblastic centrolecithal eggs : Insects.

The first divisions occur in directions which seem to be identical in the eggs of
most animals. The first plane of division which, after previous nuclear division,
separates the two daughter cells of the egg, i. e. the first two segmentation spheres,
or blastomeres, runs in the direction of the principal axis of the egg, from the
animal to the vegetative pole. By the animal pole Ave mean that portion of the
surface of the egg at which the spermatozoon entered, and near which, in telolecithal
eggs, the chief mass of the formative yolk lies. The point lying diametrically opposite
to this we call the vegetative pole. The first plane of division is, therefore, with
relation to the two poles, meridional.

The holoblastic alecithal egg falls at the first division into two similar blasto-
meres, each of which has a nucleus in its centre.

The holoblastic telolecithal egg falls generally into two blastomeres, each of which
repeats the structure of the undivided egg. Each blastomere shows polar differentia-
tion— at the animal pole lies the greatest mass of formative yolk with the nucleus,
at the vegetative pole the greatest mass of nutritive yolk (Fig. 89, B, C}.

The holoblastic centrolecithal egg (Geryonia) falls into two blastomeres, each of
which repeats the structure of a telolecithal egg, the formative yolk only appearing

FIG. 89.— 2-Blastomere stage of different eggs.— These are placed in this and all following
illustrations, so that the animal pole is directed upwards, and the vegetative pole downwards. The
nucleus is black, the formative yolk dark, the nutritive yolk light and granulated. F represents
merely a portion from the animal pole of an egg.

at the free surface of the blastomere, and not at the side which is directed towards
the plane of division. The nucleus of each blastomere lies superficially in the forma-
tive yolk (Fig. 89, D).

In the meroblastic telolecithal egg the inert lifeless nutritive yolk is so strongly
developed in comparison with the living active formative yolk, that the formative
yolk when dividing is not able to effect the division of the whole of the nutritive
yolk. The former only therefore divides, the latter remaining undivided. We have
thus a large sphere of nutritive yolk, with two masses of formative yolk divided by a
meridional furrow at the animal pole, in each of which lies a nucleus (Fig. 89, F}.

In the meroblastic centrolecithal, or rather mesolecithal, egg the central forma-
tive yolk again alone is able to divide under the influence of the nucleus, while the
remaining portions of the egg continue at first undivided (Fig. 89, E}.

The second plane of division is also meridional and stands at right angles to the
first. It divides each of the first blastomeres into two halves exactly in the same
way as the first plane divided the whole egg.

The third plane of division seems to be pretty generally equatorial. It is visible
on the exterior of the egg as an equatorial furrow. It stands at right angles to the
first two planes of division and to the chief axis of the egg, and divides the first 4

122 COMPARATIVE ANATOMY CHAP.

blastomeres into 8, of which the 4 animal portions form the animal part of the germ,
and the 4 vegetative portions its vegetative part.

In the holoblastic alecithal germ the 8 blastomeres are of equal size, and the
third plane of division shows itself outwardly as a strictly equatorial circular furrow
(Fig. 90, A).

In holoblastic telolecithal germs each of the 4 blastomeres divides into a
smaller animal blastomere, consisting almost exclusively of formative yolk, and a
larger vegetative blastomere containing the nutritive yolk and a small quantity of
formative, yolk. The greater part of the latter lies towards the animal pole of
the blastomere and contains the nucleus. Each vegetative blastomere of the germ of
8 blastomeres thus repeats the structure of a blastomere of the germ at the 4 -blasto-
mere stage. The more considerable the quantity of nutritive yolk in the germ,
the larger is the vegetative blastomere as compared with the animal. The smaller
blastomeres are called micromeres, the larger macromeres. The division of the 4
blastomeres into 4 micromeres and 4 macromeres looks as if the former budded from
the latter (Fig. 90, B, C].

In the holoblastic centrolecithal (Geryonid] germ, after the appearance of the
third plane of division, which is here strictly equatorial, the 8 blastomeres are of

FIG. 90.— A, F, Diagrams showing the 8-blastomere stage of different eggs.

equal size. Each blastomere has a peripheral layer of formative yolk, with a nucleus
imbedded in it, and a mass of nutritive yolk towards the centre of the spherical germ
(Fig. 90, D).

The germ of the Ctenophora, departs from the usual arrangement of the telolecithal
holoblastic germ, since the third process of division is not equatorial. Each of the
4 blastomeres divides, in a plane slantingly meridional, so that the germ consists
of 8 approximately equal blastomeres, of which 4 lie rather nearer the animal pole
than the other 4. The 8 blastomeres are grouped round a central space, which
lies in the principal axis of the germ. Each blastomere shows the structure of the
original Ctenophoran egg, the formative yolk lying chiefly at the animal pole of the
blastomere, and the more considerable nutritive yolk forming its larger vegetative
part.

In the meroblastic telolecithal germ, in consequence of the great development of
the nutritive yolk, the "equatorial " division takes place quite near the animal pole.
Exteriorly this division is visible as a circular furrow (polar circle) which divides the
4 prominences of formative yolk into 4 smaller central prominences with nuclei, and
4 peripheral prominences also with nuclei. The 4 central and the 4 peripheral
prominences are not completely separated below the surface, either from each other
or from the subjacent nutritive yolk (Fig. 90, F).

In the meroblastic mesolecithal germ the 8 central masses of formative yolk,
each provided with a nucleus, are not separated from the great mass of nutritive
yolk surrounding them (Fig. 90, E).

No general rule can be given for further divisions. Many meridional, equatorial,
and cross divisions follow. The determination of the processes in detail is the more

II

SEGMENTATION AND GASTRULATION

123

difficult in that divisions occur simultaneously at different points of the germ, and
the number of blastomeres considerably increases. Starting with the 8-blastomere
stage, we Avill further follow the course of development of each germ type.

1. In the holoblastic alecithal germ (Fig. 91), out of the 8 blastomeres, by
repeated meridional and equatorial division, 16, 32, etc., blastomeres of about
equal size are produced, which together form the uni-laminar wall of a sphere,
which becomes hollow by the formation of a cavity (segmentation cavity, blastoccel)
(total equal furrowing). At this stage the germ is called blastula, or cceloblastula,
because it is hollow. During repeated division of the blastomeres the blastula
becomes flattened at the vegetative pole ((7), the flattened part sinks into the seg-
mentation cavity more and more, so that the invaginated portion, which consists of
vegetative blastomeres, approaches the non-invaginated part by means of complete or
partial reduction of the cavity. We now have before us a germ consisting of two
layers of blastomeres, which have become epithelial. The outer layer is the ecto-
derm, the inner the endoderm. At the edge of the aperture of invagination or
blastopore the two layers pass into each other. The endodermal blastomeres or cells

FIG. 91. — Segmentation and gastrulation
of a holoblastic alecithal germ. B, Blastula.
D, gastrula. fh, Segmentation cavity ; e, ecto-
derm ; en, endoderm ; HI, blastopore.

FIG. 92.— Segmentation and gastrulation of
a telolecithal egg with little nutritive yolk (of
Eupomatus). mi, Micromeres ; ma, macromeres ;
en, eudoderm ; me, mesoderm ; HI, blastopore.

form together a hollow sac, the arch-enteron, the cavity of which opens externally
through the blastopore. At this stage the germ is called gastrula, and in this case
coelogastrula, because the blastopore leads into an open enteric cavity. This pro-
cess, by which a blastula becomes a gastrula, is called invagination or embole.

2. In the holoblastic telolecithal germs, the phenomena of blastula and gastrula
formation are at least apparently different according to the presence of much or
little nutritive yolk. Let us first consider a germ with little nutritive yolk, as,
for example, that of Eupomatus (Fig. 92). It consists of 4 animal micromeres
and 4 vegetative macromeres. The difference in their sizes is not very great,
because of the small mass of nutritive yolk, and the third plane of division lies near
the true equator. First the 4 micromeres divide, then the 4 macromeres. We thus
now have 8 micromeres and 8 macromeres. The micromeres continue to divide, and
the macromeres follow at a slower rate. While, however, the micromeres always
divide into two cells or blastomeres of pretty equal size, the equatorial divisions
of the macromeres are such that each divides into a micromere, which is directed
towards the animal side of the germ and contains less nutritive yolk, and into a
macromere, turned towards the vegetative pole and containing more nutritive yolk.

12-i

COMPARATIVE AX ATOMY

The number of micromeres thus increases, as in the iirst place they themselves divide,
and in the second place new microineres, produced by macromeres, become added
to them. This process of the continual production of micromeres is extremely
important for the comprehension of all following types of segmentation and gastrula-
tion. The furrowing is total, but already a little unequal. The blastula which
results is a cceloblastula, with a cap-like cover of micromeres and a lower part
formed of a few yolk-containing macromeres. In consequence, of the considerable
size of the macromeres the segmentation cavity is somewhat narrowed. The
gastrula is formed by invagination, and looks as if the layer of macromeres which sinks
in were grown round on all sides by micromeres. The gastrula is a coelogastrula,
and the arch-enteric cavity appears narrowed in consequence of the large size of
the macromeres.

Closely connected with the segmentation and germ-layer formation just described
is another process, an example of which is afforded by the developing germ of JJimcUic.
(Fig. 93). The process is essentially the same ; the apparent variations are explicable
by the fact that the mass of nutritive yolk is more considerable. The 4 large
macromeres, burdened Avith yolk, appear in consequence of their size as the fixed
resting-point round which the processes of development take place. At first,
as before, the 4 micromeres divide. The 4 macromeres are telolecithal
blastomeres. The chief mass of the formative yolk left after the first division
or budding no longer lies in the direction of the animal pole, but at some
distance from it, towards the outer edge of the 4 micromeres. The division or
budding of these 4 macromeres leads to the formation of 4 micromeres, which
take up a position externally, side by side with those already formed (A}. The 4
macromeres still contain a remnant of formative yolk, which, as it always lies at the

edge of the micromcre cap, moves along
the surface of the germ from the animal
side of the macromeres towards the vegeta-
tive pole of the germ. Thus the process
goes on (JJ, C). As the micromeres divide,
and as their number increases bv the con-
stant formation of new micromeres from
macromeres, the macromeres are at last
surrounded, everywhere except at a small
space at the vegetative pole. This process
is called epibole ; it is the growth of
micromeres over a resting mass of very
large macromeres. It will be seen from
the above description that this process and
that of invagination are fundamentally
identical. The gastrula which is formed
is a solid sterrogastrula, whose enteric
cavity is almost filled up by the large si/e
of the yolk-laden macromeres. The micro-
meres j iresent at this stage form the ectoderm ; the macromeres represent the rudi-
ments of the endodcrm and of a part of the mesoderm. The blastula stage of this
method of development is unrecognisable.

From the type of segmentation exemplified by JioncU.iu we pass on to one nearly
related which occurs in the Polijdada (Fig. 91). The formation of micromeres, and
their growth over the macromeres, is just as in JJoncUla. But here the 4 micromeres
which are first constricted yield, by division, the whole ectoderm. The micromeres,
which become constricted oil' from the macromeres in the second (and third) order,

Fie. 03. — Segmentation and gastrulation
of the egg of Bonellia, alter Spengel. mi,

Microin<;res ; -'11111.. inacrouii-'ros ; cc, cctiKluriii.

ii SEGMENTATION AND GASTRULATION 125

yield a great part of the mesoderm (musculature, sexual organs, connective tissue),
and are overgrown by the ectoderm micromeres. All micromeres which are detached
later belong to the endoderm. We have here before us a typical case of the tendency
to shift back formations to very early stages of development. A typical gastrula is
not developed in the Polyclada, since the
separation of ectoderm and endoderm occurs
as early as the 8-blastomere stage. We
shall return to this later.

The Ctenopkora exhibit an interesting
process of segmentation and gastrula tion
intermediate between the gastrulation by
means of epibole of the telolecithal holo-
blastic eggs and the gastrulation by de-
lamination of the centrolecithal eggs to be FlG- 94.- Segmentation of a Polyclad

spoken of later. The 8-blastomere stage of egg (o: Discocelif mi> First generation of

, , ,& micromeres (ectoderm - forming cells) ; ?>ui,

these animals has been described above, second generation of micromeres (mesoderm
All the 8 blastomeres are telolecithal, with micromeres) ; ma, macromeres.
the formative yolk directed towards the

animal pole. What occurs at the 4-blastomere stage of the telolecithal holoblastic
eggs hitherto described, viz. the constriction of the 4 micromeres from the 4 macro-
meres, here takes place one stage later, at the 8-blastomere stage. The 8 blasto-
meres in fact give off 8 micromeres towards the aboral pole (Fig. 95, A}. The further
segmentation is quite similar to that of Bonellia. The micromeres increase in
number (1) by division, (2) by the continual addition of new micromeres towards
the vegetative pole, by constriction from the macromeres (J5, (7). After the micro-
meres have thus grown round the macromeres, leaving a large region at the vegeta-
tive pole in which the macromeres come freely to the surface, the formation of
micromeres does not cease, as in the Polyclada. The already formed micromeres,
however, yield exclusively ectoderm ; the remaining macromeres, part of the mesoderm
and the endoderm. Here also no very recognisable gastrula is developed.

FIG. 95.— Three stages in the segmentation of a Ctenophoran egg. mi, Micromeres ;
ma, macromeres.

In all processes of the formation of micromeres the following is to be specially
noted. After a macromere has constricted off a micromere, or, what is the same
thing, after a blastomere has divided into a small micromere with little or no nutri-
tive yolk, and into a large macromere with much nutritive yolk, the portion of
formative yolk or protoplasm which remains in the macromere grows evidently by
the assimilation of nutritive yolk before the macromere can again divide.

3. Segmentation and gastrulation of the holoblastic centrolecithal germ. We
give as an example of this the Geryonid germ which has been the most carefully
investigated and is the best understood (Fig. 96). We are already acquainted with
the 8-blastomere stage. Each blastomere is telolecithal, with deutoplasm directed
towards the centre of the germ and protoplasm towards the circumference. The
8 blastomeres divide into 16, and then into 32 blastomeres of equal size, which

126 COMPARATIVE ANATOMY CHAP.

all remain telolecithal in the same way as in the 8-blastomere stage. The 32
blastomeres of the spherical germ form a single layer round a considerable central
cavity. A blastula-like stage thus occurs, though the germ has really another signi-
ficance, as the central cavity does not represent the segmentation cavity of the
alecithal germ, but, as we shall see, the enteric cavity.

When 32 blastomeres are formed, the formation of micromeres follows. From
each blastomere a micromere is constricted off on the outer side, so that the germ
now represents a double-layered hollow sphere, whose outer layer is formed of micro-
meres, and whose inner layer consists of macromeres. The micromeres increase
in number (1) by themselves dividing, (2) by the formation once more of micro-
meres which are constricted off on the outer side of the macromeres. The micromeres
form the ectoderm, the macromeres the endoderm, which surrounds a completely
closed cavity — the enteric cavity. The germ thus represents a ccelogastrula without
blastopore. "We call this a cceloplanula. The formation of the two germ layers in

FIG. 96.— Segmentation and gastrulation of the Geryonid egg. mi, Micromeres ;
ma, macromeres ; e, ectoderm ; en, endoderm.

the manner described above is called delamination. It must appear clear from our
description that this cannot be sharply distinguished from epibole. Both processes
rest on the formation of micromeres. In Bonellia, and the Polyclada, the first
formation of micromeres, or delamination, takes place at the 4-blastomere stage, in
the Ctenophora at the 8-blastomere stage, and in Geryonia at the 32-blastomere
stage.

4. The meroblastic mesolecithal germ was left at the stage where the formative
yolk or protoplasm was divided into 8 small masses, each with a nucleus, in the
centre of the undivided nutritive yolk. If we compare this stage with the 8-
blastomere stage of the Geryonid germ, we shall see that these 8 masses of
formative yolk correspond with the 8 portions of protoplasm of the latter germ,
only that in the former, in consequence of the originally different arrangement of the
nutritive yolk, the masses of protoplasm lie at the centre, while in the latter they
are peripheral. In the meroblastic germ the nutritive yolk is undivided, the
formative yolk being incapable, at the time of division, of effecting the division at
the same time of so large a mass ; in the holoblastic egg the formative yolk controls
the whole less massive nutritive yolk.

The 8 central nucleated masses of protoplasm of the meroblastic mesolecithal
germ (Fig. 97) we shall call merocytes. They are often branched, and have amoeboid
movements. Their processes penetrate the surrounding mass of yolk, and are also
connected with the thin layer of protoplasm which is found at the surface of the
germ. They feed at the expense of the deutoplasm.

The 8 merocytes divide into 16, 32, and so on, and move at the same time
centrifugally through the yolk to the surface of the germ, where they form a simple
continuous layer. It is this layer of merocytes which is called the blastoderm.
The germ has now become centrolecithal, and agrees in its structure with the Geryonid

II

SEGMENTATION AND GASTRULATION

127

germ before the formation of micromeres or delamination, except that the central
nutritive yolk is here still undivided.

The further development of the germ has unfortunately not yet been thoroughly
investigated. Some observations tend to show that at this point, as in the Geryonid
germ, delamination or formation of micromeres takes place. The merocytes at the
surface divide in such a way that the outer portion separates off completely as a
nucleated micromere layer detached on all sides, while the inner portion remains in

FIG. 97.— Segmentation and formation of blastoderm in the egg of an Insect.

the nutritive yolk as a nucleated merocyte layer. Thus the two germ layers are
formed ; the outer micromere layer represents the ectoderm ; the central mass of
nutritive yolk with the merocytes which belong to it probably represents the
endoderm.

The merocytes by increasing in number and feeding at the expense of the
nutritive yolk, become able at last to overpower and incorporate it, i.e. they can
divide it.

The gastrula is here a solid sterroplanula.

It has been observed in many cases that in the partial segmentation, i.e. in the
multiplication of the central merocytes, only
some of them move towards the surface, there
to form the blastoderm, while the others
remain in the yolk.

5. The meroblastic telolecithal germ (Fig.
98). We left this germ at the 8-blastomere
stage, with undivided mass of nutritive yolk.
We may here also call the masses of proto-
plasm merocytes. They stand in exactly the
same relation to the collection of protoplasm
at the animal pole of the blastomeres of the
holoblastic telolecithal egg as do the merocytes
of the mesolecithal meroblastic germ to the
protoplasmic portion of the blastomeres of the
Geryonid germ.

The 8 primary merocytes divide in the
following manner. The 4 central merocytes
divide into 4 secondary central merocytes in
contact with one another at the animal pole,
and 4 secondary peripheral merocytes. The
former are now entirely severed from each FIG. 98. — Segmentation of a mero-
other and from the nutritive yolk as micro- bl.astic telolecithal egg (of a shark), after
meres. The latter remain connected with the Riickert- mi> Micromeres; dp, deutoplasm;

,, „., me. merocytes.

yolk as merocytes. The primary peripheral

merocytes also divide, but their descendants remain connected with the yolk as
merocytes. Some of them even sink into the nutritive yolk as branched and probably
amoeboid merocytes.

123 COMPARATIVE ANATOMY CHAP.

The region at the animal poll' of the germ (egg) in which the inicromeres and some
of the merocytes are visible is called the germ or germinal disc.

The further course of segmentation is as follows : —

1. The micromcres already formed continue to divide independently.

'2. Towards the edge of the germ disc new inicromeres are constricted oil' from the
nuTocytes, and then associate themselves to those already existing, so that the area
of the micromcre region, or germ disc, increases.

3. The merocytes at the edge of the germ disc, as well as those which lie deep in
the nutritive yolk, constantly divide, so that not only does the germ disc grow at its
edge, hut the nutritive yolk becomes more and more crowded with merocytes, which
penetrate further and further into it.

Finally, the germ shows the following structure : at the animal pole, in the middle
of the germ disc, lies a Hat mass of inicromeres consisting of several layers, the rudi-
mentary ectoderm. Lying at the edge of the germ disc, and imbedded in the yolk,
are merocytes. The greater number of these merocytes yield the material from which,
by the formation of inicromeres in a way hereafter to be described, a large part of the
mesoderm and the endodenn are built up. By the constriction of inicromeres from
merocytes at the edge of the germ disc, and by division of the already formed ecto-
derm cells, the cap of ectoderm increases still more, so that it grows more and more
round the germ as the blastoderm.

We see that in the development of the meroblastic telolecithal egg the gastrula
form is unusually indistinct, because of the enormous mass of the nutritive yolk.
The micromere-cap respresents the ectoderm ; the yolk with the merocytes the enclo-
derm and part of the mesoderm. If a blastopore is sought for, this can only be re-
presented by that portion of the germ which is not covered by the ectoderm cap,
where the nutritive yolk comes to the surface. The edges of the blastopore are
represented by the edges of the ectoderm cap. The gastrula formation here also
occurs by means of epibole ; the gastrula is a solid discogastrula.

In the above description only the most important types of segmentation and
formation of the two primitive germ layers have been brought forward. The
processes in the animal kingdom are in reality extremely varied, yet most of them
can lie included without much dilliculty under one or other of our heads. The great
majority of investigations are still insufficient ; because in some cases, as in the
meroblastic eggs, observation is very difficult. The, dilliculty is to establish quite
clearly the distribution of the formative yolk in the egg, and to follow the division
of the blastomeres in detail.

The method of gastrulation does not afford any means of recognising the relation-
ships of animals, as it is determined by the quantity and distribution of the, nutritive
volk, and this may be quite dissimilar in animals nearly related, and similar in
widely separated forms.

The question, which sort of segmentation and gastrulation is the original, has
been much discussed. Most authorities hold, not without reason, total equal
furrowing, the formation of a cccloblastula, and the subsequent formation of a coelo-
gastrula by invagination to be the original method. This view is supported chiefly
by the following facts : — (1) the want of a nutritive yolk, which cannot have been
present in corresponding racial forms ; ("2} the similarity of the blastula with certain
1'roto/oan colonies (Mayosphcwct, Vulnw] ; (3} the similarity of the coelogastrula
with the simplest C<i:l<'iifsr(it(t (Ofynthus, Hydroids). Just this similarity makes it
possible, for us the more, easily to imagine these germ forms as independent adult
animals.

The free-swimming gastrula, which is usually ciliated on the outer surface,
possesses a mouth (blastopore) through which it can or could take food into the
enteric cavil v.

ii ONTOGENY OF GNIDARIA 129

It is important to note that in most of the higher bilaterally symmetrical animals
the bilateral symmetry appears very early. The gastrula of these animals is
bilaterally symmetrical, i.e. only one plane can be made to cut it into two exactly
similar halves. We can consequently distinguish in the gastrula upper and lower,
anterior and posterior, right and left. The gastrula of the radiate Ccelenterata, on
the contrary, is radiate or rather uniaxial. There is one principal axis round which
all the elements of the body are arranged in circles. This principal axis has unlike
poles ; at one pole, the vegetative, lies the blastopore ; the other, the animal pole, is
closed.

Bilateral symmetry is shifted back in many bilaterally symmetrical animals
to much younger stages of development — to the blastula or the segmentation stages.
In a few cases even the egg is bilaterally symmetrical, and the position of the future
principal regions of the body can be determined even in it.

In the uniaxial gastrula the blastopore is round and closes to a point. In
the bilateral gastrula, however, it has become slit-like, and closes either from front
to back or vice versa, in a line lying in the plane of symmetry or median plane of
the body.

Ontogeny of the Cnidaria. — The segmentation is everywhere complete. The
formation of the gastrula occurs by invagination, epibole, or delamination. In the
last case a cceloplanula arises direct. Where a ccelogastrula or a sterrogastrula
occurs it changes, in all cases except the Ctenophora, into a planula by the closing
of the blastopore. This is generally free-swimming and ciliated, and has a tuft of
long, mostly immobile sensory hairs at the original animal pole, which we now call
aboral. A Hydroid arises out of the blastula by the formation of the definite oral
aperture by means of a breach where the blastopore closed, the animal having
attached itself by the aboral end of the body. Round the mouth, the tentacles bud
out as hollow outgrowths of the ectoderm and endoderm.

The direct development of a Graspedote Medusa from the fertilised egg is best
known in Geryonia, with whose blastula we are already acquainted.

Between ectoderm and endoderm a jelly is formed, which constantly increases in
mass, so that the ectoderm sac is separated by a great interval from the endoderm
sac which it encloses. At one point only, the future oral pole, which probably agrees
with the vegetative pole, the endoderm sac remains in contact with the ectoderm
sac. The permanent mouth is formed at this point of junction by means of a breach,
while at the same time, at some distance from the mouth, the velum arises as a
circular thickening of the ectoderm, and the 6 tentacles as buds, into whose axes
solid processes of the endodermal sac grow. The connection of the endodermal ten-
tacle axis with the gastral sac soon ceases. The oral surface of the larva, which is
surrounded by the tentacles and the velum, sinks in and becomes the concave sub-
umbrella. The Medusa form thus gradually comes into existence. How the radial
vessels are formed in Geryonia has not been investigated.

A Scyphopolyp (Scyphula of the Acraspeda, Coral polyp) arises out of a planula
in the following way (Fig. 99). The planula (A) attaches itself by the animal or
aboral end of its body (B}. At the oral end the body of the ectoderm sinks in in the
shape of a pit and forms the ectodermal oesophagus with the external mouth (Cf).
The base of the oesophagus then breaks through in the direction of the gastral cavity
(D), and so arises the enteric aperture. The oesophagus of the Scyphula is at first
not a round but a flatly compressed tube (F). On each side of it a prolonga-
tion of the enteron penetrates between it and the ectodermal body wrall — the first
2 gastric pouches. Crosswise to these there is a further growth of 2 new pouch-
shaped invaginations of the enteron between the oesophagus and body wall. Thus
arise around the oesophagus the 4 gastric pouches of the Scyphula (G, mt). The
VOL. I K

130

COMPARATIVE ANATOMY

CHAP.

neighbouring endodermal walls of every 2 gastric pouches apply themselves to each
other and form the partition walls or septa (If, se), which are continued also, with
free axial edges, into the central gastric cavity, and there form the so-called gastric
ridges. The first 4 tentacles (E, t), arise around the oral disc as outgrowths of the
ectoderm and endoderm. The endodenn forms a solid axis in the tentacles, which
arise over the 4 gastric pouches, and which increase in number later.

In the Coral polyps the formation of the tentacles, gastric pouches, and partition
walls, if not exactly like the above, is still similar, and the endodermal axes of the
tentacles are hollow from the first.

The most important facts about the development of the Scyphula into the Acra-
specle Medusa have already been given on pp. 77, 106, 107, and 116.

The direct development of the Acraspede Medusa out of the egg is not yet suffi-

FIG. 99.— Development of the Scyphula of Aurelia Aurita out of the Planula, after GOtte.
A, Planula. J5, The same after it has attached itself. C, Invagination of the oesophagus. D, Break-
ing through of the enteric aperture. E, Scyphula. F, Transverse section through the stage C.
G, The same through stage D at the level of the line a-b. H, Transverse section through the Scyphula
(E) at the level a-b. I, the same at the level c-d. Black, ectoderm ; streaked, endoderm. s, (Eso-
phagus ; se, septa ; mt, gastric pouches ; t, tentacle ; m, gastric cavity ; st', st, septal funnels.

ciently known. Compare what was said of the development of the mesoderm of the
Coral (p. 100).

The development of the Ctenophora is in a certain way opposed to that of all
other Cnidaria. The total unequal furrowing of the telolecithal egg has already been
described, as also the formation of a two-layered germ by epibole. We here again
resume our account of the development.

After the 8 macromeres have formed a cap of micromeres which yield the whole
ectoderm, they themselves divide into 16 macromeres, which arrange themselves as
a plate on the vegetative side of the germ. Thereupon each of the 16 macromeres
constricts off one micromere on the under side, i.e. towards the vegetative surface
of the germ (Fig. 100). The 16 micromeres so formed are a part of the mesodermal
rudiment, the rudiment at any rate of the tentacle mesoderm. They may be con-
sidered, perhaps, as part of the endoderm, as an early product of separation from it.
The mesodermal axis of the Ctenophoran tentacle might then be compared with the
endodermal tentacle canals or tentacle axes of the other Cnidaria.

II

ONTOGENY OF CNIDAEIA

131

After the 16 mesoderm micromeres (me) have been formed, the macromere
plate becomes depressed, while, at the same time, the ectoderm cap spreads

FIG. 100.— Three early stages of development of a Ctenophore (Callianira), after Metschnikoff,
somewhat diagrammatic, cc, Ectoderm ; en, endoderm ; me, mesoderm ; d, gastric cavity ; st, oeso-
phagus (stomodseum).

out more and more towards the vegetative pole,
which we can already observe, besides the two
primitive germ layers, a mesodermal rudiment (It).
The latter, by the invagination of the 16 macro-
meres which represent the rudiment of the endo-
dermal gastro- canal systems, come to lie inside,
towards the gastric cavity. Later on it reaches
a position quite at the animal pole, beneath the
ectoderm (<7), its elements at the same time increas-
ing by division. The ectoderm at the vegetative
pole becomes depressed inwards round the blasto-
pore, and thus forms a stomodseum (st), the rudi-
ment of the oesophagus (erroneously stomach) of
the adult Ctenophore. The mesoderm at the aboral.
pole, viewed from this pole, assumes the shape of
a cross. Two opposite limbs of the cross stretch
out into the rudiments of the two tentacles, which
appear as outgrowths of the ectoderm. The ecto-
derm thickens at the aboral pole to form the
sensory body (Fig. 101, sk). Swimming plates
appear as fused cilia in 8 meridians, arranged in
pairs at the surface of the ectoderm. At first
only a few swimming plates are formed in each
row, but their number gradually increases. The
hollow spaces of the gastro-canals appear as fissures
penetrating from the gastric cavity to the circum-
ference. Between the endoderm and the stomo-
dseum on the one side, and the ectoderm on the
other, a clear mass of jelly is secreted. The various
elements which in the Ctenophora occupy the
jelly are, according to some investigators, derived
from cells migrating inwards from the ectoderm ;
according to others they are yielded by the
rudimentary mesoblast, which has been described
above.

A ccelogastrula thus arises, in

H .t

me/

FIG. 101.— Two further stages
of development of Callianira,
after Metschnikoff. en, Endoderm ;
me, mesoderm ; mei, mesenchyme ; t,
tentacle ; sk, sensory body ; d, gas-
tric cavity; st, oesophagus (stomo-
dseum) ; g, jelly.

132 COMPARATIVE ANATOMY CHAP, n

Literature.

E. Haeckel. Die Gastrula und die Eifurchung, in : Jenaische Zeitschrift. Bd.
IX. 1877.

F. M. Balfour. Comparative Embryology. 2 vols. London, 1880-81.

A. C. Haddon. An Introduction to the Study of Embryology. London, 1887 (Tr. )
Compare further the bibliography on p. 117.

CHAPTEE III

The organisation of the Flatworms — The life -history of the Cestoda and
Trematoda— The development of the Marine Planaria — The influence of
Parasitism.

THE TRIED KACE OR PHYLUM OF THE ANIMAL KINGDOM.

PLATODES— FLATWORMS.
Systematic Review.

CLASS I. Turbellaria.

Free-living Platodes, with ciliated body epithelium.

Order I. Polycladidse, Marine Planaria.

Large Turbellaria with flat, leaf-shaped body, with numerous ovaries and testes,
without vitellaria, mostly with 2 separate genital apertures. The intestine very
much branched, the branches anastomosing.

Tribe 1. Cotylea.

With ventral sucker. Mouth and pharynx in the middle of the body or further
forward. Tentacles wanting or found at the anterior edge of the body. Anonymus,
Thysanozoon, Yungia, Cydoporus, Stylostomum, Eurylepta, Prosthiostomum.

Tribe 2. Acotylea.

Without sucker. Mouth and pharynx in the middle of the body or further back.
Tentacles wanting, or 2 dorsal neck -tentacles present. Planocera, Leptoplana,
Trigonoporus, Cestoplana.

Order II. Tricladidae (Fresh-water, Land, and Marine Planaria).

The generally large body is long and flat. Mouth and tubular pharynx behind the
middle of the body. One common external genital aperture with 2 germaria and
numerous testes and vitellaria. The alimentary canal consists of an anterior unpaired
portion and of two lateral posterior limbs, which are again provided with side branches.
Planaria, Dendroccelum (in fresh water), Gcodcsmus, Bipalium (on land), Gunda seg-
mentata (marine).

Order III. Rhabdoccelidse.

In fresh and salt water. Small forms. Intestine, when clearly distinguishable,
a straight tubular csecum without or with very slightly pronounced lateral branches.
Body elongate, mostly cylindrical, more rarely flatly compressed.

134 COMPARATIVE ANATOMY CHAP.

Tribe 1. Alloioccela.

Intestine sharply separated from the parenchyma, often with short lateral sacs.
Numerous small testes. Female germ glands either 2 ovaries or 2 germ-vitellaria,
or separate germaria and vitellaria. Monotus, Plagiostoma, Vorticeros.

Tribe 2. Rhabdoccela.

Intestine sharply separated from the parenchyma, without lateral diverticula.
In the parenchyma there occur spaces, generally of considerable size, filled with
fluid, which form a sort of coelome. 2 large testes. Female germ glands either

1 or 2 ovaries, or 1 or 2 germaria and vitellaria, or 2 germ - vitellaria. Vortex,
Graffilla (parasitic), Macrorhynchus, Mesostoma, Prorhynchus, Microstoma arid Steno-
stoma (in these two genera the sexes are separate), Macrostoma.

Tribe 3. Accela.

"Without distinct intestinal canal ; with digesting parenchyma. "Without excre-
tory organs ; with numerous very small testes, and 2 ovaries. Nadina, Conwluta.

CLASS II. Trematoda.

Parasitic unsegmented Platodes, without covering of cilia, mostly with forked
intestine. Mouth and pharynx at the anterior end of the body. 2 testes, 1 ger-
marium, and 2 vitellaria branched or divided into numerous lobes.

Order I. Ectoparasitica (monogenetic).

With at least 3 suckers. Development direct, without alternation of genera-
tions, or heterogeny ; life-history simple. Tristomum, Diplozoon (two young, not
yet sexually mature animals early fuse in the shape of a cross, and only become
sexually mature in this condition) Polystomum, Gyrodadylus.

Order II. Endoparasitica (digenetic).

"With at the most 2 suckers. Life-history with heterogeny. Distoma hepaticum
(life-history, p. 169, Fig. 119), D. lanceolatum, both in the bile ducts of the sheep.
Distoma isostomum, Gyncecophorus hcematobius, in the blood in the portal veins
of man (in Africa), sexually separate ; the male with a channel on the ventral side
for the reception of the female. Amphistoma, Monostomum. The sporocysts and
redise usually live in water snails ; the sexual generation mostly in the intestines of
vertebrates.

CLASS III. Cestoda (Tapeworm).

Endoparasitic Platodes without cilia and without intestine, with numerous testes,

2 germaria, and 1 or 2 vitellaria broken up into lobes. With organs of adhesion only
at the front end of the body.

Order I. Monozoa.
Unsegmented individuals. Amphilina, Caryophyllceus, ArcMgetes.

Order II. Polyzoa.

Cestode stocks arising by strobilation : segmented tapeworms. With scolex and
chain of proglottides. Phyllobothrium, Tetrarhynchus, Ligula (very indistinctly seg-
mented). Bothriocephalus latus; broad tapeworm (Fig. 117, C, p. 164) in human
intestine. Over 3000 proglottides. With 2 pit-like suckers in the head. Genital
apertures on the flat surfaces. Larva ciliated in water. Scolex -like young form
in flesh of the Pike, of the Burbot, and perhaps of other fish. Schistocephalus.
Tricenophorus. Tcenia; with 4 suckers. T. Saginata (mediocancllata) (Fig. 117, A,

in PLATODES— GENERAL REMARKS 135

p. 164), without hooks on the rostellum, with over 1000 proglottides, sexual apertures
placed on the edge ; in the human intestine. The finn, a Cysticercus, lives in the
muscles of the ox. T. solium (Fig. 117, B, p. 164), with double circle of hooks on the
rostellum. Sexual apertures placed on the edge. About 800 'proglottides. Finn:
Cysticercus cellulosce in flesh of the pig (Fig. 120, p. 172). T. serrata, in intestine
of the dog. Finn : Cysticercus pisiformis, in the liver of rabbits and hares. T. crassi-
collis, in the intestine of the domestic cat and of other Felidse. Finn : Cysticercus
fasciolaris, in the liver of the mouse. T. cucumerina, in intestine of the dog, and
scolex-like finn in the body cavity of the dog-louse. T. ccenurus, in intestine of the
dog. Finn : Ccenurus cerebralis, in brain and spinal marrow of the sheep, causing
"staggers." T. echinococcus, in the small intestine of the dog; finn, Echinococcus
veterinorum in the liver of man and in that of different domesticated ungulates.

I. General Remarks.

The race of the Platodes forms a very natural division of the
animal kingdom, containing the three classes of the Turbellaria, the
Trematoda, and the Cestoda. Their fullest development is shown in the
organisation of the free living Turbellaria, which move forward by
creeping or gliding, while in the Trematoda and Cestoda many degenera-
tions and simplifications have taken place in consequence of parasitism.
The organisation of the Turbellaria more than any other therefore
supplies us with the clue to understand the systematic position of the
Platodes and their affinities.

Of all Metazoa it is the Platodes, and especially the Turbellaria,
whose general structure most agrees with that of the Ccdentemta, i.e.
of the Cnidaria. They might almost be placed near the Ctenophora
as creeping Cnidaria.

On account of the absence- of a eoelome between the alimentary
canal and the integument, and of a separate blood-vascular system,
the function of circulation is performed, as in the Cnidaria, by the
digestive system as gastro-canal apparatus. An anus is wanting here
as there. The Platodes show, however, on the other hand, many
essential differentiations of organisation which we miss in the Cnidaria,
and which are in great part to be referred to the creeping mode
of life. All the Platodes are bilaterally symmetrical ; we can dis-
tinguish in their bodies an anterior and posterior, an upper and lower,
a right and left. The aboral surface of the body of the Cnidaria
becomes in the Platodes the dorsal surface, the oral the ventral
surface, in whose centre the mouth originally lies. The sensory organs
collect chiefly at that part of the body which goes first in creeping,
i.e. at the anterior end, and the principal part of the central nervous
system also, the brain, which originally lay at the aboral pole, i.e. at
the centre of the dorsal surface, has, following the sensory organs,
moved more or less far forward.

Those organs and systems of organs which, in the Cwlenterata,
showed the tendency to leave the body epithelium and deposit
themselves below it, forming a middle layer, have in the Platodes
become markedly mesodermal, viz. (apart from the connective tissue)

136 COMPARATIVE ANATOMY CHAP.

the sexual organs, the musculature, and the nervous system. The
musculature is arranged under the body epithelium in a muscle layer
whose elements have transverse, longitudinal, and diagonal courses.
Dorso-ventral muscle fibres stretch between the ventral and dorsal
surfaces. The whole arrangement of the musculature seems adapted
for the creeping motion. The nervous system forms a plexus of nerves
lying in or on the muscle layer, which is more strongly developed
on the ventral than on the dorsal side. In this plexus single stronger
nerve trunks are more clearly developed, and they meet together from
various directions in one central point, the brain. In very many
Platodes this nerve plexus forms the so-called ladder nervous system,
in which we distinguish the following parts : — (1) the brain, lying at
the anterior end of the body ; (2) the two principal longitudinal trunks
proceeding out of it and running on the ventral side ; (3) transverse
commissures which connect these latter.

The so-called water-vascular system is very characteristic of the
Platodes. It is a system of fine canals, on the one hand ramifying in
the mesoderm, and on the other emerging"externally, which has certainly
an excretory, and perhaps at the same time a respiratory purpose. In
the Cnidaria no such system has been observed.

The Platodes are hermaphrodite. Besides the sexual reproduction
by means of fertilised eggs, there is also (in Trematoda) parthenogenetic
reproduction and (in Turbellaria and Cestoda) asexual multiplication
by fission or gemmation.

For the comprehension of the relation of the Platodes to the Cnidaria, the know-
ledge of two animal forms, which have been considered to be intermediate forms between
the Ctenophora and the Turbellaria (Polyclada), is necessary. Only one specimen
of each has till now been described. One of these forms is Cceloplatia Mecznikowi,
the other Ctenoplana Kowalevskii. Unfortunately our knowledge, especially of the
first form, is very insufficient. Their sexual organs and their development are
unknown, so that we cannot be sure whether we have to do with young stages or
with adult animals. But in any case both forms are of the greatest interest.

Cceloplana is a little animal about J of an inch long and £ broad, whose
appearance quite coincides with that of a Polydad. The body is flatly compressed
and ciliated all over ; it creeps on the ventral surface. In the centre of the dorsal
surface lies a vesicle with a mass of otoliths. Near it on each side, right and left, is
a long tentacle feathered on one side, which can be withdrawn into a special sheath.
In the middle of the ventral side lies the mouth. The gastro-canal system consists
of the quadruply-lobed stomach and numerous anastomosing canals radiating from
it. From the stomach 2 canals rise towards the dorsal surface of the body, where
they apparently end blindly in front of and behind the otolith vesicle.

Ctenoplana has in general the same body form as Cceloplana ; but besides the
general ciliation this animal also has on the dorsal surface eight short rows of stiff
plates arranged like a rosette ; these correspond with the ciliated or rowing plateis of
the Ctenophora, and lie in special grooves, out of which they can be protruded. The
arrangement of the gastro-canal apparatus is like that in the Cceloplana. In the
middle of the dorsal surface occurs a formation similar to the sensory body of the
Ctenophora. At the base of the depression containing the otoliths there is on each
side a nerve centre with nerves proceeding from it, and near these on each side a

in PLATODES—BODY FORM 137

solid tentacle with short lateral branches. In the neighbourhood of the tentacle on
each side is found an aperture leading into a system of canals which branch in the
body parenchyma, and which the discoverer of Ctenoplana compares with the water
vascular system of the Platodes. Under the body epithelium lies a basal or skeletal
membrane, under this a layer of longitudinal muscles, and under this again a layer
of transverse muscles. Besides these there are dorso-ventral muscular fibres branched
at both ends, which adhere on one side to the dorsal, and on the other to the ventral
basal membrane. There are special bundles of muscular fibres for the protrusion and
withdrawal of the ciliated plates.

These two forms agree with the Ctenophora chiefly :

1. In the possession of an aboral sensory body.

2. In the possession of 8 rows of ciliated plates (Ctenoplana}.

3. In the possession of feathered tentacles.

4. In the general structure of the body.

Cosloplana and Ctenoplana are not yet bilaterally symmetrical. The chief axis runs,
as in the Ctenophora, from mouth to sensory body. It is very much shortened. The
lateral plane runs through both tentacles ; the median or sagittal plane stands at
right angles to it. Each of these planes divides the body into 2 similar halves.
If such forms were always to move forwards in the direction say of its median plane,
and if at this end special sensory organs were to develop, or the sensory body come
more forward, they would become bilaterally symmetrical. Only one plane, viz.
the median plane, would divide the body into 2 exactly similar halves. We could
then not only distinguish upper and lower, but anterior and posterior regions of the
body.

Codoplana and Ctenoplana agree with the Polyclada : —

1. In the flatly compressed form of body, and in the capacity of moving forward
by creeping.

2. In the general ciliation of the body.

3. In the possession of a skeletal membrane (Ctenoplana}.

4. In the possession of a dermal musculature, consisting of a longitudinal and a
circular layer.

5. In the possession of dorso-ventral muscle fibres branched at both ends.

6. In the general arrangement of the gastro- canals.

7. In the possession of 2 (in Polyclada, however, unfeathered) dorsal tentacles
and of a dorsal nerve centre (?).

8. In the possession of a water-vascular system (Ctenoplana?}.

The peculiarities mentioned under 1-5 may be considered as due to the creeping
mode of life.

Cosloplana and Ctenoplana are distinguished both from the Ctenophora and the :
Polyclada by the want, of an ectodermal oesophagus, — at least no such oesophagus
has been described.

II. The Body Form.

Most of the Platodes are, as their name indicates, more or less
The Polyclada are leaf-shaped. In these all intermediate forms
between the broad oval and the long ribbon occur. The Triclada are
mostly long, lancet-shaped, with dorsal surfaces slightly arched. Among
the lanl Triclada forms occur of great length. In the Rliabdocoela
great variety prevails ; there are flat, disc-shaped, plano-convex, and
often spindle-shaped animals. Among the Trematoda the ectoparasites
and a few small endoparasites (e.g. Distoma hepaticum, D. lanceolatwn)

138 COMPARATIVE ANATOMY CHAP.

are flat and compressed. Most of the endoparasitic Distonm species,
on the contrary, are more or less cylindrical. The Ccstoda or tape-
worms are ribbon -like, and consist chiefly of a row of consecutive
segments progressively increasing in size posteriorly (proglottides). In
front, where the body thins away, the segmentation is indistinct. The
thin neck portion swells into a knot-like tapeworm-head (scolex), which
is provided with organs of adhesion.

III. The Outer Body Epithelium.

A distinct body epithelium occurs only in the Turlellaria. In the
Trematoda and Cesfoda it is replaced by a resistant, elastic, cuticular
membrane, which is generally perforated by very fine pores. It is
indeed not impossible that this integument itself is a much modified
epithelium. The body epithelium of the Turlellaria is ciliated. The
ciliary motion serves principally (1) for Respiration (by the main-
tenance of a constant circulation of the water over the naked surface
of the soft animal) and (2) for Locomotion (especially of the smaller
forms).

Unicellular dermal glands open in the skin. In special glandular
cells of the skin the so-called rod or rhabdite cells, spindle-shaped
glandular secretions occur ; the so-called rods or rhabdites, although
found in other divisions, are specially characteristic of the TurbcUaria.
Whereas the rod-glandular cells in most Polyclada lie in the epithelium
itself, in most Tridada and PihaMoccelidce they sink deep into the
parenchyma, and only remain connected with the epithelium by means
of long thin processes (rod -passages) in which the rhabdites are
conveyed outwards and deposited in the epithelium. Typical stinging
capsules also occur in the epithelium of many Turlellaria. There are
besides numerous intermediate forms between true rhabdites and true
nematocysts.

IV. The Gastro-eanal System.

This has entirely disappeared in the Oestoda, These endoparasites
are nourished endosmotically by the juices which surround them.

In the other Platodcs the gastro-canal system is well developed, and
shows in a few divisions modifications almost as important and
characteristic as those in the Ciridaria. We can everywhere dis-
tinguish in it two principal parts: (1) an oesophagus or pharyngeal
apparatus, which comes from a depression of the outer integument and
is consequently lined with ectodermal epithelium ; and (2) the endo-
dermal intestine or gastral apparatus. We will treat of these two
parts separately.

I. The pharyngeal apparatus. This opens externally through
the mouth, and internally through an enteric aperture into the gastral
apparatus.

Ill

PLATODES— GASTEO-CANAL SYSTEM

139

The position of the mouth is in no animal race so extraordinarily
variable as in the Platodes. We hold that the original position of the
mouth is the central one, in accordance with the views set forth in the
general remarks, and with its position in Ctenoplana and Codoplana.
This position is found only in Ttvrbdlaria, i.e. among the Polydada in
the families of the Anonymidce, Planoceridce, and most Leptoplanidce, and
among the Rliabdoccelidce in a few Accela and Mesostomidce. In the
remaining Platodes the mouth is placed either more to the front or
to the back, without in any case reaching the extreme anterior or
posterior end of the body. The mouth lies somewhat further forward
than the middle among the Polydada in the Pseudoceridce, among the

FIG. 102.— A-D, Diagrammatic representation of the Pharyngeal Apparatus of the Turbel-
laria. A, Of Convoluta ; B, of Mesostoma ; C, of Planocera ; D, of Prosthiostomum. de, Dorsal
body epithelium ; ve, ventral body epithelium ; o, mouth ; dm, entrance to intestine ; pt, pharyn-
geal pouch ; ph, pharynx ; d, intestine ; dep, intestinal epithelium ; s, resophagus ; p, parenchyma ; dt,
dorsal oesophageal pouch ; vt, ventral cesophageal pouch ; ms, muscular lamella.

Rhabdoccelidce in many Accela, and also in Mesostomidce, Proboscidea, and
Plagiostomida. It is found near the anterior end of the body in the
Euryleptidcv and Prosthiostomidce, among the Polydada, in many Rhabdo-
ccela, and all Trematoda. It lies somewhat behind the middle in a
few Leptoplanidce among the Polydada, all Tridada, and many
Rhaldoccelidce (especially in the Monotidce). It is met with near the
posterior end of the body in the Cestoplanidce among the Polydada, and
in many Plagiostomidce among the Alloioccda.

The structure of the pharyngeal apparatus is very varied in the
Platodes. We can only briefly describe the chief types.

The pharyngeal apparatus of the Convolutidce, the Microstomidce, and
Macrostomidce among the Rhaldoccelidce, is distinguished by its great
simplicity and its embryonic condition (in most Accela it is even

140 COMPARATIVE ANATOMY CHAP.

entirely wanting). In the 3 families mentioned it consists of a
simple resophageal tube between mouth and intestine (Pharynx
simplex, Fig. 102, A}. This simple ossophagus becomes at first
complicated by the massing of definitely arranged muscles around it.
The muscular wall of the ossophagus, then, almost always projects in
various ways more or less far into the lumen of the oesophagus, so that
we can now distinguish in the pharyngeal apparatus 2 chief parts : (1)
the cesophageal or pharyngeal pouch, and (2) the muscular cesophageal
bulb or pharynx which projects into it. When the cesophageal pouch
is not spacious, and the pharynx with its free inner surface projects
only a short way into it, the latter is generally round or barrel-
shaped, and is sharply marked off from the body parenchyma which
surrounds it by a layer of muscle (Pharynx bulbosus, Fig. 102, B).
In this form we meet it in nearly all Rhabdoccda, in the Plagiostomidce
among the Alloioccela, and further in all Trematoda.

In very many Turbellaria the pharynx projects as a circular fold
into the mostly spacious pharyngeal sheath or pouch, and takes, like
the latter, very various shapes ; it is in this case never sharply severed
from the surrounding parenchyma by a muscle layer (Pharynx
plieatus). In all Polydada, with the exception of Euryleptidce and
Prosthiostomidce, the pharyngeal pouch is very spacious, and often has
secondary pouches, which again are occasionally branched ; and the
pharynx is a flat and broad band which hangs as a circular fold from
the sides of the pouch (Fig. 102, C). Such a pharynx can be extended
through the oral aperture, and, spreading out, envelop its prey on all
sides, as in a sheet. In the Euryleptidce and Prosthiostomidce among
the Polydada, and in all the Triclada, and in the Monotidce among the
Alloiocoela, this circular fold becomes a more or less extended cylin-
drical muscular tube, which projects freely into the equally cylindrical
pharyngeal pouch from its base. By contraction of the circular
musculature this tube elongates and passes out through the oral
aperture (Fig. 102, D).

The relation of the pharynx plieatus to the pharyngeal sac is
similar to that of the variously-shaped oral or gastric peduncle of the
Acraspede Medusa to the subumbrellar cavity into which it projects.

The musculature of the Platode pharynx consists of one or more longitudinal
and circular muscular layers, and of muscle fibres arranged radially round the axis of
the pharynx.

All over the free surface of the pharynx, and chiefly at its free end, unicellular
glands (salivary glands) open. These glands lie either in the pharynx itself, as in
the pharynx bulbosus, which is sharply marked oif from the parenchyma, or, as in the
pharynx plieatus, scattered about in the parenchyma round the place of insertion of
the pharynx. In the latter case they send only their long and thin processes
(efferent ducts) into the pharynx.

The following is the rule for the position of the pharynx and the pharyngeal sac.
When the mouth lies in the middle of the body the enteric aperture is found directly
above it. The axis of the pharynx and its pouch then stand perpendicularly to the
ventral surface. If the mouth lies to the front the enteric aperture lies behind it ;

Ill

PLATODES—GASTKO-CANAL SYSTEM

141

the free end of the pharynx is directed forwards and is extended forwards. The
opposite is the case when the mouth lies behind the middle of the body. The
generally narrow round enteric aperture leads out of the pharynx or pharyngeal
pouch into the endodermal gastral system.

II. The endodermal gastral system must be separately described
for the various divisions of the Platodes.

An anus is everywhere wanting, as in the Cnidaria.

FIG. 103.— Intestinal and nervous systems of the Planocera. t, Tentacle ; g, brain ; pli,
pharynx ; o, mouth ; M, posterior end of the main intestine, the greater part of which is covered
by the pharynx.

Turbellaria.

A. Polyelada (Fig. 103). — The gastral system here is very
similar to that of the higher Cnidaria (Medusce and Ctenophora). It
consists of (1) the main intestine and (2) the gastro-canals or intestinal
branches. The longer the body the longer is the generally spacious
main or stomach intestine. Where the pharyngeal apparatus is
found in the middle of the body, the main intestine lies just above
it ; where the pharynx lies behind, the intestine in almost every case is
found in front of it. If the pharynx is placed in front the main intestine
lies nearly always behind it. From the stomach intestine the gastro-
canals arise in varying numbers. They perforate the body parenchyma,
penetrating towards the edge of the body in a horizontal direction on
all sides, and freely branching or anastomosing with each other on the
way. Pairs of gastro-canals can be distinguished, and there is one
unpaired canal. The paired are found in 4 pairs at least, and often,

142

COMPARATIVE ANATOMY

CHAP.

especially in long forms, in very numerous pairs. The unpaired
canal runs forward from the front end of the stomach intestine
in the middle line of the body.

In a few Polydada the gastro-canals open on the dorsal surface of
the body (Yungia) through pores, or else on the edge of the body
(Cydoporus). In Thysanozoon diverticula of the intestinal branches
run into the lobate processes of the dorsal surface.

B. Trielada (Fig. 104). — A distinction between main intestine and
intestinal branches cannot here be made. Three gastro-canals start

FIG. 105.— Intestinal
and nervous systems of
Mesostoma (Rhabdo-
ccele). ph, Pharynx; d,

intestine.

FIG. 104. —Intestinal and nervous
systems of a fresh-water Triclade.
va, Anterior median intestinal branch ;
ha, posterior intestinal branches ; ph,
pharynx ; pt, pharyngeal pouch.

direct from the intestinal aperture ; 1 unpaired, which runs for-
ward in the median line and often possesses lateral offshoots which
again branch, and 2 paired, which run backward at the sides of the
pharyngeal pouch and give off lateral branches towards the exterior.

C. Rhabdoecelidse. — The gastro-canal system is reduced to a simple
sac-like or tubular intestine (Fig. 105), which runs in the middle line
of the body. This occasionally (especially in the Monotidce) has
numerous short lateral diverticula.

Ill

PLATODES— ORGANS OF LOCOMOTION

143

7R4

In the so-called Acoda the alimentary canal is represented by a
mass of star-shaped branched cells, in which no intestinal cavity can
be demonstrated. Food enters this mass direct through the mouth or
oesophagus ; it has been described as " digesting parenchyma."

Trematoda. — The gastro-canal system of the Trematoda most
resembles that of the Triclada. Since the
mouth and pharynx lie to the front, the
intestinal aperture leads into the anterior
end of a short unpaired median portion of
the intestine called the oesophagus ; this
soon splits into two forked branches which
run backward (Fig. 106). In the broad leaf-
shaped Trematoda to which most of the
ectoparasitic forms and also the liver fluke be-
long, lateral branches, which again divide, run
mostly outwards from these forked branches.

In certain forms (Stichocotyle, Aspidogaster)
the intestine runs back as an unpaired
median caecum. When a median sucker is
developed at the anterior end (oral sucker)
the mouth lies in its base.

The intestinal epithelium of the Turbellaria is
covered, over larger or smaller areas, with cilia.
Intracellular taking in of food is very common
in the Turbellaria. The musculature of the gastro-
canal system is on the whole very feebly developed,
and consists of longitudinal and circular fibres. In
the Polydada single circular muscle fibres succeed
each other on the intestinal branches at regular
intervals, so that in consequence of the constrictions
thus formed the intestinal branches assume the aspect
of strings of beads.

As an anus is wanting, the fecal masses are emptied

out through the mouth. The gastro-canal system of

FIG. 106.— Intestinal and nerv-
the Turbellaria may, besides the functions of diges- QUS systems of Distoma isosto-

tion and circulation, also have a respiratory signifi- mum (Trematode), after Gaffron.

ms, Oral sucker ; ph, pharynx ;
gd, forked branches of the intesti-
nal canal ; gc, cerebral commissure ;
dn, dorsal longitudinal nerves ; sn,
lateral longitudinal nerves ; vn,

ca"nce.

Y. Supporting Organs, Passive Organs
of Locomotion.

. ventral longitudinal nerves; Is,

Inese are in most sort and delicate .rla- ventral sucker,
todes little developed. In the Turbellaria,

at least in the Polydada, the basal membrane, which is resistant and
tolerably firm and elastic, plays the part of a supporting membrane,
to which the dermal musculature is closely applied, and in which
the dorso- ventral muscle fibres are inserted. The same part is also
certainly played by the cuticle of the Trematoda and Cestoda. Besides
this, the more or less compact parenchyma lends the body a firmer

144 COMPARATIVE ANATOMY CHAP.

structure and a greater consistency, like the gelatinous tissue in
the Cnidaria.

VI. The Musculature.

The collective muscle elements of the Platodes may be brought
into 2 chief groups: (1) the general body musculature, and (2) the
special musculature of the organs, e.g. of the intestinal canal of the
copulatory organs, etc. The latter cannot here be taken into considera-
tion, as the musculature is adapted in every one of the extraordinarily
numerous cases to the special activities of the organ.

The body musculature also is by no means so uniform throughout
the race as to make a generally applicable scheme possible. It
again falls into (1) the dermal musculature and (2) the dorso-ventral
musculature. The first lies under the basal membrane of the integu-
ment or under the outer cuticle, the second runs transversely through
the parenchyma between the various organs and connects opposite
points of the basal membrane or the cuticle.

1. The dermal musculature is composed of layers, which are
generally clearly separated. In each of these layers all the
fibres run in a certain direction. We can distinguish longitudinal,
tranverse, and diagonal fibre layers. The diagonal fibre layer
is naturally always double. The longitudinal and the transverse
layers may also be double. We find the largest number of layers
— 5 or 6 — in the Polydada ; in the Triclada their number is
smaller, we here find outer circular and inner longitudinal fibres,
between which diagonal fibres may be intercalated. The same is the
case in the BJiabdoccelidce, in which the musculature is weaker than in any
other Turbettaria. It is generally much weaker on the dorsal than on
the ventral side, on which the animals creep. Certain dorsal muscle
layers may be altogether wanting. In the Turbellaria the diagonal
muscle layer seems always to lie between the others.

The sequence of layers in the dermal musculature of the Trematoda
is as follows : the circular or tranverse muscle layer lies externally ;
then follows a strong longitudinal muscle layer ; and inside comes the
diagonal muscle layer. In the Cestoda the diagonal muscle layer is
replaced by a strong inner circular muscle layer, lying pretty deep
under the skin, and divided from the outer circular layer and the
longitudinal muscle layer by a layer of parenchyma.

2. The Dorso-ventral or Sagittal Musculature. — Its fibres are
branched at both ends (Fig. 47, d, p. 47), and run through the
parenchyma from the dorsal to the ventral surface. Where intestinal
diverticula are developed, the fibres naturally run between them
as muscle septa, filling up the spaces. Where, as in the broad disc-
shaped Polyclada, the ramifying intestinal branches of the stomach
intestine radiate out towards the circumference on all sides, the septa
project more or less far from the latter towards the former, and where

in PLATODES— ORGANS OF ADHESION 145

numerous pairs of lateral intestinal branches succeed each other more
or less regularly, they are separated by equally regular muscle septa
or dissepiments. This is the case in elongated Polydada and Triclada,
and especially in the marine Triclada (Gwnda), in which the lateral
intestinal diverticula are unbranched. In all cases, when sexual maturity
is attained, the formation of septa is more or less obliterated by the
development of male and female germ glands, which are generally
placed between the intestinal branches.

VII. Organs of Adhesion.

These are very widely spread among the Platodes. One division
of the Polydada, that of the Cotylea, is characterised by the possession
of a muscular sucker, which lies about the middle of the ventral
surface, always behind the mouth and the genital apertures. By
means of this sucker the Cotylea often temporarily attach themselves
to some object on the sea bottom. Besides this, very many
Turbellaria of the most various divisions, possess special adhesive
cells with rough surfaces which serve for attachment.

In the parasitic Trematoda and Cestoda the organs for adhesion are
specially strong and variously developed. They here serve to fasten
the body either outwardly to the skin, or inwardly to the intestinal
wall of the animal inhabited, or host. They are principally pit- and
disc-like suckers, with or without stalks, whose number, form, and
arrangement are of the greatest importance in classification. We
select only the most important.

In the digenetie Trematoda there are at. the most 2 pit -like
suckers (Fig. 106), one of which, in whose base the mouth nearly
always lies, is found at the front end of the body as oral sucker (ms).
The other is either wanting (Monostoma), or lies at a variable distance
from the anterior end on the ventral side (bs) (Distoma), or at the
posterior end of the body (Amphistoma).

In the monogenetie Trematoda there are often 2 suckers or
sucker-pits on each side of the mouth. Besides these, at the posterior
end of the body, there is a very large stalked ventral sucker (Tristoma) ;
or the hinder portion of the body becomes transformed into a large
sucker-disc, which again may carry sucker-pits in varying numbers and
in symmetrical or asymmetrical arrangements (Polystoma).

In the class of the Cestoda also the presence of suckers or sucker-
pits is the rule (Fig. 117, p. 164). They here always lie at the
foremost end of the body, singly (Amphilina), two in number (BotTirio-
cephalus, Schistocephalus, Tricenophorus), or four in number (Tceniadce,
Tetrarhynchidce, Tetraphyllidce). In the Tetraphyllidce they often have
long stalks.

As a further strengthening of the apparatus for adhesion there are
often hooks, ridges, teeth, etc., as in the ventral suckers or on the
adhesive discs of many monogenetie Trematoda, in the suckers of many
VOL. i L

H6 COMPARATIVE ANATOMY CHAP.

Tetraphyllidce, or at the foremost end of the body, the apical cone, or
rostellum of many Tceniadce.

In the Tetrarliynchidce, at the foremost end 6f the body there are
4 proboscides furnished with barbed hooks, which can be protruded
from special proboscis sheaths and withdrawn again by special muscular
retractors.

\

VIII. The Nervous System.

The nervous system in the Platodes is completely detached from
the body epithelium. Nearly all its elements lie in or close under the
dermal musculature.

In the Polydada (Fig. 103, p. 141) it consists of a close network
of finer or coarser nerves, which is spread in or under the dermal
musculature over the whole body, and which, like the musculature, is
less developed on the dorsal than on the ventral side. In this plexus
specially strong nerves occur, which, converging from all sides and
thus growing thicker, unite in a nervous centre, the brain, which lies
deep in the parenchyma under the anterior median intestinal branch
between the middle and the front of the body. The more elongated
the Polydada are, the nearer the brain lies to the anterior end, and
the more conspicuously do the longitudinal nerve trunks stand out
among the nerves converging towards the brain. The most strongly
developed of these longitudinal nerves are, in order of importance,
2 inner nerves, which run on both sides of the median line, 2
lateral and 2 dorsal. From the brain special nerves proceed to the
sensory organs.

The brain in all Polydada (with one single exception) lies in front
of the mouth. Only in Oligodadus (Euroleptidce) does it lie behind the
mouth over the beginning of the pharyngeal pouch. The two inner
longitudinal trunks in this genus embrace the front end of the pharyngeal
pouch, and are connected only behind the same by means of a web of
commissures. The brain, the first part of the longitudinal trunks, and
the first transverse commissure between these, thus form together a
ring which surrounds the anterior portion of the pharyngeal pouch.

In the Tridada (Fig. 104, p. 142) the brain always lies far
forward. The ventral inner longitudinal trunks are always specially
strongly developed, and are on the one side connected together by a
web of commissures, and on the other give off anastomosing branches
outAvards. In Gunda both the branches which proceed outwards and
the commissures are very simple and regular, and correspond in number
with the successive pairs of lateral branches of the intestine. We thus
have here a typical ladder nervous system.

In the Rhaldocodidce (Fig. 105, p. 142) we generally meet with
the nervous system in a simpler form. It consists of the brain, which
lies in the anterior portion of the body, and, proceeding from it, of the
two ventral longitudinal nerves, and of several smaller nerves which

in PLATODES— NERVOUS SYSTEM 147

spread out in the anterior part of the body. Commissures between
the longitudinal nerves appear in larger numbers only in the Monotidce
among the Alloiocoela, which show near relationship to the Triclada in
other points also. In Mesostoma Ehrenbergii there is a tranverse com-
missure behind the pharynx ; in Microstoma lineare two other nerves
besides the longitudinal nerves proceed backwards from the brain,
surround the pharynx, and unite behind it.

In the A coda, according to earlier investigators, a nervous system
was wanting ; but lately a somewhat complicated nervous system has
been proved to exist. It consists of 2 ganglia lying one behind
the other in the anterior part of the body (the front ganglion
being the smaller), and six longitudinal nerves, 2 inner, 2 middle,
and 2 lateral which run along the edge of the body. The 2
inner arise from the posterior ganglion, the 2 middle and the 2 outer
from the front ganglion. All these 6 nerves are connected by trans-
verse commissures which go off at right angles and themselves again
anastomose.

The nervous system of the Trematoda (Fig. 106, p. 143) is closely
connected with that of the Triclada and Polydada. It consists of a
brain, from which proceed, besides small nerves which run to the sides
and the front, 6 nerve trunks running backwards, viz. 2 dorsal, 2
inner ventral, and 2 outer ventral or lateral. The 2 ventral are
connected together and with the lateral by more or less numerous
transverse commissures, as are the lateral with the dorsal, and the dorsal
again together; these tranverse
commissures may again anastomose
with each other.

The nervous system assumes
this form especially in ectoparasitic
Trematoda (Tristoma) and in Distoma
isostomum. In many other Trema-
toda, however, perhaps in most
species of Distoma, the commissural
system appears to have degenerated ;
and of the 6 longitudinal nerves
only the inner ventral nerves seem FlG. io7.-Nervous System in the Scoiex

to be Strongly developed, as in of a Tapeworm (Taenia serrata) after Niemic.
f>iP Phn~hrlnrn>ln Of the 8 Beaker longitudinal nerves only the

4 on one side (dm, dnj) are depicted. sn,
The brain lies above the mouth Lateral nerves ; gc, chief or cerebral commis-

in all Trematoda ; when the mouth sure-

is placed in the pit of the oral

sucker, the brain extends as a transverse bridge over the front part

of the pharynx.

The nervous system of Ampliilina quite agrees with that of the
Distomidce.

The nervous system of the typical Cestoda (Fig. 107) consists of I
two lateral longitudinal trunks, which pass through the whole body

US COMPARATIVE AXATOMY CHAP.

and are connected Ijy means of a chief commissure (brain commissure)
at the foremost end in the scolex. There are generally several more
besides, often 8 longitudinal nerves, which, however, do not stretch
backwards beyond the scolex, and further other nerves, which proceed
to the suckers, the hooks, and in the Tetrarlnjnclndn: to the 4 muscular
proboscides, etc. Apart from the brain commissure, all these nerves
in the scolex are connected together in a complicated way by circular
square, polygonal, or cross-shaped commissures. In the segments
(proglottides) no commissures between the longitudinal trunks have yet
been observed.

The brain (cerebral ganglion, cerebral commissure) of the Platodes is not to be
considered exclusively as a central nervous .system, since more or less numerous
ganglion cells exist in the larger nerve trunks also, at those points where the nerves
and commissures branch oil'. Thus the ganglion cells in many Triclndd. are lound
in the ventral longitudinal trunks of the ladder nervous system, principally at the
points of departure of the transverse commissures. AVe are thus led to conjecture
that the double ganglia of the ventral ganglionie chain of the Venues have proceeded
from these points in a ladder nervous system.

The nervous system of the Platodes is in our opinion of the greatest significance
from the point of view of comparative anatomy, because it joins on to that of the
Cni'J.aria on one hand, and on the other is, in many groups, modified in a direc-
tion which points to the nervous system of the Vcrmcs, and also perhaps of the
lowest Molluscs. If we consider the matter without prejudice the following seems
characteristic of the nervous system of the Turlcllario. and Trcmatoda. (1) The
arrangement of the nervous system in the form of a nerve plexus all over under the
skin, dorsal ly as well as ventrally, in close connection with the musculature which
has to be innervated. In this condition we recognise a great similarity to the higher
Cnldfi/ria. (2) The development of a central or^an (brain). AVe have seen in the
C/iidnria that nerve centres are formed in close connection with the sensory
organs, which is comprehensible a priori, since it is in the nerve centres that the
junction of the sensory nerves with the motor nerves or nerve fibres takes
place. Xow the finer structure of the brain of the Platodes shows, most unmistak-
ably, that the brain is nothing more than a specially developed part of the nerve
plexus, in which motor and sensory nerves unite. The sensory organs become, in
connection with the development of bilateral symmetry, more and more localised in
the anterior end of the body, •/./•. in the end which goes first in creeping ; and there-
fore the centralised nerve plexus (or in other words the brain), in which the motor
and the sensory fibres unite, must, following the sensory organs, shift more and more
towards the anterior end of the body.

In Awln we find '2 cerebral centres ; in Cextniln the many often very complicated
commissures between the nerves in the scolex must collectively be considered
as the brain ; in certain land Tr!<-l<td<t. the only part which can be called the brain
is a not very sharply demarcated tract of the, longitudinal trunks at the anterior end
of the body, in which the transverse commissures lie specially crowded together, and
into which also the sensory nerves enter.

Bearing in mind the fact that in the dtnuiplmrd, in (!n'/ojil(r.nr/., and in Ctcvo]>l«nn
the sensory body is developed in the middle of the dorsal surface, we must consider
this as the original position of the brain. As a necessary result of the development
of this central portion, the peripheral nerves, the nearer they approach to this part,
unite into increasingly massive trunks, which finally enter the brain. In consequence
of the position of the brain at the anterior end of the body, the nerves proceeding

in PLATODES— SENSORY ORGANS 149

backward are always the thicker nerves, while the lateral and front nerves are the
shorter and thinner. The greater development of the musculature on the creeping
surface has as a consequence the greater development of the ventral trunks. The
elongation of the body further causes a retrogression of the outer longitudinal nerve
trunks, so that finally the 2 inner longitudinal trunks remain as the main trunks,
which from the first are very strong, since they have to innervate the most import-
ant muscular organs which lie in the median line between them (pharynx, copulatory
apparatus, suckers). Thus in the end the regular ladder nervous system of Gunda
is deducible from a general nerve plexus in the integument, determined by the elonga-
tion of the body, the localisation of the sensory organs at its anterior end, and the
strong development of the musculature on the ventral surface.

The brain in the Platodes appears the more evidently to consist of 2 lateral halves
or ganglia connected by a bridge of fibres, the more strongly the % longitudinal trunks
of the nervous system are developed in comparison with the whole nervous system.

IX. The Sensory Organs.

The Platodes possess sensory organs, which according to their
structure may be described as eyes, auditory organs, or organs of touch.
To these belong also ciliated pits, whose function is unknown. Many
consider them to be olfactory organs. The development of the sensory
organs stands in direct relation to the mode of life. They are besf\
developed in the free-living Turbettaria. In the parasitic Tremadotal
their degeneration begins, and the Cestoda no longer possess any?
specific sensory organs.

A. Eyes.

These are found in most Turbellaria and ectoparasitic Trematoda.
They are wanting in adult endoparasitic Trematoda, but, on the other
hand, are found in their young stages,' which are free-living, at least
for a time.

All Polydada possess eyes in large numbers, often as many as several
hundreds. In their arrangement the following points are worthy of
note. A group of eyes is always found above the brain and in the
tentacles. Many forms possess, besides these, other eyes at the anterior
margin of the body or all round the margin. In the Triclada there
are either 2 eyes near the anterior end, or numerous eyes along the
margin of the whole body, or along its anterior margin. In the Rhabdo-
ccelidce there are generally 2 or 4 eyes (less frequently one unpaired
eye) directly on or over the brain ; this is also the position of the
4 eyes of the ectoparasitic Trematoda and the 2 eyes of the free-living
larvae of the endoparasitic Trematoda.

In all the Platodes (except the Accda and Microstomidce) the eyes
lie under the epithelium in the parenchyma ; in the Rliabdoccela and
Trematoda dipectly on or in the brain. Yet in the case of the Polydada
at least, it can be ontogenetically proved that the eyes at their first
appearance arise in the ectoderm of the embryo and only secondarily
sink below the surface.

150 COMPARATIVE ANATOMY CHAP.

The eyes are simple pigment spots in many Rhabdoccdidce ; in others
a refractive body is added. The eyes of the Triclada, Polyclada, and
ectoparasitic Trematoda are somewhat more complicated. They consist
of a pigment cup (often unicellular), at whose aperture lie one or more
nerve or retinal cells as perceiving elements. In the cavity of the eye-
cup lie homogeneous non-nucleated rods or clubs, apparently processes
of the retinal cells. A fine optic nerve is connected with the group
of retinal cells. In the Polyclada, where there are numerous eyes, the
sensory nerves from the brain branch, sending a single branch to each eye.

B. Auditory Organs.

These are not widely spread among the Platodes. They are only
found in the Rhabdocoelidce, and there almost exclusively in the division
of the Acoela, and among the Alloiocoda in the family of the Monotidce.
They are always found singly, and lie on the brain, and consist of a
small spherical vesicle filled with fluid ; an otolith or auditory stone
(composed of carbonate of lime and an organic substratum) is also
enclosed in this vesicle.

C. Organs of Touch.

These are universal in the Turbellaria. In the first place the skin
is everywhere very sensitive. This sensitiveness is caused by the
presence of delicate tactile hairs or tufts of immobile tactile hairs,
which are found in great numbers principally at the exposed parts of
the body, especially at the anterior edge and on the tentacles. The
tentacles may be considered in a special manner to be organs of touch ;
they are present in very many Polydada, but less frequently in Triclada
and Rhabdoccelidce. In the Polydadan family of the Planoceridce we find
on the dorsal surface, between the middle and the anterior end, two
lateral, mobile, stylet-shaped, solid tentacles (Fig. 103, t, p. 141), which
can occasionally be withdrawn into temporary depressions of the skin.
They may be directly compared with the tentacles of Ccelo- and Cteno-
plana, but are distinguished from these and from the contractile ten-
tacles of the Ctenophora (1) in that they have moved from the middle
of the back more or less far towards the front, and (2) in that they
have no lateral branches. In the Pseudoceridce and Euroleptidce there
are at the anterior edge tentacular folds of the leaf-shaped body, into
which branches of the intestine generally penetrate. In a few Triclada
also, and in Vorticeros among the fihabdocoelidce, feeler-like projections
or thickenings at the anterior margin of the body have been described.

The so-called proboscis is a highly developed specific organ of touch
which distinguishes a Rlwbdocodan family, the Proboscidea. The anterior
end of the body of many Turbellaria is very retractile ; in Mesostoma
rostratum it can be telescopically withdrawn and protruded. The
permanent arrangement in the Proboscidea can be deduced from such a

Ill

PLATODES— SENSORY ORGANS

151

condition. The proboscidal apparatus of such an animal consists of an
invagination at the anterior end of the body, on to which the outer
epithelium is continued. This invagination can be evaginated and
again withdrawn by means of special retractors. The invagination, as
well as the retractors which are inserted in it, are surrounded by a sac-
like muscular integument, by whose contraction the proboscis is pro-
truded. Fig. 108, A, B, C, represents the proboscis in various stages
of protrusion. The whole apparatus shows in every detail a similarity

K*

FIG. 108. — Proboscis of Macrorhynchus croceus. A , In a protruded condition ; B, half protruded ;
C, withdrawn (after v. Graff), re, Proboscis epithelium, which is a continuation of the body epi-
thelium (e) ; ms, muscular envelope which divides the proboscis from the body parenchyma ; m,
dermal muscle layer ; rm, muscles for withdrawing the proboscis.

which cannot be ignored to the 4 proboscides of the Tetrarhynchidce and
the proboscis of the Nemertina to be described later.

In the Trematoda the sensation of touch seems specially localised
in the suckers.

D. Ciliated Pits.

In certain Bhabdocalidce, viz. in the Microstomidce, Prorliynchidce,
and Plagiostomidce, there are 2 paired strongly ciliated integumental
pits which lie laterally on a level with the brain, and are supplied
with a nerve ring by the brain. They have been regarded as olfactory
pits. In the Tridada, also, similar strongly ciliated parts of the
epithelium of the end of the head have been observed, to which special
sensory nerves proceed. In Mpalium, a land Triclad, there are pits
supplied with special nerves in large numbers at the anterior margin
of the body which is broadened out in the shape of a crescent. Whether
the ciliated furrow which in all Polydada runs along the anterior margin
of the body in the epithelium of the ventral side belongs to the forma-
tion here described cannot yet be decided.

X. The Body Parenchyma (Retieulum).

The whole space between the body wall and the gastro-canal system,
as far as it is not filled by specific organs, is occupied by a cellular
connective tissue, the details of whose structure are very various. This
connective tissue, which corresponds with the gelatinous tissue of the
higher Cnidaria, is called parenchyma or reticulum. It often becomes

152 COMPARATIVE ANATOMY CHAP.

finely lacunar by the formation of many vacuoles filled with fluid.
The lacunae may coalesce in sinuses which conduct fluid, which, however,
generally remain small ; but in a few Ehabdoccela they become large
hollow spaces filled with a perivisceral fluid. In such cases the body
parenchyma can assume the constitution of a membrane, covering the
inner organs like an epithelium. In the Accda, where there is no
parenchyma separate from the intestine, the former, composed of star-
shaped cells filling the whole body apart from the specific organs, may
be described as digesting parenchyma.

XL The Excretory or Water-vascular System.

This is very characteristic of the Platodes, and as yet has been
found to be wanting only in the Accelo, among the Rliabdoccelidce. It
consists of a system of very fine transparent
canals (excretory capillaries) which branch out
in the parenchyma and between the muscles,
and which enter into a system of wider, equally
transparent canals, which open externally in
various ways. In the formation of the ex-
tremely thin walls of the capillaries only a few
cells take part, so that the nuclei lying in the
wall occur at long intervals, and in a trans-
verse section of a canal the wall enclosing the
central lumen belongs to a single cell. The

capillaries thus represent perforations of linear FIG 109._Excretory cell
rows of cells, and are described as intracellular. at the end of a fine excre-
The wider canals, on the other hand, in the JOI7 canal ^(fc) of a
Cestoda at any rate, appear lined by a thin epi-
thelium, and are thus intercellular. The blind wf, flame.
end of each excretory capillary is formed by one
cell (Fig. 109), which possesses fine protoplasmic processes running
into the parenchyma. In this cell, which, on the surface turned
to the lumen of the capillary, carries a tuft of fine vibrating cilia
(the flame) projecting into the lumen, excretory products (drops,
granules, etc.) collect and are emptied out of the cell into the
capillary. The excretory products are forwarded out of the capil-
laries into the wider vessels partly by the motion of the above-
mentioned cilia, and partly perhaps by the independent contractions of
the canals, and thence reach the exterior. Sometimes also in the
lateral walls of the capillaries and larger canals flames are found which
belong to the excretory cells in those walls. The larger canals some-
times have a continuous lining of cilia. It is not impossible that the
greater part of the transparent fluid which fills the canals is water
taken in from outside. If so, it is occasionally emptied out and again
taken in. In this way the water- vascular system may also perform
a respiratory function.

Ill

PLATODES— WATER-VASCULAR SYSTEM

153

The whole water-vascular system shows a decided similarity to a
greatly developed dermal gland which has sunk deep under the skin,
as is often the case in the Platodes (dermal mucus glands, rod glands,
accessory glands of the copulatory apparatus). The excretory cells
must then be considered as glandular cells, and the canals as glandular
efferent ducts. We can actually regard the water-vascular system as
a dermal gland, which has undertaken the special function of excretion.
In consequence of the strong development of the parenchyma, and
especially of the middle layer of the body, and because of the absence

FIG. 110.— Water -vascular system of various Platodes. A, of Stenostoma; B, of Meso-
stoma, after v. Graff ; C, Dendrocoelum, after Ijima ; D, Distomum divergens, after Fraipont ; E,
Phyllacanthide, after Pintner. ph, Pharynx ; o, external aperture ; ez, excretory cells (terminal
cell) ; eb, contractile vesicle.

of a body cavity, this gland is obliged to search for the products of
excretion all over the body — hence its great ramification.

Concerning the arrangement of the principal canals and their
external apertures (Fig. 110) the following is noteworthy.

We have no sufficient information as to the water-vascular system
of the Polydada.

In the Triclada (Fig. 110, C) a main canal runs on each side of
the body (Gunda has two such canals on each side, a ventral and a
dorsal, bound together by canals), and these open externally on the
dorsal side of the body by means of somewhat numerous special
branches and excretory pores placed one behind another. These pores
are arranged pretty regularly, in Gunda at least, and correspond in num-
ber with the lateral intestinal diverticula — the transverse commissures
of the nervous system, — in short with the number of those organs
which are regularly paired from one end of the body to the other.

In the Rhabdoccelidoe, we can distinguish 3 principal types.

A. There are 2 lateral principal vessels, which have separate
external openings on the ventral side, either (a) at the middle or the
front part of the body by special branches opening externally direct
(Prorliynchidce), or (b) by 2 cross branches which enter the pharyngeal
pouch (Fig. 110, B, Mesostomidce, Vortex .?), or (c) direct by two open-
ings lying at the posterior end of the body.

154 COMPAEATIVE ANATOMY CHAP.

B. There are 2 longitudinal trunks which open externally at the
posterior end of the body by means of a common terminal piece (many
J \>rtici(la>, Proboscidea).

C. There is a single median principal branch with an opening at
the posterior end of the body (Stenostoma among the Microstomidce,
Fig. 110, A\

In the Tri'iimtvdti, also we find typically 2 longitudinal trunks
which open externally, either together through a common contractile
terminal vesicle of very varying size at the posterior end of the body
(digenetic Trematoda, Distoma, Monostoma, Fig. 110, 7>), or separately,
anteriorly and dorsally, by means of 2 widened terminal portions.

In liistoma hcpaflcum there is a wider and larger median longitudinal trunk,
which stretches pretty far forward, and into which collecting canals enter from all
sides. The external aperture lies at the extreme posterior end of the body.

In the Cextodn. (Fig. 110, E) in the simplest cases there are on each
side 2 longitudinal trunks running through the whole body, which
are united anteriorly in the scolex by a loop. At the extreme
posterior end of the body (at the end of the oldest segment) all the
four branches open outwardly by means of a contractile vesicle
(Twniadie, Tetmbothridce, Tetrarhynchidce). In the Bothriocephalidce,
Ccmjophi/llidce, and Ligulidce the number of longitudinal trunks is
increased to from 10 to 24, which anastomose in a definite way.
A contractile terminal vesicle, into which all the longitudinal trunks
enter, occurs only in the end of the oldest segment of the tape-
worm ; in all other segments, after successive detachments of the older
segments, the longitudinal trunks open outward directly and indepen-
dently : some of them, however, may close, and thus end blindly. We
have an apparent exception to this rule in Tcenia cucuincrinct; in which,
as segments detach themselves behind, a new contractile terminal
vesicle is always formed in the next segment. In many Cextitdc,
besides the terminal apertures of the water-vascular system, special
secondary openings have been observed, generally in great numbers.
These are canals, which proceed at right angles from the main canals
and open outward through pores. These secondary exits are generally
found only at the front end of the body, in the scolex (in Tria'im-
plinnis, many Tn'tiia', and Tetrarhyncha), less frequently in the proglottides
also (Bothrioccphalus inindatiis and a few other forms).

The larger canals show a tendency towards the formation of islands in very many
IMatodcs, and especially in the (<rxt<i>ht. They then break up into a more or less
complicated anastomosing network. It is probable that the numerous longitudinal
trunks bound together bv anastomoses in the above-named Ccstoda have proceeded
from a few (4j longitudinal trunks by the formation of islands.

XII. The Sexual Organs.

All Platodes, with the exception of the genera Microstoma and
Stenoatonw among the PJiabdoccda and of Dixtniivi, Immatolium among

Ill

PLATODES— SEXUAL ORGANS 155

the Trematoda, are hermaphrodite. The male sexual products, however,
almost universally develop before the female — a phenomenon which
has received the name of protandrous hermaphroditism. Both male
and female sexual apparatus consist (1) of places of formation of the
sexual products which are imbedded in the body parenchyma (ovaries
and testes), (2) of special canals, efferent duets, which conduct the
sexual products away from the places where they were formed to (3)
the outer eopulatory apparatus. We will describe these three parts
in succession.

A. The Places of Formation of the Sexual Products.

1. The Female Germ Glands. — We meet with these in the Platodes
in two forms. First, and this is the simpler and no doubt also the older
condition, as simple ovaries, in which the egg germs ripen into eggs, in
whose protoplasm particles of deutoplasm or nutritive yolk occur.
Secondly, and this is the derived condition, in the double form of
germaria and vitellaria. The germaria yield the egg germs, i.e.
the young egg cells. The vitellaria, however, have undertaken the
work of supplying these egg cells with the nutritive yolk which is
necessary for their further development. Comparative research has
shown that the vitellaria are not newly formed accessory glands of the
female sexual apparatus, but that they are metamorphosed ovaries
or portions of ovaries adapted to a special function. It need not be
pointed out that the germaria are ovaries.

In the Polyclada there are only ovaries and no vitellaria. The
ovaries (Fig. 24, D, p. 29) are roundish bodies whose structure exactly
corresponds with that of the female gonades of the higher Cnidaria.
They lie in great numbers (Fig. Ill, o) on and between the intestinal
branches or gastro-canals in the lateral parts of the body.

In the Tridada, besides a few eggs in the discharged egg cocoons,
there are extraordinarily numerous yolk cells, which serve to nourish
the former. Light is thrown on this phenomenon by the fact that
in certain Polyclada and Mollusca several eggs may be deposited in
one cocoon, of which, however, only some develop ; the others
sooner or later become disorganised and serve as food for the
former. Thus perhaps the yolk cells in the Tridada cocoon may be
considered as modified egg cells, which develop no further, but serve
as food for the few fertilised egg cells which do develop. Division
of labour, therefore, has stepped in among the germ-preparing organs,
the ovaries ; some yield eggs capable of fertilisation and development ;
others yield modified egg cells laden with yolk, which serve as food
for the above, and are no longer capable of fertilisation and develop-
ment. The first are the germaria (Fig. 112, Jcs), the second the vitellaria
(ds). These two are homologous structures, and in a young condition
look quite alike. In consequence of the large number of yolk cells
which are given to the eggs, the vitellaria are far more numerous than

156

COMPARATIVE ANATOMY

CHAP.

the germaria, of which only 2 remain; but as these have only to
yield eggs without yolk, they are quite sufficient in number. They
generally lie at the anterior end of the body, whilst the vitellaria
always occur in the lateral parts of the body between the intestinal
branches.

In the Rhabdocoelidce the germ-preparing organs are considerably

cv

FIG. ill.— Sexual organs of a Polyclad
(Leptoplana). To the left only the female,
to the right only the male organs are depicted.
o, Ovaries ; ov, oviduct ; u, uterus ; h,
testes ; vd, vasa deferentia ; sb, seminal
vesicle ; p, penis ; sd, shell gland ; mo, male
sexual aperture ; wo, female aperture.

FIG. 112.— Sexual organs
of a fresh -water Planarian
(Triclad). ks, Germaria; ds,
vitellaria; h, testes; ov, ovi-
duct ; vd, vasa deferentia ; p,
penis ; go, common external
genital aperture.

reduced in number, but on the other hand are, in relation to the body,
much larger than the single ovaries of the Polydada. Many forms
possess ovaries only. The Accda, and the Macrostomidce among the
Ehabdocwla, have 2 lateral ovaries, the Microstomidce only 1 ovary.
In many Rhabdoccelidce so-called germ-vitellaria attain development,
one often clearly separated portion of the ovary yielding only egg
germs, the other only yolk.

One single germ-vitellarium is found in the Prorhynchidaz. Proxenetes among the
Mesostomidce, Schultzia among the Vorticidce, and Cylindrostoma among the Plagio-
stomidce possess two.

Ill

PLATODES— SEXUAL ORGANS 157

In the greater number of Rhabdoccda and Allcdoccela, however,
there is a complete separation into germaria and vitellaria. The
germaria are mostly small and round, the vitellaria (ds) are large, often
lobate, branched or reticulate. The vitellaria are generally double ;
where they are single reticulate branched masses, their originally double
condition can be recognised by the duplication of their efferent ducts.
The germarium is either double or single.

One germarium is found in most Mesostomidce, Gyrator among the Proboscidea,
most Vorticidce,and Solcnopharynx. Two are found in Promesostoma, most Probos-
cidea, Provortex and Graffilla among the Vorticidce, and the Alloiocosla.

Separate germaria and vitellaria are found in all Trematoda and
Cestoda. The germaria (Figs. 114 and 115, ks) are either simply
round, or lobed, or branched. The vitellaria (ds) are mostly (excepting
in the Teenies) very extensive, and branched in a reticulate manner, or
else broken up into a large number of small globular bodies or saccules.

The Trematoda possess a median germarium and 2 lateral vitellaria,
the Cestoda 2 germaria and either 2 lateral vitellaria or a small posterior
vitellarium (Tcenice).

II. The male germ elements or testes are present in the Polyclada
(Fig. Ill, h) in greater numbers than the ovaries. They always lie
in the lateral parts of the body between and under the intestinal
branches. The same holds good of the numerous testes of the Triclada
(Fig. 112, h). In Gunda segmentata the testes lie on each side in a
single longitudinal row in the dissepiments which separate the con-
secutive intestinal branches. They thus repeat themselves in the
body just as regularly as do the intestinal branches, dissepiments,
transverse commissures of the nervous system, and the external aper-
tures of the water-vascular system. In the Rhabdoccelidce there are
either 2 testes (Fig. 113, h, Rhabdocoda), or the testes are broken up
into numerous lobes or vesicles, which are scattered in the parenchyma
(Acoela, Alloiocoela). Nearly all Trematoda (Fig. 114, h) possess 2
round, or lobed, or branched testes, while in the Cestoda (Fig. 115, h)
there are numerous scattered testicle vesicles.

B. The Efferent Duets of the Sexual Products.

The Female Duets. — The female sexual glands are produced into
tubular ducts, which collect the sexual products and carry them to the
exterior. The ducts are either egg ducts (oviducts) when they pro-
ceed from the ovaries or germaria, or yolk ducts (vitello-ducts) when
they carry out material formed in the vitellaria. The anatomy of this
part is so varied that we can only select the most important points.

In the Polyclada (Fig. Ill) numerous oviducts (ov) proceed from
the numerous ovaries. These ducts frequently unite to form larger
ducts, which again enter more spacious tubes, generally running in the
longitudinal direction on both sides of the middle line. Numerous

158

COMPARATIVE ANATOMY

CHAP.

eggs collect in these tubes, which are called egg receptacles or uterus
tubes (u). As they enter the female copulatory apparatus they unite in
an unpaired terminal portion, the egg passage, into which the thread-
like efferent ducts of the numerous glands (sd) which are imbedded
in the surrounding parenchyma open. The glands whose hardening
secretions yield the egg shell which covers the egg or eggs are

010

FIG. 113.— Sexual organs of a

Rhabcloccele (Mesostoma Ehren-
bergii). To the left the testis is
omitted ; to the right the vitel-
laria and the uterus are left out.
h, Testis ; ds,' vitellaria ; «, uterus ;
p, penis ; ks, germarium.

FIG. 114.— Sexual organs of a Tre-
matode (Distoma), after Leuckart.
h, Testes ; ks, germarium ; u, uterus ;
ds, vitellariuin ; dg, yolk or vitello-
duet ; vd, vasa deferentia ; Lg, Laurer's
canal ; cb, cirrus pouch ; mo, male, wo,
female sexual aperture.

collectively known as the shell glands. They occur almost universally
in the Platodes.

In the Tridada (Fig. 112) there are 2 lateral longitudinal ovi-
ducts (00), which conduct the eggs from the 2 germaria in the front
of the body backwards to the copulatory apparatus. Apertures are
found on the way, through which the vitellaria empty their products
into them. Before passing out into the copulatory apparatus they
unite to form a short unpaired egg passage, into which the efferent
ducts of the shell glands open.

In the Rhabdoccdidce (Fig. 113) the female sexual glands are gener-
ally placed with their ends directly on the outer copulatory apparatus,
and open into it. There is often a union of the vitellaria and the
germaria before their exit into a common terminal portion. The shell

Ill

PLATODES— SEXUAL ORGANS

159

glands open either into the outer copulatory apparatus, or into a
special fold of the same into which eggs and yolk are conducted, and
which will here be called the uterus. In the A coda and Alloiocoda
no definite connection between the germ glands and the copulatory
apparatus can be demonstrated. The sexual products here make their
way through the parenchyma to the copulatory apparatus.

In the Trematoda (Fig. 114) an oviduct proceeding from the
ovarium is found, then two yolk ducts (dg) which collect the yolk from
the two lateral vitellaria. The oviduct and the yolk ducts together
enter an unpaired canal which we shall call egg passage. This passage
falls into 2 parts — one small part at the beginning, the ootype, into
which the 3 ducts mentioned enter, and a long, generally coiled

FIG. 115.— Sexual organs of Taenia saginata
(medio-canellata), after Sommer. h, Testes ; vd,
vasa deferentia ; cb, cirrus pouch ; gp, genital pore ;
ov, oviduct ; ks, germaria ; sd, shell glands ; ds,
vitellarium ; u, uterus.

FIG. 116.— Fully ripe proglottides
(segments). A, Of Taenia saginata ;
B, of Taenia solium.' The dendriform
figure represents the uterus.

portion, the uterus (u)t leading from the ootype to the female genital
aperture. The ootype receives the efferent ducts of the shell glands.
Fertilisation takes place here, and also the union of the yolk with the
egg ; and here a shell is formed round the fertilised egg. The ootype
has an exit either in the dorsal or the ventral surface of the body
through one more canal, Laurer's canal (Lg\ through which, probably,
in copulation, the sperm reaches the ootype from outside. The fertil-
ised eggs pass from the ootype into the uterus, where they often
collect in enormous numbers, at least in the Distomidce. The uterus
in these animals is consequently very long, and runs to the female
copulatory apparatus in numerous coils, which in the adult often
fill the greater part of the body.

The Cestoda (Fig. 115) are closely allied to the Trematoda, especi-
ally in forms in which, as in Bothriocephalus, the sexual apertures lie
on one of the surfaces, and in which there are 2 lateral vitellaria.
The collecting passages of the vitellaria unite in such forms into

160 COMPARATIVE ANATOMY CHAP.

2 yolk ducts, which, as well as 2 oviducts, enter the ootype by a
common terminal piece, into which the efferent ducts of the shell glands
open. From the ootype a canal proceeds to the copulatory apparatus
(ov) on one side, and on the other side arises a widened uterus filled
with eggs (Fig. 115, u; Fig. 116) running in coils, or provided with
lateral sacs ; this uterus often reaches the exterior by a special aper-
ture which recalls Laurer's canal in the Trematoda. Where there is
only one vitellarium, only one yolk duct naturally enters the ootype
(as in Fig. 115).

II. Male duets. — In the Polydada numerous very fine canalicules
enter the larger semen duets, vasa deferentia (Fig. Ill, vd), in
which the spermatozoa collect, and these canals again have their exit
in the male copulatory apparatus (p). The fine canals correspond to
the oviducts, the wider ones to the uterus of the female sexual appar-
atus. In the Triclada there are 2 lateral vasa deferentia (Fig. 112,
vd), into which some at least of the testes empty their contents
direct, while the manner in which the testes which are at a greater
distance from the vasa deferentia empty themselves is not yet fully
understood.

In the Ehabdoccela (Fig. 113) the 2 testes are often' continued
without any sharp demarcation into 2 semen ducts which enter the
male copulatory apparatus either separately or by means of a common
terminal portion. In the Accela and most Alloioccela special ducts
are wanting ; the spermatozoa reach the copulatory apparatus through
the parenchyma. Only in the Monotidce among the Alloioccela the
transmission takes place by means of special ciliated vasa deferentia.
The two testes of the Trematoda (Fig. 114) send out two semen
ducts (vd) which unite into one common duct. In the Cestoda also
(Fig. 115) many of the numerous canalicules proceeding from the
testicle vesicle enter a common vas deferens leading to the male copu-
latory apparatus.

C. The Copulatory Apparatus.

There is wonderful variety in the structure and position of the
copulatory apparatus in the Platodes. Nearly related species often
differ greatly in this point.

I. The male eopulatory apparatus is always more complicated in
structure than the female. It consists in the simplest cases of a mus-
cular pouch which projects from the surface into the parenchyma,
and into whose blind end, which is directed inwards, the semen duct,
or ducts, enter. It is found in this form in certain PJiabdoccelidce. In
most Turbellaria, however, it becomes complicated, and then we can
generally distinguish the following distinct portions: (1) a penis
sheath or penis pouch, (2) the actual penis, (3) a seminal vesicle,
and (4) a granular gland. The penis and penis sheath show a struc-
ture which is on the whole like the structure of the pharyngeal appa-

in PLATODES— SEXUAL ORGANS 161

ratus described above. The penis is, in fact, a muscular circular fold
which projects into the penis sheath from its wall in a manner similar
to that in which the pharynx projects into the pharyngeal pouch. As
the pharynx is protruded out of the pharyngeal pouch through the
mouth, so is the penis protruded through the sexual aperture. During
copulation the wall of the penis sheath also is pushed out or evagin-
ated. The penis sheath is occasionally double, or there are several
sheaths, each of which is related to the one outside it as the penis is
to the penis sheath, and the whole apparatus may be telescopically
extended and protruded. The penis is sometimes conical, some-
times cylindrical, sometimes bent, either naked or armed in various
ways. Its free end is often a hard chitinous tube. Between the
penis on the one side and the terminal portion of the semen duct
on the other, there is a vesicular expansion with muscular wall, the
seminal vesicle (Fig. Ill, sb), in which the semen collects, and
which, by its contraction during copulation, causes the ejection of
the semen through the penis canal (ductus ejaculatorius). In nearly
all Turbellaria there is, in connection with the male copulatory
apparatus, a granular gland, the structure of which differs greatly
in details. It forms a finely granular secretion, which mixes with the
semen.

The male copulatory apparatus of the Trematoda (Fig. 114, cb) and
that of the Cestoda (Fig. 115, cb) are very similar in structure.

In mechanism it corresponds with a Tetrarhynchus proboscis.
There is a cylindrical or club-shaped penis sheath. Into the inner
blind end of this penis sheath enters the unpaired terminal portion of
the vas deferens. On entering the penis sheath it generally expands
into a seminal vesicle, and then runs as a coiled thin tube through the
penis sheath to emerge at the sheath's outer end through the male
genital aperture. This tube, which is often furnished internally with
barbed hooks or covered with an elastic cuticle, is forced out as an
actual penis by the contraction of the penis sheath. The space
between the penis and the penis sheath is filled with loose connective
tissue. The penis and penis sheath are generally called cirrus and
cirrus pouch in the Trematoda and Cestoda. Glands connected with
the copulatory apparatus have also been observed.

II. The female eopulatory apparatus very often, in many Tur-
bellaria and in all Trematoda and Cestoda, consists of a simple tube of
varying length, the vagina, which connects the egg passage or the
ootype with the female sexual apparatus. This tube often serves
merely as a place for depositing the eggs, not for copulation, i.e. it
does not receive the penis. This is at least often the case with those
Polydada which have more than one copulatory apparatus, but only
one female genital aperture.

In very many Turbellaria, however, the vagina is differentiated
into a strong muscular organ, often provided with a hard cuticle, the
bursa eopulatrix, which is adapted for the reception of the penis
VOL. i M

162 COMPARATIVE ANATOMY CHAP.

during copulation. This may be developed independently as an acces-
sory organ of the female copulatory apparatus. There is in many
forms another broad, round, or pear-shaped accessory organ, the
reeeptaeulum seminis, a reservoir in which the semen is preserved
after copulation.

In the Trematoda, many Cestoda, and in Trigonoporus among the
Polydada, the ootype! or the uterus, or the egg passage, is connected
with the exterior by another special passage, Laurer's canal, already
mentioned. The physiological signification of this canal is not yet
certainly understood.

D. The Position and Number of the Copulatory Apparati and
the Sexual Apertures.

It may be considered the rule that one male and one female
copulatory apparatus are present, and that each opens externally by
its own special aperture somewhere in the middle line on the ventral
side. The two sexual apertures are generally very near each other,
and in many forms — most Trematoda, Cestoda, and Triclada, and in many
Polydada and Rhabdoccelidce — come to lie in the base of a more or less
deep depression of the outer skin — atrium genitale — so that only
one common outer sexual aperture is present.

In this point there is great variety in details, and many often striking deviations.
In the Polydada the sexual apertures always lie behind the mouth, in the Cotylea,
in particular, between the sucker and the mouth. The male aperture always lies in
front of the female. Stylochus and Stylochoplana have a common external sexual
aperture. In Anonymus there are several copulatory apparati and sexual apertures
in 2 lateral longitudinal rows. Many Pseudoceridce possess 2 male copulatory
apparati. The female copulatory apparatus and its aperture always remain single.
In Stylostomum there is one common external aperture for the pharynx and the
penis.

In the Triclada the common sexual aperture lies behind the mouth, the male
copulatory apparatus in front of the female.

In the Rhabdoccelidce the arrangements are extraordinarily varied. There are
sometimes two separate apertures, sometimes an atrium genitale, and thus a common
external sexual aperture. Sometimes the male aperture lies in front of the female,
and sometimes the reverse is the case. In Prorhynchus the male copulatory apparatus
opens in the mouth.

The genital apertures of the Trematoda, which either enter a common shallow
atrium genitale or are very near together, generally lie at the anterior end of the body ;
in the Distomidce, between the mouth and ventral sucker. Less frequently they lie
at the posterior end of the body (e.g. Gasterostomum, Opisthotrema), or, asymmetric-
ally to the left near the anterior edge of the body (e.g. in Tristomum).

In the Cestoda there is generally a common external genital pore, or else the
genital apertures are very near each other. The porus genitalis or the two genital
apertures of each proglottis either lie at the edge (Tetraphyllidce, Tetrarhynchidce,
most of the Tceniadce, Trioenophorus), or on one of the flat surfaces, which is
therefore the ventral side (Ligula, Bothriocephalus, Schistocephalus, a few Tcenice).
In Amphilina they lie at the posterior end of the body.

in PLATODES— ASEXUAL REPRODUCTION 163

Copulation is generally mutual, both the copulating individuals acting as male
and as female. Self-fertilisation, however, also seems to take place, e.g. in Cestoda,
and perhaps also in a few Trematoda and Turbellaria.

Development. — Like the atrium genitale, which is only a pit-like depression of the
outer skin, so the male and female copulatory apparati arise, at least according to
investigations made in the Polyclada, by folds from the exterior. The portion of the
female genital organs which arises from invagination apparently reaches to the egg
passage or ootype, so that not only the glands which open into the male genital
organs, but the shell glands also of the female genital organs must be considered as
modified dermal glands.

XIII. Asexual Reproduction and its Origin — The Organisation
of the Cestoda.

Many Platodes, and especially the Turbellaria, show a marked capacity for
regeneration. The body can not only re-form parts torn off, but broken off pieces of
various sizes can become regenerated into new animals. Such a capacity of regenera-
tion is very widespread, chiefly among lower stationary animals. In the Ccelenterata
it is almost universal. The great advantage of this capacity for the preservation
of the individual and of the race is evident. For attached animals, or very long
or delicate soft-bodied animals, who are more exposed than others to mutilation
and injuries to the body from enemies, etc., it is of very great importance. We
can perhaps trace back to it the power of asexual reproduction by fission and gem-
mation which occurs in the Metazoa. "We speak of such a method of reproduction
when an animal form shows the peculiarity of falling into 2 or more pieces, appar-
ently spontaneously, i.e. from causes unknown to us, these pieces becoming regener-
ated into organisms similar to the common mother animal ; or when, from unknown
causes, a smaller or larger piece of the body regularly detaches itself, the body
thus reduced again replacing the lost portion, while the detached part becomes
regenerated into a complete animal.

Thus e.g. Lumbriculus, one of the worms belonging to the Oligochceta, falls
spontaneously, or apparently spontaneously, into 2 or more pieces, each of which
can become regenerated into a whole animal. Certain marine star-fish throw off one
or more arms apparently spontaneously, which they soon replace by regeneration.
As if this were not enough, each detached arm can again be regenerated into a
complete star-fish.

This conjectural origin of asexual reproduction is, however, almost always un-
recognisable ; because different parts of an individual develop into whole individuals
before they have fully separated, or an animal replaces a part by regeneration before
this part has completely detached itself. Temporary animal stocks thus arise. If
the parts do not detach themselves typical animal stocks arise, which by division
of labour between the portions which are being regenerated into whole individuals
(i.e. between the individuals which have arisen by gemmation), and by the develop-
ment of a form and organisation in each adapted for some special function, may
become polymorphous animal stocks.

The reproduction and life-history of the Acraspede Medusa, e.g. of Aurelia,
is specially suitable for the elucidation of the above view. We know that from
the fertilised egg of this Medusa, under certain circumstances, another Medusa
may proceed, without any attached stage multiplying asexually. Generally,
however, the larva developed from the fertilised egg attaches itself and becomes
a coral-like animal, the Scyphula, and later develops into an attached young

164: COMPARATIVE ANATOMY CHAP.

Medusa, the Scyphistoma. When this Scyphistoma has developed to a certain
stage, in one case the larger portion of the body tears itself from the stem as a free-
swimming Medusa. The remaining stem can, however, become regenerated into a
complete attached Medusa (monodisc strobila), and the whole process may be repeated.
We thus have here multiplication by detachment and subsequent regeneration.
The detached piece has indeed so little to regenerate in it that the regenerative
process may be described as cicatrisation.

Or again the stem of the Scyphistoma becomes regenerated into a new
Scyphistoma before the first Medusa has detached itself, and when this regenerative
process continues without the Medusce at once fully detaching themselves we have
a polydisc strobila. We call the whole process strobilation, and it has been described
as asexual multiplication by axial budding. The polydisc strobila is a temporary
animal stock.

What has here been said helps us to understand the t

Organisation of the Cestoda Body.

In the body of the large majority of Cestoda the seolex (Fig. 117)
is distinguishable from a row of subsequent segments or proglottides

(Fig. 110, E, p. 153 ; Figs. 115, 116).
The small pear or cone-shaped seolex
itself consists of the head and neck.
The former carries the organs of
adhesion (suckers, hooks, proboscides),
by means of which it attaches itself
to the intestinal wall of the host.
In it lie the single commissures be-
tween the longitudinal trunks of the
FIG. ii7.— Three heads of Tapeworms nervous system which may be regarded
(scolices) A, Of Taeniasaginata;* of ag brain commissures. ft therefore)
Taenia solium ; C, of Bothnocephalus , . 1 . . . , /

latus. corresponds with the anterior end ot

the Trematoda body. The thinner neck

portion of the seolex is followed by the flattened segments, which are
small at first but increase in size posteriorly. The neck portion of the
seolex constantly produces new segments, which push back those already
existing. The oldest and largest segment of the whole chain is there-
fore the hindmost. In the segments the genital organs develop ; indeed,
the whole hermaphrodite genital apparatus of each segment answers
to the whole genital apparatus of a Trematode. The male genital organs
are first developed in each segment, then the female; then follows
fertilisation, and finally the segment is little else than a case which,
besides the remains of the genital organs, is almost exclusively occupied
by the extended uterus containing thousands of fertilised eggs. The
row of segments from the head to the last segment represents the row
of consecutive stages of development of the genital organs. The last
segments from time to time detach themselves singly, or several together,
and reach the exterior with the excrement of the host.

On comparing the head and the segments we find that the head has
no genital organs, and none of the segments have the organs of adhesion

in PLATODES— ORGANISATION OF CESTODA BODY 165

and the brain commissures ; or, when we compare the head and pro-
glottis with a Trematode, we find that the head has not the trunk, and
the proglottis not the head, of the Trematode body. The head and one
proglottis together, however, answer to the head and trunk, and thus
to the whole body of a Trematode, apart from the fact that an intestinal
canal is altogether wanting in the Cestoda.

We, however, know forms whose body during life consists only of
head and trunk, which is not clearly divided into a scolex and a
proglottis. Such forms are Amphilina, Caryophyllceus, and Archigetes.
These may be regarded either as intestineless Trematoda, or unsegmented
Cestoda. They are, in any case, transition forms between the Trematoda
and Cestoda. The relation existing between them and the segmented
Cestoda is somewhat similar to that between the Acraspeda which are
attached throughout life (e.g. Lucernaria), and the polydisc strobila of .
Aurelia.

A segmented tapeworm must in fact be considered as a strobila.
The young, still unsegmented tapeworm, which attaches itself to the
intestinal wall, i.e. the scolex, answers to the young stage of one of the
unsegmented Cestoda mentioned above (Amphilina, Caryophyllceus, or
Archigetes), in which the genital apparatus is not yet developed in the
slightly developed trunk, the future neck. Now follows the incom-
plete constriction of that part of the body of the scolex (the trunk or
the first proglottis) in which later the genital organs develop. Re-
generation of the constricted part then takes place ; this part is again
constricted and again regenerates, and so on. The single parts remain
connected for a longer or shorter time, and form the segments of the
tapeworm chain or strobila. Finally, like the oldest Medusa discs of
a polydisc strobila of Aurelia, the oldest segments of the tapeworm
strobila detach themselves. The points in which the process differs
in the two groups are essentially the following. The Medusae which
detach themselves from a polydisc strobila develop further, and their
sexual organs attain development only after detachment. The segments
of the Cestoda which detach themselves, however, are already more
than mature (sexually) ; they have performed their function, the pro-
duction of fertilised eggs, and they make no attempt to regenerate
the part which is wanting to make them complete Platodes, i.e. the
head. In the Medusa strobila, further, that part of the body by which
it is attached, viz. the apex of the exumbrella, is an insignificant part
of the body both physiologically and anatomically, while the part by
which the tapeworm is attached contains at least the principal part of
the central nervous system.

It does not seem difficult, in the case of the Tapeworm, to trace
back strobilation to the phenomenon of regeneration. Proceeding
from forms, like Amphilina, capable of regeneration, we can understand
that by the peristaltic movements of the intestinal canal in which the
animals lived parasitically, and by the outward movement of the ex-
crement, the trunk, with the genital organs it contains, would be torn

166 COMPARATirE AXATOMY

CHAP.

off and ejected, the head which remained being, however, able to pro-
duce a new trunk by regeneration. This process — the tearing off of
the trunk with the eggs, the continued attachment of the head, and
regeneration — must have been of the greatest use to these parasitic
forms. By the tearing off of the trunk and its ejection the greater
dispersal of the eggs was secured, and the probability of the infection
of new hosts or intermediate hosts thus increased. The attached
head could keep its ground in the already attained favourable refuge
for the parasite, and easily regenerate a new trunk and new genital
organs. The strobila consisting of man}* segments, however, offered
the immense advantage that many segments could be benefited by the
favourable nutritive conditions of parasitism, and could develop the
genital organs ; on the other hand, by the periodical tearing off of the
trunk of an unsegmented Tcenia, not only a longer time must pass
before a new sexually ripe trunk would form, but the favourable con-
ditions of nutrition would be much less utilised.

There are Ta'nicc with only very few segments (To.'nia Echinococcus,
with 3 to 4 proglottides); others possess several hundreds.

In a few Tcenia', such as Ligula and Ti*icenopJiwus, the outer segmenta-
tion is more or less indistinct ; internally, however, we find the same
repetition of the genital organs as in the typically segmented tape-
worms, from which these forms must without doubt be derived.

In freshwater Tridada, multiplication by fission has been observed.

Among the Rhaldocoela, in the genera Microstoma and Stenostoma,
we find interesting processes of reproduction by axial budding.
They can be best investigated in M. linen re. In the posterior
end of the body of an individual a double transverse partition
wall forms between the intestine and the skin. Immediately behind
this the organs characteristic of the head portion of the Microstoma, —
the pharynx and the brain, — with the nerve commissure surrounding
the pharynx, form. The two septa subsequently move somewhat apart.
An annular constriction of the body takes place between them, and the
intestine finally also becomes constricted. Only then does the spon-
taneous separation of the two pieces occur. Long before this separa-
tion occurs, however, new phenomena have appeared in both pieces.
In the first place the posterior piece grows to the same size as the
anterior. Then in the posterior part of each a head portion again
forms. The posterior parts of each principal piece thus marked
off then grow to the size of the two parts lying in front of them. The
whole body now consists of 4 pieces of equal size. This process is
repeated twice in the same way, till 16 pieces are formed, i.e. till
the worm stock consists of 16 individuals, the one most to the front
possessing the original pharynx, the original brain, etc. Then follows
generally the spontaneous separation of the individuals.

Reproduction by gemmation occurs further in the young stage of
Tcenia called the Finn, and especially in those finns which are known
as Ccenurus and Echinococcus. This will be described later on.

in PLATODES— ONTOGENY OF POLYCLADA 167

XIV. Ontogeny of the Polyelada.

As a short illustration of the development of the Platodes from the fertilised egg,
we choose the TurMlaria (Polydadidce). [The ontogeny of the Rhabdoccsla is
almost unknown, and the development of the Tridada seems to us to be markedly
coenogenetic. The eggs of these latter animals develop at the expense of the numerous
yolk cells in the midst of which they lie imbedded within the egg cocoon, and it
might with justice be said that the eggs and embryos of the Tridada live parasitic-
ally on these yolk cells, which is not the case in the Polydada.]

We have already (Fig. 94, p. 125) described and illustrated the first stages of
segmentation. The 4 micromeres which are first separated by constriction yield the
whole ectoderm ; the next 4 or twice 4 in like manner produced form a large
part of the later mesoderm. The descendants of the 4 ectoderm micromeres grow
round the whole germ by repeatedly dividing, thus enclosing not only the 4
macromeres, but also the 4 or 8 mesoderm micromeres. They thus form at
last a continuous layer of epithelial cells round the whole germ, which is only broken
through at the vegetative pole by a longitudinal slit corresponding with the ventral
medium line of the embryo. This longitudinal slit is defined as the blastopore ; it
soon completely closes. The germ is now at the stage of a bilaterally symmetrical
planula, in which there is already a formation of mesoderm between the endodermal
rudiments (the 4 macromeres, which meantime by division of one of them have
increased in number to 5) and the ectoderm.

The 4 or 8 mesoderm micromeres soon increase in number by fission, and
thus form either a ring of mesoderm cells or 4 masses of mesoderm cells (2 anterior
and 2 posterior). The macromeres, now surrounded on all sides, continue to give off
by constriction micromeres, which again increase by fission and yield the intestinal
epithelium. The yolk-containing macromeres finally become disintegrated, and the
yolk is used up by the intestinal cells. Near the original blastopore a depression of
the ectoderm occurs, the stomodseum, as the first beginning of the pharyngeal
apparatus. The germ is now, apart from the fact that it is bilaterally symmetrical,
at the stage of a Scyphula or a young Ctenophoran larva.

The mesoderm cells extend more and more between the endoderm and the ecto-
derm ; the sensory organs first appear in the ectoderm, on the side opposite the
stomodseum, near the original animal pole, but shifted somewhat along the median
plane, so that now anterior and posterior ends can be clearly distinguished. These
sensory organs take the form of 2 or 3 eyes, and of cells which carry tufts of long
hairs. In connection with these sensory organs, which in the case of eyes
soon sink down below the ectoderm and become mesodermal, the paired cerebral
rudiments arise as products of the ectoderm, and these also soon sink under the
surface and become mesodermal. The two rudiments become secondarily connected
by transverse bridges. The principal nerve trunks seem to arise as outgrowths of
the cerebral rudiments which form the so-called neural plate.

The ectodermal body epithelium becomes provided with cilia at an early stage.
In the centre of the endoderm the enteric cavity arises in consequence of the
increasing absorption of the yolk by the endoderm cells, which arrange themselves
peripherally like an epithelium. Into the base of this cavity the stomadseum soon
breaks. The enteric aperture is thus formed. The stomodseum changes in the
following way into the definite pharyngeal apparatus. A circular invagination forms
in it, the first beginning of the pharyngeal pouch. This is surrounded by a collection
of mesoderm cells. Into the pharyngeal pouch the pharynx itself grows as a
circular fold, consisting of mesoderm cells and a covering of epithelium. "The

168 COMPARATIVE ANATOMY CHAP.

body, which till now has been tolerably round, begins to flatten ; the surface
in which the mouth lies can as the ventral surface be distinguished from the dorsal
surface under which lie the eyes and brain. The mesoderm cells everywhere
spread out between the intestine and the body epithelium, and form a continuous
mass, which is thicker at the ventral side. Those layers of mesoderm which lie
close under the epithelium yield the dermal musculature ; the deeper mesoderm
cells yield the body parenchyma, and most probably also the germ -preparing organs
of the genital apparatus.

In a series of Polydada whose embryos leave the egg shell very early as free-
swimming Miiller's larvae (Fig. 118), a ring of strong and long cilia which encircles
the body arises directly in front of the mouth; this is the so-called preoral
ciliated ring, running out round 4 or 8 processes of the body, one of which lies
immediately before the mouth and one in the middle line of the back, while the

FIG. 118.— Miiller's Polyclad larva (of Thysanozoon or Yungia). A, Median longitudinal
section, g, Brain ; hd, main intestine ; en, endoderm ; ec, ectoderm ; sn, sucker ; ph, pharynx ; pt,
pharyngeal pouch ; o, mouth. B, The same from the side. The black line indicates the course of
the preoral ciliated ring.

other 2 or 6 lie laterally in pairs. These processes, with their strong cilia, are
drawn in and reabsorbed when the free-swimming larvae sink to the bottom and begin
the creeping manner of life.

The differentiation of the originally single enteron into main intestine and gastro-
canals follows in consequence of the growth of mesodermal septa from the periphery
more or less far inwards.

The position of the mouth and pharynx on the ventral surface in the adult
animal is determined by the relative growths of the anterior and posterior halves. If
they grow equally, these organs lie centrally ; if the anterior half grows the more
strongly, they lie posteriorly ; if the posterior grows the more, then they lie anteriorly.

XV. The Life-history of the Trematoda.

Whereas from the fertilised eggs of the ectoparasitic or monogenetic
Trematoda other Trematoda develop direct without their young being
assigned to another animal or host than that occupied by the adult,
the development and life -history of the endoparasitic or digenetic
Trematoda is remarkably complicated. We choose as an example the
tolerably complicated life-history of the fluke, Distoma hepaticum
(Fig. 119), which is parasitic in the liver of the sheep, causing the
" sheep rot." The eggs of the fluke leave the liver of the host by the
bile ducts, pass into the intestine, and are ejected with the excrement.
They only develop when they meet with water. In this case there
develops in the egg shell (A) a ciliated embryo, which leaves the

Ill

PLATODES— LIFE-HISTORY OF TREMATODA

169

egg shell and swims about freely (B). It is club-shaped ; at the thicker
anterior end it has a small median prominence, behind this an X-shaped
eye spot, and under this a ganglion, and further a granulated mass which
is considered to be the intestinal rudiment. We can also recognise 2
ciliated cells of the excretory system. The greater mass of the body,
however, is formed of germ cells, which are considered to be partheno-

FIG. 119.— Life-history of Distoma hepaticum, after Leuckart. A, Egg with embryo. B,
Free- swimming ciliated embryo ; o, eye spot. C, Sporocyst. D, E, and F, Redias ; ph, pharynx ;
go, birth aperture ; d, intestine. G, Cercaria ; ms, oral sucker ; n, nerve ganglia ; fcs, ventral
sucker ; gd, forked branches of the intestine ; cd, glands, whose secretion yields the cysts.
H, Encysted young Distoma ; c, cyst. I, Young Distoma in the sheep's intestine.

genetic eggs, possessing the capacity, of developing without being
fertilised. These germ cells divide (furrow) early, and become cell
spheres.

The embryos must meet with a water-snail, Limnceus truncatulus, and
penetrate into its respiratory cavity in order to develop further. They
here lose the covering of cilia ; the eyes, the ganglion, and the granulated
mass become disorganised. Their bodies represent a pouch, containing

170 COMPARATIVE ANATOMY CHAP.

in its interior a certain number of cell spheres which have developed
out of the germ cells or parthenogenetic eggs. Instead therefore of
the young animals, which are called Sporocyst s (C), developing further
into new Distonw, they not only remain at a low stage of development,
but they even suffer a considerable degeneration. It seems as if early
reproduction were the only function of this Sporocyst. The cell spheres
which they contain actually develop again into new germs, which
leave the Sporocysfs body as Redicc (D, E), the Sporocyst finally disinte-
grating, and thus never developing into a fluke. The Redicv which
have become free, being developed out of the parthenogenetic eggs of
the Sporocysts, reach a higher stage of development than their mother.
They have at the front end of their body a sucker-like formation, and
also a pharynx, a simple intestinal tube, and a birth aperture behind
two blunt processes. Here also we find numerous germ cells between
the intestine and the body wall ; these begin early to develop, i.e. to
divide. The Redice in fact, like the Sporocysts, do not grow into flukes :
they first creep about in the respiratory cavity of their host, Limnccvs
truncatulus, and then penetrate into its liver. The germs which develop
in them again become He-dice, which pass out by the birth aperture
and are parasitic in the liver with their parents. This second genera-
tion of liedice (F) again reproduces itself parthenogenetically. From
their germs, however, at a warm time of year are developed, not Redid'
again, but Iarva3 which are called Cercarice (G). These Cercarice already
show the structure of a young Distoma ; they are flat, have oral and
ventral suckers, a pharynx and a forked intestine, a double ganglion
joined by a transverse commissure in front of and above the pharynx,
both the principal branches of the excretory system, and besides these
— and this is characteristic of the Cercarice — a movable caudal
appendage. The Cercarice leave the mother body, i.e. the Redicc, by the
birth aperture, forsake their host, and reach the water, in which they
swim about for a time by means of their tail. They then settle upon
grasses or plants growing in water in flooded meadows, lose their tail
and become encysted by the help of the secretion contained in two very
large glands which lie laterally in the body. In this encysted condition
(H) they can remain a long time, and can withstand desiccation. They
reach the sheep's intestine if occasion offers in the fodder, and there
presumably the cyst is dissolved and the young Distoma enters the
liver through the bile ducts. Such a young Distoma, with the first
branchings of the intestine, is depicted in Fig. 119, /.

The life-history of other endoparasitic Irematoda runs, as far as
we know, the same course. The free -swimming Cercuria, however,
often enters into a second intermediate host, in which it becomes
encysted and loses its tail. This second host is generally an inverte-
brate animal. The encysted Cercuria enters the body of the final host
(generally a vertebrate animal) when the second intermediate host is
eaten by the latter.

Several different generations, therefore, follow each other in a

in PLATODES— LIFE-HISTORY OF CESTODA 171

regular manner in the endoparasitic Trematoda. The generation which
multiplies by fertilised eggs always reaches the full degree of organisa-
tion of the Trematoda:, the following generation, which reproduces
itself parthenogenetically and lives in other hosts, never reaches that
degree of organisation ; they are ripe extraordinarily early, and perish
after they have produced another generation, which also remains at an
early embryonic stage. The different generations are known as
Sporocysts, Redice, and Distoma generations. The regular alternation
of such generations is called Heterogeny.

XVI. The Life-history of the Cestoda.

From the fertilised eggs of the Cestoda there proceed, generally
while they still lie in their egg shells in the uterus, embryos which,
since they are provided with 6 hooks, are called the 6 -hooked embryos.
The fate of this embryo, which only in Bothriocephalus is ciliated and
swims about freely in water, is very different in different Cestoda.
In Tcenia cucumerina, which is parasitic in the intestine of the dog,
it enters the body of the dog louse, Trichodectes canis. It here gets
rid of the egg shell and reaches the body cavity, where it develops
into a small worm, at one end of which the head develops with
its rostellum and its 4 suckers, while at the other the pore of the
excretory system can be made out. The head is somewhat sunk into
the body. The body is filled with numerous calcareous granules. We
have here simply an unsegmented, not yet sexually developed, tape-
worm, which may be compared with a young Amphilina, or Caryo-
phyllceus, or Archigetes. Through the dog's habit of licking and cleaning
itself, the present host of this young form, which we may simply call
scolex, is liable to be swallowed. While the louse is digested, the
scolex withstands digestion, the calcareous granules neutralising the
acid juices of the stomach. It fastens itself to the intestinal wall,
and begins to produce, by terminal budding or strobilation, the chain
of proglottides in which the genital organs develop.

In this simple case we have one and the same individual, from
the egg to the strobilising intestinal scolex. The 6-hooked embryo,
the scolex in the body cavity of the louse, and the strobilising
scolex in the intestine of the dog, are the same individual in various
stages of development and in various habitats. In most of the
Cestoda, in consequence of peculiar complications in the develop-
ment, this is by no means so clear. In a series of Cestoda, to
which Tcenia solium and T. saginata belong, the 6-hooked embryo
in the tissues of its host changes, by the accumulation of fluid
internally, into a vesicle surrounded occasionally by a special capsule
or cyst formed out of those tissues. . From the wall of this vesicle,
which is called Finn or Cystieereus, there arises, at the base of
an invaginated hollow cone, a tapeworm head with suckers, rostellum,
etc. (Fig. 120). While most investigators regard this process as

172 COMPARATIVE ANATOMY CHAP.

one of gemmation, we hold it to be simply one of growth and
differentiation. The head with the vesicle is, according to our opinion,
a young sexless Cestode answering to the scolex of Ticnia cucumcrina in
the body cavity of the louse, only in this case the trunk or proscolex
becomes extended into a large vesicle by the accumulation of fluid
before the head of the tapeworm with its suckers, etc., forms. The
development of this vesicle ought to be
regarded as a special adaptation for the pro-
tection of the head. If such a Cysticercus
reaches, with the tissue of its host, the
intestine of a new host, not only the cyst,
but the whole vesicle dissolves, while the
head and rudimentary neck which are
evaginated resist digestion because of the

FK, 120. Icyltteercus ceiiu- calcareous bodies they contain. In other
losa. Finn of Taenia soiium, words, the young, sexless, unsegmented tape-
cut in half. The scolex, which worm ioses its vesicular trunk. The scolex

is invalidated into the vesicle, is . , P , c •, f

seen with its suckers and rostei- fastens itself, by means of its organs of

lum. After Leuckart. adhesion, to the intestinal wall, and at once

regenerates the lost portion of the body in

the form of the first proglottis, which in the developed tapeworm
chain at length becomes the last and oldest, and new segments
follow this one.

The vesicle of the Cysticercus of different tapeworms varies in size
according to the amount of fluid contained. It is sometimes a large
sphere, sometimes merely a small swelling at the posterior end of
the worm-like Cysticercus.

In a few tapeworms development is complicated by the occurrence
of an alternation of generations, the young unsegmented form in the
intermediate host, the finn, multiplying asexually by gemmation. On
the wall of the finn there thus arise, not only one rudimentary head,
but several, indeed very many heads. Such a finn is called a Coenurus.
It occurs in Ta/nia ctcnurus. In the finn known as Ecliinococcus (of
Td'itia Ecliinococcus of the dog) there arise internally in the vesicular
body by invagination of the wall numerous daughter vesicles, and
even two generations of such vesicles, on whose walls several heads
form.

We must further remark here that asexual scolices living free in
water have been observed.

The Influence of Parasitism on the Structure and Development

of Animals.

In the race of the Platodes, for the iirst time among the Mctazoa, the parasitic
manner of life is met with as a very widely spread phenomenon, Of the three classes
which form this race, the two classes of the Trcmatoda and the Ccstoda consist

in INFLUENCE OF PARASITISM ON STRUCTURE 173

entirely of parasitic forms, while most of the Turbellaria live freely. The transition
from the free life to the parasitic brings with it such far-reaching changes in the
conditions of existence that the original organisation, development, and life-history
of the animals must necessarily be strongly influenced by it. This influence can be
stated in a way which suits all cases in the animal kingdom where, in a naturally
demarcated animal group, parasitic forms appear side by side with free-living forms.
Similar variations in the conditions of existence have as a consequence similar varia-
tions in structure and development.

We can, apart from fine shades of difference in manner of life, distinguish two
principal groups of parasites : (1) the Ectoparasites, which are parasitic on the outer
surface of other animals, and (2) Endoparasites, which are parasitic in the intestine
or other inner organs. The ectoparasites in many ways form the transition from
non-parasitic to endoparasitic animals, for they still retain relations to the outer
world which the latter have entirely given up.

Parasitic life is the most convenient manner of life for the attainment of food.
Parasites feed at the expense of the juices or tissues of their hosts, which are
abundantly within their reach. Once on or in the host's body, it is of the greatest
utility for them to retain the position they have gained. Hence the numerous and
varied adaptations for the attachment of the body. In the Trematoda and Cestoda
we find suckers, hooks, and protrusible proboscides armed with barbed hooks and
other organs of adhesion. Many parasites possess a sucking apparatus to suck the
juices of the host. Trematoda suck with the oral sucker and pharynx the mucus on
the surface of the body, or the food pulp in the intestine, etc.

The ectoparasitic Trematoda possess a well- developed alimentary canal, which is
often even richly branched ; in the endoparasitic forms, which are supplied with food
already partly dissolved, the work of digestion is facilitated. The intestine in endo-
parasitic Trematoda is reduced to two main branches or to a simple caecum ; in the
Sporocyst generation, which multiplies parthenogenetically, it has become quite rudi-
mentary. Here feeding takes place simply by the diffusion of the juices of the host
through the outer skin of the parasite. The same is the case in the Cestoda, which
have entirely lost their alimentary canal. "We may therefore state that progressive
accentuation of parasitism is accompanied by progressive reduction of the gastro-canal
system, ending in its entire disappearance.

The capacity of active locomotion is generally of very little use to endoparasites.
"We accordingly find in them that those parts which serve for locomotion, locomotive
organs and musculature, are more or less reduced. On the other hand many
ectoparasites (not indeed exactly Platodes) possess a well-developed capacity of
locomotion, which is of great importance to them, chiefly for the object of infecting
new hosts (e.g. the flea). Very many ectoparasites can, in fact, live a free life for
a time. The locomotory system and its musculature are therefore generally less
degenerated in them than in endoparasites.

In consequence of the very limited locomotion of endoparasites the power of
directing themselves by special sensory organs is unnecessary, at least while para-
sitism lasts. The ectoparasitic Trematoda already are far more sparingly supplied
with sensory organs than the free-living Platodes. They still possess eyes, although
of a very simple sort. The endoparasitic Trematoda have lost even these sensory
organs, which occur only temporarily in the freely moving young stages of the
ciliated larvae and the Cercarice. In the Cestoda special sensory organs are altogether
wanting.

The degree of development of the nervous system depends (1) on that of the
musculature, and (2) on that of the sensory organs. We thus understand the
gradual simplification of the nervous system, especially the sensory portion, from
the ectoparasitic Treinatoda to the endoparasitic, and finally to the Cestoda. On

174 COMPARATIVE ANATOMY CHAP.

account, however, of the strong development of the musculature of the organs of
adhesion the nervous system in relation with them is more or less strongly developed.
Compare the strong development of the nervous system in the head of the Cestoda
with its great reduction in the segments.

Parasites seem to have a very slightly developed need for respiration. Judging
from what we find in other divisions of the animal kingdom, the respiratory organs
very often become degenerated, especially in endoparasites. The parasitic Platodes
have no covering of cilia.

The excretory system in the Platode parasites is developed at least as strongly
as in the free-living forms.

The genital organs also are as strongly developed, indeed even more strongly
developed, in the former than in the latter. Ripe Distoma or ripe segments of Cestoda
consist almost exclusively of the genital apparatus and the genital products. But to
this we shall return.

We therefore see that with increasing accentuation of the parasitic mode of life
there is a proportional reduction of the sensory organs of the nervous systems, of the
special digestive system, of the locomotory organs, and also of the respiratory organs,
and thus a degeneration of all the organs except the genital and the excretory organs
and the organs for adhesion and sucking.

The influence of the parasitic mode of life on the development, and generally on
the whole life-history of the parasite, is not less striking.

If the parasite were to remain during life and in all stages of development
parasitic on or in the same host, it would perish when the latter died, and the
whole race to which it belongs would soon also perish. There must therefore be
some provision or other for the infection of new hosts. This infection takes place in
the simplest way in most ectoparasites. Many of these retain during youth their
free mode of life, so that they can themselves seek out their hosts. Others retain
throughout their power of free locomotion, and vividly recall. in their mode of life the
beasts of prey.

In the ectoparasitic Trematoda very little is known about the manner of infection
of new hosts, but we do know — and this is very important — that the course of
their development is direct and without intermediate hosts belonging to animal
groups different from that of the final host. In the endoparasites the life-history
is, as we have seen, more complicated. But here also originally free-living young
forms must have provided for the spread of the individuals and the infection
of new hosts, and thus for the preservation of the race. The observation of free
scolices gives countenance to the presumption that originally a free-living young
form, a scolex, developed from the fertilised egg, and in some way or other again
found its way into the body of the final host. Most parasites are specialists, i.e. they
thrive only in the bodies of one or of a few definite animal species. It is certain,
however, that of their eggs or young forms only very few on the average reach the
bodies of true hosts ; many perish without reaching any host, many find their way to
the wrong place, go astray in the body of a host other than the usual one and there
perish, or they may for a time hold their ground and also, as experience shows,
develop somewhat further, never, or very seldom at any rate, attaining full develop-
ment. This perhaps throws light on the origin of development by means of inter-
mediate hosts. Carnivorous animals devour certain animals as their favourite
food ; the latter are themselves carnivorous or herbivorous. If the eggs or young of
a parasite accidentally reach the body of an animal which is the favourite
food of its proper host, and if they could there remain alive for a longer or shorter
time, the probability of their reaching in their new (intermediate) host the intestine
of their proper host would be greatly increased. This or some similar advantageous
manner of being smuggled into the body of the proper host might become established

in STROBILATION AND SEGMENTATION 175

as that most advantageous for the preservation of the race, and would finally become
the normal mode of infection.

In the systematic review the intermediate hosts of several Trematoda and Cestoda
are given, as well as the final hosts. The biological relations between the host and the
intermediate host can easily be recognised, especially in the case of the Cestoda.

Occasionally there are two intermediate hosts in the normal course of life. Free-
living young forms, e.g. the ciliated larvse of the Trematoda, often effect the transi-
tion of the parasite from host to intermediate host, or, as the Cercarice, from inter-
mediate host to definitive host.

In the Cestoda it is possible for parasitism to thrive to such an extent by the
passive transmission of various stages of the parasite from host to host that the
animals never lead a free life. The degenerating influence of the parasitic mode of
life has here told upon all stages of development.

However refined the artifices for infecting new hosts may be, the result of the
process must always^ to an extraordinary extent depend upon chance. It is a chance
when the egg or the embryo of the Distoma hepaticum reaches the water, a chance
when it meets a Limna&us truncatulus, a chance when the encysted Cercaria, with the
plant on which it lies, is eaten by a sheep. Thousands and thousands of eggs thus
miss their aim. There is therefore another way of providing for the maintenance of
the race in parasites, viz. their extraordinary fruitfulness and their highly
developed capacity of reproduction. This capacity is very easy for them, because
the conditions of existence in which they find themselves are the most favourable
possible. A Distoma, indeed a single proglottis of a Tcenia, is capable of producing
thousands, or even hundreds of thousands, of eggs and embryos. And if of all these
eggs but 1 or 2 on the average reach their aim, the maintenance of the race is
provided for. Propagation by gemmation comes to the assistance of sexual propaga-
tion by fertilised eggs in the segmented tapeworms and in the young forms known
under the names of Echinococcus and Coznurus,

In those cases also, in the Trematoda, where generations living in the so-called
intermediate host are not surrounded by conditions so favourable that they can
develop into adult Trematoda with male and female organs, they still possess the
capacity, in spite of their reduced condition, of producing at an early stage a sort of
egg, the germ cells; these dispense with fertilisation and nevertheless develop
(parthenogenetic reproduction of the Sporocysts and Redice).

When we consider the degenerated condition of the Sporocysts there is some
justification for assuming that the Dicyemidce and Orthonectidce (cf. p. 60), which are
very similar to these Sporocysts, are degenerated Trematoda from whose life-history
the typical Trematode generation has completely disappeared.

Strobilation and Segmentation.

We have seen that the bodies of most Cestoda are segmented, and
we have shown this segmentation to be the result of an axial budding
or strobilation. The whole segmented body is thus an animal stock.
In a few Turbellaria, especially in Gunda, segmentation also occurs, but
in quite another way ; this is the regular paired arrangement of the
organs which in the Polydada and Triclada are generally present in
considerable numbers. There is a repetition at regular intervals of
the transverse commissures of the nervous system (the ladder nervous
system), the male and female sexual glands, the lateral intestinal
branches, the dissepiments lying between them, and the external

176 COMPARATIVE ANATOMY CHAP, m

apertures of the water -vascular system. Such a segmented body
represents a simple Platode individual, not a Platode stock; it does
not arise by budding. Strobilation and segmentation are therefore
to be clearly kept apart.

Literature.

L. v. Graff. Monographic der Turbellarien. I. Rliabdocoelidce. Leipzig, 1882.
Arnold Lang. Die Polycladen (Seeplanarien) des Golfes von Neapel. (Fauna und

Flora des Golfes von Neapel XL) Leipzig, 1884.
Is. Ijima. Ueber Bau und Entwickelung der Susswasserplanarien (Tricladen}.

Zeitschr. f. wissensch. Zoologie. 1885.

R. Leuckart. Die Parasiten des Menschen. 2d edition not yet completed.
Numerous treatises and works by Sieboldt, Leuckart, Pagenstecher, Ercolani, Vogt,

M. Schultze, 0. Schmidt, Quatrefages, Hallez, v. Graff, Selenka, Gbtte, Wagener,

v. Beneden, Zeller, Braun, v. Kennel, Schauinsland, Thomas, Sommer, Landois

Pintner, Kiichenmeister, Fraipont, Lang, etc.

CHAPTEE IV

THE ORGANISATION AND DEVELOPMENT OF THE WORMS

(VERMES)

THE race of the worms is, even after the exclusion of the Platodes, which
till now have been included in it, by no means a natural, well-demarc-
ated division of the animal kingdom ; now, as heretofore, it is like a
lumber room, to which all those groups are relegated which cannot be
placed elsewhere. It is therefore difficult to characterise the race of
the worms in other than negative terms. All worms are bilaterally
symmetrical animals ; their detailed structure, however, is most varied.
They are raised above the Ccelenterata and Platodes by the possession
of an anus, and of a blood-vascular system which undertakes
physiologically one of the functions of the gastro-canal system of these
animals. Where these systems are wanting a secondary degeneration
has perhaps taken place. The mouth lies at the extreme anterior
end of the body, originally always on the ventral side. A body
cavity is either wanting, or is developed in varying degrees. Under
the outer body epithelium there is found in all unshelled forms a
generally strong muscular layer (derm o - muscular tube). The
nervous system is developed in very different ways. The only
constant point is the presence of a nerve centre placed above the
oesophagus (brain, supra -oasophageal ganglion). There is also
generally a nerve ring surrounding the oesophagus, the cesopb^gecc]
ring, from which longitudinal trunks run backwards in trying
number, position, and arrangement. All these portions — brain,
cesophageal ring, and longitudinal nerves — belong to the central
nervous system. Excretory organs (nephridia) are found in all
divisions, but under the most different conditions. They often perform
the function of conducting the sexual products out of the body.
Segmented body appendages (extremities) are as completely wanting
as is a specialised muscular organ of locomotion placed on the ventral
side (foot). A strictly localised central organ of the blood-vascular
system (heart) has been observed only in the ^""

VOL. I

178 COMPARATIVE AXATUMY CHAP.

THE FOURTH RACE OR PHYLUM OF THE ANIMAL KINGDOM.

VERMES.
Systematic Review.

CLASS I. Nemertina ^Rhyncoccela).

Body ciliated, externally unsegmented, elongated, generally somewhat flattened
dorso-ventrally. "\Yith out distinct body cavity, intestine straight, mostly with
lateral diverticula, anus at the posterior end of the body. Above the intestine a
proboscidal apparatus, generally emerging in front of and above the mouth. The
central nervous system consists of a brain lying between the proboscis and the oeso-
phagus, and of two lateral longitudinal trunks. Blood-vascular and excretory systems
present. Sexes separate. By regular repetition of the inner organs (lateral intestinal
divertieiiX , alar commissures of the longitudinal nerves, sexual glands) a sort of
on often arises (Pseudometamerism). Almost exclusively marine.

Order 1. Palseonemertina.

Head without deep lateral longitudinal furrows. Proboscis without stylets.
Mouth behind the brain. Carinclla, Polia.

Order 2. Schizonemertina.

On each side of the head a deep longitudinal groove. Proboscis without stylets.
Mouth behind the brain. Lincus, Borlasict, Ccrclratnlus, Lanyia.

Order 3. Hoplonemertina.

Head without deep lateral longitudinal grooves. Proboscis armed with one
stylet or several. Mouth generally in front of the brain. Ampliiporm, Drcpano-
phorus, Tetrastcmma, Ncmcrtcs.

Order 4. Malacotadellina

Head without lateral longitudinal grooves. Proboscis without stylets. One
sucking disc at the posterior end of the body. MalacoldcHa. Parasitic in marine

mussels.

CLASS II. Nemathelmia.

Body cylindrical, spindle-shaped, or thread-like, unsegmented, covered with a

fl ' cmi. le. Body cavity generally spacious. Intestine straight or wanting.

posterior end of the body. Neither blood-vascular nor excretory

<.7,ij Arable with those of any other worms. Sexes usually separate. Ner-

--em an (esophageal ring, a medio- dorsal, and a medio-veutral longitudinal

"* Vii inner metamerism is wanting, but the circular commissures of the longi-

ti rves may repeat themselves in the Nematoda with tolerable regularity.

Me vsitic.

Order 1. Nematoda.

\\*K; ; -tinal canal, without proboscis. Family Enoplidcc, mostly fret • living

in the •• «. , frequently in fresh water or on land, without resophageal 1ml f often

with ey .-. uly ^nfjuiH-nlida:, small, partly parasitic, partly free-living i-mimals.

with do ' --'•'• Tylenchus scandcns, in grains .'f wlicfl" .

A'/iguili Or, etc. Rhabditis ni</rovenos<:. in i;

IV

VERMES— SYSTEMATIC REVIEW

179

B

muddy earth. Sexes separate. The females are viviparous, and produce only a few
young (4 at most), which, after being hatched, find their way into the lungs of
frogs and toads, and there
develop into mature her-
maphrodite animals (Ascaris
nigrovcnosa}, from whose fer-
tilised eggs the free -living
Rliabditis generation again
arises. The life-history thus
exhibits a sort of heterogeny.
Sphcerularia bombi, the Rhab-
ditis-like young form lives
in the earth. The fertilised
females find their way into
the female humble bee, where
they are parasitic in the body
cavity or in the intestine.
The uterus, which is filled
with embryos, soon begins to
hang out from the female
genital aperture, and becomes
a large pouch, to which
the worm-body finally forms
merely a small insignificant
appendage. Mermithidce, with -
out anus. The young are
parasitic in the body cavity
of Insects ; they make their
way out into damp earth,
where they become sexually
mature and reproduce them-
selves ; Mermis nigrescens.

Filariidce : Filaria medin-
ensis (medina worm), may
have a breadth of 0*5-2 mm.
and length of nearly a metre ;
in the subcutaneous connec-
tive tissue of man, in warm
regions of the Old World.
The young in small Crustacea
( Cyclopidce}. Trichotrachelidce:
Trichocephalus dispar, with
swollen hinder body, in the
human csecum.

Trichina spiralis (Fig.
121) lives sexually mature
as the so - called intestinal
Trichina in the small intestine
of man and in that of many
mammals ; is viviparous ; the
female (£) is ca. 3 mm., the male (O) half as long. The young bore their way
into the intestinal wall, pass from here through the body cavity, or with the blood

FIG. 121.— Trichina spiralis (after Glaus). A, Encysted
muscle Trichina ; mf, unaltered muscle fibres ; /, fat globules ;
cy, cyst ; bh, envelope of connective tissue. B, Female intes-
tinal Trichina ; ov, ovary ; e, embryos ; zk, cell bodies (in the
post cesophageal division of the intestine) ; wo, female aperture.
C, Male intestinal Trichina ; cc, oesophagus ; h, test-is ; de,
ductus ejaculatorius.

ISO COMPARATIVE ANATOMY CHAP.

in the veins into the musculature, penetrate the muscle fibres, and there surround
themselves with a cyst or capsule which afterwards becomes calcareous (A}.
Encysted muscle Trichina cause trichinosis. Men become infected by eating
trichinous pork which is uncooked or not sufficiently cooked. Pigs are much
exposed to infection on account of their omnivorous habits. The principal carriers
of Trichina, however, are rats which happen to eat the dead bodies of infected
animals, and so secure the continued existence of the parasites.

Strongylidcc : Dock rains (Anchylostoma) duodcnalis, with strong oral capsule
armed with teeth. Female as long as 2 cm., male half as long. In the human small
intestine (Egypt, Brazil, India, the Antilles, Switzerland, Italy, Belgium). Chiefly
among labourers in pits, mines, or tunnels. Causes the so-called miner's annemia.
Eustrongylus yiga.s, female 30-100 cm. long. In the pelvis of the kidneys of the
dog and other mammals. Ascaridtc: Ascaris lu/nbricoides, male up to 25 cm.,
female to 40 cm. in length. In the human small intestine. Oxyuris vermicular is,
female up to 1 cm. in length, male half as long. In the human large intestine.
Specially common in children.

The GordiidcG hold an isolated position among the Ncmatoda, on account of
peculiarities of inner organisation which will be spoken of in the anatomical portion.
The mouth in adult animals is closed, and the intestine partly degenerated.
Gordius aquations, in the adult sexually mature condition free in fresh water.
The embryos find their way into insect larvae, where they become encysted. If these
larva? are devoured by preying insects which live in water, the embryos develop in
the body cavities of these new hosts, escaping into the water as sexual maturity
approaches. 30-90 cm. long, 1 mm. thick.

Order 2. Acanthocephala (Fig. 172, p

Mouth and intestine wanting. At the anterior end a protrusible proboscis
armed with hooks. Entirely parasitic. EcJiinorJiyncliUft : E. j/'/".s. in the small
intestine of the pig. Larva? in cockchafer grubs.

CLASS III. Annulata (Segmented Worms).

Body elongated, cylindrical, or more or less flattened dorso-ventrally. Skin soft,
or with more or less hard and rough chitinoid cuticle. Marked metamerism or
segmentation of the inner organs ; generally also outwardly visible. The body cavity,
with the exception of lliftulinm and Myzostomidou, well developed. The blood-
vascular system generally well developed, seldom quite degenerated. The intestinal
canal runs mostly in a direct lino from the mouth to the terminal anus. The
nervous s}-stem consists of a brain, an u-sophageal ring, and a usually distinctly
segmented chain of ventral ganglia. The excretory system (wanting in Myzostoiwi)
consists of segmentally arranged paired nephridia. Nephridia often perform the;
function of conducting the sexual products to the exterior.

Order 1. Hirudinea^Discophora (Leeches).

Body externally ringed. A definite number of consecutive rings correspond
with an internal segment. Round the mouth an oral sucker, at the hinder end
of the body under the anus a ventral sucker. Skin soft. Sctie wanting. Intes-
tine mostly with paired lateral diverticula. Body cavity degenerated and com-
municating with the well-developed blood-vascular system. Numerous pairs of
nephridia segmentally arranged (looped canals), which are not used for conducting
the sexual products to the exterior. Hermaphrodite : testes in several pairs, seg-

iv VERMES— SYSTEMATIC REVIEW 181

mentally arranged, with special efferent ducts ending in a single median external
aperture. A pair of germaria placed in front of the testes ; female aperture behind
the male ; both in the anterior end of the body. Parasites or carnivorous ; fresh
water, in the sea, or on land.

Sub-Order 1. Rhynchobdellidse.

With cylindrical tubular pharynx lying free in the pharyngeal pouch, and pro-
trusible through the mouth. Clepsine, Pontobdella, JBranchellion (with gill-like
appendages on the back). The two latter in the sea, parasitic on sharks.

Sub-Order 2. Gnathobdellidse.

The pharynx is a muscular thickening of the cesophageal wall, which projects
into the lumen in the shape of 3 plates or ridges occasionally toothed.

Hirudo medicinalis (the ordinary leech), Hcemopis Aulastomum (horse-leech),
Nephdis. Some Hirudinidce live on land. The remaining Gnathobdellidce are
fresh-water forms.

Order 2. Chsetopoda.

The external segmentation of the body mostly corresponds with an inner segmenta-
tion. In special segmentally arranged glandular saccules of the outer integument
(setigerous glands) arise setae, which project freely above the skin. Body cavity
well developed, separated from the blood -vascular system. The sexual products
develop in special regions of the epithelial lining (endothelium) of the body cavity,
into which they generally soon fall, and escape thence through more or less
strongly modified nephridia (vasa deferentia, oviducts, genital pouches, segmental
organs). The following division is artificial, and is only retained from practical con-
siderations.

Sub-Order 1. Oligochseta.

The setigerous glands do not open on special blunt processes of the body
(parapodia) ; outer appendages (antennae, feelers, feeler -cirri, cirri, gills, etc.) are
wanting. Hermaphrodite. Direct development. In fresh water or on land. Fam.
Aphanoneura: ^Eolosoma. Fam. Naidomorpha : Naist Dero, Stylaria. Fam.
Chcetogastridce : Chcetogaster. Fam. Discodrilidce (posterior end of body modified
into a sucker ; parasitic on Crustacea] : Branchiobdella. Fam. Enchytrceidce : Pachy-
drilus, Enchytrceus, Anachceta. Fam. Tubificidce : Tubifex, Psammoryctes, Clitellio,
Limnodrilus. Fam. Phreoryctidce : Phreoryctes. Fam. Luiribriculidce: Lumbriculus,
Rhynchelmis, Stylodrilus. Fam. Criodrilidce : Criodrilus. Fam. Lumbricidce:
Allurus, Dendrobceiut, Allolobophora, Lunibricus (earth-worm). And the related
forms Urocheeta, Eudrilus, Acanthodrilus, Perichceta, Pleurochceta, Moniligaster.

An uncertain position within the Chcetopoda is taken by the so-called Archi-
annelida (Polygordius, Protodrilus, Ctenodrilus, Histriobdella) and Saccocirrus —
forms whose organisation exhibits simple embryonic characteristics.

Between the Oligochceta and the Polychceta stand the families of the Capitellidce
(Capitetta, Notomastus, Dasybranchw), and Ophdiacca (Ophelia, Travisia, Polyoph-
thalmus). The first have no blood-vessels. In both the parapodia are much reduced.
Gills may be present or absent. Head not distinctly separated.

Sub-Order 2. Polychseta.

The setae stand on very strongly developed segmentally arranged elevations or
parapodia. On the head, antennae or feeler-cirri ; on the trunk segments, cirri, gills,

182 COMPARATIVE AXATOMY CHAP.

or other appendages, which generally stand on the parapodia. Sexes mostly separate.
Development with metamorphosis. Marine.

A. Sedentaria = Capitibranchiata (Tubicolous worms).— Pharynx (proboscis)
mostly non-protrusible ; without jaws. Eyes wanting, or small but numerous
in the head. Parapodia slightly developed, the upper with hair-like set;e ; the
lower are transverse swellings with hooked set;e. Gills chiefly on the anterior
segments, or limited to the head. Live in various kinds of tubes. Fain. Cir rat uJ idee :
Cirratiilus. Fain. Arcnicolidce: Arcnicola. Fain. Spionid"' : Spio. Fain. Arid idee :
Arii'ia. Fain. Chlorhccmidcc : Siphonostoma. Fam. Tt'reli Hula: : Lanicc (Terebella),
Potymnict, Amphitrite. Fam. Serpulidcs: Serjmla, Scibclla, SftiTographis, Myxicola,
Protula. Fam. IfcrmeUidcc : Saldlaria (Fig. 147, p. I'll). Fam. Stcrnaspidn: :
Stcrnaspis.

B. Errantia = Dorsobranchiata (carnivorous Annelids).— Pharynx protrusible,
mostly with jaws ; head distinct, mostly with a few large eyes. Parapodia well
developed. Gills generally on the dorsal parapodia. Free-swimming or creeping
animals, many of which live, at any rate at certain times, in tubes of their own
making. Fam. Aphroditca : Aphrodite, Hcrrnionc, Polynoe. Fam. Amphiiwmidce :
Ampltinomc, Eiiyhrosync, Notyryyos. Fam. Eunicidcc : Diopatra, Eunice (Fig. 124,
p. 188), Holla, Fam. Ncreidw : Nereis, Nephthys. Fam. Glyccrido: : Glyccra. Fam.
Syllidw : Hcqrtosyllis, Syllis, Evogone, Autolytus, Myrianida. Fam. ffenionidtv :
Hcsionc. Fam. Phyllodocidce : Phyllodocc. Fam. Alciopidcc : Alciope, Asteropc.
Fam. Tomoptcridcc : Tomoptcris.

Sub-Order 3. Echiuridse.

Body tubular, in adult condition unsegmented' or indistinctly segmented,
without parapodia, without cirri, without gills. In front, on the ventral side, 2
bunked seta-. In the terminal portion of the much-coiled intestine 2 anal glands
enter (modified nephridia of the anal region), which may be considered as excretory
organs. There are besides 2 or 3 pairs of nephridia, or only one nephridium.
Anterior end of the body over the mouth produced into a long variously shaped
mobile prostomium (cephalic lobe) with ventral furrow. With blood- vascular system.
Sexes separate. Development with metamorphosis. Marine animals with unknown
mode of life. EcJtiums (Fig. 148, p. 223), T/talKssnna, Londlw. The, diminutive
Turbellaria-like ciliated males of this last species live parasitic-ally in the females.

Order 3. Myzostomidse.

Body flat, disc-shaped, externally unsegmented. Margin of the body with cirrus-
or wart-like processes. On the ventral side, in two longitudinal rows, 5 pairs of
parapodia, with hooks and supporting rod. In addition to the parapodia, on each side
4 suckers. Pharynx as in the Rhynchobdeflidw among the Jliruilincii. Intestine
with branched lateral diverticnla. Body cavity reduced. Circulatory, excretory, and
respiratory organs wanting. The nervous system consists of the oesopliageal ring,
and of a ventral chord fused into a ventral ganglionic mass. The brain is reduced.
Hermaphrodites. The oviducts and the intestine together enter a cloaca. The
seminal ducts have two separate external apertures on the ventral side. Besides
hermaphrodite individuals, there are, in certain species, small males (complementary
males). Parasitic on Crinoids (Fig. 175, p. 262).

CLASS IV. Prosopygia.

Body naked or shelled, varying greatly in form. Round the mouth a circle of
ciliated tenacles, which are often inserted on a common horseshoe-shaped tentacle-

iv VERMES— SYSTEMATIC REVIEW 183

carrier (lophophore), which, may itself be produced into arms on each side. Without
parapodia, and mostly without setse. The anus is nearly always placed far to the
front. The intestine runs backwards, but bends forwards again in a loop. Body
unsegmented, or quite indistinctly segmented. Blood- vascular system wanting, or
developed in varying degrees. Number of nephridia reduced (at the most 2 pairs) ;
they occasionally serve as ducts for the transmission of the sexual products, and
emerge to the front not far from the anus. Sexes separate. Only Phoronis is
hermaphrodite. Live in the sea ; only a few forms in fresh water.

Order 1. Sipunculacea.

Body elongated, tubular, naked. The front part of the body, which is mostly
thinner, can be invaginated like a proboscis into the larger and longer hinder part
(trunk) by means of special retractors. Body cavity very spacious. Blood-vascular
system (?) much reduced or wanting. The central nervous system consists of brain,
cesophageal ring, and median ventral longitudinal trunk. Segmentation is perhaps
denoted by nerve rings, which are regularly repeated. Live in mud or lurk in
holes ; marine.

Sub-Order 1. Sipunculidse.

Anus dorsal and anterior, on the boundary between the proboscis and the trunk.
Mouth surrounded by tentacles. Usually 2 typical nephridia, opening externally
in the neighbourhood of the anus ; also serving as efferent ducts for the sexual
products. The vascular system consists principally of 2 tentacle vessels, which
accompany the fore-gut. Sipunculus (Fig. 138, p. 208), Phascolosoma.

Sub-Order 2. Priapulidae.

Anus dorsal at the posterior end. No tentacles around the mouth. No blood-
vascular system. No nephridia. Two anal glands which emerge close to the anus
and function in the young as excretory organs and in adults as genital organs.
Priapulus (at the posterior end a tuft of appendages which probably function as
gills), Halycryptus (without this caudal appendage).

Order 2. Phoronidae.

Body vermiform, in an attached chitinous tube. Numerous tentacles stand on a
horseshoe -shaped base around the mouth. Anus quite to the front, towards the
dorsal side, near the mouth. Around the mouth a nerve ring (cesophageal ring).
Two nephridia emerging to the front, serving at the same time as ducts for the
transmission of the sexual products. A simple blood-vascular system present.
Hermaphrodites. Single species : Phoronis.

Order 3. Bxyozoa.

Small animals. Anus towards the dorsal side, near the mouth. A cerebral ganglion
between mouth and anus. Nephridia, when present, in one pair of embryonic type,
emerging near the mouth and not acting as ducts for the transmission of the sexual
products. Numerous tentacles on a base, which is often horseshoe-shaped, around the
mouth. Form, mostly by gemmation, variously shaped attached stocks.

Sub-Order 1. Pterobranchia.

Tentacle-carrier lengthened out on each side into a long arm-like process directed
dorsally and posteriorly, and carrying the little tentacles in two longitudinal rows
along its whole length. Intestine limited to the swollen anterior part of the body,

184

COMPARATIVE ANATOMY

CHAP.

which is prolonged backwards like a stalk. Body cavity little developed. Colonial,
in tubes which raise themselves on a common creeping stem. Rhabdopleura,
Cephalodiscus is a related form.

Sub-Order 2. Ectoprocta.

Anal opening outside of the tentacle carrier. Tentacle carrier not produced.
Anterior body naked, posterior body shelled. Without stalk. Anterior body so
enveloped in a fold of the posterior body as to be either temporarily or permanently
surrounded by a sheath (tentacle sheath) out of which it can be protruded. Body
cavity tolerably spacious. Shell often calcareous. Colonial. Phyladolcemata.

Tentacle carrier horseshoe-shaped. In-
habit fresh water. Cristatella, Alcy-
onella, Fredericella, Lophopus (Fig.
122), Plumatella (Fig. 139, p. 208).
Gymnolcemata. Tentacle-carrier circu-
lar. Live, with the exception of Palu-
dicella, in the sea. Cellepora, Eschara,
Bugula, Fhistra, Alcyonidium, Hor-
nera, etc.

Sub-Order 3. Entoprocta.

Anal opening inside the tentacle
carrier. A tentacle sheath wanting.
Body stalked. With one pair of neph-
ridia. Body cavity reduced. Pedi-
cellina, colonial. Loxosoma, living
singly. Marine.

Fio. 122.— Anterior body of Lophopus (after
Allman), from the right side, t, Tentacles cut off
near the base ; o, mouth ; ep, epistome ; st, fore-

gut ; ga, ganglion ; an, anus ; pr, hind-gut,
free ends of the tentacle-carrier are cut off.

The

Order 4. Brachiopoda (Fig. 178, p. 269).

The dorsal and the ventral body
walls form a fold directed to the front,
so that the body is covered by dorsal
and ventral mantle folds, which may
coalesce behind and at the sides. The mantle folds secrete calcareous, and occasionally
horny shell valves, a dorsal and a ventral. The ventral valve is generally the more
concave. At the sides of the mouth are inserted the two long oral arms set with
lateral cirri ; these arms lie spirally rolled up in the mantle cavity, which is
formed by the mantle folds, and are often supported by a special calcareous skeleton
united to the dorsal valve. Anus wanting, or lies asymmetrically and anteriorly,
to the right, near the mouth. In Crania only it lies quite behind, in the dorsal
middle line. Central nervous system is an oesophageal ring with weakly developed
brain and infra-cesophageal ganglionic swellings. One pair (less frequently two pairs)
of nephridia, at the same time ducts for the transmission of the sexual products, open
to the right and left of the mouth into the mantle cavity. Blood-vascular system
probably present, with a heart placed above the intestine. The posterior end of the
body is often prolonged into an attached stalk, which emerges either between the
shell valves (Lingula), or through a hole in a posterior upward bulging of the larger
ventral valve. In many cases the stalk is wanting, and the shell is fastened to the
surface on which it rests direct by the ventral valve. Exclusively inhabitants of the
sea. The larger proportion of the species and genera are fossil. The
has maintained itself since the Paleozoic epoch.

IV

VERMES— SYSTEMATIC REVIEW

185

Sub-Order 1. Testicardines.

The shell valves are hinged together by interlocking processes. An anus is want-
ing. Terebratula, Waldheimia, Theddium (attached by the larger shell valve).
Argiope, Rhynconella (Fig. 125, p. 191), Spirifer.

Sub-Order 2. Ecardines.

Without hinge. Intestine with anus. Crania, Lingula.

m, —

CLASS V. Eotatoria (Wheel Animalcule, Fig. 123 ; Fig. 161, p. 245).

Small, mostly microscopic animals. Inner segmentation wanting. At the
anterior end a variously shaped ciliated organ (wheel organ) ; posterior end prolonged
in a frequently segmented appendage (foot-stalk). Cloacal aperture dorsal on the
boundary of body and foot. A vascular system
wanting. One pair of nephridia, of the embryonic
type, with several inner ciliated cells, emerge with
the anus and the oviduct into the cloaca. Sexes
separate. The males small, with degenerated ali-
mentary canal. Chiefly fresh -water animals. Forms
living in attached tubes or envelopes : Floscularia,
Stephanoceros, Melicerta, Lacenularia. In these the
wheel organ is produced into lobes or tentacles.
Free-living forms : Notommata, Hydatina (Fig. 161,
p. 245), Brachionus (carapaced), Asplancha, Seison
(parasitic on Nebalia).

"We must place near the Rotatoria the peculiar
species Dinophilus (Fig. 162, p. 246), which has the
same appearance as certain annelid larvae. Male
and female are alike, or the male is smaller, with-
out intestine. The whole ventral side of the body
is ciliated. Besides this there are several ciliated
rings lying one behind the other on the body. A
wheel organ is wanting. Nephridia in segmental
arrangement of the embryonic type.

-I— -met

Appendage to the Race of Vermes.

FIG. 123.— Diagrammatic repre-

VI. Chstognatha (Fig. 152, p. 227).
Body cylindrical, elongated, with horizontal organ ; 3, brain ; a, eye ; p^, pharynx;
border of fins (lateral fins, caudal fin) round the ZSZ%&.* %£%%&%
posterior body. Head tolerably distinctly separated cioaca ; M, cement gland of the
from the trunk. Body cavity spacious ; divided by foot (/).
partition walls into three consecutive parts — head

cavity, trunk cavity, and tail cavity. Mouth surrounded by setae (jaws). Intestine
straight, anus ventral, at the commencement of the tail. The central nervous
system consists of the brain, the oesophageal commissures, and a large ventral ganglion
lying in the trunk. Without vascular system. Hermaphrodites— ovaria in the trunk
cavity, testes in the caudal cavity. Paired efferent ducts (transformed nephridia ?)
emerge to the right and left at the end of the trunk and of the tail. Live in
the sea. Sagitta, Spadella.

186 COMPARATIVE ANATOMY CHAP.

Among the worms, generally near the Rotatoria, we place the little group of the
Gastrotricha. Small animals with ciliated ventral surface ; with pointed processes
arranged in longitudinal rows on the dorsal surface. Posterior end running out into
two lateral points. Intestine straight, the mouth followed by a muscular pharynx.
Anus at the posterior end. Hermaphrodites. Nephridia insufficiently known.
Blood-vascular system wanting. Ichthydium : principally inhabitants of fresh
water. The Echinodera, which are often placed near the Nematoda, are minute
marine animals without cilia, with ringed bodies ; with setse. Inner organisation
not sufficiently known.

The phylogenetic relations of the worms are still a subject of much dispute.
There are many different views. The Nemertina form a natural well-demarcated
class which in many points of their organisation (nervous system, excretory
system, absence of body cavity) recalls the Turbellaria, but they are raised
above the Platodes by the possession of a blood-vascular system and an anus. The
pseudo -metamerism of the Nemertian body is similar to that of certain Turbellaria
( Triclada). The systematic position of the Nemathelmia is quite uncertain. Their
ancestors were probably more highly developed worms, in whom adaptation to
the parasitic mode of life has led to degeneration. Perhaps the Gordiidce among
the now living Nematoda are the nearest to the ancestral form. The Annulata
form a large group extraordinarily rich in forms, in which the typical segmented
condition of the body may be regarded as primitive. The Myzostomidce, Echiuridce,
many simply organised Chcetopoda, and in some respects the Hirudinea, are to be
regarded as one-sidedly developed, partly simplified or degenerated forms. Opinions
are very divided as to the racial history of the whole class. Many investi-
gators, among whom we include ourselves, hold the segmentation (metamerism) of
the Annulate body to be a continuation of the pseudo-metamerism of animals
resembling the Turbellaria and the Nemertina. Others consider the Annulate body
as a sort of animal stock, which has arisen by axial budding. They see in the
Rotatoria the nearest approach to the unsegmented (not budding and not stock
forming) racial form, while we, on the other hand, are inclined to consider the wheel
animalcules as simplified animals which attain sexual maturity at an earlier stage
of development, so that they now no longer rise above the degree of organisation of
a young Annelid larva. The class of the Prosopygia falls into a few natural orders
to some extent sharply distinguished from each other ; their organisation is in many
respects quite comprehensible, if we refer it back to an old adaptation to a more or
less attached mode of life, and bear in mind the degenerating action of shell, case,
or tube formations on the bodies of segmented worms originally more highly de-
veloped. The systematic position of the Ohcetognatha also is very uncertain. They
are, perhaps, best considered as Annulata with a small number (3) of segments.
It has till now been impossible finally to decide the systematic position of Gastro-
tricha and Echinodera.

I. Form of Body and Outer Organisation.

The body of the Nemertina is elongated, ribbon-shaped, being more
or less flattened dorso-ventrally ; it is ciliated all over the surface,
soft skinned, unsegmented, and without outer appendages. The
mouth lies ventrally at or near the anterior end of the body, and in
the form of a longitudinal slit. In front of it, and generally quite at
the foremost end of the body, is the proboscidal aperture. Mouth

iv VEEMES—FOEM OF BODY AND ORGANISATION 187

and proboscidal aperture are united into a single external aperture
only in Amphiporus, Malacobdella, and Geonemertes palaensis. On each
side of the head there is often a strongly ciliated longitudinal furrow
or lateral cleft. The anus is terminal. In the parasitic Malacobdella
the posterior end forms, in front of and under the anus, a ventral
sucker disc.

The body of the Nemathehma, which is covered by a rough,
frequently ringed cuticle, is elongated and spindle-shaped, without
outer appendages, and with at the most small papillae or small fringes
at the anterior and posterior ends. Mouth and anus are, when present,
terminal. The Acanthocephala possess at the anterior end a proboscis
which can be withdrawn into a special sheath, and which is provided
with hooks directed backwards ; this serves for attachment to the
intestinal wall of the host. The males have at the posterior end a
protrusible copulatory organ, which, when protruded, is bell-shaped.

The various classes of the Annulata must be dealt with separately.

The body of the Hirudinea is long, and generally flattened dorso-
ventrally ; less frequently it is cylindrical (Pontobdella, Piscicola).
Round the mouth there is a small oral sucker and posteriorly, always
under the anus, a ventral sucker, which is usually larger than the oral.
The body is divided by furrows into numerous consecutive rings ;
these do not, however, correspond in number with the inner segments.
In all the Hirudinea the number of the latter corresponds with the
number of the ganglia, and is 33. The number of rings to a segment
in the central part of the body is typical for the various genera and
species. Among the Rhynchobdellidce it is 3 (Branchellion, Clepsine), or
6 (Calliobdella, Ichthyobdella, Pontobdella), or 12 (Piscicola); among the
Gnathobdellidce it is 5. In all Hirudinea the segments are shorter at
both the anterior and posterior ends of the body, the number of rings
in them being gradually reduced. Where the separate rings of the
typical central segments are in any special way externally marked,
the marks are repeated in regular succession throughout the body, i.e.
when such marked rings have not disappeared in the shortening of
the segments. These marks take the form of tactile papillae, warts,
protuberances, pigment spots, nephridial apertures, etc. The large
protuberances on the integument of Pontobdella play a most important
part in the dermal respiration. In Brancliellion, on each side of each
ring in the middle region of the body, there is an integumental
appendage which functions as a gill. The appendages on the first
ring of each segment swell at their bases and form contractile vesicles.

In the Chcetopoda the outer segmentation generally answers to the
inner. The former is most distinctly to be recognised by the arrange-
ment of the setae, which are repeated in strict accordance with the
segments. The setae generally stand in groups. Every segment,
except the oral, carries typically on each side two bundles of setae, one
dorsal and the other ventral.

188

COMPA PA TI VE A XA TOM Y

The shape, number,, and arrangement of the set* vary in details to an extra-
dmary degree, and are of the greatest significance for classification. Set* are
wanting only m a few Chcetopoda, as in the EnchytrccUUc (Anachata) and in
brachiobdella among the Oligockceta, in the so-called Archianndida (Polygordius
'odrilus, etc.), and in the Tomopteridv. In Chvtogastcr the dorsal rows of seta'
re wanting ; and in Saccocirrus also only one longitudinal row is found on each side
In the stationary Polychceta each ventral bundle of set* is developed in the form of
nsverse row of short hooks. The bundles can break up into their component
parts, the setre being arranged in Pcrichccta in
a ^single row round the segment. The setre may
disappear in certain regions of the body. They
are specially reduced in number in the Echiuridce
(see systematic review).

The bundles of setae stand either
simply in the integument, or on special
elevations of the body wall, the so-called
parapodia (Fig. 158, p. 237). The
former is the case chiefly in the Oligochceta
and Echiuridce, the latter in most of the
Polychceta. The parapodia are well devel-
oped as strong rowing and locomotory
organs principally in the Errantia, while
they are reduced in the Sedentaria, especi-
ally the ventral parapodia, which are for
the most part insignificant ridges carrying
hooks. In a few tube-worms (e.g. Scr-
pididce) the parapodia are entirely obliter-
ated, no doubt in adaptation to the
tubicolous manner of life. They are also
wanting in the Archiannelida. We do
not always find separate ventral and
dorsal parapodia ; there is often on each
side only one row of parapodia. We
then, however, find in each parapodium
a dorsal and a ventral branch. Whether
the uniserial or the biserial arrangement
is the original cannot yet be decided.

In the Polychceta the parapodia them-
selves again carry characteristic append-
ages (Fig. 124). These are the cirri,
unsegmented or segmented filaments, one
of which, in the simplest cases, occurs on
each parapodium. We can thus dis-
tinguish dorsal and ventral cirri. The
cirri may undergo the most varied trans-
formations. Thus the dorsal cirri, or their lateral branches, frequently
become gills, which are often delicately branched and provided with

Fi<;. ii>4. — Eunice limosa (after
Ehlers). Anterior and posterior ends
of the l>ody ; dorsal side, fa, Unpaired
feeler;//*, paired feelers; a, eyes;/',
feeler-cirri; /;, gills; pc, dorsal para-
podial cirri; p, parapodia; «c, anal
cirri.

iv VERMES—FORM OF BODY AND ORGANISATION 189

blood-vessels. In the Aphroditidce they become broad dorsal scales
(elytra). In some cases (e.g. Capitellidce, Glyceridce) the dorsal para-
podia may carry, besides the typical cirrus or the cirrus transformed
into a lateral organ, a gill, which, in contradistinction to the gill which
arises from the cirrus, is called a lymph gill; it is provided with haemo-
lymph by continuations of the body cavity (in the absence of a blood-
vascular system). The cirri may stand at the base of the parapodia,
or may even move away from them. They may even be retained in
those cases where the parapodia disappear.

The head of the Polychceta is characterised by special appendages,
the front ones being called feelers, the back ones feeler-cirri. They
never stand on parapodia, which are just as constantly wanting in the
true oral segment as are setae. In most of the delicate tubicolous
Polychceta a reduction of the parapodial cirri or parapodial gills goes
hand in hand with the reduction of the parapodia on the trunk ; they
are retained only in the anterior segments. In the Serpulidce all the
parapodial appendages are reduced, and therefore the head append-
ages are transformed into greatly developed tentacle gills which often
form a stately crown. In Sternaspis on each side of the anus there
is a tuft of gills.

In the Oligochceta, Archiannelida, Echiuridce, and some of the Capitel-
lidce, not only the parapodia but their appendages (cirri, gills) are
wanting. Only, Alma nilotica, a very insufficiently known Oligochcete
which is found in muddy ditches in Egypt, carries dorsal gills on the
hinder part of its body. In all these divisions the head appendages
are also wanting. Only the Archiannelida (Polygordius, Protodrilus)
possess two feelers at the extreme front of the head. In the Echiuridce
the head is produced in front of and over the mouth into a long pro-
cess provided with a longitudinal furrow or channel on the ventral
side (proboscis, prostomium); this in Bonellia is forked at its end.

The inner segmentation is reflected outwardly in most Chcetopoda,
not only by the regular repetition of the setae (and the parapodia of
the Polychceta), but also by an outer division of the body into rings,
which is caused by the occurrence of more or less distinct regularly-
repeated constrictions. These constrictions are generally found
between 2 consecutive segments, and thus the rings, in number
and position, answer to the real segments. It only seldom occurs
that each segment is again ringed. In many of the lower Oligochceta,
indeed in the Archiannelida and some of the Echiuridce, no distinct
mark of rings or segmentation is recognisable on the integument.

The body of the Chcetopoda is outwardly either homonomously
segmented, i.e. all consecutive segments of the trunk are alike, or
heteronomously segmented, when the segments in different regions
are differently developed, both as concerns their outer shape and their
provision with various setae, parapodia, cirri, gills, etc. We can in the
latter case distinguish different regions of the body (e.g. thoracic region,
branchial region, abdominal region, etc.) The integument of the Chceto-

190 COMPARATIVE ANATOMY CHAP,

poda is covered with a chitinous cuticle which is specially strongly
developed in the Polychceta Errantia. The cuticle is weaker and
much more delicate in most of the Oligochceta which live in mud,
and in the tubicolous Polychceta.

For the general form of body and outer organisation of the Myzo-
stomidce compare the systematic review.

The form of the body in the Prosopygia is extremely varied. The
most important points have already been referred to in the systematic
review. The body is as little segmented externally as internally.

In the Sipunculacea (Sipunculus, Priapulus, Halicryptus) a regular
outward ringing of the trunk occurs. The rings, at least in some
cases, correspond with the muscle bundles of the circular musculature
and with the lateral nerves which proceed from the ventral strand.
In Sipunculus there are, in addition to the circular furrows on the
trunk, still deeper and more distinct longitudinal furrows, so that the
whole skin seems divided into regular die-shaped areas. There are
similar longitudinal furrows on the "proboscis" of Priapulus. Papillae
are very wide spread on the bodies of the Sipunculacea, principally on
the proboscis.

A closer comparison of the proboscis of the Sipunculacea with the
similarly named organ of the Echiuridce (which formerly were united
with them in the class of the Gephyrea) shows great morphological
difference between the two organs. The proboscis of the Sipun-
culacea is the front portion of the body, which can be invaginated
into the hinder portion. The mouth lies at its anterior end. The
proboscis of the Echiuridce, is a prolongation of the head portion
(prostomium), which lies in front of and above the mouth and cannot
be invaginated. The mouth lies at its base. In the proboscis of the
Sipunculacea runs the fore-gut, while the fore-gut is in no way con-
nected with the prostomium of the Echiuridce.

The Sipunculacea possess a rough cuticle ; in Phoronis it is delicate,
and the skin therefore secretes a detached chitinous envelope, which
serves as a dwelling tube. The Bryozoa generally form a rough hard
cuticle (cell, ectocyst) whose aperture can be closed by a cover, and
which often calcifies. In a similar way the mantle of the Brachiopoda
secretes a bivalve shell which is generally calcareous, less frequently
horny.

This shell of the Brachiopoda (Fig. 125) cannot be compared
with the similarly bivalved shell of the Mussel (Lamellibranchiata,
Cochlidce). The two shell valves of the former are dorsal and ventral ;
each valve is symmetrical ; the median plane of the body divides
each valve into two lateral congruent halves. In the mussels, on
the contrary, we distinguish a right and a left shell. The median
plane passes between the two shell valves. Each valve is asym-
metrical. The gaping edge of the shell in the mussels is ventral, in
the Brachiopoda anterior ; the closed edge where the two valves

IV

VERMES—THE INTEGUMENT

191

are joined by a hinge is dorsal in the mussels, in the Brachiopoda it
is posterior.

We have already, in
the systematic review, said
what is most important con-
cerning the oral tentacles
and oral arms, which are
very characteristic of the
Prosopygia ; the various
positions of the anus were
also described. Eefer
to it also for the outer
form and organisation of
the Rotatoria and Chcetog-
naiha.

FIG. 125.— Rhynconella psittacea.
A, From above. B, From the left side.

II. The Integument.

The integument of the worms consists first of the outer cuticle, and
second of the subjacent body epithelium, which secretes the cuticle, and
which in most worms (just as in the Arthropoda)is called the hypodermis.

The cuticle is very variously developed. It is thin and delicate in
the soft-bodied forms, especially in the Nemertina, where it is perforated
by very fine pores to allow the cilia to pass through. Where it is
strongly developed, as in many Annulata, Prosopygia, Nematoda, and
Eotatoria, it gives protection and support to the body, and as skeleton
offers support and surfaces of attachment to the body musculature. It
consists of a substance allied to chitin, and occasionally calcifies (in
Bryozoa, Brachiopoda) into a very hard envelope or shell. It often
shows stratification, and seems to be composed of various crossing
systems of very fine adhering fibrillse. It may in general be conceived
of as a secretory product of the glandular hypodermis cells which
underlie it, or as a product of metabolism of the protoplasm of these
cells. To the same category as the cuticle belong various sorts of
tubes and envelopes which, detached from the integument, surround
the bodies of many Annelida (tube-worms) and possess a chitinous
substratum. We must further consider the setae of the Chcetopoda
as cuticular formations of certain hypodermal glands, which can be
seen, at least when they first appear, to be composed of fibrillae and
fibres in close contact and glued together.

The cellular hypodermis, which usually consists of one layer, can
best be examined in the Annulata, where it is composed of the two
following principal elements — (1) gland cells ; these are naked and
large, and yield the material for cuticular formations; over each
gland cell there is generally a pore in the cuticle; (2) thread-like
cells ; these are generally slender cells whose protoplasm is strongly
modified and falls into fibres. They often lose their nuclei, and
are arranged round the gland cells in such a way as to form for the

192

COMPARATIVE ANATOMY

CHAP.

latter a loose supporting tissue, in whose meshes the gland cells lie.
Gland cells of the hypodermis form the chief component part of the
subdermal setiparous sacs of the Chatopoda ; these are the setiparous
glands which produce the setae.

The setiparous glands may undergo important transformations.
In Polyodontes (Aphroditidce), for example, the setiparous glands of the
dorsal branch of the parapodia are changed into large spinning1 glands,
whose thread-like secretion yields the material for the structure of the
tubes they inhabit. In Aphrodite the dorsal setiparous glands produce
setae and hairs, which form the hairy felt covering the respiratory
chambers. The setiparous glands may again become simple dermal
glands. Anackcda, for example, no longer possesses setae, but, in place
of the 4 rows of setae of the related Enchytmus species, has 4 rows of
flask-shaped hypodermal glands projecting into the body cavity. The
mucous glands are peculiar dermal glands which are common, especially
in naked and soft worms (Nemertina, Hirudinea).

The hypodermis may
be very insignificant in
comparison with the
cuticle. Its elements
may fuse into a sub-
cuticular layer of proto-
plasm. In the Gordiidcr
among the Nemathelmia
we still find it clearly
developed into an epi-
thelial layer at the an-
terior and posterior ends
of the body, while in the
rest of the body it is
reduced to a subcuti-
cular finely granulated
layer containing scattered
nuclei. In this reduced
form we meet with the
hypodermis in all other
Nemathelmia, where it is

often hardly recognisable. It seems here to be almost entirely taken
up in the formation of the strong cuticle. The same is the case in the
Bryozoa.

In the Ilirudinea and in most Oli/jocha'ta, as sexual maturity begins,
the hypodermis undergoes a peculiar metamorphosis in a series of the
segments near the genital apertures (in the Hirudinea ahvays the
tenth, eleventh, and twelfth.) The gland cells here swell greatly and
come to lie in several superimposed layers, and there thus arises a girdle-
like thickening of the body which is outwardly visible (the clitellum).
In the gill-less Arnwluta (higher Oligochceta, Hirudinea) capillaries of

FIG. 120. — Transverse section through the middle part
of the body of a Nemertian, half diagrammatic. In, Lateral
longitudinal nerves ; <lii, medio dorsal nerve ; Inn, basal mem-
brane ; rm,' circular muscle layer; Im, longitudinal muscle
layer; rs, proboscis sheath ; /•, proboscis ; ?•<?, dorsal vessel ;
vl, lateral vessels ; h, testes ; p, parenchyma ; tnd, mid-^ut.

iv VERMES—DERMO-MUSGULAR TUBE 193

the blood- vascular system may penetrate as far as into the hypodermis,
and so enter the service of the general cutaneous respiration.

In Chcetopoda, a plexus of ganglionic cells lying immediately under
the hypodermis can be demonstrated, which is connected by nerve
fibrillae with the thread-like cells of the hypodermis.

In many worms (Aphanoneura, Archiannelida, Saccocirrus, Opheliacea,
various Polychceta of families widely separated, and further in the
Priapulidce, Phoronidce, and Sagitta) the central nervous system lies in
the hypodermis in such a way that no sharp distinction can be seen
between the usual hypodermis cells and the nerve elements. In many
Annulata whose central nervous system lies under the hypodermis,
the former nevertheless passes into the hypodermis at the most anterior
and posterior ends, the anterior part of the brain into the hypodermis
of the prostomium, and the most posterior end of the ventral chord
into the hypodermis of the tail segment.

The sensory organs of most worms, which will be described else-
where, belong to the hypodermis.

The hypodermis is often separated from the underlying tissues by
a thin supporting or basal membrane.

III. The Dermo-museular Tube.

Immediately under the outer integument in most worms the body
musculature lies in the form of a dermo-muscular tube, which repeats
.the shape of the body. It consists in general of two well-developed
layers, an outer layer of circular fibres and an inner layer of longi-
tudinal fibres.

These two layers are found in all Nemertina, except Cephalothrix,
where the circular muscle layer may be wanting. In the Schizonemertina,
and further in Polia and Falencinia, there is, in addition, an outer longi-
tudinal layer, which may even be more strongly developed than the
inner layer; in Carinina, Carinella, and Carinoma there is also an addi-
tional circular layer. All these layers form in the Nemertina a con-
tinuous tube, nowhere broken through in any way worth mentioning.

Among the Nemathel /ninths, all the Nematoda possess only the
longitudinal muscular layer (Fig. 127, Im). This is broken through
in 4 lines running in the longitudinal direction, and thus falls into
4 longitudinal portions. Two of the 4 longitudinal lines are median
(dorsal and ventral) and 2 lateral. In these lines of interruption the
subcuticular granulated dermal layer (hypodermis) is thickened, and
in it lie definite organs of which we shall speak later. In the Gordiidce
the ventral line only is clearly marked.

In the Acanthocephala, besides the longitudinal muscular layer,
an outer circular muscular layer is added.

In the Annulata the dermal musculature almost everywhere
appears in the typical form. The circular layer is very rarely
VOL. i o

194

COMPARATIVE ANATOMY

CHAP.

(Archiannelida) wanting. Other layers are, however, sometimes added
to the two typical layers, for instance, a layer of fibres which cross

each other diagonally in the
Hirudinea and Echiuridce. The
circular layer is everywhere
continuous ; the longitudinal
layer, however, is almost al-
ways broken through at dif-
ferent places in the Chcetopoda.
These breaches are often very
sl dissimilar in different genera
and families, so that they
cannot all be comprised in the
same description. The most
frequent are those in the dor-
sal and ventral lines, then in
the longitudinal lines formed
by the bundles of setae and the

FIG. 127.— Transverse section through a Nema- parapodia. The Various ar-

tode (Ascaris). dn, Medio-dorsal, vn, medio-ventral rangement of these latter
longitudinal nerve in the line which represents the -1-1 • , • .-,

middle line of the body; sl, lateral lines; c, cuticle; naturally caUSCS Variety in the

hy, hypodermis ; sg, lateral vessels ; ov, ovarial tubes ; arrangement of the lines Or

Im, longitudinal musculature ; Imk, cell elements of areag Qf interruption. As a
the longitudinal muscular fibres ; u, uterus ; md, , . . , A 7 r . , , ,

mid-gut. mle ln tne 4nnulata, the longi-

tudinal musculature is more
strongly developed than the circular musculature.

The dermal musculature of the Mijzostomidce is difficult to make
out. We may perhaps distinguish: (1) a system of fibres which
radiate from the centre to the circumference; (2) a system of fibres
concentrically arranged, and running parallel to the edge of the body.
The first system must represent the circular musculature of the Annu-
lata, the second their longitudinal musculature.

The various groups of muscles which serve in the Chcetopoda for
moving the bundles of setae, the parapodia and their appendages, must
be regarded as special local modifications of the dermal musculature.
In the Sternaspidce, as in the Sipunculidce, we find parts of the longi-
tudinal musculature differentiated into dorsal and ventral retractors of
the anterior introvertible portion of the body.

The general body musculature is developed in a very different
manner in the Prosopygia. The naked Sipunculacea possess a strong
and typically developed dermo-muscular tube, consisting of an outer
circular and an inner longitudinal layer (Fig. 128). Between
these two, in the Sipunculidce, a thin layer of diagonal fibres is inter-
posed. The longitudinal and circular muscles generally run in.
regular bundles or bands lying side by side, and these correspond
with the outwardly visible longitudinaj and circular ridges. The

iv VERMES— DERMO-MUSCULAR TUBE 195

longitudinal musculature supplies the retractors of the proboscis (Fig.
138, p. 208); the number of these retractors varies, and is of great
importance for classification ; they are attached to the dermo-muscular
tube on the one hand at the anterior end of the proboscis (anterior
portion of the head), and on the other at the anterior or middle por-
tion of the trunk, and run freely through the body cavity.

FIG. 128. — Portion of the musculature of the body wall of Sipunculus, diagrammatic (after
Andreae). Im, Longitudinal muscles, partly left out ; rm, circular muscles ; dm, diagonal muscles,
cut away in the middle line ; bs, ventral chord ; rn, nerve rings.

In the Phoronidce which live in chitinous tubes there is a typical
bilaminar dermo-muscular tube.

In the Bryozoa we can no longer speak of a dermo-muscular tube.
Its extreme reduction is to be referred to the development of a stiff
skeleton (shells, cells, ectocysts), which deprives a dermo-muscular tube
of its function. Only such portions of the general musculature remain
as are necessary for the withdrawal and protrusion of the soft-skinned
anterior end with its tentacles, or (Ehabdopleura) for the contraction of
the stalk of the body, which is movable within its tube (longitudinal
muscles of the stalk). In the Endoproda also there is, especially in
the stalk, a delicate longitudinal muscular layer immediately under
the skin. The elastic cuticle (which takes the place of the circular
musculature), or (Pterobranchia) a cartilaginous substance which forms
an axial strand in the stalk, serves to counteract these muscles.

The muscular apparatus which serves for the protrusion and withdrawal of the
anterior tentacle -bearing end of the body out of and into the cells, in those Bryozoa
provided with a temporary or permanent tentacle sheath, has been best observed in
the fresh-water forms (Fig. 139, p. 208). It consists esentially of 3 parts : (1) of
retractors which run in a longitudinal direction (like those of the Sipunculacca)
through the body cavity, and are attached on one hand to the anterior end of
the body near the tentacles, on the other to the body wall at the base of the cell ;

196 COMPARATIVE ANATOMY CHAP.

(2) of a system of fibres stretched between the invaginated wall of the proboscis
sheath and the neighbouring outer body wall (parieto-vaginal muscles) ; (3) of
circular muscles, generally developed only on the anterior body wall, though in
Paludicella they appear as subdermal muscular hoops in the whole body, and by
their contraction cause the protrusion out of the cells of the withdrawn anterior
end of the body with its tentacles. All these circular muscles are to be considered
as remains of the circular muscle layer ; the retractors and parieto-vaginal muscles as
remains of the longitudinal musculature of a dermo-muscular tube.

In the Brachiopoda a typical dermo-muscular tube is as little
developed as in the Bryozoa ; its absence is here also evidently to
be referred to the development of a shell. As remains of a dermo-
muscular tube, there are : (1) lying under the integument of the
mantle, weakly -developed fibres running transversely and longitu-
dinally ; (2) the arm muscles (protractors and retractors) ; and (3) the

270

FIG. 129.— Preparation of Waldheimia flavescens (after Owen), seen from the right side, to
demonstrate the musculature, the peduncle (p) and the calcareous framework (D) which serves
to support the arms. Dd, Dorsal, Dv, ventral shell valve ; mi, rat, m^, m±, muscles for opening and
closing the shell.

longitudinal muscles of the peduncle, which, in the almost universal
absence of a circular musculature, are counteracted by its elastic wall.
In the Brachiopoda a system of strong dorso- ventral muscles
passing through the body cavity (Fig. 129) serve for closing and
opening the two valves of the shell (adductors and divaricators). They
are attached to both the shell valves in the posterior region of the
body in the neighbourhood of the hinge (where this is present).
These muscles cannot be regarded as dislocated or modified portions of
the dermo-muscular tube.

In Dinopliilus there is found under the body epithelium a very
weakly developed dermo-muscular tube (circular and longitudinal
muscle layers).

The muscles of the Rotatoiia mostly run as isolated fibres in the
longitudinal direction, or circularly round the body. The longitudinal

iv PROBOSCIS OF THE NEMERTINA 197

muscles are always more strongly developed, attach themselves at both
ends to the integument, and serve chiefly for drawing in the anterior
end with the wheel organ, for shortening the tail or foot, and, in tubi-
colous forms, for withdrawing the body into its case.

The dermo-muscular tube of the Chcetognatha, which swim with
arrow-like speed, is very strongly developed, in keeping with its high
degree of activity. A circular muscle layer is, however, wanting. The
longitudinal musculature (Fig. 130, Im) is divided by 4 longitudinal
lines of interruption (2 lateral, 1 medio-
dorsal, and 1 medio-ventral) into 4 areas
(2 dorsal and 2 ventral).

"Worm-like contractions do not occur, in con-
sequence of the want of a circular muscle layer.
By alternate contractions of dorsal and ventral
musculature, and by the co-operation of the hori-
zontal fins, the elastic body is quickly propelled
forwards. It was chiefly the similarity of the
muscular arrangement of the Chcetognatha with
that of the Nematoda which caused many zool-
ogists to place the former with the Nemathelminths.

TV, +>io nrorrnc ^nnnrlellrt tlim*A i« a fTn'n FlG' 130. — Transverse section

In the genus bpadella t re is a tnm through the trunk of a Sagitta (after
layer of transverse muscle fibrillae lying o. Hertwig). ih, Body cavity ; mes,

in the body cavity, closely applied to the mesentery of the intestine ; md, mid-

ventral musculature. _ The arrangement ^±,3^' "rasoulat"re;
of the musculature in the head of the

Chcetognatha undergoes a marked complication. The most important
head muscles are those which serve for moving the seizing hooks.

IV. The Proboscis of the Nemertina and Aeanthoeephala.

These organs may be treated of together, although they arise quite
independently, and are not homologous.

In the proboseidal apparatus of the Nemertina (Fig. 131) we dis-
tinguish the following principal parts: (1) the proboscis sheath; (2)
the proboscis ; (3) the retractor muscle of the proboscis.

The proboscis sheath (rs) is a tube closed on all sides which lies
above the intestinal canal in the parenchyma of the body. Its
muscular walls consist principally of a circular and a longitudinal
muscle layer. The proboscis sheath stretches more or less far back,
often to near the hindermost end of the body.

The proboscis (r) is also a cylindrical tube. It lies invaginated
into the proboscis sheath. The space, closed in on all sides, between
proboscis sheath and proboscis is filled with fluid. The walls of the
proboscis sheath and of the proboscis join each other not far behind the
foremost end of the body. From this point a short tube (rhyneho-

198

COMPARA TIVE ANA TOMY

CHAP.

dseum, /•</) stretches to the most anterior end of the body, to emerge
generally separately from the oral aperture in front of and above it.
The whole proboscidal apparatus is entirely separate from the intestinal
canal. The rhynchoda-um thus leads through the proboscidal aperture

r

JJ

-rd

Fin. 131.— Diagrammatic representation of the proboscidal apparatus of the Nemertina. A,

Proboscis withdrawn. /;, Proboscis protruded. /•, Proboscis ; /v, proboscis slic-atli ; ?-s//, cavity of
the latter ; s/, stylet ; <nl. j>oison glanil ; mi, rotractor iiiusflc ; r<l, rliynchodieum.

into the central cavity of the prol)oscis which closes blindly behind.
The wall of the long proboscis is extremely muscular, the arrangement
of the fibres being occasionally very complicated. It is internally
lined with epithelium which is continued on to the rhynchodamm,

iv PROBOSCIS OF THE AGANTHOCEPHALA 199

and which, at the external aperture of the latter, passes ^into the outer
body epithelium.

At the blind posterior end of the proboscis a strand of muscle
fibres is inserted, the retractor (rm) of the proboscis, which runs
freely inside the proboscis sheath to its posterior end, breaking through
its walls to lose itself in the dorsal longitudinal musculature. The
proboscis can be evaginated from the proboscis sheath. This occurs
principally by a contraction of the muscular walls of the proboscis
sheath. In an evaginated condition the proboscis projects from the
anterior end of the body as a long tube, while the rhynchodseum
remains in its position. The inner wall then lies outside, the outer
wall inside ; the blind posterior end is at the extreme point of the
protruded proboscis, and its epithelium then represents still more clearly
than before a simple continuation of the outer body epithelium. The
anterior part of the extended retractor then lies in the central cavity
of the proboscis into which the fluid of the proboscis sheath penetrates.
By contraction of the retractor the proboscis is again invaginated.

At the blind end of the invaginated proboscis there is found in the Hoplonemer-
tina, a stylet (st] projecting into the proboscidal cavity, and at its side accessory stylets,
mostly small and in the course of formation. These stylets, when the proboscis is
fully protruded, come to lie at its foremost end and project freely. Further, the duct
of a sac-like gland (poison gland) (gd) to whose posterior end the retractor is
attached, is also often found at the end of the proboscis. In the unarmed Nemertina
numerous rod-shaped bodies or stinging capsules lie in the epithelium of the pro-
boscis. Only in Amphiporus, Malacobdella, and Geonemertes palaensis the probos-
cidal aperture opens into the oesophagus from above, so that the proboscis is protruded
through the mouth. Few observations have been made as to the function of the
Neinertian proboscis, which when irritated is generally so energetically projected that
it tears itself off at its edge of insertion in the rhynchod?eum. It probably serves as
a weapon both for defence and offence.

We have already in the Platodes become acquainted with organs similar to the
proboscidal apparatus of the Nemertina. The proboscis of the Proboscidea (p. 150)
among the Wiabdoccela (Turbellaria), representing an invagination of the most
anterior body wall which has become permanent, may possibly be homologous with
the Nemertian proboscis. There is a further extraordinary correspondence in general
arrangement between the proboscidal apparatus of the Nemertina and each of the
4 proboscidal apparati of the TetrarJiynchidce among the Cestoda.

The proboscidal apparatus of the Acanthocepliala (Fig. 172, p.
258) consists of the following principal parts: (1) the proboscis, (2)
the proboscis sheath, (3) the proboscis retractor, (4) the retractors of
the proboscis sheath, (5) the retinacula.

The proboscis in its evaginated condition represents the cylindri-
cal or conical anterior end of the body, narrowed and outwardly armed
with numerous hooks or stylets. The proboscis sheath is a muscular
pouch with a double wall, closed on all sides. It is attached at the
base of the proboscis to the body wall, and projects thence backwards
into the body cavity. It receives the proboscis when the latter is
invaginated. By its contraction the proboscis is protruded. The

200 COMPARATIVE ANATOMY CHAP.

proboscis retractor consists of longitudinal muscles, which, running
inside the proboscis sheath, are attached on one side to the base
(posterior , wall) of the proboscis sheath, and on the other to the apex
of the proboscis. By its contraction the proboscis is invaginated.
The proboscis retractor is continued at the posterior end of the pro-
boscis sheath into the two muscular retractors of the proboscis
sheath, which run through the body cavity, one being dorsal and the
other ventral, to attach themselves to the dermo- muscular tube.
These retractors hold the proboscis sheath in its place. At the base
of the proboscis sheath there arise further two occasionally muscular
strands, the retinaeula, which run laterally through the body cavity
into the body wall, and carry back within them lateral nerve strands
which come from the cerebral ganglion lying in the base of the pro-
boscis sheath.

Y. The Intestinal Canal.

The intestinal canal in the Worms is, in general, well developed
and provided with an anus. Only in the endoparasitic Acantliocephala
every trace of an intestinal canal has disappeared. In the males
of the Rotatoria also, and the males of certain species of Dinophilus,
the intestine is more or less completely degenerated. The intestine
of the dwarf male of Bonellia, which lives parasitically in the female,
is without mouth or anus. In the sexually mature Gordiidce also the
mouth is closed by an overgrowth of the cuticle. An anus is wanting
in various Nematoda, such as Mermithidce, Ichthyonema, and Filaria
medinensis. In the hermaphrodite generation of Allantonema mirabile
the intestine is quite reduced. A more or less far-reaching reduction
of the intestine can also be found in other Nematoda, e.g. Atractonema,
Sphcerularia. The intestine ends blindly in the Testicardines among
the Brachiopoda, and in Asplanchna among the Rotatoria. All these
defects and degenerations represent a derived condition, in contrast
with the well-developed intestine which is provided with an anus.
The degeneration can for the most part be ontogenetically established.

The walls of the intestine consist almost everywhere of 2 layers,
an outer muscular layer, which we might name the intestino-muscular
tube in contradistinction to the dermo-muscular tube, and an inner
epithelial layer turned towards the intestinal lumen.

We can, from an ontogenetic point of view, distinguish three
divisions in this intestine, the first two being already known to us in
the Ccelenterata and Platodes. (1) The fore-gut which proceeds from
the stomodaeum of the larva or embryo. Its epithelium is of ecto-
dermal origin, and it seems chiefly to supply the various adaptations
for seizing food, for reducing it into smaller pieces, and for passing it
on further (pharynx, jaws, teeth). (2) The mid-gut comes from the
mid -gut (mesenteron) of the larva; its epithelium is of endodermal
origin. It forms the principal digesting portion of the intestine. (3)

iv VERMES— INTESTINAL CANAL 201

The hind-gut, mostly short, often hardly distinguishable, comes from
the proctodaeum of the larva - or embryo; its epithelium is derived
from the ectoderm. It ejects the undigestible remains of food through
the anal aperture. We will treat of these three divisions separately,
taking the worms in order.

A. The Fore-gut.

The fore-gut is called the gullet or oesophagus in the Nemertina.
It is chiefly to be distinguished from the mid-gut which follows it by
the finer structure of its walls. Here and there glands have been
observed entering it ; these are salivary glands. There are no special
muscular swellings in the oesophagus of the Nemertina, a want which
is compensated for by the development of a special proboscidal
apparatus.

The oesophagus of the Nematoda is always distinctly separated
from the mid-gut, and lined by a strong cuticle, a continuation of the
cuticle of the outer integument. Its muscular wall is always
thickened into a generally round or egg-shaped pharynx, which
consists principally of muscle fibres placed radially to its axis. The
mouth often lies at the base of a variously shaped buccal cavity,
provided with hard teeth, lips, papillae, etc., and the pharynx itself may
fall into two consecutive parts by means of a transverse constriction.
Less frequently glandular tubes entering the buccal cavity have been
observed.

The fore-gut of the Annulata shows very various adaptations. In
the Hirudinea we can already distinguish two types. In the
Rliyncliobdellidce a pharyng'eal apparatus is developed, which agrees
even in details with that of the Tridada, Alloioccda, and many
Polydada among the Platodes. .The mouth leads into a cylindrical
pharyngeal pouch proceeding backwards, at whose posterior end a
muscular cylindrical tube, the pharynx, rises, and projects freely
forwards into the pharyngeal pouch. We have then — to use the
terminology adopted in connection with the Platodes — a tubular
pharynx plicatus, which is protruded (not evaginated) from the oral
aperture. In the Gnathobdellidce, on the contrary, the muscular wall
of the oesophagus itself is thickened and projects into the lumen,
generally in the form of three longitudinal folds or ridges. These
ridges are often very strongly developed (jaws), and at their anterior
projecting edges they are finely and sharply toothed (jaw teeth).
The use made of jaw and teeth by the medicinal leech is well known.
From the wall of the pharynx many strong muscle fibres radiate out
to the body wall. The ducts of well-developed salivary glands enter
into the pharynx.

^Eolosoma among the Oligochceta shows a very simple arrange-

202 COMPARATIVE ANATOMY CHAP.

ment. The fore-gut is limited to the head segment, and forms a single
pharyngeal cavity with a weakly developed muscular wall. In all
other Oligochceta the fore-gut stretches through several segments, and
is divided into two parts by a transverse constriction, an anterior
part, the buccal cavity, and a posterior part, the pharyngeal cavity or
pouch. The dorsal wall of the pharyngeal cavity is nearly always con-
siderably thickened, and projects into the pharyngeal cavity in the form
of a variously shaped muscular pharynx. The pharynx is attached by
muscles to the body wall, and can be protruded in order to take in
food, the pharyngeal pouch being at the same time necessarily everted.
Various glands — pharyngeal glands, salivary glands, and septal glands —
may enter the pharyngeal pouch. The part of the intestine which in
the Oligochseta follows the pharyngeal pouch and is generally termed
the oesophagus belongs, according to recent ontogenetic investigations,
to the (endodermal) mid-gut, and will thus be treated of later.

Among the Polychceta we again come upon very various forms of
the fore-gut. In most of the tubicolous forms it is a short soft-skinned
division which follows the mouth and is called the oesophagus. In
Terebellidce the oesophagus carries a ventral muscular appendage, the
oesophageal sac. Most of the Polychceta, however, are characterised by
the possession of a pharyngeal apparatus, which, especially in the
Errantia, reaches a high degree of complication and can stretch through

FIG. 132.— Diagrammatic representation of the pharyngeal apparatus of a carnivorous
Annelid, g, Brain ; ph, pharynx ; k, jaw ; m, mouth ; rt, retractors ; pt, protractors ; vt, anterior
soft-skinned portion of the pharyngeal apparatus ; p, its papillae. A, Pharyngeal apparatus in a
withdrawn condition. B, In a protruded condition. In B, ct indicates retractors ; rt, anterior soft-
skinned portion of the pharyngeal apparatus.

many segments. We may in a general way distinguish three modi-
fications of this pharyngeal apparatus.

1. The pharyngeal apparatus consists of two portions. The anterior
portion, into which the mouth leads, is a soft-skinned tube, which
is often provided internally with papillae. The wall of the posterior
portion is thick, on account of the strong development of its muscular
layers, and represents the actual pharynx (generally called proboscis).
Its anterior end carries papillae projecting inwards or a conical process,
or besides these (Errantia) two hard chitinous jaws. This pharynx can
be so pushed forward that its anterior end which is thus armed pro-
jects freely outwards, and is then everywhere surrounded by the

IV

VERMES— INTESTINAL CANAL

203

anterior soft-skinned portion, whose papillae then lie externally. The
anterior soft-skinned portion, therefore, is turned outwards like the
finger of a glove, and the pharynx proper is pushed after it. The
protrusion takes place either through pressure of the perienteric fluid
in consequence of a contraction of the dermo-muscular tube, or by the
contraction of special protractors of the pharynx. The withdrawal is
effected by means of special retractors. This is the most common
arrangement of the pharyngeal apparatus (Fig. 132).

2. The anterior soft-skinned portion is wanting, or is very weakly
developed. The mouth then leads almost directly into the muscular
pharynx, which is itself evaginated so that when protruded its inner
surface comes to lie outside (e.g. in the Capitellidce, Fig. 133).

3. The pharyngeal apparatus consists of two portions, one above
the other, the upper one of which (oesophagus) is less muscular and
forms the communication between mouth and intestine; while the
under one (jaw- carrier), which is blindly closed and enters the
oesophagus in front, is extremely muscular and carries hard jaws,
which are generally numerous, in special folds and sacs. These jaws
come to the exterior when the pharynx is protruded, and can be moved
one against the other (Euniddee, Fig. 134).

vn.

FIG. 133.— Representation of the
pharyngeal apparatus of an Anne-
lid, second diagram, g, Brain ; ph,
pharynx ; rt, retractors ; mh, buccal
cavity; vd, oesophagus. A, In a
withdrawn condition. B, In a pro-
truded condition.

FIG. 134.— Third diagram of the
pharyngeal apparatus of an Annelid
(Eunicide), withdrawn. g, Brain ;
vd, oesophagus; ph, pharynx; pht,
pharyngeal sac ; k, jaws ; mh, buccal
cavity ; m, mouth ; r, dorsal, &, vent-
ral side.

Among the so-called Archiannelida, Histriodrilus approaches nearest to the third
type on account of its ventral pharyngeal bulb armed with jaws. The pharynx
(oesophagus) of Polygordius is distinguished by the want of a muscular layer, and by
the great thickening of its epithelial wall, which is very thin only in the ventral
middle line. Under the pharynx there lies a thin-walled channel, closing blindly
and communicating by a longitudinal slit with the pharynx. In Protodrilus a
muscular accessory organ, bent- in the shape of a U, and lying under the beginning of

204 COMPARATIVE ANATOMY CHAP.

the intestine, enters the pouch-shaped non-muscular oesophagus quite to the front,
behind the oral aperture. Polygordius and Protodrilus thus approach the Terebellidce
in the arrangement of their fore-gut.

The genus Sternaspis, which possesses a strongly muscular pharynx, is distin-
guished by the fact that the seven anterior segments of the body can be invaginated.

Glands entering various parts of the pharyngeal apparatus have been observed in
many Polychceta.

In the Echiuridce the fore -gut is, relatively speaking, little dis-
tinguished from the mid-gut, and it even takes part in the formation
of the loops which the intestinal canal makes in the body cavity.
Three divisions are distinguished in it from before backward — -the
pharynx, the oesophagus, and the crop. It is the oesophagus in which
the muscular layer is specially strongly developed. It is not certain
whether any part of the fore-gut can be protruded. Food is in any
case seized by means of the prostomium (proboscis) and conducted to
the mouth along its ventral groove.

The Myzostomidce possess a pharyngeal apparatus, constructed on
the plan of that of the Rhynchobdellidce among the Hirudinea. At the
free end of the pharynx there are tentacle-like processes.

In the Prosopygia the fore-gut is in general very short and not
strikingly developed. This is evidently in keeping with the manner
of taking food of these mostly attached animals ; their feeding is
chiefly accomplished by the help of special outer appendages of the
head (tentacles, arms). (In a similar way we found that in the
Chcetopoda inhabiting tubes or living in mud and provided with a
crown of tentacles or a prostomium, the pharyngeal apparatus is far
less developed than in the other forms.) Among the Sipunculacea a
well-developed oesophageal bulb is present in the Priapulidce only ; its
inner cuticle forms numerous teeth projecting into the lumen. The
musculature is very powerful, and consists principally of circular and
radial muscles. In the Phoronidce, Bryozoa, and Bmcliiopoda the fore-
gut is represented by the connective piece between mouth and intestine
surrounded by a muscular layer, and is not very well marked ; this
is generally described as the oesophagus.

In the Rotatoria (Fig. 161, p. 245) the mouth leads first into a
narrow ciliated buccal cavity (oesophagus), in connection with which a
muscular layer is only rarely found. The buccal cavity is followed by
the distinctly separate cesophageal bulb or pharynx, which is provided
with a chitinous masticatory apparatus and strong musculature for
moving it. The masticatory apparatus consists of a middle part
(incus) and 2 lateral parts (mallei), each of which again consists of
2 pieces connected by a hinge (uncus and manubrium). The buccal
cavity and the pharynx together ought to be considered as the fore-
gut. In the pharynx of various Rotatoria salivary glands enter.

The fore-gut of the species DinopUlus (Fig. 162, p. 246), by its
ventral muscular appendage whose hollow • anterior end enters the

IV

VERMES— INTESTINAL CANAL

205

P

oesophagus directly behind the mouth, recalls in a striking manner the
arrangement found in the Archiannelida and the Terebellidce. The
ciliated fore -gut itself falls into 2 parts, a straight anterior oeso-
phagus, and a short posterior fore-stomach into which 2 lateral salivary
glands enter.

In the Cheetognatha the very simple fore-gut which lies in the head
is compressed transversely, and provided externally with a muscular
layer whose fibres run dorso-ventrally.

B. The Mid-gut.

The mid-gut comes from the endodermal mesen-
teron of the larva or embryo. The relations of the
various divisions of the gut to the embryonic gut are,
however, clearly established in only a few cases, so
that the demarcation of the mid-gut from the fore- and
hind-guts is to a great extent arbitrary. In the epi-
thelium of the mid-gut numerous gland cells are uni-
versally found.

The mid-gut of the Nemertina runs through the
whole body, generally in a straight line, from the
oesophagus to near the anus. It lies under the
proboscis. It usually has numerous unbranched lateral
sacs or diverticula, which occasionally lie one behind
the other pretty regularly, and so cause a segmentation
of the gut similar to that in the Tridada. In the
Hoplonemertina the gut sends off towards the front an
unpaired diverticulum under the oesophagus.

A musculature peculiar to the mid-gut has not yet
been discovered in the Nemertina.

The mid-gut of Malacobdella possesses no lateral diverticula ;
it has instead a serpentine course.

In the Nematoda also the mid-gut has a straight
course through the body. A special enteric muscula-
ture seems everywhere wanting. *~

With few exceptions the mid-gut of the Annulata testinai canal
also runs in a straight line through the body. It is and genital or-
almost everywhere surrounded outwardly by a muscular ^^s °J.aaNe]
layer, consisting of circular and longitudinal muscle tic. *», Mouth ;ai
fibres. auus ; d> gut ; d&,

The mid-gut of the Hirudinea shows mostly paired, &£?*}, g^ai"
lateral, segmentally arranged caeca or diverticula, the glands opening
last pair of which is often very long and extends back- outwards by ste-
wards on both sides of the hind-gut. Sometimes only
this last pair of diverticula is retained, or the diverticula may be
altogether wanting (e.g. Nephelis, Lumbricobdella).

206

COMPARATIVE ANATOMY

CHAP-

In the mid-gut of the Oligochceta we distinguish two principal
divisions, the oesophagus (which according to recent
observations also comes from the endodermal mesen-
teron) and the stomach -intestine. The oesophagus is
generally a tube with narrow lumen and weak walls.
In the terrestrial Oligochceta and the Naidomorpha the
oesophagus has somewhere in its course a muscular
swelling (seldom double), the muscular stomach
(gizzard). Besides this, 3 pairs of glandular lateral
sacs enter the posterior portion of the oesophagus of
the Lumbricidce and related terrestrial Oligochceta; these
are known as the lime-secreting glands, or Morren's
glands.

The large stomach-intestine often shows constric-
tions between the consecutive segments. In the
Lumbricidce the absorbing surface of the stomach in-
testine is increased by the fact that its dorsal wall is
folded longitudinally into the lumen of the intestine,
and so forms a tube lying in the dorsal middle line
of the gut, but open longitudinally towards the body
cavity; this is the typhlosolis (Fig. 165, p. 250, ty).

In the mid-gut of the Polychceta also we can dis-
tinguish an anterior division, mostly shorter and nar-
rower, from the posterior wider stomach-intestine. At
the boundary between the two special glands may
enter, like the Morren's glands of the Oligochceta. The
stomach intestine usually shows successive segmental
emails, a, Phar- swellings, caused by constrictions at the partitions be-
pT£ed6'iaterai di- tween tne segments. These swellings are often pro-
the longed laterally into caeca ; these are particularly long
in Aphrodite and the peculiar species Sphincter; in
Aphrodite the caeca themselves are again branched
(hepatic tubes).

In the Syllidce and Hesionidce two lateral sacs enter the anterior
end of the mid-gut (or posterior end of the fore-gut 1) ; these can be
filled with air like a swim-bladder. The mid-gut usually has a straight
course, but in the Chlorhcemidce, Amphictenidce, and above all in Stern-
aspis, it forms more or less striking loops.

The Capitellidce and a few Eunicidce are distinguished by the posses-
sion of an accessory intestine ventral to the main intestine, and
opening into it anteriorly and posteriorly in the Capitellidce, but in the
Eunicidce, as it appears, only anteriorly. In Capitella ventral ciliated
channels run forwards along the oesophagus and backwards along the
hind-gut from these openings. The constituents of food are never met
with in this accessory intestine, and it probably has a respiratory function.
In the Echiuridce (Fig. 137) a very striking mid-gut lies in
numerous loops and windings around the longitudinal axis of the body.

136.—]

of Himdo medi-

verticuia of
mid-gut; da, pos-

terior longer di-

verticuia.

IV

VERMES— INTESTINAL CANAL

207

It has an accessory intestine lying close to it which passes anteriorly
and posteriorly into longitudinal ciliated channels in the main intestine,

FIG. 137.— Enteric canal, vascular system, and nephridia of Echiuriis. sg, Lateral vessels of
the prostomium ; vd, dorsal vessel of the prostomium : vdi, dorsal vessel of the fore-gut (vda) ; Tc,
crop ; md, mid-gut, with the accessory intestine (nd) ; vv, ventral vessel ; n, nephridia ; t, their
funnel ; as, anal glands ; ed, hind-gut ; a, anus ; ml), muscles of the anterior hooked setae ; m, mouth.

and is justly considered as the connecting piece of these channels
arched over and thus separated from the intestine.

The mid-gut of the Myzost&midce (Fig. 175, p. 262), with its
branched diverticula which stretch as far as the lateral edge of the
body, recalls the mid-gut of the Annelid species Sphincter, whose body
is also broad and disc-like.

While among the Sipunculacea the mid-gut in the Priapulidce runs
in a straight line through the body, in the Sipunculidce (Fig. 138) it
forms a descending limb which runs through the trunk to the posterior

208

COMPARATIVE ANATOMY

CHAP.

hnr

Fig. 139. — Organisation of an individual of
Plumatella repens, with protruded crown of tern
tacles. t, Tentacles ; lo, lophophore or tentacle car-
rier ; ts, tentacle sheath (wall of anterior body); to,
constant invaginated portion of the tentacle sheatM
Iw, posterior portion of body wall, covered with I
cuticular envelope (cell 2) ; au, anus ; ed, hind-gut ; •
muscular (?) bands, which fasten the invaginated pal
of the tentacle sheath to the body wall ; rm, larger
retractor muscles for withdrawing the tentacle crow
into the cell ; /, funiculus ; st, statoblasts ; m, stomacB
with caecum ; Ih, body cavity ; g, brain.

FIG. 138.— Anatomy of Sipunculus with protruded proboscis, partly after Vogt and Jung ;
the body is cut open longitudinally, t, Tentacles ; g, brain ; vd, fore-gut ; rt, proboscis retractor ;
n, nephridia ; &mr, that portion of the ventral chord which lies in the proboscis and runs freely in
the body cavity ; bmrj, ventral chord of the trunk, marking the ventral middle line ; an, anus ; and,
anal glands ; ed, hind-gut ; nd, accessory intestine ; abd, descending limb ; aud, ascending limb
of the mid-gut.

iv VERMES— INTESTINAL CANAL 209

end of the body, and then bends round to run forwards as an ascend-
ing limb. The two limbs twist round each other and form a spiral.
A ciliated channel runs along the whole course of the mid-gut. Where
this ceases, at its posterior end, a blindly closed diverticulum is
attached to the gut, winding round its end ; this may be homologous
with the accessory intestine of the Chcetopoda.

The intestine of Phoronis, like that of the Sipunculacea, forms a
descending limb, and on the dorsal side of this an ascending limb.

The arrangement in the Bryozoa (Fig. 139) is closely connected with
that in the already described Prosopygia. A descending limb of the
intestine, not very sharply distinguished from the fore-gut, leads into
the expanded stomach, from which arises an ascending limb which
passes into the hind-gut. The stomach is occasionally prolonged into
a somewhat sharply demarcated csecum directed backwards.

In the mid-gut of the Brachiopoda we distinguish an anterior widened
part, the stomach, into which the oesophagus enters, and an adjoining
narrower part, the stomach-intestine. The stomach carries one or more
pairs of lateral diverticula which branch and divide into massive
glandular lobes, called the liver; these envelop the stomach on all
sides. The stomach-intestine forms either a simple or complicated coil
and then runs backwards (Crania), or bends on one side round to the
front (the other Ecardines), or ends blindly (Testicar dines},

In Dinophilus (Fig. 162, p. 246) and the Rotatoria (Fig. 123, p.
185; Fig. 161, p. 245) the mid-gut forms in the female a well-developed
pouch-shaped stomach, which is sharply divided from the fore- and hind-
guts. The epithelium of the stomach is either itself glandular, or
there are special large glandular appendages. An intestinal muscula-
ture is either wanting or very slightly developed.

The mid-gut of the Chcetognatha (Fig. 152, p. 227) takes a straight
course through the trunk cavity of the body, without lateral appendages.
A muscular layer is wanting.

C. The Hind-gut and the Anus.

The hind-gut comes from the proctodseum of the larva, and forms
in the worms a tube, often very short, but generally clearly separated
from the mid-gut ; this tube opens externally through the anus, and is
frequently called the rectum. The anus is either a separate aperture,
or else it unites with the apertures of other organs of the body. This
union is brought about through pit- or sac -like depressions of that
region of the body wall in which these apertures lie near each other ;
a cloaca is thus formed, which then opens externally by a new common
aperture, the cloacal aperture.

The juxtaposition of the external apertures of inner organs and the invagination
of the region common to these apertures are very frequent in the animal kingdom.
We will give only a few cases in illustration. In many Platodes the originally separate
male and female genital apertures come to lie at the base of a common genital
VOL. I P

210 COMPARATIVE ANATOMY CHAP.

cloaca. The male genital aperture may even be combined with the mouth (e.g. in
Stylostomum among the Polyclada and in Prorhynchus among the Rhabdocwla}.
In the Mesostomidce among the Rhabdoccela the longitudinal branches of the water-
vascular system, which generally emerge independently at the surface of the body,
enter the pharyngeal pouch, which itself represents an invagination of the outer
body wall (stomodceuni). In a few Nemertina the otherwise universally separated
proboscidal and oral apertures may be united (Amphiporus, Malacobdella, Geonemertes
palaensis).

The following are the most important combinations of the anal
aperture with the apertures of other organs. There are three kinds
of such combinations.

1. Union of the anal aperture with the apertures of the
nephridia or excretory organs. — In the Priapulidce the two anal
organs which serve primarily for excretion and secondarily as genital
ducts emerge quite near the anus. In Sipunculus slightly developed
anal glands have been observed entering the end of the hind-gut;
these are perhaps (?) homologous with the anal organs of the Priapu-
lidce. In the Echiuridce-, which in many ways occupy a position inter-
mediate between the Chcetopoda and the Sipunculacea, both the anal
glands (anal nephridia) also enter the hind-gut. In the Rotatoria also
anus and nephridia enter a common cloaca.

2. Union of the anal with genital apertures. — In the male
Nematoda the genital aperture emerges with the anus into a common
cloaca ; and the same is the case in the female Rotatoria.

3. Union of the anal with both nephridial and genital aper-
tures.— This occurs, as may be seen from 1 and 2, in the female
Rotatoria, where all three apertures open into a common cloaca.

The position of the anus and the cloaca. — The anus has a
terminal position in the Nemertina, Nematoda) and Annulata. In
the Nematoda it lies ventrally at a little distance from the posterior
end of the body ; in the Annulata generally dorsally, but always in
the last segment, except in Notopygos (Amphinomidce), where it lies
several segments from the posterior end on the back. In 'the
Chcetognatha it lies ventrally at the boundary between the trunk and
the tail regions ; in the Rotatoria dorsally at the boundary between
trunk and foot. In attached tubicolous Rotatoria it has moved some-
what forwards on the back, so that the hind-gut bends round for-
wards. The Prosopygia are distinguished by the fact that the anus
lies on the back, moved far forward ; and in the Sipunculidce it even
lies at the boundary between proboscis and trunk ; in Phoronis and the
Bryozoa quite anteriorly on the back, either (Ectoprocta) outside the
tentacle carrier or (Endoprocta) inside it. Where an anal aperture is
retained in the Brachiopoda it lies anteriorly, to the right, near the
mouth. There are, however, a few exceptions to the rule which
generally holds for the Prosopygia. In the Priapulidce and in the
Bracliiopod genus Crania the anus lies behind.

iv VERMES—BODY CAVITY 211

VI. The Body Cavity, the Musculature which passes transversely
through it, the Dissepiments and Mesenteries.

In the Nemertina we cannot yet speak of a body cavity (ccelome).
Apart from the connective tissue, which penetrates between the
muscles of the dermo-muscular tube, the space between the intestine
and that tube is everywhere filled by a gelatinous tissue which is
morphologically equivalent to the gelatinous tissue of the Codenterata
and the parenchyma of the Platodes. The various organs — genital
glands, blood-vessels, nephridia — are imbedded in this gelatinous tissue.
Dorso- ventral muscle fibres pass through it and form (1) a sort of
enteric musculature, the intestine passing between them and holding
them apart, and (2) passing between the enteric diverticula a sort of
muscular septa separating them ; these septa, just like the diverticula
themselves, lie one behind the other with more or less regularity,
and recall the septa which are developed between the gastro-canals of
the Platodes.

In the Nemathelmia there is a very spacious body cavity filled
with fluid, which, in the Nematoda, occupies the whole space between
the dermo-muscular tube and the intestine ; and in the Acantlwcepliala,
where an intestine is wanting, it is represented by the whole interior
of the body, which is surrounded by the dermo-muscular tube. In
this cavity lie the genital organs, bathed on all sides by the body fluid,
and further, in the Acanthocephala, the lemnisci. In the latter the
genital organs are attached by a muscular band or ligament to the
posterior end of the proboscis sheath, and also by lateral muscular
bands to the dermo-muscular tube. The body cavity of the Nemathel-
mia is not lined with a special epithelium (endothelium), but is
limited directly — externally by the body musculature, and internally by
the walls of the intestine.

The Gordiidce occupy an isolated position among the Nemat-
helminths in the morphological condition of the ccelome, as in many
other points of their organisation. In animals not quite sexually
mature we find between the intestine and the body wall a consider-
able mass of cells, which disappears for the most part at the time of
the development of the genital glands, and is probably used as material
for nourishing these glands. We then find, in place of the cell mass,
a spacious body cavity, which, however, in contradistinction to other
Nemathelminths, is lined on all sides by an epithelium, often of several
layers, lying on the inside of the dermo-muscular tube (Fig. 170, p.
256). This epithelium, in contrast to the epithelium of the intestinal
canal and to the outer body epithelium, we call the peritoneal endo-
thelium. This endothelium forms, in the median plane of the body,
a partition wall which separates into 2 lamellae ; these run dorso-
ventrally, having the intestine between them. At the sides of the
ventral median nerve they unite with the endothelium of the body

212 COMPARATIVE ANATOMY CHAP.

wall. By the special arrangement of this partition wall (mesentery)
the body cavity is divided into 3 principal chambers ; 2 lateral, and
1 unpaired ventral chambers, in which the intestine runs. We shall
return to some peculiarities of this mesentery later on, when treating
of the genital organs.

The presence of an endothelium and of a dorso-ventral median mesentery raises
the Gordiidce almost to the level of the higher worms, and supports the view
that they should be considered as such (probably Annulata) degenerated by
parasitism.

We cannot yet decide what should be considered as the body
cavity in the Hirudinea among the Annulata. The space between the
intestine and the body wall is filled by a connective tissue or paren-
chyma whose elements undergo the most varied transformations.
We find pigment cells, fat cells, fibres. Blood-vessels and blood-sinuses
arise by the flowing together of the contents of neighbouring cells.
The collective mass of the connective tissue elements is more strongly
developed in the Gnathobdellidce than in the Rhyncholdellidce. In the
latter division a connected system of blood-sinuses, whose walls are
not muscular but lined with an endothelium, and in which the
central nervous system lies, must be considered as a slightly developed
or else much reduced body cavity. . In the Gnathobdellidce the sinus,
which contains the ventral chord, the oesophageal commissures, and
the brain, seems to be such a reduced body cavity. This sinus,
however, is not lined with an endothelium. The fact that in
all Hirudinea the blood -vascular system is in open communication
with the sinuses makes it difficult to decide whether the canal
and sinus systems represent parts of the body cavity ; and if so,
to what extent this is the case. The sinuses are filled with a
fluid which, in the Rhynchobdellidce, contains colourless blood cor-
puscles.

Muscle fibres, branched at both ends and attached to the dorsal
and ventral body walls, run through the body parenchyma. They
form muscular dissepiments between the enteric diverticula, the
arrangement of which recalls that of the dissepiments in the Nemer-
tina and Turbellaria, and, in correspondence with the metameric
arrangement of these diverticula, are themselves metameric.

For the Chcetopoda we can establish a general morphological
scheme of the body cavity, which, however, undergoes considerable
modifications in a few divisions. Between the intestine and the
body wall there is always a body cavity filled with fluid, which is
entirely separated from the blood-vascular system. The body cavity
is divided in the following way. A dorsal mesentery connecting
the intestine with the dorsal middle line, and a ventral mesentery
connecting it with the ventral middle line of the body wall, divide
the body cavity into 2 lateral chambers, a right and a left. Muscular
partition walls, septa, or dissepiments, comparable to the dissepi-

iv VERMES—BODY CAVITY 213

merits of the Turbellaria, Nemertina, and Hirudinea, divide the body
cavity into as many consecutive chambers as there are segments.
These transverse partition walls always run between 2 consecutive
segments. They are bored through by those organs which run
longitudinally through several segments ; viz. the enteric canal, the
blood-vessels, and the nephridia. It is the septa which bring
about the segmental constrictions of the intestine. The consecutive
chambers of the body cavity are seldom completely separated from
one another, the septa being mostly perforated, so that a free com-
munication of the coelomic fluid in adjacent chambers of the body
cavity is possible.

All the walls of the body cavity and the organs lying in it are
lined with a peritoneal endothelium, which undergoes the most various
modifications. The endothelium of the body wall is distinguished as
the parietal layer ; that of the intestine as the visceral layer.

The chloragogen cells are peritoneal cells with definite excretory functions ; they
are especially strongly developed in the Oligochceta, and are attached to the dorsal
vessel and its branches, particularly to the network of blood-vessels which surround
the intestine. The brown granules which they contain are products of excretion
taken from the blood, and most probably reach the exterior through the nephridia
by the detachment and dissolution of the chloragogen cells. We find such cells also
in the Polychceta. The excretory organs probably also draw the excretory substances
direct from the blood, i.e. from the network of vessels which surround the renal
tube.

The division of the body cavity may vary greatly in details. The dissepiments
may become reduced or wholly disappear in large tracts of the body, so that con-
secutive chambers of the body cavity coalesce. Especially where a protrusible pro-
boscis is developed, the segments through which this organ stretches undergo a
reduction of their dissepiments. Dissepiments and mesenteries are generally typic-
ally developed in an early stage, even where in the adult condition great transforma-
tions take place.

The mesenteries may be reduced to isolated bands, fastening the intestine
to the body wall, and these may also be developed only in certain regions of the
body. On the other hand, the body cavity may undergo a still greater process of
division (especially in Polychceta). For example, a membrane running under the
intestine, attached on each side near the ventral chord to the body wall, often divides
the body cavity into an upper chamber containing the intestine, and a lower chamber
in which the ventral chord runs. Further, 2 lateral lamellae often run through
the body in a dorso-ventral direction, slanting upwards and outwards from both
sides of the ventral middle line ; these cut off 2 lateral cavities from the body
cavity, which may be described as renal chambers, as they generally contain the
greater part of the nephridia. These lamellae enclose transverse muscle fibres (Fig.
158, p. 237).

The dissepiments undergo a striking reduction, especially in the bodies of those
Choctopoda in which the enteric canal forms loops, and in which the segmentation of
the body is more or less obscured (Chlorhcemidce, Sternaspidce, Echiuridce). A spacious
body cavity is thus formed. In the Capitellidce the want of a separate blood-vascular
system is compensated for by a strikingly pronounced partitioning of the body
cavity.

Communication between the body cavity and the outer world

2U COMPARATIVE ANATOMY CHAP.

takes place in two ways. Firstly, by the nephridial canals, which
will be described later, and which are originally present in pairs in
each segment, and, secondly, by the dorsal pores. These have been
clearly observed in the Lumbricidcc and related land Oligockci'ta, and are
medio-dorsal perforations in the body wall, lying in the anterior end
of each segment. They are wanting in the head segment and in a
certain number of the subsequent anterior segments.

According to some writers, dorsal pores are present also in Jl/ichi/trcei<!tv ; and
cephalic pores leading into the head cavity are to he found in diiferent families of
the (>r«jo<:]t«.-tn. But the presence of these pores has again recently heen disputed.

In the Myzostomidw a body cavity filled with fluid is wanting.
The organs found between the intestine and the body wall, above
all the genital organs, are embedded in a body-parenchyma of con-
nective tissue. Still the question remains to be decided, whether the
space in which the sexual products lie does not answer to the body
cavity of other worms. Dorso-ventral muscle fibres run through the
parenchyma and form, in the same way as in the Ilirudinea, Xeme/iinn,
and Tnrlx'U<in<t) a kind of muscle septa between which secondary
muscle septa coming from the edge are intercalated, in the spaces
between the sexual organs and the enteric diverticula. In the middle
region of the body the septa leave a considerable space open in which
we find the intestine, with the uterus dorsally above it and the gan-
glionic mass of the ventral chord under it. '

In the manner of division of the body cavity the CJicetogncithct
are closely allied to the Chcetopoda. The body cavity is divided by
'2 dissepiments into 3 consecutive chambers. The first dissepiment
lies at the boundary between head and trunk, the second between
trunk and tail, and the 3 chambers separated by the septa are the
head, trunk, and caudal cavities. The enteric canal divides the head
and trunk cavities into 2 lateral portions ; in the trunk cavity the
intestine is often fastened to the body wall by a dorsal and a ventral
mesentery ; a dorso-ventral mesentery-like partition of the caudal
cavity is also found, although the intestine is there wanting. The
parietal and visceral layers of the peritoneal endothelium are continued
on to the mesenteries and the dissepiments, and form their chief com-
ponent part, as they are not provided with muscles.

The body cavity of the Prosopyyia shows very different arrange-
ments. That of the Sipimcidaccti is large and spacious, like that of
the tichinridte; dissepiments are wanting. The intestine, in the Sipnt-
niliiln; is fastened to the body wall by delicate mesenterial strands
which are wanting in the Pri<qndi<l<i'. In Priupiiln* the body cavity
is continued into the caudal appendage. A large expanse of the
peritoneal covering of the intestine is ciliated in the Sipunculidce, and,
as in the /:>A/'///'/W^<, a longitudinal muscular band runs along the intes-
tine. The co-lomic fluid generally contains amreboid lymph cells ;
and besides these, in the Sipiincnlida', the sexual products and other

iv VERMES— NERVOUS SYSTEM 215

peculiar bodies whose significance is not yet clear are found floating
in it. In Phoronis also the body cavity is well developed, and lined
throughout by a peritoneal endothelium. At the most anterior end
of the body there is a septum which separates the cavity of the
prostomium and the tentacles from the body cavity. The descending
limb of the intestine is fastened to the body wall by a ventral
mesentery, which, on the limb which ascends forwards to the anus,
becomes a dorsal mesentery. The descending limb is further connected
with the body wall by 2 lateral mesenteries. Among the Bryozoa the
body cavity in the Pterobranchia and Endoprocta, is extremely
reduced, but in the Ectoproda well developed. It is continued into
the tentacles, and is often lined with a ciliated endothelium, at least
this can be demonstrated in the fresh-water Bryozoa. The intestine is
fastened to the body wall on all sides by fibres which are considered to
be muscular. The gastric caecum is also suspended from the posterior
body wall by a strong non- muscular strand, the funieulus. In
Paludicella there is also a second funieulus. In the Brachiopoda the
cavity containing the viscera is lined by an endothelium generally
ciliated over a great part of its surface. The enteric canal is fastened
to the body wall by a more or less complete dorso-ventral mesentery,
which, when complete (e.g. in Crania), divides the body cavity into two
lateral halves. There are often lateral membranes or bands as well
fastening the intestine to the body wall — a gastro-parietal band in the
region of the stomach, and an ileo-parietal band in the region of the
hind-gut. These bands have been compared with septa, which implies
that the Bracliiopod body was originally composed of three segments.
This view is supported by other anatomical and ontogenetic facts.
The body cavity in the Brachiopoda is continued in the hollow spaces
of the mantle. An endothelial lining of the body cavities of Rotatoria
and DinophUus has not yet been proved. Distinct mesenteries and septa
are wanting. Fine fibres of connective tissue here and there connect
the organs lying in the body cavity with its walls ; no constant arrange-
ment, however, is found.

The pliylogenetic origin of the body cavity of the worms, and generally of the
higher Metazoa, is not at present certainly established ; it is also impossible to say
decidedly how far the hollow spaces in the body, called body cavities, are homologous
in the various divisions of the Vermes. In the Annulata and many other higher
animals it has long been known that special parts of the peritoneal epithelium are
the places of formation of the sexual products. We are therefore justified in asking
the question, whether the ovaries and testes of the lower acoelomous worms out of
whose germinal epithelium the sexual products are formed, do not correspond with
the chambers of the body cavity (ccelome) of the higher worms.

VII. The Nervous System.

The Nemertina (Figs. 140, 141). — The central nervous system
consists of the brain, which is placed in front of or over the oesophagus
and under the anterior portion of the proboscidal apparatus, and of

216

COMPARATIVE ANATOMY

CHAP.

Kn

Kit

two longitudinal trunks, proceeding from the brain and running some-
what ventrally in the lateral parts of the body ; these end near the

anus, or else coalesce immediately in
front of the anus above the hind-gut.
The brain consists of two large
lateral ganglia connected by a trans-
verse commissure and further united
by a commissure which runs over the
proboscis, so that the proboscis at its
anterior end is embraced by a nerve
ring. Each cerebral ganglion carries
a lobe, usually sharply demarcated
and placed behind, seldom to the
side or in front; the relations of
this with the lateral grooves (ciliated
organs, olfactory organs) will be de-
scribed later. These are known as
the olfactory lobes, and usually lie
above the first portion of the lateral
nerves. From the brain various
nerves proceed forwards to the apex
of the head, to the eyes, and so on.
Special nerves innervate the oeso-
phagus and the proboscis. In Paleo-
nemertina and Schizonemertina, between
the longitudinal and circular muscle
a continuous nerve
sheath (nerve plexus) entering the
longitudinal trunks. A nerve aris-

cerebrai lobes of the ciliated organs ; m, ing anteriorly in the dorsal cerebral
proximal portion of the proboscis nerves ; commissures often runs in the dorsal

kn, nerves of the head; sn, oesophageal -JUT v i J.T. J.T_-

nerves; sn (further down), lateral longi- middle line; beneath this nerve we

tudinal trunks ; qc, transverse commissures ; can Occasionally observe a SCCOnd

lateral branches of the longitudinal dorso-median nerve, the nerve of the

proboscis sheath. In the general
nerve sheath, which innervates the dermo- muscular tube, we can,
in the Schizonemertina, observe thicker strands, which form annu-
lar commissures between the three principal nerves running in
the longitudinal direction. In the Hoplonemertina the nerve sheath is
wanting ; the commissures run separately, and sometimes show a
markedly metameric arrangement.

In the whole structure of this nervous system a considerable agreement with that
of the Platodes (Polydada, Triclada, Trematoda) cannot be ignored. The brain and
the longitudinal trunks of the Nemertina answer to the brain and the ventral longi-
tudinal trunks of the Platodes. Whether the unpaired dorso-median longitudinal
nerve of the Nemertina answers to the two dorsal longitudinal nerves of the Platodes
cannot yet be decided. The longitudinal trunks (and also the brain) lie either

FIG. 140. — Nervous system of the an- ,

erior part of the body of a Nemertian layers, there
(Drepanophorus Lankesteri), after Hub-

trunks.

IV

VERMES— NERVOUS SYSTEM

217

imbedded in, or directly under the epithelium (Carinina, Carinella), or they are
enclosed in the musculature of the body wall (Cephalothrix, Cerebratulus}, or they
lie on the inner side of the musculature (AmpMporus).

FIG. 141.— Anterior end of the body of a Nemertian from the side, diagrammatic (partly after
Hubrecht). g, Brain ; re, proboscis commissure ; dn, medio-dorsal nerve ; Iso, cerebral lobe of the
lateral organ ; sn, lateral nerve ; mo, mouth ; r, proboscis ; d, intestine ; nr, nerve rings, commis-
sures between lateral nerves and the medio-dorsal nerve ; TO, proboscidal aperture.

In Malacobdella ganglionic swellings were formerly erroneously described as
occurring in the course of the longitudinal trunks.

Nemathelminths. — The nervous system of the Nematoda (Fig.
142) consists of a ring surrounding the oesophagus, the sides of which
often swell out into a ganglion. A dorsal and a ventral longitudinal
nerve arise out of the ring, and these run in the middle line to the
posterior end of the body. The two nerves are connected together by
transverse commissures which run under the cuticle outside the muscu-
lature. The transverse commissures of the right and left sides of the
body do not exactly correspond. Numerous nerves proceed forwards
from the oesophageal ring towards the apex of the head.

Attempts have been made to trace back the nervous system of the Nematoda to
that of the Platodes in the following way. By union of the dorsal longitudinal
trunks in the middle line the medio-dorsal nerve of the Nematoda arose, and by
the union of the ventral longitudinal trunks the medio-ventral, which in young
Ascaridce and in a Plectus is still distinctly paired. The dorsal half of the Nematodan
oesophageal ring answers to the brain of the Platodes with its often distinct lateral
swellings. The ventral half of the cesopliageal ring answers to the proximal portions
of the ventral longitudinal trunks, still separated in the region of the oesophagus,
but fused in the ventral middle line to form the medio-ventral nerve. The lateral
nerves which run a short distance at the anterior end of the Nematoda are considered
to be the remains of the lateral longitudinal nerves. The transverse commissures
of the Nematoda correspond with the transverse commissures which connect the
various longitudinal nerves of the Platodes. We prefer, however, to compare the
nervous system of the Nematoda with that of the Nemertina, because there we have
an unpaired dorso-median nerve, and the nerves corresponding with the outer lateral
longitudinal nerves of the Platodes are wanting. The lateral swellings of the brain
also are developed more clearly as lateral ganglia connected by a transverse com-
missure than is the case in the Platodes.

218

COMPARATIVE ANATOMY

CHAP.

In the Acanthocephala (Fig. 172, p. 258) there lies at the base of
the proboscis sheath a ganglion which sends several nerves forwards

to the proboscis sheath, the pro-
boscis, and the neck. Posteriorly
there proceed from the ganglion
two lateral longitudinal nerves,
which first enter the retinacula
at the posterior end of the pro-
boscis sheath and run in them to
the body wall, and then to the
posterior end of the body. A
dorso- median longitudinal nerve
has also been observed. In the
male, besides the anterior gan-
glion, another ganglion in the region
of the genital apparatus (lying
anteriorly on the base of the
withdrawn bursa) has been de-
scribed; this gives off nerves to
the genital apparatus, and is also
connected by two nerves with the
posterior ends of the lateral longi-
tudinal nerves of the body.

FIG. 142. — Diagram-
matic representation of
the nervous system of
the Nematoda, after
Biitschli. on, Upper, un,
under portion of the oeso-
phageal ring; sg, lateral
swellings of the same;
vln, medio- ventral ; din,
medio-dorsal longitudinal
nerve ; c, commissures be-
tween the two ; hsn, pos-
terior lateral nerves (bur-
sal nerves).

The nervous system of the Acantho-
cephala is up to the present time not
clearly understood. If there really is a
medio-dorsal longitudinal nerve it per-
haps corresponds with the medio-dorsal
nerve of the Nematoda, and then the
lateral longitudinal nerves perhaps repre-
sent the ventral longitudinal nerves,
which fuse in the middle line in most
Nematoda. What the relations are be-
tween the cesophageal ring of the Nema-
toda and the ganglion of the proboscis
sheath of the Acanthocephala, or whether
any such relations exist, are questions
which must be left on one side.

FIG. 143.— Cen-
tral nervous sys-
tem of Hirudo

The Gordiidce also deviate from
other Nematoda in the structure
of the nervous system. Kound
the rudimentary pharynx lies a
medicinaiis, after ganglionic mass (peripharyngeal
Hermann. ganglion) which is much thickened,

chiefly ventrally, and is produced into a ventral chord (Fig. 170, p.
256). This runs backward in the middle line and swells into a caudal
ganglion under the terminal portion of the genital ducts. In the male

IV

VERMES— NERVOUS SYSTEM

219

it divides at the most posterior end of the body into two strong
branches which run into the caudal bifurcation. The peripharyngeal
ganglion is connected on each side with the hypodermis, as is the
ventral strand at the extreme posterior end of the body, in the caudal
bifurcation, and indeed along its whole length by means of numerous
unpaired median nerves. Nerve fibres radiate forwards from the
peripharyngeal ganglion. A medio- dorsal nerve is wanting. It is
probable that the peripharyngeal ganglion and the ventral strand of
the Gordiidce are homologous with the cesophageal ring and ventral
chord of the Annulata.

.Annulata. — Here we can establish a general scheme for the central
nervous system. It consists (1) of a brain which lies over the anterior
portion of the enteric canal, and (2) of the segmented ventral chord,
running through the body in the ventral middle line. These two are

FIG. 144.— (Esophageal ring with diverging nerves of Hirudo, enlarged (after Hermann).
g, Brain ; sc, cesophageal commissure "bgi, first ganglion of the ventral chord (infra - cesophageal
ganglion) ; bg-2, second ventral ganglion.

connected by 2 commissures, the oesophageal commissures, which
embrace the fore-gut between them. The anterior end of the intestine
is thus surrounded by a nerve ring, the so-called cesophageal ring,
which enters the brain dorsally, and the first ganglion of the ventral
chord ventrally (Figs. 144, 146). The brain (supra-oesophageal gan-
glion), whose composition out of 2 lateral halves connected by trans-
verse commissures can generally be clearly seen, lies originally in the
first, the cephalic or oral segment. The anterior ganglion, or rather
double ganglion, of the ventral chord (infra -oesophageal ganglion)
probably originally belonged to the second segment of the body.
The remaining double ganglia of the ventral chord follow the infra-

220

COMPARATIVE ANATOMY

CHAP.

FIG. 145.— Terminal or (sucker) gan
glion of Hirudo, with preceding gan
glion, after Hermann.

oesophageal ganglion, one in each segment. The 2 ganglia of each
double ganglion are connected together by short transverse commis-
issures, and with the corresponding ganglia of the preceding and
succeeding double ganglia by longitudinal commissures.

Besides the cesophageal commis-
sures nerves proceed from the brain
to the integument and the sensory
organs of the head ; and from the
ganglia of the ventral chord nerves
supply the integument, sensory organs,
and musculature of the segments to
which they belong.

The musculature of the fore -gut
(pharynx, proboscis, etc.) is provided
with nerves arising either direct from
the brain or from the oesophageal
commissures. These nerves are called
cesophageal nerves (often also nervi
vagi), and the plexus they form in
the fore -gut is called the cesophageal
nervous system. In the most various divisions there is also a further
plexus of ganglionic cells and nerve fibres in the walls of the mid-gut
(visceral nervous system, sympathetic nervous system) whose con-
nection with the central nervous system requires further investigation.

Difference of opinion prevails as to the phylogenetic origin of the nervous
system of the Annulata. We prefer that view which not only derives the brain of
the Annulata from that of the Nemertina and Platodes, but also sees in the segmented
ventral chord with cesophageal commissures the further developed ladder nervous
system formed in these lower divisions by the longitudinal ventral trunks and their
transverse commissures. Ganglionic cells are found in the Platodes and Nemertina
everywhere in the larger longitudinal trunks, and are present in great numbers in the
Polyclada and Triclada at the points of divergence of the transverse commissures
and lateral branches. These latter are repeated in an extremely regular manner
and in a segmental arrangement as early as in certain Triclada (Gunda) and
Nemertina (Drepanophorus). If we imagine the ganglionic cells of the longitudinal
trunks crowded together at the points of divergence of the transverse commissures
and side branches, and their number perhaps increased still further, these points
of divergence must swell into ganglia and the longitudinal trunks become longi-
tudinal commissures between the successive ganglia. The longitudinal trunks with
their ganglia only need to move together into the ventral middle line, so as to
become the typical ventral chord of the Annulata. The mouth and oesophagus
prevent such a moving together in the anterior part of the body, and so the
first part of the longitudinal trunks becomes the cesophageal commissures of the
Annulata.

An approximation of the longitudinal trunks, which, it is true, never leads to direct
contact, may be observed as early as the Nemertina, not to speak of the Platodes.
Whereas as a rule among these forms the paired longitudinal trunks lie laterally,
sometimes even over the enteric diverticula (Amphiporus Moseleyi), they are found
in Drepanophorus placed somewhat ventrally and nearer each other. On the other

IV

VEEMES— NERVOUS SYSTEM

221

hand, the ventral chord of the Annelida can, by the separation of its symmetrical
halves, assume the appearance of a ladder nervous system (e.g. in the HermeUidce,
Fig. 147).

In very many Annulata (many Oligochccta and Polychceta] the brain stands in

7JH-*»

^
flfe>— *n

mlL

FIG. 146.— Anterior part of the body of Chaetogaster
diaphanus, from the side, after Vejdovsky. sb, Sensory
setse ; gz, ganglionic cells of the cup-shaped organ ; bga,
ladder-like ventral chord of the pharyngeal region ; disi,
dis2, first and second dissepiments ; pm, pharyngeal mus-
cles ; bgcti, ventral ganglion in the oesophageal segment ;
g, brain ; pg, pharyngeal ganglion ; md, stomach-intes-
tine ; rt, retractors (?) of the pharynx ; sc, oesophageal
commissure ; dg, dorsal vessel ; bg, ventral vessel ; ce,
resophagus ; pji, pharynx.

FIG. 147. — Nervous and nephridial
systems in the anterior portion of the
body of Sabellaria alveolata, from
the ventral side, alter E. Meyer. The
nervous system is made black, sr, (Eso-
phageal ring ; l>m, ventral chord ; t,
tentacles ; k, feeler gills ; m, mouth ;
b, bundles of setse ; hb, hooked setae ;
p, parapodia ; vn, anterior pair of ne-
phridia ; hn, posterior nephridia through
which the sexual products are dis-
charged.

direct connection with the hypodermis of the head segment. It often shows more
or less distinct lobes, always symmetrically arranged, which look like special ganglia.
The brain, which originally (and also ontogenetically) belongs to the head segment, can
in some cases stretch into the second segment ; it can even move back into the

222 COMPARATIVE ANATOMY CHAP.

second, third, or fourth segment (in certain earthworms). The infra-cesophageal
ganglion is also by no means always placed in the second segment of the body ;
here and there it lies in the head segment, or in the first and second segment, or
it moves farther back — farthest of all in Pleionc, where it lies in the fifth or seventh
segment. In the Hirudinea (Figs. 143 to 145) the infra-cesophageal ganglion fuses
with a few of the subsequent ganglia of the ventral chord, forming a considerable
infra-cesophageal ganglionic mass, in which, however, by the number and arrange-
ment of the diverging peripheral nerves, and by the arrangement of the ganglionic
cells and fibre commissures, its composition out of several ganglia can be more or
less distinctly recognised. The same is true of the large posterior (sucker) ganglion
(Fig. 145), which is composed of six fused ganglia.

As far as the ventral chord is concerned, the presence of a single ganglion in each
segment is the rule ; but there are exceptions to this rule in the most varied groups.
Sometimes one or two accessory ganglia are added to the principal ganglion,
e.g. in the Serpulidce two ganglia are typically present in each segment. On the
other hand the ganglionic masses may become very indistinct or be altogether
wanting (Archiannelida, and isolated forma in the most varied groups). In
Sternaspis the ventral chord shows ganglionic swellings only in the most posterior
part of the body. The fusion of the two lateral halves of the ventral chord is often
limited to the ganglia, whilst the longitudinal commissures run as completely
separated strands. These again may lie so close to each other that they appear
externally to be one single strand. It may even in various forms come to a fusing
of the two commissures, so that the ventral chord then consists of a single strand,
which, e.g. in the Archiannelida, does not even show the segmental ganglionic
swellings, and then recalls in a striking manner the ventral strand of the Gordiidce.
From the ganglia of the ventral chord there arise on each side nerves, generally 2 or 3,
which run upwards in the body wall and innervate the musculature, the integument,
and the segmental sensory organs, where such are present. It has been shown in some
cases that this innervation takes place by. means of a sub-epithelial nerve plexus de-
veloped all over the body. Lateral nerves may also diverge from the longitudinal com-
missures between 2 consecutive ganglia. The same is the case in the most anterior
longitudinal commissures — the cesophageal commissures.

There is often (especially in Hirudinea) in the ventral chord a delicate median
strand of longitudinal fibres which is called the intermediate nerve. The ventral
chord is nearly always covered by a single or double sheath of connective tissue
(neurilemma sheath), in which longitudinal muscle fibres run, principally in Oligo-
chceta and Hirudinea.

Certain tubes with a wide lumen and wall formed of neurilemma, which run
back in varying but always small numbers on the dorsal side of the ventral chord,
deserve special attention. These tubes, which are called neurochord strands or
giant nerve tubes, begin anteriorly in the cesophageal commissures or in the infra-
cesophageal ganglion, and stretch to the most posterior end of the ventral chord.
They were in all probability originally the neurilemma sheaths of giant nerve fibres,
arising from larger ganglionic cells which lay in the ganglia at different parts of the
ventral chord. Various observations make it probable that the nerve fibres in the
neurochord tubes in various Annulata degenerate, the neurochord tubes themselves
persisting as elastic organs of support, containing a watery fluid mixed with the
remains of the original nerve substance. Such degeneration is, however, denied by
some authorities.

The typical position of the ventral chord, and generally of the whole central
nervous system of the Annulata, is in the body cavity on the inner side of the
musculature. As the brain very often passes into the hypodermis of the head with-

OF.

VERMES— NERVOUS SYSTEM

223

out sharp limitation, so the posterior end of the ventral chord in most Oligochceta
and Polycheeta passes without sharp limitation into the hypodermis of the anal
segment. In many Chcetopoda of the most varied divisions, indeed in single genera of
different families, the nervous system even in adult animals shows an embryonic con-
dition, in so far as it lies outside the body musculature in the deeper part of the
epidermis. This position of the central nervous system is thus far from being
characteristic of the so-called Archiannelida, and we
can in general give it no special systematic signifi-
cance. We find besides in various divisions all
transition stages, from the position of the ventral
chord in the body cavity to its hypodermal position,
since it can lie in the body musculature between the
hypodermis and the body cavity. Moreover, in
Capitella, anteriorly, the ventral chord lies in the body
cavity, then passes in between the musculature, and
finally, posteriorly, comes to lie in the hypodermis.

The symmetrical halves of the ventral chord may
separate in some cases, and the ventral chord can
thus assume the form of a ladder nervous system
(Figs. 146 and 147) (Hermella, many Serpulidce,
Spinther, and in the pharyngeal region in the
Chcetogastridce). In the Aphanoneura (^olosoma)
the ventral chord is said to be entirely wanting, or
it lies in a quite embryonic condition imbedded in
the hypodermis.

The nervous system of the Ecliiuridce
(Fig. 148) deserves special attention. A
distinctly marked supra - cesophageal gan-
glion or brain is wanting. The cesophageal
ring is very much elongated, in keeping
with the great length of the prostomium.
Its two limbs (cesophageal commissures),
which run laterally in the prostomium (pro-
boscis), and which coalesce at its anterior
end, give off numerous branches towards
both the exterior and interior ; those
branches which run dorsally inwards may
represent transverse commissures between the

limbs of the oesophageal ring. Below and be- FIG. us.- Nervous system of
hind the mouth the limbs of the cesophageal Echiurus, diagrammatic, us, The
ring coalesce to form an unpaired ventral S^ ^"Sfflft

Strand, which in adult animals has no gan- the prostomium and enter the ex-

glion swellings, but is supplied with ganglionic treme anterior end of the ventral

chord (6m) behind the mouth ; nr,
Cells throughout its whole COUrse; this nerve ring8; vb, anterior hooked

strand runs in the ventral middle line on setae ; 7j*> *he two posterior circles
the body wall to the posterior end of the ^pMdL^ ^nuT^™ °f the
body, and gives off to the right and left

at short intervals lateral branches which correspond with the rings
of the outer integument. The corresponding right and left lateral

224

COM PA It A 77 T ~E A XA TOM } '

CHAP.

branches pass into each other dorsally, and thus form in the body wall

numerous consecutive nerve rings.

The nervous system of the MirJistoiuiJn' (Fig. 149) is now very

exactly known. It consists of an a-sophageal ring surrounding the base

of the pharynx, whose dorsal lateral
portions are somewhat thickened, and
so represent a weakly developed brain.
Close to the oesophageal ring lie gangli-
onic cells. Further forward in the
pharynx a second nerve ring (pharyn-
geal ring) is found, which is connected
with the oesophageal ring by several
longitudinal nerves, and gives off nerves
to the tentacles at the free anterior
edge of the pharynx. On the ventral
side, under the integument, lies a large,
elongated, ganglionic mass, from whose
anterior end the two commissures
(limbs) of the oesophageal ring arise.
The ganglionic mass represents a ven-
tral chord which consists of several
(probably G) fused pairs of ganglia, and
in which an intermediate nerve is to
be found. From this ganglionic mass
1 1 alternately stronger and weaker
nerves radiate on each side towards the
circumference of the disc-shaped body ;
these nerves branch in a complicated

FIG. i40.-central nervous system of mannei- and innervate especially the
•zostoma, after Nansen. </<\ Ce bral , . r , T , i i i

„„ . musculature of the parapodia, the hooks,
the the cirri at the edge of the body, and
^ the integument and its musculature.
The existence of a sympathetic nervous
system also seems probable.

Prosopygia.— The nervous system of the Sipumulacea is in
many respects closely related to that of the EcU'mridce. As there is
no long prostomium, the u'sophageal ring is also not elongated
anteriorly, but forms a simple ring at the anterior end of the enteric
canal. This ring in the PnnpnJiiln' is only a little thickened dorsally,
while in the Sipuncnlidce it enters a well -developed brain. Nerves
diverge from the brain to the anterior end of the body and to the
tentacles (where the latter are present). Xerves also diverge from the
oesophageal ring ; 2 of these in Sipunculus supply the intestine,
forming a ganglion on each side, while in Priapulu* 4 penetrate the
pharynx. The ventral strand, which is covered with ganglionic cells
along its whole course, runs in the ventral middle line of the body to
its extreme posterior end, where it swells somewhat. In its whole

Myzostoma

cominissurc ; pr, pharyn-x-al nervo
bym, ventral ganglionic mass, will
proximal part of the diverging nerve
cesnphageal coniniissnres ; tn, tei
nerves ; ;/.t, ganglionic cells.

iv VERMES— NERVOUS SYSTEM 225

course it gives off, just as in the Echiuridce, corresponding right and left
lateral nerves, which, running in the body cavity, unite dorsally to
form nerve rings (at least in Sipunculus). Each pair of nerves corre-
sponds with one of the circular muscle bundles and with one of the
more or less distinct rings of the outer integument. In the Priapulidce
it is said that there are weak but regularly recurring swellings of
the ventral chord which correspond with the circular muscle bands.
While in the Sipunculidce the central nervous system lies in the body
cavity, in the Priapulidce it is in close connection with the hypodermis.

Whereas in the Echiuridce the ventral chord is distinctly segmented in the early
stages of development, such a segmentation is wanting in the Sipunculacea even in
those stages. The ventral strand of the latter is nevertheless homologous with that
of the Echiuridce and with the ventral chord of the Annulata in general. The lierve
rings of the Sipunculacea evidently correspond with those of the Echiuridce. There
can be no doubt as to the homology of the brain and cesophageal ring. We have
also seen that the ventral chord, even in true Chcetopoda, may present no ganglionic
swellings, and may fuse into a single median ventral strand. It is a question whether
the nerve rings of the Echiuridce and Sipunculacea are repeated segmentally. In
the Echiuridce several body rings with smaller papillae alternate with one with
larger papillae. The setse only occur on the rings which carry the larger
papillae. We also find that the nerve rings in the body rings with the larger
dermal papillee, are stronger than the others. It seems probable that several rings
go to one segment, and that the number of these rings with larger papillae and
stronger nerve rings corresponds with the true number of the segments. In the
Chcetopoda also (apart from the Hirudinea] one segment is sometimes externally
divided into 2 or more rings, and more than one ganglion of the ventral chord
may lie in one segment. In the Hermellidce and Serpulidce we find typically
2 pairs of ganglia in each segment, and further 2 transverse commissures and 2
lateral branches on each side, which ascend in the body wall towards the dorsal
middle line.

The nervous system in other Prosopygici is remarkably weakly
developed. The ventral chord is reduced to the infra -oesophageal
ganglion, and even this may be wanting. The small development of
the nervous system is probably to be traced to the stationary manner
of life, which results in a reduction of the specific sensory organs,
and — by the development of tube-dwellings, shells, or cases — of the
general body musculature (dermo-muscular tube).

Plioi*onis possesses a nerve ring surrounding the mouth at the base
of the tentacles. The anus lies outside the nerve ring. From the
dorsal part of the ring a nerve arises which runs backwards asym-
metrically on the left side, through about two-thirds of the length of
the body, and with a tube (notochord ?) passing through it. The
morphological significance of this nerve is unknown. The whole
nervous system lies in the hypodermis outside the basal membrane.

The nervous system of the Brachiopoda (Fig. 150) is weakly

developed. It consists of a delicate O3sophageal ring, whose upper

portion is only very slightly swollen into a supra-cesophageal ganglion.

The infra-cesophageal ganglion is indicated by a thickening of the

VOL. I Q

226

COMPARATIVE ANATOMY

CHAP.

ventral portion of the oesophageal ring, and in some cases two lateral
swellings can be made out.

From the supra-oasophageal ganglion two nerves run into the arms;

from the infra-

-oes

tar

&**

gan-
glion nerves run
also to the arms,
to the mantle, and
to the opening and
closing muscles of
the shell. The
arm nerves are
connected with a
plexus of gangli-
onic cells and
nerve fibres, which
spread out in the
supporting sub-
stance of the arm
walls close under
the epithelium.
Each of the man-
tle nerves divides
into a dorsal
branch for the
dorsal fold of the

°" oV

FIG. 150.— Preparation of Terebratula vitrea for demonstration
of the nervous system, the ovaries, and the nephridia, after van
Bemmelen. Anterior body wall after removal of the arm apparatus,
the enteric canal, and the closing muscles, spread out and seen from . -i -j

within. Above lies a part of the dorsal body wall ; below, a part of the T UG>
ventral body wall, g, Brain ; sc, oesophageal commissures ; usg, infra-
O3sophageal ganglion ; san, arm nerve proceeding from the supra-oeso-
phageal ganglion ; ian, ditto from infra-cesophageal ganglion ; ces, oeso- rp-i
phagus ; dmn, dorsal mantle nerves ; hn, nerves proceeding backwards
from infra-oesophageal ganglion ; n, nephridia (oviducts) ; nt, nephridial
funnel ; gf, genital folds ; ov, ovaries ; dm, dorsal mesentery ; vm, ven-
tral mesentery.

ventral branch for
the ventral fold,
are again
richly branched,
and their branches
anastomose form-
The nervous system of the Bracliiopoda, lies under the

ing a plexus,
integument.

In the Bryozoa, only the supra-oesophageal ganglion of the oesoph-
ageal ring is retained ; it lies as a generally inconsiderable mass (occa-
sionally with thickened lateral parts) dorsally over the fore-gut, between
mouth and anus, under the integument. From this ganglion nerves
run chiefly into the tentacles, and further to the two ciliated pits,
when such are present.

Rotatoria. — Over the oesophagus there lies a ganglion (supra-oesoph-
ageal ganglion) which sends off nerves to the wheel organ, the cutaneous
organs of touch, and the muscles. It lies under the integument.

In Dinophilus also, in front of and above the mouth (in the pro-
stomium), is found a mass of fibres surrounded by ganglionic cells, which
represents the supra-oesophageal ganglion. In Dinophilus gigas two longi-

IV

VERMES— NERVOUS SYSTEM

227

— m.

ovd

tudinal nerves rise out of it, which pass by the mouth, and run on both
sides of the body
immediately be-
neath the integu-
ment to its pos-
terior end. These
longitudinal nerves
must represent the
separated lateral
halves of the ven-
tral chord of the '•• //j^. \{
Annulata. Trans- L>- v ' J^.,jLkl

verse commissures FIG. 151.— Head of Sagitta bipunct-

.bewanting. ata> seen from above' wiih closed

0 seizing hooks, after O. Hertwig. g,

Cil36tOg>na,tn3, Brain ; gh, seizing hooks ; sc, commissure

3. 1 5 1 andl 52). between brain and ventral ganglion ; an,

rp-i optic nerve ; au, eye ; ro, anterior portion

8 of the olfactory organ ; rn, olfactory nerve.

.system is here

well developed. The central nervous system
and the peripheral nerves lie, with the exception
of a single portion, external to the musculature
in the body epithelium. The brain or supra-
oesophageal ganglion lies dorsally in the head
segment, while the infra-cesophageal ganglion lies
ventrally in the trunk segment and surpasses the
cephalic ganglion in size. The cephalic and ventral
ganglia are connected by 2 long commissures.
Besides these 2 commissures, the supra-oesophageal
ganglion gives off 2 strong nerves which penetrate
the mesoderm forwards and downwards, and which
we may call motor nerves, 2 lateral nerves which
.supply the integument of the head, 2 outer pos-
terior nerves which, after a short course, reach
the 2 eyes behind the supra-oesophageal ganglion
(nervi optici), and 2 inner posterior nerves which
supply the unpaired sensory organ lying behind
the eyes which is supposed to be the olfactory
organ (nervi olfactorii). A great number of nerves
radiate from the ventral ganglion, among which
the continuations of the 2 oesophageal commissures, the ventral side, after
after running through the ventral ganglion, are °-HeJtwis- m> Mouth ;*>

,. , , . ' intestine ; sc, oesophageal

tound as. 2 strong longitudinal strands, which, commissure; ig, ventral

after giving off numerous lateral nerves, them- ganglion; fl, fins; o»,

selves end in fine nerve fibres. All nerves diverg- °^^^\Q^^°^

ing from the ventral ganglion and the posterior tore ; a, anus ; ho, testes ;

longitudinal nerves pass finally into a plexus of sh' tail cavity ; sl> sPerm

!• • •,-, , „, -i • -i • i T T duct: sb. snfirm vpsiplp •

gangliomc cells and nerve fibres, which is developed

— d

*

228 COMPARATirE AX ATOM V CHAP.

in the epithelium all over the body. The motor nerves of the supra-
cesophageal ganglion form a ganglion each on the sides of the fore-gut
in the cephalic segment (lateral cephalic ganglia), with small accessory
ganglia. The musculature of the head and the fore-gut are supplied with
nerves by these ganglia. This mesodermal part of the nervous system
of ^iy/tt<i. recalls the tesophageal nervous system of other worms. We
do not yet know how the trunk and caudal musculature are supplied
with nerves.

VIII. Sensory Organs.

All the different kinds of sensory organs are found in the worms —
organs of touch, sight, hearing, smell, and taste. Besides these, in a
few divisions we meet with sensory organs which cannot at present be
classed in any of the above categories — the lateral organs of the
(.'ltiffnjii,</ii. the lateral eyes of Polynphthcilmux, &i\(\. the segmental organs
of the ILii'ii<liix-«. It must not be forgotten that the functions
of the sensory organs of the worms are as little experimentally
established as are those of most other invertebrate animals ; and that
it is almost entirely their position and structure which lead us to
consider them auditory, olfactory, etc. The function of the organs of
touch indeed is more surely established. That the worms in some
way or other see with the sensory organs which have been indicated
as eyes may also be considered certain, though we do not know what
and how they see.

The sensory organs are most numerous and best developed in
animals leading a free aquatic life (Polycho.'ta, Ernintui, ChcctognatJia), and
among these again the good swimmers take the first place. The
worms which are attached at the bottom of the water and those
which lurk in holes are not quite so fully provided. In worms living
in mud and sand or in earth the sensory organs are much reduced,
and this is the case in the highest degree in parasitic and attached
animals. In the latter, however, the strongly developed organs of
touch form an exception. Where the sensory organs are reduced in
adult, stationary, or parasitic animals we often meet with them well
developed in their young stages, when they as larvae move about freely.
In order of frequency we have the organs of touch, which are universally
distributed, then the eyes, then the olfactory organs and organs of
taste. Organs of hearing have been certainly observed only in a few
<"i<es (slrenit'iilidt.?, Scrpnhiccn, Terclelloidea^.

A. Organs of Touch.

K very where, except in the worms provided with a thick shell, the
entire integument is the seat of a highly developed sense of feeling
or touch. This sense is served in a special manner by epithelial sen-
sory cells, which curry at their free end sensory hairs or seta?, and at
their basal ends are continued as nerve fibres, which are themselves

iv VERMES— SENSORY ORGANS 229

generally processes of peripheral ganglion cells. A plexus of nerve
fibres and ganglion cells lying immediately under the body epithelium
can now be proved with certainty to exist in various worms, especially
in the Chcetopoda and Chcetognatha. In parasites with a thick outer
cuticle papillae, rod- or seta-like processes of this cuticle, which are
principally developed at the anterior end and near the genital apertures,
play the part of organs of touch. As a rule the tactile cells are most
numerously present in that part of the integument of the worm where
the body has most points of contact with its immediate surroundings.
Such points are, above all, the anterior end of the body, the neighbour-
hood of the mouth, and the various appendages. There are some
such appendages which, on account of their position and their speci-
ally rich provision of tactile cells, may be regarded as specific
organs of touch. We may mention in illustration the feelers on
the heads of the Chcetopoda, especially of the Polychceta, the cirri of
the parapodia, the prostomium of the Uchiuridce, the tentacles of
the Prosopygia (cirri on the oral arms of the Brackiopoda), and the
wheel organ of the Rotatoria. It is obvious that such organs of
touch in worms inhabiting tubes or shells can only attain development
at the anterior end of the body, which carries on the relations with
the outer world ; and it is equally intelligible that in such worms
these same organs should carry on other relations to the outer world
as well (prehension of food, respiration). Hence the strong develop-
ment of the tentacular apparatus in the tubicolous and shelled worms
(tubicolous Annelids, most Prosopygia, and the tubicolous Rotatoiia).

The sense of touch is very strongly developed in the Hirudinea.
The tactile cells, each of which is provided with a fine hair, form
groups (tactile cones), which are arranged in 18 longitudinal rows.
They are developed on the warts or papillae, when such occur (Fig. 156).

B. Eyes.

Their oeeurrenee, number, and arrangement. — In many genera
and species of Nemertina the eyes are wanting. In others small
eyes occur in varying numbers (2, 4, or many) at the anterior end
of the body. In the Nematoda the absence of eyes is the rule, the
presence of 2 simple eye-spots at the anterior end of the body (in
some of the free -living forms) the exception. The Acanthocephala
are without eyes. In the Annulata the presence of eyes is the
rule, their absence the exception. The Hirudinea possess 1 to 5 pairs of
eyes in the anterior rings of the body. Among the Oligoclueta, which
mostly live in mud or earth, only the Naidomorplia possess eyes — one
pair in the head segment. In the Arcliiannelida which have a similar mode
of life, or are, like Histriodrilus, parasitic, the eyes are either wanting
or reduced in the adult animal. Most of the Polychceta are provided
with eyes, which lie, with very few exceptions, in the head segment.
Most of the Errantia have 1 or 2 pairs of well-developed eyes, which

230

COMPARATIVE ANATOMY

CHAP.

in the Aldopidce reach a remarkable size ; many Sedentaria are eye-
less ; where, however, eyes occur they lie as small so-called eye-spots,
generally in great numbers, in the prostomium, at the part where
the brain is connected with the hypodermis. In Fabrida eye-spots
occur at the posterior end of the body; in some species of Sabella
on the tentacle gills. For the so-called lateral eyes of Polyoplithalmus
see below. The Echiuridce-, which live in mud or lurk in holes,
and the parasitic Myzostomidce, are blind. The absence of eyes
is characteristic of the whole class of the Prosopygia. The Piotatwia
possess an unpaired or a paired eye- spot lying on the brain, but this is
wanting or rudimentary in the adult condition in stationary forms.
DinopJiilus has 2 eye- spots in the prostomium. In the Chcetognatha
2 eyes lie on the dorsal side of the head, behind the brain, in the body
epithelium.

Structure of the eyes. — A comparative morphology of the eyes
of worms is at the present time a desideratum. At any rate it now
appears certain that the eyes in the various divisions need not be
homologous. The eyes which appear in pairs are perhaps homologous,
such as those developed in larvae of the Trochophom type in close
connection with the cerebral rudiment; these very often disappear
through metamorphosis, or degenerate. As a detailed account of
these always more or less complicated organs is here impossible, we
select a few for brief description, viz. the eyes in Capitella, in Aldope
(the most highly developed eye among worms), in a Chcetognathan, and
in Hirudo.

1. The eye of Capitella. — The numerous eyes (ocelli) of Capitella lie
in the prostomium at the part where the brain fuses with the hypo-
dermis. The following are the
elements of a single eye. We
find slipper -shaped refracting
cells, whose outer portion is
homogeneous and transparent,
while the inner part contains
pigment. Each of these cells
is continued as a nerve fibre
towards the brain, this nerve
fibre entering a ganglion cell of
the central optic lobe. The
refracting cells of the eyes are
connected together by thread
cells of the hypodermis. The
cuticle is arched over the
eye.

2. The eye of Aldope (Fig.
153). — The two eyes of Aldope
stand out spherically, one on
They are covered by a thin layer of hypo-

fl

FIG. 153.— Section through the eye of an Alciope
(Callizona Grubei). hy, Hypodermis ; c, cornea ; I,
lens ; fl, eye fluid ; p, pigment of the retina ; r, retinal
cells ; st, rods ; go, ganglion opticum of the brain
(after Carriere).

each side of the head.

iv VERMES— SENSORY ORGANS 231

dermis with its cuticle, which form the outer cornea over the centre
of the protruding eye. The eye itself is a vesicle whose posterior
thicker wall forms the retina, while the anterior thinner wall is the
inner cornea. The elements of the retina are long cells standing
closely pressed together, in which three parts can be distinguished : (1)
towards the brain, the cell body with a nucleus ; (2) the rod, which is
directed towards the hollow of the bulb ; and (3), between these two,
a thin layer of pigment. Under the cornea lies the spherical lens.
The rest of the eye is filled with fluid. The retinal cells are continued
into nerve fibres, which soon enter the ganglion cells of the optic lobe
(ganglion opticum) ; the latter is connected with the brain by a mass
of nerve fibres.

3. The Chcetognathan eye (Fig. 154) is spherical,
of the sphere lie 3 bi-convex lenses

imbedded in pigment ; to the out-
side of each of these 3 lenses a
third part of the whole retina is Jsr^**

•-"-- ' 1 . 4Bfc-"*"*.X> L5^^** ""~**>Sfcl ' ..

applied in such a way that the three f;

parts together form the wall of the [

sphere. The retina consists of cells ; |

the portion of each of these cells *

which is in contact with the lens f

is rod -like, and the part which f|

is turned outwards is the cell-body Pia i54.-Section through the eye of sag-

with its nucleus. At the circum- itta hexaptera, after O. Hertwig. ep, Body

ference of the sphere, each retinal ^eiium ; z, lens ; p, pigment ; s« rods;™,

retinal cells.

cell is continued as a nerve-fibre.

All the nerve fibres unite in the nervus opticus (Fig. 151, p. 227).
The Chcetognathan eye may be considered to have come from 3 simple
fused ocelli.

A comparison of the three eyes just described shows how greatly
the eyes of worms may vary in structure.

4. The eyes of Hirudo (Figs. 155, 156) lie in the anterior rings
of the body and vary in number. They are cylindrical and stand at
right angles to the somewhat modified hypodermis with which they
are in contact. The optic nerve enters at the base, its fibres passing
into long sensory cells which lie in the axis of the eye. Around
the axis are arranged large clear cells, each containing a nucleus and
a refractive substance. The whole organ is imbedded in strongly
pigmented connective tissue. Our present knowledge of the structure
of these organs hardly justifies us in calling them eyes ; morphologically
they are transformed tactile organs.

C. Olfactory Organs (Ciliated Organs).

In many worms of the Nemertian and Chcetopodan divisions there
are found, at the anterior end of the body, 2 lateral strongly ciliated

232

COMPARATIVE ANATOMY

CHAP.

parts of the hypodermis, the so-called ciliated organs, ciliated clefts,
ciliated pits, ciliated prominences, which are regarded as olfactory

FIG. 156. — Section through a tactile
sensory organ of Macrobdella, after
Whitman, c, Cuticle ; ep, hypodermis ;
p, large clear cells ; gz, ganglion cells ;
n, nerve. The sensory cells are here
clearly seen to be long hypodermis cells,

»'• FIG. 155.-Section through the eye the tactile hairs which they carry are

of a land leech, c, Cuticle ; ep, hypo- not depicted.

dermis ; p, large clear cells ; g, ganglion

cells ; n,\ nerve ; dz, cutaneous gland

cells ; pi, pigment (after Whitman).

organs. The body epithelium at these points consists of ciliated sensory
cells, whose bases are prolonged as nerve fibres. These nerve fibres are
connected under the sensory epithelium with a plexus of ganglion cells,
which is itself again connected with the brain. In the Chcetopoda
the ciliated parts just mentioned are frequently depressed in the form
of pits or sacs, and are often protrusible. A special olfactory lobe
(ganglion olfactorium) may be developed on the brain in close proximity
to the ciliated organ, this lobe bearing the same relation to the
olfactory organ that the lobus opticus or ganglion opticum bears to
the eye. In the Nemertina these olfactory lobes are very strongly de-
veloped, and often sharply separated from the brain (cf. p. 216 and Figs.
140, 141). The ciliated organs here are pits which open outwardly by
means of longitudinal clefts at the sides of the head, and which in the
Schizonemertina are continued inwardly as ciliated canals, penetrating
to the interior of the cerebral olfactory lobes. It is probable that all
these ciliated organs are homologous formations, and correspond with
the ciliated grooves at the extremity of the head in the Turbellaria.

In the Chcetopoda such organs have been found more or less developed in the
Capitellidce, Eunicidce, Nereidce, Phyllodocidce, Syllidce, Opheliacea, Typhloscolecidce,

iv VEEMES— SENSORY ORGANS 233

Sabellidce, Archiannelida, Tomopteridce, Ctenodrilus, and Aphanoneura. Sensory
organs of similar structure and in a similar position have been observed in Bryozoa
(Loxosoma Rhabdopleurci) and in Phoronis.

In the Chcetognatha (Fig. 151, p. 227) a circular band, partly con-
sisting of ciliated epithelial cells, lies like a ridge on the ordinary
epithelium cells. It is considered to be an olfactory organ, and lies
behind the eyes between the head and the trunk; this unpaired
sensory organ is innervated by a pair of olfactory nerves running
between the nervi optici.

D. Organs of Taste (Cup-shaped Organs).

There occur also certain sensory prominences of the body
epithelium, essentially similar in structure to the olfactory organs just
described, and called cup-shaped organs from the fact that they can be
withdrawn into pit- or cup-shaped depressions of the integument. They
always occur in large numbers and widely scattered. As, however, they
are specially numerous at the edge of the mouth, in the oral cavity,
and also in the pharynx, they are held to be organs of taste.

The structure and distribution of these organs in the Capitellidce
is well known. In Notomastus, Dasybranchus, and Heteromastus
they occur only in the prostomium, thorax, and pharynx, in Masto-
branchus .and Capitella on the abdomen also. Similar sensory organs
are found also in OligocJiceta (Lumbricidce, Chcetogastridce, Enchytrceidce),
especially numerous in the head, chiefly on the upper lip. Among
the Polychceta they have been observed in the Nereidce (Nepliihys) and
the Eunicidce (on the pharynx and in the buccal cavity). In the
Hirudinea, where they were first observed and exactly described, they
are always found on the lips. Cutaneous sensory organs, which are
found in great numbers in the Sipunculidce and JEchiuridce on the
papillse not only of the body, but of the proboscis as well (often
arranged, like the papillae themselves, in longitudinal or transverse
rows), probably also belong to the category of cup-shaped organs.

E. Lateral Organs.

These retractile sensory organs only occur in the Chcetopoda and
agree essentially in structure with the cup- shaped organs. The
numerous thread-like sensory cells of these organs carry sensory
hairs, and are connected on the one hand with transverse muscle
fibrillae which together form a retractor for the organ, and on the
other hand with a plexus of ganglion cells which is again connected
by a special nerve with the ventral chord. The above is the case in
those Capitellidce which have been most carefully examined in this
connection. The lateral organs are most clearly distinguished from
the cup-shaped organs by their strictly segmental arrangement. There
is a pair in each segment, one on each side between the dorsal and the

234 COMPARATIVE ANATOMY CHAP.

ventral parapodia. They are found not only in the Capitellidce, but
also in Polyophthalmus and the Amphictenidce, and among the Oligochceta
in the Lumbriculidce, There are many reasons for considering the
lateral organs to be homologous with the dorsal cirri of the ventral
parapodia of other Polychceta, and in the family of the Glyceridce we
can follow, almost step by step, the transformation of these cirri
into lateral organs. The cirri, being sensory organs, their gradual
reduction into papillaB causes the tactile cells scattered in their
hypodermis to collect together to form the compact sensory epithelium
of the lateral organs.

Strands of lateral ganglion cells appear always to occur in the
lateral lines of the Oligochceta, entering the brain anteriorly. They are
closely connected with the hypodermis, and in the Lumbriculidce supply
the lateral organs with nerves. They are probably also connected
with the intestinal nervous system.

The function of the lateral organs is at present an unsolved
problem.

F. Auditory Organs.

There is a remarkable absence of auditory organs in the Vermes.
They only occur in the Polychceta, and there also only occasionally in
a few families, viz. in the Arenicolidce, in the Terebellidce (Lanice), and
Serpulidce (Myxicola, Amphiglene, Fabricia). Their occurrence has also
been proved in several Terebellid Iarva3, in the larvae of Eupomatus
(Serpulidce), and in a nearly related Chcetopodan larva. They are
paired, and lie in Arenicola on the cesophageal commissures in
the head segment, and in this case receive their nerves from the
brain. In other forms they lie, as it appears, in the first trunk
segment, and are supplied with nerves by the infra -cesophageal
ganglion. In this point, as well as in their development, they recall
the auditory organs of the mollusca. In adult animals they are
vesicular (otocysts), the wall being formed by epithelial cells (sensory
cells, auditory cells). The vesicles contain a fluid in which one or more
otoliths are suspended.

G. The Lateral Eyes of Polyophthalmus.

Eye-like organs are found in strictly segmental order somewhat
beneath the insertion of the transverse muscular bands in the lateral
line of Polyophthalmus. They occur in P. pidus in the 8th to the
19th body segments, and are closely connected with the hypo-
dermis which is free from pigment, and which with the cuticle
covers each eye. Each eye consists of a lens, a pigment cup, and
a body which consists of prismatic cells placed in this cup. The
pigment cup and cell body together perhaps form a sort of retina (?)
It must be expressly noticed that besides the lateral eyes, Polyoph-
thalmus possesses cephalic eyes as well (3 in number) and lateral organs,
and that it is very doubtful whether the lateral eyes are visual organs.

IV

VERMES— NEPHRIDIA

235

IX. Excretory Organs — Nephridia.
(Occasionally Ducts for the transmission of the Sexual Products.)

Nemertina. — The excretory apparatus, which is always paired,
consists of canals lined with epithelium and mostly ciliated, which as
a rule rise in the blood sinuses of the body and open externally.

It is limited to the anterior portion of the bod
always lie laterally over the longitudinal trunks
of the nervous system. Its arrangement differs
greatly in details. In all Nemertina, the epithelial
walls of the excretory canals (nephridia) are
glandular.

In Carinella (Paleonemertina) an excretory
portion separates itself from the lateral vessels
of the blood-vascular system, and falls into two
parts, a glandular part and a reservoir. On the
one side it is connected with the lateraL vessels
at two points, on the other with a canal which
opens externally. In Carinoma the nephridium
on each side consists of (1) a very short longi-
tudinal canal which communicates with the lateral
vessels at three points, and (2) an efferent duct
which opens externally. In Carinina also the
nephridial system on each side consists of two
parts : (1) a compact mass of small canals, which
projects inwards towards the blood sinus of the
cesophageal region, and which (2) is connected
with a nephridial cavity narrowing at its posterior
end into a canal which opens externally.

While in the Paleonemertina (with the excep-
tion of Carinina) the nephridia are in__gpea- com-
munication with the blood- vascular- system, such anterior

The efferent ducts

TUf

portion of the
a communication has till now not been proved in body of a Nemertian, dia-

.-, -,r ,. -r ,1 n i • •• i grarurnatic. n, Longitudinal

other i\emertma. In the bcluzonemertum on each canais of the nephridial
side of the anterior region of the body there system; np, lateral apertures
is either a single longitudinal canal or else a

of the same ; na,
ducts ; vd, dorsal vessel ; vl,
longitudinal network of canals, m which, however, lateral vessel ; gs, transverse

one principal canal can generally be distinguished, vessels between the dorsal
These longitudinal canals, which lie in or on the a
blood sinuses or lateral vessels on the inner side of the longitudinal
muscle layers, reach the exterior on each side either (1) posteriorly
through an efferent duct, or (2) through two ducts opening near the
middle of their course, or (3) through several often very numerous
lateral ducts which are more or less metamerically arranged.

In the Hoplonemertina and Malacobdellidce also, on each side, in
the anterior region of the body, a longitudinal canal is found. This

236 COMPARATIVE ANATOMY CHAP.

gives off, along its whole length, branches which again ramify. The
longitudinal canals and their branches lie neither on nor in the blood
sinuses, but are directly imbedded in the gelatinous connective tissue
(body parenchyma). The longitudinal canals either open externally
through several lateral efferent ducts, or on each side by one lateral
canal, which may branch off from the longitudinal trunk either
anteriorly, in the middle, or posteriorly.

The nephridial system of the Schizonemertina and Hoplommertina shows a certain
agreement with that of the Platodes. Here, as there, we meet with lateral longi-
tudinal trunks which open externally on each side, either through one aperture
(e.g. in the Platodes among the Rhdbdoccdida : Derostoma, Prorhynchus, Gyrator,
Mesostoma ; and among the Trematoda : Polystomidce), or through numerous lateral
ducts in more or less segmental arrangement (e.g. in Platodes among the Tridada).
The longitudinal trunks of the Nemertina may even, as in the Platodes, be broken
up into a plexus, or they may be present in numbers. Where the nephridial system
in the Nemertina (Hoplonemertina) is neither in direct nor indirect communication
with the blood-vascular system, it belongs to the branched type, as in the Platodes,
which have no blood-vascular system. In the Nemertina, it is true, the nephridial
system lies only in the front or foremost portion of the body. Terminal excretory
cells have not been proved to exist, and the canals are lined with a ciliated epi-
thelium, while in the Platodes (everywhere and in all divisions ?) each canal runs
within cells arranged in a single row (intracellular). These differences, ho\vever,
ought not to prevent a recognition of the homology between the excretory systems
of the Platodes and the Nemertina.

Nemathelmia. — The Nematoda and Acanthocephala must be
described separately. In the first, longitudinal canals occur in the
lateral lines ; till now no inner lining of epithelium has been proved
to exist. The two longitudinal canals unite at the anterior end
of the body to form a longer or shorter unpaired canal, which opens
externally near the brain by a ventral median pore. The homologies
of these canals, which are considered to be excretory canals, are quite
uncertain. In the Gordiidce they are wanting. The canal-like space
(section of the body cavity) which surrounds the intestine of these
animals is said to divide (in Gordius Preslii) in front of the cloaca into
two branches, and these perhaps open into the cloaca. Whether this
canal is an excretory tube is, however, quite uncertain.

In the Acanthocephala, in the subcuticle of the integument, a
system of canals is found which will be described in the section on
the vascular system. The anterior part of this canal system, which is
quite separated from the posterior part, was formerly claimed as an
excretory system ; but there are difficulties in the way of accepting
this view, chiefly because it has no external aperture.

Annulata. — The following scheme may be given as of general
application to the excretory or nephridial system. It consists of
paired tubes (nephridia), open at both ends, which are repeated
segmentally. Each nephridium is in open— communication with the
body cavity or blood sinuses by an inner aperture ; the external
aperture lies in the integument. The nephridia therefore form an

IV

VERMES— NEPHRIDIA

237

open communication between the body cavity and the exterior, and
serve chiefly for conducting the waste products of metabolism out
of the body/ Since the genital products are, in many Annulate,
developed out of the endothelium of the body cavity, then free them-
selves from the matrix, and ripen when floating freely in the ccelomic
fluid, an opportunity is given to them also of reaching the exterior
through the nephridia. The nephridia thus frequently undertake, in
addition to their purely excretory function, the transmission of the
genital products to the exterior. This secondary function may often
become the principal function in some of the nephridia, which
may then undergo a complete transformation, and in the Polycliceta
are called genital tubes.^x

As already explained, paired nephridia originally occur in all the

OD fan

FIG. 158.— Transverse section through a trunk segment of a carnivorous Annelid,
diagrammatic. &, Setae ; ac, aciculum (supporting seta) ; Im, longitudinal musculature ; vd, dorsal
vessel ; k, gill ; dc, dorsal cirrus ; dp, dorsal parapodium ; vp, ventral parapodium ; vc, ventral
cirrus ; tm, transverse muscles ; bm, ventral chord ; vo, ventral vessel ; rm, circular musculature ;
ov, ovary ; np, nephridia ; tr, nephridial funnel ; md, mid-gut. In the body cavity are eggs.

segments of the Annulate body, even in the cephalic or oral segment.
It may, however, happen that the nephridia in a smaller or greater
number of segments do not attain development. It is further very
generally found that some of the nephridia begin to form early in the
embryo or larva or young animal, and function as embryonic or larval
kidneys, but afterwards entirely disappear when the permanent
nephridia attain development. We shall call those which temporarily
appear in the ontogenetic development provisional or embryonic
nephridia. We can again distinguish two sorts of such provisional
nephridia. (1) Those which appear in that region of the embryo or
larva which corresponds with the subsequent head segment, and lie at
the anterior end of the cell mass (mesoderm streaks), from which the
most important organs of the segmented mesoderm come ; these are

238 COMPARATIVE ANATOMY CHAP.

the embryonic head nephridia or the head kidneys. (2) Those
which appear in the trunk segments ; these are the provisional trunk
kidneys. The permanent nephridia, on account of their frequently
strict segmental arrangement, are often called segmental organs, or
on account of their looped or winding course (in Oligochceta and
Hirudinea) looped canals.

We will first describe the three sorts of nephridia separately, and then discuss
their morphological significance and their relations to each other.

A. The embryonic head nephridia (head kidneys). — These appear temporarily in
the larva or embryo, and are paired. Their inner end lies in the embryonic head
cavity. They have been observed in many Oligochceta and Polychceta. They are
ciliated canals, which are not in open commnnication with the head cavity. The
lumen of these canals is intracellular, i.e. the nephridia are rows of consecutive cells
perforated to form a canal. In this point the embryonic head nephridia agree with
the permanent nephridia of the Oligochceta and Hirudinea, and with the canals of
the water- vascular system of the Platodes. They are occasionally branched (e.g. in
the larvse of Echiurus and Polygordius), like the water-vascular system of the Platodes
and the nephridia of many Nemertina. Lateral branchings of the principal canals
also occur in the permanent nephridia of the Hirudinea and Oligochceta. Terminal
cells provided with bundles of cilia (flames) often occur at the inner ends of the
branched or simple nephridia ; the flames project into the lumen of the canal, in
which they undulate. These terminal cells resemble those of the water- vascular
system in the Platodes.

B. The embryonic or provisional trunk nephridia. — These have till now
been observed in comparatively few cases ; it is, however, probable that they are
widely distributed. They occur (like permanent nephridia) in strictly segmental
arrangement as paired canals in those generally anterior^ segments, in which in
adults permanent nephridia are wanting.

Among the Oligochceta it has been proved that in Rhynchelmis, in the 5
anterior trunk segments in which nephridia are wanting in adult animalg,
5 pairs of provisional nephridia, which degenerate later, attain development
in the embryo. In the Capitellidce the nephridia are always wanting in
a large number of anterior segments (thorax and anterior part of the
abdomen), but this is only the case in adult animals. In the young animals,
however, we meet with provisional nephridia in most of these segments, which are
the better developed the younger the animal, and the further forward the segment
to which they belong. In other words, the nephridia arise first anteriorly, and their
degeneration proceeds in order from before backward (in the thoracic and in
some of the anterior abdominal segments) in proportion as the permanent
nephridia of the abdominal region attain development. In Nereis cultrifera
(Fig. 159) there are found in the larval stage 5 pairs of provisional or larval
trunk nephridia in the 5 anterior segments, in which in adult animals we meet
with no nephridia. As in most Oligochceta nephridia are wanting in some of the
anterior segments, it is probable that provisional nephridia occur in these segments
in the larval forms. The same may be conjectured of the Polychceta. In the
Oligochceta the nephridia are wanting in the genital segments, except in the
Lumbricidce. But here also provisional nephridia attain development in these
segments at early stages. In the Hirudinea provisional nephridia are developed in
early embryonic or larval stages (2 pairs in Nephelis, 3 pairs in Hirudo, 4 pairs
in Aulostoma) which disappear early. As the larvae are, at the time when pro-
visional nephridia are present, still unsegmented, it cannot be certainly decided what

IV

VERMES—NEPHRIDIA

239

is their morphological significance. A certain number of the anterior and of the
posterior segments of the adult Hirudo are without nephridia. This fact favours
the conjecture that the larval nephridia of the Hirudinca are the provisional nephridia
of the anterior trunk segments. Possibly
the foremost pair of larval nephridia of the
Hirudinea represent the embryonic head
nephridia (head kidneys) of other Annulata.

Concerning the structure of the pro-
visional trunk nephridia, the following
may be said. In the Capitellidce and
Oligochceta they show in general the same
structure as the permanent trunk nephridia.
In Xereis they are distinguished from the
permanent nephridia by the want of an
inner aperture opening into the body
cavity, i.e. of a funnel ; both by this fact
and the fact that the nephridial canal is
intracellular they recall the larval head
nephridia of many Annulata. The larval
nephridia of the Hirudinca have neither
inner nor outer aperture.

C. The permanent nephridia. — In
every Annulate nephridium, if we for the
time ignore the numerous complications
and modifications presented by the differ-
ent divisions, the following three portions
may be distinguished : — (1) an inner 'cili-
ated aperture opening into the body cavity
or into a blood sinus ; this from its shape is
often called the funnel ; (2) a canal con-
nected with the above, which is generally
ciliated, and often has glandular walls ;
and (3) a terminal portion opening exter-
nally. The central part or nephridial canal

is intracellular in the Hirudinea and FlG. 159._Diagram of a very y0ung sped-
Oligochceta, and generally much coiled men of Nereis cultrifera, after Edward Meyer,
(looped canal) (Fig. 160). In the Polychceta g, Brain ; au, eyes ; In, larval trunk nephridia ;
it is usually intercellular (lined with a Ph> pharynx with jaws ; di, dissepiments ; d,

many-celled epithelium) and not coiled in iniestin1e ; <™> ventral ^> ed' ^ld:gu*; ^
J parapodia with cirri and setae. On the head are

a complicated manner. The portion of the tentacles and sensory cirri,
the Annulate nephridium which projects

into the body cavity is outwardly covered by a continuation of the peritoneal endo-
thelium.,

In the Hirudinea the permanent nephridia are wanting in a number of the
anterior and posterior segments. In the rest of the body they are found in strictly
segmental arrangement, one pair in each segment. The position of the funnels
in the body varies 'very much ; they lie either in the ventral blood sinus
(Clepsine\ or in those sinuses in which the testes lie (Hirudo, Aulostoma), or in
other blood sinuses of the body. The nephridial canal has many windings and
loops which lie close together, the finer details of which it is extremely diffi-
cult to make out. Finally, it opens externally either directly without terminal
swelling (Clepsine), or it opens into a vesicle lined with epithelium (ciliated in H.

240

COMPARATIVE ANATOMY

CHAP.

•it

medicinalis), which opens externally through a pore. That part of the nephridial
canal which comes after the funnel is distinguished by the fact that fine, branched,
and often anastomosing canals enter its intracellular principal lumen. The cells of
this portion are in fact perforated by branching canals. In Hirudo the funnel is
closed towards the blood sinus in which it lies.

The nephridia in the genera Pontobdella, Branchellion, and Piscicola differ
much from the above, as they form in each segment a complicated network of
canals which are always intracellular, this network opening outwardly by 2 apertures
and entering the blood sinuses of the body through 2 funnels.

The nephridia of other Oligochceta show (Fig. 160) great correspondence with

those of the Hirudinea. The funnel of a ne-
phridium projects from the anterior wall of each
dissepiment into the cavity of the segment lying
anterior to that in which the rephridium lies.
Starting from the funnel, the nephridial canal,
which is everywhere intracellular, first passes
through the dissepiment, forms more or less
complicated loops in that segment of the body
cavity which lies posteriorly (in these coils we
can generally distinguish several different por-
tions), and finally emerges through a terminal
portion into a vesicle which opens outwardly.
This vesicle is often provided with muscular
walls. The funnel and external aperture of a
nephridium ' thus always lie in two different
segments Jvthe two external apertures of a pair
of nephridia lie in the same segment as the
inner funnels of the pair of nephridia which
come next in order posteriorly. This position
of the inner and outer apertures of the nephridia
in 2 consecutive segments is maintained even in
those cases where, as in the middle body seg-
ments of Phreatothrix, the nephridial canal
passes through several dissepiments, running
back from its ciliated funnels through several
segments ; it then forms a loop and bends for-
wards again. In the nephridia of the Chceto-
gastridce the ciliated funnels are wanting. In
them, as in the Hirudinea, numerous branched
and anastomosing intracellular canals enter the
central canal. In a species of Acanthodrilus
typically 4 pairs of nephridia in each segment (even in

FIG. 160.— Nephridium of an Oli-
gochaete, diagrammatic, tr, Funnel ; dis,
dissepiment; ng2, glandular; ngi, non-
glandular portion of the nephridial duct ;
eb, terminal vesicle ; Iw, body wall (partly
after Vejdovsky).

(Lumbricidce) there are

the genital segments). There is said to be a similar arrangement of nephridia in
the anterior segments of Perichceta mirabilis.

•The permanent nephridia of the Polyehceta are tubes with cellular walls ; their
often ciliated central canal is thus as a rule, in opposition to that of the
Oligochceta and Hirudinea, intercellular. (Intracellular nephridial canals, however,
also occur. ) The nephridial tube is almost always so bent that we can distinguish
in it two limbs, one centripetal, at whose inner end lies the funnel, which is mostly
wide open and provided with cilia, and another centrifugal, which opens outwardly
by breaking through the body wall. The nephridia lie in the nephridial or renal
chambers of the body cavity which have already been described (p. 213), and they may

iv VERMES— NEPHRIDIA 241.

lie wholly in one segment, or else, as in the Oligochceta, each pair of nephridia belongs
to two consecutive segments.

Although there are many Polychceta in which the pairs of nephridia are repeated
throughout the greater part of the body with great uniformity and in strictly seg-
mental arrangement, there are, on the other hand, many groups in which we find great
deviations from this arrangement. We can only mention the most important.

In the Capitellidce permanent nephridia occur, as a rule, only in the abdominal
region, either in the greater part of that region or only in the anterior, or only in the
posterior portion. There is either one pair in each segment or several (Capitella] ;
even as many as 6 pairs may occur in each segment. In most Capitellidce more or
less numerous pairs of nephridia are changed into genital tubes, which will be
further described below, and it is always the anterior pairs of nephridia which
undergo such a transformation. The permanent nephridia of Capitella are distin-
guished by the fact that they possess, as a rule, more than one funnel.

In the Terebelloidea the nephridia only occur in the thoracic region, and in
strictly segmental arrangement. In this region the dissepiments are wanting, with
the exception of a strongly developed diaphragm which divides the coelome of the
thoracic region into an anterior and a posterior cavity. The nephridia of the anterior
thoracic cavity function as organs of excretion ; those of the posterior conduct
the genital products to the exterior. In Lanice conchilcga there is a very striking
nephridial arrangement. The 3 pairs of nephridia of the anterior thoracic
cavity do not emerge externally direct, but the 3 nephridia of each side enter a short
nephridial duct which has a single external aperture. In a similar way the 4
nephridia of the posterior thoracic cavity enter on each side a longitudinal nephridial
duct, which, however, has not 1 opening but 4. It may therefore be said that
the 4 nephridia on each side are connected by a longitudinal canal.

In the Cirratulidce, Serpulacea, and ffermella (Fig. 147, p. 221) an anterior
sterile region and a posterior genital region may be distinguished, the genital products
attaining development in the latter. In the anterior sterile region, which consists
of a varying number of segments, only one pair of nephridia occurs^ This pair alone
has an excretory function. They are long, and extend through several segments.
In the Cirratulidce they emerge ventrally by separate apertures in the third segment ;
in the Serpulacea and Hcrmella, however, they unite anteriorly to form an unpaired
duct, which reaches the exterior near the extreme anterior part of the body in the
dorsal middle line. In the genital region the nephridia are repeated in strictly
segmental arrangement, and serve as genital tubes for conducting the sexual products
to the exterior.

In Sternaspis, two brown lobate bodies lying in the 5th and 6th segments are
regarded as nephridia ; these possess neither a lumen nor an internal aperture, and
end in the integument between the sixth and seventh segments.

In the Ecliiuridce (Fig. 137, p. 207) there are two sorts of organs
which have been considered as nephridia, the so-called segmental organs
and the anal tubes. The segmental organs have quite the structure
of the permanent Polychsetan nephridia, and we can hardly doubt their
homology with the latter. They occur either in 2 pairs (JEchiurus),
or 3 pairs (Thalassema), or unpaired and singly (Bonellia), and
possess well -developed internal funnels. Their outer apertures lie
behind the anterior hooked setae. Their principal function is the
transmission of the genital products out of the body 'cavity. The
anal tubes are 2 long tubes which on the one hand enter the
VOL. I R

242 COMPAEA TIVR ANA TOMY CHAP.

hind-gut, and on the other are in open communication with the body
cavity by means of numerous ciliated funnel apertures, one of which
always lies terminally. An excretory function is ascribed to them.
Whether they represent a pair of modified nephridia cannot at
present be decided. The fact that they are supplied with numerous
funnel apertures ought not to stand in the way of sach a view, as
typical permanent nephridia in the Polychceta (Capitella) and Oligochceta
(Anachceta} may be provided with accessory funnels.

Organs which may with certainty be pointed to as nephridia have
until now not been observed in the Myzostomidse.

The problem of the morphological relations between the permanent nephridia,
the provisional trunk nephridia, and the embryonic head nephridia, is still unsolved.
It is closely connected with the questions as to the significance of segmentation in
the Annulate body and the morphological significance of the body cavity and the
mesoderm. It is very probable that the permanent segmentally-arranged nephridia in
all the Annulata are homologous. The histological difference between the nephridia
of the Hirudinea and Oligochceta on the one hand and those of the Polychceta on
the other, which consists in the fact that in the former the nephridia are perforated
rows of cells and in the latter as a rule tubes with epithelial walls, would in that
case be unessential, as also would be the absence or presence of branchings. It is
further probable that the provisional trunk nephridia are morphologically equivalent
to the permanent nephridia. They are distinguished from the latter only in that
they appear earlier and disappear in proportion as the permanent nephridia appear
and assume their functions. The homology between provisional trunk nephridia
and permanent nephridia is supported by the circumstance that no case has as yet
been known in which permanent nephridia have attained development in a segment
where provisional nephridia have previously appeared and then disappeared. Capi-
tella, in which in the tenth and eleventh segments both provisional and permanent
nephridia develop, is an exception, but only an apparent exception, for we have seen
that in this animal several pairs of permanent nephridia occur in one segment.
It is probable that the embryonic head nephridia are homologous with the
trunk nephridia (provisional and permanent). The whole nephridial apparatus
would then have to be judged of in the following way. Originally a pair of nephridia
occurs in each segment of the segmented Annulate body, even in the head segment.
All the pairs of nephridia are segmentally homologous with each other. The larva
or embryo of the now living Annulata consists of the embryonic head segment and
the unsegmented rudiment of the trunk. The differentiation of the trunk occurs
from before backward ; the first and oldest segment to be developed is the first trunk
segment ; then follows the second trunk segment, and so on. In correspond-
ence with this, the pair of nephridia of the head segment first appears (em-
bryonic head kidneys), then the pairs of nephridia of the anterior trunk segments,
several of which (provisional trunk nephridia) may degenerate gradually as new
nephridia, the permanent trunk nephridia, begin to form behind them. The cause
of the disappearance of the head and the provisional trunk nephridia in the
course of development is perhaps to be found in the fact that the foremost body
segments in the adult animal are crowded with organs (pharynx, brain, etc.) Besides
this, in the whole animal kingdom organs which attain development very early and
function during larval or embryonic life show a tendency to degenerate early, as
if they were soon worn out. Further in the Oligochceta (excepting the LumbricMce)
the nephridia in the genital segments degenerate when the various sexual organs
attain development.

iv VERMES— NEPHRIDIA 243

The nephridia as organs of excretion. — The original and most
general function of the nephridia is that of excretion. The discharge
of the excretory products occurs in two ways. The nephridia may
either take up the excretory products direct out of the body fluid
or out of the blood by means of the open funnels, or the excretory
products may collect in the walls of the nephridia and thence reach the
nephridial cavity. The nephridia are often surrounded by a rich net-
work of blood-vessels which yield up to the nephridial walls the
excretory material they contain.

The nephridia as duets for the sexual products. — Since the
nephridia establish an open means of communication between the body
cavity and the outer world, an opportunity is offered to the sexual pro-
ducts floating in the coelomic fluid to choose this way of reaching the
exterior. , In the Polychceta the nephridia in fact act at the same time
as sperm ducts and oviducts. In the simplest cases this new function
does not bring about any marked variation in the form and struct-
ure of the nephridia, the funnels at most showing slight enlargement
at the time of sexual maturity. It often happens, however, that some
of the nephridia act almost exclusively as sperm ducts and oviducts.
Their funnels are then strikingly enlarged, while at the same time the
(excretory) nephridial canal diminishes in size and becomes simplified.
Some of the nephridia in the Capitellidce undergo even more profound
modification. The funnel becomes enormously enlarged, and connected
with the exterior by means of a new canal, which breaks through the
body wall and opens outwardly through a genital pore, while at the same
time the nephridial canal may become reduced and may quite disappear.
Thus arise the genital tubes ; they act as ducts for the sexual products
and as copulatory organs. The sperm ducts and oviducts of the Oligo-
chceta have also long been considered as modified nephridia, and it can-
not be denied that they show great agreement with nephridia in their
structure and composition (funnel, canal, or duct and terminal vesicle)
as well as in their development. Nevertheless it is a striking fact
that in the Oligoclmta^ where the nQphridia occur in strictly segmental
arrangement (one pair in each segment), either permanent (iMmbritidce)
or provisional nephridia (other Oligochceta) are found in the genital
segments also, side by side with the sperm ducts and oviducts. In
Acanthodrilus, where 4 pairs of nephridia occur typically in each
segment, there are also 4 pairs in the genital segments. These are
difficulties which cannot be ignored in the way of establishing a
homology between the oviducts and sperm ducts of the Oligochceta and
nephridia. The ducts for the sexual products in the Hirudinea and
Myzostvmidce can certainly not as yet be considered as modified
nephridia.

The phylogenetic origin of the nephridial system of the Annulata is still
quite uncertain. There are three different views. According to one of these, the
whole nephridial system of the Annulata corresponds with the water -vascular
system of the Platodes and with the excretory system of the Nemertina, which (in the

244 COMPARATIVE ANATOMY CHAP.

Triclada and certain Nemertina] already shows a more or less distinct segmentation,
in that the efferent ducts are segmentally repeated. In all Platodes and Nemertina,
however, longitudinal canals are present throughout the whole body or in the
anterior part of the body, which can open externally through more or less distinctly
paired and segmentally arranged ducts, whilst in all Annulata the nephridia are
separate at their first appearance, and with a very few exceptions (Lanice) remain
separate during life. While the water- vascular system of the Platodes is markedly
branched, such branching is less marked even in the Nemertina, and in the Annulata
(Oligochceta, Polychceta) is generally altogether wanting. This may be explained by
the fact that in the parenchymatous Platodes the excretory organs have to seek
out the excretory products all over the body, while the development of a blood-
vascular system and a body cavity affords spaces for collecting these products, out
of which the nephridia can take them direct. The nephridial funnel in the Annulata,
which ontogenetically originates quite separately from the other parts of the nephri-
dium, would then be a new adaptation, a collecting apparatus, suited for taking up
the excretory material out of the blood sinuses or body cavity, and for discharging
it through the nephridial canals. According to a second view, only the embryonic
head kidneys of the Annulata correspond with the water-vascular system of the
Platodes, with which they certainly often show a great structural resemblance. A
third supposition is that the head kidneys of the Chcetopoda and the embryonic
kidneys of the Hirudinea answer to the excretory organs of the Nemertina, while
the permanent nephridia may have arisen from the efferent ducts of the ovaries and
testes of the Nemertina. This last conjecture is opposed by the fact, which is:
unanimously supported by all recent investigations, that the original function of
the Annulatan nephridium is excretory, and that only secondarily some of the
nephridia undertake the transmission of the sexual products.

Prosopygia. — The number of the nephridia is everywhere in this
class very small ; there are never more than two pairs. In the Sipun-
culidce (Fig. 138, p. 208) there are two large tubular nephridia like
the permanent nephridia of the Polychceta, especially of the Echiuridce.
They emerge laterally at the limit between the proboscis and the
trunk (near the anus). Less frequently (Phascolion) only one nephri-
dium occurs. The nephridia, besides their excretory functions, serve as
ducts for the transmission of the genital products. In Sipunculus
anal tubes which enter the hind-gut have been observed, which are
perhaps homologous with the anal tubes of the Echiuridce and the
Priapulidce. The Priapulidce have only two richly-branched anal tubes
emerging near the anus. At the blind ends of the branches are found
terminal cells with long flagella projecting into the canals, simi-
lar to those which are characteristic of the water- vascular system of
the Platodes and of some of the embryonic head nephridia and the
provisional trunk nephridia of the Annulata. The anal tubes of the
Priapulidce' are said to act as excretory organs in youth, and in later
stages as places of formation and ducts for the transmission of the
sexual products. Phoronis possesses one pair of nephridia which open
outwardly and anteriorly by two lateral apertures, and besides their
excretory function also undertake the transmission of the genital
products out of the body cavity. Among the Bryozoa nephridia
have till now been found only in the Endoprocta. They are paired

IV

VERMES— NEPHRIDIA

245

nt

canals like the embryonic head nephridia of the Annulata, which issue
between mouth and anus rc~

into the so - called vesti-
bulum. The Brachiopoda
(Fig. 150, p. 226) possess
one pair, less frequently
(Rhynchonella) two pairs
like those permanent ne-
phridia which in the Poly-
chceta, Sipunculidce, and
Phoronis discharge the
sexual products. They
emerge to the right and
left of the mouth into the
mantle cavity.

Rotatoria and Dino-
philus. — Dinophilus gyro-
ciliatus(Fig. 162) possesses
five pairs of nephridia, In
which show a remarkable
agreement with the provi-
sional trunk nephridia of
certain Polyclmta (Nereis
cuUrifera, Fig. 159, p. 239).
They lie one behind the
other in the trunk region
in those segments which
are outwardly indicated
and demarcated by ciliated
rings. The nephridia of
the Rotatvrici (Fig. 161)
consist of two looped, and
in certain places much con-
voluted canals, which run
longitudinally near the
intestine ; these open
into the cloaca, generally
forming a contractile
terminal vesicle. The
longitudinal canals usually
have short accessory
hrflnrhps whosp ends FlG< ^--Organisation of Hydatina senta, after Plate.
j 11^ -1 -i ro' Wheel or§an ; nt> nePMdial ciliated cells 5 n> nephridia ;

(ciliated lobes, Vlbratlle Ph, pharynx ; md, gastric glands ; m, stomach ; ds, vitel-
Seem tO be COn- laruim; fcs, germarium ; c&, outline of the contractile vesicle;
li'l^-P thp PTirk of ^.hind-Sut;«.uterus;a.anus;/^cementorPedalS1ands;
Ui It, lateral feeler ; In, nerve of the same ; e, advanced egg.

the water-vascular system

of the Platodes, and of the embryonic head nephridia of the

246

COMPARATIVE ANATOMY

CHAP.

ill /

Annulata. This correspondence indeed is apparent in the whole

nephridium.

The only organs which can perhaps be pointed to as transformed
nephridia in the Chcetognatha are the ovi-
ducts and the sperm ducts; these are
paired tubes, the oviducts opening out-
wardly, at the posterior end of the trunk
segment, the sperm ducts in the tail
segment by paired lateral apertures.
The latter are provided with funnel-like
ciliated inner apertures.

X. Respiratory Organs.

In many worms no specific respiratory
organs are developed. Respiration is
performed by means of the integument,
and also often by the walls of the intestine.
The general ciliation of the body in the
Nemertina is of great assistance for respira-
tion in water. In the Hirudinea and
many Oligochce-ta cutaneous respiration is
facilitated by the presence in the integu-
ment of numerous fine blood-vessels. But
we can only speak of specific respiratory
organs where definite organs have been
developed whose exclusive or most im-
portant function is respiration. Among
the Hirudinea we see in Pontobdella the
integumental capillaries of the blood -
FIG. 162.— organisation of Dino- vascular system already localised in eleva-

md

pharynx ; 'phd, pharyngeai glands ; n, therefore be called the branchial papillae.
nephridia; md, mid-gut; wk, ciliated jn Bmnchellwn each ring carries on each

rings ; o. ovarium ; an, anus ; ed, hind- . -, •, -IT •> • , i ' t

gu1r side a branched appendage into which

blood-vessels enter. Specific respiratory

organs are wanting in the Oligochceta and the Archiannelida. Among
the Oligochceta, only the insufficiently known Alma nilotica possesses
branchial appendages at the posterior part of the body. The Polychceta,
on the contrary, are pretty generally supplied with gills, which
are usually branched appendages of the Parapodia (Fig. 124, p. 188 ;
Fig. 158, p. 237); these in some cases, however, may emancipate
themselves from the parapodia and may be independently inserted
on the back. The comparative morphology of the gills of the Polychceta
has not yet been thoroughly worked out. We may perhaps at once
distinguish two sorts of gills — lymph gills and blood gills. The
lymph gills are processes of the parapodia which are distinct from

iv VERMES— RESPIRATORY ORGANS 247

the parapodial cirri ; they may occur simultaneously on the ventral
and dorsal parapodia. They are found in the Capitellidce and the
Glyceridce, and in the absence of a separate blood-vascular system are
provided with continuations of the body cavity carrying haemolymph.
The blood gills, on the contrary, are often branched appendages of
the parapodia which are penetrated throughout by blood-vessels. We
can again distinguish two sorts of blood gills — dorsal gills and cephalic
gills. The dorsal gills are transformed cirri of the dorsal parapodia, or
transformed lateral off-shoots of such cirri. They occur, like the para-
podia themselves, in segmental order, but may often attain development
only in certain regions of the body (branchial regions). The cephalic
gills, on the other hand, which we meet with specially in tubicolous
Annelids, are transformed tentacles or feeler-cirri of the head, and often
form a beautiful crown of gills or tentacles projecting above the aperture
of the tube. These crowns are at the same time also the seat of a fine
sense of touch, and, further, organs for the drawing in of nourishment.
In the Sabellidce, (Branchiomma) eyes may be developed on the cephalic
gills. Where gills attain development they are almost always either
ciliated or mobile, so that a constant exchange of the respirable
medium, oxygenated water, is secured. In the thoracic membrane of
the Serpulidce there is a rich network of vessels, and this no doubt has
a respiratory significance. In Sternaspis, on each side of the anus, there
is a tuft of filamentous gills. We do not find special respiratory organs
in the Myzostomidce, the Chcetognatha, and the Rotatoiia, in whom a blood-
vascular system also is wanting. In all these forms cutaneous respira-
tion must take place, facilitated in the wheel animalculae by the activity
of the wheel organ. The cutaneous respiration in the Echmridce is
principally localised in the prostomium. In the Prosopygia the tentacles
which stand round the mouth, or the oral arms furnished with cirri
(Brachiopoda) chiefly act as gills, and these organs are either penetrated
by blood-vessels (Phoronis, Brachiopoda?) or supplied with canal or
vessel-like processes of the body cavity (Sipunculidce, Bryozoa). In the
Brachiopoda the inner surface of the mantle has in all cases an addi-
tional respiratory significance. The Priapulidce among the Sipunculacea
possess no oral tentacles; in them respiration takes place through the
integument. Besides this, however, there can be little doubt that
in Priapulus the deeply-lobed caudal appendage into which, in the
absence of a blood- vascular system, a continuation of the body cavity
extends, may be regarded as a respiratory organ.

Just as the cutaneous respiration may be concentrated in certain
localised parts of the integument, in which places the principle of in-
increase of surface is applied, accompanied by a richer vascularisation,
the enteric respiration also may be localised. The accessory intestine
which is found in some Polyclmta and Sipunculacea is said to represent
such a respiratory portion of the enteric canal. Respiratory organs
are wanting in the Nemathelminths.

248 COMPARATIVE ANATOMY CHAP.

XL Blood-Vascular System.

There is no organic system in the worms which is so variable as
the blood-vascular system. It is sometimes wanting, sometimes highly-
developed. We even find that it may be wanting in certain groups
whose nearest relations possess it. Thus, as we find a blood-vascular
system sometimes present, sometimes absent, in the most different
orders of the most different classes, its small morphological worth is
evident. Even where worms are supplied with a circulatory apparatus
it varies so much in structure that no morphological comparison is
possible, at any rate at present.

Nemertina (Fig. 157, p. 235). — In this class of the Fermes
for the first time in the animal kingdom we meet with a blood-
vascular system. In the Paleonemertina (excepting Falendniidce and
Polliidce) it consists of 2 lateral vascular trunks, uniting posteriorly
above the intestine, and anteriorly entering a lacunar system,
which likewise establishes communication between them. In the
Schizwiemertina (with the Polliidce and Valendniidce) there are 3 longi-
tudinal vessels, 2 lateral, and 1 medio- dorsal which lies above the
intestine, in the proboscidal region between the proboscis and the
intestine. The 3 vessels become lacunar anteriorly, and communicate
above and below the proboscis sheath. In the rest of the body they
are connected together by transverse vessels. The same is the case in
the Hoplonemertina, only a lacunar system is here wanting, and the
vascular system is completely closed. The blood is colourless or
contains red blood corpuscles. The vessels are lined with endothelium
and occasionally have muscular walls.

Nemathelminthes. — The Nematoda have no vessels. In the
Acanthocephala, throughout the whole subcuticle, a peculiar network
of canals extends, whose morphological and physiological significance is
still little understood. The system of canals, which are hollow spaces
without walls of their own, running in the very much thickened
subcuticle, consists of 2 completely separate parts — the canal system
of the trunk and the canal system of the neck, the proboscis, and
the lemnisci. In the trunk canal system we find 2 specially distinct
longitudinal trunks which run either laterally, or dorsally and ven-
trally. The neck, head, and lemniscal canal system enters a circular
canal situated at the base of the neck. The lemnisci (Fig. 172, /, p.
258 ) are two pouches, generally of a brown colour, which hang from
the base of the neck into the body cavity and are processes or
appendages of the subcuticle of the neck. A canal enters each
lemniscus from the circular canal, and divides into two branches
directly after entering ; these branches run longitudinally through the
lemniscus. Besides these, narrower canals also occur in the lemnisci.

Annulata. — The Hirudinea and the Chcetopoda are separated by
sharp and radical differences in the blood-vascular system. In the
Myzostomidce blood-vessels are altogether wanting.

IV

VERMES— BLOOD-VASCULAR SYSTEM

249

dm.

Hirudinea. — In describing the body cavity we have already drawn
attention to the difficulty of distinguishing it from the blood-vascular
system. We are, besides, not certain that these are two originally
separate systems, and it
seems almost necessary to
consider the whole to-
gether. There occur al-
most universally 4 longi-
tudinal vessels (Fig. 163)
—1 dorsal, lying over the
intestine, 1 ventral, in
which the ventral chord
lies, and 2 lateral, which
in many cases pulsate. Of
these 4 vessels the ventral
one (ventral sinus) may
best be considered as the

, ,. -, -j FIG. 163.— Transverse section through Hirudo, diagram-
principal part 01 a reduced matic> m> Circular;musculature ; Im, longitudinal muscular
body cavity. The dorsal layer ; vl, lateral vessels ; np, looped canals (nephridia) ; vd,
Vessel is wanting in Ne- dorsal vessel ; ^m> dorso-ventral musculature ; enp, terminal
° vesicle of the nephridia ; 6m, ventral chord ; vv, ventral
pliehs and some land vessel ; A, testes ; vd, vas deferens ; md, mid-gut.

leeches. The longitudinal

vessels are connected together, chiefly at the anterior and posterior
ends of the body, by fine vascular branchings. Such a connection also
takes place in various ways in other parts of the body. The peripheral

vascular system consists principally of
2 well -developed systems of branched
and often anastomosing capillaries, one
of which lies in the integument and
penetrates into the body epithelium, the
other spreading out over the intestine.
The excretory and sexual organs are
richly supplied with blood-vessels. In
-w Nephelis (Fig. 164) and land leeches
there are, in connection with the anas-
tomoses between the lateral and ventral
vessels, ampullae or blood vesicles in
segmental arrangement, one on each side
(land leech) or two together (Nephelis).

FIG. i64.-vascuiar system in 4 In Branchellion on each side, at the base
segments of the middle part of the of every third gill, there is a blood sinus,

Ventral vMsef^S lateraf Basel's • *a' ^^6116(1 into a Vesicle. The blood 6V6ry-

ampuiise. ' where contains colourless amoeboid cor-,

puscles, and often free nuclei. In the

Gnathobdellidce the blood is red. Haemoglobin is found dissolved in
the blood plasm.

By the presence of one dorsal and two lateral vessels, the blood-

250

COMPARATIVE ANATOMY

CHAP.

vascular system of the Hirudinea recalls that of the Nemertina. They

also agree in the possession of blood sinuses.

The vascular system of the Chcetopoda is strikingly different from

that of the Hirudinea. It is entirely separate from the body cavity.

Both the lateral vessels of
the Hirudinea are wanting.
The most important and
constant parts of the Chceto-
podan circulatory apparatus
are : (l^M^medio-dorsal
longitudin^Bj^ssel and (2)
a medio-venpil longitudinal
vessel (Fig. 165). The first
is mostly contractile ; in it
the blood streams from
behind forwards ; it lies
over the intestine, some-
times nearer the latter,
sometimes nearer the body
wall. The second is not

FIG. 165.— Transverse section through a Lumbricus,
diagrammatic. Ih, Body cavity ; eg, chloragogen cells ; contractile ; in it the blood
rm, circular musculature ; lm, longitudinal musculature ; gleams from before back-
dv, dorsal vessel ; ty, typhlosolis ; vt, typhlosolis vessel ;
np, nephridia ; vv, ventral vessel ; vln, lateral vessel of
the ventral chord ; bm, ventral chord with the neurochord
tubes ; vvn, sub-neural vessel ; bm, 2 setae of the ventral
row ; U, 2 of the lateral row.

ward. It lies in the body
cavity, below the intestine
and above the ventral chord,
approaching sometimes the
one, sometimes the other. In the details of the arrangement, develop-
ment, and course of the vessels there is extraordinary variety, which
makes it impossible to describe them briefly and comprehensively. In
a very simple case the ventral vessel divides at the anterior end of
the body into 2 branches, which, surrounding the fore-gut, enter the
anterior end of the dorsal vessel. The dorsal and ventral vessels are
further connected in each segment by lateral vascular loops ; it is from
these especially that branches proceed to the body wall. Vessels
coming from a vascular network surrounding the intestine also enter
the dorsal vessel ; this network in many cases may be replaced by
a blood sinus lying between the epithelial and the muscular walls
of the intestine. The blood in the vessels of the Chcetopoda is generally
red, and contains colourless corpuscles.

The following are brief descriptions of the. blood- vascular systems of an Oligochcetan
and of a Polychcetan (arbitrarily selected).

Lumbricus (Fig. 166) (as an example of the Oligochceta). — There are 5 longi-
tudinal vessels ; first a medio - dorsal vessel ; — second and third 2 medio - ventral
vessels, one of which lies under the intestine and above the ventral chord, and
represents the ventral vessel which is always found in the Chcetopoda, while the
other is much finer and runs under the ventral chord, the former is known as the
ventral vessel, and the latter as the sub-neural vessel ;— fourth and fifth, 2 delicate

VERMES— BLOOD-VASCULAR SYSTEM

251

vd

vessels which accompany the ventral chord throughout its whole length, running to
the right and left of it ; these are the lateral vessels of the ventral chord. The
sub-neural vessel is connected with the lateral vessels of the ventral chord at intervals
by transverse anastomoses. The dorsal
vessel, near the mid-gut, has segmental
swellings, so that it here assumes the
form of a string of beads. In the genital
segments it is connected with the ven-
tral vessel by 5 pairs of wider pouch-like
contractile vascular loops, the so-called
hearts. In the region of the mid-gut
the dorsal vessel gives rise in each seg-
ment to 3 pairs of vessels. The first
pair run laterally in the body cavity and
enter the sub -neural vessel. They give
off in their course vessels to the body
wall and the integument, and further,
in the lower part of their course, anasto-
moses to the ventral vessel and the
lateral vessels of the ventral chord. The
two posterior pairs run on the intestine
where they break up into an extremely
rich and close network. The typhlo-
solis and the muscular stomach are also
supplied from the dorsal vessel. An-
teriorly, between the third and fourth
pairs of hearts, there arises from the
dorsal vessel on each side a vascular
trunk, whose complicated branches sup-
ply the anterior part of the body, the
intestine, the body wall, the first pair
of Morren's glands, the pharynx, the
oesophagus, etc. ; these are also con-
nected with the ventral and sub -neural
vessels. In each segment the ventral
vessel gives rise to a lateral pair of ves-
sels, which branch in the body wall and
the integument. These branches anas-
tomose in that part of the body which
contains the stomach intestine with the
branches arising in each segment from
the first pair of lateral vessels of the
dorsal vessel, which latter maintain a
communication between the dorsal ves-
sel and the sub-neural vessel. The ventral and dorsal vessels give off branches to
the extreme anterior end of the body, which ramify in the body wall and pharynx.
The ventral vessel further divides at its anterior end into 2 branches, which pene-
trate to the brain and so form an oesophageal ring. The contractile part of the
blood-vascular system possesses muscular walls.

Nephthys scolopendroides (Fig. 167) may be taken as an example of a Polychcctan
with homonomous segmentation of the body. We find here the two most important
typical vessels, the dorsal and the ventral. Both lie close to the wall of the

FIG. 16*5.— Anterior portion of the body of
Lumbricus terrestris, opened, to show the vascu-

sperm sacs ; ph, pharynx ; aft, seminal vesicles ; vd,
dorsal vessel ; vv, ventral vessel ; vsn, subrieural

vessel; vln> lateral vessels of the ventral chord;
bun, ventral chord; h, contractile vascular loops

(hearts) between the dorsal and ventral vessels.

252

COMPARATIVE ANATOMY

CHAP.

intestine. From the posterior end of the pharynx to the end of the body the dorsal
vessel gives rise in each segment to a pair of lateral vessels, which run to the body
wall and especially to the dorsal branch of the parapodia and there ramify. Soon

after its rise out of the dorsal vessel
each lateral vessel gives off a branch
to the enteric wall, this branch there
forming a rich vascular network. At
the foremost end of the mid-gut the
dorsal vessel shows a slight swelling ;
it then leaves the intestine and runs
along the inner side of the body wall
in the region where the pharynx lies
to the extreme anterior end of the
body. It here divides into 2 branches,
which run backwards, to pass into the
anterior end of the ventral vessel at
the posterior end of the pharyngeal
region. The departure from the
general plan of the principal vessels
in the pharyngeal region is evidently
caused by the great development of
the pharynx. • If they lay as close to
the pharynx as they do to the intes-
tine in the rest of the body, they
would be broken by its protrusion.
The ventral vessel gives rise in each
segment to a pair of lateral vessels,
which run along to the base of the
ventral branches of the parapodia,
passing into the latter and there rami-
fying. A connection between the lateral branches of the dorsal vessel and the
lateral branches of the ventral vessel takes place in the following way. Between
the dorsal and ventral branches of a parapodium there runs on each side a vessel
which ramifies on one side in the dorsal branch, on the other in the ventral branch.
These ramifications of the connecting vessels anastomose with the ramifications of
the dorsal or ventral lateral vessels. On the ventral side of each dorsal parapodial
branch a sickle-shaped gill is inserted. Two vessels from the vascular system of the
dorsal parapodial branch enter the gill, and with many windings and anastomoses
run through its whole length. Besides the dorsal and ventral vessels, Nephthys
possesses also two extremely delicate lateral vessels of the ventral chord, which are
connected by fine branches with the lateral branches of the ventral vessel.

This arrangement of the blood- vascular system can, however, only to a very limited
extent be considered typical of the Polychceta. In the different families, indeed, very
great variations are to be found. The blood- vascular system is extraordinarily plastic.
It follows the smallest modifications in the structure and arrangement of the other
organs of the body. Its special form is above all dependent on the arrangement and
development of the gills. In the Terebellidce the dorsal vessel forms in the anterior
part of the body a tubular pulsating heart, which is narrowed anteriorly, and which
supplies the gills lying on the most anterior body segments with blood-vessels. At
its posterior end it splits into two branches, which encircle the intestine. These
branches unite below the intestine to form a sub-intestinal vessel which runs a
certain distance backwards over the ventral vessel. Then the sub-intestinal vessel

FIG. 167.— Part of a transverse section through
a segment of the body of Nephthys, diagrammatic,
to demonstrate the arrangement of the blood-vessels
(partly after J.aquet). dp, Dorsal ; vp, ventral para-
podium, with bundles of setae and supporting seta ; vd,
dorsal vessel ; vv, ventral vessel ; gin, lateral vessels
of the ventral chord ; vg, connecting vessel between
the dorsal and ventral parapodia ; d, intestine ; Inn,
ventral chord ; k, gills ; kg, branchial vessels.

iv VERMES— BLOOD-VASCULAR SYSTEM 253

again divides into two branches, which encircle the intestine and again unite in the
dorsal middle line to form a dorsal vessel running to the extreme posterior end of
the body. In the Cirratulidce (Chcetozone}, at the posterior region of the body where
there are no gills, a separate dorsal vessel and an enteric vascular plexus are both
wanting. They are here replaced by a blood sinus which continuously surrounds
the intestine inside its muscular wall. This sinus is continued anteriorly into a
strong pulsating dorsal vessel (heart) which runs through the whole branchial region
and gives off vessels to the gills directly or indirectly. In the Serpulidce the enteric
canal is generally embedded along its whole length in a blood sinus, and a dorsal
vessel is wanting. Lateral vessels frequently run along the sides of the intestine,
and are connected by segmentally arranged transverse loops with the ventral vessel,
and in the case of Serpulidce provided with a thoracic membrane these lateral
vessels give off branches segmentally to this membrane which break up in it into
extraordinarily numerous and fine ramifications. In the Serpulidce, where the gills
are developed exclusively in the head segment, each branchial filament is penetrated
by a single vessel, and the afferent and efferent blood passages first separate at the
bases of the gills. In nearly all other Polychccta we can distinguish afferent and
efferent vessels (branchial arteries and branchial veins) which pass into each other
at the ends of the branchial filaments. In the Capitellidce and Glyceridce a blood-
vascular system separate from the body cavity is wanting. The blood mingles with
the ccelomic fluid (hsemolymph). In Mastobranchus alone there are still found
rudiments of a' blood sinus surrounding the intestine. In many Polychccta there is
found, in that broader part of the dorsal vessel which is called the heart, a strand,
generally of a brown colour, lying freely in its lumen. This is called the heart body.
Its function is not yet clear. In Chcetozone 3 such bodies are found.

The blood-vascular system of the Echiuridce (Fig. 137, p. 207) is very simple. It
consists of a . ventral vessel running through the whole body and lying above the
ventral chord. Behind the mouth this vessel divides (like the ventral chord) into
2 branches which, embracing the mouth between them, pass along the 2 sides
of the prostomium to its extreme anterior end, where they unite (like the 2 limbs
of the oesophageal ring). A dorsal vessel arises from the point of junction,
which, running backwards, traverses the prostomium and then runs along the fore-gut
to the posterior end of the crop. Here it divides into 2 limbs, embracing the
intestine, and these enter the ventral vessel together. The dorsal vessel is thus con-
nected with the ventral vessel by 2 pairs of loops— one anterior, represented by the
lateral vessels of the prostomium, and one posterior, at the limit between the fore-gut
and the mid-gut. The dorsal vessel is not continued back over the fore-gut.

Prosopygla. — A vascular system is wanting in the Priapulidce
among the Sipunculacea, and in all Bryozoa. In Sipunculus there are
two vessels, one dorsal and the other ventral, which accompany the
fore-gut. Both end posteriorly near the place where the proboscis
retractors are attached to the body wall. Anteriorly they enter a
sinus which lies in front of the brain at the base of the tentacle crown,
encircles the oral cavity, and is in communication with the inner cavities
of the tentacles. By the contraction of the dorsal and ventral vessels
the fluid they contain is forced into the circular sinus, and from it into
the tentacles, which consequently extend and expand. The elements
which float in the coelomic fluid are met with in the vascular fluid,
so that an open communication between the body cavity and the
vascular system probably exists. In Phoronis there is a closed blood-

254 COMPARATIVE ANATOMY CHAP.

vascular system with red blood corpuscles. A dorsal blood-vessel
conducts the blood into a vessel which runs along the base of the
tentacle crown and gives off branches to the tentacles. Other
branches conduct the blood back into a second vascular ring which runs
on the outer side of the former. A vessel rises from each side of the
external vascular ring, which unites with that from the other side
under the oesophagus, and runs back as ventral vessel asymmetrically
in the left chamber of the body cavity; this vessel has -numerous
lateral caeca. Besides these vessels there is a blood sinus around the
stomach intestine. All the vessels are contractile.

The opinions of the most recent investigators as to the blood- vascular
system of the Bmcliiopoda vary greatly. Some deny the existence of
any circulatory apparatus. There is only, they say, a system of sinuses
belonging to the body cavity. Other investigations confirm the old view
according to which, in some Brachiopoda a contractile tubular heart lies
above the stomach, and a vein arising from the heart over the fore-gut.
There are also said to be vessels in the arms, and also vessels called
genital arteries.

A blood-vascular system is wanting in the Rotatoria, Dinophilus,
and the Chcetognatha.

XII. Genital Organs.

Division of the sexes generally prevails among the worms. The
exceptions to this rule, apart from single cases, are the Hirudinea,
Oligochceta, Myzostomidce, Chcetognatha, Phwonis, and many Bryozoa.

Nemertina. — The genital apparatus (Fig. 135, p. 205) is here very
simple. The ovaries in the female and the testes in the male are
present in large numbers, and are found in the shape of small sacs in
adult animals throughout that region of the body through which the
mid-gut runs. They lie in the parenchyma (jelly) under the mus-
culature. Each ovary and each testis, at the time of sexual maturity,
becomes directly connected with the exterior by means of a special
duct. The genital glands on each side generally lie in a longitudinal
row in such a way that between two consecutive diverticula of the
intestine there always lies an ovary or a testicle. They are therefore
more or less regularly metameric in their arrangement, in cor-
respondence with the more or less regular metameric arrangement
of the enteric diverticula themselves. In some Nemertina there are
also genital glands scattered about the parenchyma, each gland, however,
has an independent external aperture. In this arrangement we find
great agreement with the Turbellaria (especially the Polydada and
Triclada). In the latter, however, the oviducts and sperm ducts
arising from the genital glands unite to form common channels of exit.

Nemathelminths. — 1. Nematoda. — The male genital apparatus is
unpaired, and emerges at the posterior end of the body in the cloaca ;
the female apparatus is paired, and emerges externally on the ventral

IV

VERMES— GENITAL ORGANS

255

side in front of the anus, generally near the middle of the body. The
male genital apparatus (Fig. 168) is a single continuous tube,
lying in the body cavity, which falls into several divisions between
its blind inner end and its external aperture. The testicle division

FIG. 168.— Ascaris lumbricoides,
male genital apparatus, after Vogt
and Yung, sf, Lateral lines ; h,
testes ; d, intestine ; sb, sperm vesi-
cle ; Tel, cloaca ; de, ductus ejacula-
torius.

FIG. 169.— Ascaris lumbricoides,
female genital apparatus, after
Vogt and Yung, wo, Female aper-
ture ; v, vagina ; sf, lateral lines ; d,
intestine ; u, uterus ; o, ovaries ; ov,
oviducts.

forms a thin much twisted tube into which a solid axis (rachis)
projects from the blind end ; attached to this rachis the sperm cells
are found at different stages of development. The germ layer lies at
the blind solid end of the testicle tube. The nearer we approach
to the efferent duct the more advanced are the stages of develop-
ment and ripeness of the spermatozoa. The testicle division passes
into a shorter, wider, pouch -shaped division, the sperm vesicle,

256

COMPARATIVE ANATOMY

CHAP.

and this emerges by a short and narrow terminal piece, the duetus
ejaeulatorius, into the cloaca. Above the duetus ejaculatorius there
often lie two elongated sacs, invaginations of the cloaca. Each sac
contains a chitinous spiculum. In copulation the spicula may be
protruded from the cloacal aperture by special muscles attached to the
spicular sacs. In the female genital apparatus (Fig. 169) each of
the paired genital tubes repeats in essentials the divisions of the male
apparatus. The extraordinarily long ovarial tube lies in numerous
windings in the body cavity. In its axis there also lies a rachis, to
which the young eggs are attached. The ovarial tube is continued
into a wider part, the uterus, at whose commencement fertilisation
generally takes place, that is if the eggs have not already been fertilised
in that terminal portion of the ovarian tube into which the rachis does
not reach. The uterus contains fertilised eggs in the first stages of
their segmentation, and even young embryos. The two uteri unite at
their ends and pass into a short common terminal division, the vagina,
which is often muscular, and opens outwardly in the ventral middle
line through the female genital aperture.

2. Gordiidse. — The genital apparatus of the Gordiidce is quite
different from that of the Nematoda. The male apparatus also is
paired. The male and female genital apertures emerge into the last
division of the intestine (cloaca). The female genital glands (Fig.
170) are lobed ovaries which lie in large numbers in pairs one behind

the other on both sides of the
dorsal mesentery. They are
developed very late at the
expense of the mesodermal
cells, which in the young
Gordiidse almost entirely fill
the body cavity. Some of
the eggs which ripen in the
ovaries reach the body cavity,
entirely filling its lateral
chambers and so forming
those masses of eggs which
were formerly erroneously
considered to be ovaries. The
further fate of these eggs is
not known. Another portion
of the eggs, however, pass
out of the ovaries into two
tubes lined with epithelium which lie in the dorsal mesentery. These
tubes, which might be called uteri, run backwards, and when approach-
ing the posterior end of the body become narrow and bend round as
oviducts to the ventral side, where they enter a pear-shaped glandular
accessory organ of the small degenerating cloaca, the atrium. There
is further an unpaired vesicle, the receptaeulum se iis, placed under

^__iOT_^^^ md

BD

Fio. 170.— Transverse section through Gordius,
after Vejdovsky. Ih, Body cavity ; c, cuticle ; hy, hypo-
dermis ; u, uterus ; Im, longitudinal musculature ; et,
endothelium of the body cavity ; ov, ovarium ; mes,
mesenteries ; md, mid-gut ; bm, ventral chord.

VERMES— GENITAL ORGANS

257

li

the posterior part of the uterus ; this is connected with the atrium,
and during copulation becomes filled with spermatozoa. In the male
genital apparatus the actual testes have not yet been found. The
other parts correspond with the various divisions of the female appar-
atus. We can distinguish sperm sacs (corresponding with the uteri
of the female), sperm duets, and the well-developed cloaca, a flask-
shaped organ whose terminal part can be evaginated. There is no
organ corresponding with the receptaculum of the female.

3. Acanthoeephala ; Female Apparatus (Fig. 171). — In quite
young animals the ovaries are enclosed in the ligament by means of
which the genital apparatus is suspended
from the base of the proboscis sheath.
Masses of eggs (swimming ovaries) early u — I- j
reach the body cavity, probably by the
bursting of the ligament, and in sexually
mature animals are there present in great
numbers together with single detached eggs.
From the body cavity they are transmitted
to the exterior by a muscular apparatus
which is of very complicated structure,
although consisting of a limited, but for
each species definite and constant, number
of muscle cells. We distinguish in it first
an apparatus for swallowing the ova, the
uterus bell, to the bottom of which the
posterior, end of the ligament is attached.
It is in open communication with the body
cavity by means of a large anterior and
a small posterior aperture. It alternately
expands and contracts, and thus draws in
the eggs which float in the coelomic fluid.
From the uterus bell the eggs pass into
the anterior apertures of two short canals

which are called oviducts. The oviducts FlG 17L_Female genitai appar-
emerge into a tubular unpaired division, atus of an Echinornynchus, <iia-
the uterus, and this opens externally at grammatic. ug, uterus bell, vog,

, i • j £ , i v T i i . anterior aperture of the same : hog.

the posterior end of the body by a short posterior ditto; vod) anterior apertu£
terminal piece, the vagina. Through the of one of the two oviducts (<xi) ; u,
posterior aperture of the uterus bell the uterus ; v> vagina ; u'°> female aper'

1 , . , , , . . , ture ; o, ovaries ; U, ligament. The

Unripe eggS Which have been drawn in With arrows indicate the course taken by

the others are returned to the body cavity. the esgs in passing from the body

Male Apparatus (Fig. 172, p. 258).-

Two, or less frequently three, testes lie in the ligament. Each is
continued in thj form of a sperm duet. Each of these sperm ducts
has three pouch-like invaginations along its course (sperm vesicles).
Posteriorly they -unite to form one common muscular vas deferens,
which enters the ~^irsa at the point of a conical muscular projecting

VOL. I S

258

COMPARATIVE ANATOMY

CHAP,

penis. The bursa is a spacious sac-like invagination of the posterior
end of the body which can be evaginated in copulation. Three pairs

of cement glands are generally con-
nected with the male apparatus, their
ducts entering the terminal division of
the vas deferens (ductus ejaculatorius).
Annulata. — The Hirudinea are
hermaphrodite. Their male genital
apparatus (Fig. 173) has the following
general structure. Several testes lie
in pairs segmentally arranged in the
middle region of the body, generally
between the consecutive lateral diver-
ticula of the mid-gut, within the muscu-
lar septa separat-
ing these diverti-
cula • the testes
thus divide these
septa into anterior
and posterior lam-
ellae. A short effer-
ent duct arises
from each testis.
The ducts of all
the consecutive
testes of one side

emerge into a vas

FIG. 172.— Representation, partly dia- deferens which

grammatic, of the organisation of a male Qei

Echinorhynchus. r, Proboscis with barbed runs through the

hooks, protruded ; rm, retractor muscles of -whole length of

the proboscis; rs, proboscis sheath; rn, ,, fpqfl-pip rpo-irm

proboscis nerves ; g, cerebral ganglion ; re, tne testicle region,

retinacula; li, ligament; sn, longitudinal 111 front of the

nerves; h, testes; vd, vas deferens; *6, most anterior tes-
vesiculse seminales'; Ted, cement glands ; gg, . ,

common vas deferens ; gag, genital gan- tlS the tWO

glion ; bs, bursa ; mo, male genital aperture ; deferentia
Ih, body cavity ; p, penis ; I, lemnisci.

FIG. 173.— Genital or-
gans of Hirudo. p, Penis ;
mo, male ; wo, female geni-
tal aperture ; ov, ovaries ;
h, testicles; vd, vas de-
ferens.

vasa
con-
verge towards the

middle line, here to open outwardly by a common aperture, or in some
cases by a common unpaired copulatory apparatus (penis).

The number of testes may in some cases be greatly increased (Nephelis,
Lumbricobdella}, and this probably results from the subdivision of each testis into
several vesicles. Complications often occur in the last portion of the vasa deferentia.
For example, in Hirudo, the vas deferens, before entering the penis through a short
terminal passage, forms a tangled coil on each side. Glands are often connected with
the penis. The male genital aperture in Hirudo is in the 10th segment, between the
30th and 31st rings.

The female genital apparatus (Fig. 173) lies behind the male
aperture, and generally between the vasa deferentia. It consists of

iv VERMES— GENITAL ORGANS 259

two ovaries. From each ovary an oviduct arises, which, joining
that of the other side, passes into a sac -like muscular terminal
division, the vagina (Gnathobdellidce), or opens outwardly direct
(Rhynchobdellidce).

In the land leeches (Lumbricobdella, Cylicobdella] the oviducts remain separate
until they enter the vagina. The ovaries here occur in numbers as swellings of the
oviducts. In the Rhynchobdellidce the ovaries are elongated sacs ; in the aquatic
Gnathobdellidce they are short and lobed, and usually contained in a round sac. In
Hirudo the ducts unite, some distance before entering the vagina, to form a common
oviduct, into which numerous glands (albumen glands) enter. The female aperture
of Hirudo lies in the llth segment, between the 35th and 36th rings.

The arrangement of the hermaphrodite genital apparatus of the
Hirudinea vividly recalls the conditions which prevail in the Turbellaria
(Polydada and Triclada).

Oligoehseta. — The sexual, but more especially the transmitting,
organs of the Oligochceta are so completely different from those of the
Hirudinea, that it is not possible at present to refer the two to one
common type. Testes and ovaries are always very few in number ; the
former occur in one or two pairs, the latter in a single pair. The former
always lie in front of the latter. The collective male and female genital
apparatus occupies a limited number of segments in the anterior part of
the body. The segments lying between the 9th and 1 4th generally form
the genital zone. Less frequently (Aphanoneura, Chcetogastridce) the
genital organs lie farther to the front. While in the Hirudinea, Nemertina,
and Nematoda the egg and sperm passages are, as in the Platodes,
direct canal-like continuations of the ovaries and testes, they are, in
the Oligochceta, separated from the first from the germ glands, and show
much agreement with nephridia in their structure. They are there-
fore pretty generally regarded as nephridia which have assumed the
function of conducting the genital products out of the body. We must
not, nevertheless, forget that the sperm ducts and oviducts, even if
they really are transformed nephridia, must in the segments referred
to represent supernumerary nephridia, since, in the adult Lumbricidce
and in the young stages of other Oligochceta, besides the oviducts
and sperm ducts, typical nephridia occur in the genital segments
as well as in the rest of the body. In the male genital ap-
paratus we must distinguish three parts, viz. the testes, the sperm
sacs/ and the sperm ducts. In the lAimbricidce there are almost always
two pairs of testes, while the rest of the Oligochceta only possess one
pair. The testes seem everywhere to break up at an early stage into
the formative cells of the spermatozoa, so that in adult sexually
mature animals they are retained at the best as rudiments. The
sperm formative cells are early gathered into special sperm sacs, in
which they develop further and produce the ripe spermatozoa. These
sperm sacs, formerly regarded as testes, are large vesicles which
develop on the dissepiments of the testicle segments as sac -like

260

COMPARATIVE ANATOMY

CHAP.

outgrowths of their posterior lamellae. They are divided internally
by lamellae into numerous compartments and chambers, in which
the developing spermatozoa lie, and are in open communication
with the body cavity. The manner in which the sperm formative
cells pass out of the testes into the sperm sacs is not known.
The sperm sacs, even in nearly related genera and species, differ
considerably in their number and special arrangement. In some
Lumbricidce a middle unpaired portion is formed, a sperm capsule, and
the sperm sacs then appear merely as paired appendages to this
capsule. In the Chcetogastridce alone no sperm sacs are developed.

FIG. 174.— Lumbricus agricola. Genital organs, after Vogt and Yung. To the right the
sperm sacs and a part of the unpaired sperm capsule are removed, fern, Ventral chord ; sti, st«,
receptacula seminis ; ^, U, sperm funnels ; sbu, unpaired sperm capsule ; di, dissepiments cut off
at their bases ; vd, vas deferens ; to, funnels of the oviducts ; o, ovaries ; ov, oviduct ; dti, part of
the dissepiment between the 13th and 14th segments ; s&i, sb-2, s&3> paired sperm sacs ; fcj, ho,
testes ; VIII-XV, 8th to 15th segments.

The sperm formative cells in this family develop fully in the body
cavity, and the ripe spermatozoa are collected direct out of the coelomic
fluid by the funnels of the sperm ducts. Each sperm duet consists
of a preseptal funnel, and of a duct which penetrates the septum
and opens externally through a widened terminal division, the
atrium (arising by an invagination of the outer integument). Where
only one pair of testes is present there is generally only one pair of
sperm ducts. This is the case in the Naidomorpha, Chcetogastridce,
TuUficidce, and Enchitraeidce. In other Oligochceta, however, especially
those which live on land, and in most Lumbriculidce, there are two pairs
of sperm ducts. In this case all the four ducts are either entirely

iv VERMES— GENITAL ORGANS 261

distinct from each other, or else the anterior and posterior ducts on
each side enter a common atrium (many Lurnbriculidce), or the ducts of
each side unite to form a common duct, which opens outwardly without
the formation of an atrium (Lumbricidce). The sperm funnels lie in
the same segments as the testes. In those Lumbricidce which have a
median sperm capsule they lie in it.

Female Genital Apparatus. — This consists of the two ovaries,
the two oviducts and reeeptaeula seminis. The eggs either
ripen in the ovaries, or the ovaries fall into a few groups of
egg cells. Only one cell in each group then develops into an egg.
The eggs occasionally pass out of the ovaries into special egg sacs
corresponding to the sperm sacs, and pass out from these through
the oviducts. In many of the lower Oligochceta there are no
special oviducts. It is not really understood in what way the eggs
are here transmitted to the exterior. The reeeptaeula seminis are
paired sacs which open outwardly in special segments of the genital
zone. They arise as invaginations of the integument, and occur in
most Oligochceta in one pair, in the Lumbricidce, however, in two (less
often three) pairs. They are closed towards the body cavity, are in no
way connected with the rest of the genital apparatus, and are during
copulation filled with sperm from without.

Fig. 174, which depicts the genital apparatus of Lumbricus agricola, will help to
elucidate the above. In the 9th and 10th segments we see the 2 reeeptaeula seminis,
in the 10th and llth the sperm capsule divided by a transverse partition wall into an
anterior and a posterior part, with its 3 appendages (the sperm sacs) to the left ; to
the right the 3 sperm sacs and the cover of the sperm capsule are removed. We see
the anterior and posterior testes of the right side of the body, and further the 2 right-
hand sperm funnels and their ducts, which in the 12th segment unite to form the
unpaired vas deferens emerging in the loth segment. In the 13th segment the 2
ovaries lie at the 2 sides of the ventral chord. Behind these, on the anterior side of
the dissepiment between the 13th and 14th segments are the funnels of the oviducts
which open outwardly in the 14th segment.

Polyehseta. — The description here may be brief. With isolated
exceptions, the sexes are separate. The matrix from which,
generally only at certain periods, the ovaries or testes are de-
veloped, is the endothelium of the body cavity. The position of
the germ glands varies greatly. They are sometimes found on the
so-called genital plates, sometimes on the dissepiments, or on the
mesenteries, or they may be outgrowths of the endothelial covering of
the ventral vessel, etc. The ovaries or testes are generally repeated
in many, or at any rate in several, segments. Their form varies as
much as their position; they are sometimes cellular thickenings,
sometimes massive knobs, or tufts of strands, etc. The egg or sperm
cells sever themselves sooner or later from the ovaries and testes, and
ripen when floating freely in the coelomic fluid. From this they are
discharged through nephridia which are more or less strongly modified

262

COMPARATIVE ANATOMY

CHAP.

at the time of sexual maturity, or through nephridia which are
permanently transformed into genital tubes.

In Sternaspis, in the female one ovary is found, and in the male one testis. The
genital gland in both sexes has the form of a four-lobed pouch, which lies between
the loops of the intestinal tube, and passes into 2 efferent ducts opening outwards,
between the 7th and 8th segments. The genital products do not enter the body
cavity, but pass from the cavity of the genital gland to the exterior through the
efferent ducts. The question as to whether these ducts represent modified nephridia
must be decided by further research.

The Myzostomidce (Figs. 175 and 176) are hermaphrodite. Their
sexual apparatus does not easily admit of comparison with that
of the other Annulate,, but rather recalls in many points that of

FIG. 175.— Organisation of Myzostoma cirriferum, after v. Graft. To the left the parapodia
(p), the suckers (sri), and the male genital apparatus are represented. To the right the enteric
branches (da) and ovaries (o). c, Cirri ; php, pharyngeal tentacles ; ph, pharynx ; pht, pharyngeal
pouch ; m, stomach (mid-gut) ; ed, hind-gut ; u, uterus ; wo, female genital aperture, which enters
the cloaca; Mo, aperture of cloaca; pm, muscles for moving the parapodia; mo, male genital
aperture ; si, sperm vesicle ; vd, vas deferens ; h, testes ; p, parapodia with hooks and supporting
rod.

the Platodes. It is, however, probable that this is more a case
of analogy than of homology. In the sexually mature animals
the parenchyma is filled with numerous egg cells which lie in

VERMES— GENITAL ORGANS

263

masses between the branches of the intestine, especially on the dorsal
side. These masses of eggs are regarded as ovaries. It is, however,
possible /that they consist of eggs which have left the real ovaria.

genital

FIG. 176.— Section through Myzostoma, after v. Graff, ov, Ovaries ; da, intestinal branches ;
u, uterus ; dm, dorso-ventral muscle fibres ; sn, suckers ; li, testes ; qm, transverse muscles ; l>m,
ventral ganglionic masses ; pa, parenchyma ; md, intestine.

The origin and places of formation of the eggs are not known with any
certainty. The ripe eggs pass into a spacious uterus which lies on the
dorsal side of the stomach -intestine. One dorsal and two lateral
oviducts connect the uterus with the cloaca. The male
apparatus is paired. On each side
branched solid testicle strands
lie in the parenchyma below the
intestine and its branchings. On
each side an anterior and a pos-
terior vas deferens collect the
sperm from the testes. Both
vasa deferentia enter a muscular
lateral sperm vesicle, which lies
between the 3d and 4th suckers,
and which opens outwardly at the
edge of the body.

Prosopygia. — The places of
formation for the eggs and sperma-
tozoa of the Sipunculidce, Phoron-

idee, and Bracliiopoda are definite
points of the endothelium of the
body cavity. The sexual products

fall into the coelomic fluid, and

. , , , ,1 FIG. 177.— Diagrammatic representation of

are emptied OUt thence by the the organisation of a Brachiopod, from the dor-

nephridia, as Was the Case in the sal side ; the dorsal halves of shell and mantle are

Polychseta.

removed, a, Arms ; mh, mantle cavity ; m, mouth ;
I, liver ; ma, stomach ; n, nephridia ; an, anus ;
In the Sipunculidce the formation of 9, genital glands ; vs, part of the ventral shell

the germs takes place for the most part whic? proje(;ts ^ckwards over the dorsal shell;

, , v, anterior ; h, posterior ; r, right ; I, left,

at the base of the ventral proboscis

retractors. Phoronis is hermaphrodite. The male and female sexual products arise

264 COMPARATIVE ANATOMY CHAP.

on the asymmetrical ventral vessel. According to recent research it is probable that
in most, perhaps in all, Brachiopoda the sexes are separate. The germ glands are
outgrowths of the endothelium of the body cavity, and lie as branched or reticulated
strands, in the Tcsticardincs in pairs in the body cavity of both mantle folds, in the
Ecardines in the mantle folds and in the body cavity of the trunk, or in the latter
alone (Fig. 150, p. 226 ; Fig. 177, p. 263).

In the Priapulidce the anal tubes, which in early stages act as excre-
tory organs, become later the places of formation of the genital products.
They are then at the same time germ glands and efferent ducts.

In the Bryozoa there seems to be no definite rule ; separation
of the sexes is sometimes found, and at other times hermaphrodit-
ism. The ovaries and the testes arise, in the Edoproda, as cell out-
growths on the inside of the body wall or on the funiculus. The
first position is most characteristic of the ovaries, the second of the
testes. In the fresh-water Bryozoa, however (Fig. 139, p. 208), there
also arise on the funiculus the statoblasts, which are to be regarded as
parthenogenetic eggs. Special efferent ducts are wanting. The eggs
and spermatozoa fall into the body cavity. We do not yet know for
certain how they reach the exterior. Many observers maintain that
in hermaphrodite Bryozoa self -fertilisation takes place in the body
cavity. In many marine Edoprocta (Cliilostomata and some Cyclostomata)
the ripe (fertilised) eggs are taken up into special capsule-like foldings
of the body wall, the so-called ooeeia or ovieells, and these are
regarded as modified individuals which have arisen by gemmation.
This view does not as yet rest on sufficient foundation. The sexe&
seem to be separate in the Endoproda, but it may perhaps be the case
that the ovaries and testes do not develop at the same time. Two
testes lie between the stomach and body wall, and pass into 2
sperm ducts, which emerge into a sperm vesicle, the latter opening by
a pore into the vestibulum (the depression between the tentacles at
the bottom of which the mouth lies). The female genital apparatus
shows a similar arrangement : 2 ovaries, 2 oviducts, and an un-
paired terminal division which also opens into the vestibulum by a
pore. The genital apparatus of the Pterobranchia is not yet sufficiently
known ; under the 2 eyes of Cephalodiscus lie 2 ovaries.

Rotatoria (Fig. 161, p. 245). — Separation of the sexes here pre-
vails. The usually unpaired, seldom paired, female germ gland lies
near, generally below, the intestine, and consists of two parts — a
germarium, which yields the egg germs, and a vitellarium, which
richly supplies the young cells with yolk. The germ gland, which we
may call the germ-vitellarium, is surrounded by a membrane which
passes into a fine oviduct opening into the cloaca. The last part of
the oviduct, where the eggs often remain for some time, may be called
the uterus. In the male a testis with a vas deferens and protrusible
penis are found, which latter lies at the posterior end of the body ;
through it also the contractile terminal vesicles of the nephridia open
outwardly.

iv VERMES— GENITAL ORGANS 265

In Dinophilus (Fig. 162, p. 246) the sexes are separate. An
ovary lies on the outer surface of the intestinal wall, ventrally
on the boundary between the mid- gut and the hind-gut. It may
perhaps be developed out of an endothelium, which, however, has
not yet been proved to exist. The ripe eggs fall into the body
cavity and are emptied out through a pore which forms temporarily
in the body wall in front of the anus. The male sexual organs are
not yet sufficiently known.

In the male of D. apatris, at the posterior end, a conical organ
(penis) is found, which lies in a sac out of which it can be protruded
and into which it can be withdrawn.

The Chcetognatha (Fig. 152, p. 227) are hermaphrodite. The
ovaries are two long tubes which lie in the posterior part of the
two lateral chambers of the trunk cavity. Throughout their whole
length they are attached laterally by a mesentery to the body wall,
this mesentery enclosing them all round like a sac. On the outer side
of each ovary, and enclosed in its mesentery, lies a long oviduct,
ending blindly to the front, and opening outwardly behind near the
septum which separates the trunk and the tail cavities. It is not yet
known how the eggs reach the oviduct and are discharged. The two
testes lie as lateral cell thickenings or germ masses projecting into the
body cavity from the body wall of the tail cavity. Groups of
young sperm formative cells sever themselves and fall into the two
lateral chambers of the tail cavity. The ripe spermatozoa are dis-
charged through the sperm duets. In each sperm duct we can dis-
tinguish an inner ciliated funnel, a duct, and a vesicle ; these open
externally at the sides of the caudal segment. Thus in the arrange-
ment of the reproductive organs, especially in that of the male genital
apparatus, we find great similarity between these animals and the
Chcetopoda, Sipunculidce, Phoronis, and the Bracliiopoda.

Sexual Dimorphism in Worms. — Apart from the difference between the sexual
organs and outer copulatory organs, there are, in many worms in which the sexes
are separate, insignificant external differences between the male and the female.
In some worms, however, the differences in inner organisation cause a remarkable
sexual dimorphism in outer appearance and in size between the two. This is especi-
ally the case in the Rotatoria, Dinophilus, and Bonellia. It is always the male
which, in comparison with, the female, appears reduced and generally dwarfed. The-
Rotatorian male is smaller than the female, with degenerated enteric canal and
simplified wheel apparatus. The males are known only in the minority of genera
and species, and are much rarer than the females. In Seison alone the males and
females are alike. The fertilisation of the Rotatorian eggs has till now not been
observed. The females usually produce two sorts of eggs — delicate-skinned summer
eggs, and hard -shelled eggs which last through the winter. In Dinophilus
vorticoides the two sexes are not different ; in D. apatris the male is smaller, with
ciliated rings. Mouth, intestine, and anus are wanting. The males of Bonellia,.
which are ciliated all over, and are in appearance not unlike the Rhabdocaelan
Turbellaria, are minute in comparison with the females ; these males live parasitic -
ally on the female in varying numbers on the proboscis, in the oesophagus, or in the

266 COMPARATIVE ANATOMY CHAP.

nephridium which serves as uterus and oviduct. The intestinal canal is without
mouth or anus, a vascular system is wanting, nor have they the prostomium
which in the female is strongly developed as a proboscis. In short, apart from the
.sexual organs, they remain on the larval level. In some Myzostomidce there are so
called complementary males, which are considerably smaller than the hermaphrodite
individuals on which they live. Their organisation is like that of the hermaphro-
dites. As rudiments of ovaries and oviducts have been discovered in their bodies,
we can only consider these complementary males as originally hermaphrodites with
one-sided development of the male genital apparatus, or as young hermaphrodites
with the female genital apparatus not yet developed.

XIII. Parthenogenesis.

Reproduction by means of unfertilised eggs certainly takes place in the Rotatoria
and Bryozoa. It was formerly thought that only the summer eggs of the wheel
animalcules could develop without fertilisation, while the winter eggs must be fer-
tilised, but the act of fertilisation has not yet been observed in connection with these
latter. The statoblasts of the fresh -water Bryozoa are parthenogenetic eggs, and
such eggs are also found in some marine Bryozoa. These eggs are enclosed in hard
cases provided with many arrangements (air rings, processes, etc. ) serving to disperse
them in the air or water.

XIV. Asexual Reproduction by Gemmation and Fission.

Many worms, especially the Nemertina, Chcetopoda, Sipunculidce, Phoronis, and
the Bryozoa, are distinguished by a highly developed capacity of regeneration, which
is of the greatest use for the maintenance of the individual and of the race. Such
portions of the body as have been lost through adverse external circumstances, .
broken off, bitten off, etc., are quickly regenerated, even such as contain the most
important organs, e.g. the anterior part with the brain. Isolated broken off pieces
may occasionally be regenerated into whole animals. As already indicated, we may
perhaps some day be able to refer back the capacity shown throughout the animal
kingdom for asexual reproduction by gemmation and fission to such an accidental
multiplication by voluntary or enforced falling to pieces of the body with subsequent
regeneration.

Among the Vermes this form of reproduction occurs in the Polychceta, Oligochceta,
and Bryozoa.

Polychaeta. — One of the Capitellidce, Clistomastus, constricts off (most probably
periodically) the posterior part of the body, which contains the sexual products,
reforming it again by regeneration. In a Syllis, Haplosyllis spongicola, which lurks
in holes at the bottom of the sea, the parapodia and setae of a number of the
posterior segments become more strongly developed as sexual maturity approaches.
The group of segments thus modified, containing the sexual products, severs itself,
and swims about freely in the sea as a sexual swimming bud, dispersing the sexual
products. In other Syllidce (Syllis, Autolytus] at the anterior end of the swimming
bud a new head is formed with highly developed eyes, and this takes place even
before it is severed. The detached swimming bud then represents a complete
individual (person), in which the sexual products ripen. The individual from which
the swimming bud has severed itself forms no sexual products, but is able at its
posterior end to produce new swimming buds. The detached swimming buds or
sexual animals may, apart from the fact that they contain the sexual organs,
be distinguished from the mother animal by other, chiefly external, points of

iv VERMES— STOCK FORMATION 267

their organisation. We thus have before us an alternation of generations. A
mother animal which remains asexual produces asexually successive daughter
animals which differ from the mother animal externally ; these detach themselves and
reproduce sexually by means of fertilised eggs ; we thus have alternating asexual and
sexual generations. In Myrianida, (also a Syllis) new buds arise on the mother animal
even before the hindermost has detached itself. There thus arises a chain of buds of
which the hindermost is the oldest and the foremost the youngest. This is a case
of axial gemmation which is very similar to strobilation.

Oligochseta. — In autumn Lumbriculus falls into pieces which are all able to
regenerate into complete animals. In the genera ^Eolosoma and Ctenodrilus no sexual
organs and no sexual reproduction have as yet been observed, only asexual reproduc-
tion. In Ctenodrilus monostylos the body becomes constricted in the middle, and finally
separates into two pieces, each of which may again divide. These pieces regenerate
into normal animals after detachment. It is otherwise in Ct. pardalis. Here, in
each segment (except the foremost) behind the dissepiment of the preceding segment,
a budding zone appears, in which brain, oesophagus, etc., form. The development of
these budding zones takes place from before backward. The segments thus trans-
formed finally separate. In each of them the fore-gut and hind-gut become con-
nected with the mid -gut and the typical segmentation is developed, so that each
becomes a complete individual. In ^Eolosoma, as in the Syllidce, several posterior
segments are included in the first bud. While the head forms anteriorly in this bud,
it as well as the mother body increases in length, and the latter develops other
buds posteriorly before the first and oldest, i.e. the hindermost bud detaches
itself. The processes of gemmation are most complicated in Nais and Chwtogastcr,
since here in the (anterior) mother individual as well as in the daughter individual
new phenomena of gemmation appear before they become detached from each other.
Chains of several individuals varying in age and stage of development thus arise. The
age and degree of development may be given in a formula. A indicates the fore-
most and oldest individual, in which at first the daughter individual B appeared ;
then the bud C began to form in the individual B, and so on. An order of develop-
ment of buds which has been observed in Nais barbata (from before backward) is as
follows : A, F, D, £, E, C. Finally, the chain breaks up into its separate parts, which
no longer multiply asexually, but increase the number of their segments, and as
sexual individuals can develop sexual products. There is therefore here also a kind
of alternation of generations, since sexual and asexual reproduction mutually exclude
one another.

XV. Stock Formation.

The peculiar Syllis ramosa, which lives in deep-sea sponges, forms
by means of lateral gemmation much-branched stocks, in which, as in
most other Syllidce, special sexual individuals develop and detach
themselves. This is the only case of lateral gemmation in the Chcetopoda.
Among the Bryozoa, animal stocks of the most various shapes arise
by lateral gemmation. They are sometimes tree -like, sometimes
tufts, or they may be spread out like webs or crusts ; sometimes
many single animals rise from a creeping stem. We thus find repeated
in the Bryozoa the forms assumed by the Hydroida. In Loxosoma
alone the buds detach themselves, so that a permanent stock is never
formed. On the stocks of many Chilostoman Bryozoa peculiar
appendages, so-called vibraeularia and avieularia, are found. The

268 COMPARATIVE ANATOMY CHAP.

former are prominences on the wall of the ordinary individuals
(zooecia), each of which carries a long seta, by means of whose free
movement particles of food are brought within reach, and the water
surrounding the colony is kept in motion. The stalked avicularia
are catching apparati which hold small animals fast till they die. The
mechanism is similar to that of a crab's forceps or a bird's beak ; a
movable mandible is hinged upon an immovable beak, and is
worked by special muscles, so that the forceps can open and shut.
Avicularia and vibracularia are, like the above-mentioned ooecia and
ovicells, regarded as metamorphosed individuals without intestine.
This assumed polymorphism of Chilostoman stocks is, however, as yet
weakly supported. In marine Bryozoa all the individuals of a stock
are connected together by a network of nerve threads, forming what
has been called a colonial nervous system.

XVI. Ontogeny of the Worms.

The eggs of worms are either holoblastic alecitlial, or holoblastic telolecithal.
In the latter case the yolk may vary very much in quantity. The course of seg-
mentation and gastrulation varies in correspondence with the constitution of the egg.
There is sometimes (in the holoblastic alecitlial egg, example Sagitta) a total and
tolerably equal furrowing, forming a cceloblastuia, and then by invagination a ccelo-
gastrula. Sometimes the furrowing is more or less unequal, and often connected with
the formation of micromeres (cf. p. 124, segmentation of Bonellia). We find, always
according to the quantity of yolk stored in the egg, all the stages from a coeloblastula
to a sterroblastula, from an invagination to an epibole, and from a coelogastrula to a
sterrogastrula. Alecithal eggs or telolecithal eggs with little nutritive yolk, ccelo-
blastulae and coelogastrulae, are found in those groups of worms in which a free-
swimming and independently feeding larva develops very early. This is the case in
very many marine worms, especially in Neinertina, Polychceta, Sipunculidce, Bryozoa,
Phoronis, Brachiopoda, and Chcetognatha. Telolecithal eggs with much nutritive yolk,
sterroblastulse and sterrogastrulse, are found in those cases where the developing
animal only begins to move and to feed very late, i.e. in an almost adult condition,
and especially where a so-called direct embryonic development takes place, e.g.
Oligochceta, Hirudinea, Eotatoria.

Before passing on to describe the more important larval forms among the Vermes,
we will briefly describe the development of Eupomatus uncinatus (Serpulidae) (Fig. 178,
cf. also Fig. 92, p. 123). The blastula has a small blastoccel. The cells of the upper
(animal) half are smaller and more numerous than those of the lower (vegetative) half.
The former form the ectoderm, the latter the endoderm. At an early stage, at one side
of the blastula, which we may call the anal side, 2 round cells appear with remarkable
distinctness at the limit between the ectoderm and the endoderm. By the rise of
these primitive mesoderm cells the blastula becomes bilaterally symmetrical.
Besides an aboral or animal side and an oral or vegetative side, we can also distinguish
an anterior and posterior (where the 2 primitive mesoderm cells lie), and likewise
a right and a left, and a median plane. The two primitive mesoderm cells lie to the
right and left of the median plane. The vegetative or endodermal wall of the
blastula then becomes invaginated into the segmentation cavity to form the arch-
enteron while at the same time the ectoderm grows out over the invaginated part.
The process thus stands half way between invagination and epibole. The primi-
tive mesoderm cells, lying at the posterior edge of the blastopore, sink between

IV

VERMES— ONTOGENY OF THE WORMS

269

endoderm and ectoderm, i.e. deeper into the segmentation cavity. This gastrula
formation takes place in such a way that the blastopore is a median ventral
longitudinal slit. This closes from back to front, until anteriorly, i.e. excentrically,
only a small aperture remains. An equatorial ciliated ring, the preoral ciliated
ring, appears early in the gastrula larva. In the cell thickening which carries this
ring of cilia a circular nerve is developed. At the aboral pole the ectoderm thickens
to form the neural plate, which carries a tuft of cilia. Directly round the narrowed
blastopore the ectoderm becomes invaginated like a funnel, and forms the larval
oesophagus or the stomodseum, which gradually approaches the anterior ventral
edge of the preoral ciliated ring. The archenteron elongates downwards and back-
wards. Several smaller cells are severed by fission from the primitive or pole cells of
the mesoderm, and these are spread out in the segmentation cavity (primary body
cavity) and form various larval organs, e.g. muscle fibres and the larval head

FIG. ITS.— .4, B, C, Three stages of development of the larva (Trochophora) of Eupomatus,
from the side, m, Polar cells of the mesoderm ; md, mid-gut ; fh, segmentation cavity ; sp, neural
plate ; ivk, preoral ciliated ring ; st, stomodseum ; wki, postoral ciliated ring ; n, larval head nephri-
dium ; ot, otolith ; an, anus (after Hatschek).

nephridia. The primary body cavity lengthens. That half of the body which
lies behind and below the preoral ciliated ring assumes a conical form. The
point of the cone is the posterior end. From the posterior end to the mouth
the body becomes flattened. On the dorsal surface the posterior end of the
intestine opens by means of a small invaginatioii of the ectoderm, the proc-
todseum with an anus ; this occurs not far from the posterior end of the larva.
Diametrically opposite to this, at the neural area, which is surrounded by the preoral
ciliated ring, lies the neural plate, which represents a part at any rate of the rudiment
of the brain. A pigment spot (eye) arises asymmetrically in a cell of the neural
area. Behind the mouth a slighter postoral ciliated ring appears. The whole
ectoderm between the preoral and postoral ciliated rings is covered with short cilia,
and thus forms an adoral ciliated zone. A narrow medio- ventral ciliated band forms
from the mouth to the posterior end. Two auditory vesicles are developed out of two
ectodermal cells immediately behind the postoral ciliated ring, and sink beneath the
surface later. Two cell streaks or bands are developed from the two pole cells of

270

COMPARATIVE ANATOMY

CHAP.

the mesoblast which lie near the posterior end ; these lie close to the ectoderm
ventrally on each side, and develop anteriorly by constricting off smaller cells and by
the continuous division of the same ; they are the two xnesoderm streaks. The larva
has now reached what is known as the Trochophora stage. It swims about freely
by means of its ciliated rings. The hindermost smaller portion of the body with
the intestine, proctodseum, and the posterior part of the mesoderm streaks, represents
the rudiment of the afterwards segmented trunk plus the anal segment ; the whole
of the larger anterior portion contains the rudiment of the head or oral segment.

Larvre of the Trochophora type occur not only in worms, but are also common
among Molluscs.

In addition to the Trochophora the following are some of the most important
kinds of worm larvse.

The free-swimming larva of the Sipunculidce (Fig. 179) is already much further
developed when hatched than the Polychcetan Trochophora, with which, however,
it has much general similarity. In the Sipunculus larva the characteristic
preoral ciliated ring of the Trochopliora larva is wanting, but such a ring, weakly

md

sp

FIG. 179. —Larva of Sipunculus, after Hatschek. o, Mouth ; sp, neural-plate ; sto, stomodseum
wk, post-oral ciliated ring; rr, retractors of the anterior body (proboscis); md, mid-gut; n
nephridium ; an, anus.

developed, is said to occur in the Phascolosoma larva. A postoral ciliated ring, on
the other hand, is strongly developed. The intestine consists, as in the Trochophora
larva, of stomodseum, mid-gut, and proctodaeum. The latter is moved somewhat
from the posterior end on to the back, and the postanal part of the body growing
faster than the preanal part, the anterior position of the anus which is characteristic
of the adult animal comes about. The neural plate with 2, and later 4, larval
eye-spots is found in the same place as in the Polychcetan Trochophora. A larval
head kidney is not formed. The mesoderm, on the contrary, is far more developed
than in the Polychcetan Trochophora. The retractors of the anterior body (proboscis),
and the two trunk nephridia, have already begun to form. There is a spacious
body cavity, but this does not correspond with the primary body cavity of the
Trochophora, but rather with the secondary body cavity of the Annulata which
appears within the mesodermal streaks, so that under the integument and around
the intestine there is already a layer of mesoderm cells answering to the parietal
and visceral layers of the mespderm of the Annulata. While, however, in the
Annulata the mesodermal streaks become segmented and the body cavity is thereby
divided into consecutive pairs of chambers, in the Sipunculidce there is neither
segmentation of the mesoderm nor of the body cavity.

IV

VERMES— ONTOGENY OF THE WORMS

271

Bryozoa. — It is very difficult to describe the different, and for the most part still
insufficiently known, larval forms of the Bryozoa, and to establish their relations to
the adult animals. We must confine our-
selves to a description of the larva of
Pedicellina (Endoproda) (Fig. 180). A
ciliated ring on a circular elevation divides
the larval body into two regions. In
the oral region, which is ciliated all
over, lie the mouth and anus, the latter
on a conical prominence. Behind the
mouth lies a prominence with a tuft of
cilia. Between the mouth and anus there
is a depression, the vestibular pit. The
whole oral region can be withdrawn into
the aboral, so that the mouth and anus
come to lie at the base of a depression
whose edge is formed by the ciliated ring,
and which is called the vestibulum. In
the middle of the aboral region rises a
ciliated tuft. From the base of this ciliated
tuft an organ consisting of long ectoder-
mal cells, the so-called cement gland, pro-
jects into the interior of the larva. There
is also in the aboral region another organ
projecting inwards like a sac, into which
a short canal enters from outside, this is
the so-called dorsal organ. Both these

organs are said to disappear in later metamorphosis. In examining the inner organis-
ation we find a stomodseum, a sac-like mid-gut, and a hind-gut rising up to the anus.
The wall of the intestine turned towards the vestibulum is much thickened and is
called the liver. Muscles arranged in various ways serve for retracting the oral region.
Between the stomach -intestine and the body epithelium of the oral region lies a mass
of mesoderm cells, and connected with this on each side a small ciliated canal (nephri-
dia of the adult animal ?). If we wish to compare these Endoproctan larvae, which show
considerable resemblance with the Edoprodan larva of Meinbranipora known as Cypho-

FIG. 180.— Larva of Pedicellina (after Hats-
chek), from the side, o, Mouth ; 71, nephridium ;
wk, ciliated ring ; do, dorsal organ ; kd, " cement
gland" ; ws, ciliated tuft ; I, "liver"; m, meso-
derm cells ; pd, proctodseum ; an, anus ; vs,
vestibular pit ; st, stomodseum ; md, mid-gut.

FIG. 181.— A, B, C, D, Four stages of the metamorphosis of the attached larva of Pedicel-
lina. do, Dorsal organ; M, "cement gland"; st, stomodseum; md, mid -gut; pd, proctodteum ;:
v, vestibulum ; vs, vestibular pit ; t, rudiment of tentacles ; pe, peduncle. The arrows indicate
the direction from mouth to anus, ve, Invagination of the body wall towards the vestibulum
(after Barrois).

nautes, with a Polychcetan Trochophora, we must consider the ciliated ring as equiva-
lent to the preoral ciliated ring of the latter. Then the so-called cement gland

272

COMPARATIVE ANATOMY

CHAP.

with the ciliated tuft would answer in position to the neural plate of the Trocho-
phora. The further development of the Pedicellina larva involves a peculiar
metamorphosis (Fig. 181, A-D). The larva attaches itself by the oral region,
while at the same time the vestibulum closes by the growing together of its free
edges over the mouth and the anus. Thereupon the whole enteric canal, with the
altered vestibulum turns round, inside the sac-like ectoderm which encloses them,
so that later, reversing the larval order, the vestibulum, with the stomodsemn and
proctodseum entering it, is turned to the free end, i.e. the original aboral region
of the larva. The vestibulum becomes connected with the exterior by means of a
new invagination of the ectoderm ; at the point of junction the tentacles appear.
According to this ontogenetic observation the anus would lie, not dorsally, but
ventrally, behind the mouth, and so would the ganglionic mass, which would thus
not be homologous with the brain of other worms. Further investigations, however,

FIG. 182.— A, B, C, Three stages of the development of the larva of Phoronis (Actinotrocha),
from the side, sp, Neural plate; wk, ciliated organ; st, stomodaeum ; t, larval tentacles; ti,
definitive tentacles ; md, mid-gut ; ra, rudiment of the trunk (stalk), invaginated in the larva A, pro-
truded in B, developed into the trunk in C ; a, anus ; awk, anal ciliated ring ; hd, hind-gut (partly
after Metschnikoff).

especially as to the development of the nervous system, are needed to elucidate these
points.

The larva of Phoronis (Fig. 182) is known by the name of Actinotrocha. The
mouth and anus lie at opposite ends of the ciliated larval body. Over the mouth a
large prostommm hangs down, whose edge carries stronger cilia which probably
correspond with the preoral ciliated ring of the Trochophora. A larval ganglion
(neural plate) lies in the ectoderm of the prostomium, and is in one species provided
with 4 eye-spots. Behind the mouth lies a ring of larval tentacles, and im-
mediately behind this the rudiments of the definitive tentacles, at whose bases the
nerve ring of the adult Phoronis begins to form. Around the anus we find a strongly
ciliated ring. Behind the definitive tentacles on the ventral side lies the rudiment
of the trunk, invaginated into the larval body. The secondary body cavity is well
developed. In front of the invaginated rudiment of the trunk a nephridium like the

VERMES— ONTOGENY OF THE WORMS

273

head nepliridia of the TrotJiophora lies on each side. These become the permanent
nephridia of the adult Phoronis. The Adinotrocha thus formed sinks to the bottom
of the sea ; the invaginated trunk protrudes and grows quickly, the mid-gut at the
same time entering it and forming a loop with ascending and descending limbs.
The whole prostomium, with the neural plate and the larval tentacles, are thrown
off and devoured by the young Phoronis. Through all these processes the body
has approached the adult stage ; it is quite evident that by the protrusion and
rapid growth of the trunk, and the comparatively slight growth of the rest of the
original larval body, the anus conies to lie dorsally near the mouth. This process
readily allows of being referred back to the similar process in Sipunculus, only there
the rudiment of the trunk is never invaginated into the larval body.

Brachiopoda. — The free-swimming larva of Argiope (Fig. 183) consists of three
consecutive divisions, which are called the anterior, middle, and posterior segments.

FIG. 183.— ^1, free - swimming, li,
attached larva of Argiope, from
above (after Kowalewsky). vs, An-
terior segment ; ins, middle segment ;
As, posterior segment ; m, mantle ;
md, mid-gut ; ivk, ciliated organ.

FIG. 184.— Vertical median
longitudinal section through
an advanced embryo of Lin-
gula, after Brooks, m, Mantle
folds ; sd, dorsal, sv, ventral
shell ; t, tentacles ; rnh, man-
tle cavity ; Ih, body cavity ;
st, stomoda?um ; o, mouth ;
md, mid-gut ; hd, hind-gut ; m
(below), shell muscles.

The anterior segment is umbrella - shaped, and carries anteriorly 4 eyes. The
margin of the umbrella has longer cilia than the rest of the body. The middle
segment has dorsal and ventral folds directed posteriorly, and covering the posterior
segment. At each side of the free edge of the ventral fold two bundles of setae are
VOL. I T

( 'OMI'A RA Tl \ *E A X. 1 TOM Y

CHAP.

found. The mid-gut is the only part of the intestine which is developed. Between

tlic intestine and the outer integument of the larva are found the paired secondary
!>ody cavity, mesenteries, and muscles. The larva attaches itself by the end of the
posterior segment which grows out into a stalk. The two folds of the middle
segment bend forward like valves and form the mantle, the reduced anterior segment
coming to lie in the mantle cavity. The bundles of seta' are thrown oil'. The
stomodieum is formed by an imagination of the bodv wall of the anterior segment,
whose base breaks through into the anterior end of the mid-gut. It lies a little
below the eyes which afterwards degenerate. The rudiments of the nervous system
and of the nepliridia have not been observed. In Tercbratnlina tentacles develop
as buds on the circular edge of a disc which projects from the dorsal mantle fold.
The tentacles increase in number and are grouped in the shape of a horse-shoe. The
tentacular disc is then anteriorly prolonged into two processes, the arms, on which
the double row of tentacles become the arm cirri. AYe here recognise great agree-
ment with the Jjt-yo'.na and PJiOi'onis. Assuming that the point Avhere the eyes lie
on the umbrella-shaped anterior segment in the .Jirucltinjiod larva corresponds with
the neural plate of ^-Iffiimf rue/at, the agreement between their courses of development
is considerable. The posterior segment oi t\\& Bmcltiopod larva, perhaps answers to
the evaginating trunk part (stalk) in A<:tuiotrudia. The oral nerve lings in the
I!r(i<-lt!i>jivd<i and in P/toro/d* must lie homologous.

A five-swimming larva which occurs among \\\G 2?e inert ina is called Pilidium (Fig.

1 sr. ). It is helmet-shaped. AVe can dis-
tinguish in it a convex aboral region and
a somewhat concave oral region. The
ciliated ring is found on the boundary
between these two, which corresponds
with the edge of the helmet. Along the
base of the ciliated ring runs a nerve.
The ciliated ring, which answers to the
Trochophvran preoral ciliated ring, is
produced to the right and left into two
pendent lobes, which are comparable
with the ear-Haps of manv helmets. At
the highest point of the aboral region
is a depression ; in it the ectoderm is
thickened, and- carries a ciliated tuft.
I'm. is:,. -Pilidium- Larva ol a Nemertian T!lis answers in position to the neural
hum the side (after Salensky). >'/s Neural thick- plate of the T I'ocln^iJnu'd. The mouth
i-iiiiiH with ciliated tuft; icfr, ciliated riii- ; rn, lies in the middle of the oral region.
n.Tvcof the same; m, muscles; st, stomod*um ; lf ,(,1(]s j]lto t]u. stomodieum5 and this
u/'l, niid-L'ut ; c.s- ectoderm sacs. . ,., .-, i • i v

into a sac-like mid-gut winch lies ex-

eentricallv (behind). A proctod.enm is wanting, and does not attain development

during larval life. The space between the intestine and the ectoderm is a
>pacioiis segmentation cavity or primary body cavitv. In it lie muscle fibres, and
Lfeiiera'] v branched star-like mesoderm cells. The ectoderm of the oral region is
invaginated into the ]irimary body cavity at I ]ioints, forming 'J ] tail's of sacs. One
pair of these sacs lies in front of and the other behind the stomod;eum. The
further development of these sacs is as follows (Fig. 18(5, A, ]>, C, ]>}: They sever
themselves from the ectoderm of the larva. They then become connected in
pairs, then the, anterior fused pair unites with the posterior pair, so that now,
on the oral side of the Pil'«l iti.m, in its body cavity, a hollow plate arises with
inner and outer walls. The inner wall grows round the enteric canal on all sides,

IV

VERMES— ONTOGENY OF THE WORMS

275

and forms the permanent outer integument of the young Nemertian. The outer,
thinner, \vall then forms inside the Pilidium integument a sac -like covering for
the young Nemertian, the amnion. This integument with ciliated ridge and neural
plate, i.e. the primary ectoderm, falls away together with the amnion when the young
Nemertian issues from the Pilidium. At an early stage we find on the inner side
of the ectoderm sac a layer of cells, which in the Pilidium arise out of the ecto-
dermal wall of the sac itself, but in the related Dcsor's larva are said to be derived
from the mesoderm cells which lie in the primary body cavity: The 4 layers of cells
thus arising represent the rudiment of the definitive mesoderm. The central nervous
system arises out of 2 ectodermal outgrowths which unite to form the brain in the
young Nemertian. They grow out posteriorly into two strands, the lateral nerves.
The proboscidal apparatus rises firsi^?. out of an invagination of the ectoderm above
the brain, and secondly out of parts of the mesoderm which surround this in-
vagination.

The Pilidium larva takes up a position intermediate between the young Tur-
bellarian larva of the Polydada (cf. p. 167, Mailer's larva} and the typical Trockophoran
larva. It agrees with the former in the absence of a proctodseum. The 4

FIG. 186.— A, B, C, D, Four diagrammatic transverse sections behind the mouth through a
Pilidium larva during metamorphosis, to illustrate the method of formation of the mesoderm
and secondary (definitive) ectoderm, e, Larval ectoderm ; ei, definitive ectoderm ; a, amnion
ectoderm ; es, ectoderm sac ; m, mesoderm ; I, lateral lobes of the larva ; md, mid-gut ; ml, larval
mesoderm.

rudiments of the mesoderm probably answer to the 4 mesoderm masses of the
young Polyclad larvae or embryos.

A Nemertian larva related to the Pilidium, Desor's larva, shows the larval
characteristics less developed. The ciliated ring and the ectodermal thickening with
the ciliated tuft which corresponds with the neural plate are wanting. But the
definitive ectoderm is formed, as in Pilidium, of discs which detach themselves from the
primary ectoderm. Many Nemertina develop without metamorphosis.

If we glance over the larval forms of worms we see that all are distinguished
by peculiarities of structure which are explained by the circumstance that the
animals feed independently and swim about freely at a very early stage of develop-
ment. The simplest organs of locomotion which can be developed early are the
cilia. They appear in all larvae. Ciliated rings are universally present ; the most
constant is the preoral, which is provided with a special nervous system (nerve
ring). Almost everywhere we find a spacious larval body cavity filled with fluid.
The larvae are hydropic, their specific gravity being nearly that of water. They
are provided with a functional hollow enteron and other functional parts — nervous
system, sensory organs, muscles, excretory organs. It is pretty generally found
that those parts of the larval body which function at an early stage are thrown off
or reabsorbed at the end of larval life, and that the organs of the adult animal consist
of cell material which is present in the larva as undifferentiated germ material perform-

276 COMPARATIVE ANATOMY CHAP.

ing no specific functions. A comparison of the larva with the adult animal shows
that the body of the former corresponds with the anterior portion of the body of the
latter, i.e., in segmented animals with the head segment. The smaller portion of
the larva corresponds with the posterior end of the adult animal, while the trunk,
with the exception of the mid-gut, remains in its embryonic condition in the larva.
This is again to be explained by the fact that the embryo, which is provided with
little or no nutritive yolk, must early develop organs necessary for independent
feeding and locomotion, and these most indispensable organs lie chiefly at the
anterior end of the body. This also explains why the locomotory organs of the
larva, the ciliated rings, lie near the mouth, mostly somewhat in front of it ; and
why, besides this, there sometimes appears a preanal ciliated ring ; why also, in
certain (Polytrochati) Annelid larvae, other segmental ciliated rings appear as the
segments of the trunk develop.

Direct Embryonic Development. — This is chiefly found in fresh-water worms.
The embryo is provided with enough nutritive yolk, generally stored in the endoderm
cells, to enable it to develop direct (usually within an egg shell). Hence it follows
that the organs necessary for free independent locomotion and feeding as an embryo
are unnecessary. The comparison of the direct development of the embryo, say of
Lumbricus, with that of a pelagic larva, e.g. a Trodiopliora, is very instructive.

Fio. 187.— Embryo of Lumbricus (after Wilson). Optical median longitudinal section, e, Ecto-
derm ; pme, pole cells of the mesodenn ; m, mesoderraal streaks ; sh, rudiments of the segmental
body cavities in the mesodenn ; nb, neuroblast cells ; bm, rudiment of the ventral chord ; vm,
visceral layer ; pm, parietal layer of the mesodenn somites ; ug, rudiment of the infra-ossophageal
ganglion ; g, rudiment of the brain ; kh, head cavity ; o, mouth ; st, stomodseum ; md, mid-gut ; en,
endoderm.

The egg of Lumbricus is supplied with little yolk. The egg and the embryo
which develop out of it are nourished in another manner. In every capsule
there are several eggs in the midst of a mass of albumen which nourishes them.
The embryo (Fig. 187) is at a certain stage egg-shaped, and surrounded on all
sides by a thin unciliated ectodermal epithelium. The mouth lies anteriorly,
somewhat near the ventral side ; it is surrounded by an epithelial thickening, and
leads into a short stomodseum. This again opens into a very spacious mid-gut,
whose epithelial wall lies close to the body wall all round. On the ventral side
only, masses of cells, the two germ streaks, to be described later, are intercalated
between the intestine and the body epithelium. A proctodaeum is wanting, it

iv VERMES— ONTOGENY OF THE WORMS 111

develops only at a later stage. Anteriorly, above the mouth and directly under
the ectoderm from which it is derived, lies a mass of cells — the rudiment of the
brain. In many other worms with direct development the mid-gut is solid and
not hollow, in consequence of the large masses of yolk contained in the endoderm
cells (primitive macromeres).

Since embryonic head nephridia have been observed in some OligocJiceta, the
chief differences between them and the Polychcetan Trochophora are the following :
Cilia are wanting in the former, and especially a pre-oral ciliated ring. Sensory
organs (eyes, ciliated tufts) are wanting. A body cavity is wanting, and also a proc-
todreum. On the other hand, as we shall see later, the mesoderm is much more
developed than in the Polychcetan Trochopkora above described. The parts of the
body present in the embryo represent the rudiments of the definitive organs, they
do not, as in a free-swimming larva, fulfil specific functions, and therefore seldom
disappear in the course of further development. The Gnathobdellidce form an ex-
ception, for in them a metamorphosis takes place, and, according to recent research,
the larval integument with its muscular and nervous systems, and further the
provisional trunk nephridia and the oesophagus, are said to disappear.

Development of the Outer Integument. — We may say, generally, that the outer
integument or body epithelium of the worms is derived from the embryonic or larval
ectoderm. The ectoderm either changes, without suffering any loss worth mentioning,
into the body epithelium ; or some parts of the larval ectoderm which perform
specific functions are thrown off ; or else certain portions of the larval ectoderm sever
themselves, and unite on all sides of the body within the larval ectoderm to form
the permanent ectoderm or body integument (e.g. Sipunculidce, Nemertinct}. We
consider these processes as a kind of ecdysis. In the Gnathobdellidce (Hirudinca}
alone it is said that the secondary ectoderm is not formed from the larval
ectoderm, and that the latter is quite lost. The larval and definitive sensory
organs, the larval and definitive nervous systems, and the larval and definitive seti-
parous sacs (of the Chcetopoda) are products of differentiation of the ectoderm. The
last mentioned are groups of glandular hypodermis cells, which sink under the
integument, the setse arising in them as secretions. The ectodermal setiparous glands
are enveloped by mesodermal elements which supply their musculature.

Development of the Mesoderm and the Mesodermal Organs. — The mesoderm is
for us a topographical conception. All that lies in the adult animal between the
outer integument and the intestinal epithelium belongs to the mesoderm. The
relations of the various mesodermal organs and systems of organs to the outer
integument are very varied. These relations are closest in the nervous system in
consequence of its dependence on the (ectodermal) sensory organs. As already
described (cf. p. 223), in many worms the central nervous system remains in the
integument, even in the adult animal, and is thus not mesodermal. The mesodermal
position of the nervous system is, however, the general rule, as it is even as early as
in the Platodes. As, however, it nearly always develops entirely separate from the
rest of the mesoderm, we shall describe its development first.

The Brain. — In some cases the brain, or a part of it at any rate, develops out
of the ectodermal neural plate (many Annelids, Sipunculidce}. The elements of
the neural plate probably arise in connection with the provisional or definitive sensory
organs of the head (eyes, neural tuft, tentacles), though this cannot be established in
detail. There are thus different parts which unite together to constitute it a sensory
nervous centre. The neural plate must thus represent an organ similar to the sensory
body of the Ctenophora or the marginal centres of the Medusce. It is often thrown
off with the larval integument (e.g. in Phoronis, Pilidium), and the oral nerve rino-
or the brain arises anew out of the secondary ectoderm.

278

COMPARATIVE ANATOMY

CHAP.

npr

npr

lib

Ventral Chord of the Annulata.— The ventral chord seems always to begin to
form separately from the neural plate. It arises either as a continuous thickening
of the ectoderm in the Ventral middlp line, or as a pair of thickenings one on each side
of this middle line. The differentiation of the rudiment of the ventral chord into
the definitive ventral chord goes hand in hand with the development of the rest of
the trunk, and proceeds from before backward. It either remains, like the brain,
connected during life with the ectoderm, or it becomes constricted off from it and
takes up a position either in the musculature of the body wall or still deeper in the
body cavity. At the posterior end of the body it almost always retains its embry-
onic condition, as it here remains throughout life in its place of formation, the
integument. The Hirudinea and Lumbricus among the Oligochceta differ very
much from other Annulata and Sipunculidce. The ventral chord here does not arise
in situ in the ectodermal integument ; but two ectodermal segmentation spheres
(micromeres) appear veiy early near the posterior end, and take up a position under
the ectoderm, lying symmetrically on each side near the middle line. New cells are

continually constricted off anteriorly from
these neuroblasts (Fig. 188, nb), which
again divide, and a cell strand thus arises
on each side of the ventral middle line,
immediately beneath the integument. The
two cell strands, which form part of the
germ streaks of the Hirudinea and of Lum-
bricus, represent the rudiment of the ventral
chord, which, beginning behind the mouth,
becomes continuously differentiated from
before backward. It is evident from this
that the rudiment of the ventral chord is
unusually localised, and at the same time
is to be referred to a very early stage of
development. The connection of the ventral
chord with the brain through the cesophageal
commissure seems everywhere to take place
secondarily. It may perhaps in time be
proved that the central nervous system in
the Worms and Platodes proceeds onto-

streaks in a Lumbricus embryo (after Wil- genetically and phylogenetically from two
son), pme, Pole cells of the mesoderm (ineso- cMef parts> viz< first from the sensory part
blasts); n* pole cells of the ventral chord ^ ^ united

(neuroblasts) ; npb, pole cells of the nephnclial

rows (nephroblasts) ; x, pole cells of the cell centres or sensory ganglia of the anterior
rows (xr) of unknown significance ; ec, ecto- end of the body, and second from the motor
derm ; ez, large ectoderm cells ; npr, nephridial central nervous system, i.e. the ventral
cell rows ; nr, neural cell rows ; m, mesoblast chordj ^ 3^^^ commissure, and the

motor part of the brain of the Annulata,

the longitudinal trunks and the motor part of the brain of the other worms and
Platodes. In the Nemertina, however, the lateral nerves are said to grow out from
the brain posteriorly. This may perhaps here, and also in the Turbellaria, point
to a concentration of the whole central nervous system into one single rudiment.

In the development of the other component parts of the mesoderm, we find that,
just as the rudiment of the ventral chord in the Hirudinea and Lumbricus is shifted
back to an early stage, and is condensed into two germ cells (the neuroblasts), so the
rudiment of all other mesodermal organs in worms are extremely condensed and
localised and shifted back to early stages, so that generally a few germ cells or a

FIG. 188.— Superficial aspect of the germ

IV

VERMES— ONTOGENY OF THE WORMS

279

limited germ zone represents the condensed rudiment of all mesodermal organs, with
the exception of the nervous system.

We accordingly find in the Hirudinea (Cl-epsine) and in Lumbricus, in early
stages of development, even during
segmentation, on each side of what
morphologically corresponds with
the posterior end of the embryo, 4
or 5 micromeres, which soon sink
down under the ectoderm - micro -
meres, or are grown round by them.
The 2 which lie posteriorly are the
largest ; they lie close to each other,
and in front of them 3 or 4 lie on
each side of the ventral median line
(Fig. 188). The 2 inner micromeres
we already know ; they are the
neuroblasts. Just as 2 rows of cells
develop anteriorly from the neuro-
blasts, forming the rudiment of the
ventral chord, rows of cells also
develop anteriorly from the remain-
ing polar cells. The single or double
cell rows which lie near the neural
rows are the nephridial rows ; the
polar cells from which they develop
are the nephroblasts. They yield
the material for the nephridia, which
become differentiated from before
backward. The most anterior ne-
phridia are temporary ; they are the
larval or embryonic nephridia. It
is not yet known what part of the
mesoderm is formed by the lateral
rows of cells with their posterior
polar cells. The cell rows which
proceed from the most posterior lar-
gest polar cells represent the rudi-
ments of the body musculature, the
endothelium, septa, and mesenteries.
We shall return to their further
development.

All these cell rows taken together
are known as the paired germ streaks.
They lie at the two sides of the ven-
tral middle line between the intestine
and the integument. The germ

pme

FIG. 189.— Surface view of the germ streaks of a
somewhat older Lumbricus-embryo, after the disap-
pearance of the anterior polar cells (after Wilson).
pme, Polar cells of the mesoderm (m) ; x, xr, cell rows
of unknown significance ; nphr, nephridial cell streaks ;
streaks thus yield all the mesodermal nbr, neural cell streaks, rudiment of the ventral chord
organs, nervous system, nephridia, (bm) ', d^ dissepiments ; ib, inner ; ab, outer rows of
muscles, endothelium, etc. In the setiParous S^ds; nph, rudiments of the nephridia;
.. T 7 7 *, . 7 i .1 11 mu> longitudinal musculature.

Ghiathobdefiidce, where the whole

larval ectoderm is lost, the germ streaks are said even to produce the definitive
body epithelium as well. This statement, however, requires confirmation.

280 COMPARATIVE ANATOMY CHAP.

There are germ streaks in many worms similar to those in the Hlrudinea, and
Lumbricus. But it appears that these do not consist of contiguous rows of cells, but
that each germ streak is the product of a single polar cell at the end of the embryo
or young larva. We have described these two polar cells in Eupomatus as primitive
mesoderm cells. The germ streaks which proceed from them are called mesoderm
streaks ; they seem to yield the whole mesoderm with the exception of the nervous
system. Here, therefore, the rudiments of almost all the mesodermal organs are
localised and condensed into two blastomeres wrhich appear at an early stage at the
posterior edge of the blastopore.

In the larva of Lopadorhynchus (Polychwta), as the trunk grows, what is called the
postoral ventral plate of the ectoderm is said to split from front to back into an
inner muscular plate and an outer neural plate, the ventral chord being principally
formed from the latter.

Further Development of the Germ Streaks (Figs. 187-189). — In Lumbricus
the two cell rows of the germ streaks, which are formed from the two large pos-
terior polar cells, soon begin to be differentiated from before backward. Their
cells divide. The simple cell rows thus become solid plates or strands. They
extend forwards on both sides and meet dorsally above the mouth. In this
cephalic portion of the germ streaks a central cavity appears which enlarges and
becomes the cavity of the head. The outer layer of cells attaches itself to the outer
integument of the head, and becomes the outer musculature and endothelium of the
head segment. The inner layer forms the musculature of the pharynx and its
endothelial covering. Behind the head on each side clefts also occur in the above-
mentioned cell strands, and these, forming in segmental order from before backward
and increasing in size, form the rudiments of the paired segmental chambers
of the coelome. They separate the cell strands in each segment into a parietal
layer contiguous to the integument and a visceral layer contiguous to the intestine.
The cell strands are thus divided into paired segmentally consecutive portions.
In each segment on each side both the parietal and the visceral layers grow up
between integument and intestine till they meet in the dorsal middle line. The
partition wall which is here formed in this way, and which divides the two lateral
chambers of the body cavity, is the rudiment of the dorsal mesentery which is often
temporary. A ventral mesentery arises in a similar manner. The cell material,
which separates the consecutive chambers of the body cavity forms the dissepiments.
The parietal layer forms the musculature of the body wall and the parietal
endothelium ; the visceral layer forms the muscular layer of the intestine and the
visceral endothelium.

Development of the Blood-vascular System. — This system arises (Tercbella,
Psygmobranchus] as a cavity filled with fluid between the epithelial intestinal wall
and the contiguous visceral layer of the mesoderm. The chief vessels arise as longi-
tudinal bulgings of the mesodermal walls of this enteric sinus, which finally become
constricted off from it, passing from the form of grooves to that of closed canals. In
Lumbricus and other Oligochceta the dorsal .vessels arise in a similar way as clefts
between the enteric epithelium and the enteric muscle layer. They are at first
paired, but generally unite to form the unpaired dorsal vessel. In a few Lumbricidce,
however, even in adult animals, over a larger or smaller region of the body, they
may remain double.

Development of the Nephridia. — In the anatomical section we distinguished
three parts in each nephridium : (1) the funnel ; (2) the nephridial duct ; ^and (3) the
terminal portion which opens outwardly, which in the Hirudinea and Oligochceta is
often widened out into a vesicle. It appears that in Lumbricus the nephridia
develop in the following way. The nephridial ducts develop in pairs in each

IV

VEEMES— ONTOGENY OF THE WORMS

281

segment as outgrowths of the nephridial cell rows. Each duct consists at first of one
cell or of a few cells, and later of a row of cells bent in the shape of the letter U,
which projects into the body cavity and is thus provided with an outer endothelium.
One limb of the cell row remains in contact with the integument, the end of the
other attaches itself to the posterior wall of the dissepiment which lies in front of it.
At this point an inner canal first appears in the solid row of cells. The terminal
portion arises by an invagination of the integument. The funnel begins to form
from one cell on the anterior wall of the dissepiment at the point where, on the
posterior wall, the end of the nephridial duct lies. This cell only secondarily
becomes a hollow ciliated funnel, which then unites with the nephridial canal
through the dissepiment. The funnel thus arises separately from the nephridial
canal out of the epithelium of the body cavity, and not out of the nephridial
rows of cells. In the Polychceta also the funnel and the nephridial duct of each
nephridium are said to arise separately.

Development of the Sexual Glands. — It may be considered certain that in the
Annulata and Prosopygia the ovaries and testes are developed from special parts of
the endothelium of the body cavity.

The Development of the Mesoderm in the Chsetognatha (Fig. 190).— The nervous
system here lies in the integument of the body and does not belong to the mesoderm.

FIG. 190.— A, B, C, Three early stages of development of Sagitta (after O. Hertwig). a, Gas-
trula ; U, blastopore ; ud, arch-enteron ; g, primitive cells of the sexual organs ; vm, visceral layer ;
pm, parietal layer of the mesoderm ; d, rudiment of mid-gut ; cs, ccelome sacs ; st, stomodteum ; d,
intestine.

In this case the mesoderm develops in a manner different from that in the worms as
yet described. A ccelogastrula forms whose principal axis answers pretty accur-
ately to the longitudinal axis of the adult Sagitta. The aboral pole of the gastrula
corresponds with the future anterior end of the body. Two large cells which soon
divide enter the base of the archenteron from the endoderm at an early stage.
These 4 cells are the rudiment of the testes and ovaries. Then on each side
there arises out of the base of the archenteron a fold of the endoderm, which grows
into the archenteric cavity towards the blastopore. These 2 folds divide the arch-
enteric cavity into a central cavity and two lateral cavities, which communicate at
the free edges of the folds. The central -cavity is the definitive enteric cavity ; its
epithelial walls, i.e. the inner epithelial lamellae of the folds, represent the rudiment
of the definitive enteric epithelium. The folds close dorsally and ventrally to form
the enteric tube, the latter carrying the 4 sexual cells at its freely projecting end.
The two lateral cavities, which may almost be looked upon as 2 sac-like invaginations
of the archenteron (ccelomic sacs), form the commencement of the body cavity.
Each sac has an outer epithelial wall in contact with the ectoderm, and an inner
wall in contact with the enteric tube. The former is the parietal, the later the
visceral layer of the mesoderm. The former probably forms the musculature and
the endothelium of the body wall, the latter the enteric endothelium. At the aboral

282 COMPARATIVE ANATOMY CHAP.

pole of the larva a small depression of the ectoderm becomes connected with the
enteric tube. The permanent mouth and the stomodteum thus arise. The primitive
mouth closes. The enteric tube becomes a solid strand, which continues to grow till
it reaches the closed primitive mouth, i.e. the posterior end of the body. An enteric
cavity does not again appear till a later stage. On the ventral side, a little in
front of the posterior end of the body, an anus forms. The postanal intestine
degenerates, its 2 visceral layers of the mesoderm forming the septum of the caudal
segment. From the 4 genital cells the testes and ovaries develop. We thus see
that in Sagitta the mesodermal organs have a double origin : first, 2 (afterwards 4)
large endoderm cells lying at the base of the archenteron, which form the rudiments
of the sexual glands ; secondly, the whole middle and oral epithelial wall of the
archenteron of the ccelogastrula ; this represents a germ zone, continued at the
edge of the blastopore into the ectoderm, and at the aboral portion of the larva
into that part of the wall of the archenteron from which the definitive intestine is
produced. This germ zone forms at an early stage two lateral cceloine sacs as a
consequence of the formation of the folds above mentioned ; the cavities of these
sacs produce the body cavity, and their walls the endothelium of the body cavity,
and probably also the musculature of the body wall.

Two hollow lateral coelome sacs of the archenteron appear at an early stage in
one of the Brachiopoda, Argiope, much in the same way as do those in Sagitta ; these,
constricting themselves off from the intestine, are said to form the body cavity and
the mesoderm. The development of the mesodermal organs has, however, up to the
present time been insufficiently observed.

Various theories have been put forward as to the phylogenetic significance of the
different processes of development to which the mesodermal organs owe their rise
ontogenetically, all these theories resting upon the assumption that the ontogenetic
process exactly repeats, sometimes in one point sometimes in another, the phylo-
genetic development. These theories rest upon weak foundations as long as com-
parative anatomy knows of no series of animal forms which shows us the gradual rise
of the mesodermal organs in a manner similar to that seen in the successive onto-
genetic stages of development. It is very doubtful whether the whole mesoderm,
except the nervous system, can be derived phylogenetically from such simple organs
as the ccelome sacs, or from cell groups such as the polar cells of the mesoblast,
once present in simple gastrula-like racial forms. Observations are increasing in
number which tend to show that ontogenetically also there is no single rudiment of
the whole mesoderm, but rather several rudiments for the different mesodermal
organs. Our review of the history of development of the worms supports this latter
view. It may perhaps in time be established that the manner in which the various
mesodermal organs appear in the Cnidaria moving from their places of formation, the
ectodermal body epithelium and the endodermal enteric epithelium, into the deeper
parts of the body wall, is essentially the same as that in which the mesodermal organs
originally arose in the ancestors of the Platodes and the Vermes. The ontogenetic
development of the mesoderm would then represent this process very much abbreviated
and localised, pushed back to very early stages. If, in the Hirudinea (in Clepsine at
least) and in Lumbricus as opposed to the other Annulata, the ventral chord does not
arise in situ in the ectoderm but is formed by two blastomeres, the neuroblasts, which
arise at an early stage, it is difficult to see why the polar cells of the other mesodermal
organs (nephroblasts, mesoblasts, etc.) should not represent similar early developing
condensed and localised rudiments. And why should not these different rudiments
themselves be pushed back to, and localised and condensed in, a rudiment such
as the early developed primitive mesoderm cells or zones 1 In the Polyclada we
see at an 8-micromere stage (Fig. 94, p. 125) the rudiment of the whole ectoderm,

iv VERMES— LITERATURE 283

with the nervous system localised and condensed in 4 micromeres, the rudiment of
the mesoderm in 4 other micromeres, and the rudiment of the whole endodermal
enteric system in the 4 macromeres.

Development of the Intestine. — The intestine of the worms consists as a rule of
three parts of different origin — the fore-gut, the mid-gut, and the hind-gut. From
the" endodermal archenteron of the gastrula only the mid-gut is derived. At an
early stage in larval life an anterior ectodermal invagination, the embryonic or larval
oesophagus or the stomodieum, becomes connected with it, while a similar invagina-
tion at the posterior end yields the proctodseum. "While the proctodaeum becomes in
a direct manner the hind-gut, which is often very short, the fore-gut does not always
proceed direct from the stomodseum. Sometimes, e.g. in the Hirudinea, the
stomodaeum disappears and a new oesophagus arises independently in its place.
Sometimes the definitive oesophagus begins to form out of the stomodaeum which
as such disappears. This is the case in many Polychceta. The musculature
of the fore-gut, which is often very strongly developed as the pharynx, seems every-
where to be formed from the cephalic portion of the mesodermal or germ streaks.
The origin of the muscular wall of the mid-gut and its endothelium, where this is
present, from the visceral layer of the mesodermal streaks has already been described.

Literature.

Nemertina.

W. C. M'Intosh. A Monograph of the British Annelids. Part I. Nemerteans.

London, 1872-74.
J. v. Kennel. Beitrdge zur Kenntniss der Nemertinen. Arbeiten d. zool. Instituts in

Wurzburg. Bd. IV. 1878.
A. A. W. Hubrecht. Report on the Nemertea, in Report on the Scientific Results of

the Voyage of H. M.S. Challenger. Zoology. Vol. XIX. Part LI V. London,

1887.

Treatises by Quatrefages, Hubrecht, Oudemans, Moseley, L. v. Graff, Dewoletzky.
Ontogeny : Hubrecht, Salensky, Gotte, Barrois, etc.

Nematoda.

A. Schneider. Monographic der Neinatoden. Berlin, 1866.
R. Leuckart. Die Parasiten des Menschen. New edition in preparation.
Works and treatises by Meissner, Eberth, Biitschli, Schneider, Leuckart, Glaus,

E. van Beneden, de Man, Rohde, Schnlthess, etc.
Ontogeny : Biitschli, Gotte, Hallez, etc.

Gordiidse.

F. Vejdovsky. Zur Morphologic der Gordiiden. Zeitschr. f. w. Zool. Bd. XLIII.

1866.
The same. Studien iiber Gordiiden. Ibid. Bd. XL VI. 1888. With Bibliography.

Acanthocephala.

Carl Baltzer. Zur Kenntniss der Echinorhynchen, in Arch. f. Naturgeschichte.

1880.
A. Safftigen. Zur Organisation der Echinorhynchen, in Morphol. Jahrbuch von

Gegenbaur. 10 Bd. 1885.
Treatises by Greef, Schneider, Andres, etc.

S+

284 COMPARATIVE ANATOMY CHAP.

Hirudinea.

A. Moquin-Tandon. Monographic de la famille des Hirudinees. With Atlas. 2d

edition. Paris, 1846.

E. Leuckart. Die Parasiten des Menschen. New Edition in preparation.
A. Gibbs Bourne. Contributions to the Anatomy of the Hirudinea. Quart. Journ.

Microsc. Soc. Vol. XXIV. 1884.
St. Apathy. Analyse der Aussern Korperform der Hirudineen. Mitth. Zool.

Station Neapel. Bd. VIII. 1888.
Treatises by Brandt and Ratzeburg (Medicin. Zoologie, 1829), Leydig, 0. Schtiltze,

Hermann, Whitman, Jaquet, v. Kennel.
Ontogeny : Rathke, Robin, Biitschli, Bergh, Whitman.

Chaetopoda.

Franz Vejdovsky. System und Morphologic der Oligochaeten. Prague, 1884.
Ed. Claparede. Recherches anatomiques sur les Oligochetes. Geneva, 1862.
The same. Les Ann6lides Chetopodes du golfe de Naples. Geneva, 1868. Supple-
ment, 1870.

The same. Recherches sur la structure des Annelides sedentaires. Geneva, 1873.
Ernst Ehlers. Die Borstenwiirmer (Annelida Chaetopoda}. Leipzig, 1864-68.

A. de Quatrefages. Histoire naturelle des Anneles. With Atlas. Paris, 1865.

C. Semper. Die Vemvandtschaftsbeziehungen der gegliederten Thiere. Arbeiten aus
dem zool. Institute in Wurzburg. Bd. III. 1876.

B. Hatschek. Studien uber Entivicklungsgeschichte der Anneliden. Arbeiten des

zool. Instituts zu Wien. Bd. I. 1878.
J. Fraipont. Le genre Polygordius, in Fauna und Flora des Golfes von Neapel.

XIV. Berlin, 1887.
H. Eisig. Die Capitelliden des Golfes von Neapel, in Fauna und Flora des Golfes

von Neapel. XVI. Berlin, 1887.
E. Meyer. Studien uber den Korperbau der Anneliden. I. Mitth. a. d. zool.

Station zu Neapel. 7 Bd. 1887.
H. de Lacaze-Duthiers. Recherches sur la Bonellie. Annales d. Sciences natur.

1858.
J. W. Spengel. Beitrdge zur Kenntniss der Gephyreen (i.e. the Echiuridce). I.

in Mitth. a. d. zool. Station zu Neapel. 1879. II. in Zeitschr. f. iviss. Zoologie.

Bd. XXXIV. 1880.
R. Greef. Die Echiuren (Gephyrea Armata), in Nova acta. Bd. XLI. Halle,

1879.
Numerous treatises and works by d'Udekem, Audouin et Milne-Edwards, Hering,

Gegenbaur, Claparede, Grube, Ratzel, Leydig, Perrier, Vejdovsky, Tide, Jaquet,

A. Agassiz, Greef, Spengel, Eisig, E. Meyer, Pruvost, Vignal, Kiikenthal,

Ehlers, Hatschek, v. Graff, Riesch, Foettinger, v. Kennel, Billow, and others.
Ontogeny : Loven, Krohn, Claparede, Metschnikoff, A. Agassiz, Salensky, Kowal-

evski, Kleinenberg, Vejdovsk^, Wilson, Hatschek, Gotte, and others.

Myzostomidse.

L. v. Graff. Das Genus Myzostoma. Leipzig, 1877.

J. Beard. On the Life-history and Development of tJie Genus Myzostoma. Mitth. zool.

Station Neapel. 5th vol. 1884.
Fridtjof Nansen. Bidrag til Myzostomernes Anatomi og Histologi. With English

resuirM. Bergen, 1885. With bibliography.

iv VERMES— LITERATURE 285

Sipunculacea.

W. Keferstein. Beit-rage zur anatomischen und systematischen Kenntniss der Sipun-

culiden. Zeitschr. f. iv. Zool. Bd. XV. 1865.

E. Selenka (with de Man and C. Billow). Die Sipunculiden. Wiesbaden, 1883.
J. Andreae. Beitrage zur Anatomic und Histologie des Sipunculus nudus. Zeitschr.

f. w. Zool. 36. Bd. 1881.
W. Apel. Beitrag zur Anatomie und Histologie des Priapulus caudatus (Lam. ) und

des Halicryptus spinulosus (v. Sieb. ) Zeitschr. f. w. Zool. 42. Bd. 1885.
Treatises by Grube, Krohn, Ehlers, A. Brandt, H. Theel, Spengel, Selenka, Hatschek

(Ontogenie von Sipunculus], Horst, Slniter, Scharff, Schauinsland.

Phoronis.

A. Kowalevski. Anatomie und Entiuicklungsgeschichte von Phoronis. 1867.

W. H. Caldwell. Note on the Structure, Development, and Affinities of Phoronis. Proc.

Roy. Soc. 1882.
Further : Metschnikoff and Schneider.

Bryozoa.

G. J. Allmann. A Monograph of the Fresh-water Poly zoa. Ray Society. 1856.
H. Nitsche. Beitrage zur Kenntniss der Bryozoen. Zeitschr. f. w. Zool. Bd. XX.

1869. Bd. XXI. 1871. Bd. XXV. Suppl. 1875.
Ed. Claparede. Beitrage zur Anatomie und EntwicTcelungsgeschichte der Seebryozoen.

Zeitschr. f.w. Zool. Bd. XXL 1871.
L. Joliet. Contributions a I'histoire des Bryozoaires des cotes de la France. Arch.

Zool. experim. Tome V. 1877. Tome VI. 1878. Tome VIII. 1880.
Th. Hinks. A History of the British Marine Polyzoa. London, 1880.
Karl Kraepelin. Die deutsche Susswasser Bryozoa. Eine Monographic. I. anatomische

systematischer Thcil. Hamburg, 1887.
Treatises by Dumortier, P. J. van Beneden, Repiachoff, Smitt, 0. Schmidt, C. Vogt,

Salensky, Vigelius, Waters, Haddon, Allnian, Sars.
Ontogeny : Schneider, Barrels, Hatschek, Haddon.

Brachiopoda.

E.Owen. On the Anatomy of the Brachiopoda. Transact. Zool. Soc. London, 1835.
G. Vogt. Anatomie der Lingula anatina. Neue DenTcschr. d. schweiz Gesellsch. f.

Natunvissensch. Bd. VII. 1845.
T. H. Huxley. Contributions to the Anatomy of the Brachiopoda. Annals and

Magaz. of Nat. Hist. 2d series. Vol. XIV. 1854.

A.Hancock. On the Organisation of the Brachiopoda. Phil. Trans. 1858.
J. F. van Bemmelen, Untersuchungen uber den anatomischen und histologischen Bau

der Brachiopoda Testicardinia. Neue Zeitschr. f. Naturw. Bd. XVI. 1883.
L. Joubin. Recherches sur T anatomic des Brachiopodes inarticules. Arch, de Zoo-

logie experimentale. 2d serie. Tome IV. 1886.
H. de Lacaze Duthiers. Histoire de la Thecidie. Ann. Sc. naturelles. 4ieme

serie. Tome XV. 1861.
Further treatises and works by Cuvier, Davidson, Carpenter, Gratiolet, Morse,

Shipley, Schulgin, Beyer.
Ontogeny : Lacaze Duthiers, Kowalevski, Morse, Brooks, Shipley.

286 COMPARATIVE ANATOMY CHAP, iv

Rotatoria.

Ehrenberg. Die Infusionsthierchen als vollkommene Organismen. Leipzig, 1838.
Fr. Leydig. Ucber den Ban und die, systematische Stellung cler Rdderthicre. Zeitschr.

f. 10. Zoologie. Bd. VI. 1854.
Karl Eckstein. Die Rotatorien der Umgegend von Giessen. Zeitschr. f. w. Zool.

Bd. XXXIX. 1883.
Ludwig Plate. Beitrdge zur Naturgeschichte der Rotatorien. Jena. Zeitschr. f.

Natumv. Bd. XIX. N. F. XII. 1885.
Further treatises and works by Dujardin, Weisse, Dalrymple, Leydig, Gosse, Cohn,

Claparede, Metschnikoff, Semper, Mbbius, Plate, Zelinka, Salensky, Hudson,

Joliet, etc.

Dinophilus.

Eug. Korschelt. Ucber Bau und Entivickelung dcs Dinophilus apatris, in Zeitschr.

f. w. Zool. 37 Bd. 1882.
Further : 0. Schmidt, Mereschkovsky, Weldon, E. Meyer.

Echinoderes and Gastrotricha.

E. Metschnikoff. Ueber einige wcnig bekannte niedere Thierformen. Zeitschr. f. w.

Zool. Bd. XV. 1865.

H. Ludwig. Die Ordnung Gastrotricha. Zeitschr. f. w. Zool. Bd. XXVI. 1875.
W. Reinhard. Kinorhyncha (Echinoderes), ihr anatomischer Bau und ihre Stellung

im System. Zeitschr. f. w. Zool. 45 Bd. 1887.
Cf. further : Dujardin, Greef, Schulze, Biitschli.

Chaetognatha.

0. Hertwig. Die Chaetognathen. Eine Monographic. Jen. Zeitschr. f. Naturw.
Bd. XIV. 1880.

G. B. Grassi. / Chetognati, in Fauna und Flora des Golfes von Neapel. 5. Mono-
graphic. 1883.

Further, the works of Krohn, Wilms, Kowalevski, etc.

CHAPTER V

The first division of the Arthropoda— The organisation and development
of the Crustacea.

THE FIFTH RACE OR PHYLUM OF THE ANIMAL KINGDOM.

ARTHROPODA— ARTICULATA.

Bilaterally symmetrical animals with chitinous exoskeleton, seg-
mented body, and paired jointed extremities on all or some of the
segments. With brain, oesophageal commissures, and segmented ventral
chord. With heart lying above the intestine. Sexes separate, with
one pair of sexual glands and originally paired ducts from these glands.

First Sub-race or Sub-phylum.

Branchiata. — Aquatic animals. With the exception of the anterior
antennae all the appendages are morphologically biramose. Respi-
ration cutaneous or by means of gills, which are almost always
appendages of the basal joints of the limbs.

SINGLE CLASS. Crustacea.

First appendage to the Sub-race Branchiata : The Trilobites, Gigantostraca,
Hemiaspiclse, and Xiphosura.

Second appendage to the Sub-race Branchiata: The Pantopoda or Pycno-
gonidae.

Second Sub-race or Sub-phylum.

Tracheata. — Land animals. Limbs not biramose, consisting of a
single row of joints. Respiration by means of tracheae (tubular
or book-leaf tracheae).

CLASS I. Protracheata.

CLASS II. Antennata (Myriapoda and Hexapoda).
CLASS III. Chelicerota sive Arachnoidea.

Appendage to the Phylum of the Arthropoda.
The Tardigrada or Bear animalcules.

288

COMPARATIVE ANATOMY

CHAP.

THE CRUSTACEA.
Systematic Review.

Sub-Class I. Entomostraca.

Fig. 191.— Branchipus stagnalis, male. 01, Anterior an-
tenna; ; a-2, posterior antennae, seizing antennae with accessory
appendages ; ua, unpaired eye ; I, liver ; md (above), mandible ;
sd, shell gland ; h, heart or dorsal vessel ; oh, slit-like apertures
(ostia) of the heart ; md (below), intestine ; p, penis ; br, branchial
sac ; bri, branchial leaflet ; pa, paired stalked eyes (after Glaus).

The trunk consists of a
varying number of segments.
AVe can here often distin-
guish an anterior division
bearing limbs from a pos-
terior division which has
no such appendages. Each
division, however, consists of
a varying number of seg-
ments. The genital aper-
tures usually lie between the
two divisions of the trunk.
A dorsal shield is often
present, and is developed in
various ways. The limbs
are very variously shaped.
Besides the usual lateral
eyes, the unpaired frontal
eye of the Nauplius larva is
retained by the adult animal.
A masticatory stomach is
wanting. A Nauplius larva
is hatched from the egg.
Mostly sinall animals.

Order I. Phyllopoda.

With swimming feet
which carry branchial sacs,
mandibles without feelers,
and reduced maxillse.

Sub-Order 1. Branchiopoda.

Body distinctly seg-
mented with numerous
trunk segments, and numer -
ous pairs of swimming feet.,
Carapace seldom wanting,
either flat and shield-shaped
or in the form of a bivalve
shell. Heart an elongated
dorsal vessel with numerous
pairs of ostia. In fresh
water. Branchipus (Fig.
191) (without shell), Apus
(with flat carapace), Es-
theria, Limnadia (with bi-
valve shell).

CRUSTACEA— SYSTEMATIC REVIEW

289

Sub-Order 2. Cladocera (Daphnidse), Water Fleas.

Body small, Avith few indistinct segments and 4 to 6 pairs of swimming feet. The
posterior antennae are large rowing feet.

•/

Fig. 192.— Daphnia similis, young female (after Glaus). al5 Antennule ; a2, second (rowing)
antenna ; Z, hepatic caecum ; a?«, eye ; d, intestine ; sd, shell gland ; h, heart ; bru, brood cavity ;
ab, abdomen ; br, branchial sac ; frf5, trunk feet ; g, brain.

a-.,

Fig. 193.— Cypridina mediterranea, female, from the side (after Glaus). OL Anterior, a2,
posterior antennae ; fs, frontal organ ; oc, unpaired eye ; au, paired eye ; h, heart ; in, stomach ; ~s,
shell ; /2, cleaning foot ; g, sexual organs (?) ; /;, first foot ; mx.2, second maxilla ; sm, shell muscle,'
mxi, first maxilla ; 06, upper lip ; md, mandible.

Branchial sacs may be wanting. With bivalve shell. Head freely projecting.
Female with dorsal brood cavity between shell and trunk. Heart sac-shaped with
VOL. I

290

COMPARATIVE ANATOMY

CHAP.

at*.

e*i-

one pair of ostia. Mostly in
fresh water. Daphnia (Fig.
192), Sida, Moina, Lynceus,
Polyphemus, Leptodora, Evadne
(marine).

Order 2. Ostracoda.

Body small, consisting of
few segments indistinctly seg-
mented, with bivalve shell.
Besides the 5 pairs of well-
developed limbs to be attri-
buted to the head, viz. the
antennae, mandibles, and max-
illae, all or some of which may
be developed as creeping or
swimming feet, we find only 2
pairs of trunk limbs. The
heart may be present or want-
ing. Fresh-water form : Cypris.
Marine forms: Cythere, Halo-
cypris, Cypridina (Fig. 193).

Order 3. Copepoda (with
biramose or rowing feet).

Sub-Order 1. Eucopepoda.

Body small, mostly dis-
tinctly segmented, without
shell fold. The trunk consists
of 10 segments, the 5 anterior
carrying 5 pairs of biramose
rowing feet, while the 5 pos-
terior are limbless. The fore-
most trunk segment is fused
with the head. Antennae,
mandibles, and maxillae (the
two branches of the posterior
maxillae separated from one
another) are well developed, at
any rate in the free -living
forms. The mouth parts in
the parasitic forms either suck
or pierce. Heart sometimes
wanting ; when present it is

Fig. 194 — Clausocalanus mastigophorus (Glaus), female, from the ventral side (after an
original drawing by W. Giesbrecht). Only the extremities of the left side of the body are
depicted 01 Anterior, 03, posterior antennse ; md, mandible with masticatory ridge fc; TO*,,
anterior maxilla ; mx^a, mx^b, anterior and posterior maxillipedes = endopodite and exopodite oft
second maxilla ; trtt, rowing feet (biramose), ts, rudimentary ; /, frontal organ ; r, rostrum ; au,
eye • o upper lip; m, mouth; u, under lip; c+J, head and 1st trunk segment; JI-A', 2d-10th
trunk segments; I-V, limb-bearing segments (thoracic segments) ; Vl-X, limbless segments (abdo-
minal segments); VI+VII, genital double segment; go, genital aperture; es, ovisac (unpair
/, fur ex.

CRUSTACEA— SYSTEMATIC REVIEW

291

sac-shaped. The females carry about the fertilised eggs in a paired or unpaired
ovisac. Gills are wanting. Free-living or commensal Copepoda : Cyclops, Cantho-
mptus, in fresh water ; Cetochilus, Clausocalanus (Fig. 194), marine ; Notodelphys,
commensal in the branchial cavity of the Asddians. Parasitic Copepoda : Corycceus,
Sapphirina (some of which are only occasionally or temporarily parasitic), Chondra-
canthus, Caligus, Lerncea, Lernceocera, Penella, Lernanthropus, Lernceascus, Achthercs,
Anchorella.

Sub-Order 2. Branchiura (Argulidse), Carp Lice.

Body consists of the flattened shield-shaped cephalo-thorax and the small flat
abdomen (caudal fin) divided longitudinally. In front of the oral suctorial tube a

ff, V

<••/ aO-^C 1"

ifi. ' '

FIG. 195.— Argulus foliaceus young male (after Glaus). «i, Anterior, a2, posterior antenna ; pa,
paired eye ; ua, unpaired eye ; r, beak or suctorial tube enclosing the mandibles and maxillse ; kfi,
anterior maxillipede with the adhering disc ; kfz, posterior maxillipede ; sd, shell glands ; (?, intes-
tine with its lateral branched diverticula ; &j, &2> &s, &4> thoracic feet ; ab, abdomen ; t, testes.

long protrusible stylet. Four pairs of long cirrus-like biramose swimming feet. Two
large compound lateral eyes. Testes in the caudal fin. Heart present. Females
without ovisacs, attach the eggs to foreign objects. Argulus (Fig. 195), on the carp.

Order 4. Cirripedia,

Characteristics of the attached forms : body indistinctly segmented, attached by
the head end, surrounded by a mantle which generally calcifies and then forms a

292 COMPARATIVE ANATOMY CHAP.

shell or case. Anterior antennae (adhering antennae) minute, posterior antennae
reduced. Oral limbs small, partly reduced. Six (less frequently 4) pairs of long
biramose tendril-like feet. Without heart. Hermaphrodite, occasionally with dwarf
males, less frequently sexes separate and dimorphic. Live in the sea.

Family 1. Lepadidse (Pedunculata).

Head end elongated into an attached peduncle. Lepas (Figs. 204 and 205),
Conchoderma, Scalpellum, Pollicipes, Ibla.

Family 2. Balanidse.

Peduncle wanting. Body surrounded by a ring of calcareous plates. Balanus
(Figs. 206 and 207), Tubicitiella, Coronula.

Family 3. Alcippidse (Abdominalia).

Body surrounded by a flask-shaped integumental mantle, with 3 or 4 pairs of feet,
corresponding with the last 3 or 4 pairs of other Cirripedes. Live in the calcareous
shells of other Cirripedes and Molluscs. Alcippe, Cryptophyalus.

Family 4. Proteolepadidse (Apoda).

Body maggot-like, without tendril-like feet. Anterior (adhering) antennae ribbon-
shaped. Mouth a sucker. Enteric canal rudimentary. Parasitic in the mantle of
other Cirripedes. Proteolepas.

Family 5. Rhizocephala (Kentrogonidse), perhaps to be separated as a

special sub -order or order from the other Cirripedes.

Body pouch-shaped, answers to the cephalic portion only of related Crustaceans.
Integument split into 2 lamellae ; between them is a brood cavity which opens out-
wardly by means of an aperture in the outer lamella. Enteric canal wanting. Limbs
wanting. Hermaphrodites, with dwarf males. Parasitic on the abdomen of Decapoda.
The pouch-shaped body has an adhering peduncle from which spring the branched
"roots" which penetrate everywhere between the viscera of the host and convey
nourishment into the body of the parasite. The larval stages (Nauplius- and Cijpris-
like larvje) are like those of other Cirripedes. Sacculina (Fig. 208), Peltogaster.

Sub-Class II. Malacostraca.

The body consists of 3 regions with constant number of segments. (1) The
head, originally formed of 5 segments ; (2) the thorax, consisting of 8 segments, of
which the anterior segment or segments, or all the segments, may fuse with the head
to form an incomplete or a complete cephalo-thorax ; (3) the abdomen, which
(reckoning the telson) consists of 7 segments, (in Nebalia alone, including the
terminal segment, of 8). All the segments of the body except the last (and in
Nebalia the last but one) carry limbs. The most anterior thoracic feet often
move into the neighbourhood of the mouth to serve as foot-jaws and to assist in
taking in food. The sixth pair of pleopoda (abdominal limbs) almost always differs
in shape from the rest, and often forms with the telson a caudal fin. A shell fold
springing from the posterior cephalic region is very common. A pair of compound
lateral eyes is always found, as is also a masticatory stomach. The female genital
apertures lie in the 6th thoracic segment, the male in the last. Development
sometimes with, sometimes without, metamorphosis. The larva hatched from the
egg is rarely a Nauplius. In many Thoracostraca the larvae pass through the Zocea

Legion I. Leptostraca.

An extremely important group, which of all living Crustaceans stands the nearest

CRUSTACEA— SYSTEMATIC REVIEW

293

...u

Ml

to the racial form of the Malacostraca, and is often placed as a special sub-class
between the Entomostraca and the
Malacostraca. Body slender, covered
with a bivalve shell, which extends
back, leaving only the last 4 abdo-
minal segments free. Besides this
there is a movable cephalic plate.
Head, with the 5 typical pairs of
appendages, distinct from the thorax.
All the 8 segments of the short
thorax are distinct, with 8 pairs of
similar biramose lamellate feet. On
the basal joint of the protopodites of
these segments there is a large epi-
podial lamella functioning as a gill.
The 4 anterior pairs of pleopoda
are strong biramose rowing feet, the
2 posterior pairs are short and uni-
ramose. The last segment of the
abdomen carries two furcal processes.
On the head are 2 stalked compound
lateral eyes. Heart elongated, with
7 pairs of ostia, stretches through
the thorax and the abdomen as far
as into the 4th abdominal segment.
Masticatory stomach present.

Single order and family, Ncba-
lidce: Nebalia (Fig. 196), Parane-
balia, Nebaliopsis, marine forms.

The fossil Paleozoic forms Cera-
tiocaridce (Archoeostraca), Hymen-

ocaris, Ceratiocaris, etc., are probably related to the Lepto-
straca.

Legion II. Arthrostraca (Edriophthalmata).
A shell fold is wanting, except in the division of the
Anisopoda. The first thoracic segment (less frequently the
second also) is fused with the head, and the foremost pair
of thoracic feet are transformed into foot -jaws. The 2
lateral eyes are sessile.

Order 1. Anisopoda. Fia 19(3.— Nebalia Geoffroyi, male (after Claus).

r, Rostral plate ; c, cephalic region ; km, masticatory

First and second thoracic seg- stomach ; md, mandible ; sm, shell muscle ; mxt, feeler
ments fused with the head. Cephalo- of the anterior maxilla (cleaning foot) ; I- VIII, thoracic

thorax with lateral shell fold, which seSments.; «. testes5 s> shell«/> heart5 d> intestine;
... ai> anterior, a-->, posterior antenna ; mt, mandibular

on each side covers a respiratory feeler . 6r/j thoracic feet ; prP6} pleopoda ; «, eye.
cavity. Both the pairs of maxillse

have feelers. The feelers of the anterior -pair project into the respiratory cavity
as cleaning appendages. The maxillipede has an epipodial appendage functioning
as gill. The pair of limbs belonging to the second thoracic segment, which is
also fused with the head, are developed as powerful forceps. Abdomen with
biramose swimming feet. Heart in the thorax, generally with 2 pairs of ostia (the
heart of Apseudes has only 3 ostia). Apseudes, Tanais, Leptochelia.

294

COMPARATIVE ANATOMY

CHAP.

Order 2. Isopoda.

Body broad, often flattened dorso-ventrally. Only the foremost thoracic seg-
ment is fused with the head, the other 7 are free. No freely projecting shell fold.

ctop

--Ad

•.ed

FIG. 197.— Organisation of Orchestia cavimana, male (after Nebeski). c+ 1, Head+lst thoracic
segment; II-VIII, free thoracic segments with their extremities; Pi-p?, abdominal segments; cti,
anterior, a2, posterior antenna ; a, eye ; ce, oesophagus ; fc/, foot-jaw ; br, gills ; Im, ventral chord ;
g, brain ; aoa, anterior aorta ; sm, cesophagal stomach ; ltd, unpaired intestinal csecum ; od, egg-
bearing part of the germ glands ; h, heart ; t, testis ; vs, vesica seminalis ; de, ductus ejaculatorius ;
elid, entrance of the urinary gland (hd) into the intestine ; aop, posterior aorta ; I, ends of the
hepatic tubes ; ed, posterior end of the intestine.

The 2 pairs of maxillae without feelers. Abdomen generally short, often reduced,
mostly consisting of 6 segments with biramose lamellated pleopoda, whose branches,
especially the endopodites, function as gills. Heart in the abdomen, generally
stretches as far as into the posterior thoracic region, with 1 to 2 pairs of ostia.

CRUSTACEA— SYSTEMATIC REVIEW

295

Cymothoidea, hermaphrodite, some living free, others parasitic on fishes : Cymothoa,
Anilocra, Cirolana, Nerodla, Aega, almost exclusively marine forms. Sphceromidm,
free living, mostly marine, Spharoma. Pranizidce, free in the sea, the 3 anterior

FIG. 198.— Caprella acutifrons, after
P. Mayer. A, male from the side ; B, from
the back. 01, Anterior, a2, posterior an-
tenna ; &2> &3, &6> &7» &8. 2(i to 8tn thoracic
feet, the 2<1 moved on to the throat; br,
gills in the place of the 4th and 5th thor-
acic feet; cth+I+II; cephalo-thorax = head
+lst and 2d thoracic segments ; 1 1 1- VI II,
free thoracic segments; ab, truncated
abdomen.

FIG. 199.— Diastylis stygia, male (after
G. O. Sars). aj, Anterior, a^, posterior
antenna; cth, cephalo-thoracic shield ; IV-
VIII, free thoracic segments ; abi-abj, ab-
dominal segments ; PI, 1st pleopod ; pfi, 6th
pleopod ; en, endopodite ; ex, exopodite.

thoracic segments fused with the head. Anceidce, female parasitic, male free-living,
Anceus. Idotheidce, free living, principally marine, Idothea. Asellidcc : Aselhis,
fresh water. Oniscidce, on land : Oniscus, Porcellio. The divisions of the Bopyridce.

296 COMPARATIVE ANATOMY CHAP.

and Cryptoniscidce contain parasites which are chiefly hermaphrodite with dwarf
males. Body of the female deformed. Bopyrus (sexes separate), Gyge, Entoniscus,
Cryptoniscus.

Order 3. Amphipoda.

Body laterally compressed. In the typical Amphipoda only the foremost thoracic
segment is fused with the head, in the Caprellidce and Cyamidce the two anterior
segments. The gills are found on the thoracic feet as pouch-shaped epipodial
appendages. Where the abdomen is well developed it carries 6 pairs of biramose
feet, of which the 3 anterior, generally more strongly developed, serve as swimming
feet, the posterior, directed backwards and often stylet-shaped, as springing feet.
Heart in the thorax with 3, seldom 1 or 2 pairs of ostia.

Sub-Order 1. Crevettina.

Head and eyes small. Foot-jaws with their limb-like feelers form a large under
lip. Marine forms : Corophium, Talitrus, Orckestia (Fig. 197), Lysianassa. In fresh
water, Gammarus.

FIG. 200.— Squilla, from the side, cth, Cephalo-thoracic shield ; VI, VII, VIII, the 3 posterior
free thoracic segments ; 01-07, the segments of the abdomen ; a7, the telson ; au, eye ; a1} anterior ;
02, posterior antenna ; 1-8, the 8 thoracic feet, of which 1 is the 1st foot-jaw and 2 is the 2d foot-
jaw or the large seizing foot, 3, 4, 5 are the posterior seizing feet, 1-5 are called oral feet, 6, 7, 8,
the 3 biramose rowing feet of the 3 posterior thoracic segments ; p, penis ; p\-p§, the pleopoda
(swimming feet), p§ forming with the telson the powerful caudal fin ; br, the branchial tufts on
the exopodites of the pleopoda.

Sub-Order 2. Hyperina.

Head and eyes large, the latter often divided into frontal and lateral eyes. Foot
jaws form a small lower lip without feelers. Marine forms : Hyperia (eyes not divided),
Phronima, Platyscelus, Oxyceplialus.

Sub- Order 3. Lamodipoda.

Abdomen truncated. The 2 anterior thoracic segments fused with the head. Gills
on the 2d and 3d free thoracic segments ; limbs on these segments often reduced.
Marine forms : Caprellidce, body veiy slender and thin — Caprella (Fig. 198), Proto,
Protella ; Cyamidce, body broad and flat, parasitic on the skin of whales, Cyamus.

Legion III. Thoracostraca (Podophthalmata).

With a shell fold which covers a larger or smaller part of the thorax and fuses
with the dorsal integument of all or some of the anterior thoracic segments, always,
however, projecting freely laterally and covering the respiratory cavity.as branchio-
stegite. A varying number of anterior thoracic segments, or all the thoracic segments
fused, at least dorsally, with the head to form an incomplete or a complete cephalo-

CRUSTACEA— SYSTEMATIC REVIEW

297

...ctfl,

thorax. The 2 lateral eyes stalked (except in the
Cumaceci}.

Order 1. Cumacea.

Shell (cephalo - thoracic shield) small, leaving
the 5 posterior free thoracic segments uncovered.
Eyes sessile, close together or fused into one,
weakly developed, occasionally wanting. Two pairs
of foot-jaws. The first with a very large epipodite
carrying a gill. Of the 6 subsequent pairs of
thoracic feet, the first 2 always carry exopodites as
well as endopodites, the next 3 also often have
exopodites, but this is never the case with the last.
Abdomen long and slender. In the female the
pleopoda are wanting, excepting the last pair.
Marine form, Diastylis (Fig. 199).

Order 2. Stomatopoda.

Cephalo-thoracic shield rather small, not cover-
ing the 3 distinct posterior thoracic segments. Body
elongated, flattened dorso - ventrally. Abdomen
large and strong. The 5 anterior pairs of thoracic
feet (oral feet because moved to near the mouth),
are holding or seizing feet with epipodial lamellae,
but without exopodites. The 3 posterior pairs of
thoracic feet are biramose limbs without epipodial
appendages. The 5 anterior pairs of pleopoda are
strong lamellated swimming feet, whose exopodites
carry branchial tufts. The 6th pair of pleopoda
forms with the telson a powerful caudal fin. The
heart with several pairs of ostia is elongated into
a dorsal vessel running through the abdomen.
Ovaries and testes in the abdomen. Marine, Squilla
(Fig. 200).

Order 3. Schizopoda.

Cephalo-thoracic shield well developed, like a
delicate integument covering the whole thorax. FlG. 201.— Slriella Thompson!!,

The dorsal integument of the last 5 thoracic seg- male (after G. O. Sars). cth, Cephalo-

ments, or of the last thoracic segment, is not united thoracic shield ; VII, VIII, 7th and

with it. The 8 pairs of thoracic feet are formed 8th thoracic segments ; ab^afy, abdo-

,,.,••, ,. , .,, ,., minal segments ; pi-ps, pleopoda ; 6r,

pretty much alike, and are biramose (with exopodite gjlls

and endopodite) ; we can, however, generally describe

the 2 anterior pairs of thoracic feet as foot-jaws, as they may have masticatory
ridges. Abdomen strong, slender. Pleopoda very small in the female, strongly
developed in the male. The last pair of pleopoda, well developed in both sexes,
forms with the telson a rowing or swimming fin. Marine.

Family 1. Mysidse.

Thoracic feet without gills, the first pair with large vibratile epipodial lamellse.
Last 5 thoracic segments free under the dorsal shield. Auditory organs in the
endopodites of the 6th pair of pleopoda. My sis, Siriella (Fig. 201) ; in the male,
gills on the pleopoda.

298

COMPARATIVE ANATOMY

CHAP.

Family 2. Lpphogastridse.

With branchial tufts on the thoracic feet ; last 5 thoracic segments as in the
Mysidce. Lophogaster.

Family 3. Euphausidae.

With branchial tufts on the thoracic feet. Only the last thoracic segment is
free under the dorsal shield. Euphausia, Thysanopod.a.

\UL

FIG. 202.— Astacus fluviatilis, male, from the side, cth, Cephalo-thorax ; ab, abdomen ; kd,
branchiostegite of the cephalo-thoracic shield ; XIV, first, XIX, last but one abdominal segment ;
r, rostrum ; 0,1-0%, 1st and 2d antennae ; 8, third foot-jaw or maxillipede ; 9, chelate foot ; 10,
11, 12, 13, the 4 remaining ambulatory feet ; 19, the pleopoda of the 6th abdominal segment,
which with the telson or terminal segment form the caudal fin (after Huxley).

Order 4. Decapoda.

Cephalo-thoracic shield large, generally firm and hard, calcareous, covering the
whole thorax, and fused with the dorsal integument of all the thoracic segments.
Exopodite of the 2d maxilla forms a vibratile plate which regulates the streaming of
the water in the branchial cavity. The 3 anterior pairs of thoracic feet developed as
foot-jaws or maxillipedes, the 5 posterior, some of which are armed with pincers, as
ambulatory feet (hence " Decapoda "). In the adult these ambulatory feet consist
only of protopodite and endopodite, while the exopodite is almost always wanting ;
auditory organs on the basal joints of the inner antennse. Development director
with metamorphosis. In the latter case a Nauplius is seldom (Carididce) hatched
from the egg ; the larvae hatched are generally further developed, and pass through a
Zocea and a Mysis stage. This order is very rich in forms.

Sub-Order 1. Macrura.

With well-developed abdomen, which is at least as long as the cephalo-thorax.
Mostly with the full number of pleopoda, the last pair of which forms with the
telson a powerful caudal fin. Carididaz (shrimps) : Penceus, Palcemon, Crangon, Pon-
tonia, Alpheus, Sergestes, Lucifer, in the sea. Astacidce : Astacus flumatilis (Cray-
fish, freshwater ; Figs. 202 and 203), Homarus (lobster), Nephrops, Callianassa, Gebia,
marine. Palinuridce : Palinurus, Scyllarus, in the sea.

Sub-Order 2. Anomura (this division cannot be sharply demarcated).

Abdomen moderately large ; caudal fin mostly reduced. The hindermost pair of

ambulatory feet or the two posterior pairs reduced. Third foot -jaws limb-like.

Paguridce, hermit-crabs, marine, in empty shells of Molluscs ; abdomen soft skinned,

asymmetrical, with degenerated pleopoda serving as clinging organs. Pagurus, Eupa-

CRUSTACEA— SYSTEMATIC REVIEW

299

gurus, Birgus (in holes in the ground). Hippidce, marine, live in mud ; posterior
body hard skinned, the half of it bent forwards. The Galatcidce (Galatea) approach
the Macrura, and the Porcellanidce (Porcellana) the Brachyura.

FIG. 203.— Astacus fluviatilis, from the ventral side. A, male ; B, female. In the male the
8th to 13th and 15th to 18th extremities of the left side of the body are removed ; in the female the
2d and 4th to 13th of the right side. 3-13, Extremities of the cephalo-thorax ;\ara^ anterior and
posterior antennae ; gd, aperture of the antennal glands ; wo, female ; mo, male genital aperture ;
PrPe, pleopoda ; en, endopodite ; ex, exopodite of the last pair of pleopoda ; an, anus ; t, telson
(after Huxley).

Sub-Order 3. Brachyura (Crabs).

Body flattened. Posterior body without anal fin, reduced, bent round on the
ventral side of the cephalo-thorax. In the male only the two anterior pairs
of pleopoda are usually retained. Notopoda, in the sea : Dromia, Dorippe, Lithodes.
Oxystomata, round crabs : Calappa, Ilia, in the sea. Oxyrhyncha, • triangular crabs :
Maja, Pisa, Stenorhynchus, Inachus, Lambrus, in the sea. Cydometopa : Telphusa

300 COMPARATIVE ANATOMY CHAP.

(fresh water), Cancer, Xantho, Pilumnus, Eriphia, Portunus, Carcinus, in the sea.
Catometopa, square crabs : Pinnoteres, Ocypoda, Grapsus, in the sea, Gecarcinus,
land-crabs.

I. Outer Organisation.1

Nowhere in the animal kingdom does the study of outer organ-
isation afford so much of interest to the comparative anatomist as
among the Arthropoda. The more or less hard chitinous envelope which
outwardly covers the body and all its limbs not only serves as a
protection to the inner organs, but also represents the skeleton to
which the musculature is attached inside. Herein lies the chief cause
of the specially close relations between inner and outer organisation
in the Arthropoda.

In describing the outer organisation of the Crustacea we must take
into consideration in turn (1) the body, (2) the extremities, and (3) the
gills.

A. The Body.

We denote by this term the whole animal minus its appendages.
It consists of a number of consecutive joints (segments, metameres,
somites).

The consecutive segments are movably articulated together. The
chitinous integument investing the whole animal remains thin and soft
between adjoining segments and forms intersegmental membranes.

The study of comparative anatomy leads us to suppose that
originally each segment except the last possessed a pair of limbs, so
that the number of the limbs answered to the number of somites.
We may diagrammatically represent the Crustacean body as consisting
of a great number of segments, as is the case in the Annulata. The
most anterior or head segment is distinguished by the possession of the
eyes, the mouth, the brain, and a pair of extremities, which as feelers
differ from all the other extremities ; these latter resemble one another
more or less closely ; their special modifications will be described below.
The anus lies in the hindermost segment, which has no limbs.

This diagrammatic representation of the segmentation of the
Crustacean body is not exactly realised in any known Crustacean.
In fact we everywhere find important deviations even in those which
are considered to stand nearest the racial form.

We find first of all that in all Crustaceans the anterior region of
the body is outwardly unsegmented, and, in opposition to our typical
Crustacean, carries not one but five pairs of limbs. We are inclined
to assume that this region has arisen by the fusing of a head segment
with the four following segments, this assumption being supported by
similar phenomena observed in various groups of Crustacea.

Thus the most anterior unsegmented region of the body, which

1 In order to emphasise the great morphological significance of the Leptostracan
body, Nebalia is treated of in this division apart from the other Malacostraca.

v CRUSTACEA— OUTER ORGANISATION 301

carries 5 pairs of limbs, is known as the head, as apart from the trunk,
i.e. the whole of the remaining segmented body.

The trunk of the Entomostraca consists of a very varying number
of segments, which in different regions may differ greatly in many
respects (heteronomous segmentation of the trunk). The trunk of the
Malacostraca always consists of a constant number of segments, viz.
fifteen. It always falls into two sharply distinguished regions, each
with a constant number of segments, an anterior thoracic region,
consisting of 8 segments, and a posterior abdominal region (pleon)
containing 7.

The trunk of the Leptostraca (Nebalid), which in classification takes
a place half way between the conjectural racial forms of the
Entomostraca and those of the Malacostraca, though really more nearly
related to the latter, also consists of (1) a thorax of 8 segments
(which exactly answers to the thorax of the Malacostraca), and (2) an
abdomen of 8 segments.

There is nothing to hinder us from assuming that in all Crustaceans
the segments which are numbered alike correspond, for instance the
2d, 6th, and 10th trunk segments of an Isopod with the 2d, 6th, and
10th trunk segments of a Phyllopod.

As in the Annulata, so also in the Crustacea, the somites become
differentiated ontogenetically in regular order from before backward,
so that the youngest segment always appears posteriorly in front of
the anal segment, this latter containing the formative material for the
segments which are to appear in the course of development. There is
therefore also nothing to hinder us from assuming that the anal
segments of all Crustaceans correspond, of however many segments
the trunk may consist.

Apart from the above-mentioned conjectural fusing of the 5 most
anterior primary segments to form the unsegmented head of the
Crustacean, the metamerism of the trunk, either of the whole tffunk
or of single regions of it, may be obscured or even entirely obliterated.
Such obscuration or obliteration may in almost all cases be referred
to one or more of the following causes : —

1. To the appearance of a shell or carapace as a fold arising from
the dorsal side of the posterior head region, which spreads in varying
form more or less far back over the body, covering or enveloping the
same. Such a carapace serves for the greater protection of the body,
and also is often closely connected with the respiratory functions.
Since a shield -like integumental fold is formed in an essentially
similar manner in the most various groups of Crustaceans and their
larval forms, we have reason for assuming that it represents a primi-
tive peculiarity of the Crustacean body. By the concrescence of the
shield or shell with the integument of all or a portion of the trunk
segments, the outer metamerism of the body is obliterated to a greater
or less extent.

2. To the fusing of the anterior trunk segments with the head in

302 COMPARATIVE ANATOMY CHAP.

order to associate the limbs of these segments with the limbs already
present on the head in the work of taking in food.

3. ' To the fact that the extremities in certain regions of the body
give up their various functions (which are almost always associated
with movement), and consequently become reduced or quite disappear.
The independence of the segments which carry such reduced limbs
is then more or less lost, that is if the said region does not in any
other way gain a locomotory significance.

4. To the loss of the capacity for active locomotion by adapta-
tion to the parasitic mode of life ; this adaptation leads to the reduc-
tion of the extremities and the more or less complete obliteration of
the metamerism of the whole body.

Having prefaced these remarks by way of elucidation, we proceed
to sketch in outline the external morphology of the body in the prin-
cipal groups of the Crustacea.

I. Entomostraca.

Phyllopoda. — In Branchipus (Fig. 191) the trunk appears distinctly segmented,
and falls into an anterior and a posterior region which are called thorax and abdomen
respectively. The 11 segments of the thorax carry 11 pairs of limbs ; the abdomen,
which consists of many segments, is limbless, and ends in 2 so-called furcal plates.
A shield or shell fold is wanting. The head carries 2 stalked movable lateral
eyes. In Apus the body is covered for the most part with a flat shell fold, not fusing
with any of the anterior thoracic segments. The trunk consists of a large number of
segments, the most posterior being limbless. In the Esthcridce the body is quite
enveloped in a bivalve shell, which also covers the limbs laterally. The trunk
consists of numerous segments. The posterior segments, which have no limbs, form
a short abdomen, as opposed to the thorax of many segments. A bivalve shell also
occurs in the Cladocera ; here also the shell covers the limbs. In this case, however,
it leaves the well-marked head uncovered. In the trunk the segmentation is
obliterated. It carries 4 to 6 pairs of limbs, and ends in an unsegmented abdomen
devoid of appendages.

Ostracoda. — The body is entirely enveloped in a bivalve shell, into which also
the limbs can be withdrawn. The whole body is unsegmented. Besides the 5 pairs
of cephalic limbs there are only 2 other pairs, so that the trunk appears extraordin-
arily reduced.

Copepoda. — A shield or shell-fold is here in all cases wanting. The manner of
life of the animal leads to very various modifications of the body. In most free-
living gnatkostomatous Copepoda the trunk is distinctly segmented (Fig. 194). We
can distinguish in it a thorax of 5 segments carrying appendages, and an abdominal
region also of 5 segments, but without appendages. The most ^ anterior thoracic
segment is fused with the head to form an incomplete cephalo-thorax. The abdomen
ends in 2 spine- or bristle-carrying processes which diverge in a fork (furcal pro-
cesses). In certain female Notodelphydce each of the 4 free limb-carrying segments
has on the dorsal side an unpaired wing-like fold.

In the siphonostomatous Copepoda and Argulidcc we observe an advancing oblitera-
tion of the segmentation of the trunk and a reduction of the abdomen as the parasitic
manner of life becomes more marked. The body, in a few of the parasitic Siphon-
ostomata (Lernceidce, Lernceopodidce, Chondracanthidoe), assumes such various and
extraordinary forms that no similarity to other Crustaceans can be recognised.

Cirripedia. — These, in an adult condition, are attached or parasitic animals.

CRUSTACEA— OUTER ORGANISATION

303

The similarity to Crustaceans in outer appearance is only retained in the free-moving
young stages, while in the adult forms this similarity is hardly recognisable. If we
first consider the attached Lepadidw (Figs. 204 and 205), we can distinguish out-
wardly an attached peduncle and a flattened shell, with a slit-like aperture on one side
carried by the peduncle ; this shell consists (in Lepas} of 5 calcareous plates. If we
open the shell we find within it the indistinctly segmented body, which carries
long tendril-like feet (Fig. 205) and is attached to the shell near the point of inser-
tion of the peduncle. The tendril-like feet are found on that side of the body turned
away from the peduncle. A thorough ontogenetic and anatomical study of Lepas
has now proved that these parts are to be described as follows. The peduncle

FIG. 204.— Lepas ana-
tifera after Darwin.
Seen somewhat diagon-
ally from the carinal side,
c, Carina ; t, tergum ; s,
scutum ; p, peduncle.

FIG. 205.— Organisation of Lepas, after
Glaus. The right half of the fold of the
body integument with its calcareous plates
removed, t, Tergum ; s, scutum ; c, caring ;
TO, closing muscle of the scuta ; I, liver ;
od, oviduct ; ov, ovarium ; cd, cement
glands ; a1, anterior (adhering) antenna ; t,
testes ; vd, vas deferens ; p, cirrus-shaped
penis. ^

corresponds with the prolonged anterior portion of the Crustacean head, which has
become attached, and which still carries on its anterior end the much-simplified
anterior antenna in the shape of very small adhering organs (Fig. 205, a1). The shell
is an integumental fold arising from the posterior head region. It answers to the
shell or shield of other Crustaceans. In the chitinous integument of this fold in the
Cirrcpedia, however, there arise by calcification the various calcareous plates, to
which we shall again refer. The part of the indistinctly segmented body enclosed
by the shell which lies anteriorly next the peduncle corresponds with the posterior
head regjon, while the remainder, which carries the tendril-like feet, answers to the
thorax of an Untomostracan ; and in addition to these parts there is a small trun-
cated portion representing a reduced abdomen. The abdomen carries a long

304

COMPARATIVE ANATOMY

CHAP.

FIG. 206. — Case of Balanus
Hameri, from the side (after Dar-
win), s, Scutum ; t, tergum.

appendage, the male copulatory organ (p). This is bent forwards on the ventral side,
and lies between the tendril-like feet. The thorax consists of 6 indistinct segments.
In Lepas we find, as already mentioned, 5 shell pieces or calcareous plates of the
integumental fold (mantle), one unpaired and four paired. The unpaired piece (c) lies
on the dorsal side and is called the carina. The paired pieces (s, t) lie to the right
and left ; the anterior are called the scuta, the posterior the terga. The cleft in the
mantle or shell lies posteriorly and ventrally. Accessory shell pieces, impaired as
well as paired, may be found in addition to the above.

In the Balanidce (Figs. 206 and 207), in contradistinction to the Lepadidce, the
attached anterior portion of the body is not prolonged
like a stalk. The mantle forms by calcification several
strongly united shell pieces which surround the body
like a rampart, to which the scuta and terga form a
kind of movable lid.

In the Abdominalia, which live in the shells of
Cirripedes and Molluscs, the number of the thoracic
segments is reduced, and the body is enclosed in a
flask-shaped mantle which does not calcify, as the
shell of the host affords sufficient protection. The
Apoda live parasitically, like the Abdominalia, in
the mantle of other Cirripedes. The mantle fold
does not here attain development. The body of 11
segments assumes the form of a fly -maggot, hav-
ing lost the tendril-like feet. The Crustacean body suffers the greatest degree of
degradation in the Rhizocephala, which live parasitically on the abdomen of
Decapoda. In this case we

find only an unsegmented •v" \ "\ o

sac (Fig. 208), entirely devoid
of limbs, containing the
viscera (testes, germarium,
cement glands, ganglion),
and itself enveloped in
another outer sac-like mem-
brane. This outer membrane
which surrounds the brood
cavity has been considered,
erroneously it appears, as the
mantle. An aperture in it
leading to the exterior is
termed the cloaca. The body
is attached to that of the
host by means of a short
peduncle. On this stalk
of attachment arise long
branched filaments which
penetrate the body of the
host and conduct nourish-
ment from its body to the

or;

FIG. 207.— Balanus tintinabulum, after removal of the
right half of the calcareous ring, o-o, Edges of the aperture
of the ring sk ; sc, scutum ; t, tergum ; aj, anterior (adhering)
antennae ; ov, ovarium ; ovi, oviduct ; wo, female genital aper-
ture ; m, muscles for moving the scuta and terga ; ad, musculus
adductor scutorum (after Darwin).

parasite in a manner similar to that in which the roots of a plant convey nourishment
out of the earth.

The Rhizocephala are classed as a special order of the Entomostraca (Kentrogo-
nidce). It is from thfeir ontogeny alone that we learn that they are Crustaceans at

CRUSTACEA— OUTER ORGANISATION

305

all and nearly related to the Cirripedes. In the course of their development they
pass through stages similar to those of the Cirripedes, free-swimming and provided
with Crustacean limbs.

In many Cirripedes there are by the side of the ordinary hermaphrodite
individuals, complementary dwarf males differently formed, which will be referred

to later.

II. Leptostraca.

In the segmentation of the trunk, the Leptostraca (Fig. 196) take an inter-
mediate place between Entomostraca (especially Phyllopoda) and Malacostraca. As
in the latter, the thorax which follows the head consists of eight limb - bearing
segments, here very short but distinct. After the thorax comes the well-developed

-mb

FIG. 208.— Sacculina carcini in situ on the host (after a somewhat diagrammatic original drawing
by Professor Delage, Paris), br, Branchial region ; I, hepatic region ; d, intestinal region of the host
(Carcinus) ; ks, body ; p, peduncle of the Sacculina ; mb, basilar membrane out of which the roots of
the parasite proceed.

powerful abdomen consisting of 8 segments, i.e. of one more segment than the
typical Malacostracan abdomen of 7 segments. The abdomen is further followed
by the 2 so-called furcal processes. Only the 6 anterior abdominal segments carry
limbs. The 6 limb-bearing abdominal segments must correspond with the 6 anterior
abdominal segments of the Malacostraca. On the dorsal side of the posterior head
region an integumental fold arises, which in the shape of a delicate laterally
compressed bivalve shell covers the thorax and the 4 anterior abdominal segments,
but does not fuse with them. This shell also, like the corresponding shells of many
JSntomostraca, covers a large part of the limbs (Fig. 196). The head carries on each
side a stalked compound (facet) eye.

III. Malacostraca.

There are two peculiarities specially to be noticed in the segmentation of the body
of this large division.

VOL. I X

306 COMPARATIVE ANATOMY CHAP.

First. — Apart from the unsegmented head, which carries the typical number
of limbs (5 pairs), the trunk always consists of two regions, the thorax and the
abdomen. They both have a constant number of segments, the former 8, the
latter 7. Each of these, with the exception of the last abdominal segment, is •
typically provided with a pair of limbs. The limbs of the 6th abdominal segment
often form with the 7th or terminal segment a caudal or rowing fin.

Second.— The segments of the thorax show a tendency to fuse with the head.
Either the first segment or several anterior segments fuse with the head to form an
incomplete cephalo-thorax, or all the thoracic segments unite, and with the head
form a complete cephalo-thorax, which then shows external signs of the original
segmentation only on the ventral side.

Arthrostraca (Amphipoda, Isopoda, and Anisopoda). — The foremost thoracic
segment is fused with the head. Seven thoracic segments thus remain free (Fig.
197). The eyes are sessile (Edriophthalmata). In the Amphipoda, the body is
laterally compressed. The Caprellidce (Amphipoda) (Fig. 198) present many
important peculiarities. The 2d thoracic segment is also fused with the head,
so that only 6 thoracic segments are left free. The abdomen is reduced to a
stump. The body of the Isopoda is dorso-ventrally flattened. In a few Isopoda
even more thoracic segments become fused with the head. Thus in the Pranizidce
(Anceus) the 3d thoracic segment is comprised in the incomplete cephalo-thorax.
A shell-fold, though as a rule wanting in the Arthrostraca, is found in the Anisopoda,
although slightly developed. In these latter the 2 anterior thoracic segments
are fused with the head. In parasitic Isopoda the metamerism of the body may be
indistinct and obliterated, and the body itself become asymmetrical.

Thoracostraca.— A shell-fold is everywhere developed, proceeding from the head ;
this, as a cephalo-thoracic shield, fuses dorsally with the integument of a larger or
smaller number of thoracic segments, but, unlike the corresponding integumental fold
of the Entomostraca and Leptostraca, it only covers the thorax, never the extremities
and the abdomen. On the head (except in the Cumacea) there are 2 stalked facet-
eyes (Podophthalmata).

I. Cumacea (Fig. 199). — The cephalo-thoracic shield. remains small ; the cephalo-
thorax includes the 3 or 4 anterior thoracic segments ; the 4 or 5 posterior ones
remain free and distinct from one another. The abdomen is long and slender,
and distinctly segmented. In the female it carries no feet ; the 6th segment only
has on each side a pair of biramose stylet-like limbs. The two eyes (when present)
are fused into one unpaired eye, or are very close to one another.

II. Stomatopoda (Fig. 200).— The cephalo-thoracic shield is fused with the most
anterior thoracic segments and covers the thorax with the exception of its 3
posterior segments which remain free. The broad abdomen is very strongly
developed, and longer than the cephalo-thorax. The strongly developed lamellate
limbs of the 6th abdominal segment form with the limbless terminal segment (telson)
an imposing caudal plate (swimming fin).

III. Schizopoda (Fig. 201).— The soft-skinned cephalo-thoracic shield generally
covers the whole thorax, and fuses with the dorsal integument of a varying number of
its segments. One or more segments, however, always remain unfused. In the
Mysidce the last 5, in the Euphausidce only the last thoracic segment, remains
unfused. Abdomen elongated and strong, ending in a swimming fin.

IV. Decapoda (Figs. 202 and 203). — The strong cephalo-thoracic shield, which
occasionally becomes through calcification of its chitinous integument as hard as
stone, generally covers the whole thorax, and is also fused with the dorsal integu-
ment of the thoracic segments. A complete cephalo-thorax is thus formed. The
pleura of the cephalo-thorax, which project freely downwards at the sides, cover the

v CRUSTACEA— OUTER ORGANISATION 307

two respiratory cavities, and are called gill covers or branchiostegites. The form
and dimensions of the abdomen are very various. In. the Macrura the abdomen
is strongly developed. In the good swimmers (e.g. Carididce) the whole body
is laterally compressed, while in those Macrura which generally crawl, or only swim
occasionally and not well (Astacidoe, Palinuridcc, Galatheidce, Thalassinidce), it is
more or less flattened dorso-ventrally. The abdomen always ends in a strong caudal
fin (terminal segment with the limbs of the 6th abdominal segment). In the Paguridtu
(hermit crabs), which live in empty mollusc shells, the last thoracic segment is
separate and not fused with the cephalo-thorax ; the abdomen, which is covered
by the mollusc shell, is soft-skinned, the caudal fin reduced and bent round for-
wards. The Brachyura are distinguished by the fact that the abdomen is -reduced
to a small plate, which is bent forward on the ventral side of the cephalo-thorax, so
that in looking at these Crustaceans from above only the cephalo-thorax can be seen.
The abdominal limbs are reduced in number and form. The caudal fin is atrophied.

B. The Extremities.

According to the scheme sketched above of the segmentation of
the Crustacean body, every segment except the last is provided with a
pair of jointed extremities which articulate ventrally and laterally
with the body. In order to complete the scheme with reference to
the limbs, we must distinguish between the limbs of the most anterior
segment and the rest. The first pair are not biramose, but consist of
a single row of consecutive joints. All the other pairs are biramose.
In such a biramose limb we distinguish 3 parts ; the shaft or stem
(protopodite), the inner branch (endopodite), and the outer branch
(exopodite). The shaft consists of 2 joints, a proximal joint articu-
lating with the body, and a distal joint carrying the 2 branches.
The 2 branches themselves are again jointed ; the inner branch is
turned towards the median plane of the body, the outer branch away
from it.

The limbs of the Crustacea undergo the most various transforma-
tions according to the special functions they perform. They can
always, however, be referred back to the typical forms, the first pair
to the unbranched (uniramose) form, and all the rest to the biramose
form.

This typical arrangement is found in the Nauplius, which is the
youngest Crustacean larva, universally found among the Entomostraca
and occurring also in a few Malacostraca. The Nauplius, which
hatches from the egg, is unsegmented, and always possesses 3 pairs of
appendages, the first uniramose, the second and third biramose.

The limbs of the Nauplius become in all cases the three anterior
pairs of limbs of the adult animal. The most anterior pair becomes
the anterior antennae, the second the posterior antennae, and the
third the mandibles of the adult. In the course of the metamor-
phoses of the Nauplius, which are accompanied by numerous moults,
the larval body elongates into the adult animal, and behind the
appendages of the Nauplius new appendages bud from the body as a
rule in order from before backward. All these newly formed

308

COMPARATIVE ANATOMY

CHAP.

appendages originate as biramose limbs ; the biramose character may,
however, be more or less indistinct, or may even be lost as the limbs
develop into the corresponding limbs of the adult.

In attempting a comparative review of the limbs throughout the
class of the Crustacea, only the most important points can be touched
upon. The setae with which they are often provided, and whose form
and arrangement are extremely important for classification, cannot
here be taken into consideration.

1. The Limbs of the Head.

In all the Crustacea the head carries 5. pairs of limbs, which are
called, following the order from before backward, the anterior antennae,
posterior antennae, mandibles, anterior maxillae, and posterior maxillae.
The 3 anterior pairs correspond with the 3 pairs of limbs of the

Nauplius.

a. The Anterior Antennae (Antennules) (Fig. 209).

These lie in front of the mouth, and consist typically of a single
row of joints. As a rule they function as organs of touch, but usually
also carry the olfactory organs, and occasionally the auditory organs.

FIG. 209.— Anterior antennae (antennules) of various Crustaceans. A, Of Astacus (after
Huxley) ; o, auditory sac. B, Of Munnopsis typica, Isopod <$ (after Sars). C,OfNebaliaGeo/royi 6
(after Claus), without the setae ; pi, plate. D, Of Cyclops serrulatus 6 (after Glaus) ; rf, olfactory
hairs. E, Of Daphnia pulex (after Leydig) ; rf, olfactory hairs ; g, ganglion.

Entomostraca. — In all Entomostraca the antennules consist typically of a single
row of joints. In the Phyllopoda (Fig. 209, E] they are small, carry numerous
olfactory hairs, and are called feelers or olfactory antennse. In the Ostracoda (Fig.
193) they are strongly developed and occasionally provided with olfactory hairs, but
chiefly function as locomotory organs for crawling and swimming. The anterior
antennse of the Copepoda are strongly developed as swimming feet in the free-swim-
ming forms, and are longer than any of the other limbs (Fig. 194). They carry
olfactory hairs, and serve in the males as organs for seizing and holding the female
during copulation (Fig. 209, D}. In the parasitic forms they are usually much

v CRUSTACEA— OUTER ORGANISATION 309

shortened. The anterior antennae of the Cirripedia (Figs. 205 and 207) are very
small, and can no longer be called limbs. The cement glands, whose secretion serves
for fastening the body to the surface it rests on, open on them. These antennae, as
well as all other limbs, are wanting in the Rhizocephala. In all Cirripedes, however,
even in the Rhizocephala, they are well developed in the free-swimming young
forms (the Nauplius and the so-called Cypris-like larva).

Leptostraca. — In Nebalia (Fig. 209, C) the antennules are well developed. They
consist of a 4 -jointed shaft which carries two appendages, one in the form of a plate ;
the other, which is slender and flagellate, has many joints and carries olfactory
filaments. These two appendages can in no wise be considered as the exopodite and
endopodite of a biramose limb, as these latter always arise from the 2d (distal) joint
of the shaft or protopodite. The shaft, with its many-jointed flagellum, corresponds
with the undivided uniramose antenna. The plate is a new formation.

Malacostraca. — Here also the anterior antennas are well developed and provided
with olfactory filaments. They usually consist of a 3- or 4-jointed shaft and 2
flagella, one of which (accessory flagellum) is a secondary production of the
antennule (Fig. 209, A}. There are sometimes 2 accessory flagella, and some-
times they are altogether wanting (Isopoda), and in this latter case the antennule
shows its typical uniramose form (Fig. 209, B}. The form of the Malacostracan
antennules varies very much in details ; it shows more or less considerable variations
in the two sexes. That the antennules of the Malacostraca also were originally
uniramose as opposed to all the other biramose appendages, and that the accessory
flagella are new formations, is principally proved by the Nauplius larva which
occurs in some of the Malacostraca ; its first pair of limbs (the later antennules)
being always uniramose.

b. The Posterior Antennse (Fig. 210).

These correspond with the 2d pair of limbs of the Nauplius, being
its first pair of biramose limbs, and often serve as feelers. They' con-
sist typically of the 2- jointed shaft or protopodite, an outer branch
(exopodite) and an inner branch (endopodite). They appear in this
form in many Entomostraca.

Entomostraca. — Among the Phyllopoda the posterior antennae appear in the
Cladocera as strong biramose rowing antennae (Fig. 192). In Apus they are reduced,
and in Branchipus transformed into pincers. Among the Ostracoda in the Halocy-
pridve, and Cyprinidce they are biramose swimming feet. The exopodite is, however,
considerably reduced, and in the male supplied with seizing hooks. In the Cypri-
didce and Cytheridce, however, they are uniramose, i.e. without exopodite. The
posterior antennae of most Copepoda are clinging organs. In a few free-living forms
they are typically biramose (Fig. 210, D\ in others uniramose, consisting of several
joints (Fig. 210, C}. In the parasitic Copepoda, however, they appear degenerated
into short simple clinging hooks (Fig. 210, JE). The posterior antennae are always
wanting in adult Cirripedes.

Leptostraca (Nebalia) (Fig. 196).— The posterior antennas consist of a 3-jointed
shaft and a many-jointed flagellum, which is unusually long in the male. The exo-
podite is wanting.

Malacostraca.— In this division the posterior antennae is very commonly a
5-jointed shaft and a thin (ringed) many-jointed flagellum, the 2d joint of the
shaft carrying a scale (squame). This structure of the posterior antennas is to be
explained as follows. The first 2 joints of the shaft answer to the protopodite,

310

COMPARATIVE ANATOMY

CHAP.

the 3 other joints together with the flagellum to the endopodite, and the scale
to the exopodite, of a typical biramose limb. This is clear from the fact that the 2d
antennae of the Malacostracan larvae (Nauplius, Protozocea) show the typical biramose
structure, the exopodite (which is often still jointed) being transformed in the course
of development into the squame of the antenna in the adult. The 3 distal joints of
the shaft are thus only the 3 proximal joints of the endopodite enlarged.

Arthrostraca. — Amphipoda, the squame (exopodite) is wanting. In the female
Hyperidfx, the antennae are rudimentary. Isopoda, squame wanting. In Bopyridw
and Entoniscidas the antennae rudimentary. Anisopoda, squame present in Apseudes.

FIG. 210.— Second or posterior antennae of various Crustaceans. A, lolanthe acanthonotus,
(after Beddard)'. B, Eulimnadia texana, larva (after Packard). C, Cyclops signatus
(after Uljanin). D, Pseudocalanus elongatus (after Brady). E, Trebius caudatus, parasitic
Copepod (after Kroyer). F, Eulimnadia Agassizii, adult Phyllopod (after Packard). G-I,
Euphausia pellucida (after Sars). G, Last Furcilia stage ; H, first Cyrtopia stage ; I, young Euph-
ausia. K, Astacus fluviatilis (after Huxley); go, aperture of the green gland (antennal gland);
er, exopodite (squame) ; en, endopodite with flagellum ; I, proximal ; II, distal joint of the proto-
podite.

Thoracostraca. — Cumacea, antennae without squame, in the male with unusu-
ally long flagellum, in the female rudimentary. Stomatopoda, with large squame.
Schizopoda, with well -developed squame (Fig. 210, G-I). Decapoda, except the
Brachyura, with squame (Fig. 210, K), the outer being the posterior antennae.

The antennal glands which have been observed both in the Entomostraca and the
Malacostraca show a constant relation to the 2d antennae. That is to say, they
always open on the basal joint of their protopodites.

c. The Mandibles (Fig. 211).

The mandibles correspond with the 3d pair of extremities (the 2d
pair of biramose feet) of the Nauplius. They lie to the front at
the side of the mouth and serve for mastication. They are origin-

CRUSTACEA— OUTER ORGANISATION

311

ally typically biramose, but appear in various ways transformed in
consequence of having undertaken the function of mastication. It
is always the basal joint of the protopodite, as that lying nearest the
mouth, which changes into a variously formed, hard, masticatory
portion (corpus mandibulare), and is often toothed on the side turned
towards the mouth ; the rest of the limb is more or less degenerated.

Entomostraca. — In the Phyllopoda (Fig. 211, G) the mandible is reduced to the
horny masticatory portion. The mandibles of the Ostracoda (H} have retained the
typical form. The strongly developed masticatory joint is followed by a segmented
"feeler," whose first joint (corresponding with the distal joint of the protopodite)
may carry a little fan-like plate. This represents the exopodite, while the feeler,
with the exception of its first joint, represents the endopodite. The mandibles in
the free -living Copepoda (E, F) are toothed masticators (hence Gnathostomata)

FIG. 211.— Mandibles of various Crustaceans. A, Lucifer, Nauplius (after Brooks). B, Nebalia
9 (after Glaus). C, Campy laspis nodulosa, Cumacean (after Sars). D, A larva of Branchipus, 0.8
mm. long (after Glaus). E, Notodelphys Almannii (after Thorell). F, Cyclops tenuicornis (after
Glaus). G, Apus lucasanus (after Packard). H, Xestoleberis aurantia, Cytherid Ostracod (after
Dahl). I, Astacus fluviatilis (after Huxley) ; I, proximal, II, distal joint of the protopodite ; ex,
exopodite ; en, endopodite (feeler) ; fc, masticatory part or ridge.

and carry feelers. The first joint of the feeler (2d joint of the protopodite) may
carry a segmented exopodite. In most parasitic forms the mandibles are changed
into stylet-shaped organs for sucking and piercing (Siphonostomata). Among the
Cirripedes the mandibles are wanting in the Ehizocephala, and in other groups are
developed as masticatory portions without feelers.

Leptostraca (B) and Malacostraca (A, C, /). — The exopodite is everywhere
wranting ; it is only present in the Nauplius stage of a few Malacostraca. The
mandible consists of the basal masticatory or cutting joint, and a frequently 3-
jointed feeler, whose first joint belongs to the protopodite, while the last two
represent the joints of the endopodite. The feeler may here and there be wanting ;
it is entirely wanting in the Cumacea (C].

We see from the above review that among all Crustacea only the Ostracoda and

312

COMPARATIVE ANATOMY

CHAP.

the Copepoda (more especially the latter) still retain in the structure of the man-
dibles the original typical biramose form, since they alone retain the exopodite in the
adult animal.

d. The Anterior Maxillae (Fig. 212).

These lie, in all Crustaceans, close to the mouth, and serve chiefly
for mastication, like the mandibles and the posterior maxillae. The
biramose character is much more commonly retained in them than in
the mandibles, the exopodite being more frequently present.

Entomostraca. — Phyllopoda (D), the anterior maxillae are reduced to simple
unjointed masticatory ridges without feelers. In the Ostracoda (B, C] also the mas-
ticatory ridge is the principal part, but there is a feeler as well, and in the Cypridce

FIG. 212.— Anterior maxillae of various Crustaceans. A, Notodelphys agilis (after Brady).
B, Cypridina stellifera (after Glaus). C, Cythera viridis (after Zenker). D, Daphnia similis
(after Glaus). E, Euphausia pellucida, last Calyptopsis stage (after G. O. Sars). F, Astacus
fluviatilis (after Huxley). G, Euphausia pellucida, adult (after G. O. Sars). H, Paranebalia
longipes (after G. O. Sars). ex, Exopodite; en, endopodite ; Ar, masticatory ridge; fc], inner;
fc>, outer masticatory ridge.

and Cytheridce an exopodite, in the form of a fan-like plate, which is vibratile, and
when the maxilla moves promotes respiration. The anterior maxillae of the free-
living Copepoda (A) have masticatory ridges, feelers, and sometimes also fan-like
exopodites ; in the parasitic forms, on the contrary, these parts are much reduced.
The anterior maxillae of the Cirripedia are simple masticatory ridges without feelers ;
they are wanting in the Rhizocephala.

Leptostraca. — The anterior maxillse of Nebalia (H] are provided with two
masticatory ridges (lacinise), and carry a long, jointed, whip-like appendage, which
is regarded as an endopodite. This is bent backward dorsally, at least in the
female, and serves for cleaning the inside of the shell fold.

Malacostraca (E, F, G}.— The maxillae are flatly compressed. The exopodite is
often wanting. The distal joint of the protopodite carries a masticatory ridge
(lacinia interna), and so does the basal joint of the endopodite (lacinia externa).
The remaining one or two joints of the endopodite form the feeler (palp).

CRUSTACEA— OUTER ORGANISATION

313

Ontogeny and comparative anatomy enable us to trace back the anterior maxillte
of the Malacostraca to the typical biramose foot. In those Malacostraca which pass
through free Nauplius and Protozocea stages the maxillae are distinctly recognisable
as consisting of a protopodite with a masticatory ridge on the distal joint, an en-
dopodite of two or more joints with a masticatory ridge on the basal joint, and an
exopodite in the form of a fan-like plate. The exopodite is retained as a vibratile
fan -like plate in most Mysidce (Euphausia, Thysanopus, My sis}) and in a very
reduced form in many Decapoda.

e. The Posterior Maxillae (Fig. 213).

The posterior maxillae have the same general typical structure as

3 C

FIG. 213.— Second or posterior maxillae of various Crustaceans. A, Lernaea branchialis
(5 (after Glaus) ; ms, edge of the mouth ; md, mandible ; ml5 anterior maxilla ; d, chitinous
ridge ; in^a, m^b, anterior and posterior maxillipedes. B, Penaeidan larva (AcetesT) (after Claus).
C, Eucopia australis (after G. O. Sars). D, Paranebalia longipes (after G. O. Sars). E, Astacus
fluviatilis (after Huxley). F, Cypridina messinensis (after Claus). G, Cirolana spinipes (after
Schioedte). H, Cyclops coronatus ; mo«, inner ; m2&, outer maxillipede (endo- and exo-podite of the
2d maxilla). I, Limnocy there incisa, anterior limb (after Dahl). K, Lysianassa umbo (after Goes).
L, Lysiosquilla maculata (after Brooks). I, Proximal, II, distal joint of the protopodite ; ki,
lacinia interna ; Jce, lacinia externa ; a, b, divisions of the same,; en, endopodite (palp, feeler) ; ex,
exopodite (fan plate). In G : fcl5 Lacinia interna ; fc2, k3, divided lacinia externa ; k, masticatory
ridge (lacinia). In L : 1 and 2, joints of the endopodite. :. ,

the anterior, and like the latter serve for mastication. But they

314 COMPARATIVE ANATOMY CHAP.

often show more clearly than the anterior maxillae the biramose
character ; e.g. in the Malacostraca the exopodite is almost everywhere
retained as a vibratile plate.

Entomostraca. — In the Phyllopoda, the posterior maxillse are, like the anterior,
reduced to simple masticatory ridges. In the Cladoeera they are indeed only to be
found in the embryo.

The posterior maxillse of the Ostracoda (F, I) show very various arrangements.
They sometimes function almost exclusively as masticatory organs, sometimes they
are locomotory organs as well, sometimes only the. latter. In the first case the
masticatory ridge is well developed, the endopodite (feeler) small and 2-jointed,
the exopodite (fan plate) either rudimentary (Cypris) or very strongly developed
(Cypridina). In the second case the endopodite is longer and many -join ted. In the
third case the maxilla is formed like an ordinary limb and the fan plate has dis-
appeared. The arrangement of the posterior maxillse in the Copcpoda is very in-
teresting (A, H). The endopodite and exopodite are here retained as appendages,
which are usually jointed. Instead, however, of their being placed on a protopodite,
they are inserted direct on the body, so that we might be tempted to consider them
as special limbs. They have been called anterior and posterior maxillipedes. In
the parasitic Copepoda they serve as clinging organs and end in hooks. In the
ArgulidcK (Fig. 195, p. 291) each of the anterior maxillipedes is changed into a
large adhering disc. The posterior maxillse of the Cirripedia are small, much
reduced, and fused together into a sort of lower lip. They are wanting in the
Rhizocephala.

Leptostraca (D). — The 2d maxillse of Nebalia are biramose, with protopodite,
endopodite, and exopodite. The protopodite carries 3 lobate masticatory ridges.
The endopodite has 2 joints. The exopodite is unjointed and narrow, and is a
transition form between the jointed branch and the broad flat fan plate.

Malacostraca (B, E, G, K, L). — The posterior maxillse are more easily recognised
as metamorphosed biramose feet than the anterior maxillse, in that they (except in
the Arthrostracd) have retained besides the protopodite and the endopodite the exo-
podite as the so-called fan plate. The protopodite generally carries 2 masticatory
ridges (lacinise), one on the proximal, the other on the distal joint. These lacinise
are often divided.

The posterior maxillae of the Arthrostraca are very much simplified, most of all
in the Amphipoda (K), where the exo- and endo-podite are wanting, and both the
masticatory ridges of the protopodite are simple. In the Isopoda the lacinia of the
distal joint of the protopodite is divided ; the exo- and endo-podite are wanting. In the
land Isopoda and the parasitic forms even the protopodite, with the masticatory
ridges, is more or less degenerated.

In the Thoracostraca the posterior maxillse of the Schizopoda (£) show the arrange-
ment above described as characteristic of the Malacostraca. Both masticatory ridges
are divided (Thysanopus, Uuphausia), or the proximal remains undivided (My sis,
Lophogaster, Siriella, Eucopia). In the Cumacea the endopodite (palp) is wanting,
and the exopodite is small. In the Stomatopoda the exopodite is wanting, but on the
other hand fan-like lobes are developed on the 2-jointed endopodite. The proximal
masticatory ridge is undivided, the distal divided. In the Decapoda (B, E] the
proximal as well as the distal masticatory ridge is divided ; the endopodite (palp
or feeler) is small and unjointed (in the larva only has it two or more joints), and
the exopodite is well developed in the shape of a fan plate, with a crescent-shaped
process directed backwards, which regulates the streaming of water in the branchial
cavity.

CRUSTACEA— OUTER ORGANISATION
f. The Parag-natha.

315

We may here in passing notice certain peculiar processes which in
the Thoracostraca and a few Entomostraca (Ostracoda and Copepoda) rise
independently between the mandibles and maxillae on the ventral
integument of the head, and are called paragnatha. They cannot be
considered as separate limbs, as they are never innervated by special
ganglia. They may perhaps represent the proximal masticatory ridges
of the anterior maxillse which have become independent. In Apseudes
(Anisopoda) a ridge-like portion is marked off' from them.

2. The Limbs of the Trunk.

The limbs of the trunk can be deduced from biramose feet. Their
number varies in the Entomostraca, but is constant in the Mcdacostraca.
In the latter we always distinguish, in correspondence with the
segmentation of the trunk, 8 pairs of thoracic limbs and 6 pairs of
abdominal limbs or pleopoda.

There is nothing to prevent us from assuming that the trunk
appendages of the Entomostraca correspond with those of the Mala-
costraca, pair with pair from before backwards.

a. Entomostraea (Fig. 215).

Phyllopoda. — The Branchiopoda, and Cladocera must be described separately.
Branchiopoda (Fig. 214 ; Fig. 215, C) : the numerous (10 to 36) pairs of trunk limbs are
formed pretty much alike. They are
wanting on a varying number of pos-
terior (abdominal) segments. All the
limbs are leaf - shaped swimming feet
with branchial appendages. They also
serve for whirling food within reach.
Their structure is essentially as follows.
An unjointed or indistinctly jointed stem
carries on its inner side (that turned to
the median plane of the body) 6 appen-
dages or lobes (endites) and on the outer
side a flat respiratory plate and a pouch- or
sac-like branchial appendage (epipodite).
It is at present impossible, without
straining, to deduce all these parts from

FIG. 214. — Apus. Transverse section in the
neighbourhood of the 7th or' 8th pair of feet.
h, Heart ; d, intestine ; ov, ovaries ; 5m, ventral
chord ; ah, respiratory cavity between the shell
(s) and the body ; 1-6, endites ; 6r, gills ; ex, respir-
there are 10 to 2/ pairs of swimming feet. atory plate (after Packard).

The respiratory plate is divided into two.

The Apodidcc (Apus) usually possess 35-50 pairs of swimming feet. The endites
are jointed, and may be described as flagellate appendages ; they are very long (the
5th especially) on the 1st pair of swimming feet. The llth pair of feet carry on

the typical component parts of a biramose
limb. The respiratory plate is generally
held to be an exopodite. In the Limna-
diadce (Limnetis, Estheria, Limnadia)

316

COMPARATIVE ANATOMY

CHAP.

each side a basin-like brood capsule serving for the reception of the eggs, formed out
of the shaft and its respiratory plate. Each of the limb-bearing trunk segments lying
behind the llth carries several (up to 6) pairs of swimming feet, gradually diminishing

in size, an arrangement which is
not yet sufficiently explained. The
Branchiopoda (Branchipus] gener-
ally possess 11 pairs of swimming
feet.

The Cladocera (Fig. 215, A\ in
contradistinction to the Branchio-
poda, are distinguished by the
small number (4 to 6) of their trunk
limbs. The special form of these
trunk limbs in some genera recalls
the swimming feet of the Branchio-
poda, and they carry branchial
appendages, especially in Sida and
Daphnia. The most anterior trunk
feet may, however, become slender
and more leg -like, and finally
through degeneration of the
branchial appendage and the re-
spiratory plate may become long
seizing feet (Polyphemus, Lepto-
dora).

FIG. 215.— Trunk feet of some Entomostraca. A, Ostracoda.—The reduction in
Daphnia similis (j> , 2d limb (after Glaus). B, Limno- the number of limbs here goes
cythere incisa, last (3d) limb, i.e. 2d trunk limb (after even further than in the Cladocera.
Dahl). c, Apus longicaudatus £, 1st limb (after We find only 2 pairs, which (Fig.
Packard). D, Notodelphydaa, Doropygus porcicauda 215 £) are long and many-jointed

d

m, . . .
The anterior pair

serve as creeping or clinging feet,
the posterior as cleaning feet. In
Cypridina the latter are inserted dorsally on the trunk, and are here long many-
jointed appendages (cf. Fig. 193, p. 289). Locomotion is effected chiefly by the limbs
of the head.

Copepoda (Figs. 194 and 195, pp. 290 and 291 ; Fig. 215, D).— The 4 or 5 pairs of
feet are limited to the anterior part of the trunk, which as thorax is opposed to
the limbless abdomen. The most anterior pair is inserted on the 1st thoracic
segment, which is fused with the head, and is generally unlike the other pairs in its
form. The thoracic limbs, as rowing feet, cause the swimming movement of the
Copepoda. They, unlike those of the Phyllopoda, exhibit in a fine typical manner
the biramose character, as they consist of a protopodite of 2 joints, an exo- and
an endo-podite. The exo- and endo-podites generally have 3 joints (in the Argulidce
they are long and many jointed) and function as flat oars. Adaptation to a
parasitic mode of life in the Copepoda leads to the reduction and occasionally to
the disappearance of the thoracic feet, e.g. in the Chondracanthina the 3d, 4th, and
5th pairs of thoracic feet are wanting, and in the Lernceopodidce all the thoracic
feet have disappeared.

Cirripedia (Fig. 215, E}. — The trunk extremities of these Crustaceans are biramose ;
their exo- and endo-podites are long and many -jointed, and are described as tendril-

? , swimming foot of the 4th pair (after Brady). E,
Balanus perforatus, 2d cirrus (after Darwin). 1,2,3, j ,.,

4, 5, Bndites ; ex, respiratory plate or exopodite ; to, gill ; and exopodites.
I and II, joints of the protopodite.

CRUSTACEA— OUTER ORGANISATION

317

like feet. They are alternately protruded and withdrawn through the shell- or
mantle-cleft, and serve for taking in particles of food and at the same time
as respiratory organs. Six pairs of tendril-like feet are found in the Lepadidce and
Salanidce, 3 or 4 pairs in the Abdominalia. In the Proteolepadidce and Rhizowphala
the tendril-like feet entirely disappear.

6. Leptostraea.

Nebalia, in the morphology of its trunk limbs, represents in many respects a
transition form between the Entomostraca and the Thoracostraca. We can distinguish
thoracic feet and abdominal feet (pleopoda) corresponding with the division of the
trunk into a thorax of 8 segments (exactly answering to the Malacostracan thorax)
and an abdomen (pleon) of 8 segments. The 8 pairs of thoracic feet are similar to
one another, as in most Entomostraca. They are
(Fig. 216) lamellate, strongly recalling the leaf-
shaped limbs of the Phyllopoda, between which and
j;he thoracic feet of the Malacostraca they form a
connecting link. Each thoracic foot of Nebalia
consists of the 3 parts characteristic of a biramose
foot, viz. a protopodite of 2 joints, an exopodite,
and an endopodite. The proximal or basal joint
of the protopodite carries a doubly-tipped branchial
lamella (epipodite), probably corresponding with the
branchial appendage of the trunk feet in the Phyllo-
poda. On the one side of the distal segment is
inserted the 5-jointed endopodite as a direct pro-
longation of the protopodite, and on the other the
unjointed exopodite in the form of a branchial
lamella, probably homologous with the respiratory
plate on the trunk feet in the Phyllopoda. FlG- 216. -Nebalia, leaf-shaped

In the abdomen (pleon) only the 6 anterior Seg- %£* ^0^1, ££
ments carry limbs. The 6 pairs of limbs probably podite ; ex, exopodite ; ep, epipodite.
correspond with the 6 pairs of pleopoda in the Mala-
costraca. The 4 anterior pairs (Fig. 225, D) serve for swimming ; they are typical
biramose feet (with proto-, exo-, and endo-podites), and show some similarity to the
swimming feet of the Copepoda. An epipodite is wanting. The last 2 pairs of
pleopoda (Fig. 225, E] are short and uniramose, and consist of one or two joints.

c. Malaeostraea.

The trunk feet fall into thoracic feet and abdominal feet (pleopoda) in correspond-
ence with the division of the trunk into a thorax of 8 segments and an abdomen
(pleon) of 7. We find 8 pairs of thoracic feet and 6 pairs of pleopoda, the last
abdominal segment being always devoid of appendages. It is better to consider the
extremities of the thorax and of the abdomen separately.

The Thoracic Limbs.

As a varying number of the anterior thoracic segments may fuse with the head,
a varying number of the anterior thoracic feet often enter into close relations with
the mouth, as accessory organs for the taking in of food (foot-jaws, maxillipedes).
The thoracic foot of Nebalia described above may be considered as the primitive
form of the thoracic feet in the Malacostraca.

In a typical Malacostracan thoracic limb the proximal joint of the protopodite

318

COMPARATIVE ANATOMY

CHAP.

carries an epipodite, while the exopodite and the 5-jointed endopodite are attached
to the distal joint. Very often both the exopodite and the epipodite disappear, and
the thoracic foot is then an unbranched, 7 -jointed limb. The proximal joint of the
protopodite occasionally fusing with the skeleton of the thorax, the distal joint alone
is recognisable.

Arthrostraca (Fig. 217). — Here, where the most anterior thoracic segment is
fused with the head, the most anterior pair of thoracic feet are associated with the

FIG. 217.— Thoracic limbs of some Arthrostraca. A, 1st pair of thoracic feet (maxillipedes) of
Amphithoe penicillata (Costa). B-D, Apseudes. B, 1st right thoracic foot ; C, 3d thoracic foot ;
D, 2d thoracic foot (after Boas). E and F, Asellus. E, 3d thoracic foot ; F, 1st thoracic foot (after
Boas). J, II, Joints of the protopodite ; 1-5 joints of the endopodite ; k, fcj, masticatory ridges ; ep,
epipodite ; ex, exopodite ; en, endopodite.

oral limbs as a pair of maxillipedes. The absence of the exopodite is the general
rule for the thoracic feet of the Arthrostraca. The proximal joint of the protopodite
often fuses with the thoracic skeleton. On the basal joint of some thoracic feet
there is, in the female, a lamellate appendage, the brood plate or lamella. The brood

plates cover over a cavity, the brood pouch,
on the ventral side of the thorax, and into
this pouch the eggs enter and there develop
(Fig. 218).

The first thoracic foot (maxillipede) is
characterised by the fact that the distal joint
of the protopodite, and in the Amphipoda the
proximal joint of the endopodite as well, car-
ries a masticatory plate (lacinia).

The Amphipoda are distinguished by the
fact that the middle and posterior thoracic
feet carry on the basal joints of their proto-
podites pouch-like gills (epipodites) (Fig. 218),
FIG. 218.— Corophium longicorne (Am- which, however, by the fusing of the basal
phipod). Transverse section through the joints with the skeleton of the trunk may be
thorax (after Delage.) d, Intestine; h, inserted directly on the latter. They are

eavity ; bf, thoracic feet.

istic of epipodial appendages, but they rise
from the inner side of the basal joints. The
gills of the Caprellidce (Fig. 198, p. 295) are generally limited to the 4th and 5th
pairs of thoracic feet, and these thoracic feet are then reduced to the proximal joint
of the protopodite. In the Isopoda gills are wanting on the thoracic feet, except on

CRUSTACEA— OUTER ORGANISATION

319

the maxillipede, on which an epipodial appendage in the shape of a firm plate has
been retained (Fig. 217, F). The Anisopoda deviate in many ways from other Isopoda
(Fig. 217, B-D], especially the genus Apseudes. The anterior thoracic foot (maxilli-
pede) (£) possesses a large epipodial appendage, which by its vibration causes a
constant current of water in the respiratory cavity formed by the shell-fold. On
the 2d and 3d thoracic feet (0, D) of Apseudes there are rudimentary exopodites,
a fact of great importance in tracing back the thoracic feet of the Arthrostraca to
biramose feet. The thoracic feet of the second pair are transformed into strong
chelate feet in the ordinary manner, i.e. the ultimate (5th) joint of the endopodite
is opposable to a distal process of the penultimate (the 4th).

FIG. 219.— Thoracic feet of Diastylis stygia ? (after G. O. Sars). A, 1st, B, 2d, C, 4th, D,
6th thoracic feet, ep, Epipodial plate ; &r, gill on the same ; en, endopodite ; ex, exopodite, which
in A is a hard lamella ; Irp, brood plate.

Thoracostraca. — In the Cumacea (Figs. 199 and 219) the most anterior thoracic
foot has become a maxillipede. The remaining thoracic feet are long. Brood-
lamellae occur on the basal joints of the 2d to the 6th pairs in the female. The
exopodite is wanting in the 1st (?), 2d, and 8th pairs ; in the female usually also in
the 6th and 7th ; on the other feet it is present and serves for swimming. The
endopodite has 5 joints. An epipodial appendage is developed only on the 1st
thoracic foot (maxillipede), but here is very large. It has numerous branchial tubes.
The distal joint of the protopodite of the maxillipede carries a masticatory ridge.

Among the Stomatopoda (Figs. 200 and 220) the 5 anterior pairs of thoracic feet
are formed very differently from the 3 posterior pairs. The latter arise from the
3 free posterior segments of the thorax, which are not covered by the cephalo-

320

COMPARATIVE ANATOMY

CHAP. V.

thoracic shield. The 5 anterior pairs, which have moved to the neighbourhood of
the mouth, have no exopodite in an adult condition, though it is to be found in
their larval stages (Fig. 220). They all possess a disc-shaped epipodial appendage

which serves for respiration. The endo-
podite and the protopodite together only
contain 5 joints. These 5 anterior pairs
of feet are armed with prehensile hooks.
Such a prehensile hook is formed by the
last joint of a foot moving upon the last
joint but one, like the blade of a knife
upon its handle. Very powerful hooks
of this kind are developed in the 2d pair
of limbs for catching prey. The last 3
pairs of thoracic feet serve as ambu-
latory feet. They are biramose feet with
somewhat reduced endopodites ; the
exopodite here forms the limb - like
prolongation of the protopodite. Epi-
podial appendages are wanting.
FIG. 220.— Thoracic feet of a Squilla larva, In the Schizopoda the thoracic feet
after Claus. A, 2d maxillipede. B, one of the are very interesting, connected on the
subsequent 3 thoracic feet, en, Endopodite : ex, -i j • , •• ,v ? j • ,1 T *

exopodite one han(^ Wltn tllose found in the Lepto-

straca, and leading on the other to those

of the Decapoda. All the 8 pairs of thoracic feet are still more or less similarly
formed, and are biramose. It is best to describe the Euphausidce first, then the
Lophogastridce, and lastly the Mysidce.

In the Euphausidce (Fig. 221, F-I) the thoracic feet consist of the 2-jointed
protopodite, a 5-jointed endopodite, and an exopodite which is composed of a one-
jointed shaft and a flagellum which is frequently ringed. All the 8 pairs of thoracic
limbs have epipodial appendages on the basal joints of the protopodites ; these
appendages are simply pouch-like on the first pair, but are more or less branched on
the other pairs, and form gills. The 2 anterior pairs of thoracic feet are slightly
different from the subsequent pairs, the proximal joint of the protopodite having
a ridge-like process. Herein we see the beginning of the metamorphosis of these
thoracic feet into maxillipedes. In Euphausia the endopodite is wanting in the
last two pairs, in Thysanopus in the last pair.

In the Lophogastridce (Fig. 221, A and B] the 1st thoracic foot has already become
a maxillipede, and the 2d also approaches the form of one. On the 1st pair the
epipodial appendage becomes a broad vibratile plate. In the female brood-lamellse
are found on the basal joints of the protopodites of the other feet. Gills arise near
this basal joint, but from the integument of the thorax itself; these correspond
with the epipodial gills of the Euphausidce, and may be described as appendages
dislocated from the proximal joint of the protopodite.

In the Mysidce (Figs. 201, 221, C-E) both the 2 anterior pairs of thoracic limbs are
maxillipedes with masticatory ridges. The first maxillipede carries a vibratile
epipodial plate. Some or all the other feet may in the female carry brood-lamellse
(Fig. 222). Branchial appendages are wanting on the thoracic limbs of the Mysidce.
The Decapoda (Figs. 223, 224, 202, 203) have the three anterior pairs of thoracic
feet transformed into maxillipedes, which no longer serve for locomotion but only for
taking in food. The exopodite is well developed in them, and flagellate ; the epipod-
ite lies in the branchial cavity in the form of a long lamella. In the Brachi-
ura it is whip-like and resembles a cleaning foot. The anterior maxillipedes have

FIG. 221.— Thoracic feet
of ScMzopoda (after G. O.
Sars). A and B, Lophogas-
ter typicus. A, 1st thoracic
foot (rnaxillipede) ; B, 2cl
thoracic foot of the <? , with
brood-lamella. C-E, Mysis
flexuosa. C, 1st thoracic
foot (maxillipede) ; £, 2d
thoracic foot (maxillipede) ;
E, 3d thoracic foot. F-I,
Thysanoessa gregaria. F,
1st, G, 5th, H, 7th, I, 8th
thoracic feet. I, II, Joints
of the protopodite, 1-5, of the
endopodite ; en, endopodite ;
ex, exopodite ; br, gills ; ep,
epipodial plate ; brl, brood-
lamella ; k, masticatory
ridge.

FIG. 222.—
Boreomysis
scyphops 9
(after G. O.
Sars). The
free portion
of the ceph-
alo-thorax
cut off at x
to show the
free thoracic
segments II-
FIII lying
beneath with
the branchial
folds ~br. aul,

Cup-shaped ophthalmic lobe without eye (without pigment or visual apparatus) ; ctj, 0-2, anterior
and posterior antennae ; mdt, mandibular feeler ; mx, maxillse ; &1; 1st thoracic foot with exopodite
(ex) and epipodite (ep), the latter in the uncovered branchial cavity; U, brood lamellze of the
2d-8th removed thoracic feet ; ab\, 1st abdominal segment ; cth, cephalo-thorax.

VOL. I Y

322

COMPARATIVE ANATOMY

CHAP.

well-developed masticatory ridges. In contradistinction to the maxillipedes, the 5
posterior pairs of thoracic feet are described as ambulatory feet. The Decapoda owe
to them their name. They are distinguished by the want of the exopodite, so

that the appendage consisting of the
protopodite and endopodite is a simple
limb of 7 joints. The exopodites, how-
ever, may be present in larval stages,
and in a few cases may be retained as
rudiments in the adult. The basal
joint of the ambulatory feet carry gills,
which project up into the respiratory
cavity. The Decapodan gills, however,
must be described in a special section.
The anterior ambulatory feet are often
ch elate ; the first pair are generally very
powerful. In the cray-fish the 3 an-
terior pairs of ambulatory feet are
ch elate ; the most anterior ambulatory
foot is provided with the well-known
large forceps.

The Abdominal Feet (Pleopoda)

(Fig. 225).

The 6 anterior of the 7 ab-
dominal segments typically carry
limbs, while the last segment,
the telson, is always limbless.

Arthrostraca. — In the Amphipoda,
setting aside for the present the aber-
rant group of the Caprellidce, we find
that the pleopoda are well developed
as typical biramose feet. The 3 an-
terior pairs of pleopoda are directed
FIG. 223.-01der larva of Calliaxis in advanced forwards and are strong swimming
Mysis stage. Gills and extremities of the cephalo- ,,.,,' , -, g

thoracic region (after Glaus), md, Mandible ; mx,, feet with many -jointed exo- and endo-
anterior, mx* posterior maxillre ; I-VIII, thoracic podites. The 3 smaller posterior pairs
feet, of which m/i=lst maxillipede, m/2=2d maxilli- of pleopoda (G\ also mostly biramose,
pede, m/3= 3d maxillipede, grg*>,= ambulatory feet; are directed backwards and pointed;
ex, exopodites ; en, endopodites ; a, epipodites

= podobranchife ;
branchiae.

arthrobranchise ; c, pleuro-

they often serve for springing. The
pleopoda are rarely lamellate. In the
Caprellidce, where the abdomen is
usually rudimentary, it carries at the most 3, at the least 2, much-degenerated limbs,
which are better developed in the male (as copulatory feet) than they are in the
female.

Isopoda. — The pleopoda of the Isopoda are biramose limbs, whose endo- and
exo-podites are unjointed, and generally delicate - skinned, lamellse, which serve for
respiration. The last (6th) pair of pleopoda either forms together with the last
abdominal segment a rowing fin (F)t as in the marine Isopoda, or is sty let- shaped,
as in fresh-water and land Isopoda. In the parasitic Bopyridce and Cryptoniscidce
the pleopoda are reduced or entirely wanting. In the land Onisddce the outer lamellse

OR USTA CEA — GILLS

323

of the anterior pleopoda contain air chambers. In the Anisopoda the pleopoda are
biramose swimming feet, not serving for respiration.

Thoracostraca. — In the Stomatopoda we find on the strong abdomen 6 pairs of
well-developed typical biramose pleopoda. The 5 anterior pairs are swimming feet,
each of whose outer lamellae carries a branched gill. The 6th pair of pleopoda form
with the telson (7th abdominal segment) a strong caudal fin.

In the Cumacea the 6th pair of pleopoda (A, p6) consists of long bifurcated
processes. The pleopoda of the 5 anterior abdominal segments are wanting in the
female ; in the male they are swimming feet (H), and either present in their full
numbers or in 2 to 3 pairs.

In the Schizopoda the 6th pair of pleopoda (C) forms with the telson a caudal fin.
Where there is an auditory organ it lies on the inner lamella of this pleopod. The
5 anterior pairs of pleopoda are well-developed biramose swimming feet, in the male

FIG. 224.— Thoracic feet of Astacus fluviatilis (after Huxley). A, 2d thoracic foot (2d maxil-
lipede). B, 1st thoracic foot (1st maxillipede). C, Part of transverse section through the thorax,
showing an ambulatory foot and gills. D, 3d thoracic foot (3d maxillipede). I, II, Joints of
the protopodite ; 1-5, of the endopodite en ; ex, exopodite ; pob, podobranchiae ; ab, arthrobranchiae ;
plb, pleurobranchise ; k, masticatory ridge.

at least. The 2 anterior pairs serve as copulatory organs in the Euphausidcc. In
the male of Siriclla the pleopoda carry gills (£).

Dccapoda (1-N}. — The development of the pleopoda stands in direct relation to
the development of the abdomen itself. In the long-tailed Dccapoda (Macrura) 6
pairs of biramose pleopoda are generally found. The 6th pair with the telson forms
a strong caudal fin. The 5 anterior pairs play no very important part in locomotion.
In the Brachyura the pleopoda, in correspondence with the great reduction of the
abdomen, are reduced. A caudal fin is usually wanting. In the male only the 2
anterior pairs of pleopoda are found, in the female 4 pairs. The caudal fin is
generally reduced in the Anomura, and the pleopoda are truncated and only,
developed on one side. The 2 anterior pairs of pleopoda very generally serve in
the Decapodan male as copulatory organs. In the female the pleopoda often carry
the eggs after their discharge and fertilisation.

324

COMPARATIVE ANATOMY
C. The Respiratory Organs — Gills.

CHAP.

Respiration always takes place in Crustaceans by means of the
outer integument. In small Crustaceans, in which the cuticle lying
over the hypodermis is thin and delicate, the whole body surface

Fio. 225.— Pleopoda (abdominal feet) of Leptostraca and Malacostraca. A, End of the
abdomen of Diastylis stygia (after G. O. Sars) ; V, VI, VII, abdominal segments ; VII, telson ; p$,
pleopoda of the 6th segment ; en, endopodite ; ex, exopodite. B, 2d pleopod with gills, exo- and
endo-podite of Siriella Thompson!! (after G. 0. Sars). C, End of the abdomen (caudal fin) of Siri-
ella gracilis (after G. O. Sars) ; VI, 6th abdominal segment ; VII, 7th abdominal segment (telson) ;
en, ex, endo- and exo-podites of the 6th pleopoda, which together with the telson form the caudal"
fin ; g, auditory organ. D, An anterior pleopod of Nebalia (after Glaus) ; ex, exopodite ; en,
endopodite. E, Nebalia, 6th pleopod of the 9 (after Glaus). F, Anilocra (Isopod), caudal tin ; VI,
6th abdominal segment ; VII, telson ; p6, 6th pleopod with exopodite (ex) and endopodite (en) (after
Delage). G, Lysianassa producta (Amphipod) ; end of the abdomen with the 4th, 5th, and 6th
pleopoda, p±, p5, p$ ; IV, V, VI, VII, abdominal segments ; en, endo-, ex, exo-podite (after Goes).
//, Diastylis stygia, 1st pair of pleopoda ; ex, exo-, en, endo-podite (after G. O. Sars). " I-N,
Astacus fluviatilis. I, 3d pleopod of male ; K, 1st pleopod of female ; L, 1st pleopod of male ;
M, 3d pleopod of the female; N, 2d pleopod of the male. 1, Anterior surface ; 2, posterior
surface ; ex, exopodite ; a, the rolled-up plate of the endopodite ; 6, jointed end of the endopodite
(after Huxley).

breathes. This is especially the case in Ostracoda, Copepoda, many

Cladocera, and many Cirripedia, which have no specific respiratory organs.

In the great majority of Crustaceans, however, the respiration

is particularly active at definite parts of the body, even if in addi-

OB USTA CEA— GILLS

325

tion the whole integument or a large portion of the same is capable
of respiration in a lesser degree.

The respiratory functions in the Crustacea, as elsewhere, are
promoted and facilitated by various adaptations. Among these
we must note (1) the increase of the respiratory surface, the
outer integument, (2) the flowing of the blood to and from the
respiratory parts, (3) adaptations which serve to bring about a
continual change of the aerated water which bathes the respiratory
parts, and (4) adaptations designed for the protection of the neces-
sarily delicate-skinned respiratory organs.

The principal increase of the respiratory surface is due to the
integumental fold which rises so generally from the head and appears
in very different forms as mantle, dorsal carapace, bivalve shell, or

FIG. 226.— Euphausia pellucida 9 (after G. O. Sars), cephalo-thorax (cth) and first abdominal
segment (a&j) from the side, h, Heart ; ovd, oviduct ; ov, ovary ; I, liver ; m, stomach ; era, eye ; a-[,
anterior, ao, posterior antenna ; ex^-ex^ exopodites of the 6 anterior thoracic limbs ; en\, en%,
endopodites of the 2 anterior thoracic limbs ; the 4 following are not drawn ; endo- and exo-podites of
the 7th and 8th thoracic limbs rudimentary ; &rr&r8, gills on the protopodites (I-VIII) of the
thoracic limbs ; 6r1? first gill, a small epipodial appendage. •

cephalo-thoracic shield. Wherever this fold remains soft-skinned it
always helps in the respiration. It is often only the inner wall of the
fold which remains soft-skinned, and we frequently meet with adapta-
tions which serve for the purpose of setting in motion the water found
in the cavity between the fold and the body (respiratory cavity).
There are in Malacostraca (Zocea larvce, Tanaidce, Mysidce, Stomatopoda)
epipodial appendages of the maxillae or of the anterior thoracic feet,
which, during the movements of the limbs which carry them, vibrate
and cause a stream of water in the respiratory cavity in which they lie.
In a similar way in the Ostracoda the movement of the fan-like plates,
the so-called branchial appendages, which may occur on the 4th,
5th, and 6th pairs of limbs, bring about a constant stream of water
over the body surface. In the higher Malacostraca, especially in the

326

COMPARATIVE ANATOMY

CHAP.

flu

Dempoda, the integumental fold, which becomes very hard and thick,
loses its respiratory significance, and as branchiostegite becomes a pro-
tective cover to the delicate gills which lie laterally in the respiratory
cavity under it.

In the Balanidce a pair of fold-like projections of the mantle stand
out into the mantle cavity. These folds may again form numerous
lateral folds, so that the mantle surface is very much increased.
These formations have been assumed to be gills.

In some Cyprididce (Ostracoda) the body carries on each side near

the dorsal middle line, under the
shell, a row of small branchial
lamellae.

The function of respiration
in the large majority of Crusta-
ceans is performed by the limbs
or their appendages. This is
comprehensible, since the change
of the water caused by the move-
ment of the limbs is of the greatest
use in promoting respiration.

In many Lepadidce among the
Cirripedia the tendril-like feet have
cylindrical or lancet-shaped ap-
pendages which with doubtful
accuracy have been indicated as
gills. They are found on the first
pair or pairs, and occasionally on
all of the tendril-like feet, and are
usually inserted on them near the
base.

The leaf-shaped swimming feet
FIG. 227. -Transverse section through the of ti.p PJiuJJnnnJa arp vprv wpll
cephalo-thorax of the Cray-fish near the heart, Ol . me rO^UOpOOa ar
diagrammatic, kd, Branchiostegite ; fc, gills ; Teh, SUlted for respiration. , The ap-
respiratory or branchial cavity ; ep, lateral wall of pendagCS which are Called gills and

£2tt^i55±K;*3E respiratory plates in this order

abm, ventral longitudinal muscles running to the have already been described. The

abdomen ; dim; dorsal longitudinal muscles run- g^fe Qf faQ pMlopo^a are epipodial

ning to the abdomen ; 6m, ventral chord ; sit, c ,

subneural vessel ; bf, ambulatory foot; vs, ventral appendages, perhaps COrreSpOnd-

sinus ; ov, ovary. The arrows indicate the direc- ing with the gills of the TIlOTCl-

tion in which the blood flows (after Huxley and „_/„._

Plateau). COStl aca"

In the Leptostraca (Nebalia),

besides the delicate shell -fold, the lamellate thoracic feet function in
a special manner as respiratory organs. Their two branchial plates
(epipodite and exopodite) correspond with the branchial and respiratory
plates of the Phyllopoda. The branchial plate (epipodite) of the
proximal joint of the protopodite is morphologically equivalent to a
Decapodan gill.

CRUSTACEA— GILLS

327

In the Isopoda the delicate-skinned lamellae of the pleopoda serve
for respiration, either both the lamellae of a foot or only the inner, the
outer one being hard-skinned and serving as a covering plate to protect
the inner one.

The gills of the Stomatopoda (Squilld) are found as branched

FIG. 228.— ^4 and B, Gills of Astacus fluviatilis. In A the branchiostegite is removed. The gills
are seen in their natural position. In B the podobranchiae are cut off, and the outer arthrobranchise
turned back downwards. Twice the natural size, oj, a2, 1st and 2d antennae; 3, mandible ; ep5,
epipodite of the 2d maxilla ; 6, 1st maxillipede ; 7, 2d maxillipede ; 8, 3d maxillipede ; 9, forceps
(cut-off) ; 10-13, the 4 succeeding ambulatory feet ; pli, 1st pleopod ; a&i, a&2 1st and 2d abdominal
segments ; pdb, podobranchige ; arl>, inner, orb1, outer arthrobranchise ; plb, pleurobranchise ; the
numbers attached show the appendages to which the gills belong, in order from before backward,
beginning with the anterior antenna as No. 1 ; gh, articular membrane between the body and basal
joint of the protopodite (after Huxley).

appendages on the outer lamellae of the abdominal swimming feet
(pleopoda). The arrangement of the gills in Sirietta (Mysidce)
is similar; they here also occur in the males as appendages of the
pleopoda, but on their inner branches.

The gills of the Amphipoda, Schizopoda (except the Mysidce), and

328

COMPARATIVE AX ATOMY

CHAP.

Dccapoda are always originally epipodial appendages on the basal
joints of the protopodites of the thoracic feet. They may be considered
as homologous formations, and perhaps correspond with the gills of the
Phyllopuda and the basal branchial plates (epipodites) of Xclalia.

The pouch-shaped Amphipodan gills have already been considered.

SehizopuJa, — The branchial tufts of the EujJtausi/hc (Fig. 226), whose branches
are feathered, project freely from the basal joints of the protopodites of the thoracic
limbs into the surrounding water without being covered by the lateral lamella? of the
cephalo-thoracic shield. The brandling of the gills becomes increasingly complicated
from the anterior to the posterior thoracic limbs. On the most anterior thoracic
limb the gill is a simple appendage.

In Loplwyastcr we find 2 to 7 such gills. These consist of 3 feathered
branches, the upper one of which lies in a branchial cavity covered by the lateral
lamella of the cephalo-thoracic shield. The gills are said not to rise direct from the
basal joint of the protopodite, but close to it from the body. None the less they
should be considered as dislocated epipodial appendages.

The gills of the Dccvpoda (Figs. 227, 228) deserve more detailed description.
Over the sides of the thorax to the right and left there is always an arched extension
of the cephalo-thoracic shield, which, as branchiostegite (Fig. 227, M), covers a
respiratory cavity (kh) in which the gills (/„•) lie. The branchiostegite extends
ventrally to the points of insertion of the thoracic limbs, where the respiratory cavity
on each side communicates by means of a longitudinal slit with the surrounding
medium. AVe distinguish in the first place, according to their manner of insertion,
three sorts of gills — podobranchire, arthrobranchiffi, and pleurobranchia?. The podo-
branchise arise from the basal joints of the thoracic limbs, the arthrobranchise from
the articular membranes between the basal joints and the body, and the pleuro-
branchise from the body itself, but directly above the basal joints of the thoracic
limbs to which they belong. All gills are to be considered phylo-genetically as
epipodial appendages of the basal joints of the protopodites, the arthro- and pleuro-
branchire having moved from their original places. Again, we distinguish two sorts
of gills, according to their special form, viz. trichobranelme and pliyllobrancliire.
In the trichobranchise numerous branchial filaments stand round a common stem

cf <•£

^ eH

FIG. L^y.— Birgus latro. Diagrammatic transverse section in the region of the heart (after
Semper). M, Branchial or lung cover ; 7;, la-art ; /;, gills ; ah, respiratory cavity ; p, pericardium ;
>:k, branchial blood-oanals leading to the heart ; "i, u*, ";!, ('.4, lung or shell vessels leading from the
heart ; /';, respiratory tufts ; d, pulmonary vessels leading to the heart ; el\, the same near their
entrance into the pericardium.

or a common axis, like the bristles of a bottle brush. In the phyllobranchiae the
branchial filaments are small lamellse arranged in two rows on the stem, like the

CRUSTACEA— GILLS

329

barbs on the shaft of a quill. Phyllobranchise and trichobranchise, between which
there are many transitory forms, are not found together in the same species. Tricho-
branchise are found in the Macrura (with the exception of the Sergestidce, Carididce,
and the genera GeUa and Callianassa) ; phyllobranchise are found in all Anomura
and Brachyura, and in those Macrura which do not possess trichobranchife.

Podobranchise, arthrobranchise, and pleurobranchise may occur together, even on
the same thoracic segment. They undergo many modifications and degenerations.
The arrangement of the branchial apparatus in the various Decapodan genera and
species may be given in branchial formulae. We shall here give the branchial
formulae of Astacus fluviatilis and of Cancer pagurus.

Branchial Formula of Astaeus fluviatilis (Cray-fish),
after Huxley.

Arthrobranchise.

Thoracic segments and limbs.

Podo-
branchiae.

Pleuro-
branchice.

Total.

Anterior. ' Posterior.

VI. 1st maxillipede

0(ep.)

0 0

0

= 0(ep.)

VII. 2d

1

1 0

0

= 2

VIII. 3d

1

1 1

0

= 3

IX. 1st ambulatory foot

1

1 1

0

= 3

X. 2d

1

1 1

0

= 3

XI. 3d

1

1 1

Rudim.

= 3 + Rudim.

XII. 4th

1

1 1

Rudim. =3 + Rudim.

XIII. 5th

0

0 , 0

1 =1

6 + ep. + | 6 + 5 +

l+2Ru.

= 18 + ep.+2R.

Branchial Formula of Cancer pagurus (after Huxley).

Arthrobranchise.

Thoracic segments and limbs. branch*'^

i

Pleuro-
branchise.

Total.

Anterior.

Posterior.

VI. 1st ) J0(ep.)
VII. 2d }- maxillipede. 1 1

0
1

0
0

0
0

= 0(ep.)
= 2

VIII. 3d J Jl

1

1

0

= 3

IX. 1st 1 10

1

1

0

= 2

X. 2d 1 0

XL 3d I ambulatory Q

0
0

0
0

1
1

-t

= 1

XII. 4th 1 foot' 0

0

0

0

= 0

XIII. 5th J | 0

0

0

0 =0

j2 + ep.

+ 3

+ 2

+ 2 = 9 + ep.

While in other Decapoda the water enters the respiratory cavity through the
lower longitudinal slit, in the Brachyura the water passes in and out only through
certain small apertures, which are variously placed. Many Brachyura and Anomura
may live for a longer or shorter time or almost exclusive on land, and show various
adaptations which make it possible for them to retain water in the respiratory cavity,
• or to draw back into that cavity water which leaves it, or to breathe air direct. It
would lead us too far to describe all these adaptations in detail. We shall only
consider the respiratory organs of Birgus latro (Anomura), which lives in holes in the
earth (Fig. 229). The respiratory cavity of this animal falls into two parts, an

330 COMPARATIVE ANATOMY CHAP.

upper partly closed, and a lower more open, division, the lower edge of the branchio-
stegite being bent inwards and somewhat upwards. The reduced gills lie in the lower
division. The upper division holds air and functions as a lung. The integument of
the outer cover (branchiostegite) of this upper division carries a great number of
branched tufts, projecting into the air -filled cavity, and containing complicated
arrangements of vascular spaces. For the special relation of the vascular system to
the respiratory organs in Birgus, see the section on the circulatory system.

The manner in which in the Crustacea the blood which has become poor in
oxygen is conducted out of the body to the respiratory organs, and the blood rich in
oxygen conducted out of the respiratoiy organs into the body, will be described in
the section on the blood-vascular system.

II. The Integument.

The chitinous secretion of the hypodermis (body epithelium) which
we met with in the Annulata is even more strongly developed in the
Arthropoda. It covers the whole surface of the body and its append-
ages as a cuticle, serving not only for protection, but also as outer
skeleton (exo-skeleton) for the attachment of the muscles. This chitin-
ous envelope is very differently constituted in the different divisions of
the Crustacea, sometimes even in different parts of the same animal. In
most Decapoda and Stomatopoda, many AmpUpoda and Isopoda, and on
the shell -fold of Ostracoda and Cirripedia, however, the usually thick
chitinous cuticle becomes very hard and firm by deposits of lime salts
(carbonate and phosphate of lime) ; in certain Brachyura, Cirripedia,
and Ostracoda it is even as hard as stone. The cuticle is, however,
always comparatively thin, delicate, and flexible between the joints
and on the respiratory surfaces.

The constitution of the exoskeleton presents obstacles to the growth
of the body. These are overcome by moulting or ecdysis. Under the
old exoskeleton a new one is developed, which is at first soft and ex-
tensible ; but after the first has been thrown off it soon hardens. The
metamorphoses of the Crustaceans take place by means of several
repeated moults. In this process not only the, outer chitinous integu-
ment of the body, but the cuticular lining of the alimentary canal also
is removed and renewed.

The cuticle, of whose further structure we cannot here speak more
in detail, is penetrated by fine perpendicular pores.

Dermal glands, which take part in excretion, a fact which may be proved by
feeding with carmine, are very common, especially in soft-skinned Crustaceans.

It is hardly possible as yet to give a detailed comparative account of the structure
and distribution of the various forms of glands. But we may give a few cases which
for one reason or another are specially interesting.

The segmentally arranged ventral and leg glands of Branchipus consist of small
groups of dermal gland cells, found on the segments of the middle body. In every
segment a pair of ventral glands are found on the outer sides of the double ganglion of
the ventral chord, and a pair of leg glands in the basal lobes of the leg. The seg-
mental repetition, the character as dermal glands, the position (in the region of the
legs), and their rod-like secretion entirely justify the opinion that these glands are

CR USTA CEA—INTEG UMENT— MUSCULATURE

331

homologous with the spinning and setiparous glands of the Annelida and the coxal
glands of other Arthropoda.

In the basal joints of the 8 pairs of thoracic limbs of Nebalia hypodermal
glands have been observed ; it is probable that these perform excretory functions.
Their position recalls the leg glands of Branchipus.

We may here further mention — the dermal glands which occur in the basal joints
of certain limbs of the Phronimidce ; and the beautifully constructed dermal glands
in the limbs of the CoropJiiidce ; the unicellular dermal glands scattered in various parts
of the body of Orchestia ; the scat-
tered dermal glands of the Anisopoda
(Tanais, Apseudes] ; the hook glands
of Caprellidce ; and the so-called
cement glands of female Decapoda,
which lie on the ventral side of the
abdomen, and whose hardening secre-
tion serves for the attachment of the
eggs.

In Tanais and the Corophiidce the
secretion of the above - mentioned
glands hardens in water, and prob-
ably helps, by cementing together
foreign particles, to form the tubes

inhabited by these animals.

Special interest is claimed by the
uni- and multi-cellular dermal glands

lying scattered under the chitinous

cuticle of the Corycoeidae, (Copepoda),

because in them the connection of

the gland cells with nerve fibres can

be easily observed. A unicellular

dermal gland with cuticular duct,

which opens through a pore of the

chitinous integument of the body,

is in the Corycceidcc generally coupled

with a terminal ganglion cell lying

under a sensory seta. The nerve

which runs to this pair of cells divides

into 2 threads, one of which is con-
nected with the gland cell and the

other with the ganglion cell.

III. The Musculature.

FIG. 230. — Diagrams to demonstrate the
mechanism of the motion of the segmented body
in the Arthropoda. One larger segment (ct) and 4
smaller. The exoskeleton is denoted by black lines,
A COntinilOUS derniO-mUS- tne inte particular membranes by dotted lines. The

Cular tube such a s i s r h. a ra of Pr hinges between consecutive segments are marked a. t,
,UU« LUD6, SI Cnaracter- Tergal (dorgal) skeleton . ,f sternal (ventral) skeleton ;

IStlC 01 the worms in general, d, dorsal longitudinal muscles = extensors (and flexors

and Specially of the Amiulata in an uPward direction); v, ventral longitudinal

i« wnnfirrr ,« f^^ n™,,^ ' muscles = flexors. In B, the row of segments is

S Wanting m the Crustaceans, stretched ; In A, by the contraction of the muscle d,

and indeed in the Arthropoda bent upwards; in C downwards, tg, Tergal; $g,

generally. The development sternal interarticular membranes.

of the cuticular integumental covering into a much firmer exoskeleton

makes a greater localisation of the musculature possible.

332 COMPARATIVE ANATOMY CHAP.

We may assume, that in the homonomous segmented ancestors of
the Crustacea, whose conjectural organisation we have already in many
points diagrammatically sketched, 4 strongly -developed longitudinal
muscles ran through the body. Two of these muscles ran dorsally and
two ventrally on each side of the middle line. They were segmented
in correspondence with the segmentation of the body, and in such a
manner that the single muscle segments or myomeres lay interseg-
mentally, i.e. with one end attached to the integument of one segment,
and the other to the integument of the next following or next pre-
ceding segment.

We have no difficulty in assuming that the dorsal and ventral pair of longi-
tudinal muscle strands answer to the 4 similarly placed masses of longitudinal
muscles which occur in the Polychceta. We can find nothing, certainly, correspond-
ing with the circular musculature of the Annelida. Traces of the latter may
perhaps, however, be seen in the muscles which run at right angles to the longi-
tudinal muscles, and which are attached on the one hand to the integument of the
body, and on the other to the basal portion of the appendages which they move.

Setting aside at first the muscles belonging to the inner organs, we can arrange
the whole musculature of the body in three principal groups, viz. (1) the muscu-
lature of the body, (2) the musculature of the limbs, and (3) the musculature common
to both.

The formation and segmentation of the body and limbs in differ-
ent divisions varies so much in details that a comparative review of
the musculature cannot be attempted. We may, however, explain
the principle according to which the muscles are arranged, and per-
form their functions not only in all Crustaceans, but universally in all
Arthropoda. We have to bear in mind — (1) that the passive loco-
motory organ, the skeleton, is in the Arthropoda an exoskeleton,
which is in each segment of the body and in each joint of its extrem-
ities a chitinous tube ; (2) that the muscles lie on the inner side of the
skeleton, and are attached to the same from within ; (3) that the
muscles are stretched intersegmentally, i.e. between the consecutive
segments ; (4) that the chitinous integument between two segments is
thin and flexible, forming the interarticular membrane ; (5) that the
tubular exoskeleton of two adjacent segments or joints is hinged
together in each case at two transversely opposite points.

The arrangement and working of the musculature is illustrated by
Fig. 230, which shows us 5 segments, one larger (ct) and 4 smaller, in
vertical projection. The firm hard portions of the exoskeleton are
marked by strong outlines, the delicate and flexible interarticular
membranes (tg, sg) by dotted lines. The hinges between 2 consecutive
skeletal segments are marked a. Such a hinge is found on each side ;
in the projection the corresponding right and left hinges are seen as
one. A dorsal muscle (d) is attached to the larger segment (ct), and
runs through the smaller segments, being inserted in the dorsal or
tergal skeleton (t) of each by means of a bundle of fibres. A ventral
muscle (v) does the same on the ventral or sternal side (s).

CRUSTACEA— MUSCULATURE

333

The skeletal segments may be compared to a double-armed lever
whose fulcrum lies in the hinges. If the dorsal muscle contracts, it
draws the dorsal arm of the lever (the tergal portion of the skeleton)
in the direction of the pull towards the larger segment ; the tergal
interarticular membranes become folded, the ventral stretched, and
the 4 segments bend upwards (Fig. 230, A). If the ventral muscle
contracts, while at the same time the dorsal slackens, the row of
segments will be bent downwards (Fig. 230, C).

It is evident that the same effect would be attained if the dorsal

FIG. 231. — Two abdominal segments of the Cray-fish, diagrammatic, t, ti, t%, Tergal ; s, s^ s%,
sternal exoskeleton ; x, hinges ; b-c, t>i-ci, articular facets, which when the row of segments is
straightened take the position shown in B; ai-bi, a2-&2, c-aj, Ci-a2, e-di, ej-d2, interarticular
membranes ; tm, tergal ; sm, sternal longitudinal muscle. A, Row of segments nearly straightened ;
B, Row of segments bent ventrally by the contraction of the sternal longitudinal muscle,
stretching the intertergal membranes.

and ventral muscles, instead of collectively running from the smaller
segments to the larger segment, only ran from one segment to the next
following, as we assumed to be the case in the racial form, and as is
really the case in the thorax and abdomen of many Crustaceans (Branchi-
pus, many Isopoda, Amphipoda, etc.) The row of segments would then
likewise bend dorsally on the contraction of the dorsal myomeres, and
ventrally on the contraction of the ventral.

In reality only the ventral bending takes place in the Crustacea ; the compara-
tively weakly developed dorsal longitudinal muscles serve only for straightening the

334

COMPARATIVE ANATOMY

CHAP.

body, as we see from the movements of the abdomen of the Cray-fish (Figs. 231 and
232). In its normal position the abdomen is somewhat stretched, find lies more or
less as a straight posterior prolongation of the cephalo-thorax. Each dorsal (tergal)
skeletal segment (Fig. 231, A, t, ti, t.2) has its most anterior somewhat thinner portion
(b-c, bi-Ci) pushed some distance under the posterior edge of the preceding segment.

The interarticular membrane (c-a1} Ci-«2) being
bent backwards and outwards. The ventral
hard skeleton consists of relatively narrow
transverse segmental stripes (d-e, d^-e^ d^-c.^,
connected by large intersegmental membranes
(e-d^ e^d*), which in a state of rest are
somewhat stretched. The pair of dorsal
muscles (tm) are attached on the one hand
anteriorly to the lateral walls of the cephalo-
thorax (corresponding with the large segment
of our diagram, Fig. 230), and inserted, on
the other hand, by just as many pairs of
bundles as there are abdominal segments, to
the inner surfaces of the terga, a pair of
muscles being inserted in each.

The pair of ventral or sternal muscles (sm)
are attached anteriorly to the ventral side of
the cephalo-thorax to a row of processes of
the exoskeleton directed inwards, and partly
bound together by transverse ridges, which
roof over the thoracic portion of the ventral
chord and the sub-neural vessel. Posteriorly
the sternal longitudinal muscles are succes-
sively inserted on the inner sides of the sternal
skeleton of the abdominal segments. The
fibres of this muscle are twisted like the
strands of a rope.

If the pair of sternal or flexor muscles con-
tracts, the abdominal row of segments cor-
responding with our diagram bends ventrally
(Fig. 231, B) till, as is the case in swimming,
the telson touches the ventral side of the
cephalo-thorax. While in this position the
intersternal membranes of the abdomen are
FIG. 232.— Astacus fluviatilis. Longi- folded, the whole tergal integument is
tudinal section to represent the most stretched, and the tergal articular facets and
important muscles and their relation to . , , . , , , , -

the exoskeleton (after Huxley), em, Ex- "Particular membranes are drawn out from
tensor; fm, flexor of the abdomen; adm, under the terga which cover them, and
adductor of the mandibles ; cth, cephalo- come freely to the surface. If noAv the sternal
thorax; abrab^ abdominal segments; t, muscle slackens, and the pair of tergal or

££: iS; ££:.£L£ the cephal°- «*-« — <Fi?- ™>. f • ** f > •*

tracts, the abdomen is straightened, the terga

pass under one another like the tiles of a roof, and the intersternal membranes are
stretched. A dorsal bending of the body is impossible, first because th^e intersternal
membranes do not allow of further stretching, and secondly, because the terga can
only be pushed one beneath the other as far as the posterior limit of their articular
facets. This is best exemplified by the illustration (Fig. 231, A).

v CRUSTACEA— MUSCULATURE 335

The mechanism for the movement of the appendages is in principle the same as
that of the body. It is, however, evident that if the hinges of the joints of an
appendage were arranged at equal distances along two parallel straight lines the row
of joints of the extremity could only bend in one plane, as is indeed the case in the
row of body segments. But the two hinges of the consecutive joints are in reality so
placed as to make free movement possible. The exoskeleton of the body segments is
related to the exoskeleton of the basal joints of the appendages which it carries just
as a large joint of a limb is to a smaller. The muscles which on the one hand are
inserted in the exoskeleton of the first joints of the extremities, by preference attach
themselves on the other to the tergum of the corresponding segment.

The mechanism of the forceps (chelse) of the Cray -fish (Fig. 233) is as follows.
The forceps is formed by the two terminal joints of the ch elate foot. The last
joint but one is produced into a pointed process (zf). The last joint (eg) articu-
lates with it in the ordinary way by means of two opposite hinges.

Round the hinges the two joints are connected by means of a thin and flexible
interarticular membrane. Two muscles serve in the way shown in the illustration

FIG. 233.— Forceps of the large chelate foot of the Cray-fish, diagrammatic. A, closed. B,
Open, eg, Terminal joint ; vg, last joint but one with the pincer process zf; a, hinge on which the
terminal joint moves ; gh, interarticular membrane ; o, opening muscle (abductor) ; s, closing
muscle (adductor).

for moving the terminal joint. By the contraction of the smaller the movable joint
describes an arc away from the fixed process (Fig. 233, £), opening the forceps. If
the much stronger muscle which lies on the other side of the hinge contracts the
chela closes (A).

The muscles of Crustaceans are often attached to the exoskeleton
by means of sinewy and even chitinous terminal pieces. In the latter
case we can speak of an endoskeleton. Both arrangements serve for
the increase of the surface of attachment of the muscles.

In those Entomostraca in which the strongly -developed shell -fold
surrounds the whole body as a bivalve shell (Ostmcoda, Estheridce), a
strong shell-musele connecting the two valves transversely serves for
closing the shell. We find such a shell-muscle among the Malacostraca,
also in the Leptostraca (Nebalia).

As in all Arthropoda, the body musculature is transversely
striated.

336 COMPARATIVE ANATOMY CHAP.

IV. The Enteric Canal.

The intestine of the Crustacea has a simple straight course through
the body. The mouth lies on the ventral side of the head, bordered
and protected by an upper and an under lip (paragnatha) ; it is
surrounded by appendages which serve for the taking in of food
(mandibles, maxillae, and maxillipedes). The anus is found in the
terminal segment of the body. We must distinguish, according to
their ontogenetic origin and structure, 3 sections in the intestine : the
fore-gut, the hind-gut, and the connecting mid-gut. The fore- and hind-
guts, proceeding respectively from the ectodermal stomodaeum and
proctodseum of the larva or embryo, are lined internally by a chitinous
cuticle (intima). This cuticle, secreted by the hypodermis, passes at
the mouth and anus into the chitinous exoskeleton. Only the epithelium
of the mid-gut, which proceeds from the mesenteron, is of endodermal
origin. The mid -gut in almost all Crustaceans is distinguished by
the possession of diverticula which play the part of a hepato-
panereas.

As in other divisions of the animal kingdom, so also in many Crustaceans, pro-
nounced parasitism has brought about a degeneration of the enteric canal. In the
parasitic Cirripedia we find various stages of degeneration leading to the condition of
the Rhizocephala, in which an enteric canal is wanting not only in the adult animals,
but also in the free-swimming larvse. In the parasitic Isopoda the hind-gut with the
anus, and sometimes a large part of the mid-gut as well, may entirely disappear.

A. The Fore-gut.

There is a characteristic difference between the condition of the
fore-gut in Entomostraca and Malacostraca. In the former it is a simple
alimentary tube (oesophagus), which, passing between the cesophageal
commissures, runs dorsally, to pass later into the mid-gut which runs
backwards. In the Malacostraca, on the contrary, it falls into at least
two divisions, that which follows the buccal cavity, the ascending
narrower oesophagus, and the broader masticatory or fore-stomach
which lies in the head, and which leads to the mid-gut. Special
salivary glands entering the oesophagus seem generally to be wanting
in the Crustacea (such glands have only been observed in Cray-
fish), but glands emerging on the upper lip, in the buccal cavity, and
on the maxillae, are common; these are usually called salivary glands.
They probably belong to the category of the leg- and the other dermal
glands.

For the present we set aside the special modifications which
the fore-gut may undergo ; its structure is essentially the following.
Its wall consists of a layer of hypodermis cells, which rest upon

v CRUSTACEA— ENTERIC CANAL 337

an outer, often chitinous, basal membrane, and secrete, like the
hypodermis of the integument, internally, a chitinous cuticle (intima).
The fore -gut is embraced by circular muscular hoops, serving to
narrow its lumen ; groups of muscles, mostly paired, attached on
one side to the intestinal wall, on the other to a portion of the
neighbouring integument, effect its expansion and other move-
ments. The hypodermis of the oesophagus is often difficult to
discover.

Entomostraca. — The end of the oesophagus often projects in the form of a cone or
funnel into the first part of the mid-gut, somewhat as does the uterus of a mammal
into the vagina. This projection, which in Branchipus is bi-lobed, and set with small
cuticular papillse, perhaps .corresponds with the masticatory stomach of the Malaco-
straca.

The intima of the oesophagus is generally longitudinally folded, which permits of
its enlargement.

In the Ostracoda the oesophagus may be enlarged before passing into the mid-gut
into a so-called crop.

In the Leptodora (Daphnidce) two divisions have been distinguished in the fore-
gut, an ascending pharynx and a strikingly wide oesophagus running backwards.
This latter may perhaps belong to the mid-gut.

Malacostraca. — The possession of a masticatory or fore-stomach is characteristic
of the Crustacea belonging to this second principal division. It is found in the
Leptostraca, though in a somewhat simpler form than in the other Malacostraca. The
masticatory stomach is a spacious sac of varying form. It is chiefly distinguished
by the fact that its wall projects into its cavity in the form of definitely arranged
folds, ridges, valves, plates, lamellae, and other prominences, on which the chitinous
intima is specially strongly developed, so as together to form a very complicated
framework. The special form of the masticatory stomach, with its projections,
pouches, etc., and its chitinous framework, is extremely important in classification.
A more detailed account, however, would take far too much space. The masticatory
stomach generally falls into two divisions, an anterior cardiac division, into which
the oesophagus enters, and a posterior pyloric division, which opens into the mid-gut.
In the anterior division the food which has already been torn in pieces by the
oral appendages is still further cut and ground up by the masticatory framework,
and the digestion takes place chiefly in the posterior division, into which the
secretions of the glands of the mid-gut enter.1 The special formation of the wall
with its chitinous framework, in this second division, both hinders too rapid
passage of the food to the mid-gut, and prevents its return into the masticatory
stomach.

The parts of the masticatory framework are moved by suitably arranged muscles,
which are attached to the neighbouring integument.

In many parasitic Isopoda, which suck in food in a fluid condition, the masticatory
stomach is much simplified.

The intestine of the Entoniscidce will be described at the end of this section
(p. 341).

In the Dccapoda, on the anterior wall of the cardiac division of the mas-
ticatory stomach, are found two concretions, principally consisting of carbonate
and phosphate of lime ; these are the so-called crab's eyes or gastroliths. In

1 See section on the mid-gut of the Dccapoda (p. 340).
VOL. I Z

338

COMPARATIVE ANATOMY

CHAP.

the Cray-fish they are developed in summer and are largest just before the skin is
cast. During this process they reach the cavity of the masticatory stomach, are
there ground up, dissolved, and re -absorbed. It is highly probable that they
yield the calcareous material for the hardening of the new after the casting of the
old skeleton.

The whole intima of the fore-gut, to-
gether with the complicated chitinous
framework of the masticatory stomach, is
removed when the integument is shed ; it
is vomited from the mouth.

B. The Mid-gut.

This represents the endodermal portion
of the enteric canal, and its epithelium shows
clearly its endodermal origin even in the
adult, being differently constituted from the
g , ectodermal epithelium of the fore- and hind-
guts. The mid -gut again differs from
the two other divisions of the intestine
in having no chitinous intima answering
to the outer chitinous cuticle of the
body. It is nearly always distinctly
marked off from the fore- and hind-guts by
constrictions or valve -like arrangements.
Besides this its anterior end, and occasion-
ally its posterior end also, is marked by
the entrance of glandular diverticula which
represent invaginations of its wall. These,
among which the hepatopancreas of the
higher Malacostraca shows the greatest
development, we shall describe in the sec-
tion on the divertieula of the mid-gut.
The form and extension of the mid-gut
are very varied, while in the Entomostraca
and many Malacostraca it represents by far
the largest portion of the enteric tube,
running from the head to near the posterior
end of the body ; it has in other Mala-

FIG. 234.— Astacus fluviatilis, median section of the body, seen from the right side. The
thoracic feet and antennae of the left side are incompletely drawn (after Huxley), oa, Ophthalmic
artery (aorta cephalica) ; aa, antennal artery ; cth, cephalo-thorax ; o, lateral ostium of the heart ; h,
heart ; sa, sternal artery ; t, testes ; vd, vas deferens ; em, extensor muscles of the abdomen ; oaa,
upper abdominal artery (arteria abdominalis) ; db, abdomen ; hd, hind-gut ; an, anus ; t, telson ;
g, brain. (supra-cesophageal ganglion) ; 6/4, 4th thoracic foot (chelate foot) ; oe, oesophagus ; cs, cardiac
portion of the stomach ; ps, pyloric portion of the stomach ; bd, right aperture of the hepato-
pancreas into the stomach ; bm, ventral chord ; md, mid-gut ; Ir, liver (hepatopancreas) ; vdo, male
genital aperture ; pli, pl^, pl$, 1st, 5th, and 6th pleopoda ; uaa, lower abdominal artery ; fin, flexor
muscles of the abdomen.

v CRUSTACEA— ENTERIG CANAL 339

costraca (at least in the Decapoda, Isopoda, and probably also in the
Anisopoda) almost entirely disappeared as a special division of the
enteric tube. It is here used up in the formation of its strongly devel-
oped glandular diverticula, the hepatopancreatic tubes. In these cases
the hind-gut which proceeds from the ectodermal proctodseum represents
by far the largest portion of the enteric tube, running through the
body from the masticatory stomach to its hindermost end.

The fact that in the Isopoda and the Decapoda the whole enteric tube with the
exception of the point of entrance of the hepatopancreas, proceeds from the embry-
onic or larval stomodseum and proctodseum is ontogenetically established ; in the other
Malacostraca ontogenetic investigations as to the limit between mid- and hind-guts
have yet to be made. Up to the present time conclusions as to the extent of the
hind-gut have been based exclusively upon that of the chitinous intima.

The walls of the mid -gut and of its diverticula show the same
general structure. The distinct epithelium is placed on a basal mem-
brane and sometimes shows on the surface turned towards the lumen of
the intestine a (non-chitinous) cuticular limiting membrane. On the
outer side of the basal membrane the mid-gut and its diverticula are
encircled by hoop-like, regularly repeated, circular muscles, which are
seldom transversely striated. Longitudinal muscles are more rare, and
where they occur are not numerous. They lie on the inner side of the
circular muscles. In life we can observe, not only in the mid-gut but
also in its diverticula, rhythmical waves of contraction, which are often
very strong ; these also bring about, especially in the smaller Entomo-
straca without hearts, a sort of circulation of hsemolymph in the lacunar
system of the body.

The Mid-gut of the Entomostraea.

This generally falls into an anterior widened division (stomach,
chyle stomach, stomach-intestine) and a posterior narrower division
which we might designate the small intestine. The diverticula of the
mid-gut, present usually in a single pair, open into the former division.

The details of the arrangement of the mid-gut diverticula of the Entomostraea
are very varied. The two diverticula of the Branchiopoda are themselves sub-
divided. They vary greatly in size. In Apus they have lateral branches beset with
numerous glandular lobes. In the Cladocera (Fig. 192, p. 289) two short horn-like
diverticula are generally found, which are directed forwards. In the Ostracoda the
two diverticula are so long that they often project on both sides into the shell fold.
Diverticula of the mid-gut are wanting in a good many Copepoda ; in others they are
present singly or in pairs, simple, or else complicated by the formation of accessory
coaca. The arrangement of the two diverticula of the disc-shaped flattened Branchiura
(Argulus) (Fig. 195, p. 291) recalls that in many Branchiopoda. Each of the two
diverticula divides first into an anterior and a posterior branch, each of which again
branches, and the branches penetrate as far as the lateral edges of the cephalo-thorax.
In the generally longitudinally folded stomach of the non-parasitic Cirripedes diver-
ticula also not unfrequently enter ; Balanus has 8 diverticula, which may be branched
(B. pcrforatus}.

340

COMPARATIVE ANATOMY

CHAR

OS
km

The Mid-gut Divertieula of the Malaeostraea.

Among these we shall first distinguish those which enter at
the anterior end of the mid-gut from those which enter at the posterior
end. The former are universal. They correspond with the mid-gut
diverticula of the Entomostraca, and are generally called livers. Taking
into account their physiological activity, however, the name of hepato-
panereas is more suitable.

In the Leptostraca we find 4 pairs of hepatic tubes, 3 pairs of which, one upper,
one lateral, and one lower, are very long and run along the mid-gut, reaching far
into the abdomen. The short 4th pair stretches to the front of the head. The
tubes join on each side to form a short wide sinus, and these
sinuses, uniting at the two sides, enter by a common aperture
the posterior end of the masticatory stomach ventrally.

In the Arthrostraca also there are 1 to 3 pairs of diverticula
entering the beginning of the mid-gut which is occasionally
widened so as partly to surround the masticatory stomach.
Among these there are often 2 (in the normal Amphipoda 4)
tubes running backwards along the intestine, which, according
to their function and the structure of their epithelium, specially
deserve the name of a hepatopancreas.

The arrangement of the glands of the mid -gut in the Schizo-
poda, the Cumacea, and the larvae of the Decapoda is similar
to that in the Leptostraca and Arthrostraca. They are distin-
guished by 3 pairs of long hepatic tubes entering the most
anterior portion of the mid-gut. In the Stomatopoda, distri-
buted along the whole length of the mid-gut, there are 10 pairs
of branched tufts of hepatic tubes. The adult Decapoda are
distinguished by the possession of a paired hepatopancreas,
which to the right and left enters the posterior and lower end
of the masticatory stomach. By means of much branching the
liver assumes the character of a very voluminous tubular gland
filling a large part of the cephalo-thorax, and falling on each
side into 3 lobes — an anterior, a lateral, and a posterior. If
we examine only the extreme forms of the cells which unite to
form the epithelium of the Malacostracan liver, we can distin-
guish two sorts of cells : first, ferment cells, whose varied
secretions (which may be liquid or solid, coloured or colourless)
digest fibrine ; second, hepatic cells, whose fatty secretion con-
tain a colouring matter related to the gall pigment of verte-
brates. In consequence of these observations we cannot describe
the glands of the mid-gut simply as a liver, but rather as a
hepatopancreas. We cannot, also, carry out a sharp distinction
of the cells into ferment cells and hepatic cells ; many transition forms occur.

The glandular caeca which enter the posterior end of the mid-gut are found in
the Amphipoda, i.e. in the Caprellidce and the Crevettina. They occur as one pair,
except in Melita, which has only one such glandular tube. Physiologically they must
(in the Crevettina) be considered as urinary glands. Morphologically they cannot
be compared with the Malpighian vessels of the Tracheata, since they do not belong
to the hind-gut, but to the mid-gut.

In the posterior portion of the mid-gut of Nebalia there is at the inner side of

ft

FIG. 235. — Enteric
canal of Asellus aquat-
icus (after G. O. Sars).
oe, OZsophagus ; km,
masticatory stomach ;
d, mid-gut ; a, anal-gut
(rectum) ; I, hepatic
tubes (hepatopancreas).

v CRUSTACEA— ENTERIC CANAL 341

the dorsal enteric wall a longitudinal channel, which, at the end of the mid-gut, is
continued into a csecum, reaching into the anal segment posteriorly, and ending in
two lateral projections.

C. The Hind-Gut.

The hind-gut in Crustaceans is as a rule short and limited to the
last segment or segments of the body. Its epithelium is lined with a
frequently very strong chitinous intima. Its wall is almost always
provided with well-developed hoop -like circular muscles. Special
muscles or groups of muscles (dilators) are stretched between the
hind-gut and the neighbouring integument (and widen the former by
their contraction). In those Isopoda, Anisopoda, and Decapoda in which
the hind-gut is very long, taking the place of the small intestine of
other Crustacea, we find these dilators only at the posterior differen-
tiated division of the hind-gut, called the rectum.

Among the Amphipoda, Orchestia has a hind-gut which is strikingly
long for this group, reaching as far forward as into the 7th thoracic
segment.

The anus lies in the terminal segment — dorsally in the Entomo-
straca, ventrally in the Malacostraca.

Special glands or diverticula entering the hind-gut are wanting in (
the Crustacea.

In the Lynceidce among the Cladoccra, however, a glandular contractile csecum i
found ventrally in the hind-gut, which in Pleuroxus is prolonged into a long vermi-
form appendage wound round the gut.

In the Stomatopoda it is said that 2 glandular saccules enter the anal gut. As
other excretory organs are wanting in the adult animals these are supposed to have
excretory functions.

In the meantime it is not certainly proved that these glands of the Lynceidce and
Stomatopoda really belong to the hind-gut and not rather to the terminal division of
the mid-gut.

The widening of the hind-gut on the contraction of the dilators draws in water
through the anus, and subsequent contraction of the lumen of the gut expells it again
with faecal particles. It has been stated without sufficient foundation that these
sucking movements, at least in certain Entowwstraca (Phyllopoda), imply a respiratory
function in the hind-gut.

The chitinous intima of the hind-gut is ejected through the anus when the
integument is shed.

The peculiar modification which the enteric canal undergoes in the females of
many parasitic Isopoda can here only briefly be described, taking as an example
Portunion mcenadis (Entoniscidce}. The narrow oesophagus leads into a large
sac composed of two lateral sacs. The inner wall of the sac projects into its
lumen in the shape of numerous long processes covered with chitinous intima. In
this division of the gut, which has been called the cephalogaster, the absorption
of the food takes place. The cephalogaster is continued posteriorly into a second
division, the typhlosolis, whose wall, much thickened dorsally, projects into the
lumen in such a way that the latter in a transverse section is crescent-shaped,
the concave side being directed upwards. A strong cuticle lines the typhlosolis
and carries closely - placed long and stiff hairs which, projecting from the

342 COMPARATIVE ANATOMY CHAP.

opposite walls, mingle in such a way as to form a fine sieve apparatus, only
admitting of the passage of fluids. The typhlosolis, which is fastened to the body
wall by a pair of strong muscles, is followed by a third extremely muscular division,
which is called Rathke's organ, and carries on energetic rhythmic contractions.
A small tubular blindly closed terminal division, which is the only representative
of the mid-gut, receives the ducts of two large contractile hepatic tubes. A hind-
gut and anus are wanting. The whole intestine stretches only to the 3d thoracic
segment. Portunion mcenadis sucks the blood of its host (Carcinus). The peculiar
structure of the enteric canal, in which Rathke's organ and the cephalogaster alter-
nately contract and expand, seems adapted to this sucking process.

The enteric canal of the minute male of Portunion which lives in the body of the
female does not show the peculiar transformation which it undergoes in the female.
It is straight, and possesses two hepatic tubes, a hind-gut, and an anus.

V. The Nervous System.

The nervous system of the Crustacea is constructed on the same
type as that of the Annulate^ and must be derived from the latter.
The result of research in comparative anatomy and ontogeny justify
us in giving the following diagrammatic representation of its general
structure and original constitution. This scheme stands in direct
relation to the generalised plan of the segmentation of the body of
the Crustacean sketched above (p. 300). In the most anterior segment
of the body (head segment) the brain (supra-cesophageal ganglion),
consisting of two symmetrical lateral halves, lies in front of and over
the oesophagus, giving off nerves to the unpaired eye, the anterior
antennae and the frontal sensory organs (see below). Each of the
other segments of the body possesses two ganglia (a double ganglion)
lying very near each other in the ventral middle line. The two
ganglia of each segment (the two symmetrical halves of each double
ganglion) are connected together by means of a transverse commissure,
and with the corresponding ganglia of the preceding and subsequent
segments by longitudinal commissures. The two most anterior
longitudinal commissures which connect the double ganglion of the
second segment with the brain embrace the oesophagus. These are
the cesophageal commissures. The whole central nervous system thus
consists, as in the Annulata, of the brain and the segmented ventral
chord (ventral ganglionie chain), whose segmentation corresponds
with the segmentation of the body. From each double ganglion of
the ventral chord nerves proceed to the body musculature of the
segment to which it belongs, and to the musculature of the limbs
with which that segment is provided. There are therefore one
double ganglion for the 2d pair of antennae (in the 2d segment), a
similar one for the 2 mandibles (in the 3d segment), 2 pairs of
ganglia for the anterior and posterior pair of maxillae (in the 4th and
5th segments), and so on, a pair of ganglia in each segment for
the limbs which belong to it.

It must, however, be specially noted that the assumption of a special ganglion for

v CRUSTACEA— NERVOUS SYSTEM 343

the posterior antennae is not yet securely established. Such a ganglion would have
to be compared with the infra-cesophageal ganglion, and the segment corresponding
with it to the first trunk segment of the Annulata. In most Crustaceans, however,
the nerves for the posterior antennae do not arise from a distinct ganglion but from
the cesophageal commissures ; their places of origin in the higher Crustacea have
indeed moved as far forward as the brain. There are, however, many important
reasons, to be brought forward later, in favour of the assumption of an originally
distinct pair of ganglia and a special segment for the posterior antennae.

All the more important and striking deviations from the typical
Crustacean nervous system can be classed under the following
heads.

1. Approximation of the two ganglia of a double ganglion
by the shortening of the transverse commissure which unites them,
and finally the fusing of these two ganglia into one ; its composition
out of two lateral ganglia can, however, often be seen externally, and
always internally, on examination of its finer structure.

2. Approximation and contact of the longitudinal commissures
connecting the consecutive ganglia.

3. Approximation and contact of the consecutive ganglia by the
shortening of the longitudinal commissures. This may lead to the
fusing of the consecutive ganglia into one ganglionic mass, in which
the original composition out of several ganglia can sometimes be clearly
recognised ; at other times, however, this is very indistinct, or even
quite unrecognisable.

4. Longitudinal displacement and shifting of the ganglia,
generally from behind forward. Ganglia may be displaced from one
segment into another.

5. Shifting of the points at which nerves leave the ventral
chord. These displacements, however, apparently never affect the
real origin of the nerves in the centres of the nervous system.

6. Entire disappearance of ganglia. This is perhaps only to be
proved with certainty of the most posterior ganglia of the body.

All these changes go hand in hand with transformations of the
whole organism, especially with modifications in the segmentation of
the body and in the development of the extremities. In the young
stages of many Crustaceans it often happens that certain ganglia are
still separate which in adult animals are fused. We see from this how
important the knowledge of ontogeny is for a right comprehension of
the morphology of the nervous system.

The concentration of the nervous system (the fusing of originally
distinct ganglia to form larger ganglionic masses) can be observed in
most of the natural divisions of the Crustacea.

For the sake of clearness we shall in each group of the Crustacea
place the most conspicuously segmented nervous system in the fore-
ground, while the description of the deviating, concentrated, or
simplified nervous systems will be distinguished by the use of small
print.

344 COMPARATIVE ANATOMY CHAP.

Entomostraea.

The nervous system of the Phyllopoda (Fig. 236, D\ especially
that of the Branchiopoda (Branchipus, Artemia), best corresponds with
the scheme of the Crustacean nervous system given above. The
transverse commissures between the ganglia of the ventral chord are
tolerably long, and so the whole ventral chord has the character of a
ladder nervous system. The two ganglia of a double ganglion are
connected together by two transverse commissures. We must note as
specially important that the pairs of ganglia for the oral appendages,
i.e. for the mandibles and maxillae, have remained distinct. Behind
these 3 pairs of ganglia come (in Branchipus) the 1 1 pairs of ganglia
of the limb -carry ing segments, then the 2 pairs of ganglia in the
genital segments. Behind these, ganglia only occur as rudiments in
the two following segments.

The nerves for the posterior antennae arise out of the cesophageal
commissures, which are provided with a layer of ganglionic cells ; this
layer can the better be considered the ganglion of the 2d antennae,
since in front of the mandibular ganglion, and behind the oesophagus,
the cesophageal commissures are connected by a double tranverse
commissure, corresponding with the double transverse commissures of
the other ganglia of the ventral chord.

The nerves for the unpaired frontal eye, for the paired stalked
eyes, and for the anterior antennae spring from the brain.

In correspondence with the greatly reduced segmentation of the whole body in
the Cladocera (Phyllopod] its central nervous system is also much less pronounced.
The ladder-like ventral chord consists of 7 pairs of ganglia, the foremost of which
(infra-cesophageal ganglion) innervates the mandibles and maxillse, and the other 6
the 6 pairs of limbs. In front of the infra-cesophageal ganglion we find here a
transverse commissure connecting the cesophageal commissures. The nerve for the
2d antennae arises from the infra-cesophageal ganglion. In Leptodom the 6 ventral
ganglia in the adult animals are fused into one ventral ganglionic mass, while in the
young animals they are still tolerably distinct.

The nervous system of the Ostracoda deserves further investigation.
The ventral chord of Cythere which follows the brain and oesophageal
commissures is said to consist of an infra-oesophageal ganglion and 4
subsequent ventral ganglia. The infra-cesophageal ganglion is said to
show its composition out of two ganglia and to innervate the jaws,
while the 3 subsequent ganglia give off nerves to the limbs, and the
last ganglion nerves to the most posterior divisions of the body and
the genital apparatus.

In contrast with the above, the ventral chord of Halocypris appears much con-
centrated. It consists of an infra-cesophageal ganglion with nerves to the jaws and
maxillipedes, and a small ventral ganglion. Out of the latter arise 2 pairs of nerves,
which probably innervate the musculature of the limbs and the abdomen.

In the various divisions of the Copepoda the central nervous system

v CRUSTACEA— NERVOUS SYSTEM 345

shows several degrees of centralisation, from a more or less decentralised
condition to the almost complete fusion of brain and ventral chord to
form one ganglionic mass, pierced through by the esophagus. The
segmentation is most complete in the free-swimming Copepoda, from
which all the others are to be derived. In the Calanidce, for example,
we have (apart from the brain) a ventral chord consisting of 7 ganglionic
swellings which stretches more or less far into the abdomen.

In other free-living Copepoda the number of ganglia is reduced, and the abdominal
ganglia become small or disappear. But in the Corycceidce (Fig. 236, H) we already
have only one single ganglionic mass surrounding the cesophagus, from which nerves
radiate to the sensory organs, extremities, the musculature of the body, etc. The
nervous system in various delicate parasitic Copepoda shows a similarly con-
centrated, though partly also reduced, condition.

The nervous system of the Carp-lice (Argulidce, Fig. 236, G), which
are closely connected with the true Copepoda, is relatively highly
developed. The concentrated ventral chord consists of 6 ganglia with
much -shortened longitudinal and transverse commissures. The 4
posterior ganglia supply the 4 pairs of limbs, the 2 anterior the jaws,
maxillipedes, and clinging feet. At the points at which the cesophageal
commissures join the brain there are 2 ganglionic swellings, from which
nerves go to the 2d antennae.

Among the Cirripedes the nervous system of the Lepadidce is the
most richly segmented. They possess a brain, long oesophageal com-
missures, and 5 or 6 ventral ganglia. The nervous system of the so-
called Cypris-like larvae, that is, of those which develop into herma-
phrodite individuals, is similar.

The so-called complementary males of the Lepadidce, on the contrary, and their
Cypris-like larvce, only possess a cerebral ganglion (supra-oasophageal ganglion) and a
thoracic ganglion, which alone represents the whole ventral chord. In the Balanidce
the ventral ganglia are fused into one large ventral ganglionic mass. Degeneration
goes very far in the whole nervous system of the Rhizocephala (Sacculina, Peltogaster),
which are so much degenerated by parasitism ; we here find it in the form of one
single ganglion, from which various nerves radiate (cf. Fig. 248, p. 373). This
ganglion is said not to correspond with the larval supra-oasophageal ganglion, but to
arise anew in the development of the adult animal.

Malaeostraea.

I. Leptostraea. — The interesting genus Nebalia, which of all living
Malaeostraea stands nearest to their racial form, possesses an extra-
ordinarily richly segmented nervous system. If this pronounced
segmentation recalls, on the one hand, the nervous system of the Pliyl-
lopoda (Branchiopoda), it shows on the other (especially in the structure
of the brain) a decidedly Malacostracan character. The ganglia for the
posterior antennae are moved far forward on the cesophageal commissures,
and form, as in all Malaeostraea, the most posterior division of the brain.
The transverse commissure, which corresponds with them, however,

FIG. 236.

CHAP, v CRUSTACEA— NERVOUS SYSTEM 347

FIG. 23(5.— Central nervous systems of various Crustaceans. A, Of Euphausia pellucida
(after G. O. Sars). B, Of Astacus fluviatilis (after Vogt and Yung). C, Of Apseudes Latreillii
(combined from several figures by Glaus). D, Of Limnadia (after Klunzinger), anterior portion.
E, Of Asellus aquaticus (after G. 0. Sars). F, Of Maja squinado (after Milne Edwards). G, Of
Argulus Corregoni (after Glaus). H, Of Sapphirina Edwardsii (after Haeckel). gg, Brain; cm,
nerves of the paired eyes ; ua, unpaired eye with its nerve ; go, ganglion optjcum ; oj, nerve of the
1st antenna ; o2, of the 2d antenna ; sc, oesophageal commissures ; y, post-oesophageal tranverse
commissure of the same (commissure of the antennal ganglia of 2d antenna ? ) ; a*>g, ganglion of the
2d antenna (in D) ; md, maudibular ganglion ; mxi, mxv, ganglia of the 1st and 2d pairs of maxillae ;
I-V1II, thoracic ganglia; l>g, sub-oesophageal ganglion, consisting of several fused ganglia; 1-6,
abdominal ganglia ; s, sympathetic nervous system ; sg, ganglion of the same ; eg, commissural
ganglion ; TO, stomach (in F) ; ab, fused abdominal ganglia (in E). Jn G, g% signifies 2d ventral gan-
glion ; %i-?i4, nerves for the 4 pairs of limbs ; Iff, nerve for the clinging foot. In H, gm signifies the
ganglionic mass (brain fused with ventral chord) broken through by the oesophagus ; cl, corneal lens ;
I, lens ; bn, posterior strong nerves, the lateral pairs of which go to the limbs.

has remained separate behind the oesophagus and in front of the
mandibular ganglia. The ganglia of all the pairs of limbs, even
those of the oral limbs, have remained separate on the ventral
chord, which has 17 ganglionic swellings. We thus find, counting
from before backward : (1) a mandibular ganglion, (2) and (3) 2
maxillary ganglia; (4) to (11) 8 thoracic ganglia, and (12) to (17)
6 abdominal ganglia. It is a significant fact that in the larva the 6
abdominal ganglia are followed by a small 7th swelling, belonging to
the limbless 7th abdominal segment ; this, however, entirely disappears
at a later stage. This fact is rightly interpreted to mean that there
must originally have been more than 6 pairs of pleopoda.

In contradistinction to the ladder-like nervous system of the PTvyllopoda, the
ganglia in Nebalia are fused in the middle line to form a double ganglion, and the
longitudinal commissures are moved nearer each other. The latter are very much
shortened in the thoracic region in correspondence with the abbreviation of the
thoracic segments.

II. Arthrostraea, Anisopoda. — The richly segmented nervous
system of Apseudes (Fig. 236, C) is closely connected with that of
Nebalia. The brain and oesophageal commissures are followed by an
infra-oesophageal section, in which we can clearly distinguish the
separate ganglia for the mandibles, the two pairs of maxillae, and the
pair of maxillipedes. It is important to note that the maxillipedal
ganglion, which corresponds with the first thoracic ganglion of Nebalia,
is here more closely connected with the preceding ganglia, in keeping
with the commencing transformation of the first pair of thoracic feet
into a pair of maxillipedes. The 4 infra-oesophageal ganglia are
followed by 7 double ganglia for the thorax and 6 for the abdomen,
the last and largest of which is probably composed of two or more
fused together. The ganglion for the posterior antennae is moved
towards the brain, but the transverse commissure corresponding with
it 4s clearly visible behind the oesophagus, in front of the mandibular
commissure. The ganglia of the ventral chord are clearly double, and
connected by two separate longitudinal commissures.

In Tanais fusings and displacements seem already to have taken place in the
ventral chord. The ventral chord here has, it is said, only 12 ganglia.

348 COMPARATIVE ANATOMY CHAP.

Isopoda. — Among the true Isopoda several genera (such as Sphceroma,
Idothea, Glyptonotus) are, in the rich segmentation of their nervous
systems, closely connected with Apseudes. In Sphceroma we even find a
7th abdominal ganglion. The double nature of the central nervous
system is everywhere more or less distinctly marked. The transverse
commissure corresponding with the ganglia of the 2d pair of antennae
seems to be more or less completely fused with the mandibular
commissure.

In many Isopoda there is fusion in the ventral chord and displacement and re-
duction of the pairs of ganglia. The mandibular, maxillary, and maxillipedal ganglia
in the first place fuse to form an infra-cesophageal ganglionic mass. Then a reduction
in the number of the ganglia takes place chiefly in those of the abdomen. In some
Isopoda 5 abdominal ganglia occur, in others (Porcellio, Oniscus, Asellus, Fig. 236, E)
we find as the remains of the abdominal ventral chord only one ganglionic swelling
attached to the last thoracic ganglion, and in others even this is wanting. The
number of the separate thoracic ganglia is less frequently reduced.

In those Entoniscidce (Portunion mceimdis) which are specially strongly modified
through parasitism we find, besides the brain (which is everywhere retained), 2
thoracic ganglia and one abdominal ganglion at some distance from them, under
the heart ; while in the not less strongly modified parasitic Bopyridce 7 thoracic
ganglia are said to occur.

Amphipoda. — Here in all cases we have a fusing of the anterior
ganglia of the ventral chord, so that the nervous system no longer
shows at any point that segmentation which we meet with in many
Isopoda. In the nervous system of the Gammaridce we can still
distinguish, apart from the brain, an infra-oasophageal ganglion
consisting of several fused ganglia, and further 7 thoracic ganglia in
the 7 free thoracic segments, and 4 abdominal ganglia in the 4
anterior abdominal segments.

In Phronima the centralisation goes still further. There are only 5 thoracic
ganglia behind the infra-cesophageal ganglionic mass, which consists of 6 fused
ganglia. The last of the 4 abdominal ganglia is formed by the fusion of 3 ganglia
which are still separate in the embryo. Not only the points of divergence of the
nerves for the 2d antennse, but also those of the nerves of all the oral appendages
are shifted forward on to the cesphagea'l commissures. The nervous system of
many Hyperidce is similar to that of Phronima. In others, however, concentration
goes still further, as not only the last 2 thoracic but also the last 2 abdominal
ganglia may fuse together. In the most extreme cases we thus find only an infra-
cesophageal ganglionic mass, 4 thoracic and 3 abdominal pairs of ganglia. In the
Caprellidce the reduced abdomen retains no ganglia. Besides the brain and the
sub-cesophageal ganglion which supplies the mandibles, maxillse, and maxillipedes,
we find 7 thoracic ganglia, the 7th of which lies behind the 6th in the last thoracic
segment but one. Three small ganglia belonging to the abdomen follow close
behind the 7th thoracic ganglion. In young animals 4 more pairs of abdominal
ganglia begin to form, and then fuse with the 3 small ganglia of the adult animal
above mentioned.

III. Thoraeostraea. — In many Schizopoda (Euphausia, Fig. 236,
A, Boreomysis) all the ganglia for the oral and thoracic limbs, 11 in
number, seem to have remained separate.

v CRUSTACEA— NERVOUS SYSTEM 349

In others the number is evidently reduced by the fusing of originally separate
ganglia. In Gnaihophausia we find besides the infra-oesophageal ganglion (which
most probably consists of the fused mandibular and maxillary ganglia) 8, in Eucopia
however only 6, thoracic ganglia. The most anterior thoracic ganglion has in this
latter case probably united with the infra- cesophageal ganglion, and the last thoracic
ganglion with the last but one. This is made probable by the fact that in the last
thoracic segment no ganglion occurs. In Mysis (relicta) the thoracic ganglia are
even said to be fused into a longitudinal strand. The cesophageal commissures are in
many Schizopoda (Euphausia, Boreomysis] connected directly behind the oesophagus
and in front of the most anterior ventral ganglion by a transverse commissure, which
perhaps corresponds with the commissure of the ganglia of the posterior antennae often
mentioned above.

All the Schizopoda seem to possess 6 abdominal ganglia. The
nerves which supply the body musculature of the abdomen diverge
from the longitudinal commissures half-way between 2 consecu-
tive ganglia. This arrangement seems to be characteristic of all
Thoracostraca.

In the Cumacea (Diastylis) the ventral chord consists of 16 pairs of
ganglia ; the three most anterior of these, which have moved near each
other, supply the oral limbs; 7 thoracic, and 6 abdominal pairs of
ganglia follow.

The nervous system of the Stomatopoda in its segmentation shows
very clearly a close relation to the metamerism of the body. In the
cephalo-thoracic portion of the ventral chord only the 3 ganglia of the
3 most posterior thoracic segments have remained separate, i.e. of
those segments which, uncovered by the cephalo-thoracic shield,
carry the biramose ambulatory feet. All the other preceding ganglia
are united into a large infra-oesophageal ganglion. The cesophageal
commissures are very long, arid show behind the oesophagus the
transverse commissures often mentioned above. The 3 posterior
thoracic ganglia are followed by the 6 abdominal ganglia which are
characteristic of the Thoracostraca.

Deeapoda. — We here find many grades of concentration, from the
still tolerably richly segmented nervous system of the Macrura to the
nervous system of the Brachyura, in which all the ganglia of the
ventral chord have fused into one single thoracic ganglionic mass.
Taking as a type of the Macrura the Cray-fish, Astacus fluviatilis (Fig.
236, B\ the highly developed brain gives off the nerves for the eyes, the
anterior antennae, and from its posterior part for the posterior antennae.
The cesophageal commissures are of considerable length and connected
behind the oesophagus by a transverse commissure. In the course of
each cesophageal commissure lies a small ganglion, the so-called com-
missural ganglion. From these 2 commissural ganglia various nerves
diverge, among which the visceral nerves and the nerves of the
mandibles are to be specially noted. The latter indeed have their roots
in the infra-cesophageal ganglion, but are united with the cesophageal
commissures as far as the commissural ganglia. The cesophageal
commissures enter an infra-oesophageal ganglionic mass which consists

350 COMPARATIVE ANATOMY CHAP.

of 6 united ganglia, i.e. of the mandibles, 2 maxillae, and 3 maxillipedes.
The last swelling of this infra -oesophageal ganglionic mass (the 3d
maxillipedal) is pretty clearly marked off. The 5 distinctly separate
large thoracic ganglia for the 5 posterior thoracic segments and their
extremities (the ambulatory feet) follow, and of these the 4th and
5th ganglia are very near each other. In the abdomen we find 6
ganglia, the last is the largest, and, as in all Malacostraca, may well
represent 2 or even more originally separate ganglia. In the Cray-fish,
as in nearly all Thoracostraca, the 2 ganglia of the originally double
ganglion and the longitudinal commissures between the consecutive
ganglia of the ventral chord are so closely connected in the middle
line that their double nature is not outwardly perceptible. The
longitudinal commissures separate from each other only between the
penultimate and antepenultimate thoracic ganglia (6th and 7th), so as
to let the sternal artery pass between them.

In a few Macrura the 2 most posterior thoracic ganglia are fused, or there is a
close approximation of all the cephalo-thoracic ganglia (Carididce, Palinurus). Here,
however, the 6 abdominal ganglia remain separate, while in the Paguridce only 1
abdominal ganglion is present, in correspondence with the great reduction of the
abdomen. We finally come to the Brachyura (Fig. 236, F), in which, in
correspondence with the great reduction of the abdomen and the concentration of the
cephalo-thorax, the whole ventral chord is fused together into a great thoracic knot,
from which numerous nerves radiate out to all sides.

Sympathetic Nervous System.

This seems to occur in all Malacostraca, at least it has been observed
with considerable uniformity in representatives of the three principal
divisions, the Leptostraca, Arthrostraca, and Thoracostraca. In the
Cray-fish a nerve with double roots arises on each side out of the
commissural ganglion, proceeds forwards to the upper lip and mounts
upwards at the sides of the oesophagus. The two nerves unite on the
upper side of the stomach to form a median nerve which swells into a
ganglion. From this ganglion a branching nerve runs backwards and
spreads out in the wall of the stomach and gives off branches to the
liver and also probably to the heart. The sympathetic nervous
system is further connected by an unpaired nerve with the posterior
portion of the brain.

So highly developed a sympathetic nervous system seems wanting
in the Entomostraca. But it is noteworthy that in the Phyllopoda
(Branchipus) a nerve arises on each side out of the cesophageal com-
missures which runs to the upper lip. The 2 nerves unite to form a
labial ring, which is connected with a median ganglion and gives off
nerves to the upper lip, the muscles of the oesophagus, etc.

Structure of the Brain. — The brain of the Crustacea rises in the Malacostraca to a
very high degree of complication. This complication, which reaches its highest point
in the Decapoda, is seen in the complicated arrangement of the ganglionic cells and

v CRUSTACEA— SENSORY ORGANS 351

of the courses of the fibres, and shows itself externally by the formation of lobes. It
is probable that the brain of the ancestors of the Crustaceans contained the centres
for the unpaired frontal eye, for the anterior antennae and perhaps also for the frontal
sensory organs, together with the fibrous connections between these centres them-
selves, between these centres and the cesophageal commissures, and lastly the
anterior transverse connections of the cesophageal commissures. A higher complica-
tion is caused in most of the Crustacea now living (all Malacostraca and many Ento-
mostraca), first by tie occurrence of the paired eyes, and second by the fact that the
originally infra-cesophageal ganglia of the posterior antennae unite with the brain from
behind. We can in these cases distinguish three principal regions in the brain.
First, an anterior region (fore-brain), with the centres for the unpaired eye (where
this persists) and for the paired eyes. The optic nerves of the latter, whose
fibres in the brain of the higher Crustacea form a chiasma, enter on each side a
frequently very large optic ganglion, which is the largest accessory lobe of the fore-
brain. Besides this the fore-brain has, especially in the higher Malacostraca, other
lobate formations. Second, a mid-brain which adjoins the fore-brain, and contains
the centres for the nerves of the anterior antennae. Third, the hindermost region
of the brain (hind-brain), which is formed by the ganglia for the posterior antennae,
whose transverse commissure is to be found behind the oesophagus, where it has,
as already described, often remained as a separate transverse commissure between
the cesophageal commissures.

Neurochord strands or giant nerve tubes, like those with which we became
acquainted in the Annulata, occur in the ventral chord of the Thoracostraca. An
intermediate nerve also, of varying extent, has here and there been observed in the
ventral chord.

VL The Sensory Organs.
A. Eyes.

These are, as a rule, well developed in the Crustacea, and often
show a high degree of complexity, especially in the Malacostraca,
occasionally also in certain Entomostmca. Visual organs are either
wanting or very much reduced in the adult condition of most of the
parasitic and attached Crustacea (Cirripedia), also in many deep-sea
forms and in others which live in dark places. Setting on one side
a few divergent forms of Crustacean eyes, we can distinguish 2 kinds,
which may occur simultaneously in the same animal : the unpaired
frontal eye (accessory eye) and the paired lateral eyes (principal eyes).
Both belong to the head. The unpaired eye lies above the brain, the
paired at its sides.

The unpaired eye occurs in the young larval forms of all Crustaceans
(Nauplius eye) : it is always retained in the adult Entomostraca, some-
times well developed, sometimes in a reduced condition. In the
Malacostraca it degenerates in the course of development. From the
universal distribution of the unpaired eye in the Entomostraca and
young Malacostraca we may conclude that it was present in the
ancestors of the Crustaceans.

Paired eyes are found in all Malacostraca and many Entomostraca.
They are either movable stalked eyes or fixed sessile eyes. The first
may be imagined to arise out of unstalked eyes by the raising of that

352 COMPARATIVE ANATOMY CHAP.

part of the head immediately surrounding the eye, by its articula-
tion, and further growth into a stalk. Formerly the stalked eyes
were often erroneously considered to represent a pair of extremi-
ties. In the development of the Crustaceans the paired eyes always
appear much later than the unpaired eye, and we have reason for
assuming that the unpaired eye is phylogenetically older than the paired.
The following is a short review of the occurrence ar.d distribution of
the paired or principal eyes.

Entomostraca. — Phyllopoda : in the Estheridce, and Apusidce the paired eyes have
moved towards each other in the middle line. The Branchiopoda have movable
stalked eyes. The two principal eyes of the Cladocera have fused in the middle
line to form a trembling frontal eye, which, however, is wanting in Monospilus.
Ostracoda : the Cipridinidce possess besides the unpaired eye paired movable
lateral eyes. In the Cypridce and Cytheridce also paired eyes occur, which may fuse
together to form one unpaired eye. Whether these correspond with the paired eyes
of other Crustaceans is not known. Copepoda : the Carp-lice, which are nearly
related to the Copepoda, possess besides the unpaired eye 2 large lateral eyes. In the
true Copepoda the latter are generally wanting, but paired eyes do occur, e.g. in
Pontellidce, and these perhaps correspond with the lateral eyes of other Crustacea.
The paired eyes of the Corycaeidce on the contrary do not easily admit of such a com-
parison. It is, however, not improbable that the racial forms of the Copepoda
possessed paired compound eyes, which have been again lost. Cirripedia : in the
adult animals the paired eyes are wanting throughout, but on the other hand the
Cypris-like larvce of the attached forms are provided with large lateral eyes.

Malacostraca. — Lateral eyes occur everywhere. They are stalked in the
Leptostraca and all Thoracostraca except the Cumacea. In the last case the sessile eyes
are generally fused in the middle line ; they may, however, be altogether wanting.
In the Arthrostraca, which have also been called the Edriophthalmata (in contra-
distinction to the other Malacostraca, which are known as the Podophthalmata), the
eyes are sessile. The facts that movable stalked eyes occur in the Phyllopoda
(Branchipus), and that the eyes of Leptostraca (Nebalia), which in every way stand
nearest the racial form of the Malacostraca, are similarly stalked, make it appear
probable (other facts also being taken into consideration) that the paired eyes of
the Arthrostraca were once stalked.

Among the Amphipoda the Phronimidce show peculiarities. They possess two
pairs of compound eyes, one frontal pair and one in the region of the cheek. The
two eyes of the same side, however, must have proceeded by division from the single
eye of the Amphipoda which in the Hyperidce is very large superficially.

Structure of the Eyes. — The unpaired eye was formerly described
as an x- shaped eye -spot, with or without a refractive body. On
account of its general distribution in the Entomostraca, it is also called
the Entomostracan eye. Its structure will be best illustrated by
means of an example.

The frontal eye of Calanella mediterranea (Fig. 237), a free-living
Copepod, consists of 3 single eyes united together, 1 unpaired median
and ventral, and 2 lateral and dorsal. Each single eye is composed
of a pigment cup and a strongly refractive transparent " lens " laid in
and on it. The term "lens" is, however, not applicable. It is
composed of several cells, each of which is connected, whether at its

CRUSTACEA— SENSORY ORGANS

353

outer or inner side is not yet quite certain, with a fibre of the optic
nerve, and must therefore be considered as a retinal cell.

The great similarity in structure between
such a single eye and the eyes of Platodes
should not be overlooked. The three -fold
structure of the unpaired Crustacean eye seems
to be characteristic. Occasionally, e.g. in
Branchipus, 3 separate nerves leave the brain
to run to the 3 single eyes.

The structure of the paired lateral
eyes of the Crustacea (stalked and un-
stalked) is much more complicated.
We have here the compound eye so
characteristic of the Arthropoda. Even
though, in single divisions of the
Crustacea, it presents important modi-
fications and complications in its
structure we nevertheless evidently pia 237_Eye of Calanella mediter

have to do (With a tew exceptions to ranea ? juv, from below (after Gren-
be mentioned later) with homologous acher). p, Pigment plates of the paired ;plt

visual organs. Let us take for descrip- ^ti'— S —' ""* ''
tion the paired eye of Brancmpus,

which presents in a tolerably simple manner the typical structure
of the compound eye. The movable stalk of the hemispherical eye of
Brancliipus (Fig. 238, B) contains the optie nerve ; this swells
in the stalk into a ganglion, the ganglion optieum, which must
be reckoned as belonging to the brain. The optic ganglion is
followed at the distal end near the base of the eye by a second
ganglion, the retinal ganglion. Nerve fibres radiate towards the eye
from the nerve cells of this retinal ganglion. The eye itself is
separated from the cavity of the eye stalk by a thin basal membrane.
The nerve fibres which radiate from the retinal ganglion penetrate this
membrane to enter the retinal cells immediately on the other side
of it. The eye represents the half of a hollow sphere with thick
walls, whose outer spherical surface corresponds with the outer surface
of the eye, and whose inner (concave) surface corresponds with the
basal membrane. It consists of numerous closely packed single eyes
arranged radially. In each single eye or ommatidium (Fig. 238, E)
we distinguish three chief constituents : —

1. The Retinula, i.e. that portion of the whole retina of the
compound eye which belongs to each of the single eyes. This is the
proximal portion coming next to the basal membrane.

2. The crystal cone, and

3. The hypodermal elements with the super jacent chitinous
cuticle, the cornea of the Arthropodan eye.

A. The

VOL. I

Retinula consists of 5 lono:

cells regularly grouped
2 A

354

COMPARATIVE ANATOMY

CHAP.

round a central axis : into the proximal ends of each of these cells
a fibre from the optic nerve enters. The central axis is a tubular
rod called the rhabdom. The 5 cells of the retinula contain pigment
in the immediate neighbourhood of the rhabdom, and this pigment
occurs in such large quantities in the thicker nucleated portions of the
cells, that in each retinula a distal pigment layer is formed. Above

FIG. 238.— Compound Crustacean eye. A, 2 single eyes (ommatidia) of Palaemon Squilla.
The pigment is removed from the ommatidium on the right hand. C, Isolated crystal body
of an ommatidium, consisting of 4 pieces. D, Transverse section through a retinula, about
the middle of its length ; re, retinal cells ; rh, rhabdom consisting of 4 pieces. B, Section
through the stalked eye of Branchipus. E, 2 ommatidia of the same animal, on a larger
scale (figs. A, C, D, after Grenacher, B and E, after Claus); c, cornea (cuticle) ; d, corneal lens ;
hy, hypodermis cells ; fc, crystal cone ; fcj, outer crystal body ; Jcz, crystal cells ; p, pigment ; pj, in
fig. A, pigment strands running between the retinulse from the layer of nerve fibres ; re, retinula ;
rh, rhabdom ; nf, nerve fibres ; 6m, basal membrane ; m, muscle ; rg, retinal ganglion ; go, ganglion
opticum ; no, nervus opticus.

this layer, terminal unpigmented portions project. These 5 ends
together embrace the proximal ends of

B. the refractive crystal cells. These are 4 in number and
together form a cone, which contains in its distal portion a solid crystal

v CRUSTACEA— SENSORY ORGANS 355

body secreted by the crystal cells. The layer of the crystal cells of
the compound eye is covered by

C. The transparent smooth ehitinous integument (cornea) with
subjacent hypodermis cells, a continuation of the general body
integument.

The principal distinctions to be pointed out between the
stalked eye of the higher Crustacea and that of Branchipus are in the
cornea and the optic ganglion. In the former the cornea (ehitinous
cuticle) is somewhat thickened over each single eye, forming for each
a eorneal facet or corneal lens, which is convex either on its inner or
outer side or on both sides (Fig. 238, A, d). The cornea then appears,
when viewed from the surface, to be divided into regular polygonal
areas, each of which corresponds with a corneal lens of a single eye.

Again, whereas the ganglion opticum is simple in Bmnchipus, in
the stalked eyes of the Malacostraca it falls into 3 ganglia also placed
in the eye stalk.

There are many important variations in detail in the structure of the Crustacean
compound eye ; these chiefly concern the number of cells in a retinula, the number
of crystal cells in a single eye, the number of single eyes in the whole eye, and the
specific arrangement of the elements. The Decapoda and Isopoda possess 7 retinular
cells, the Branchiopoda and Amphipoda 5, the Schizopoda 4. The single eyes of the
Cladocera have 5 crystal cells, the Decapoda and the Branchiopoda 4, the Isopoda,
Amphipoda, and Schizopoda 2. In the Isopoda only a few single eyes are found
which are not closely packed (4 in Aselhis, 20 in Porcellio}. The isolated corneal
lenses (not flattened into polygonal facets) are here biconvex.

Each of the paired eyes of the Corycaeidce consists of one single eye, which in
Corycaeus is strikingly large and long. It is in many respects markedly different from
the ommatidia of the compound eyes of other Crustacea.

In the Euphausidce among the Schizopoda there are, besides the two
compound stalked eyes, other " accessory " eyes. These are found on
the basal joints of the second and penultimate pairs of thoracic feet,
and further, one in the ventral middle line of the abdomen between the
pleopoda of each of the 4 anterior segments. Whether these organs
belong to the category of visual organs is very doubtful ; we only
know for certain that they are luminous.

B. Other Sensory Organs.

Among the other sensory organs of the Crustacea the most
widespread are the tactile, and what are generally supposed to be
olfactory, organs. In many Entomostraca we find, in addition, the
so-called frontal sensory organs of unknown physiological significance.
Auditory Organs occur in all the Decapoda, and have also been
observed in isolated cases in other divisions. Other structures which
have been described as sensory organs with unknown functions must
be passed over in silence, because of their sporadic occurrence and also
because too little is known about them.

356

COMPARATIVE ANATOMY

CHAP.

1. Specific Organs of Touch. — The points of the limbs, especially
of such as serve for locomotion or for holding food, possess a finer
sense of touch than the other parts of the surface of the body.

Special Tactile Setae are the principal organs of touch. These are
found chiefly on the antennae, but also on other extremities, and occa-
sionally also on the body itself. These setae are distinguished from
other setae, spines, etc., whose function is almost entirely mechanical,
by the fact that one or more ganglion cells lie at their bases connected
by nerve fibres with the general nervous system (Fig. 239, D).

2. Olfactory Organs are found in the shape of pale delicate knobs,
filaments, tubes, or points (Fig. 239, A, B), which are often grouped
in bundles or transverse rows, and occur on the anterior antennae.

FIG. 239.—^!, 7, 8, 9, joints of a 13-jointecl flagellum of the anterior Antenna of Nebalia (male)
with the olfactory tubes. B, Two Olfactory tubes, more strongly magnified. C, Feathered sen-
sory seta (auditory hair) from the antepenultimate pair of thoracic limbs of Apseudes, with cuti-
cular basal capsule. D, Tactile hair (ib) of Branchipus. c, Body cuticle ; hy, hypodermis cells of
the seta ; gz, ganglionic cell ; n, nerve fibre (after Glaus).

Less frequently similar structures are also found on the second
antennae. They always occur in greater numbers in the male than in
the female. The chitinous cuticle of these olfactory processes is thickest
at their bases ; towards their free end it is thin and delicate. At the
base of each olfactory process a nerve fibre enters without forming a
ganglionic cell, and continues into the interior of the process, running
through it and filling it to its free end. The nerve fibres originate in
ganglionic cells which, lying in the same or preceding joint of the
antenna, belong to the antennal nerve.

Whether the so-called Calceoli of the Amphipoda are olfactory
organs, or whether, perhaps, they represent a kind of auditory organ,
must remain undecided.

3. Frontal Sensory Organs. — The characteristic position of these
organs which occur in one pair is the frontal region very near the
brain. They are projecting filaments, cones, rods, or other cuticular

v CRUSTACEA— SENSORY ORGANS 357

appendages, into which 2 nerves, the frontal nerves, enter, generally
forming ganglionic cells. In Branchipus, in place of the cuticular pro-
cess, there is only an inconsiderable 'thickening of the cuticle with
large hypodermis cells surrounded by ganglionic cells lying under it
Frontal sensory organs have been observed not only in the Ento-
mostraca and Entomostracan larva, but also in the Malacostmcan larva
(i.e. when it is a Nauplius), and from this we are justified in concluding
that they were original structures present in the Crustacean racial
form.

4. Auditory Organs of the Deeapoda. — These lie on the basal
joint of the anterior antennae (antennulse). In all Deeapoda they
occur as pit- shaped depressions of the chitinous integument, which
generally remain open, but in a few cases (Hippolyte) may close and
form a vesicle. In the open auditory vesicles the aperture is often
covered by a compact row of setse projecting from one of the edges,
less frequently by a thin projecting fold. The auditory pits contain
sand particles taken in from outside, which probably function as
otoliths, like the concretions of fluorcalcium in the closed auditory
vesicles. At the base of the auditory pits, or on the inner wall
of the auditory vesicles, feathered hairs arise; these are (1) otolith
hairs, which carry the otoliths, and (2) free hairs, projecting
into the cavity of the auditory pit. The distinctly marked, swollen
base of the auditory hair is extremely delicate and thin-walled,
and permits a considerable oscillation of the hair in response to the
waves of sound. The auditory nerve, which branches from the
antennal nerve and has its root in the brain, first sends off fibres to
single ganglionic cells. Filaments from these ganglion cells enter the
auditory hairs at their bases and are attached near their points to rod-
shaped bodies.

It is usually said of the two closed auditory vesicles of the Mysidce (Schizo-
poda), that they lie in the tail. More correctly, they lie in the inner lamella (endo-
podite) of the last pair of pleopoda, which, together with the telson, form the caudal
fin. Their structure does not deviate essentially from that of the Decapodan
auditory organs. They are innervated from the last abdominal ganglion.

Oxycephalus (Amphipoda) possesses two auditory vesicles lying above the brain
and containing otoliths.

There is some justification for classing the sensory organs of the Deeapoda, Schizo-
poda, and Ampkipoda above described as auditory organs ; but we must not fail to
mention that more recent experimental investigations greatly support the view that
they also serve for regulating the position of the body, and for maintaining its
equilibrium.

Feathered setae, which in structure show great agreement with the auditory hairs
of the Deeapoda, occur on the antennae and also in other parts of the body in many
Malacostraca, and have often been considered as auditory organs. This is, however,
still an undecided point. It is, however, probable that the auditory organs of the
Deeapoda have developed phylogenetically by the localisation of feathered sensory
hairs and by the pit-like depression of those parts of the integument which carried
them. The utilisation of foreign particles of sand as otoliths also supports this view.

358 COMPARATIVE ANATOMY CHAP.

VII. The Blood- Vascular System and the Body Cavity.

In the Crustacea (and in the Arthropoda generally) we find no
closed blood-vascular system. Those portions of the circulatory
system which are provided with special walls, stand in open com-
munication with blood lacunae. These lacunae have no special walls,
but are only spaces between the various organs of the body, and repre-
sent the body cavity.

Scheme of the Circulatory System. — According to what is now
known of the Crustacea we may imagine the circulatory system in the
racial form of these animals to be essentially as follows : A contractile,
tubular, dorsal vessel (heart) runs longitudinally through the body
in the middle line above the intestine. The direction of the flow of
blood in this dorsal vessel is from behind forward, as in the dorsal
vessel of the Annulata. In each body segment the dorsal vessel
possesses a pair of lateral slit-like apertures, the so-called ostia, through
which its interior is in open communication with the blood sinus
surrounding it ; this is the perieardial sinus, a part of the body cavity.
The blood fluid (hsemolymph) enters the dorsal vessel through an aper-
ture at its posterior end as well as through the lateral ostia from the
perieardial sinus, and flows out at its anterior end. It then runs back-
ward through the lacunar system more or less constantly in contact
with the integument of the body and limbs where respiration takes
place, and. finally re-enters the perieardial sinus.

Entomostraea. — The scheme just sketched corresponds more
exactly with the circulatory system of the Brancliiopoda (Phyllopodd)
than with that of any other known Crustacean. The contractile
dorsal vessel (heart) of Brancliipus (Fig. 191, p. 288) runs through
the whole trunk and possesses a pair of ostia in all segments
except the last, in which there is one terminal ostium. Anteriorly
the heart is continued into an Aorta without ostia which enters
the head and opens into the lacunar system of the body. In
this latter system a ventral principal stream from before backward can
be distinguished incompletely separated from the perieardial sinus by
a septum stretched transversely over the enteric wall. The respiration
takes place in the whole surface of the delicate integument of the body
and limbs, but is apparently specially active in the branchial sacs.
From the ventral principal stream an accessory stream runs into each
limb down one side to the point, there to bend round and to run back
up the other side to rejoin the principal stream.

The heart of the other Entomostraea (Figs. 192 and 193, p. 289)
(where such an organ occurs) is always much shortened, sac or pouch-
shaped and only supplied with one pair of ostia. Anteriorly, in front
of the anterior terminal ostium, the heart is sometimes (in many
Copepoda, Branchiura, and Cladocera) continued into a longer or shorter
aorta. A posterior ostium is added to the heart of the Copepoda. The
ostia through which the blood flows into the heart are generally pro-

v CRUSTACEA— BLOOD-VASCULAR SYSTEM 359

vided with valves to prevent its return into the pericardial cavity on
the contraction of the heart.

The heart is always placed above the intestine in the most anterior
trunk region.

With regard to the presence of the heart in the Entomostraca, all
Cladocera possess hearts. Among the Ostracoda it is only found in
the Halocypridce and Cypridinidw, among the Copepoda only in the
Calanidce, Pontellida, and Branchium. In the latter it lies far back in
front of the so-called caudal fin and is continued anteriorly in a long-
aorta as far as the brain. In the Cirripedia a separate blood-vascular
system is altogether wanting.

It would be a mistake to assume that those Entomostraca, which appear to be
simple forms because of the want of a heart and generally of a separate blood-
vascular system, are therefore also original forms. As in the worms, so also in Crus-
taceans, the want of this system of organs must certainly be considered as a derived
condition. The causes why a reduction of the heart goes as far as complete dis-
appearance are indeed only known to a very small extent. Small size of body may
here and there have some influence, occasionally the rhythmical movements of other
inner organs (e. g. the stomach of many Copepoda) seem to suffice to set in circulation
the blood or ccelomic fluid in the lacunar system ; a heart is thus rendered superfluous.

"We have one remarkable exception to all that has been said above as to the
circulatory system in the Crustacea. One genus of parasitic Copepoda (Lernan-
thropus) possesses a richly branched blood- vascular system widely spread in the body
and its appendages, and completely closed from the body cavity. A heart is wanting.
The yellowish red blood is propelled along the principal vessels by the movements
of the enteric canal, and flows forwards in two ventral longitudinal trunks and back-
wards through an unpaired dorsal vessel.

Malaeostraea, Leptostraea. — A knowledge of the circulation in
the Leptostraea (Nebalia, Fig. 196, p. 293) is of great importance;
it recalls in many respects that of the Branchiopoda, and in others
points to that of the Malaeostraea. The long tubular heart stretches
from the most posterior head region through the whole thorax
into the 4th abdominal segment, and possesses 7 pairs of ostia. The
3 most anterior of these lie laterally in the heart in the posterior
part of the cephalic region, the 3 following dorsally in the 2d, 4th,
and 5th thoracic segments, and the 7th and largest in the 6th thoracic
segment. In the last 2 thoracic segments, and in the abdomen the
heart has no ostia. These pairs of (venous) ostia through which the
blood enters the heart from the pericardial sinus, are, as in all Crusta-
ceans, provided with valves. The heart is continued into an anterior
and a posterior aorta, through which the blood flows out into
the body. Valves hinder its return from the two aorta into the
heart. Besides the aorta, branched arteries occur in both pairs of
antennae and in the abdomen. The principal portions of the lacunar
blood -vascular system are the pericardial sinus and a sinus lying
under the intestine.

The respiration is in all cases specially active at the thin inner

360

COMPARATIVE ANATOMY

CHAP.

surface of the shell (which is kept clean by long maxillar feelers), and in
the lamellate exo- and epipodites of the thoracic feet. In these parts
a brisk circulation takes place. The blood which flows through the

shell comes from the anterior aorta
and re-enters the heart by the dor-
sally placed ostia.

Arthrostraea. — First of all we
must here point out an important
difference in the position of the
heart in the two principal divisions
of the Arthrostraea, the Isopoda and
AmpUpoda. In the Isopoda by far
the largest portion of the heart,
which is provided with 1 to 2 pairs
of ostia, lies in the abdomen ; in
the Amphipoda, however, the tubu-
lar elongated heart which is almost
everywhere provided with 3 pairs
of ostia lies in the thorax. This
difference may be explained by
supposing that the Isopoda have
retained only the abdominal por-
tion, and the Amphipoda only the
thoracic portion of the primitive
Malacostracan heart which ran
through nearly the whole length
of the body, and was provided
with many pairs of ostia. The
localisation of the respiration has
probably played the chief part in
bringing about this differentiation,
since in the Isopoda respiration
takes place in the rami, and prin-
cipally in the endopodites of the
abdominal feet (pleopoda), but in
the Amphipoda chiefly in the pouch-
like branchial appendages of the
thoracic feet.

In the Anisopoda the heart lies
as in the Amphipoda in the thorax.

FIG. 240.— Diagram of the circulatory system of the Isopoda, seen from the side. The right
thoracic and cephalic walls removed. A part of the intestine (d) cut away (after Delage). Arterial
system red, venous system blue, nervousvsystem black, ani, Anterior ; an^, posterior antennae ;
ctb, cephalo-thorax ; II-VIII, 7 free thoracic segments ; 01-07, 7 abdominal segments ; br, gills
(pleopoda) ; g, brain ; d, intestine ; h, heart ; o, ostium of the heart ; pc, pericardium ; va, anterior
aorta ; la, lateral arteries ; t, thoracic arteries ; ha, hepatic artery ; si, lateral sinuses of the thoracic
region ; sa, abdominal sinus ; obl^oints of insertion of the thoracic feet ; pg, subneural vessel (the
dotted line should stop at the rect line); bp}>branchio-pericardial vessels ; zg, vessels leading to
the gills ; aa, abdominal aorta ; os, ostia (?) of the lateral sinuses.

pc,

aucL-

CRUSTACEA— BLOOD-VASCULAR SYSTEM

361

The following is a rather more detailed description of the circula-
tion of the Arthrostraca.

Isopoda (Figs. 240 and 241). The heart, which lies for the greater part in the
abdomen, and is provided with 1 to 2 pairs of lateral ostia, is closed blindly behind.
Out of it 11 arteries arise, viz. (a) a medio-dorsal thoracic aorta running to the head
and the eyes, and supplying the cerebral ganglia, and the 2 pairs of antennae, (&) one
pair of lateral arteries for the anterior
thoracic segments, and the posterior
cephalic region together with the ex-
tremities of these regions, (c) 3 pairs
of thoracic arteries for the 3 posterior
thoracic segments and their extremi-
ties, (d) 1 pair of abdominal aortse for
the abdomen and its limbs, which
function as gills. The thoracic aorta
forms anteriorly in front of the brain
a ring which embraces the oesophagus,
from which a subneural artery runs
longitudinally under the ventral chord
the whole length of the body. This
also gives off branches to the limbs.
Besides the blood lacunae which lie
between the viscera, there is generally
in the thorax a large, paired, ventral
blood sinus, which in the abdomen FIG. 241.— Conilera cylindracea (after Delage).
becomes unpaired. 5 pairs of vessels Transverse section of the abdomen. Most of the

conduct the venous blood out of the letters have the same meaninS as in FiS- 24°- &>

, . . , n Heart : x and y, muscle layers (muscle lamella) for

abdominal sinus into the pleopoda moving the gilS (pieopoda)! flexor and extensor ;,,

which function as gills. 5 pairs of is laid back . aTO> the abdominal arteries which sup-

efferent vessels (veins) conduct the ply them ; ex, outer ; en, inner branchial lamella of the

blood which has become arterial in pleopoda (exo- and endopodites) ; ag, efferent ; zg,

*"*'' "'

the giUs into the pedcarfial sinus;

it passes thence through the ostia

into the heart, and by the contraction of this latter it is again dispersed through the

arteries.

Amphipoda (Fig. 242). The Amphipoda present a striking contrast to the Iso-
poda in that in the latter the vascular system, in the former the lacunar system, is
the most pronounced. The long tubular heart which generally lies in the 5 or 6
anterior free thoracic segments, usually possesses 3 pairs of ostia, less frequently one
(Corophium} or 2 (Platyscelidce). It is continued into an anterior and a posterior
medio-dorsal aorta, which pour the blood either direct or through further arterial
branches into a large ventral sinus which lies between the integument and the
intestine and runs through the whole length of the body. Special afferent blood
streams (vessels ?) conduct the mixed blood into the extremities of the thorax and
abdomen, thus also into the branchial pouches of the thoracic limbs. Special
efferent streams collect the blood in these extremities (thus also the blood which has
become arterial in the gills) and conduct it back through 7 vascular loops in the
thorax and through 6 in the abdomen into the pericardium, which stretches back
beyond the heart to the end of the abdomen. In Corophium the abdominal portion
of the pericardium and the abdominal vascular loops are wanting. The blood-
vascular system of the Caprellidce agrees in the main with that of other Amphipoda,
allowance being made for the reduction of the abdomen.

362

COMPARATIVE ANATOMY

CHAP. V

It follows, from the above description, that in the Amphipoda the arterial blood
cannot in any way be sharply distinguished from the venous blood.

Anisopoda. — This division of the
Arthrostraca, though in many points
approaching more nearly the organ-
isation of the Isopoda than that of
the Amphipoda, shows in its blood-
vascular system greater similarity to
the latter. Two abdominal aortae,
however, arise from the posterior end
of the thoracic heart, and the peri-
cardial sinus is continued into the
abdomen. The heart of Apseudes
possesses two. ostia on the left side
and only one on the right ; in youth
the 2 pairs of ostia are present. As
to the significance of this fact see
below (p. 367).

In all Arthrostraca the paired
ostia, as well as the points of origin
of the aorta are provided with valves.

Thoraeostraea, — The cir-
culatory system here is linked
on to that of the Isopoda. Start-
ing with the Stomatopoda, the
circulatory system of the older
larvae of SquUla, known as
A lima and Erichthus, which can
hardly be distinguished from
that of the adult, have been
the most carefully investigated.
The heart (Fig. 243) extends
as a many -chambered dorsal
vessel from the maxillar region
(behind the stomach) through

the thorax and the abdomen to
fi1p pTiri nf fi1p K^ -hr]™™'™!
tllG end Ot ™6 5™ abdominal
Segment. 1WO divisions Can

be distinguished in it, a short,
lnrio. Tin<,
long Pos'

terior division. Probably the

anterior division alone cor-

FIG. 242.— Diagram of the circulatory system of
theAmphipoda,fn>mtheside(afterDelage). Mostof
the lettering as in Fig. 240. pt, Pericardial vessels
rising from the epimeres (ep), extremities (brf), and

gills (br) of the thorax ; pa, pericardial vessels of the ... anfpr:or flTir]
abdomen ; s, ventral sinus ; ep, epimeres. The epi- W1Cle anterior> ancl
meres of the thoracic segments IV and V partly cut

off ; pi, pieopoda.

FIG. 243.— Circulatory system of an older Squilla larva before it has passed into the SquUla
form (after Glaus), h, Heart, continued posteriorly into the many - chambered dorsal vessel
which is richly provided with pairs of ostia (o) ; ac, cephalic aorta ; ao, optic artery ; adi, oao,
arteries of the two pairs of antennae ; am, marginal artery of the dorsal shell ; ast, arteria sternalis ;
al, hepatic artery ; as, shell artery ; en, 1st lateral artery of the dorsal vessel ; aabi to o«66, lateral
arteries of the abdomen ; dr, glandular saccules on the hind-gut ; I, hepatic lobes in the telson ;
pfi, 6th pleopod (uropod) ; Tcb, branchial leaves (epipodial appendages of the oral feet).

FIG. 243.

364 COMPARATIVE ANATOMY CHAP.

responds with the heart of the Deeapoda. It reaches to the
posterior limit of the first maxillipedal segment, possesses a large
pair of ostia, and gives off the following vessels : (a) an anterior un-
paired cephalic aorta (ac), (b) and (c) an anterior weaker and a posterior
stronger pair of arteries. The second division, the many-chambered
dorsal vessel, possesses 12 pairs of ostia, and gives off 13 pairs of
arteries and an unpaired posterior aorta. To complete the whole picture
of the circulatory system and to show its relation to that of the Isopoda,
we must add that a median subneural vessel runs under the ventral
chord through the whole body, that the whole venous system is
laeunar, and that there are two principal venous sinuses, one ventral,
and the other dorsal. The arterial system, on the other hand, is
composed of richly-branched vessels having walls of their own and
breaking up into capillaries.

As to the more detailed arrangements, the cephalic aorta supplies the eyes, the
two pairs of antennse, the brain, and the anterior lateral regions of the shell. The
most anterior pair of arteries supplies the mandibles and maxillae, and the central
part of the shell. The large 2d pair of arteries probably supplies the maxillse and
maxillipedes ; one artery passes between the longitudinal commissures of the ganglia
of the 1st and 2d maxillipedal segments to connect itself with the subneural vessel
(compare the sternal artery of the Schizopoda and Deeapoda). The subneural
vessel gives off primarily vascular loops to the ganglia of the ventral chord, but it
also gives off branches to the limbs. The 13 pairs of arteries of the many-chambered
dorsal vessel supply the thorax and abdomen with their extremities in such a way
that the parr of arteries belonging to a pair of ostia spread out, not in the segment
in which the ostia occur, but in that immediately in front of it. The whole heart
seems to have been shifted back one segment, so that the pair of ostia lying in the
first abdominal segment (there are 7 pairs of ostia in the abdomen) originally belonged
to the most posterior thoracic segment. The posterior aorta richly supplies the
telson with lateral- branches. In the venous system paired lateral blood sinuses
conduct the blood out of the extremities and other organs into the large ventral
sinus. The blood streams thence [into the pericardial sinus and through the ostia
back into the heart. It is only in the abdomen, whose limbs carry the gills, that
the blood which has become arterial appears to flow direct back into the pericardial
sinus, avoiding the ventral sinus.

A comparison with younger Squilla larvae of the so-called Erichthoid stage makes
it highly probable that 2 pairs of anterior ostia of the dorsal vessel there present
disappear in the course of development. While, as a rule, in the many-chambered
dorsal vessel a pair of ostia lies over each outgoing pair of arteries, there are no
ostia to correspond with the two most anterior pair.

Among the Thoracostraca a chambered dorsal vessel provided with many pairs of
ostia and reaching into the abdomen is only found in the Stomatopoda. This more
primitive condition has here been retained, evidently in connection with the localisa-
tion of the respiration in the branchial tufts of the abdominal limbs.

The blood-vascular systems of all the other Thoracostraca, at least
of the Schizopoda and Deeapoda, closely resemble one another, and must
be contrasted with that of the Stomatopoda. [The blood-vascular system
of the Cumacea is not yet thoroughly investigated ; it probably agrees
to a great extent with that of the Schizopoda and Deeapoda]

CRUSTACEA— BLOOD-VASCULAR SYSTEM

365

The heart of the Schizopoda, Decapoda^ and Cumacea appears, as
contrasted with that of the Stomatopoda, extraordinarily shortened and
provided with very few ostia (2 to 3 pairs). It always lies in the
thorax and never stretches into the abdomen. This shortening
was evidently caused by the
localisation of the respiration in
the thoracic region (gills of the
thoracic feet, cephalothoracic
shield as respiratory organ), and
by the more or less extensive
fusing of the thoracic segments.

Among the Schizopoda the
heart is still elongated in Siriella,
where it runs through nearly the
whole thorax into the last thor-
acic segment. It is progressively
shortened in Mysis and Mysodopsis.
In Euphausia it has the concen-
trated form of the Decapodan
heart, and has, like the latter,
3 pairs of ostia, one dorsal, one
lateral, and one ventral, while in
the other Schizopoda and in the
Zocea-larvce of the Decapoda, there
are but 2 pairs of ostia.

The Circulatory system of
Astaeus (Fig. 234, p. 338, and

Fig. 244) may be taken as an the heart, diagrammatic." M, branchiostegite ;

example. The following vessels
rise out of the heart : (a) an-
teriorly the unpaired cephalic
aorta, which supplies with its

rich branchings the brain and the vessel ; &/, ambulatory foot ; vs, ventral sinus ; ov,
eyes. (&) TWO anterior lateral ovarium. The arrows give the direction of the

arteries (also called antennal flow of blood (after Huxley and Plateau)"
arteries). These give off branches to the stomach, the antennal glands,
the anterior and posterior antennae, and the cephalothoracic shield,
(c) The two hepatic arteries. These arise at the anterior and lower
edge of the heart and branch in the liver, (d) The sternal artery.
This arises from the lower and posterior end of the heart, which
is produced in the shape of a bulb, descends on the right or left
side of the intestine, passes between the longitudinal commissures
of the penultimate and ante-penultimate thoracic ganglia, to enter the
sub-neural vessel below the ventral chord. This must be considered
as a modified lateral artery of the heart (see Stomatopoda). (e) The
posterior aorta arises out of the posterior end of the heart, and runs
over the intestine backwards through the abdomen, giving off in each

FlG- 244.— Transverse section through the

cepnalotnorax of tne Cray.fisn in the re|ion of

k> Sills • **• respiratory or branchial cavity ; ep,
lateral wall of the cephalothorax ; pc, pericardium;
7tj heart . sa> sternal artery ; I, hepatopancreas ; d,
intestine ; abm, ventral longitudinal muscles to
the abdomen ; dbm, dorsal longitudinal muscles to
the abdomen ; bm, ventral chord ; sn, sub-neural

366 COMPARATIVE ANATOMY CHAP.

segment a pair of lateral arteries which supply the intestine, integu-
ment, and musculature of the abdomen. The sub-neural vessel which
receives its blood from the sternal artery, is already known to us in
the Isopoda and Stomatopoda, here, however, and in all Decapoda it plays
a much larger part, since lateral branches from it supply all the limbs
from the maxillae to the last pleopoda. It also serves for nourishing
the ventral chord, this being its almost exclusive function in the
Stomatopoda, the limbs there being more often supplied by branches of
the lateral arteries of the heart. All the arteries branch richly and
pass into arterial capillaries, which open into the venous laeunar
system of the body. Even though the flow of blood in this laeunar
system is, as in all Crustaceans, regular and constant, and though
venous blood canals often come into existence, we do not find among
the Decapoda any veins with walls of their own. The system of venous
cavities, taken as a whole, represents the body cavity. We can only
give a few details as to the course of the venous blood. It nearly all
flows together into a large ventral blood sinus in the cephalothorax,
from the lateral parts of which canals conduct it into the gills,
while other canals convey the blood which has become arterial in
the gills away from them to the pericardia! sinus. The respiratory
organs are therefore here placed in that part of the circulatory
system which conveys the blood out of the body back to the heart,
and this is the case in all Arthropoda. The arterial blood coming
from the gills leaves the pericardium, and, mixed with the blood
which flows back out of the cephalothoracic shield, enters the heart
through its ostia. The ostia as well as the points of origin of the
arteries of the heart are provided with valves. The valves of the
former prevent the return of the blood into the pericardium, those of
the latter its return out of the arteries into the heart. When the
heart contracts (systole) the blood contained in it is driven into the
arteries, and when it again expands (diastole) it sucks in blood out of
the pericardium through the ostia.

In the Mysidce (with the exception of Euphausia) there are 2 or 3 unpaired
hepatic arteries, springing from the ventral wall of the heart. The abdominal portion
of the sub-neural vessel is wanting. In the male of Siriella, which carries gills on
its pleopoda, the latter receive their blood from branches of the lateral arteries of
the posterior aorta. In the Schizopoda, Cumacea, and larvce of Decapoda in which
the gills are not yet developed, the blood circulation in the cephalothoracic shield
with its fold is very brisk, and there is no doubt that in those forms which have no
special gills the respiration principally takes place in it. In Siriella and Mysis and
perhaps also in other Schizopoda it is most probable that the integument forming the
inner wall of the respiratory cavity has also a respiratory function (comp. Fig. 222,
p. 321). The channels which convey the blood out of the thoracic feet back to the
heart cause ridge-like projections in this integument which may be called branchial
ridges. The constant vibrations in the respiratory cavity of the epipodial appendages
of the first thoracic foot create a constant exchange of water in it.

We have already described (p. 329) the respiratory organ of the air-breathing
Eirgus latro. We will here briefly describe the circulation of the blood in connection

v CRUSTACEA—BLOOD-VASCULAR SYSTEM 367

with it. The respiratory organs are branched tufts which arise on the inner surface
of the branchiostegite. The shell circulation which is found in all Thoracostraca
and which also plays a great part in the respiration of many water - breathing
forms, here brings about air-breathing. In the branchiostegite (here better
"lung cover") and its tuft -like appendages there is a rich meshwork of blood
sinuses, spread out between the vessels which conduct the blood in and out.
The blood passes through a large vessel out of the venous blood sinus of the

V M

FIG. 245.— Birgus latro. Diagrammatic transverse section in the region of the heart (after
Semper). M, Gill or lung cover ; h, heart ; fc, gills ; ah, respiratory cavity ; p, pericardium ; ek,
branchial blood channels leading to the heart ; «i, 03, 03, 04, lung or shell vessels leading from the
heart ; Ib, pulmonary tufts ; el, pulmonary blood channels leading to the heart ; eli, the same near
their entrance into the pericardium.

head into the lung cover. This vessel divides into 4 branches, 3 of which run
to the upper and 1 to the lower portion of the lung cover, and break up into
the meshwork of blood sinuses above mentioned. From this the blood which
has become arterial is collected by vessels which unite into a great trunk, running
first along the edge of the lung cover backwards, then upwards, and finally forwards,
to unite before entering the pericardium with the vessel coming from the small gills.

The blood of the Crustacea is usually colourless ; it is occasionally,
however, slightly yellow, green, or red. In the latter case, e.g. in the
Branchiopoda, the colouring material of the blood is haemoglobin.
The colourless blood corpuscles are almost always able to change their
form in an amseboid manner.

Judging from the varying position and form of the Crustacean
heart, we come to the conclusion that the original Crustacean heart
was, as in Branchipus, a long, many -chambered dorsal vessel pro-
vided with many segmental pairs of ostia. All other forms of
heart have been developed from this by reduction of the anterior
or the posterior portion, and by the disappearance of numerous
pairs of ostia. These reductions have had for their principal causes
the localisation of the respiration, the various differentiations of the
different portions of the body, and the fusing of segments. The
already mentioned ontogenetic fact that pairs of ostia may disappear
in the course of development (Apseudes, Stomatopoda), agrees with this
view. In many Isopoda the ostia lie alternately to the right and left
in the heart ; this arrangement perhaps comes into existence by the
disappearance of alternate ostia in the heart at first provided with
paired ostia.

368 COMPARATIVE ANATOMY CHAP.

The fact that in many Crustacea, e.g. the Branchiopoda, which go
through a long series of metamorphoses from the Nauplius larva, the
pairs of ostia are formed with the heart in order from before backward,
cannot be brought up as an objection to the above. For this method
of development corresponds in general with the manner of ontogenetic
differentiation of the Arthropodan and Annelid body which takes place
in that order. The whole question is closely connected with the view
to be treated of later as to the phylogenetic significance of the Crus-
tacean larval forms.

VIII. The Excretory Organs (Antennal Glands, Shell Glands).

Although comprehensive comparative investigations as to the
methods of excretion are still wanting, we at least know that this
function is performed in very various ways and by very various organs.
We shall here take into consideration only two of these organs, the
shell and the antennal glands. Certain intestinal appendages and
dermal glands which also seem to serve for excretion, are mentioned
in the sections on the intestine and the integument. It must not,
however, be thought that this exhausts the number of parts of the
body which have some share in excretion.

Confining ourselves to the antennal and shell glands, we note : —

(1) Number and Position. Each of these glands occurs as a single
pair. The former emerge at the basal joint of the posterior antennae.
The gland itself lies either entirely in this basal joint or more or less in
the adjoining cavity of the head. The shell gland lies in the shell
fold or in the cephalothoracic carapace in a region which originally
corresponds with the 2d maxillar segment. Its aperture lies on or
near the posterior maxillae.

(2) Occurrence. — The antennal gland is widely distributed in the
Crustacea. It seems wanting only in the Isopoda. ' While in the
Malacostraca it is, as a rule, best developed in the adult, in the Ento-
mostraca it only appears in the larval stages, and it but rarely per-
sists in the adult 6ven as a rudiment. In the Decapoda the antennal
gland has been called the green . gland. The shell gland is widely
distributed among adult Entomostraca. Among the Malacostraca it has
been observed in Nebalia, and further, in the Isopoda, Anisopoda,
Cumacea, and in the larvae of some forms (Sergestes, Euphausia) in whose
adult condition it is wanting. In Nebalia it is found in a reduced
condition.

(3) Structure and Development. — The structure of the antennal
gland (Fig. 246) is everywhere essentially the same. We distinguish
in it (1) a terminal saccule, (2) a coiled urinary canal which emerges
through (3) a urinary bladder on the basal segment of the posterior
antennae. The constitution of the epithelial wall is different in the
terminal sac and in the urinary canal. The wall of the latter is often
shown in transverse section to consist of one single cell, its lumen thus

CRUSTACEA— EXCRETORY ORGANS

369

being intracellular. In the higher Crustacea, however, the cells appear
in greater number and more closely crowded, the lumen being
intercellular. The urinary canal
is very long in the Malacostraca,
and lies in complicated coils. At
its distal end (near the opening)
it widens into the urinary bladder.
The terminal sac, as well as the
urinary canal, may be further
complicated in the higher Crusta-
ceans by the formation of lateral
invaginations.

The shell gland has essen-
tially the same structure as the
antennal gland.

That we have to do in the
antennal and shell glands with
excretory organs is shown by the
fact that when the animals are
fed with carmine, carmine par-
ticles are after a time met with
in the glandular sacs, at least in
those of the antennal glands.

According to observations
made on the Cladocera, the shell
glands are said to be of meso-
dermal origin. The antennal FIG. 246.— Left Antennal Gland of M/sis

glands (Of the Cray-fish), On the Jf?* Grobben). re Urinary canal; hb, urinary
1/17 bladder ; es, terminal sac ; oer, blood lacunse ; ea,

Contrary, are Said to Come from urinary passage (efferent duct).

a dermal depression, and so

belong to the dermal glands. These statements, however, require

further confirmation.

Morphological Importance. — Leaving out of consideration the different origins
mentioned above attributed to these glands it appears probable, from their essential
agreement in structure, that the shell and antennal glands are segmental homologous
formations. From their wide distribution in the Entomostmca, and Malacostraca
or their larvae, we may further conclude that these glands are, phylogenetically,
very ancient organs, derived from the racial forms of the Crustacea. The view that
they are homologous with the nephridia of the Annulata may be supported by many
facts in their coarser and finer structure. This view would gain greatly in probability
if it could be shown that both are developed out of the mesoderm. This homology
has not, however, yet been established.

This is, perhaps, the place to mention the cement glands of the Cirripedia,
which emerge on the last joint but one of the small adhering antennae (anterior
antennse). The hardening secretion of these glands serves to attach the animal to
the surface on which it rests.

Besides these, certain glandular tubes of the Cirripedia, which emerge on the
outer maxillae and were formerly taken for olfactory organs, have recently been
VOL. I 2 B

370 COMPARATIVE ANATOMY CHAP.

pointed out as formations equivalent to the nephridia of the Annulata (segmental
organs). They are said to be in open communication with the body cavity.

IX. The Connective Tissue.

The connective tissue found throughout the Crustacean body can here receive
only brief attention. Plates, membranes, etc. of connective tissue lie close under
the hypodermis and envelop the enteric canal and the sexual organs, and, as
neurilemma, the nervous system. Connective tissue strands, fibres, mesenteries, in
various places bind the inner organs together and attach them to the integument
The lacunar blood - vascular system, the body cavity, is lined to a great extent,
though certainly not continuously, with connective tissue.

A special form of connective tissue, widely found in Crustaceans, is the fat body,
which varies greatly in details. This often envelops the intestine and the heart.
In the connective tissue cells of the fat body are found fat-drops, fat-globules, and
also often protein granules. The fat body evidently plays a part sometimes larger,
sometimes smaller, in metabolism. It is generally variously developed at different
ages and times of the year, and also in the two sexes. In the larval forms it is often
strongly developed before the moult, which accompanies a metamorphosis, and forms a
reserve of nourishment for the process of transformation. In a few Crustaceans,
which take no food at the time of sexual ripeness or of the hatching process, it is
strongly developed before this time and much reduced after it.

Connective tissue cells may often become star-like, branched, and occasionally
contractile pigment cells. Pigment also occasionally occurs in hypodermal and
intestinal cells.

X. The Sexual Organs.

The sexes are separate in the Crustacea, except in a few cases
which will be duly mentioned.

The male and female sexual organs are constructed on one type and
have a similar position in the body. They are, as a rule, paired.
More than one pair never occurs. We can distinguish corresponding
divisions in male and female organs ; viz. first, the germ-preparing
organs (ovaries in the female, testes in the male) ; second, the duets
of the genital glands (oviducts, in the female, vasa deferentia in the
male) ; third, terminal divisions of these ducts, sharply distinguished
anatomically and ontogenetically from the preceding (vulva, vagina,
reeeptaeulum seminis in the female, muscular duetus ejaeulatorius
in the male) ; and fourth, outer eopulatory organs.

The ovaries and testes cannot in their earliest stage be distin-
guished. They can early be recognised as distinct cell groups in
the mesoderm, their rudiments can sometimes be traced back to one or
two segmentation cells.

The oviducts and vasa deferentia arise out of the mesoderm
apart perhaps from the rudiments of the germ glands.

The terminal sections of the ducts arise by invaginations of the
outer integument.

The outer eopulatory apparatus consists either of transformed
limbs or appendages of limbs, or of processes, folds, prominences, etc.
of the integument.

v CRUSTACEA— SEXUAL ORGANS 371

There is no doubt that the sexual organs in all Crustacea were
originally paired. Some of them may, however, become unpaired,
either (as in most Copepoda and Thoracostraca) by the two germ glands
becoming connected by an unpaired uniting portion, or by the two
ducts uniting over a greater or smaller extent to form a common un-
paired oviduct or vas deferens, or by the ducts emerging through a
common aperture. We can, however, always recognise the double
nature of the sexual apparatus in some one (generally the larger)
portion of it.

The ovaries and testes are either simple or branched or coiled
tubes or sacs which occupy in the body a dorsal position on each side
of the intestine, often between the heart and the intestine. They lie
in the trunk sometimes more to the front sometimes more to the back,
and sometimes along nearly its whole length. Where there are con-
necting portions between the germ glands of the two sides they lie
dorsally above the intestine.

The genital apertures are found on the ventral side except in the
Cladocera and some Copepoda, where they lie dorsally.

In the Entomostraca, setting aside the Cirripedia, the apertures lie,
as a rule, immediately behind the limb-bearing anterior division of the
trunk, at the limit between this and the limbless terminal division
called the abdomen. The single or double segment in which they emerge
is called the genital segment. There is thus in the Entomostraca no
definite constant segment of the body in which the genital apertures lie.

In the Malacostraca, on the contrary, apparently including the
Leptostraca, the position of the genital apertures is definite and constant.
The male genital apertures everywhere lie in the most posterior
(i.e. the 8th) thoracic segment, usually (Thomcostraca) in the basal joint
of the 8th pair of thoracic limbs. The female apertures are in the
third from the last (the 6th thoracic segment, if we reckon in those
fused with the head), and mostly in the basal joint of the protopodite.
There are no exceptions to this rule.

The Spermatozoa of the Crustacea are often distinguished by their
remarkable size and shape. In the Decapoda they are provided with
radially arranged processes, and are, as also in other divisions of the
Crustacea, immobile.

Numerous spermatozoa are often enclosed in a common envelope
(spermatophore) formed by the secretions of the glandular portion of
the male ducts. The eggs of many Crustacea possess besides a yolk
membrane other accessory envelopes secreted by the female ducts.
On adaptations for the care of the brood, see p. 379.

Should the view prove correct, that the oviducts and vasa
deferentia and also the antennal and shell glands correspond with the
Annulatan nephridia, then, considering the different position of the male
and female genital apertures, several pairs (4 at the least) of the
segmental nephridia of the Annulata have been retained in the
Crustacea.

372

COMPARATIVE ANATOMY

CHAP.

Entomostraca. — In the Uranchiopoda, the germ glands are paired ; in Branchipus
(Fig. 191, p. 288) they are tubular and lie in the abdomen ; in Apus they are richly
branched and lie in the limb-bearing division of the trunk. In Branchipus the ovi-
ducts are widened out at their ends and enter a uterus in which the eggs remain for a
time receiving a shell yielded by special uterine glands. The uterus lies in a pouch
formed by a concrescence of genital prominences, the modified limbs of the two genital
segments (12th and 13th trunk segments).

The vasa deferentia are also widened out at their ends (sperm vesicles) and enter
a muscular ductus ejaculatorius, which, when the penis is evaginated, is drawn out

R

FIG. 247.— Genital organs
of Lernanthropus (after
Heider). A, Female, B, male
organs. OP, Ovary ; od, ovi-
duct ; kd, cement glands ;
rs, receptaculum seminis ;
go, sexual aperture ; p, pore
canal to the spermatophore ;
o/t, brown body ; sp, sperma-
tophore ; t, testis ; vd, vas
deferens ; st, spermatophore
pouch.

with it inside it. The penis is an ectodermal outgrowth of the second genital
prominence.

In the Oladocera, the ovaries and testes are simple paired tubes. The oviducts
emerge dorsally into the posterior end of the brood cavity (see on Care of the Brood,
p. 379). The vasa deferentia with the ductus ejaculatorius emerge by paired
apertures on the ventral side of the limbless posterior division of the body.

In the Ostracoda the germ glands are paired, as also are the vasa deferentia and
the oviducts. The testes often fall into several tubes. The ductus ejaculatorius with
the sperm vesicle may be unpaired (Cypridina). The copulate ry apparatus of the
male is to some extent complicated (transformed hindermost limbs). In the female,
two genital prominences corresponding with those of the male contain the receptacula
seminis. In Cypris the ovarial tubes extend into the shell-fold.

CRUSTACEA— SEXUAL ORGANS

373

ol

ai
at

HOD

cd.-

Copepoda. — The germ glands are mostly unpaired, and placed symmetrically in the
anterior trunk segment dorsally on the intestine. That they were originally double
is occasionally apparent. In many parasitic Copepoda (Fig. 247) the germ glands are
distinctly paired and not connected by any transverse bridge. In Sapphirina a
transverse bridge occurs. The oviducts are paired, and generally branched. Their
ends are glandular or provided with glandular invaginations (cement glands), whose
secretion yields the material for the egg sacs. A receptaculum seminis common to
the two oviducts is often found. The paired apertures lie in the first abdominal
.segment (sometimes at its posterior
edge) either ventrally, laterally, or
(rarely) dorsally. The sperm passages
are paired or unpaired ; in the latter
case they are generally on one side.
They are provided with a glandular
division, which yields the envelope of
the spermatophore, and often with a
wider portion functioning as spermato-
phore pouch. The apertures in the
genital segment are paired or unpaired ;
in the second case frequently asym-
metrical.

Argulidse. — Two pairs of testes
occur in the caudal fin, and there are
2 vasa deferentia with common sperm
vesicle. A glandular tube, coming
from the anterior part of the body,
enters each vas deferens. The two
vasa deferentia unite under the intes-
tine into a common ductus ejacula-
torius, which opens at the end of the
last thoracic segment on a papilla-like
projection. The ovary is unpaired,
and even from its first appearance
asymmetrical ; it lies in the thorax.
The oviduct first appears paired, but
it is afterwards atrophied on one side,
and emerges at the base of the caudal
fin. Two receptacula seminis, entirely
separate from the female genital appar-
atus occur on the under side of the
caudal fin.

Cirripedia. — The strikingly lobate
ovaries are paired in the Balanidce,

ID

FIG. 248.— Longitudinal section through a
mature Sacculina carcini externa, at right angles
to the plane of symmetry (after Delage). co, Cloacal
aperture ; sp, sphincter of the cloaca (cl) ; g, ganglion ;
ai, outer integumental lamella, covering the brood
cavity ; at, female genital atrium, into which the un-
paired portion (uov) of the ovary and the cement
glands (cd) enter ; "bh, brood cavity, shown empty to
the left, with egg sacs (es) containing the developing
and lie deep down in the shell ring eggs to. the right ; pov, the paired part of the ovary;

(Fig. 207, p. 304) in that part of the U> inner integumental lamella covering the body
. , , . , 5 . , ,. proper or visceral sac ; p, stalk entering the aperture

body^cayity which^ extends _ into _the in the shell carapace (cp) of the host ; ^attachments

of the roots on the stalk ; da, central lacuna of the
stalk continued into the lacunae of the outer integu-
mental lamella, the roots, etc., representing the body
cavity ; t, testes.

mantle fold. In the Lepadidce (Fig.

205, p. 303) the ovaries, which are to

some extent united, lie in the anterior

cephalic portion of the body called

the peduncle. In both the Balanidce and the Lepadidce the terminal division of the

oviduct emerges on a projection on the basal joint of the anterior pair of tendril-like

374

COMPARATIVE ANATOMY

CHAP.

feet. This position deserves special notice, because the sexual apertures in no other
Crustacean lie so far forward.

In the RMzocephala (Fig. 248) the ovaries, in the shape of two lobate united
masses, fill the greatest part of the visceral sac of the body which corresponds with
the head of other Cirripedes. They open on each side into an atrium (at), into which
the cement glands (cd) also enter, and which itself opens into the brood cavity (Wi).
The ovary in Sacculina is said at its first appearance to be unpaired.

The testes in the Balanidce and Lepadidcc (Fig. 205, t) occur as two richly-branched
tubes at the sides of the intestine and are continued as 2 vasa deferentia, which swell
out into sperm vesicles before uniting at the base of the cirrus at the extreme posterior
end of the body to form the common ductus ejaculatorius. The 2 testes of the
Rhizocephala (Fig. 248, t) are simply tubular, and their vasa deferentia emerge at

FIG. 249.— A, female, B, male genital apparatus of Asellus aquaticus (after 0. Sars). ov,
Ovary ; od, oviduct ; t, testicle lobes ; vd, vas cleferens ; pi, 1st and 2d pairs of pleopoda.

that part of the brood cavity where the visceral sac is produced into the stalk. See
also below as to the sexual relations in the Cirripedia.

There is much more uniformity in the genital organs of the Malacostraca than in
those of the Entomostraca. While the two ovaries and the two testes remain sepa-
rate in the Leptostraca and Arthrostraca, in the Thoracostraca, with few exceptions,
they are joined above the intestine by an intermediate piece.

Leptostraca. — The ovaries and testes are long tubes which in the sexually mature
animal run dorsally at the sides of the intestine from near the masticatory stomach
to the last abdominal segment. The two short sperm ducts of the male emerge
in the manner typical of the Malacostraca on a projection on the basal joint of the
8th pair of thoracic feet. The aperture of the oviduct also probably lies, as in other
Malacostraca, in the third thoracic segment from the last.

Arthrostraca. — The testes and ovaries are nearly always simple paired tubes,
which sometimes ran through the largest portion of the thorax and abdomen, some-

CRUSTACEA— SEXUAL ORGANS

375

times are limited to the thorax, or to one portion of the thorax. In most Isopoda
the testis generally falls on each side into three pouches (Fig. 249, B\ which, however,
possess a common vas deferens. The oviducts open in the antepenultimate thoracic
segment into the brood cavity, receptacula seminis often developing at their ends.

OD

FIG. 250.— Genital organs of Squilla mantis
(after Grobben). A, Male, B, female organs, ct,
Posterior end of the cephalo - thoracic shield ; VI,
VII, VIII, the 3 hindermost free thoracic segments ;
t, testis ; vd, vas deferens ; d, appended glands ; ov,
ovary ; od, oviduct ; rs, receptaculum seminis.

FIG. 251.— Sexual organs of As-
tacus. A, female, B, male organs, ov,
Ovary ; u, unpaired portion of the
same ; od, oviduct ; oe, genital aper-
ture ; t, testes ; vd, vas deferens (after
Huxley).

The female genital apertures very often (Isopoda, Anisopoda) only appear at the time
of the formation of the brood pouch. Peculiar phenomena appear at the time of
reproduction in the Oniscidce. The two receptacula (representing invaginations of
the outer integument) are at first not in open communication with the ends of the
oviducts. Only after sperm has, during copulation, entered the receptacula, do they

376 COMPARATIVE ANATOMY CHAP.

pass into the oviducts by the bursting of their walls, and thus bring about the
fertilisation of the eggs in the ovaries. The animal then casts its skin, and with
the skin the receptacula seminis, so that the two genital apertures are no longer
present. The fertilised eggs pass from the ovarium into the body cavity and thence
through a newly-formed unpaired birth aperture in the last thoracic segment but one
into the brood cavity. A new batch of eggs is fertilised later in the ovary by sperm
left over from the first copulation, and this reaches the brood cavity in the same
way. After this second batch of eggs has been developed in the brood cavity and
the young that are hatched have left it, the animal moults, and then again appears
as it was before copulation.

Thoracostraca. — The genital organs of the Cumacea and in some respects also
those of the Schizopoda need more thorough investigation. The paired sexual glands
of the Thoracostraca are united by an unpaired piece which lies In the thorax in the
Decapoda and Schizopoda, and in the telson in the Stomatopoda (Fig. 250), and always
above the intestine. This piece is wanting only in the Cumacea (?) and Paguridce.
Except in the Stomatopoda and Paguridce where the ovaries and testes lie in the
abdomen, the germ glands are entirely or for the greater part restricted to the thorax.
They everywhere lie between the intestine and the heart. The testis on each side is
a tube which either remains straight (Stomatopoda}, or is coiled up in a complicated
manner (e.g. in Paguristes, Carcinus), or is provided with simple lateral invaginations
(Palinurus), or is much branched and carries small sacs at the ends of the branches
(Astacus). It is enclosed in an envelope of connective tissue. The long and much
coiled vasa deferentia run, like the oviducts in the female, posteriorly where the
germ glands are in the thorax, and anteriorly where they are in the abdomen.
They fall into two divisions, a proximal portion lying near the testes and a distal
glandular portion often provided with small invaginations, this latter division being
continued into the strongly muscula^ ductus ejaculatorius. This opens outwardly
at the basal joint of the last pair of thoracic feet either on a slight -swelling
(Macrura) or at the point of an elongated tubular penis (Brachiura, Schizopoda}.

In the Stomatopoda there emerges at the point of the penis, besides the ductus
ejaculatorius, the duct of a paired tubular accessory gland (Fig. 250, A, d) which lies
in the free thoracic segments, and the two parts of which are connected anteriorly
by an unpaired intermediate piece.

The ovaries agree in general with the testes in position and shape, but they are
simpler inasmuch as they are simple tubes or vesicles. In Squilla they have seg-
mental bulgings. The oviducts are shorter and not so much coiled as the sperm
ducts. They emerge at the typical point in the antepenultimate thoracic segment,
in Squilla immediately at the side of a median receptaculum seminis.

We cannot here enter on the subject of egg and sperm formation in the Crus-
tacea. 'But the egg formation in the Cladocera, as it is peculiarly interesting,
must be briefly described. Successive groups, each consisting of four germ cells,
sever themselves from the germ layer when the production of summer eggs takes
place, but only one cell out of each group becomes an egg, the others being used up
as nourishment. In the production of winter eggs, however, only one cell out of
every second group of germ cells becomes an egg, while the remaining 7 cells of
the two groups serve as nourishment for the one egg.

XL Sexual Dimorphism.

This is more or less marked in all Crustacea. There are indeed no outer or
inner portions of the body which in some one species, genus, or order of Crustacea
are not differently constructed in the two sexes, and these differences have great
biological significance and are of great importance in classification. We can here
only select the most important and most widely distributed.

CRUSTACEA— SEXUAL DIMORPHISM

377

The sexual differences are all to be ultimately explained as adaptations for
ensuring reproduction and for preserving the young. Adaptations facilitating the
copulation of the male and female are principally found in the body of the male.
Adaptations for securing the
favourable development of
the eggs are met with in the
female.

Male Sexual Character-
istics.—(a) The males of the
Crustacea are throughout
smaller and often also more
agile than the females. This
distinction of size k .specially
remarkable in parasitic and
attached Crustacea, where
the minute males (as in the
Cirripedia and parasitic Iso-
poda) are described as dwarf
males. On the other hand,
in all cases where the females
are so deformed and degraded
by parasitism as to be hardly
recognisable or even alto-
gether unrecognisable as
Crustaceans, the males ap-
pear less degenerated ; they
are usually still able to move
freely, are provided with dis-
tinct limbs, and show some
resemblance to the nearest
related free - living forms.
This slighter degeneration on
the part of the males may
be considered as the persist-
ence of a larval stage, the
parasitic Crustacea under-
going, as we shall see,
a striking metamorphosis.
Free-living larval forms with
Crustacean characteristics,
however, bring about the
infection of new hosts The FIQ 252._Lernaeascus nematoxys. Aj Femal B ^
degradation and crippling of more highly magnmed. alt Anterior; 03, posterior antenna;
md, mandibles ; mf, maxillipedes ; 6j, b%, &3, ti, t%, t& thoracic

feet ; od, oviduct ; ov, ovary ; Ted, cement gland ; ab, abdomen ;
d, intestine ; III, IV, V, thoracic segments ; aj, 1st abdominal
segment; gp, genital plate; vd, vas deferens ; t, testes (after
Glaus).

the body only take place after

the larva has attached itself

on its definitive host. In

the males this degradation

does not occur, or not to

the same extent, because they are obliged to retain their power of free locomotion in

order to ensure the possibility of copulation and of fertilisation of the females.

Since, however, the sole work of the male consists in the seeking out of, and the

fertilisation of the female, we often find (as in the parasitic Cirripedia) great

378

COMPARATIVE ANATOMY

CHAP.

degeneration of the parts not immediately connected with reproduction. The
intestine is thus wanting in the dwarf male of the Cirripedia which on reaching its
destination, i. e. the body of the female, there leads a semi-parasitic life. If the male
does not reach this destination, he has failed in his life-work and perishes. To this
subject of dwarf males we shall have to return (p. 382).

The accompanying figures (Figs. 252, 253, 254) illustrate the great sexual
dimorphism found in certain parasitic Copepoda and Isopoda.

(b) The olfactory filaments (Riech- und Spur- faden) on the anterior antennae are
always present in far greater numbers in the male than in the female.

FIG. 253.— Portunion Maenadis. Adult mature female (after Giard andBonnier). A, With the
brood cavity partly opened in the ventral median line and the brood lamellae'separated. The
abdomen (db) is so placed that the ventral side is seen. Ir, The anterior middle and posterior lobes
of the first brood lamella on the right side ; II, the same of the first brood lamella on the left ; II r
and II I, 2cl brood lamella (right and left) : Illr and III I, 3d brood lamellae (right and left) ; IV, 4th
brood lamellae; Vr and VI, 5th brood lamellae (right and left); pi, pleural lamella of the 1st
abdominal segment ; ex%, exopodite of the pleopod of the 2d abdominal segment ; en^, endopodite
of the pleopod of the 3d abdominal segment ; ov, ovary ; eg, cephalogaster ; ae, outer ; ai, inner
antennae ; mf, maxillipede. B, Adult female, brood cavity not opened. The abdomen ab is seen
slantingly from above ; th, thorax ; eg, cephalogaster ; Ji, cardial prominence.

(c) In the males of the most different divisions, apart from the actual copu-
latory organs, there are limbs transformed into " accessory organs of copulation"
for the seizing, grasping, and holding fast of the female. Such are the posterior
antennse of Branchipus, the seizing hooks in the anterior pair of limbs of the
Estheridce, the adaptations for holding the female in the 2d antennse or the
maxillipedes of the Ostracoda, the anterior (seizing) antennse of the Copepoda, etc.
In the Amphipoda the seizing hooks on the anterior thoracic feet are more strongly
developed in the male than in the female. In the Anisopoda ( Tanais dubius) 2 forms
of males have been observed, both of which seem to be peculiarly well organised for
catching and holding the female. The one form may be called " scenters," the other

CRUSTACEA— SEXUAL DIMORPHISM

379

"seizers." The former have numerous olfactory filaments on the antennae, the
latter much larger and very movable pincers on the chelate feet. In the Decapodan
males, the most anterior pair or the two most anterior pairs of pleopoda seem trans-
formed in order to assist in copulation. In the crayfish for example, they serve as
tubes or channels for conducting the spermatophores away from the genital aperture
to their destination. The other pleopoda which in the female carry the fertilised
eggs are reduced in the male, or may be, as in the Brachyura, entirely wanting. In

FIG. 254.— A, Adult male of Cancrion miser (nearly related to Portunion Fig 253, after
Giard and Bonnier), r, Rostrum ; 01, anterior antenna ; th, thorax ; I, liver ; 7i, testis ; Tie, heart ;
ab, abdomen. B. Hatched embryo of Portunion Maenadis (after Giard and Bonnier) from the
ventral side. t2, limb of the 2d ; ti, of the 7th thoracic segment ; 03, 2d antenna ; pZ1; 1st ;
p?6, 6th pleopod ; au, eye.

the Decapodan males also the chelae of the chelate feet are more strongly developed
than in the female.

Adaptations for the Care of the Brood.— It rarely happens in the Crustacea that
the female simply lays the eggs, attaching them to some foreign object and leaving
them to their fate. We find almost everywhere, on the contrary, that the females
retain the eggs on or in their own bodies in such a way that they are protected and
often also nourished by the mother body. The eggs develop under the protection
of the mother body, till the larvae or young Crustaceans are hatched, and even
these occasionally remain for some time in their birthplace.

380

COMPARATIVE ANATOMY

CHAP.

In the Cirripedia the eggs are concealed in the interior of the shell between the
mantle and the body of the animal. In the Rkizocephala the integument splits into
an outer lamella (mantle) and an inner lamella (wall of the visceral sac). Between
the two a brood cavity arises (Fig. 248, p. 373), into which the eggs, emerging
from the female atrium, enter, and in which they develop. The eggs are enclosed
in a richly branched sac, formed of a chitinous membrane, and exactly repeating
the form of the cement glands which enter the female atrium. The sac is in
reality nothing but the inner cuticular lining of the cement glands which is ejected
when the eggs are laid and becomes filled with the eggs as they leave the ovary.
The Nauplii which develop out of the eggs in the brood cavity reach the exterior by
means of its aperture called the cloaca. In the Branchiopoda there are various arrange-
ments for the care of the brood. In the shelled forms, the eggs are concealed under
the shell, either in appendages of certain pairs of limbs which are transformed into ovi-
sacs or brood pouches (Apus), or on filamentous appendages of such limbs (Estheridce).
In the Cladocera the eggs develop in a brood cavity (Fig. 192, p. 289), which forms
dorsally between the shell and the body, becomes entirely closed towards the exterior
by special arrangements, and contains a fluid for the further
nourishing of the brood. In some Cladocera, a saddle-shaped
thickening of the dorsal integument of the shell (Ephippium)
covers every one or two winter eggs, and is cast off with the eggs
as a protection during winter. This ephippium is often provided
with adaptations which facilitate its passive distribution in
space. In the Copepoda, the eggs which emerge from the
genital apertures reach the interior of ovisacs which stand out
freely from the body (in the genital segment), and which are
formed from the secretion yielded by the cement glands.
Where the two genital apertures lie somewhat far apart later-
ally or dorsally in the double genital segment there is a pair
of ovisacs ; where they lie very near each other on the ventral
side, one unpaired median ovisac is formed .(Fig. 194, p. 290).
These ovisacs are so characteristic of the Copepoda, that by this
means the most deformed parasitic Copepodan females may be
recognised (Fig. 255). In the Notodelphydce alone the eggs pass
into a brood cavity enclosed by the integumental folds. The
female of the Leptostraca shelters the eggs and hatched larvae
between the lamellated thoracic feet. In the female of the
Arthrostraca, Schizopoda, and Cumacea the brood lamellae on
the basal joints of the thoracic feet already described develop
at the approach of sexual maturity. These brood lamellae, by
locking into one another from right and left, form the base
of a brood cavity, whose cover is the ventral (sternal) integu-
ment of the thorax (Fig. 218, p. 318). The eggs reach this
brood cavity and develop in it. The hatched young or larvae
often stay some time in it. The females of the Decapoda attach
the emerging eggs to the pleopoda by means of the secretion

FIG. 255.— Lernae-
ocera esocina, female.
ua, Frontal eye ; tj, t%,
t3, t4, rudimentary thor-

acic feet ; d, intestine ; , ., ,.

od, oviduct; es, egg of tne cement glands already mentioned on the under side or

sacs ; A, arm processes the abdomen. In the Brachyum, whose shield-shaped abdomen

at the anterior end of js ^ent round on the sternal side of the cephalo-thorax, the

the body (after Claus). abdomen is generally decidedly larger and broader in the

female than in the male, and more adapted for covering and protecting the egg

masses. The same difference, though not so pronounced, may be seen in the
Macrura also.

CR USTA CE A— HERMAPHRODITISM

381

XII. Hermaphroditism in the Crustacea.

Hermaphroditism is a rare phenomenon in the Crustacea, and only found in
attached and parasitic forms, viz. in the attached and parasitic Cirripedia and
in parasitic Isopoda.1 The sexual relations in these groups are very interesting and
must be further briefly described.

The commoner Balanidce and Lepadidce are hermaphrodite. There are, however,
Lepadidce (Ibla and many species of Scalpella) in which, besides the hermaphrodite
individuals, dwarf males occur. These latter live parasitically on the bodies of
the hermaphrodite individuals, generally in a fold of the mantle at the closing edge
of the scutum. In their structure and form they are nowise like the hermaphrodites.
They do not advance beyond the so-called Cypris stage, their body is almost vermiform
and possesses besides the antennae only 4 pairs of reduced tendril-like feet. The
oral limbs are wanting. A mouth is wanting, the enteric canal is rudimentary,
and the testes unpaired. It is evident that these reduced dwarf males provide
occasionally for the cross fertilisation of the hermaphrodite individuals.

There are again some species of Scalpella (Sc. ornatum, regium, parallelo-
gramma, nymphocola], and further the genera Cnjptophialus and Alcippe, in which
dwarf males occur, but in which the individuals which correspond with the herma-
phrodite individuals of the related Cirripedia are not hermaphro-
dite, but only female. Here, therefore, separation of the sexes
prevails with marked dimorphism. The majority of the Cirri-
pedia, however, seem to be hermaphrodite without dwarf males.
The Rhizocephala are hermaphrodite with dwarf males, which
remain at the Cypris stage.

The hermaphroditism of certain parasitic Isopoda is of another
sort. The Cymothoidea are protandrously hermaphrodite, i.e.
in youth they are male, later, the male copulatory segments
are lost and the adult animals are exclusively female (Fig. 256).

The sexual relations in the Entoniscidce (Portunion] are pro-
bably the following. These large, characteristically-deformed
parasites (Fig. 253) are protandrously hermaphrodite, but there
are, besides, small males (Fig. 254, A] which have remained in
a larval stage, and besides these again other degenerate so-called
complementary males. Out of several larvae which reach the
host, those which have the best place on its body and are best
nourished probably develop into adults which function as females,
the second-best nourished larvae remain as males in a larval
stage, and all the others become degenerate complementary
males.

The Bopyridse, which are parasitic in the branchial cavity
of the Carididce, are sexually separate and strongly dimorphic ;
the dwarf males live on the bodies of the females.

The origin of all these peculiar sexual relations is still very
uncertain. Most free living Crustacea are sexually separate, and
so are also the free living forms related to the hermaphrodite
Crustacea.

FIG. 256.— Herma-
phrodite sexual ap-
paratus of a young
Cymothoa cestroides
(after P. Mayer),
somewhat diagram-
matic, t, Testes ; ov,

From this, and from the fact that hermaphrodites ovary ; od> oviduct ;

e -. ,, •jj»jjLtj^ v®' vas deferens : p.

are found among the parasitic and attached Crustacea, we may, penis

with some probability, conclude that hermaphroditism in the

Cirripedia and Isopoda is an acquired condition, perhaps brought about by the small

1 The Apusidoe (Phijllopoda] have also lately been shown to be hermaphrodites,
with the occasional presence of males.

382 COMPARATIVE ANATOMY CHAP.

chance of fertilisation, which the attached or parasitic mode of life offers to sexually
separate animals. To explain the occurrence of a hermaphrodite condition in
general, we must assume (an assumption not without foundation) that the rudiments
of the germ glands are indifferent, that in one case, under certain unknown circum-
stances, they may develop as testes, in another as ovaries, and in others again pro-
duce ovaries as well as testes.

In the Cirripedia, the attached and parasitic modes of life are evidently extremely
old. If the view that they are descended from Copepoda-HkQ forms is correct, then
the ancestors of the Cirripedia, when first the attached or parasitic modes of life
appeared, were probably sexually separate, and dimorphic with small, free-moving
males, as indeed is still the case in many parasitic Copepoda. The attached and
parasitic modes of life then became more and more pronounced in the female, and
caused the appearance of hermaphroditism. The males, in the meantime, who had
also taken on the Cirripede character, remained as dwarf males, and so the possibility
of occasional cross-fertilisation was preserved. In most Cirripedia the males have
probably in time disappeared, and the purely hermaphrodite condition has become
fixed. In others, the dwarf males proving under certain conditions sufficient for
ensuring fertilisation, have apparently led to the disappearance of hermaphroditism
and the reappearance of sexual dimorphism.

In the Isopoda the sexual relations have probably developed in quite another way.
How protandrous hermaphroditism arose in the Cymothoidea is indeed, at present,
uncertain. But we can perhaps imagine the rise of the sexual relations in the Crypt-
oniscidce and Entoniscidce in this way ; these animals were originally, like the
Cymothoidea, protandrously hermaphrodite, then in time some of the larvae developed
only to the male stage and became larval or degenerate males.

In the gill-inhabiting Bopyridce the male stage, in the originally hermaphrodite
individuals, must have been suppressed, as the dwarf males sufficed.

In the Amphipodan species Orchcstia, the curious fact has been established that
a certain part of the germ layer of the testes of the male produces eggs, while the
other parts produce spermatozoa. The eggs, however, never, or only in exceptional
cases, reach the exterior, and in any case do not develop further. The above fact,
which does not stand alone, is at present unexplained.

XIII. Parthenogenesis— Cyclic Reproduction.

Parthenogenesis occurs in the Crustacea only in the Phyllopoda, viz., in Esthcria
and Apus (see note, p. 381) among the Branchiopoda and in the Cladoccra. The males
are much rarer than the females, and in the Cladocera appear only periodically in
autumn.

The thin-shelled summer eggs develop parthenogenetically, and in many Phyllo-
poda in summer there is a succession of generations of females multiplying partheno-
genetically. The larger hard-shelled winter eggs, on the contrary, which are supplied
with more nutritive yolk and are laid in autumn, require fertilisation.

XIV. Ontogeny.

We can here present only a selection of the observations on the ontogeny of the
Crustacea, which are so numerous and have such an important bearing on general
morphological and biological questions. We shall first briefly describe the develop-
ment of the outer body form of some few Crustaceans which go through a long
process of metamorphosis, and then give a sketch of the development of their inner
organisation.

CR USTA CEA— ONTOGENY

383

A. The larval history of the Crustacea — The Development of Apus
(Order Phyllopoda, Family Branchiopoda).

1st Larval Stage, Nauplius. — Out of the egg is hatched an oval larva narrowing
like a pear posteriorly, with a median frontal eye and three pairs of limbs, the most
anterior of which is rod-shaped, while the two posterior are biramose. Its form and
its setae are illustrated in Fig. 257, A.

On the dorsal side of the body the dorsal shield has begun to form. The anus
lies in an indentation of the posterior margin of the body. This first stage is called

FIG. 257.— Larvae of Apus (after Claus). A, Nauplius, just hatched, with the rudiments of the 5
anterior trunk segments I-IV. B, 2d larval stage, with the rudiments of the anterior maxillae and
the first 7 trunk segments. C, 4th larval stage. L, Liver ; s, shell ; fs, frontal sensory organ.

the Nauplius larva : it is met with in essentially the same form as the first product of
the egg in many Crustaceans ; we shall therefore enumerate its general characteristics.
Body unsegmented with median frontal eye, with dorsal shield and frontal
sensory organs (filaments, etc.) with 3 pairs of limbs, the first of which is simply
rod-shaped, i.e. consists of a single row of joints, while the 2 posterior are
biramose, i.e. consist of a protopodite, endopodite, and exopodite (shaft, inner
and outer branch). In all Crustaceans the first pair of limbs of the Nauplius
become the anterior antennae, the second the posterior antennae, the third the
mandibles of the adult animal.

The Nauplius of Apus (Fig. 257, A} is distinguished from the typical Nauplius

384 COMPARATIVE ANATOMY CHAP.

larva only by the fact that in the posterior third of the larval body, in front of its
posterior end, the rudiments of the 5 anterior trunk segments (/- V} and their limbs
can be recognised under the integument.

2d Larval Stage. — The Nauplius casts, its skin and the larva in the 2d stage
(Fig. 257, B] shows considerable modifications. The anterior body is broadened, the
posterior body elongated in the shape of a cone. There are two frontal projecting
stylets (frontal sensory organs). The dorsal shield has increased in size. On the
basal joint of the 3d pair of limbs (mandibular limbs) a masticatory process has
formed. Behind the mandibular limbs are the rudiments of the first pair of maxillae
(4). The 5 anterior segments of the trunk can be more clearly distinguished, and
also the rudiments of the 3d and 4th pair of trunk limbs, the latter as transverse
ridges. Later a 6th segment arises, and behind it the rudiments of the 2 subsequent
segments. The bulgings at each side of the anus have become elongated into con-
siderable furcal processes.

3d Larval Stage. — This is entered on at the 2d moult. There are 6 anterior
pairs of lobate trunk limbs, whose size and degree of differentiation decreases from
before backward ; these already clearly show the characteristic form of the Phyllo-
podan swimmerets, with their endites, exopodites, and branchial sacs. Behind the 6
anterior trunk segments 2 more, and later 3 more, can be distinguished, and behind
the anterior maxillae the posterior have begun to form. The dorsal shield at first
covers only the two anterior trunk segments.

The 4th Larval Stage appears with 7 anterior pairs of lobate trunk limbs (Fig.
257, C], Three to four anterior segments are covered by the dorsal shield. The 8th
and 9th pairs of limbs show the beginning of lobate formations ; the 10th to 13th
pairs of trunk limbs just arising. Rudiments of the paired eyes are visible.
The rowing antennae (2d antennae) are provided with large jaw hooks.

5th Larval Stage. — The 9 anterior pairs of trunk feet are lobate, the 10th is in
the act of forming a lobe, the llth, 12th, 13th, and 14th pairs of feet are forming.
Behind these are to be seen the rudiments of 6 new segments. Locomotion,
hitherto caused by the 2 anterior pairs of limbs, is now chiefly produced by the trunk
feet (swimmerets).

The mandibular foot is, as compared with the earlier stage, very much
reduced, its principal part now being the masticatory process.

Further Larval Stages. — Frequent moults follow. New swimmerets are con-
tinually formed behind those already developed, and become differentiated from before
backward. The mandibular foot is more and more reduced, till nothing remains but
the masticatory process. The rowing limbs of the larvae (2d antennae) also become
reduced. The dorsal shield continues to widen. The form of the adult animal
develops very gradually.

From this larval history we see, 1st, that the body and its appendages in general
become differentiated quite gradually from before backward, that new segments and
limbs progressively, though also occasionally irregularly, form behind those already
developed, and that these differentiations originate in the most posterior region of
the body.

2d. That there occur deviations in details from this manner of formation. The
maxillae are an instance of belated appearance in the order of the limbs from before
backward. This fact is of importance because in the adult Phyllopoda as compared
with other Crustaceans the maxillae are very much simplified.

3d. The mandibles, which in the adults are masticatory ridges without feelers,
are in young larvae well developed biramose limbs.

4th. The posterior antennae, which are reduced in the adult, are, as large biramose
rowing arms, the chief organs of locomotion in the young larva.

CR USTA CEA— ONTOGENY

385

Development of Cetochilus (Order Copepoda, Family Calanidse).

1st Larval Stage. Typical Nauplius (Fig. 258, A).— At the basal joint of the
2d antenna there is a masticatory process characteristic of most Crustacean Nauplii.
The mouth is overhung by an enormous upper lip, which is also characteristic of
many other Nauplius larvae. The anal aperture is not yet formed.

2d Larval Stage. — The body, and especially the posterior division, has grown
longer. Anteriorly there is a small shell-fold. At the end of this stage the first
pair of maxillae appear as small biramose feet behind the pair of mandibular feet.

3d Larval Stage. Metanauplius (Fig. 258, B and C). — Behind the first pair of
maxillae the 2d, and behind these the 2 anterior pairs of trunk limbs, are forming,

FIG. 258.— Larvae of Cetochilus septentrionalis (after Grobben). A, Nauplius ; B, Metanau-
plius with the rudiments of the first 2 thoracic limbs from the side. C, older Metanauplius.
ol, Upper lip ; m, mouth ; g, brain ; gz, genital cells ; an, anus ; me, primitive mesoderm cells ;
5, the two maxillipedes of the left side (exo- and endopodite of the posterior maxillae).

all as biramose feet. The dorsal shield covers the anterior part of the body as far as
and including the 2d maxillar segment. Masticatory ridges have developed on the
basal joints of the maxillipedes

Further Larval Stages. — During several moults a third trunk segment forms.

1st Cetochilus Stage. — The furcse at the end of the body are completed. A 4th
trunk segment and the 3d trunk limbs begin to form.

From what we find in other Copepoda, the only parts needed to complete the form
of the adult animal are the remaining trunk segments and the 2 posterior trunk feet,
all of which form during successive moults. The 2 pairs of maxillipedes which are
separately inserted on the body, and are characteristic of the adult Copepoda in
reality only correspond with the branches of the 2d pair of maxillae from which they
come, and thus strictly speaking represent but one pair of extremities.

VOL. I 2 C

386

COMPARATIVE ANATOMY

CHAP.

We thus again see that in the Copepoda also the body with its limbs becomes
progressively differentiated from before backward. Inasmuch as all the limbs of the
adult animal are fully and typicalty developed, wTe find in this case, during the
development, no reduction of limbs once (i.e. in larval stages) strongly developed.

Development of Sacculina (Order Cirripedia, Sub-order Rhizocephala).

The comparison of the process of development in free-living Entomostraca with
that in the parasitic forms is very instructive. Let us take that Crustacean form

FIG. 259.— Various larval stages of Sacculina Carcini. A, Nauplius after the first moult. B,
Cypris stage from the side. C, The same, 3 hours after the larva has by means of its adhering
antennae attached itself to a seta of the host. D, Formation of the Kentrogon larva. E, The
same completed, the Cypris larval shell thrown off. F, The arrow has bored through the chitinous
carapace of the host. The contents of the sac begin to pass into the body cavity of the host through
the arrow, fs, frontal sensory organ ; ua, Nauplius eye ; gl, glands of the frontal horns ; ov, rudi-
ment of the ovary ; /, fat globules ; &, seta of the host, to which the parasite has attached itself by
means of its adhering antennae ; pf, arrow of the Kentrogon stage ; ab, abdomen (after Delage).

which in an adult hermaphrodite condition is the most degraded and deformed, viz.
Sacculina (Figs. 208, 248). Although the adult animal cannot be recognised as a
Crustacean, the series of larval forms through which it passes during its individual
development most clearly proves it to be such, and related to the Cirripedia.

Nauplius Stage (Fig. 259, A}. — A typical Nauplius with its 3 characteristic pairs

v CRUSTACEA— ONTOGENY 387

of limbs is hatched from the egg. The shield-like dorsal integument forms on each
side anteriorly a process (frontal horn) at whose base glands emerge. There are 2
frontal filaments (frontal sensory organs) and an unpaired frontal eye. On the
under side of the head is a median projection of considerable size in the place of the
large upper lip of other Cirripede larvee. At the posterior end of the body are 2
separate jointed caudal processes. Mouth, intestine, and anus are wanting. Near
the Nauplius eye is a cerebral ganglion. About the middle of the body lies a mass
of cells, the rudiments of the ovary. The Nauplius now moults 3 times and under-
goes during these processes a series of transformations preparatory to

The Cypris Stage (Fig. 259, B], which it enters after its 4th moult. In this stage
we find a laterally compressed shell entirely enclosing the body, and consisting of 2
lateral valves which pass into each other, in the dorsal middle line, without articulat-
ing. The body consists of 3 regions, the large head, the trunk, and a rudimentary
terminal portion (the abdomen). The shell arises from the head. The head contains
the rudiment of the ovary, and carries the Nauplius eye, the frontal filaments and
one pair of antennae which have proceeded from the anterior uniramose pair of limbs
of the Nauplius. The two pairs of typical biramose limbs of the Nauplius (2d
pair of antennae and pair of mandibular feet) have disappeared. The trunk con-
sists of 6 segments, which have formed during the last Nauplius stages behind the
head portion, and it has 6 pairs of typical biramose limbs which cause the swimming
movement of the larva. The short abdomen carries one pair of short appendages
provided with setae. Mouth, intestine, and anus are wanting. The larvse still feed at
the expense of the nutritive yolk derived from the egg, which is thus gradually
absorbed.

The Kentrogon Stage. — After a free life of at least 3 days the Cypris-like larva
fixes itself by means of one of its two antennae to the base of a seta on the back or
on a foot of a very young crab. It then throws off the whole trunk, so that only the
head is retained (Fig. 259, C, D}. The organs retained in the head become indistinct
and, to a certain extent, fuse into a spherical mass which surrounds itself with
a new hollow cuticle under the old one. The shell is then thrown off, and another
new cuticle forms round the sac-shaped body within the old one, and in a crater-
like depression of this new cuticle a hollow arrow-like process is formed (Fig. 259, E).
The crater-like depression is then evaginated, the hollow arrow or borer is in this
way pushed forward through the antennae and pierces it and the soft chitinous cuticle
at the base of the seta of the host, and thus penetrates the body of the latter (Fig.
259, F). Through this hollow arrow the whole contents of the pouch now pass over
into the body cavity of the host, and after becoming surrounded with a new cuticle
are known as

"Sacculina Interna." — All the organs of the adult Sacculina are formed out of
the cell masses which have in this way passed over through the arrow. Among
others the testes are now first developed, and are thus, as compared with the ovaries,
very late. The Sacculina interim lies on the abdominal intestine of the host, and feeds
by means of numerous root-like processes proceeding from its surface and penetrating
the viscera of the host. As the Sacculina increases in size it exercises pressure on
the musculature and integument of the host, which die away on the under side of
the abdomen in the immediate neighbourhood of the parasite, thus allowing the latter
to pass out, while the roots (now proceeding from a stalk) remain inside the host.

Sacculina Externa. — The cloaca, till now closed, opens, and at its edge dwarf males
are always found ; these have been shown to be animals which have remained at the
Cypris-like larval stage, but can be distinguished from the female Cypris-like lame
by the fact that they develop no arrow.

The attached Cirripedia (Lcpadidce and Balanidce) like the Sacculina pass through

388

COMPARATIVE ANATOMY

CHAP

a Nauplius and a Cypris stage in the course of their developments. These larvae differ
from the corresponding larvae of Sacculina not only in the possession of an enteric
canal, but in a few other points as well. The Nauplius larva of the Cirripedia is
characterised by a dorsal shield with frontal horns and posterior pointed processes, and
by a large upper lip. Frequent moults lead to the Cypris-like larva (Fig. 260) ; this
has a bivalve shell with shell muscles ; its anterior antennae have become adhering

ab

Tf

FIG. 260.— Cypris-like larva of Lepas fasciculata (after Glaus).

antennae. The posterior antennae and mandibular feet of the Nauplius larva have
disappeared, and so has the upper lip. By the side of the median eye a paired com-
pound eye has arisen. Behind the mandibular feet of the Nauplius simple anterior
and perhaps also posterior maxillae have begun to form. The trunk consists of 6
segments with 6 pairs of biramose feet serving for swimming. The abdomen has
2 furcal appendages. A cement gland opens on the 2d joint of the adhering antenna,
which is provided with a sucking disc.

ca

rf

FIG. 261.— Pupa of Lepas pectinata, in optical section (after Glaus). In Figs. 260 and 261, the
same letters are used : pa, paired eye ; ua, unpaired Nauplius eye ; rf, trunk feet, in Fig. 261 with
the tendril-like feet beginning to form inside ; 1, anterior (adhering) antennae ; L, liver ; sm, occlusor
muscle of the scuta ; ab, abdomen ; ca, carina ; t, tergum ; sc, scutum ; cd, cement gland.

The Cypris-like larva attaches itself by means of its adhering antennas. A pupa
stage follows (Fig. 261), during which the organs of the adult Cirripede form under
the larval skin. Within the maxillae and the 6 pairs of trunk feet of the Cypris-like
larva the mouth parts and the 6 pairs of tendril-like feet of the adult Cirripede begin
to form. In the Lepadidce the head grows out anteriorly into the stalk which
carries the minute adhering antennae, and under the Cypris shell the 5 shell pieces

v CRUSTACEA— ONTOGENY 389

form on the mantle-fold. The Cypris shell is thrown off, and the paired eyes
disappear.

The history of the development of the attached and parasitic Cirripcdia is in
many respects exceedingly interesting. The Cypris-like larva shows, apart from the
absence of posterior antennae and mandibular feet, a distinctly Entomostracan
character. Its 6 typical biramose feet recall the swimmerets of the Copepoda. The
remarkable transformation of the free-swimming Cypris-like larva into the attached
sexual animal must be referred to its adaptation to an attached mode of life ; this
adaptation consists in the formation of a hard shell or framework serving for protec-
tion, the transformation of the swimmerets into tendril-like feet suited for bringing
food within reach, and the reduction of the paired eyes which are of no great use to
the adult animal. A commencement of this adaptation to the attached mode of life
may perhaps be seen even in the Cypris stage, viz., in the degeneration of the
posterior antennae and mandibular feet, which in the Nauplius larva had played an
important part, especially as organs of locomotion, whereas in the Cypris-like larva
the trunk feet serve that purpose. The transformation of the Cypris-like larva into
the parasitic Sacculina involves far more radical changes than those which take
place in the development of the Lepadidce and Balanidce. The development of
Sacculina may be described as a strikingly retrogressive metamorphosis. To explain
the reappearance of the typical Nauplius and Cypris-like larvae in the course of
development, notwithstanding the degradation of the adult animal, it is necessary
not only to emphasise the power of inheritance but to remember that free-moving
young forms are extremely useful to parasites for the infection of new hosts and
the maintenance of the race. Nevertheless, even in the free-swimming larvae of
Sacculina, we recognise distinct signs of degeneration such as the absence of an
alimentary canal. This degeneration could take place without damage to the
maintenance of the race, because the adult Sacculina, in consequence of its exceed-
ingly favourable conditions of nutrition, can give its eggs for their development so
much nutritive yolk that the larvae proceeding from them are under no necessity of
obtaining food from without. The occurrence, at first sight so remarkable, of an
endoparasitic stage in the development of the Sacculina is not difficult to under-
stand, for by passing through such a stage the parasite avoids the danger of being
thrown off by the moulting of the host.

Larval History of the Euphausidse (Order Schizopoda, Fig. 262).

1. Nauplius Stage. — Typical.

2. Metanauplius Stage. — The masticatory ridge of the mandible develops, while
the mandibular foot itself is reduced. The upper and under lips form. Behind
the mandibles the rudiments of the 2 pairs of maxillae and of the 1st pair of thoracic
feet (maxillipedes) appear as buds. The cephalothoracic shield is distinctly
developed, and the paired eyes first appear.

3. Calyptopis Stage (somewhat corresponding to the Protozocea stage of Penazus]
(B, C}. — The thorax and abdomen are demarcated, and the latter already elongated.
Segmentation appears in the thorax, and later in the abdomen. No new extremities
except the last pair of pleopoda (uropoda) begin to form.

4. Furcilia Stage. — The paired eyes become stalked. The most anterior pairs of
thoracic and abdominal feet begin to form in succession from before backward.

5. Cyrtopia Stage. — The antennae are transformed and no longer serve for
locomotion. The posterior pairs of thoracic and abdominal feet and the gills
appear.

6. Post-larval Stages. — The adult animal is gradually formed, and the caudal
fin definitely developed. It is hardly necessary to remark that all the thoracic feet

390

COMPARATIVE ANATOMY

CHAP.

and pleopoda are originally biramose and in this animal retain this character during
life. No great importance should be attached to the names of the developmental
stages (Calyptopis, Furcilia, Gyrtopia) ; they are referable to a time when the larvre
so named were thought to be different genera. We thus again see, from the
larval history of the Euphausidce, that the body with its limbs is differentiated
from before backward. We notice, however, special and important exceptions
to this rule. In the first place the rudiment of the last pair of pleopoda

FIG. 262.— Larvae of Euphausia. A, Nauplius, last form before moulting (after Metschnikoff).
B, Protozoaea. C, the same somewhat older (after Glaus), th, Thoracic segments ; 06, abdomen ;
(ara6\ abdominal segments ; an, anus ; fs, frontal sensory organ ; 1-5, limbs of the head ; OQ, 6th
pair of pleopoda.

appears before those of the other pleopoda, before even those of the thoracic
feet. This is noteworthy on account of the special form of and the important
part played by these pleopoda as part of the caudal fin in the older stages of develop-
ment and in the adult. We further note that although the thorax becomes
segmented sooner than the abdomen, and although on the thorax as on the abdomen
the extremities become differentiated in succession from before backward, the rudi-
ments of the extremities on the thorax and the abdomen are almost simultaneous
and sometimes they even occur earlier on the abdomen.

CR USTA CEA— ONTOGENY

391

Larval history of Penaeus (Order Decapoda, Sub-order Macrura, Family
Carididse (Shrimps), Figs. 263 and 264).

1. Nauplius Stage.— A typical unsegmented Nauplius (Fig. 263, A) is hatched
from the egg. The body possesses no dorsal shield ; it carries 2 setse posteriorly.

2. Metanauplius Stage. — The dorsal shield begins to form. The 3d pair of
Xauplius limbs (mandibular limbs) shows

the rudiments. of the masticatory ridge.
Behind this appear the rudiments of the 4
subsequent limbs (maxillre and 2 anterior
pairs of maxillipedes).

3. First Protozosea Stage (Fig. 263, B).
— The cephalothoracic shield has grown
large. The posterior division of the body
elongates till it is as long as the anterior
part. The 2 pairs of maxillae and the 2
anterior pairs of maxillipedes have devel-
oped and are capable of functioning ; the
latter are biramose limbs with endo- and
exopodites. The division which follows
behind these is divided into 6 segments
without any trace of extremities, and these
segments are the rudiments of the 6 pos-
terior thoracic segments (III-VIII). These
are followed by the posterior body which
is not yet segmented and shows no trace
of extremities. The mandibular feeler has
disappeared. The posterior body ends
with two furcal processes. Near the
median eye the paired eyes appear.

4. The Second Protozosea stage (Fig.
264, A] is very similar to the first, but on
the abdomen the rudiments of the 5
anterior abdominal segments (a^-a^) are
visible. Behind the 2d pair of maxilli-
pedes on the first of the 6 newly formed
thoracic segments the rudiments of the 3d
pair of maxillipedes appear (III).

5. First Zosea Stage. — The paired eyes
stand out as stalked eyes. The 3d pair of
maxillipedes has also become biramose.
On the 5 subsequent pairs of thoracic seg-
ments the rudiments of the 5 pairs of
"ambulatory feet" appear (Fig. 264, B,
IV- Till). On the segments of the
abdomen also formations appear which are
probably the rudiments of the pleopoda
(ai-cts). In any case the beginnings of the
last pair of pleopoda which are destined to

(H-M)

FIG. 263.— Young larva of Penaeus (after
F. Muller). A, Nauplius ; B, Protozoaea. III-
VIII, Rudiments of the 3d-8th trunk segments.
ab, Abdomen ; 1-5, limbs of the head ; I, II,
of the thorax. In all subsequent figures the
head limbs are denoted by Arabic, the thoracic
limbs by Roman numerals, the abdominal limbs
(pleopoda) by «i, a^ etc., the telson by t, the
exopodite by ex, and the endopodite by en.

form with the telson the caudal fin, are distinctly visible on each side as bi-lobed
formations under the integument.

6. Second Zosea Stage (Fig. 264, (7).— The last pair of pleopoda project freely.
On the two pairs of maxillse the small fan-plates (exopodites) have formed. The 5

392

COMPARATIVE ANATOMY

CHAP.

posterior pairs of thoracic feet (ambulatory feet) project freely as doubly- tipped pro-
tuberances. The formations on the abdomen of the 1st Zocea stage, which were
indicated as the rudiments of pleopoda, are no longer to be seen. The cephalothoracic
shield has a pointed process projecting anteriorly in the middle line. The animal still
moves chiefly by means of the 2 pairs of antennse.

7. Mysis or Schizopoda Stage (Fig. 264, D). — This is so called because all the
thoracic feet are developed, like those of the Schizopoda, as long biramose limbs
(with exo- and endopodite), and here also serve for swimming. The branchial appen-

FIG. 264.— Older larvae of Penaeus. A , Older Protozoaea, dorsal view. B, 6 posterior thoracic
segments, and abdomen, with the rudiments of the feet of a somewhat older larva. C, Further
advanced Zoaea. D, Mysis stage of a Penaeus, from the side, fs, Frontal sensory organ ; L, liver ;
ab, abdomen ; IV-VIII, thoracic segments ; (c^-og), abdominal segments ; t, telson. In C to the
left, the 3d thoracic foot (3d maxillipede) is covered by the 2d (after Glaus).

dages of the thoracic feet appear. The pleopoda grow further during this period, the
most anterior pair first, then the 2d and then the others almost simultaneously.
The feelers undergo important transformations, which bring them nearer the adult
form. Alterations likewise appear in the cephalothoracic shield. The auditory sac
forms at the base of the antennae. The mandible receives a feeler. The jaws
approach their definitive form. By degrees, through several moults, the larvae reach

8. The Penaeus form, the exopodites of the thoracic feet becoming more or less
reduced and the pleopoda developing further.

The Penaeus development also shows us that the body with its appendages

v CRUSTACEA— ONTOGENY 393

becomes differentiated from before backward. But here also the last pair of pleopoda
advances more rapidly than the others.

Larval history of the Stomatopoda (Fig. 265).

Unfortunately we do not know the whole series of larval forms in any of the
Stomatopoda. They belong to two types, one of which is called the Erichthus, the
other the Alima type. "We shall only consider the first.

A. Youngest known Erichthoid larva (A). — Three regions can be distinguished
in the body, an anterior, a middle, and a posterior. The anterior corresponds
with the head, and from it arises as a fold of the integument a large dorsal shield
which covers the second region as well. The head carries besides a median eye the
two large stalked eyes, the two pairs of antennse, the pair of mandibles and the two
pairs of maxillae. The second region consists of 5 segments, corresponding with the
5 anterior thoracic segments, and carries 5 pairs of biramose swimmerets (7- V), the
last 3 of which decrease in size from before backward. The 5 pairs of limbs answer
to the 5 pairs of oral limbs in the adult Stomatopoda. The third region consists of 3
short limbless segments (also covered by the dorsal shield) corresponding with the 3
posterior thoracic segments, and a very large caudal plate, also devoid of appendages.

B. In a somewhat older 2d larva a new segment (the most anterior abdominal
segment) with one pair of limbs has formed in front of the caudal plate. The 2d
pair of thoracic feet shows alterations preparatory to its transformation into large
seizing feet.

C. In a third larva (E) 2 new segments with the rudiments of their limbs, and
in older larvse all the abdominal segments with their pairs of limbs excepting the 6th,
have begun to form in front of the caudal plate, while the 3 posterior thoracic
segments are still limbless.

D. In a fourth Erichthoid larva the 2 anterior pairs of thoracic feet have lost
their exopodites, but on the other hand the rudiments of epipodites (gills) appear on
them. The three subsequent pairs of thoracic feet are reduced, and the last three
thoracic segments are still limbless. On the 6th abdominal segment the rudiments
of the limbs (uropoda) appear.

E. In the subsequent stages, the 3rd, 4th and 5th pairs of thoracic limbs are
completely reduced, or else are represented only by small sac-like prominences ( (7).

F. There now follows, after various preparatory intermediate stages, the
completely developed Erichthus larva. (D] The 3d, 4th, and 5th pairs of thoracic
limbs again appear in their definitive form, so that now the 5 anterior trunk limbs
are developed as seizing or oral limbs. On the last three thoracic segments the
rudiments of the biramose ambulatory limbs appear. By this time, not only the
full number of segments, but the full number of limbs of the adult animal is reached.

In this larval histoiy of the Stomatopoda we note

(1) That the segments of the body become differentiated from before backward.

(2) That the limbs also as a general rule follow the same order. We observe
this principally in the case of the abdominal limbs, since in the youngest Erichthoid
larva the 5 anterior thoracic limbs are already developed. Of the pleopoda, the last
(uropoda) appeared last, in opposition to the Decapoda, in which this latter takes
precedence of all the others and even of some of the thoracic limbs.

3. The last three weakly developed thoracic limbs (ambulatory limbs) form the
chief exception to the rule stated under (2), in that they first appear after the
pleopoda.

4. It is a striking fact that the 3d, 4th, and 5th pairs of trunk limbs which are
early developed as biramose rowing feet become completely reduced, and are then
again formed in their definitive shape in the oldest Erichthus stage.

394

COMPARATIVE ANATOMY

FIG. 265.— Stomatopodan larvae of the Erichthus type. A, Youngest known Erichthoid
larva. B, Somewhat older larva, from the side. C, Young Erichthus larva (Zocea). D, Older
Erichthus larva with complete number of limbs, an, Anus ; ua, nauplius eye ; br, rudiments of
the gills on the pleopoda ; ab, abdomen. The Roman numerals in brackets denote the corresponding
limbless thoracic segments. I-V, Oral feet ; VI-VIII, ambulatory feet ; t, telson (after Glaus).

v CRUSTACEA— ONTOGENY 395

Development of Palinurus and Scyllarus (Decapoda, Macrura, Fam. Loricata).

The early larval development is carried on within the shell ; during this
development 2 stages, a Nauplius stage and a very important further stage which
might be called the embryonic Phyllosoma stage, are passed through. The
characteristics of this latter stage are as follows : There are 2 pairs of antennae,
mandibles, and 2 pairs of maxillse. The thoracic limbs have all begun to form, i.e.
3 pairs of maxillipedes and 5 pairs of ambulatory limbs. The 2 most posterior
(ambulatory limbs) are only present as minute buds. The 6 anterior are biramose,
but the exopodites of the 2d and 3d pairs of maxillipedes are degenerated even
during embryonic life. There are two stalked lateral eyes, and the median Nauplius
eye. In the body 3 regions are to be distinguished. (1) Head with dorsal shield ;
(2) thoracic region, in which the segmentation is indicated ; (3) a distinctly segmented
but limbless abdominal region ending in a fork.

Before the larva is hatched, however, in addition to the 2 posterior pairs of
ambulatory feet which remained undeveloped, the 1st pair of maxillipedes dis-
appear, and the 2d pair of maxillse and the 2d pair of antennae degenerate.

FIG. 26G.— Phyllosoma of Palinurus (after Glaus), ad, Abdomen ; I, liver.

In the larva hatched from the egg the above-mentioned regions of the body can be
recognised — it is much flattened dorso-ventrally (like a leaf), and as transparent as
glass ; it is called the younger Phyllosoma (Fig. 266).

The older Phyllosoma larva is distinguished from the younger by the following
characteristics. The 1st pair of maxillipedes has formed anew, and the 2 posterior
pairs of ambulatory feet have developed. The 2 posterior maxillipedes again develop
exopodites, and on the ambulatory limbs the rudiments of the gills appear. The
abdomen is more elongated and shows the rudiments of the pleopoda. The larva
thus already has, apart from its strange form, typical Decapodan characteristics.
Its transformation into the sexual form has not been observed.

In the development of the Loricata we note : —

1. That during the processes which go on in the last part of embryonic
life the body with its extremities becomes differentiated typically from before
backward.

2. That before the hatching of the Phyllosoma larva certain extremities or parts
of extremities degenerate, to appear anew on the older Phyllosoma larva ; this

396

COMPARATIVE ANATOMY

CHAP.

is especially the case with the 1st maxillipedes, and the rudiments of the last 2
ambulatory feet. The Phyllosoma shows in a specially striking manner the character
of a pelagic larva.

Development of the Brachyura.

A very characteristic Zosea larva (Fig. 267) is hatched from the egg. Its dorsal
„ ^. shield is marked by the

possession of long spine-like
processes, among which a
frontal, 2 lateral and 1 dorsal
are never wanting. All the
head limbs are present. Of
the thoracic limbs we find
only the 2 anterior pairs of
maxillipedes ; the other thor-
acic limbs, as well as that
part of the thorax to which
they belong, are wanting or
are only found in their first
rudiments. The abdomen is
segmented, and ends in a
fork, but has no appendages.
In the later Zosea stages
the 3d pair of maxillipedes
appears, the 5 ambulatory
feet arise as imiramose limbs
(without exopodites), and
the pleopoda begin to form.

FIG. 267.-Zoaea of Maja, after its moult (after Glaus), h, Heart. The Zocea swims principally

by means of the 2 biramose

anterior pairs of maxillipedes, and also by means of the abdomen which, in com-
parison with that of the adult, is

elongated and well developed.

The older Zocea changes into a

Megalopa larva (Fig. 268).

This larva, but for the more

strongly developed abdomen and

the pleopoda, already resembles the

sexual form. The maxillipedes and

ambulatory feet appear as in the

adult condition, and it must be n

specially remarked that the ambu- /A

latory feet are never biramose, and

that the Brachyura thus do not

pass through any Schizopoda

stage.

Through many moults the Mega-
lopa is gradually transformed into

the sexual form.

We need say only a few words

about the development of the Fia 268.— Megalopa larva of a Portunus with abdo-

other Malacostraca. The Lcptos- men straightened out, dorsal aspect (after Glaus).

CR USTA CEA— ONTOGENY

397

traca (Nebalia}, the Amphipoda, and a few Decapoda (e.g. the Cray-fsli), leave
the egg in a form like that of the sexually ripe animal. On the other hand in
the Isopoda, Mysidce, and Lophogastridce among the Schizopoda and the Cumacea, the
young form hatched from the egg may be very little developed, and may even
resemble a maggot-shaped Nauplius or Metanauplius ; but there are no early free-
swimming larval forms, as the young undergo their metamorphoses in the brood
pouch of the mother. The development of the parasitic Isopoda is interesting, since
here the free-moving young or larval forms which go in search of hosts, resemble the
typical Isopoda in form and development of the body and limbs much more distinctly
than do the sexually mature animals.

R B

FIG. 269.— Moina rectirostris. Four early stages of development. A, Blastosphere, seen
from the vegetative pole. B, Gastrula stage. C, somewhat older stage, with neural plate, closed
blastopore and sexual cells sunk below the surface ; D, Nauplius stage, B, C, D, in optical
longitudinal section. In C, D, the neural plate, and in D, the rudiments of the Nauplius limbs are
projected on to the section, ec, Ectoderm ; me, mesoderm ; g, primitive sexual cells ; en, endoderm ;
sp, neural plate ; lp, blastopore ; nd, nutritive yolk ; m, mouth ; st, stomodteum ; cq, anterior ; fro,'
posterior antennae ; md, rudiments of mandible (after Grobfoen).

B. The Arrangements of the Germ Layers and the Development of the
Inner Organs

may here be described by means of a few examples.

I. Moina rectirostris (Order, Phyllopoda ; Sub-order, Cladocera ; Fig. 269).
The segmentation of the mesolecithal egg is superficial, and somewhat unequal.
In the 32 blastomere stage a blastomere lying at the vegetative pole is distinguished

398 COMPARATIVE ANATOMY CHAP.

by a specially large nucleus. This divides, and as the primitive sexual cell produces
the rudiments of the germ glands.

A blastomere in contact with this sexual cell yields the endoderm, and the
blastomeres, which surround the primitive sexual cell, and the primitive endoderm
cell, yield the Mesoderm (except the sexual organs and nervous system). In a
later stage (Fig. 269, A] we find 4 genital cells, 32 endoderm cells, and 12 mesoderm
cells ; all the remaining blastomeres represent the ectoderm. First the 12 mesoderm
cells sink down, then the endodermal plate becomes invaginated (B}. In place
of the gastrula mouth, which probably closes, the definitive mouth appears later.
The genital cells increasing in number to 8, then sink down ((7). Nutritive yolk
remains in the segmentation cavity. Two paired frontal groups of cells appear as
rudiments of the neural plate. Later on this plate consists of several layers and
yields anteriorly the brain, and posteriorly the retina. The oesophageal commissures
and the ventral chord arise in situ as thickenings of the ectoderm. The mesoderm
spreads itself out on the inner side of the ectoderm, on the surface of the rudiments
of the genital glands, and around the mid-gut which is at first solid : this mid-gut
alone proceeds from the endodermal invagination. The shell gland is of mesodermal
origin and only opens secondarily at the point of the 2d maxilla. The compound
eye is from the first paired.

II. Cetochilus septentrionalis (Order, Copepoda).

The segmentation is total and yields a blastula with a small segmentation cavity.
Under the blastomeres at a certain stage 2 symmetrically placed cells can be
distinguished as primitive mesoderm cells and a few others also symmetrically
placed as endodermal cells. The dividing mesoderm cells sink deeper inwards. In
still later stages we can recognise at the posterior end of the mesoderm, on each side,
a large primitive mesodermal cell. The endoderm also becomes invaginated into the
segmentation cavity. The gastrula mouth closes in a median line from before back-
ward. The dividing mesoderm cells fill the segmentation cavity. The stomodseum
and proctodaeum arise by means of ectodermal invaginations. At a later (Nauplius}
stage there arises at each side behind the Nauplius eye an ectodermal thickening
connected with the brain, which severs itself from the integument but degenerates
later. These are to be considered as the rudiments of the paired compound lateral
eyes with their optic ganglia, which are wanting in most Copepoda. A pair of meso-
derm cells lying on the ventral side of the intestine represent the rudiments of the
sexual glands. The 2 cells move higher on the 2 sides of the intestine, and become
surrounded with smaller mesoderm cells, which yield the rudiments of the ducts,
which at first are solid. In the 1st Cetochilus stage the paired genital rudiments
fuse to form an unpaired dorsal sexual gland. The heart develops out of a paired
rudiment of mesoderm cells.

III. Branchipus (Phyllopoda).

In the hatched Nauplius larva under the cuticle, the segments of the first maxilla
and the first two trunk segments, with their limbs, have already begun to form. An
elongated portion follows, in which the segmentation of the mesoderm streaks
has begun. This does not reach as far back as into the anal segment. The two
mesoderm streaks unite posteriorly directly in front of the anal segnient to form a
ventral plate which forms a budding zone. Its cells rapidly increase, and as the
development of the larva progresses new mesoderm segments continually become
demarcated from the budding zone and spread out under the ectoderm to the dorsal
middle line. But this segmented germ streak merely represents the parietal layer

v CRUSTACEA— ONTOGENY 399

of the mesoderm ; the visceral layer never becomes segmented, and differentiates
quite apart from the parietal layer. (In the anal segment on each side there lie 2
large cells, which, however, do not divide and do nothing to increase the budding
zone by contributing cells.)

The cell material of the mesoderm segments, which are successively formed at
the anterior end of the budding zone, begins to group itself into three divisions, and
this group is the more distinct the further the segment is removed from the budding
zone, i.e. the older it is. The dorsal division yields the rudiment of the cardial
chamber of the segment and that part of the dorsal longitudinal musculature which
also belongs to it ; the lateral division yields the musculature of the limbs, the ventral
the segmental division of the ventral musculature as well as the neurilemma of the
ganglia. The limbs begin to form as outgrowths and bulgings of the ectoderm, into
which cell-growths of the mesoderm penetrate. The ventral chord becomes differ-
entiated from before backward. In each segment in which a pair of limbs begins to
form an ectodermal thickening appears on each side of the ventral median line. The
two thickenings in a segment represent the rudiments of the double ganglion, which
at first are not united by a transverse commissure. These rudiments free themselves
from the ectoderm at a later stage. We thus find in an older Branchipus larva the
whole ventral chord from back to front in all stages of differentiation. Posteriorly,
where new pairs of ganglia are continuously formed with the new segments, these
are still ectodermal thickenings. In the formation of the heart (many-chambered
dorsal vessel) only one longitudinal row of perpendicularly arranged muscle cells on
each side takes part. The two rows grow together in such a way as to form a hollow
tube. The heart becomes differentiated from before backward. In many larval
stages a greater or smaller number of segmental cardial chambers are already developed
anteriorly and silre&dy pulsate, while posteriorly new cardial chambers are in process
of formation.

The rudiments of the compound eye appear in the Metanauplius larva on each
side as a hypodermal thickening, which is then said to divide into two layers, a
superficial layer which yields the cornea and crystalline cone cells, and a deeper
layer which yields the retinulre with the rhabdoms. Another ectodermal thickening
connected with the first and belonging to the secondary brain chiefly yields the
material for the optic and the retinal ganglia.

IV. Astacus fluviatilis (Figs. 270-280).

The segmentation is superficial, and leads to the formation of a spherical blasto-
sphere (blastula), in which the central nutritive yolk, which is divided in radially
arranged yolk pyramids, is enveloped all round by the blastoderm as by an
epithelium.

Stage A. At one point of the blastosphere the constitution of the blastoderm
undergoes a change. The cells at this part are longer and stand closer together.
This part corresponds with the future ventral side and is symmetrical ; it can be
denominated the ventral plate. This ventral plate consists of the following portions :
the 2 cephalic lobes, the 2 thoraco-abdominal rudiments, and farthest back, the
unpaired median endoderm disc (Fig. 271). The whole ventral plate is of one layer
except at one point between the endodermal disc and the thoraco-abdominal rudiments.
Large cells have sunk down inwards from the blastoderm as the primitive mesoderm
cells (B, M). "With the exception of the endoderm disc, which becomes invaginated
later, and the primitive mesoderm cells, the ventral plate and the whole remaining
blastoderm represents the ectoderm. The latter yields at a later stage the sides and
back of the thoracic shield, but in the early stages of development is nothing more
than a sac surrounding the yolk.

400

COMPARATIVE ANATOMY

CHAP.

FIG. 270.— Astacus fluviatilis, section through an egg, after
completed formation of blastoderm, ch, Chorion ; hs, stalk of
attachment ; U, blastoderm cells. The nutritive yolk is black in
this and the following Figs, (after Reich enbach).

FIG. 271.— Astacus fluviatilis, part of the surface of an egg
with embryo beginning to form.— Stage A. K Cephalic lobes,
with the rudiment of the eye ; TA, thoraco-abdominal rudiments ;
BM, formative zone of the mesoderm ; ES, endoderm disc (after
Reichenbach).

Stage B. Embryo, with
semicircular gastral fur-
row. — Only the ventral
plate alters. At the an-
terior edge of the endoderm
disc a semicircular furrow,
or a fold projecting in wards,
appears.

Stage C. Embryo, with
circular gastral furrow. —
The thoraco - abdominal
plates unite in the middle
line. The semicircular
furrow has become a fur-
row round the whole cir-
cumference of the endoderm
disc and is thus circular.
The middle part of the en-
doderm disc sinks down, so
that it now becomes a de-
pression with a somewhat
raised floor (gastrula inva-
gination). This depression
represents the rudimentary
midgut, and the outer edges
of the original circular
furrow are the edges of
the blastopore or gastrula
mouth. In front of the
anterior edge of the blasto-
pore lie the primitive meso-
derm cells in active growth,
and giving off cells towards
the centre of the blasto-
sphere (Fig. 275).

Stage D. The primitive
mouth is in the act of
closing. — In the centres of
the two cephalic lobes are
the rudiments of the eyes.
Between the cephalic lobes
and the thoraco-abdominal
rudiments the first traces
of the mandibles and an-
tennae appear. The primi-
tive mouth closes from be-
fore backward, its lateral
edges growing together in
the middle line. The cells
•of the primitive intestine
begin to consume nutritive
yolk.

CR USTACEA— ONTOGENY

401

FIG. 272.— Astacus fluviatilis, embryo in the Nauplius stage.
— Stage F. A (above), Rudiment of eye ; I, upper lip ; G, cerebral
ganglion ; «j, anterior antennae ; grao, ganglion of the segment of the
2cl antennae (03) ; gm, ganglion of the segment of the mandibles (in) ;
Ta, thoraco-abdominal rudiments ; A, (below) anus (after Reichen-
bach).

Stage E. Embryo,
with mandibles begin-
ning to form. — In the
middle of the thoraco-
abdominal disc the (ecto-
dermal) rudiments of the
anus and hind-gut (proc-
todseum) appear in the
form of an anal pit.
The mesodermal elements
spread out below the
ectoderm.

Stage F. Embryo in
the Nauplius stage
(Figs. 272 and 276).—
The 3 most anterior pairs
of extremities (Nauplius
extremities) can be dis-
tinctly made out. Be-
tween the anterior an-
tennae the first (paired)
traces of the brain appear,
and immediately behind
there, is a pit-like de-
pression, the ectodermal
rudiments of the mouth
and fore - gut (stomo-
dfeum). In the posterior
antennal and the mandi-
bular segments the rudi-
ments of ganglia appear
as ectodermal thicken-
ings in the same way as
in the formation of the
brain. The thoraco-
abdominal ridge, sur-
rounded by a deep fur-
row, projects anteriorly.
The anus has moved
forward on it and reaches
the inner side of its an-
terior edge, i.e. the later
ventral side of the end
of the body. The ecto-
derm and mesoderm form
a budding zone on the
thoraco-abdominal ridge,
and from this time new

FIG. 273.— Astacus fluviatilis, embryo with thoracic feet beginning to form.— Stage H. .A,
Eyes ; g, brain and ganglia of the anterior antennae (aj) ; «2, 2d antennae ; m, mandible ; mx^, mx2,
anterior and posterior maxillae ; <i-£g, thoracic feet, of which ti-t-j are maxillipedes ; ts, rudiment
of thoracic shield ; db, abdomen, bent back on the anterior portion of the thorax ; T, telson ; I,
upper lip ; go, ganglion opticum (after Reichenbach).

VOL. I 2 D

402

COMPARATIVE ANATOMY

CHAP.

segments with the rudiments of their extremities are continually differentiated
anteriorly (but apparently posteriorly, since the rudiments of the thoraco-abdomen
are bent forward). At this (F) stage the rudiments of the maxillar segments and
of the first maxillipedal segment are already present. The hind-gut has become
longer, and with its blind end touches the primitive enteric sac (mid-gut). On each

Fio. 274.— Astacus fluviatilis, embryo with the rudiments of all the limbs.— The thoraco-
abdomen, with the last 4 pairs of thoracic feet cut off and laid back, ad, Antennal glands ; t4-t$,
4th-8th pair of thoracic feet (ambulatory feet) ; t±, chelate feet ; ab, abdomen ; T, telson (after
Reichenbach).

side the thoraco-abdomen is bordered by a curved integumental wall, the rudiment
of the cephalo-thoracic shield.

Stage G. Embryo with rudiments of maxillipedes (Fig. 277). — The thoraco-
abdomen elongates in a longitudinal direction and shows at the end which is

v CRUSTACEA— ONTOGENY 403

directed forwards (telson) a deep indentation. The budding zone, from which new
segments have formed, borders directly on this indentation. The penultimate
abdominal segment (6th) appears, and thus precedes in order of development a large
number of segments which lie in front of it. The rudiments of its limbs are already
visible.

Stage H. (Fig. 273 and 278) Embryo with rudiments of ambulatory feet.— All
the segments of the Crustacean body are formed ; they have developed in succession
from the budding zone in front of the telson, so that the anterior and most
differentiated are the oldest and the posterior least developed are the youngest
(with the exception of the 6th abdominal segment). The budding zone itself has
been exhausted in the formation of the segments. The rudiments of the eyes project
as spheres. The ambulatory feet have developed. The telson appears deeply forked.
The abdomen has become slender, and is so bent forward along the thorax that the
telson almost touches the upper lip. The hind-gut has opened into the mid-gut, the
epithelial cells of the latter having for some time become swollen by assimilating
yolk, and being at this stage columnar. The yolk between the mid-gut and the
ectoderm is absorbed, and the wall of the mid-gut is almost in contact with the
cephalothoracic ectoderm. The ventral chord with its ganglia differentiates from
before backward, having arisen out of paired lateral strands, and a middle strand
formed by invagination. The lateral strands are ectodermal thickenings with
segmental swellings, which sever themselves from the ectoderm in order from before
backward (cf. also Fig. 279).

Stage J. Embryo with rudiments of abdominal feet. — The cephalothoracic
shield has developed greatly, its lateral parts project freely as rudimentary branchio-
stegites.

Stage K then follows, with developed eye-pigment and the rudiments of the gills.
Thereupon the young Crustacean, which is already tolerably similar in appearance to
the adult, is hatched from the egg. The fusing of the anterior thoracic ganglia to
form the infra-cesophageal ganglion has begun. The forked telson has become a
round plate, and the abdomen resembles that of the adult. The liver is formed
almost exclusively by processes of folding of the wall of the mid-gut (cf. Figs. 274
and 280).

FIGS. 275-280.— Astacus fluviatilis. Median longitudinal sections through embryos at
different stages of development. In Figs. 275, 276, 277, and 279 only the ventral side of the
embryo is depicted, in Figs. 278 and 280 the whole embryo in longitudinal section. In Fig. 280 the
position of the embryo is the reverse of what it is in the other figures (all Figs, after Reichenbach

A ^

FIG. 275. — Stage C. TA, Thoraco-abdominal rudiments ; ene, endodermal invaginatiou (gastrula
invagination) ; m, mesodenn ; hi, cephalic lobes.

FIG. 276.— Nauplius Stage. Stage F. md, Mid-gut, the endoderm cells have assimilated
nutritive yolk; pr, proctodseum (hind-gut); A, anus; TA, thoraco-abdominal rudiment; m,
mesoderm ; st, beginning of invagination of the stomodseum ; ec, ectoderm.

FIG. 277.— Stage G. en, The endodermal cells laden with yolk ; h, rudiment of heart ; pr,
proctodseum ; kz, germ zone ; A, anus ; m, mesoderm ; I, upper lip ; st, stomodseum ; ec, ectoderm.

m

cfa T

FIG. 278.— Stage H. Lettering the same. In addition : ss, sternal sinus; ga, ganglia of the
ventral chord ; T, telson ; g, supra-oesophageal ganglion.

FIG. 279. — Stage J. Section not quite median. Lettering as before. In addition : cts, cephalo-
thoracic shield ; ms, mesoderm segments ; tgi, 1st thoracic ganglion (of 1st maxillipede) ; mg,
mandibular ganglion.

nut

Fig. 280.— Eipe embryo. Lettering as before, mu, Mouth ; 6w, ventral chord ; msc, muscu-
lature ; aa, arteria abdominalis ; as, arteria sternalis ; Id, diverticula of the mid-gut (hepato pancreas) ;
h, heart ; pc, pericardium ; md, extensions of the mid-gut ; mm, muscles of the masticatory stomach ;
pk, pyloric valve.

406 COMPARATIVE ANATOMY CHAP.

Development of the eye. — Each eye consists of three elements : (1) a hypodermal
layer, (2) an ectodermal invagination, and (3) a nerve mass.

The hypodermal layer becomes many-layered ; some of its elements combine to
form groups, each of which consists of 8 cells, 4 of which (Semper's cells) secrete a
cuticular corneal facet, while the 4 others produce the crystalline cone.

The ectodermal invagination (optic fold) deepens in the Nauplius stage
and is constricted off in stage (G) as a solid mass of cells. This mass develops
into a fold, opens outward and upward, with an outer and inner wall several layers
thick. The outer wall becomes connected with the layer of the crystalline cones, its
cell elements form groups of 6 to 8 and become the retinular cells. The inner wall
yields the nervous connections of the retinular layer with the ganglion opticum.

The ganglion opticum arises like any other ganglion of the central nervous system
as an ectodermal thickening.

Between the hypodermal layer and the optic fold, mesodermal elements penetrate
and grow, uniting to form a fenestrated mass' which secretes a large quantity of
pigment.

XV. — The morphological significance of the most important Crus-
tacean larval forms, and the Phylogeny of the Crustacea.

That' the Crustacea form a single class is evident to the student of comparative
anatomy and ontogeny. It is probable that they can ultimately be traced to one
racial form. Basing our conjectures for the present simply on the comparative
anatomy and classification of the group, we feel justified in describing this racial
form as follows.

The original Crustacean was an elongated animal, consisting of numerous and
tolerably homonomous segments. The head segment was fused with the 4 subsequent
trunk segments to form a cephalic region, and carried a median frontal eye, a pair
of simple anterior antennae, a second pair of biramose antennae and 3 pairs of biramose
oral limbs, which already served to some extent for taking food. From the posterior
cephalic region proceeded an integumental fold which, as dorsal shield, covered a
larger or smaller portion of the trunk. The trunk segments were each provided with
one pair of biramose limbs. Besides the median eye there were 2 frontal sensory
organs. The nervous system consisted of brain, cesophageal commissures and
segmented ventral chord, with a double ganglion for each segment and pair of limbs.
The heart was a long contractile dorsal vessel with" numerous pairs of ostia seg-
mentally arranged. In the racial form the sexes were separate, the male with a
pair of testes, the female with a pair of ovaries, both with paired ' ducts emerging
externally at the bases of a pair of trunk limbs. The excretory function was carried
on by at least 2 pairs of glands, the anterior pair (antennal glands) emerging at the
base of the second pair of antennae, the posterior (shell glands) at the base of the
second pair of maxillae. The mid-gut possibly had segmentally arranged diverticula
(hepatic imaginations).

This conjectural racial form shows a considerable correspondence with the Annelida,
and this correspondence would be increased if it could be satisfactorily proved that
the biramose limbs of the Crustacea with its exo- and endopodite answered to the
dorsal and ventral parapodia,1 and the epipodial gills of the Crustacea to the dorsal
gills (of the dorsal parapodia) of the Annelida. The setiparous glands of the Annelida
would correspond with the segmental leg glands of certain Crustacea (Phyllopoda}.

1 It must here be remembered that the question whether in the Polychceta the uniserial
or the biserial arrangement of the parapodia is the original is not yet satisfactorily settled.

v CRUSTACEA— PHYLOGENY 407

The proof of the homology of the antennal and shell glands of the oviducts and
vasa deferentia with Annulate nephridia would also be of the greatest importance.

Among the now living Crustacea there are two orders of the JZntomostraca, viz.,
the Phyllopoda (the Branchiopoda especially) and the Copepoda, whose organisation
best recalls that of the racial form, the former in the rich homonomous segmentation
of the trunk and the structure of the nervous system and the heart ; the latter in
the form of the limbs, especially the oral limbs, which still most clearly show their
original biramose character. The Cladocera and perhaps the Ostracoda might be
derived from animals like the Branchiopoda by the shortening of the body and the
reduction of the number of segments. The ancestors of the Cirripedes are probably
nearly related to the immediate progenitors of the Copepoda.

At first sight the results of research in comparative ontogeny do not seem to
harmonise with the view just stated. Out of the egg of the Entomostraca and many
Malacostraca the unsegmented Nauplius larva is hatched with -only 3 pairs of
extremities, without a heart, and without a segmented ventral chord. On account
of this the Crustacean racial form was formerly universally held to be a Nauplius-
like animal. It was assumed that from this racial form the Crustaceans of to-day
developed phylogenetically through a series of gradual transformations, in a manner
similar to that in which they develop ontogenetically at the present time from the
Nauplius by a series of metamorphoses.

We consider these views incorrect, both for general and special reasons.

General Reasons. — (1) We know of no animal form which in an adult sexually
mature condition resembles a Nauplius. (2) We are not justified, without further
proof, in concluding that the early larval stages of an animal form closely resemble
the ancestors of that form.

Special Reasons. — The assumption of a racial form like a Nauplius leaves the
problem of the rise of the typical Crustacean organisation, of the segmentation of the
body, the segmented ventral chord, and the segmented dorsal heart quite unexplained.
It must also be emphasised that the ducts of the sexual organs, except in one single
case, emerge in regions of the body which in the Nauplius are not at all developed.

On the other hand, the Nauplius, as characteristic larval, not racial form of
the Crustacea, is explained without difficulty by the assumption of a Crustacean
racial form resembling the Annelida. Just as the racial form of the Crustacea is to
be traced back to the Annelida, so is the larval form of the Crustacea to the larval
form of the Annelida. We have already mentioned the tendency in the animal
kingdom to shift back to earlier stages of development the characteristics of the adult
animal. Hence in the Crustacea, Crustacean characteristics (such as limbs and dorsal
shield) appear as early as in the larva which corresponds with the Trochophora of the
Annelida. The method of moving by means of the limbs makes the old method
needless, so that the ciliated circles of the Trochophora larva are no longer produced.
. In all Crustacea the 3 pairs of Nauplius limbs become the 3 most anterior extrem-
ities of the adult animal (2 pairs of antennae and the mandibles). The fact that, in
the Nauplius, just these pairs of limbs appear first can again be explained without
difficulty. As in the Annelida so in the Crustacea also, the body with all its organs
becomes differentiated from before backward. At the posterior end of the body a
growing or formative zone continually constricts off new segments anteriorly. It is
thus evident that the most anterior extremities must appear first. But why just the
3 most anterior, and neither more nor less ? A natural answer may perhaps also be
found to this question. In a young larva which, like the Nauplius, is hatched early
from the egg, only a few of the organs most necessary for independent life and inde-
pendent acquisition of food can be developed.. The 3 most anterior pairs of limbs
which serve for swimming may be described as such most necessary organs. The 3d

408 COMPARATIVE ANATOMY CHAP.

pair perhaps belongs to this category, because as mouth parts, generally provided
with masticatory processes, they serve not only with the others for locomotion, but
also for conducting food to the oral aperture. Where the egg is provided with a
somewhat richer supply of nutritive yolk, the larva which hatches from it may, as a
Metanaiqjlius, be provided with the rudiments of other limbs.

The Nauplius, however, as typical Crustacean larva, certainly shows many
primitive Crustacean characteristics, such as the dorsal shield, the median eye, the
frontal sensory organs, the special form of the posterior antennae and mandibles
which, like the extremities which follow them, are developed as typical biramose
limbs.

The Nauplius is thus to be traced back to a Trochophora larva, in which we
already find Crustacean characteristics; it is unsegmented, contains the rudi-
ments of the anterior cephalic portion of the adult Crustacean with the mouth,
and the rudiments of the most posterior end of the body with the anus. Between
the two lies an embryonic formative zone, from which, in the further development
of the larva, the rest of the body begins to form and becomes differentiated, as in
the Annelida, from before backward. The Nauplius is a typical Crustacean larva ;
the ancestors of the Crustacea did not as yet possess a typical Nauplius larva,
still less did they come from a Nauplius-like racial form.

As for the Malacostraca, the testimony of comparative anatomy is unequivocal,
that the Leptostraca not only stand nearest to the common racial form of this whole
sub-class, but also retain many primitive characteristics of the common racial form
of all Crustacea. The Leptostraca appear as genuine Malacostraca, first on account
of the formation of the regions of the trunk, which consists of a thorax of 8 segments
and an abdomen ; the latter, although it has one more posterior segment than the
typical Malacostracan abdomen, still carries the same number of pleopoda. Both
the oral and thoracic limbs show the typical Malacostracan character. The intestine
proves itself to be a Malacostracan intestine by its masticatory stomach, and the
special form of its hepatic glands, and the apertures of the genital organs have the
position which is characteristic of the Malacostraca. On the other hand, by the
possession of a bivalve shell - fold, the simultaneous development of the 8 free
thoracic segments and their appendages, by the rich segmentation of the nervous
system, and the elongated heart with many pairs of ostia, the Leptostraca show them-
selves to be very primitive Malacostraca, whose ancestors must have been racially
related to the Phyllopoda.

The relationships of the other orders of Malacostraca are, according to the present
state of our knowledge, to be considered as follows : the Stomatopoda form an order
standing quite by itself, which, though showing further and somewhat peculiar
development (e.g. in the possession of gills on the pleopoda, sexual organs in the
abdomen, numerous hepatic appendages, special form of the thorax and its extrem-
ities) has still retained, in many points of its organisation, primitive characteristics,
as, for instance, the elongated dorsal vessel with many pairs of ostia, and the shell-
fold which leaves several thoracic segments free.

Of the other Malacostraca, the Schizopoda, and especially the Euphausidce have,
for the most part, retained characteristics belonging to the common racial form of
the Malacostraca. Of these the chief are the special form of the biramose thoracic
feet, their epipodial appendages which function as gills, the dorsal shell, etc. The
Arthrostraca have evidently a common origin with the Scliizopoda. Among the
former the Anisopoda must occupy the most primitive place, as well on account of
the occurrence of a small dorsal shell as of that of small exopodites on the second
and third pairs of trunk feet (Apscudes}.

The Arthrostraca are otherwise characterised by degeneration of the shell-folds,

v CRUSTACEA— PHYLOGENY 409

transformation of the stalked eyes into sessile eyes, and disappearance of the exopo-
dites on the thoracic feet.

The difference in the position of the heart in the AmpTiipoda and Isopoda is to be
explained, as already pointed out, by the fact that in the former only a thoracic
portion, and in the latter only an abdominal portion of the primitive elongated dorsal
vessel has been retained, in both cases, clearly in connection with the localisation of
the respiration. The Cumacea stand nearest to the Schizopoda, but exhibit a few
Isopodan characteristics. It is pretty certain that the Decapoda come from Schizo-
poda-like ancestors ; the Macrura, and especially the Carididce, appear to be the
most original forms, while the Brachyura and Anomura appear as one-sidedly, but
very highly, developed branches of the order.

We are again met by the question, what is the phylogenetic significance of the
various larval forms of the Malacostraca ? Just as the Nauplius larva, which is
characteristic of the whole class of the Crustacea, was held to be a stage of develop-
ment repeating the common ancestral form, so a further developed Malacostracan
larva, the so-called Zocea, was considered to be a larva characteristic of the Mala-
costraca, and assumed to correspond with one of their racial forms.

These Zocea larvae have been already thus characterised : a large cephalothoracic
shield, 2 compound stalked eyes, a median Nauplius eye, head with 5 pairs of
limbs, only the most anterior part of the thorax with thoracic feet (2-3) in a
rudimentary condition, the remainder of the thorax wanting, or else rudimentary
and limbless. Abdomen with full number of segments but without appendages.
Tail bifurcated.

Those investigators' who wished to find in this Zooea larva a larval form correspond-
ing more or less with the racial form of the Malacostraca had to assume that in the
Malacostraca now living the last 5 or 6 thoracic segments are new formations, since
they are wanting in the racial form. "When therefore the structure and systematic
position of the Leptostraca were better appreciated, and the larval history of the
Euphausidce and the Carididce better and more completely understood, these authors
were obliged to take refuge in the following explanation. The earliest ancestors of
the Malacostraca possessed the full number of trunk segments and the full number
of thoracic feet, but a later form lost the last 5 or 6 trunk segments with their appen-
dages, while, however, they were still retained by the larva. Finally these segments
and their appendages again appeared (in the living Malacostraca known to us or in
fossil forms). In support of these very forced views the larval history of certain
Stomatopoda was cited ; in these the 3d, 4th, and 5th pairs of thoracic feet are
present in the young larva, but disappear later, to be finally reformed. In the larvre
of Sergestes also, the last 2 pairs of thoracic feet are reduced and appear again later.

So many objections, however, can be brought against this assumption that it
must necessarily be relinquished.

In the first place it must be pointed out that of all the manifold known forms of
Crustacea, and especially in that series (which indeed is not without 'gaps) beginning
with the Phyllopoda and ending with the Brachyuran Decapoda, of which Branchipus,
Nebalia, Euphausia, Penaeus are the most important, there is nothing to give the
slightest indication that whole regions of the body with their extremities can
disappear and reappear again later, or that between already existing segments new
segments with their limbs can be intercalated.

We must therefore ask, to what extent the Zocea larva above characterised is
distributed among the Malacostraca. In its typical form it is found only in the
Brachyura and (though slightly debating in form) in the Stomatopoda. This fact
alone warns us to be careful, since the Brachyura are decidedly the most recent and
most specialised Decapoda, and in their development the Schizopodan stage with the

410 COMPARATIVE ANATOMY CHAP.

characteristic biramose thoracic feet does not once occur. The larval history of the
Euphausidce, Stomatopoda, and Carididce, the early developmental processes in the
Loricata which take place within the egg, and the so-called direct development of so
many Malacostraca show that the Malacostracan body becomes gradually differen-
tiated in the typical manner (as in the Entomostraca] from before backward. The
larval history of Penaeus shows further that, with the exception of the 6th pair of
pleopoda, the extremities also begin to form and differentiate from before backward.
In the Euphausidce the extremities of the posterior thoracic region differentiate
almost simultaneously with those of the abdomen from before backward.

It is of the greatest importance for the comprehension of the typical Decapodan
Zocea, and especially of that of the Brachyura, to bear in mind the established fact
that in that larva, in spite of the apparent absence of the thoracic region comprising
the 4th to 8th thoracic segments, the thoracic ganglionic mass through which the
sternal artery runs has already begun to form.

Not less important also is the fact that the Stomatopodan larva possess an
elongated dorsal vessel with many pairs of ostia, while the Decapodan Zocea and the
other Thoracostracan larvae which resemble it possess a compact thoracic heart,
generally with 1 or 2 pairs of ostia. These facts sufficiently show how varied are the
Thoracostracan larvae which have been artifically combined into a Zocea group.

In all future attempts to understand the morphology of Crustacean larval forms
the following considerations will probably have to be taken into account.

The grade of development and physiological importance of a section of the body
or of a pair of limbs in the adult animal may be recognised by the earlier or later
appearance of their rudiments ; e.g. the extraordinarily early' appearance of the last
pair of pleopoda in the Thoracostraca, i.e. of that pair of pleopoda which in the
adult differs so strikingly from the others, forming together with the telson the
caudal fin, which is the chief organ of locomotion. A contrast to this is afforded by
the late appearance of the reduced maxillae of the adult Phyllopoda and Cirripedia.

Not only the special form of the adult animal, however, but also that of one or
more of its larval stages may have some influence on the early processes of develop-
ment, i.e. when these larval stages are not merely phases of development, but
animals feeding themselves and leading an independent life and playing an important
part in the life economy of the species, more or less like adult animals though unable
to reproduce themselves. Such larvae show throughout their organisation independ-
ent adaptation to the special conditions of their existence, and it is the first and
chief object of the preceding developmental stages to prepare their organisation,
not that of the adult animal. From this point of view we shall perhaps some day
be able to explain the Brachyuran Zocea with its reduced thorax and failing 4th
to 8th ambulatory feet, which in the adult animal are the only organs of locomotion,
but in the Zocea are perhaps useless or even a hindrance. The larvae lead a marine
life, and their organisation is no doubt more or less adapted to this life. "We see,
therefore, how important it is to ascertain what are the functions of the different
parts of each larva, in order to arrive at any sure conclusion as to the significance of
the changes it undergoes.

One of the most important and most interesting problems of ontogeny is the
reduction and subsequent reappearance of the same portions of the body, of
which the larval history of the Malacostraca yields us so many examples. (Compare
especially the larval history of the Stomatopoda and Loricata. ) For the solution of
this problem also the above points of view may prove of service. In'the Stomatopoda
the last 3 pairs of thoracic feet, the so-called \ ambulatory feet, like the 5 pairs of
ambulatory feet in the Brachyura, first appear at the end of larval life. The 5
anterior pairs, however, develop during the first period of larval life (youngest

v CRUSTACEA—LITERATURE 411

known Erichthoid larva) partly to disappear in the 2d period and finally to re-
appear in their definitive form. The developmental history of the Loricata shows
similar phenomena. The first attempt, later frustrated, to form all or most of the
typical Malacostracan extremities, which we here observe is no doubt to be
ascribed to the force of heredity. The temporary disappearance of some of the
extremities is most probably a phenomenon of adaptation to the special conditions of
existence of the larva, so different from those of the adult. If, however, these first
fruitless and useless attempts were in the course of time to become gradually weaker,
and finally entirely cease, then we should meet in the Loricata and Stomatopoda with
phenomena quite similar to those in the Brachyuran development, where the
formation of the last 5 thoracic segments, and their extremities occurs so extra-
ordinarily late. In this way, also perhaps we can explain the fact that the
Brachyuran ambulatory feet appear in their definitive form and not as biramose
feet.

Review of the most important Literature.

Especially recommended as an introduction to the study of the Crustacea.

T. H. Huxley. The Cray-fish: An introduction to the study of Zoology ("Inter-
national Scientific Series," 1884).

Comprehensive or more general works.

J. E. V. Boas. Studien iiber die Verwandtschaftsbeziehungen der Malacostraken, in

Gegenbaur's Morphol. Jahrbuch. 8 Bd. 1883.
Justus Carriere. Die Sehorgane der Tkiere, vergleichend-anatomisch daryestellt.

Mimchen und Leipzig, 1885.
C. Glaus. Untersuchungen zur Erforschung der genealogischen Grundlage des

Crustaceen-systeins. Ein Beitrag zur Descendenzlehre. "VVien, 1876. Important.
The same. Neue Beitrage zur Morphologic der Crustaceen, in Arb. aus dem Zool.

Institute zu Wien. 6 Bd. 1886.
J. Dana. Crustacea of U. S. exploring expedition under Capt. Charles Wilkes. 2

vols. with Atlas. Philadelphia, 1852.
H. Gerstaecker. Arthropoden in Bronris Klassen und Ordnungen des Thierreichs.

5 Bd. /. Abfh. First half 1866-1879. Second half not yet completed.
A. Grenacher. Untersuchungen iiber das Sehorgan der Arthropoden. Gb'ttingen,

1879.
Carl Grobben. Die Antennendruse der Crustaceen. Arb. aus dem Zool. Institute zu

Wien. 3 Bd. 1880.
Milne Edwards. Histoire naturelle des Crustaces. Paris, 1834-1840. Vol. III.

with Atlas.
Fritz Miiller. Fiir Darwin. Leipzig, 1864.

Anatomy of single divisions.

G. Bellonci. Sistema ncrwso e organi dei sensi dello Sphaeroma serratum, in

Memorie Accad. Lincei. Vol. X. 1881.
Carl Claus. Neue Beolachtungen uber Cypridinen. Zeitschrift f. wiss.' Zoologie.

23 Bd.
The same. Zur anatomie und Entwickelungsgcschichte der Copepoden, in Archiv. f.

Naturgeschichte. 24 Jahrg. 1 Bd. 1858.
The same. Beitrage zur Kenntniss der EntomostraJcen (Copepoden, Esther ten}.

Marburg, 1860.

412 COMPARATIVE ANATOMY CHAP.

The same. Diefrei lebenden Copepoden. Leipzig, 1863.

The same. Die Metamorphose der Squilliden. Abhandl, d. Konigl. Gesellsch. d.

Wissenschaften. Gottingen, 1871.
The same. Zur Kenntniss des Baues und der EntwicTcelung von Branchipus stagnalis

und Apus cancriformis. Abhandl. d. Konigl. Gesellsch. d. Wissensch. Gottingen,

1873.
The same. Schriften zoologischen Inhalts. I. Die Familie der Halocypriden.

Wien, 1874.
The same. Ueber die Entwickelung, Organisation und systematische Stellung der

Arguliden, in Zeitschr. f. wissensch. Zoologie. 25 Bd. 1875.
The same. Zur Kenntniss der Organisation und des feineren Baues der Dapliniden

und verwandter Cladoceren, in Zeitschr. f. wissensch. Zoologie. 27 Bd. 1876.
The same. Der Organismus der Phronimiden, in Arb. aus dem Zool. Institute zu

Wien. 2 Bd. 1879.
The same. Kreislauforgane der Stomatopoden, Schizopoden und Decapoden. Ibidem.

5 Bd. 1884.

The same. Ueber Apseudes Latreillei Edw. und die Tanaiden. Ibidem. 5 Bd. 1884.
The same. Untersuchungen iiber die Organisation und Entivickelung von Branchipus

und Artemia. Ibidem. 6 Bd. 1886.
The same. Ueber Lernaeascus nematoxys und die Familie der Philichthyden. Ibidem.

7 Bd. 1887.
The same. Ueber Apseudes Latreillei Edw. und die Tanaiden. II. Ibidem. 7 Bd.

1887.
The same. Ueber den Organismus der Nebaliden und die systematische Stellung der

LeptostraJcen. Ibidem. 8 Bd. 1 Heft. 1888.

Ch. Darwin. A monograph of the Cirripedia. Hay Society. 2 Vols. 1851, 1853.
Yves Delage. Contribution a I'etude de Tappareil circulatoire des Crustaces

edriophthalmes marins. Arch, de Zoologie experimentale. 1°. Vol. IX. Paris,

1881.
Anton Dohrn. Untersuchungen iiber Bau und EntwicTcelung der Arthropoden.

Leipzig, 1870.
A. Giard el J. Bonnier. Contributions a Vetude des Bopyriens. Travaux de I'lnst.

zool. de Lille. Tome V. 1887.
C. Grobben. Die Geschlechtsorgane von Squilla mantis. Sitz-Ber. ATcad. Wissensch.

Wien, 1876.
The same. Beitrdge zur Kenntniss der mdnnlichen Geschlechtsorgane der Decapoden.

Arb. aus dem Zool. Institute zu Wein. 1 Bd. 1878.

Ernst Haeckel. Ueber die Gewebe des Flusskrebses, in Mutter's Archiv. 1857.
The same. Beitrdge zur Kenntniss der Corycaeiden, in Jenaische Zeitschr. 1 Bd.

1864.
C. Heider. Die Gattung Lernanthropus, in Arb. aus dem Zool. Institute zu Wien.

2 Bd. 1879.
V. Hensen. Studien iiber das Gchororgan der Decapoden, in Zeitschr. f. wissensch.

Zoologie. 13 Bd. 1863.
P. P. C. Hoek. Report on the Cirripedia collected by H. M. S. Challenger. Anatomical

Part. Zool. Chall. Exp. Part XXVIII. 1884.
Lemoine. Eecherches pour servir a I'histoire des systemes nerveux, musculaire et

glandulaire de Vecrevisse. Annales des sciences naturelles. Serie IV. Vol. XV.
1861.

Fr. Leydig. Naturgeschichte der Daphnidcn. Tubingen, 1860.
P. Mayer. Carcinologische Mittheilungen. Mitth. d. Zool. Station zu Neapel. 1 und
2 Bd. 1878-1881.

v CRUSTACEA— LITERATURE 413

The same. Die Caprelliden des Golfes von Neapcl, in Fauna und Flora dcs Golfcs

von Neapel. 6 Bd. Leipzig, 1882.
Otmar Nebeski. Beitrage zur Kenntniss der Amphipoden der Adria. Arb. aus

dem Zool. Institute zu Wien. 3 Bd. 1880.
A. S. Packard jr. A monograph of North American Phyllopod Crustacea.

Washington, 1883. (Geolog. Survey of the Territories, 12 Annual Report.}
G. 0. Sars. Histoire naturelle des Crustaces d'eau douce de Norvege. Christiania,'

1867.
The same. Carcinologiske Bidrag til Norges Fauna. I. Mysider. Christiania,

1870-1872.
The same. Report on the Schizopoda, in Report on the scientific results of the voyage

ofH. M. S. Challenger. Zoology. Vol. XIII. 1885.

The same. Report on the Phyllocarida (Leptostraca}. Ibid. Vol. XIX. 1887.
The same. Report on the Cumacea. Ibid. Vol. XIX. 1887.
R. Walz. Ueber die Familie der Bopyriden. Arb. aus dem Zool. Institute zu Wien.

4 Bd. 1882.
Max Weber. Die Isopoden, gesa?nmelt wahrend der Fahrten des " Willem Barents"

in das nordliche Eismeer, in Bijdragcn tot de Dierkunde. Amsterdam, 1884.
August Weismann. Ueber Bau und Lebenserscheinungen von Leptodora hyalina, in

Zcitschr. f. wissensch. Zoologie. 24 Bd. 1874.
The same. Beitrage zur Naturgcschichte der Daphnoiden. Sieben Abhandlungen.

Leipzig. 1876-1879. Separat-Abdruck aus Zeitschr. f. wissensch. Zoologie.

27-33 Bd.
W. Zenker. Monographic der Ostracoden, in Archiv. filr Naturgeschichte. 20 Jahrg.

1 Bd. 1854.
Many other works and treatises of von Blanc, Boas, Brooks, Berger, Burmeister,

Brady, Bullar, Braun, Brocchi, Bellonci, Brandt, Beddard, Claus, Delage, Dietl,

Friedrich, Gegenbaur, Grube, Giesbrecht, Gruber, Herbst, Kossmann, Kraepelin,

Krieger, Klunzinger, Kroyer, Leydig, Lilljeborg, Leach, Lacaze-Duthiers, Fritz

Miiller, W. Miiller, P. Mayer, A. Milne Edwards, Noll, v. Nordmann, Plateau,

Patten, Pelseneer, Rathke, Sye, Semper, M. Sars, G. 0. Sars, Spangenberg,

Stuhlmann, C. Vogt, v. Willemoes-Suhm, Vitzou, Yung, Zaddach, etc. etc.

Ontogeny (besides those mentioned above).

N. Bobretzky. Zur Embryologie des Oniscus murarius, in Zeitschr. f. unssensch.

Zoologie. 24 Bd. 1874.
W. K. Brooks. Lucifer, a study in morphology, in Philos. Transactions of the

Royal Society I. London, 1882.
C. Claus. Beitrage zur Kenntniss der Ostracoden. Entwickelungsgeschichte von Cypris.

Marburg. 1868.
Yves Delage. Evolution de la Sacculine, Crustace cndoparasite de Vordre nouveau

des Kentrogonides. Arch, de Zoologie cxperimentale de H. de Lacaze-Duthiers.

2°. 2dVol. Paris. 1884.
C. Grobben. Die Entwickelungsgeschichte der Moina rectirostris. Arb. aus dem

Zool. Institute zu Wien. 2 Bd. 1879.
The same. Die Entwickelungsgeschichte von Cetochilus septcntrionalis. Ibidem. 3

Bd. 1881.
Heinr. Reichenbach. Studien zur Entwickelungsgeschichte des Flusskrebscs, in Abh.

d. Senkenb. Nat. Gesellsch. zu Frankfurt. 14 Bd. 1886.
In addition, treatises of E. van Beneden, Bobretzky, W. K. Brooks, Claus, Dohrn,

Grube, Hoek, Joly, Lereboullet, Krohn, Kossmann, P. Mayer, Fritz Miiller,

Pagenstecher, Richters, Uljanin, Spence Bate, Thompson, Metschnikoff, etc.

4U

COMPARATIVE ANATOMY

CHAP.

First Appendage to the Class of Crustacea.
The Trilobites, Gigantostraea, Hemiaspidse and Xiphosura.

I. The Trilobites.

These extinct Articulata are found only as fossils, and only in
palaeozoic formations.

Body (Fig. 281). The integument of the upper side was hard;
but on the under side soft. The body falls into 3 divisions,
cephalic shield, thorax, and caudal shield (pygidium). Each of these
divisions is again divided by 2 almost parallel longitudinal dorsal
furrows (the thorax most distinctly) into an arched middle area

FIG. 282.— Restored trunk segment of a Trilobite,
transverse section (after Walcott). ?, Rhachis ; p,
pleura ; ep, epipodial appendages ; en, endopodite ;
ex, exopodite ; d, intestine.

FIG. 281.— Cheirurus Quenstedtii,
dorsal view.

(rhachis) and 2 lateral areas (pleura). The cephalic shield is
unsegmented, semicircular, or crescent-shaped, with the rounded part
to the front; it generally carries 2 large compound eyes. The
occurrence of simple eyes is very doubtful. The thorax consists of a
varying (usually rather large) number of freely moving segments.
The caudal shield seems composed of a varying number of segments
more or less completely fused together. Nearly all (or all ?) Trilobites
were able to roll up their bodies like woodlice, so that the anterior
edge of the cephalic shield and the posterior edge of the caudal shield
touched one another.

Limbs (Fig. 282). — These are retained only in rare cases. They
are slender and long, and more or less like one another, segmentally
repeated from the cephalic shield to the end of the caudal shield.
Under the cephalic shield there are 4 pairs of limbs reckoned as

v GIGANTOSTRACA 415

maxillipedes, the most posterior of these being more developed than
the rest ; the most anterior is inserted behind the upper lip. The
limbs of the trunk and pygidium are biramose with long endo-
podite and short exopodite, and with bifurcated epipodial appendages
on the basal joint, which may be either filamentous or ribbon-shaped,
simple or spirally twisted ; these may safely be assumed to be gills.
These gills also seem to occur on the limbs of the pygidium, but in
a reduced condition. The enteric canal runs through the body in a
straight line to the end of the caudal shield. In front of the mouth,
at the lower and anterior edge of the cephalic shield, is found a shell-
piece which is called the upper lip.

In some of the genera of TriloUtes a pretty complete series of successive stages of
development (larval stages) has been discovered. In the youngest stages the cephalic
shield is present, but the trunk is still quite incomplete. The development of the latter
generally occurs in such a way that the pygidium takes precedence of the thorax,
and that new segments continually become differentiated at the anterior end of the
pygidium. In other words, the thorax differentiates in the order from before back-
ward.

The available material suffices to show : (1) that the Trilobites are Arthropoda,
and (2) that they are most nearly related to the Crustacea. Their Crustacean
character is unmistakably supported by the fact that the trunk feet are biramose,
and carry epipodial appendages. A closer comparison with distinct orders of
Crustacea, however, cannot be carried out, because in the Trilobites no limbs have
been found placed in front of the mouth, and comparable with the anterior antennae.
Whether these were entirely wanting, or rudimentary, or not capable of petrifaction,
is quite uncertain. Should they still be found, then the 5 cephalic limbs could
without difficulty be referred to the 5 typical limbs of the Crustacean head, and the
Trilobites might then be regarded as original Untomostraca, to be derived from the
same racial form as the Phyllopoda. We are prevented from comparing them with
the Malacostraca on account of the different • segmentation of the body and the
inconstant number of the segments which does not agree with that of the Mala-
costraca.

The Trilobites were marine. Agnostus (with only 2 thoracic segments),
Trinucleus, Olenus, Paradoxides, ConocepJialites, Sao, Calymene, AsapJius, Bronteus,
Phacops, Cheirurus, Acidaspis, Lichas, Proetus, Harpes.

II. The Gigantostraea (Merostomae, Eurypteridse).

These are also extinct. They lived during the palaeozoic epoch.

Body (Fig. 283). — The elongated scale-covered body of these, the
largest of all Arthropoda, falls into head (cephalothorax ?), thorax, and
abdomen. The unsegme'nted head is relatively small, and carries 2
compound lateral eyes, and very near the median line 2 ocelli. The
thorax and abdomen each consist of 6 segments. The 6th abdominal
segment is followed by a caudal stylet or a fin - shaped terminal
segment.

Extremities. — The division described as the head carries 6 pairs
of extremities, composed of simple rows of joints, and thus not

416

COMPARATIVE ANATOMY

CHAP.

biramose. The anterior pair lies in front of the mouth, and in
Eurypterm is a small finely jointed pair of feelers, while in Pterygotus

it is a pair of long chelate
limbs, with powerful pin-
cers. The basal joints
of the five subsequent
pairs of limbs are inserted
round the mouth, and
have masticatory ridges
directed inwards, which,
in Pterygotus, are specially
strongly developed on
the last pair.

The last pair of ceph-
alic limbs is much more
strongly developed than
the rest; they are oar-
shaped, and evidently
served as swimming feet.
On the under side of the
thorax of 6 segments are
found 5 plates (leaf-like
limbs) consisting of 2
lateral halves ; these
:e tiles, and
cover leaf -like gills. The
most anterior largest plate is called the operculum. The abdomen
is devoid of limbs. Behind the mouth is a large oval plate, the
metastoma.

The systematic position of the Gigantostraca is not clear. In the number and
position of the cephalic feet they agree with the Xiphosura. The leaf-like limbs of
the thorax also somewhat recall those of Limulus, and above all the operculum in
the 2 groups seems to be homologous. Their relation to the other Arthropoda,
especially to the Crustacea and the Scorpionidce, whom they resemble in appearance,
has often been pointed out, but nevertheless is not clear, because the morphological
significance of the limbs, and especially of those which lie in front of the mouth, is
not known with anything like certainty. Our knowledge of the structure of the
Gigantostraca and Trilobites has during the last decade received such unexpected
additions through palseontological investigations (e.g. discovery of the limbs of
Trilobites} that we may hope for further advance.

Eurypterus, Pterygotus. •

III. The Hemiaspidse.

These are extinct palaeozoic forms apparently related to the
Xiphosum and perhaps forming a sort of connecting link between them
and the Gigantostraca.

The body falls into 3 regions : a head of considerable size covered

FIG. 283.— Pterygotus osiliensis. Upper Silurian. Under
side restored, and smaller than natural size (after F. Schmidt).
a, Epistoma; me, metastoma ; 1-6, feet; 1, chelicer; 6, rowing
foot with large masticatory ridge, Tel ; I-V, ventral plates ; i

au, eye. Overlap

v XI P HO SUE A 417

by a shield, a thorax consisting of 5 or 6 free (rarely fused) rings, and
an abdomen consisting of 3 or more segments, and followed by a
strong caudal stylet. The cephalic shield often has 2 compound
lateral eyes, ocelli are wanting. Two dorsal longitudinal furrows give
the thorax an appearance like that of the Trilobites. The extremities
are unknown.

Bunodes, Hemiaspis, Belinurus.

IV. The Xiphosura (Pceeilopoda, Limulidse).

The body, which is covered by a hard and thick chitinous carapace
falls into 2 principal divisions, the cephalic shield (cephalothorax) and
the hind body (abdomen). The hind body is followed by a large and
long post-anal caudal spine which can move independently of the
body. The cephalothorax is very large, almost crescent-shaped, with
2 lateral horns directed backwards. On its dorsal side are the 2
compound lateral eyes, and in front of them, nearer the median line, 2
ocelli. The flat segmented abdomen, which articulates with the
cephalothorax, is almost hexagonal.

The upper side of the cephalothorax is arched, and the under side
is concave.

Limbs (Fig. 284). — There are 7 pairs of limbs on the cephalothorax.
The most anterior lie in front of the mouth ; they are small and end in
pincers. They are innervated from the anterior part of the oesophageal
commissure, and (if a comparison with Crustacea is allowed) should
perhaps correspond with the 2d pair of antennae, while limbs
corresponding with the anterior antennae are wanting. The 1st pair of
limbs, which are called ehelieerse, are followed by 5 strong long pairs
of limbs, also ending in pincers. They arise at the sides of the slit-
like oral aperture, and each has a masticatory ridge on the basal
joint. The masticatory ridges of the 4 anterior pairs of feet are
armed with spines, those of the 5th pair, however, have a sharp cut-
ting inner edge. The 5th pair of feet is further distinguished by
a different formation of the terminal joint and by an appendage
on the basal joint, which has been considered to be an exopodite.
Behind the mouth there are two stylet - shaped processes, the
ehilaria.

The 7th pair of limbs of the cephalothorax, which arises at its
posterior edge, is quite differently formed from the preceding, and
agrees far more with the abdominal limbs. It is called the opereulum,
and consists of 2 plates, which are united in the middle line, and
cover the subsequent abdominal limbs. The latter, of which there
are 5 pairs, are leaf-like, and shaped like the opereulum. In all the
leaf -like feet 2 rows of areas are marked by sutures, an outer row of
large areas (exopodites) and an inner row of smaller areas (endopodites).
* The last areas or joints project freely. The leaf -shaped abdominal
feet on each side carry on the upper surface (i.e on the surface turned

VOL. I 2 E

418

COMPARATIVE ANATOMY

to the body) a gill consisting of numerous integumental folds,
arranged like the leaves of a book.

FIG. 284.— Limulus polyphemus, young specimen from the ventral side (after Packard).

The leaf-like feet serve for swimming as well as for breathing.
A sternal endoskeleton is present.

XIPHOSURA

419

Nervous System. — The central nervous system consists of a
ganglionic mass lying in the cephalothorax and surrounding the ceso-

ft

feU cffi. A

C

fe

FIG. 285.— Median longitudinal section through a young Limulus polyphemus (after
Packard), cth, Cephalothorax ; ab, abdomen ; ss, caudal spine ; vm, anterior stomach ; m
stomach; 7i, heart; ss, sternal cartilaginous endo-skeleton ; d, intestine; g, brain; itg, infra
oesophageal ganglionic mass ; &m, ventral chord ; o, operculum ; «i-«5, gill carrying abdominal
appendages ; mu, mouth ; an, anus.

phagus ; it is continued into the abdomen as a ganglionic ventral chord.
The circular ganglionic mass of the cephalothorax is found to consist
of the brain, which lies in front of the oesophagus, and gives off nerves
to the lateral eyes and the ocelli, and of 7 postoral pairs of ganglia with
their transverse commissures, these ganglia lying near each other.
The latter yield the nerves for the cephalothoracic limbs. The ventral
chord of the abdomen consists of 6 ganglia, the last of which is the
largest. The nerves to the leaf -like feet are given off by these
latter.

The compound eye of Limulus (Fig. 286) deserves special descrip-
tion. The chitinous cara-
pace is thickened over each
of the two compound eyes ;
outwardly its surface is
smooth, but inwardly, by
the formation of conical
processes, it pushes in the
subjacent hypodermis to
form towards the interior
an equal number of papillae ;
each of these chitinous cones
may be considered as a
single lens. ^ A completely FlG> 2S6.-Part of a section through the eye of Lim-
Separate retinula probably Ulus. cp, Chitiuous carapace ; cv, papilla-like thickenings

conm'stinp- of 10 rells with of the same over each single eye: hy' hvP°dermis; ret,
us, \A iLn retinul8e . %j nerves of the single eyes>

rhabdom, pigment, and

nerve, corresponds with each of these single lenses. The retinulae lie

420 COMPARATIVE ANATOMY CHAP.

in the hypodermis. The compound eye of Limulus is thus seen to be
composed of numerous independent unilaminate single eyes crowded
together. Each single eye corresponds with the unilaminate eye of
certain Myriapoda and Scorpionidce, and the whole compound eye cor-
responds with the sum of the eyes on one side of these Arthropoda,
except that the chitinous carapace in Limulus forms a thickening
common to all the single eyes.

Enteric Canal. — From the large mouth a long oesophagus rises
upward and forward, to enter a muscular masticatory- or fore-stomach
placed in the anterior part of the cephalothorax ; the chitinous intima
of this stomach projects in numerous longitudinal folds into its lumen.
The fore-stomach is followed by a long straight mid-gut widened at its
commencement ; this runs through the cephalothorax and abdomen and
opens externally at the base of the caudal spine through a ventral anal
aperture of the short hind-gut. The mid-gut receives the 4 ducts of 2
pairs of hepatopancreatic glands which branch freely in the cephalo-
thorax. All through the intestine, except in the mid-gut, a chitinous
intima is found.

•Circulatory System. — The heart is an elongated dorsal vessel
provided with 8 pairs of ostia which can be closed by valves.

Sexual Organs. — The sexes are separate. The male, which is
smaller than the female, is further distinguished externally by the fact
that the most anterior or the two anterior postoral pairs of limbs do
not end in pincers, but in claws.

The 2 ovaries are tubes forming a network of branches, thos« of
the two sides communicating with each other at various points. The
two oviducts form a sac-like wider portion before they emerge. The
female sexual apertures lie on the inner side of the opercular plates (the
side turned to the body), at their bases, to the right and left of the
median line. The two testes consist of a large number of vesicles
dispersed throughout the body and attached to sperm ducts which
branch and anastomose freely. The male apertures have the same
position as the female.

Coxal Glands. — On each side of the cephalothorax lies a red gland
of considerable size whose outer aperture has only been found in
young animals on the basal joint of the fifth pair of limbs. It is un-
certain whether these coxal glands correspond with the shell glands of
the Crustacea (which also emerge on the 5th pair of extremities, i.e. on
the 2d maxillae). We have no right to assume that the 5th pair of
extremities of Limulus answers to the 2d pair of maxillae of the
Crustacea ; it is indeed improbable that this is the case.

Ontogeny. — The 6 anterior pairs of limbs appear first and simultaneously, then
follow the 7th pair (operculum) and the 8th (first gill-carrying abdominal limbs).
On the cephalothorax there are indications of segmentation. The young Limulus
hatched from the eggs shows a complete rudimentary cephalothoracic shield, the
segmentation having then entirely disappeared. The abdomen, on the contrary,
appears distinctly formed of 8 segments, but these are not movable upon each other.

v XIPHOSURA 421

The caudal spine is still a very short and simple plate. The 2 compound eyes
and the 2 ocelli are already present. Behind the 1st pair of abdominal feet the
rudiments of the 2d are visible. At this stage the larva has the appearance
of a Trilobite, and this similarity is increased by two dorsal longitudinal furrows.
The gradual transition from the Trilobite stage to the Limulus stage is brought about
by the appearance in order from before backward of the abdominal leaf-like feet. The

A

I

FIG. 287.— Limulus polyphemus in the so-called Trilobite stage. A, Dorsal side ;
B, ventral side (after Kingsley).

leaf-like feet become biramose. Gills develop on them ; in the Trilobite stage rudi-
ments of gills were found only on the most anterior pair of abdominal feet. The
abdomen loses its segmentation. The caudal plate elongates by degrees into the
caudal spine.

Systematic position. — The relationship of the Xiphosura to the Crustacea is in
any case very distant, since it is at present impossible to homologise the extremities
of Limulus with typical Crustacean extremities, or to compare in detail the segmenta-
tion of Limulus with that of any Crustacean. The biramose character of the leaf-like
feet and of the 6th pair of thoracic feet is the only specific Crustacean characteristic
shown by the Xiphosura, if we leave the gills out of consideration. The relation
between the Xiphosura and .the fossil Hemiaspidce and Gigantostraca is evidently
much closer. The Xiphosura, Nemiaspidce, and Gigantostraca are themselves again
perhaps racially connected with the TriloUtcs. In any case, however, in the present
state of science, it seems probable that all these groups are only connected at their
roots with the Crustacea; The relations of the Gigantostraca and Xiphosura to the
Arachnoidea, especially to the Scorpions, which is assumed by some observers, will
be discussed later.

Single genus Limulus. — Marine, L. moluccanus, Molucca, Sunda Islands ; L.
polyphemus, East coast of North America.

Most Important Literature.

For general guidance.

Karl A. Zittel. Handbuch der Palceontologie. 1. Abth. 2. Band. Mollusca und
Arthropoda. Miinchen und Leipzig, 1881-1885.

Trilobitze.
H. Burmeister. Die Organisation der Trilobiten. Berlin, 1843.

422 COMPARATIVE ANATOMY CHAP.

J. Barrande. Systeme silurien du centre de la Boheme. Vol. I. Prague, 1852.

Supplement. 1874.
J. W. Salter and H. Woodward. A Monograph of British Trilobites. Palcconto-

grapliical Society, 1867-1884.
Fr. Schmidt. Revision der ostbaltischen silurischen Trilobiten. 1. Him. de I'Acad.

imp. de St. Petersburg. Ser. VII, tome 30. 1881.
C. D. Walcott. The Trilobitce. New and old evidence relating to its organisation.

Bull. Mus. Comp. Zoology, Cambridge, Mass. Vol. VIII. 1881.

Gigantostraca and Hemiaspidse.

J. Nieszkowski. Der Eurypicrus remipes aus den obersilurischen Schichten der
Insel Oesel, in Archiv f. Naturgeschichte Liv-, Est-, und Kurlands. 1. Ser.
2. Band. 1859.

H. Woodward. A Monograph of British fossil Crustacea belonging to the order
Mesostomata. Palwoniographical Society. Parts I-V. 1866-1878.

Works of Huxley, Salter, Woodward, Baily, Schmidt, etc.

Xiphosura.

Alph. Milne-Edwards. Eecherches sur I' Anatomic des Limules, in Ann. Sciences

naturelles, 5° serie, t. XVII. Paris, 1873.
A. J. Packard. The Anatomy, Histology, and Embryology of Limulus polyphcmus,

in Mem. Boston Society Natural History. Boston, 1880.
E. Ray Lankester. Limulus an Arachnid, in Quart. Journ. Micr. Science, vol.

21. London, 1881.
J. S. Kingsley. Notes on the Embryology of Limulus, in Quart. Journ. Micr.

Science, vol. 25. London, 1885.
Treatises of J. van der Hceven, Gegenbaur, Packard, Dohrn, R. Owen, Gulland,

etc.

Second Appendage to the Class of the Crustacea
The Pantopoda (Pyenogonidse).

The body, in comparison with the long and slender limbs, is
extremely reduced, and falls into three divisions — proboscis or beak,
trunk, and hind -body. The proboscis articulates with the most
anterior trunk segment. At its point lies the mouth surrounded by
three lips, and it contains internally the greater part of the fore-gut
(" fish trap " apparatus). It consists of three pieces lying side by side
longitudinally, an upper median piece and two lower lateral pieces.
The trunk consists of 6 segments, the three anterior of which are
always fused together ; it has lateral outgrowths on to the ends of
which the limbs are hinged. The hind-body is unsegmented, short,
truncated, and devoid of limbs.

Extremities. — 7 pairs of extremities occur typically. The first
extremities (1), the chelicerse, are innervated from the brain, and in
the young animal end in pincers ; in the adult they are often reduced
or altogether wanting. The following extremities (2-7) are inner-
vated from the ganglia of the ventral chord, 2 and 3 from the most
anterior ganglion, which consists in the larva of two separate ganglia.

PANTOPODA

423

The 2d pair of extremities is generally shorter than those which follow,
and in several genera is wanting. The 3d pair of extremities is
developed in the males of all Panfopoda, and functions in them as
egg-carriers. In several genera it is wanting in the females. The
extremities 4-7 are never absent, they consist of nine joints, and
end in claws, and are in comparison with the body exceedingly
long, giving the animal a spider-like appearance. All the extremities
are uniramose.

The nervous system in the adult animal consists of a supra-
cesophageal ganglion, an oesophageal commissure, and a ventral chord.
The latter has 4 or 5 pairs of ganglia, from which arise the nerves
for extremities 2 - 7 ; it ends posteriorly with one or two pairs of
reduced ganglia, the last of which gives off nerves to the abdomen.

FIG. 288.— Nymphon hispidum, male, ventral side (after Hoek). The setae are omitted.
1-7, Limbs ; (1, chelicerae ; 3, egg-carriers) ; s, proboscis ; ab, abdomen.

Since the most anterior ganglion of the ventral chord consists of
2 or 3 pairs of ganglia, distinct in the larva, the complete number
of the pairs of ganglia in the ventral chord is 8, 7 of which
belong to the trunk and 1 to the abdomen. According to this the
number of trunk segments and also of pairs of limbs must originally
have been 8. From the supra-cesophageal ganglion the nerves for the
eyes arise, and also those for the first pair of limbs and some of those
for the proboscis. Some of the other proboscis nerves have their roots
in the anterior portion of the first ventral ganglion. The nervous
system of the proboscis with its ganglion is very complicated.

Four eyes, each with a cuticular lens and a retina surrounded by
pigment, lie on a prominence on the dorsal side of the first trunk
segment.

The enteric canal has 3 divisions : fore-gut, mid-gut, and hind-gut.
The fore-gut, placed in the proboscis, has a complicated inner frame-
work and a "fish trap" apparatus. The straight mid-gut is provided

424 COMPARATIVE ANATOMY CHAP.

with long eceea which project into extremities 1, 4, 5, 6, 7, some-
times as far as the terminal joint.

The anus lies at the end of the hind-body. Special respiratory
organs are wanting. The heart has 2-3 pairs of ostia; its dorsal
wall is fornled by the dorsal integument.

The sexes are separate. The sexual glands are paired tubes, which
extend through the trunk at the sides of and above the intestine, and
are connected behind the heart by an unpaired piece. They give off
accessory tubes into the extremities 4-7, which emerge on the 2d joints.
In the male, however, the genital apertures are wanting in the 4th
pair of extremities, and generally also in the 5th. In Pycnogonum and
Rliyncliothorax only one aperture is found on each side in the females,
and this is in the 7th pair of extremities. In the males, in the fourth
joint of extremities 4-7 cement glands (coxal glands?) are found,
whose secretion glues the eggs which issue from the female genital
apertures into balls, which are carried about by the male on the 3d pair
of extremities, transformed into egg-carriers.

We find glands which are considered as excretory organs in the
2d and 3d pairs of extremities, emerging on the 4th or 5th joints.

Ontogeny. — Most Pycnogonidce pass through a more or less complicated metamor-
phosis. The youngest unsegmented larva carries 3 pairs of extremities, correspond-
ing with extremities 1, 2, and 3 of the adult. The first ends in a claw. In spite
of the agreement in the number of extremities this larva shows no near agreement
with the Nauplius larva, and the extremities themselves, since they all consist of
only one row of joints, do not show the character of the Nauplius limb. In the next
stage new segments appear at the posterior end of the body and differentiate in the
order from before backward. The enteric cceca at first do not project into the
extremities.

The cause of the entrance of enteric cceca and lateral tubes of the sexual glands
into the interior of the limbs must be sought in the extraordinary reduction of the
trunk and in the great longitudinal development of the limbs.

The Pantopoda seem to occupy an isolated position among the Arthropoda. On
account of the want of a typical Nauplius or Zocea larva we are not justified in
placing them near the Crustacea, and they show no evident relation to any other class
of Arthropoda. Many zoologists consider the Pantopoda to be related to the spiders,
and establish the following homologies for the limbs. Extremity 1 = chelicerse or
falces ; extremity 2 + the paired piece of the proboscis = lower jaw and pedi-
palps ; extremities 3-6 = the 4 pairs of legs of the spider. Extremity 7 is want-
ing in the adult spiders, but it is pointed out that in a few Arachnoidea the
rudiments of paired extremities temporarily appear on the abdominal segments
during embryonic development. On the other hand it must be remarked that the
connection of the two paired pieces of the proboscis of the Pantopoda with the
second extremity, and the homology of the two parts taken together with the lower
jaw and pedipalp of the Arachnoidea, is by no means proved. The inner organisa-
tion and the development give little footing for a special comparison of the Pantopoda
with the Arachnoidea, since the coeca of the mid-gut have no great morphological
significance.

The Pantopoda are exclusively marine. Nymphon, Pallene, Phoxichilidium,
Ammothea, Pycnogonum. Collossendeis gigas is a gigantic form in the deep seas.

v PANTOPODA 425

The longest extremity sometimes measures 30 cm., while the whole body only attains
a length of 8 cm.

Most important Literature.

Anton Dohrn. Die Pantopoden des Golfes von Ncapd. Eine Monographic in Fauna

und Flora des Golfes von Neapel. 3. Band. 1881.
P. P. C. Hoek. Report on the Pycnogonida, in Report on the scientific results of the

voyage of H. M.S. Challenger. Zoology, vol. III. 1881.
In these works further bibliography is given.

CHAPTER VI

The second division of the Arthropoda. — Organisation and Development of the Air-
breathing Articulata (Tracheata).

Systematic Review.
CLASS I. Protracheata (Onychophora).

Body vermiform. A pair of preoral feelers at the point of the head. In the
oral cavity a pair of horny jaws, at the side of the mouth oral papilla? (slime papillae),
numerous pairs of short almost truncated limbs. Eespiration through tubular
tracheae, whose outer apertures are scattered over the whole body. Numerous seg-
mentally arranged pairs of nephridia. Coxal glands on the legs. Heart a long
dorsal vessel with numerous pairs of ostia.

CLASS II. Antennata (Myriapoda and Hexapoda).

One pair of preoral feelers, 3 pairs of oral limbs. Trunk either homonomously
segmented with numerous pairs of jointed legs, or heteronomously segmented with
the limbs limited to the three segments of the anterior region of the trunk, the
thorax ; the posterior region, the hind-body, being limbless. The head everywhere
distinctly marked off from the trunk. Kespiration by means of tubular tracheae, whose
outer apertures (stigmata) are segmentally arranged. The heart in the homono-
mously segmented Myriapoda is a long dorsal vessel, supplied with many segmentally
arranged pairs of ostia ; it runs through the trunk longitudinally : in the Hexapoda
it is restricted to the abdomen.

CLASS III. Chelicerota sive Arachnoidea.

No preoral limbs comparable with the antennae of the Antennata or Protracheata.
Several anterior body segments (7, including the frontal lobes) are fused to form an
unsegmented region, called the cephalothorax. This carries 6 pairs of extremities,
the most anterior of which lies in front of the mouth. The two anterior pairs are
developed as oral limbs. The first are called the jaw-feelers (chelicerae), and the
second jaw-palps (pedipalps). The 4 remaining pairs of extremities are jointed
legs, generally long. Abdomen segmented or unsegmented, or fused with the
thorax, Avith no developed limbs. Respiration either exclusively through book-leaf
tracheae, or at the same time through book-leaf and tubular tracheae, or exclusively
through tubular tracheae. Number of stigmata limited, at the most 4 pairs.
The stigmata almost always lie in the abdomen. Heart, seldom wanting, restricted
to the abdomen.

CHAP. VI

PROTRACHEATA

427

CLASS I. The Protraeheata (Onyehophora).

An accurate knowledge of the organisation of the only genus belong-
ing to this division, Peripatus, is of the greatest importance, because it
combines unmistakable Annelid with unmistakable Tracheatan char-
acteristics. Of all living Arthropoda, Peripatus has perhaps best
preserved the original organisation of the ancestors of the Traclieata.

Body. — The soft-skinned body is long and in
appearance strikingly recalls certain Annelida,
e.g. Hesione. The body is slightly flattened
dorso-ventrally, with arched dorsal side and

tolerably flat ventral side. The

integument is

ft °f

4

Fio. 289. — Peripatus
Novae Zealandiaa (after
Sedgwick).

FIG. 290. — Peripatus Ectwardsii, Head from the under side.
a, Basal portions of the antennae; op, oral papillae. The figure
shows also the papillae surrounding the entrance to the oral cavity,
and the jaws within the same.

transversely ringed. The limbs are the only
external indications of segmentation. The head
is fairly distinct from the trunk. Over the
whole body are scattered the wart-like papillae,
each of which carries a hollow spine at its point.
On each side of the head dorsally there is an eye,
and in the middle of the ventral side is the
mouth. The anus lies terminally at the posterior end of the body.

Extremities. — The head carries anteriorly and dorsally two ringed
and somewhat slender antenna. In the buccal cavity on each side lies
a sickle-shaped jaw, consisting of two chitinous plates toothed along
their inner sharp edges. On each side of the mouth arises a short
ringed process beset with papillae. This process is called the oral or
slime papilla. The trunk carries a varying number (14 to 42, accord-
ing to the species) of similarly formed limbs. These extremities
are placed laterally, at that part where the arched dorsal side bends
in to meet the flat ventral side. Each extremity is truncated,

428

COMPARATIVE ANATOMY

CHAP.

with transverse rows of papillae, which give it a ringed appearance,
and each falls into two parts, a larger proximal conical leg and a
smaller, narrower, distal foot, ending in two chitinous claws.

Integument. — The body epithelium (hypodermis) is externally
covered by a delicate chitinous cuticle. The spines on the papillae and
claws on the feet also belong to the cuticular formation secreted by the
hypodermis. Under the hypodermis lies a peculiar subepithelial layer,
formed of fibres running in various directions.

Musculature (Fig. 291). — Peripatus has a strongly developed dermo-
muscular tube, which consists of the following layers : (1) an external
layer of circular muscular fibres ; (2) a double layer of fibres crossing
each other diagonally; (3) internally a strong longitudinal layer, con-

t/m

rt

FIG. 291.— Transverse section through the antepenultimate segment of a female Peri-
patus Edwardsii (after Gaffron). n, Longitudinal trunks of the nervous system ; h, heart (con-
tractile dorsal vessel) ; 7m, longitudinal muscles ; ut, uterus ; d, intestine ; on, outer apertures of
the nephridia.

sisting of various bundles, whose arrangement on a transverse section
of the body is shown in Fig. 291. In addition to these layers there
are numerous sagittal or transverse muscle fibres corresponding with
the dorso-ventral or transverse muscle fibres of the Annulata. Some
of these fibres run through the body cavity in such a way as to divide -
it into a median and two lateral divisions, the former containing the
heart, the intestine, and the sexual organs, the latter the longitudinal
trunks of the nervous system and the segmental organs. The muscul-
ature of the extremities chiefly proceeds from the layer of diagonal
fibres and the sagittal musculature. Special muscles serve for moving
the jaws, claws, etc.

None of the muscle fibres of Peripatus, except those of the
jaw muscles,1 are transversely striated.

The alimentary canal runs nearly straight through the body.

1 This exception, however, does not hold good for P. Leuckartii.

VI

PROTRACHEATA ,

429

It falls into the following divisions : buccal cavity, pharynx, oeso-
phagus, mid-gut or stomach-intestine, and rectum. The buecal cavity,
in whose base the mouth proper lies, .arises ontogenetically by the
growing together of a row of papillae surrounding the mouth ; the
mouth and jaws are thus enclosed within a circular wall. In front
of the mouth, within the buccal cavity, lies a median prominence, the
tongue. At the back of the buccal cavity, where it passes into the
pharynx, i.e. at the posterior edge of the
mouth, is an invagination directed back-
wards, into which the unpaired terminal
portion of two salivary glands enters ;
these glands are long tubes running
through the greater part of the body
longitudinally in the lateral divisions of
the body cavity. At the anterior end,
near the bend towards the common ter-
minal division, each salivary gland has
a ccecal vesicular appendage. The
pharynx, which reaches to the region
between the first and second pair of
legs, possesses a very thick muscular
wall : its lumen in a transverse section
is Y-shaped. The oesophagus is shorter.
Its wall, which consists of an outer
longitudinal and an inner circular mus-
cular layer, is much thinner than that of of Peripatus capensis, ventral side, laid

open (after Balfour). a, Antenna; z,
tongue ; fc, jaw ; dg, salivary gland ; gs,

3s

FIG. 292.— Anterior end of the body

are

common terminal portion of the two
salivary glands ; ph, pharynx ; ce, oaso-
phagus ; I, the lip papillae surrounding
the buccal cavity ; op, oral or slime
papillge ; sld, duct or reservoir of the
slime glands.

the pharynx. These three divisions
lined by the chitinous cuticle.

The stomach - intestine stretches
from near the two pairs of legs almost
to the end of the body. Its wall is in
folds, and its muscular layer (outer cir-
cular and inner longitudinal, i.e. the reverse of what obtains in all
anterior sections of the canal) is exceedingly thin. It is nowhere
fastened to the body wall by mesenteries. The rectum, which is
distinctly separated from the mid-gut, is considerably narrower, with
a tolerably well-developed muscular wall.

An endothelium (peritoneal epithelium) covers the outer muscular
wall of the enteric canal and the other organs lying in or forming the
boundary of the body cavity.

The nervous system of Peripatus (Fig. 293) consists of a large
brain placed in the head in front of and over the pharynx (supra-
oesophageal ganglion), and of two ventral longitudinal nerve trunks
proceeding from the brain, which run far apart in the lateral divisions
of the body cavity to its posterior end. In each segment, i.e. in each
part of the body corresponding with a pair of extremities, the longi-
tudinal nerves are connected by several transverse commissures (9 -10 in

430

COMPARATIVE ANATOMY

CHAP.

Peripatus capensis). The longi-
tudinal trunks show slight seg-
mental swellings corresponding
with the extremities ; they are
also provided with a continuous
layer of ganglion cells. At the
posterior end of the body they
pass over into each other above
the rectum by means of a con-
necting portion in which the
layer of ganglion cells is want-
ing. Numerous nerves go off
laterally from the longitudinal
trunks along their whole course ;
these diverge at regular inter-
vals and more or less correspond
with the transverse commissures.
Each extremity is supplied with
two such lateral nerves. The
nerves for the jaws diverge at
the points where the longitud-
inal trunks enter the brain ;
rather farther back (or at the
posterior end of the cesophageal
commissure) arise the nerves for
the oral papillae. Besides small
nerves, the brain gives off strong
nerves to the antennae. From
its lower portion two more nerves
arise which run to the tongue
and the dorsal wall of the
pharynx, and unite at the com-
mencement of the oesophagus into
an unpaired mediodorsal nerve
forming a sympathetic nervous
system.

The two eyes correspond to
some extent in their structure

FIG. 293.— Anatomy of Peripatus cap-
ensis (after Balfour). The enteric canal
behind the pharynx is cut off and removed.
g, Brain ; a, antenna ; op, oral or slime papil-
lae ; sd, slime glands ; sr, slime reservoir, which
at the same time acts as duct to the glands ;
§04, sog, SOQ, SOQ, nephridia of the 4th, 5th, 6th,
and 9th pairs of limbs ; cd, elongated coxal
gland of the last pair of feet ; go, genital aper-
ture ; an, anus ; ph, pharynx ; n, longitudinal
trunk of the nervous system.

VI

PROTRACHEATA

431

with the Aldopidan eye of the Chcetopoda described on p. 230. But the
space containing fluid between the lens and the rod layer is wanting.
The Peripatw eye proceeds ontogenetically from a hollow invagination
of the cephalic ectoderm near the rudiments of the brain. The
invagination closes and becomes the optic vesicle. The connection
with the brain is said to arise later by the growing out of the optic
nerve from the brain.

The circulatory system consists of a contractile dorsal vessel or
heart running through the body from the first segment to the last but
one. This heart is supplied with paired ostia arranged segmentally
and provided with valves. It lies in a pericardial sinus imbedded on
its ventral side in a tissue comparable with the fat body of the Insecta.
This sinus is incompletely separated from the underlying body cavity
by a horizontal septum. The septum, which is formed of endothelium
and muscle fibres running transversely, is fenestrated on both sides of
the middle line. A median longitudinal nerve runs on the dorsal wall
of the heart, as in the lulidce. Besides the heart a very fine medio-
ventral longitudinal vessel is said to occur.

Excretory organs (nephridia). Each trunk segment of Peri-
patus is provided with a pair of nephridia. The nephridia lie in
the lateral divisions of the body cavity and
emerge on the under side of the extremities
near their bases. Each nephridium (Fig.
294) consists of the 3 following parts : —
(1) a terminal vesicle which opens out-
wards through the external aperture ; (2)
a looped nephridial canal bent back upon
itself, which ends in (3) a funnel opening
into the body cavity and placed near
the terminal vesicle. We here find then
the typical structure and arrangement
of an Annulatan nephridium. In the 4th
and 5th segment the nephridial canal is
distinctly longer and forms many loops.
The nephridia of the first three trunk* seg-
ments are much smaller than the others ;
their canal is short and without loops.
According to the species, the nephridia
are apparently wanting in the penultimate
or antepenultimate segment, i.e. the last or
penultimate limb-bearing segment. It has,
however, been proved that the duets of
the sexual organs which emerge here
are transformed nephridia. Transformed
nephridia are also found in two so-called anal glands, which
open in the last body segment (the anal segment without extremi-
ties) near the anus. These glands are wanting in the adult female,

FIG. 294.— A nephridium of Peri-
patus Edwardsii (after Gaffron).
tr, Funnel ; sg, looped canal, or ne-
phridial duct ; eb, terminal vesicle of
the nephridium.

432 COMPARATIVE ANATOMY CHAP.

but appear temporarily in the course of development. It has also
been ontogenetically proved that the salivary glands are the trans-
formed nephridia of the segment carrying the oral papillae
which has fused with the head ; this is of extreme importance.
Two blindly ending canals which open into the oral cavity near the
base of the jaws have been pointed out as nephridia of the jaw
segment reduced to their eetodermal portion.

Respiratory Organs. — It is a fact of great importance that Peripatus
possesses the respiratory organs which are characteristic of the Tracheata,
and which occur only in them. They consist of long, very fine, and
thin chitinous tubes filled with air and widely dispersed through the
body ; in Peripatus they are not branched ; they emerge united into
tufts at the base of a flask -shaped depression of the integument.
The outer aperture of such a depression may be called, as in the
Tracheata, the stigma. In P. Edwardsii the stigmata occur in great
numbers and quite irregularly all over the surface of the body. In
P. capensis, on the contrary, at least some of the stigmata show a definite
arrangement, viz. in longitudinal rows — on each side two, one dorsally
and one ventrally. The stigmata in a longitudinal row are, however,
more numerous than the pairs of legs.

Leg Glands (eoxal glands) and Slime Glands. — In Peripatus
capensis, in both sexes, there are paired glands emerging on the under
side of the extremities, and only wanting in the first pair of trunk
limbs. Every such coxal gland consists of a sac lying in the lateral
division of the body cavity, and of a duct. The coxal glands of the
last pair of feet are extraordinarily long in the male, and stretch far
forward to near the middle of the body (Fig. 293, cd). In P. Edwardsii
leg glands occur only in the males, not in each segment, but only in a
certain number of segments lying in front of the genital segments ;
one, two, or three glands may occur in each limb.

There are two largely -developed thickly -branched slime glands
(Fig. 293, sd), which must be considered as transformed leg glands,
reaching far back into the body cavity. Their ducts run forward to
emerge at the ends of the oral papillae. When the animals are
irritated, these glands forcibly eject a secretion consisting of a tangle
of viscid threads.

Sexual Organs. — The sexes are separate. The best known sexual
organs are those of P. Edwardsii. Those of other species seem in many
points to be differently formed. (1) Female sexual apparatus (Fig.
295). This is as a rule paired. The two lateral halves, however, are
connected at two points ; first, where the ovaries pass into the uteri,
and second, to form the unpaired terminal division (vagina) leading to
the exterior. The two ovaries are imbedded in a common envelope
of connective tissue, and suspended to the pericardial septum in the
median line by a ligament consisting of two muscles. They lie in the
posterior portion of the body cavity. They are continued into the two
uteri, which, close to the ovary, are united by an unpaired portion ;

VI

PROTRACHEATA

433

separating again and winding about the intestine, they run first for-
wards then outwards, and finally backwards towards the median line,
where they reunite to form the unpaired vagina, whose outer aperture
lies ventrally between the penultimate pair of legs. Each uterus
has in the part nearest the ovaries two appendages, a reeeptaeulum
seminis and reeeptaeulum ovorum. The former is a sac opening into
the uterus by means of two canals which unite at their mouths.

This peculiar method of junction of the reeeptaeulum seminis and
the uterus is explained by ontogeny. The reeeptaeulum is originally

FIG. 295.— Female sexual organs of
an older embryo of Peripatus Edwardsii
(after Gaffron). U, Ovarial ligament; ov,
ovarium ; ro, funnel portion of the reeepta-
eulum ovorum ; rs, reeeptaeulum seminis ;
ut, uterus ; va, vagina.

FIG. 296.— Male genital
apparatus of an adult Peri-
patus Edwardsii spread out
(after Gaffron). t, Testis;
vs, vesicula seminalis ; vd, vas
deferens ; de, ductus ejacu-
latorius.

only a U-shaped loop of the uterus. The two limbs of the loops unite
at a later stage (Fig. 295, rs), fuse together at their ends, and open
into each other, the partition wall disappearing. The limbs of the
U-shaped bend thus become the two connecting canals, and the median
piece (the bend of the U) becomes the sac of the reeeptaeulum seminis.
The reeeptaeulum ovorum, which enters the part of the uterus lying
between the ovary and reeeptaeulum seminis, consists of a funnel,
which enlarges at its free edge into a blindly-closed sac of connective
tissue filled with fertilised eggs.

Peripatus is viviparous. The eggs develop in the uterus, in which all stages of
VOL. I 2 F

434

COMPARATIVE ANATOMY

CHAP.

development are met with, the youngest embryos close to the ovary, the oldest
near the vagina. The younger embryos are fastened in a peculiar manner to the
wall of the uterus, in the older embryos this placenta-like connection ceases, but
they lie in a closed sac formed by the epithelium of the
uterus. The position of the embryo in the uterus is
externally marked by constrictions of that organ. As
the embryos cannot travel along the uterus, the latter
itself grows at the part nearest to the receptaculum
seminis, while its last chamber shortens and degener-
ates ; in this way space is provided for the attachment
of new embryos to its wall. In copulation, which pro-
bably takes place only once, the semen enters the
receptaculum seminis. The eggs from the ovary which
have reached the receptaculum ovorum pass thence into
the uterus.

In other species of Peripatus either the
receptaculum seminis or the receptaculum
ovorum may be wanting, and the eggs, which
from the first have been better provided with
nutritive yolk, do not attach themselves to
the wall of the uterus.

2. Male genital apparatus (Fig. 296). —
This is paired, with the exception of the ter-
minal portion, which opens outwardly at the
FIG. 297.— Part of a uterus same place as the vagina in the female. Each

<* the two tubular te^ is continued into a

cavity, and the embryo r(e) con- short vas effepens, which opens like a funnel
tained in it with its placenta (ep), into a vesieula seminalis. From this again

arises a fine coiled vas deferens, which, united

with its companion, enters a long coiled terminal portion, the tubular
duetus ejaeulatorius. In the proximal part of the latter an envelope
of complicated structure is secreted round the masses of spermatozoa,
and a spermatophore is formed.

Ontogeny. — The development of Peripatus Edwardsii is complicated by the
attachment of the embryos to the uterus wall, the latter undergoing considerable
changes and forming a closed brood chamber (Fig. 297) round each embryo. In
the case of each embryo an umbilical cord and placenta are formed, serving for its
nourishment. Attached by the cord the embryo projects freely into the brood
chamber. The side of the embryo turned away from the navel cord (which is a
process of the dorsal side of its future head) becomes the ventral side. Around the
embryo an envelope yielded by itself and called the amnion is formed, and is attached
to the inner surface of the uterus. As the embryo grows older, it gradually curls
up within the brood cavity.

In Peripatus all those parts of the body which are metamerically or segment-
ally repeated (the mesoderm segments, extremities, nervous system, coxal glands,
etc.) develop and differentiate in the manner universally characteristic of the
segmented animals, i.e. progressively in order from before backwards.

The Mesoderm is differentiated into two ventral symmetrical mesoderm streaks,
which unite posteriorly (at the edge of the blastopore) in a median zone, and in this

VI

PROTRACHEATA

435

zone, throughout embryonic development, active processes of growth go on. Besides
this the cell material of the two mesoderm streaks themselves increases by a continuous
process of division. Segmented cavities appear in the mesoderm streaks in continuous
succession from before backward, and these separate into mesoderm segments or
mesoderm sacs with Avails, which are at first unilaminar. The further differentiation
of these mesoderm sacs occurs in such a way that each falls into three cavities, one
of which becomes the nephridial funnel (Fig. 299, A-C] Avhile the others disappear as
distinct cavities, and the cell material of their Avails yields the mesodermal por-
tions (endothelium, muscles, connective tissue) of the trunk and of the extremities.
The extremities arise as outgrowths of the body Avail. The first pair of rudimentary

A

FIG. 298.— Embryo of Peripatus Edwardsii,
with growths beginning round the jaws. An-
terior end of the body from the ventral side (after
v. Kennel), fc, Jaws ; p, papillae, embracing the
jaws laterally ; op, oral papillae ; no, nephridial
aperture of the segment of the oral papillae.

FIG. 299.— A, B, C, Diagrams to elucidate the
development of the nephridia of Peripatus
Edwardsii (after v. Kennel). Only one side of
the body is represented in the transverse section.
I, II, III, The three divisions into which each
mesoderm sac falls ; II, the division which forms
the rudiment of the funnel. In A the rudiment
of the nephridial canal (nc) has appeared as an
invagination of the ectoderm, in B it has united
with the funnel rudiment (II) ; m, mesoderm ; Ib,
body cavity ; n, longitudinal trunks of the nervous
system ; d, intestine.

extremities, after the antennse, develop into the jaws ; the second into the oral papillae
(Fig. 298). The two segments corresponding with them fuse with the primitive head
segment to form the later secondary head.

The pharynx and oesophagus (stomodseum) and the hind-gut (proctodseum) form
by imaginations of the ectoderm which open later into the endodermal mid-gut.
The buccal cavity arises by the growing up of a rampart round the oral region, and
in this cavity the jaws come to lie.

^^Thile the coxal glands (including the slime glands), which proceed exclusively
from ectodermal invaginations, are clearly dermal glands, the nephridia (and the
salivary glands, genital ducts, and anal glands, which are homologous with- them)
arise out of paired rudiments. The funnel comes, as has already been mentioned,
from one part of a mesoderm sac and only later becomes connected with an ectodermal
invagination which yields the terminal vesicle, and, as it appears, the whole nephridial

436

COMPARATIVE ANATOMY

CHAP.

duct also, however long and coiled it may be. The salivary glands in the young
recently hatched animal still function as nephridia and have funnels opening towards

the body cavity, which afterwards close but are
still retained in the adult animal as vesicular
appendages. Their nephridial ducts grow out far
beyond the funnels to form blindly ending tubes
posteriorly. Their apertures approach the middle
line by the growing up of the oral enclosure and
thus reach the floor of the buccal cavity. A new
median invagination of the integument then yields
their impaired duct. The genital ducts represent
the nephridia of the penultimate limb -bearing
segment and develop in a similar manner (Fig. 300).
At first their outer apertures lie laterally, and
apart. Later they approach the middle line. The
uteri in the female and vasa deferentia in the male,
which correspond with the nephridial canal of the
typical nephridia, are joined by a new median
unpaired invagination from without, and from this
proceed, in the male the ductus ejaculatorius, and
in the female the vagina.

The brain and longitudinal nerve trunks arise
as paired thickenings of the ectoderm, which differ-
entiate from before backward and separate from
the ectoderm. The rudimentary ganglia of the jaw
segment fuse at a later stage with the rudimentary
ganglia of the head segment to form the brain.

FIG. 300.— A, B, C, Diagrammatic
representation of the development
of the female sexual apparatus of
Peripatus Edwardsii in transverse
sections (after v. Kennel), ov, Ovary,
proceeding from the median portion
of the mesoderm sac ; ml, that part of
the female genital apparatus which
corresponds with the nephridial fun-
nel, and from which proceed chiefly
the receptaculum ovorum and the
piece which connects the uteri with
the ovaries ; ee, paired ectodermal in-
vaginations, which become the uterus.
In B the two invaginations have ap-
proached each other in the middle
line, and in C at this point a new
unpaired ectodermal invagination (va)
has appeared, the rudiment of the
vagina ; d, intestine ; n, longitudinal
trunks of the nervous system.

The Systematic Position of Peripatus.

If we try to estimate what has been said about
the organisation and development of this animal
from a comparative point of view, we come to the
conclusion that Peripatus unites typical Annulatan
characteristics with typical Arthropodan and espe-
cially Tracheatan characteristics. The following
are its Annulatan characteristics : (1) segment-
ally arranged nephridia of the type of the per-
manent trunk nephridia of Worms ; (2) segment -
ally arranged coxal glands, which are undoubtedly
homologous with the Chsetopodan setiparous
glands ; (3) a dermo-muscular tube, which most

nearly approaches that of the Hirudinea. The

truncated form of the extremities and the structure of the eye are less significant.
The following are the Arthropodan and especially Tracheatan characteristics of Peri-
patus:— (1) the respiratory organs developed in the form of tracheae ; (2) the dorsal
heart lying in a pericardial sinus and supplied with many pairs of ostia, and
the lacunar circulatory system ; (3) the transformation of extremities into mouth
parts (jaws) ; (4) the specific form of the salivary glands.

The nervous system deviates in a characteristic way from the somewhat similar
nervous system of the Annulata and of the Arthropoda, by the lateral position of the
ventral longitudinal trunks, the slight development of the ganglia, and the large
number of transverse commissures in each segment. The nervous system of Peripatus

vi PROTRACHEATA 437

is a ladder nervous system, which shows striking similarity with that of the Amphi-
neura, Placophora, and Zeugobranchia among the Mollusca, and that of certain Platodes
and Nemertina. There is no doubt, however, that the ladder nervous system of
Peripatus is homologous with the brain and ventral cord of the Annulata and
Arthropoda. Its specifically deviating form may be regarded in two ways : (1) The
ladder nervous system of Peripatus has arisen out of a typical ventral cord by the
moving apart of its symmetrical halves and the increase of the transverse commissures ;
(2) in contrast with the ventral cord of the Annulata it represents a more primitive
condition. The latter view seems to us the more plausible, since we adhere to the
opinion that the ventral cord of the Annulata itself proceeded from a ladder-like
nervous system by the moving together of the longitudinal trunks towards the ventral
middle line. Peripatus, according to this view, would be related only to the typically
segmented racial form of the Annulata. The fact is perhaps not without significance
that the Phyllopoda also (wThich are held to stand nearest of all the living Crustacea
to the racial form) possess a ladder-like nervous system. The large number of trans-
verse commissures in each segment must be a secondary condition. In a few
Annulata we find more than one transverse commissure, also probably as a secondary
condition ; the same is also the case in the Phyllopoda.

The tracheae of Peripatus may perhaps be regarded as dermal glands transformed
by adaptation to life on land, glands similar to those long mostly unicellular dermal
glands which in certain Hirudinea and many Turbellaria spread far through the
body parenchyma.

From the point of view of Comparative Anatomy it is of the greatest importance
to have proved that the salivary glands and genital ducts are transformed nephridia,
helping us as it does to understand the morphological significance of these organs in
the Trachcata. No less important is the almost certain proof that the slime and
coxal glands are homologous, and that these dermal glands are homologous with
the setiparous glands of the Annulata, especially with reference to similar glands
in the Traclieata.

It cannot be certainly proved that the antennse, jaws, and oral papillae of Peripatus
.correspond with the antennse mandibles, and one pair of maxillae of the Tracheata.

Single genus : Peripatus. Animals avoiding light, and living on land in damp
places, under the bark of old trees, under stones, etc. P. capensis : on the wooded
slopes of Table Mountain, Cape of Good Hope. P. Edivardsii : Venezuela ; related
species in Trinidad. P. Novce Zealandice ; P. Leuckartii : Australia.

Literature.

H. N. Moseley. On the Structure and Development of Peripatus capensis, in Philos.

Transactions. Vol. CLXIV. 1874, and in Proc. Roy. Soc. No. 153. Vol.

XXII. 1874.
F. M. Balfour. The Anatomy and Development of Peripatus capensis. Quart. Journ.

Microsc. Science. Vol. XXIII. 1883.
J. v. Kennel. Entwicklungsgeschicte von Peripatus Edivardsii und P. torquatus. I. u.

II. in Arbeiten Zool. Inst., Wiirzburg. Vols. VII. and VIII. 1885 to 1886.
E. Gaffron. Beitrdge zur Anatomie und Histologie von Peripatus, in Schneider's

Zool. Beitrdge. Vol. I. 1883. 1885.
A. Sedgwick. A Monograph of the Development of Peripatus capensis, in Studies

from the Morphological Laboratory in the University of Cambridge. Vol. IV.

London, 1888. Also in Quart. Journ. of Micros. Science.

In these Treatises also further Bibliography.

438 COMPARATIVE ANATOMY CHAP.

CLASS II. Antennata.

Systematic Review.

Sub-Class I. Myriapoda. Millipedes.

Trunk homonomously segmented, segments usually numerous, of nearly equal
size, and, except the last, provided with feet. "Without compound eyes. With
numerous ocelli.

Order 1. Symphyla.

With not more than 12 leg-bearing trunk segments. One pair of branched
tracheae, whose stigmata lie in the head. Unpaired genital aperture in the 4th
segment. Scolopendrella (Fig. 301, p. 444).

Order 2. Chilopoda.

Body more or less flattened dorso-ventrally. Each body ring carries only one
pair of limbs and answers to a segment. The two pairs of maxillae are separate.
The first pair of trunk feet moved on to the head as maxillipedes with poison glands
emerging on the terminal claw. Unpaired genital aperture on the penultimate
segment. Fam. Scutigeridce : with compound eyes. Trunk consists of 15 leg-bearing
segments. Scutigera. Fam. LithoUidce : trunk consists of 15 leg-bearing segments.
No compound eyes, but ocelli. LithoUus (Fig. 323, p. 464), Henicops. Fam.
Scolopendridce : with 21 or 23 leg-bearing trunk segments (the maxillipedal segment
not included). Body elongated. Scolopendra, Cryptops. Fam. Geophilidce : body
very long, with 31-173 leg-bearing trunk segments. GeopMlus, Himantarium.

Order 3. Diplopoda (Chilognatha).

Body mostly arched. From the 5th segment onwards each ring has 2 pairs of
legs and thus corresponds \vith a double segment. The two pairs of maxillae are
fused to form the so-called gnathochilarium. Without maxillipedes. Paired genital
apertures between the 2d and 3d pairs of legs. The legs of the 7th ring in the male
are changed into copulatory organs. Fam. Polyxenidce : 15 pairs of feet. Gnatho-
chilarium rudimentary. Copulatory feet wanting. Polyxenus. Fam. Glomeridce :
11-14 rings. Glomeris. Fam. Polydesmidce : without eyes. 19-20 trunk rings, 29-
31 pairs of feet. Polydesmus, Brachydesmus. Fam. Chordeumidce : 30 trunk rings,
45-50 pairs of feet. Atractosotna, Craspcdosoma, Chordeuma. Fam. Lysiopetalidce :
number of rings large, indefinite. Lysiopetalum. Fam. lulidce : 30-70 or more
rings. lulus. Fam. Polyzonidce : gnathochilarium reduced. Number of the rings
inconstant, 30-100 or more. Polyzonium.

Order 4. Pauropoda.

Feeler with several flagella. Only one pair of weakly developed maxillae. 10
trunk segments. 9 pairs of legs. Tracheae as yet not discovered. Genital apertures
at the base of the second pair of legs. Pauropus.

Sub-Class II. Hexapoda. Insecta.

Trunk heteronomously segmented ; the almost constant number of segments of
unequal size, the body is divided into head, thorax of 3 segments, and hind body of
10 segments. Each of the 3 thoracic segments has a pair of legs. Abdomen limbless.
Compound eyes as well as ocelli almost always present. Apertures of the sexual
organs always at the end of the hind body.

vi ANTENNATA— SYSTEMATIC REVIEW 439

Legion I. Apterygota.

Without wings. With rudiments of abdominal limbs, at least in the Thysanura.
Without metamorphosis.

Order 1. Thysanura.

With 10 abdominal segments and 2-3 jointed bristle- shaped appendages (cerci) on
the anal segment. Compound eyes and ocelli may be present or absent. Machilis,
Lepisma, Nicoletia, Campodea (Fig. 302, p. 444), lapyx.

Order 2. Collembola.

With 6 abdominal segments or fewer. Nearly always a springing fork at the end
of the hind-body. Without compound eyes. Occasionally with ocelli. Sminthurus,
Podura, Isotoma, Macrolomct.

Legion II. Pterygota.

With a pair of wings on both the 2d and 3d thoracic segments. There are
unwinged forms, which, however, are descended from winged ancestors.

Order 1. Dermaptera (Forficulidse), Earwigs.

Insects with gradual metamorphosis, and with biting mouth parts. The last
abdominal segment has unjointed appendages (cerci), which form a pincer. Fore- wings
short, changed into horny wing covers. Hind-wings large, delicate skinned, fan-
shaped, can be folded longitudinally and transversely. Paired genital apertures, of
which one may be rudimentary. Forficula, Labidura.

Order 2. Orthoptera.

Insects with gradual metamorphosis, with biting mouth parts ; with 2 pairs
of membranous or parchment-like wings, sometimes wanting. Fore-wing generally
shorter and more chitinous than the hind-wing. Variously shaped cerci on the hind
body. Sexual apertures unpaired. Embidce, Blattidce (cockroaches : Pcriplancta,
Blatta}. Mantidce (Mantis, praying insect). Phasmidce (stick or spectre insects,
Bacillus, Phasma, Phyllium}. Saltatoria, including the 3 families, Acridiidce
(grasshoppers : Acridium, (Edipoda, Mecostcthus, Stenobothrus, Tcttix, etc. ) ; Locustidce
(Locusta, Thamnotrizon, Platycleis, Decticus) ; and Gryllidce (Gryllus, Gryllotalpa,
Myrm ccophila).

Order 3. Ephemeridse.

Insects with incomplete metamorphosis. Mouth parts somewhat reduced, of
the biting type. Hind -wing small or wanting, fore -wing large, wings finely
membranous. Hind body with 3 (rarely 2) long cerci. Paired genital ducts,
emerging separately. Larvte apneustic, Thysanura-like, with tracheal gills, and with
biting mouth parts, live in water. Ephemera, Palingcnia, Chloe.

Order 4. Odonata (Libellulidse).

Insects with incomplete metamorphosis, and with biting mouth parts. Hind
body with 2 unjointed anal processes. Both pairs of wings are large, and glass-like,
with a rich network of veins. Thoracic legs moved forward. Larvae in water with
various sorts of tracheal gills, apneustic. Libcllula, ^Jschna, Caloptcryx, Agrion, etc.

440 COMPARATIVE ANATOMY .CHAP.

Order 5. Plecoptera (Perlaria).

Insects with incomplete or gradual metamorphosis, with biting mouth parts.
Hind body generally with 2 long cerci. Both pairs of wings large, veins forming
large meshes, the hinder pair often broader than the front pair, and partly foldable.
The Thysanura-like larvre live in water, with tracheal gills, apneustic. The tracheal
gills often persist in the imago. Pcrla (Fig. 317, p. 456), Nemura.

Order 6. Corrodentia.

Insects without, or with gradual metamorphosis, with biting mouth parts.
"Wings often wanting. In the Termites they are finely membranous, and in the
sexual animals deciduous. They are wanting in the workers. Some Psoeidce and the
Mallopliagct are wingless. The compound eyes are wanting in the Mallopliaga. The
Avings of the winged Psoeidce are glassy, areolate, and like those of the Hymen-
optera. Young forms Thysanura-like. Termitidce (white ants, which form communities,
Termcs, Calotermes) ; Psoeidce ( Troctcs, Psocus, book lice ; Mallophaga, parasites feed-
ing on the fur of animals or feathers of birds ; Trichodectes, Philopterus, bird lice ;
Liotheum).

Order 7. Thysanoptera sive Physopoda.

Insects with gradual metamorphosis, the larval form very like the imaginal. The
last larval stage goes without food. Sucking mouth parts. The claws of the short
feet with the adhering lobes of the tarsus changed into a protrusible vesicular
apparatus. Wings very narrow, with reduced veining, with long fringed edges,
often wanting or rudimentary. Only 3 or 4 pairs of stigmata, one or 2 on the thorax,
one on the first and one on the eighth ring of the hind body. Nervous system con-
centrated. Thrips.

Order 8. Rhynchota.

Insects with gradual metamorphosis (in the males of the Coccidce complete meta-
morphosis). Mouth parts form a proboscis adapted for piercing and sucking.
Compound eyes are wanting in the parasitic Rhynchota.

Sub-Order 1. Phytophthires.

With two pairs of membranous wings. Generally wanting in the female. The
Coccidce have only fore-wings, the hind-wings being changed into halteres. Fam.
Psyllidce : with 2 pairs of wings (fore-wings parchment-like) ; Psylla, Lima. Fam.
Aphidce : with 2 pairs of membranous wings, generally wanting in the female ; Aphis,
Chermes, Schizoneura, Phylloxera. Fam. : Coccidce, scale insects ; Coccus, Lecanium,
Aspidiotus.

Sub-Order 2. Pediculidse (Aptera), Lice.
Without facet eyes and without wings. Pediculus, Hcematopinus, Phthirius.

Sub-Order 3. Heteroptera (Hcmiptera), Bugs.

Four wings (seldom wanting). The anterior horny wing-covers are membranous at
their points. Hind-wings membranous. Geocores (land bugs : Hydrometra, Halo-
bates, Pentatoma, Coreus, Corizus, Alydus, Pyrrhocoris, Lygaeus, Miris, Capsus,
Acanihia [bed-bugs], Reduviis, etc.) Hydrocores (water -bugs : Nepa, JRanatra,
Naucoris, Corixa, Notonccta, etc.)

vi ANTENNATA— SYSTEMATIC REVIEW 441

Sub-Order 4. Homoptera.

Fore-and hind-wings similar in shape and membranous, but the fore-wings are
always somewhat harder. Cicada, Fulgora, Pscudophana, Centrotus, Aphrophora,
Tettigonia, Ledra, etc.

Order 9. Neuroptera.

Insects with complete metamorphosis and biting mouth parts. 2 pairs of

membranous glassy wings, closely reticulate. Fam. Megaloptera : Myrmeleon,

Mantispa, Hcmerobius, Chrysopa. Fam. Sialidce : larvse mostly in water, with
trachea! gills. Sialis, Corydalis, Raphidia.

Order 10. Panorpata.

Insects with complete metamorphosis and biting mouth parts. 2 pairs of narrow
membranous wings, widely reticulate. Larvse catterpillar-like. Panorpa, Bittacus,
Boreus (wings rudimentary).

Order 11. Trichoptera (Phryganidae), Caddis-flies.

Insects with complete metamorphosis. Mandibles rudimentary. Maxillse form
a membranous blunt proboscis. Body mostly hairy, less frequently scaly. Hind-
wings generally larger than the fore-wings, folding like a fan. The larvse, which
resemble those of cockchafers, live in tubes or cases chiefly in the water, have
tracheal gills, and are apneustic. Phryganea, Limnophilus, Halesus, Hydropsyche,
Mystacides, etc.

Order 12. Siphonaptera sive Aphaniptera, Fleas.

Insects with complete metamorphosis, with piercing and sucking mouth parts.
No wings. No facet eyes. Parasites. Pulex, Sarcopsylla, Ceratopsyllus.

Order 13. Coleoptera, Beetles.

Insects with complete metamorphosis and biting mouth parts. Fore-wings as
horny wing cases (elytra). Hind-wings membranous, can fold transversely and longi-
tudinally, serve exclusively for flight. Larvse variously shaped, often Thysanura-like,
occasionally like the cockchafer larvse, seldom limbless (Curculionidce), with biting
mouth parts. Several thousand genera with over 80,000 species.

Sub-Order 1. Cryptotetramera.

The tarsi are four -jointed, one joint being rudimentary. Fam. Coccinellidce,
EndomychidcK.

Sub-Order 2. Cryptopentamera.

Tarsi five -jointed, one joint being reduced and hidden. Fam. Chrysomelidce,
Cerambycidce, Curculionidce, Bostrychidce, etc.

Sub-Order 3. Heteromera.

Tarsi of the two anterior pairs of legs five-jointed, those of the posterior pairs
four-jointed. Fam. Meloidce (Cantharidce), Rhipiphoridoe, Tenebrionidce, (Edemeridce,
Cistelidce, etc.

442 COMPARATIVE ANATOMY CHAP.

Sub-Order. 4. Pentamera.

Tarsi as a rule five-jointed in all legs. Fam. Xylophaga, Malacodermata, Elate-
ridce, Buprestidce, Lamcllicornia, Silphidce, Pselapliidce, Staphylinidce, HydropMlidce,
Dytiscidce, Carabidce, Cicindelidce, etc.

Order 14. Lepidoptera.

Insects with complete metamorphosis and sucking mouth parts, forming a pro-
boscis which can generally be curled up. Body covered with scales. Both pairs of
wings similar, membranous, covered with scales, rarely foldable. Hind-wings
generally somewhat smaller than fore -wings. The larvae are caterpillars, with anal
feet, with biting mouth parts, rarely (Micropteryx) footless.

Sub-Order 1. Microlepidoptera.

Fam. Pterophoridce, Tineidce, Pyralidce, Tortricidce.

Sub-Order 2. Geometrina.
Fam. Phytometridce, Deiidrometridce.

Sub-Order 3. Noctuina.
Fam. Ophiusida, Plusiadce, Agrotidce, Cuculliadce, Acronyctidce, etc.

Sub-Order 4. Bombycina.

Fam. Bonibycidce, Saturnidce, Psychidce, Zygcenidce, Cossidce, Liparidce, Eupre-
piadce, Notodontidce.

Sub-Order 5. Sphingina.
Fam. Sesiadce, Sphingidce.

Sub- Order 6. Ehopalocera.
Fam. Hesperidce, Lyccenidce, Satyridce, Nymphalidce, ffeliconiidce, Equitidce.

Order 15. Hymenoptera.

Insects with complete metamorphosis, with mandibles adapted for biting and
maxillae generally adapted for sucking. Usually 4 membranous, transparent, slightly
veined wings. Various sorts of caterpillars — those of the Tenthredinidce and Uro-
ceridce are footless, i.e. maggot-like.

Sub-Order 1. Terebrantia.

Female with ovipositor (borer or tube). Fam. Tenthredinidce, Uroceridce, Cyni-
pidce (gall flies). The larvae of the Pteromalidce, Braconidce, Ichneumonidce, JSvaniadce,
are generally parasitic in the larvae of other insects.

Sub-Order 2. Acnleata.

Female with poison sting and poison glands. Fam. Formicidce (ants), Fossoria
(sand wasps), Vespidce (social wasps), Apidce (bees).

vi ANTENN ATA— OUTER ORGANISATION 443

Order 16. Diptera.

Insects with complete metamorphosis, with sucking and sometimes also piercing
mouth parts. Fore-wings membranous, transparent. Hind-wings transformed into
halteres. Larvse maggot-shaped (without legs), with or without head.

Sub- Order 1. Pupipara.

Viviparous. The larvse are born shortly before entering ^ the pupal state.
Parasites. Wings often rudimentary. Melophayus, Braula, Nyctcribia.

Sub-Order 2. Brachycera, Flies.

Feelers short, generally three -jointed. Many families: Muscidce, Conopidce,
Oestridce, Syrphidce, Empidce, Asilidce, Bombyliidce, Therevidce, Tabanidce, Leptidce,
Xylophagidce, Sir atiomy idee.

Sub-Order 3. Nemocera (Tipularia), Gnats.

Feelers long, many -jointed, in the male often plumose. Fam. Bibionidce,
Fungicolce, Noctuiformes, Culiciformes, CuUtidce, Gallicolce, Limnobiidce.

Of the above enumerated orders that of the Dermaptera is usually placed as a
family of the Orthoptera. The Epliemeridce, Odonata, Plecoptera, Corrodentia, and
Thysanoptera are often united into the order of the Pseudoneuroptera, and the
Panorpata incorporated with the Neuroptera.

I. Outer Organisation.1

A. The Body.

Myriapoda.

The body consists of a head and a large number of uniform trunk
segments, the anterior 3 of which correspond with the 3 thoracic
segments of the Hexapoda (Insecta). The head has almost certainly
arisen from at least 4 fused segments.

Symphyla (Scolopendrella). — The trunk in this division, which
probably stands nearest to the common racial form of the Myriapoda
and Hexapoda, consists of 12 distinct leg-bearing segments, and an
anal segment with 2 processes which may be described as spinning
stylets (Fig. 301). Two feelers, which lie in front of these spinning
processes, are perhaps transformed legs, and indicate the existence of
a 13th pre-anal segment. If so, the whole number of trunk segments
would be 13 or 14, and would then almost exactly correspond with
the original number of segments in the Hexapodan trunk, in which
the thorax has 3 and the abdomen 10 (perhaps 11) segments. The

1 This representation of the outer organisation can only be of the most general char-
acter. For details, which are of great zoological importance, we must refer the reader
to systematic works on Entomology.

444

COMPARATIVE ANATOMY

CHAP.

Pauropoda have the smaller number of 10 trunk segments (including
the anal segment). In the Chilopoda (Fig. 323, p. 464) and Diplopoda
the number of trunk segments is larger and often very considerable (in
Himantarium there are as many as 173). It is possible that this large
number of segments is not an original peculiarity of the Chilopoda and

FIG. 301.— Scolopendrella immaculata
(after Latzel).

FIG. 302.— Campodea staphylinus, without
the setae and hairs (after Lubbock).

Diplopoda, but secondarily acquired, as in the serpents. In the
Diplopoda only the 4 or 5 anterior trunk rings represent single
segments, each subsequent ring is a double segment.

Hexapoda.

The body of the Hexapoda falls typically into 3 parts quite distinct
from each other : head, thorax, and hind body (abdomen). The un-
segmented head probably originally consisted of 4 segments. The
thorax is composed of 3 segments : prothorax, mesothorax, and
metathorax, answering to the 3 anterior trunk segments of the
Myriapoda. The typical number of segments of the hind-body is 1 0 or
11. The thorax and the abdomen together form the trunk, which
may be compared with the trunk of the Symphyla. Among the

vi ANTENNATA— MOUTH PARTS 445

Apterygota the Thysanum possess 10 abdominal segments, and the
Collembola a varying number, but always less than 1 0. In the Pterygota
the number of abdominal segments in the adult animals varies, and is
generally less than 1 0. This diminution is caused by the fusing of those
segments which are connected with the genital apparatus and lie in
front of the last, and secondly by the fusing of the anterior abdominal
segments (usually only the first) with the thorax. On the other hand
in a few insects (Macrolepidoptera, Diptera, and PJiynchota) the last
(3d) thoracic segment is joined with the abdomen.

B. The Limbs.

The limbs of the Insecta consist of single rows of joints. We
distinguish the limbs of the head from those of the trunk. It is
certain that each trunk segment was originally provided with a pair of
limbs (as is now the case in Peripatus and the Myriapoda). In the
Hexapoda, however, only the limbs of the 3 anterior trunk segments,
i.e. of the thorax, have been retained.

1. The Limbs of the Head.

There are, typically, 4 pairs of cephalic appendages, which are
called, in the order from before backward, the Antennae (feelers)
Mandibles, anterior and posterior Maxillae.

Comparing these cephalic limbs with the analogous limbs of the Crustacea, we
see that in the Myriapoda and Insecta one pair of antennse is wanting.

The cephalic limbs themselves are divided into 2 groups, the
feelers, and the oral limbs or mouth parts (mandibles and maxillse).

The feelers (antennae) of the Myriapoda and Hexapoda are always
found in one single pair, and are pre-oral, springing from the forehead ;
they are long and slender, many jointed, very variously formed in
details, and v.ery often different in the two sexes. They are organs of
touch, and at the same time carry the olfactory organs. They are
innervated from the brain.

The oral limbs (mouth parts) vary extraordinarily in form,
according to the special functions to which they are adapted, these
functions being chewing, triturating, biting, sucking, and piercing, etc.
The tracing back of all these variously transformed mouth parts of
the Hexapoda to 3 pairs of oral limbs (mandibles, and anterior and
posterior maxillse) is one of the greatest achievements of comparative
anatomy. We can only take into consideration the principal forms of
these oral limbs. The mouth parts of the Orthoptera form the best
starting-point in our review, because in them the composition of the
lower lip (labium) of 2 lateral pieces (posterior maxillae) is most
evident. The whole apparatus (Fig. 303) is as follows.

446

COMPARATIVE ANATOMY

CHAP.

1. The upper lip (labrum Ibr) is an unpaired piece which covers
the oral aperture in front and from above, and has nothing* to do with
the limbs.

2. The Mandibles (upper jaws, md) consist on each side of a
powerful but unsegmented masticatory plate with toothed edge.

3. The anterior maxillae (lower jaw). Each of these 2 maxillae
consists of a 2-jointed basal portion (mx^, which carries first a
5-jointed feeler (pin, palpus maxillaris), and second, 2 un jointed

771 X.

Fig. 303.— Mouth parts of Blatta (Orthoptera, after Savigny). Ibr, Labrum (upper lip) ; md,
mandible; mx1} anterior pair of maxillae; mx.2, posterior pair of maxillae = lower lip (labium); st,
stipes (stem) ; m, mentum ; sm, submentum ; mi and me, mala interna and externa, inner and outer
ridges of the 1st and 2d pairs of maxillae ; pm, palpus maxillaris, feeler of the anterior maxillae ; pi,
palpus labialis, feeler of the posterior maxillae.

masticatory ridges, one outer (me, mala externa) and one inner (mi,
mala interna).

4. The posterior maxillae together form the lower lip (labium, mx2).
Each posterior maxilla consists of the same parts as the anterior
maxillae (basal part, 3-jointed feeler pi, outer and inner masticatory
ridges me and mi), but the 2 basal parts on each side have grown
together behind and below the mouth in the middle line.

These mouth parts are adapted for biting and chewing.

» We shall now describe the most important modifications of the above type in
systematic order.

VI

ANTENNATA— MOUTH PARTS

447

Myriapoda.

Symphyla. — Mouth parts for chewing. Upper lip, mandibles, and 1 pair of
maxillse with only 1 masticatory ridge and rudimentary feeler. The Pauropoda
have similar mouth parts also weakly

developed. The mouth parts of both •/y%k. *rK / V

groups require further investigation.

Chilopoda (Fig. 304).— The mouth
parts, apart from the upper lip and
the hypopharynx which belongs to
the lower cesophageal wall, consist of
the typical limbs, mandibles, anterior
and posterior maxillse. The anterior
pair of maxillse has well developed
masticatory ridges, but has no feeler
or only a rudimentary one. The
feelers are well developed on the 2d
pair of maxillse, but the masticatory
ridges are wanting. The basal por-
tions of these maxillae are sometimes Fi- 304.-Lithobius validus. The head from

below after removal of the maxillipedes (after Latzel).
separate, sometimes fused. a> Antenn83 ; sk> frontal portion of the cephalic shield ;

Diplopoda. — The mouth parts are Oc, grouped ocelli; pi, feeler of lower lip or of the 2d
here complicated and difficult to ex- pair of maxillse ; stl, stems of the same fused in the
plain. The powerful upper law is middle line ; sto, stems of 1st pair of maxillse ; me, ?m,
,. -11 I, ,, , ,. / .1 outer and inner ridges of the same,

followed by the lower lip (gnatho-

chilarium, Fig. 305). This lower lip is said by some observers to consist of only
1 pair of maxillse. Others explain the pieces represented in the figure in such a way

that the paired halves of the middle
piece, each of which is provided with
a masticatory ridge, correspond with
the stem pieces of the posterior
maxillse (lower lip), and the 2 lateral
pieces each provided with 2 masti-
catory ridges with the stem portions
of the anterior maxillse, the palps
being absent. Although this last
view, which rests upon analogous
modification of the 2 pairs of maxillse
in certain beetle larvse (Elateridce),
is preferable from the point of view
of comparative anatomy, it is not
yet quite certainly established. The
developmental history, as far as it is
as yet known, seems rather to support
the first view, since the mandibles
and the gnathochilarium of the
Fig. 305.-The Gnathochilarium of Lysiopetalum Dipl^poda are said to come from the
carmatum (Diplopoda, after v. Rath), mx,, Stem of *.*

the anterior ; mz2) of the posterior maxillae (?) ; me and rudiments of 2 pairs of feet. A corn-
mi, outer and inner masticatory ridges of the anterior parison of the mouth parts of the
maxillae ; m, masticatory ridge of the posterior maxillse Myriapoda on a new ontogenetic
Cower lip). basig ig urgently needed.

Hexapoda.
Apterygota. — The mouth parts of the Apterygota are adapted for mastication

mx

448

COMPARATIVE ANATOMY

CHAP.

and agree in all essential points with the Orthopteran type above described. The
composition of the lower lip out of 2 maxillse is especially clearly shown in the

Aptera. Both pairs of maxillse possess
well developed palps.

Pterygota. — The mouth parts of the
Orthoptera were described and illustrated
above. As, however, .the mouth parts
of other orders of the Insecta deviate
markedly from these, it is necessary to
describe the more typical forms or
arrangements.

A knowledge of the mouth parts of
a small family of the MicroUpidoptera,
the Micropterygina, throws light on the
mouth parts of the Lepidoptera. We
here still find the typical parts : (1)
toothed mandibles, capable of mastica-
tion ; (2) anterior maxillae, with separate basal portions, with 6-jointed palps and 2
e masticatory ridges ; and (3) a lower lip (posterior maxillfe) whose basal

cm.

FIG. 306.— Mouth parts of a Macrolepidop-
tera larva (Ocneria). Lettering as in Figs. 303
and 309.

FIG. 307.— A, Mouth parts of the Macrolepidoptera. B, The lower lip (2d pair of maxillfe),
isolated. Lettering as before, sr, Sucking proboscis, corresponding with the fused ridges of the
1 st pair of maxillae.

portions are fused into one common piece, but carry 3-jointed palps and masticatory
ridges still distinctly separate. The 2 inner ridges have grown together and form
a short tube. In the other Microlepidoptera the mandibles lose their teeth and
become rudimentary. On the anterior maxillae only 1 ridge is found. The ridges

VI

ANTENNATA— MOUTH PARTS

449

of the 2 pairs of maxillae fit together to form a sucking proboscis which can easily
be coiled up. In the Macrolepidoptera the mandibles have disappeared, but the
sucking proboscis formed by the 2
ridges of the anterior maxillae is on
the contrary very strongly developed c
and capable of being coiled. The
maxillar and labial palps are nearly
always retained, the former generally
in a very reduced condition (1 -jointed
in the Sphingina and many Rhopa-
locera). In some of the latter the
maxillar palp has, however, al-
together disappeared.

A series analogous to that of the
Lepidoptera is afforded by the Hy-
menoptera. At the head of the series
stands the Tenthredinidce, whose mouth parts show great agreement with those of
the Micropterygina. Besides the mandibles which, as in the other Hymenoptera,

fini

FIG. 308.— Mouth parts of a Tenthredo larva.
Lettering as before.

FIG. 309.— -4, Mouth parts of the Hymenoptera (Apis melifica). E, The two pairs of maxillae.
au, Facet eye; a, antenna; Ibr, upper lip; md, mandible; ep, epipharynx ; mx^, anterior maxillse ;
pm, palp of the same ; mm, the fused ridges of the same ; prg, paraglossa = outer ridge of the
posterior maxillee (labium or lower lip) ; li, tongue (glossa)= inner ridge of the posterior maxillae ; c,
cardo ; sm, submentum ; m, mentum ; stm, stem (stipes) of the anterior maxillae.

are adapted for biting, we find anterior maxillae, on whose basal portions 6 -jointed

palps and 2 maxillar ridges are quite distinct. On the posterior maxillae (lower lip)

VOL. I 2 G

450

COMPARATIVE ANATOMY

CHAP.

the basal portions are fused, the 2 4 -jointed palps are well retained, the outer
masticatory ridges are separate, but the 2 inner ridges fuse to form a tube. In the
other Hymenoptera (Fig. 309) the mandibles are always retained in a condition
capable of masticating or biting ; both the maxillae go to form sucking or licking
mouth parts. The palps on the anterior maxillae become reduced, the basal
portions elongate and the masticatory ridges grow together on each side into a
long piece (mm). On the under lip also the basal portion elongates, the feeler
remains well developed, and slender, 2-4 jointed ; the inner ridges together form the
long tongue, and the outer ridges small lateral appendages to it, called the accessory
tongues (paraglossa).

The mouth parts of the Diptera (Fig. 310) are adapted for piercing and sucking,
and together form a peculiar proboscis. The bristle-shaped mandibles in the male,

A 3

FIG. 310.— Mouth parts of the Diptera. A, of Tabanus. B, of Culex. Lettering as before.
a, Antenna , an, facet eye ; oc, simple eye (ocellus).

and occasionally in both sexes, are wanting as separate pieces, and are then no doubt
fused with the upper lip. The proboscis is principally formed out of the following
parts much elongated : first, the upper lip ; second, the basal portion of the lower lip,
2 lips (labella) at the end of this representing the transformed palps ; and third,
a prolongation of the lower oesophageal wall (hypopharynx), developed into a pierc-
ing setae, at whose point the salivary vessels emerge. The anterior maxillae form 2
slender setae, which lie together with the seta-like mandibles in the sucking proboscis.
Their 1-5-jointed palps are mostly well developed.

The mouth parts of the Rhynchota (Fig. 311) together form a proboscis adapted
for piercing and sucking. The elongated, generally 4 -jointed lower lip (posterior
maxillae) forms a channel in which lie the mandibles and interior maxillae, trans-
formed into setae covered at their basal part by the- upper lip (labrum).

ANTENNATA— MOUTH PARTS

451

The mechanism for sucking and stinging, which is occasionally very complicated,
and to which certain adaptations in the oesophagus (pumps, " fish trap " apparatus,
etc. ), belong, cannot be here more exactly described.

The mouth parts of the other Hexapoda must be referred to one or other of the
types depicted.

The mouth parts of the Coleoptera are for biting and masticating, similar to those
of the Orthoptera ; the masticatory ridges of the anterior maxillae are rarely trans-
formed into a sucking tube.

The mouth parts of the Dermaptera, Epheineridce, Odonata, Plecoptera Corrodentia,

FIG. 311.— Mouth parts of the Hemiptera. A, of Pentatoma. B, of Pyrrhocoris.

as before.

Lettering

Neuroptera, and Panorpata are also adapted for biting, and belong with various
deviations to the type of those of the Orthoptera and Coleoptera.

The mouth parts of the Thysanoptera (Physopoda) hold a position intermediate
between the biting mouth parts of the Orthoptera and the sucking mouth parts of
the Rhynchota. The mandibles are changed into piercing setae, and come to lie
within a short tubular proboscis, which arises by the growing together of the upper
lip with the anterior maxillae and the lower lip (posterior maxillae). The 2 pairs
of maxillae have distinct palps and otherwise generally show the typical parts
variously modified. In the Trichoptera the mandibles are rudimentary, the 2
pairs of maxillae together form a sort of proboscis (for piercing and sucking), the 4
palps remaining separate on it. The palps, however, as well as the proboscis itself,
may disappear.

In the Aphaniptera (Fleas) the mouth parts are for piercing and sucking. The
mandibles are toothed ridges which, together with the upper lip form the sucking

452 COMPARATIVE ANATOMY CHAP.

tube. The anterior maxillae are short and ear-shaped with 4-jointed palps. The 2
many -jointed palps of the small lower lip lie on the sucking tube laterally.

The various piercing and sucking mouth parts found among Insects have no
doubt developed independently of one another from masticatory mouth parts.

The special morphology of the mouth parts is therefore necessary for a knowledge
of the relationships of the members of one and the same order, but not for a know-
ledge of the phylogeny of the Insect-orders themselves.

It sometimes occurs that the larvae of certain Insects (Megaloptera among the
Neuroptera) have sucking mouth parts while the adults possess biting mouth parts.
This is an interesting fact, which shows how within a small group the larvae may
develop sucking mouth parts in adaptation to special conditions of existence. In
those Lepidoptcra, Diptera, Aphaniptera, and certain Hymenoptera which are provided
with sucking mouth parts, those of the larvae, when not degenerated, are of the biting
type.

2. The Limbs of the Trunk.

In the ancestors of the Antennata (Myriapoda and Hexapoda) in
every case each trunk segment was certainly provided with a pair of
limbs, as is still the case in Peripatus and in the Myriapoda.

In the Hempoda only the 3 pairs of limbs of the 3 anterior trunk
segments have been retained, these 3 segments together forming
the thorax. Rudiments of extremities, however, are not wanting, as
we shall presently see, on the segments of the hind-body even in the
Hexapoda.

The trunk limbs are throughout distinctly jointed and consist of

several parts, whose number
and constitution is extremely
important in classification.
The legs may be variously
formed according to their
special functions. We thus
distinguish ambulatory,
springing, swimming, seizing
legs, etc. Among the Myria-
poda the 1st pair of trunk
feet in the Chilopoda moves to
the head as a pair of maxilli-
pedes (Fig. 323, p. 464).'
These are very strong, and
shaped like pincers. Their

basal segments are fused to-
FIG. 3l2.-Anterior end of the body of a female ther into ft kte j ^
Polydesmus complanatus, from the ventral side (after .-,-,1v A . i

Latzel). «, Antenna ; st, stems of the mandibles ; v, middle line. A poison gland
vulvse (apertures of the female sexual organs) ; «, lying in the maxillipede itself

emerges at its terminal claw.

In the Diplopoda (Fig. 312) the double segments (i.e. the rings
following the 4th or 5th trunk rings) have each 2 pairs of legs, while

vi ANTENNATA— TRUNK LIMBS 453

the 4 or 5 anterior rings are only provided with 1 pair each.
One of the 4 or 5 anterior rings — in the lulidce it seems to be the 4th
— may be limbless. The extremities of the 7th ring are usually trans-
formed in the male into copulatory organs.

/
Rudiments of abdominal limbs in the Hexapoda.

In order to prove the existence of such rudiments we must recall the coxal glands
emerging on the legs of Peripatus. Similar glands emerge in the GMlopoda on the
coxae of the 4 or 5 last pairs of legs, on the pleura of the last leg-bearing segment, and
on the anal segment. In the Diplopoda these glands apparently correspond with
the protrusible warts which occur in the Lysiopetalidce on the coxal joints of the 3d-
16th pairs of legs, and also with the pores on the coxae of the Chordeumidce. A
knowledge of these organs in the Symphyla (Scolopendrella), which perhaps of all
living Antennata stands nearest the common racial form of the Myriapoda and the

FIG. 313.— Posterior end of body of Scolopendrella immaculata, from the ventral side (after
Latzel). pu llth, piz 12th undeveloped, pairs of legs, p13, transformed legs (13th pair) carrying
organs of touch (so) ; sg, spinning processes with the duct (dg) of the spinning gland ; cd, coxal
gland : hs, coxal spur of the llth pair of legs.

Insecta, is extremely important. On the coxal joints of the legs in Scolopendrella
protrusible saccules, apparently glandular (Fig. 313, cd), can be distinctly made out,
especially on the 3d-llth pairs. Laterally from these saccules, which must be homo-
logous with the coxal glands of other Myriapoda and of Peripatus, there is a stylet-
shaped appendage hs, which must be considered as a modified process of the coxal
joint (coxal spur). In addition to these coxal saccules, Scolopendrella possesses 2 spin-
ning glands, which emerge externally (dg) at the point of the spinning processes (sg),
on the terminal segment of the body. These glands also probably belong to the
category of coxal glands, and thus the spinning processes probably represent the last
pair of limbs considerably transformed. The coxal saccules of Scolopendrella and
the coxal glands of the Myriapoda and Protracheata (Peripatus) now throw much light
on similar arrangements in the lowest Hexapoda, the Apterygota. In Campodea there
are in the first abdominal segment two indistinctly jointed appendages which are
rudimentary'extremities. In the subsequent abdominal segments as far as to the 8th
there occurs on each side ventrally a protrusible saccule on whose outer side lies a mov-
able pointed process. These saccules evidently correspond with the coxal glands of
Scolopendrella, and are to be considered as degenerated coxal glands, while the

454

COMPARATIVE ANATOMY

CHAP.

pointed process answers to the coxal spur of Scolopendrella. The coxal saccules
and spurs of Campodea must therefore be regarded as remains of coxal joints of
abdominal limbs, in short, as rudiments of coxse. Similar organs are also found in
other Aptera (Figs. 314 and 315), principally in the Thysanura. On the other hand,
the coxal rudiments may be wanting or be limited sometimes to the saccules and
sometimes to the spurs. It is interesting also to relate that the above-mentioned

L— V-

FIG. 315.— A ventral shield of Machilis maritima,
with two protrusible saccules (cb) on each side. On
the left the saccules are withdrawn, on the right pro-
truded, hs, movable appendages (coxal spurs), muscles
of the same and of the protrusible saccules (after
Oudemans).

ft-

FIG. 314.— Ventral side of the hind-body of a female Machilis maritima (after Oudemans).
The left half of the 8th ventral shield is removed. I-IX, segments of the abdomen ; c, bristle-like
jointed appendages (cerci) of the 10th abdominal segment ; cb, protrusible saccules = coxal glands in
the act of degenerating ; hs, movable appendages=coxal spurs, conjectural rudiments of abdominal
feet ; Ir, ovipositors.

processes on the abdominal segments of the Thysanura were by many observers at
once assumed to be degenerated abdominal feet.

In the winged Hexapoda (Pterygota) rudiments of abdominal feet have also been
observed. They appear at certain embryonic stages exactly like the rudiments of
the thoracic feet, i.e. as prominences or stumps on the most anterior, or on several
anterior, or on all the abdominal segments, sooner or later again to disappear. They have
been observed in Coleoptera (Hydrophilus, Fig. 316, A and JB, Melolontha), Orthoptera
(Gryllotalpa, Mantis, Periplaneta, (Ecanthus, JBlatta), and Trichoptera (Neophalax
concinnus}. In a few forms (Gryllotalpa, (Ecanthus, Periplaneta, Blatta, Melolontha)
the rudiments of the 1st pair of abdominal feet, before the hatching of the embryo,

VI

ANTENNA TA— WINGS

455

become short stalked vesicles of considerable size, which may be compared with the
protruded coxal sacs of the Thysanura. A respiratory function has without sufficient
foundation been ascribed to both these structures.

Considering the widespread occurrence of rudimentary abdominal feet in the
Embryos of winged Insects we are justified in asking the question, whether the

B

— (L

-P9

FIG. 316.— A and B Hydrophilus embryos with the rudiments of extremities (after Heider).
In the somewhat older embryo, B, the rudiments of abdominal feet, which disappear later, can be
very distinctly seen ; a, anal aperture ; an., antenna; g, rudiment of the ventral ganglionic chain; m,
oral aperture, md, Mandible ; mxi, 1st maxillse ; mx%, 2d maxillae (rudiment of the lower lip) ; pi,
f>ii P& thoracic pairs of legs ; p±, p^, p?, p9, rudiments of extremities of the 1st, 2d, 4th, and 6th
abdominal segments ; st, stigmata ; vie, procephalon.

truncated feet (anal feet) of the larvse of butterflies and wasps are not rather the
remains of real limbs than new formations.

C. The Wings.

Wings are altogether wanting in the Myriapoda. Among the
Hexapoda the Apterygota, as their name implies, are entirely wingless.
Since neither the adult Apterygota nor the embryos at any stage of
their development have wings or organs belonging to wings, we are
justified in assuming that their ancestors also were wingless, in short
that the wingless state is as much the original condition here as in the
Myriapoda and Protmcheata. This assumption is not without support
from other points in their constitution. All other Hexapoda, however,
are typically provided with wings, and originally indeed with 2 pairs,
and although within the different orders of the Pterygota the wings

456 COMPARATIVE ANATOMY CHAP.

may be reduced to 1 pair, or may be entirely wanting (in both sexes
or only in the female), we here have to do with a derived con-
dition and with animals which have lost the wings once possessed by
their ancestors. In such insects the rudiments of wings or of organs
belonging to wings can often still be pointed out.

The wings are thin lamellate un jointed folds of the body wall, speci-
ally of the integument. The 2 lamellae of a wing fold lie close to each
other. The wings are veined like the leaf of a plant. The veins
for the most part are thickenings of the chitinous cuticle. Within
the narrow interior space of the wing, nerves and especially tracheae
enter, branching like the veins. Blood-vessels also accompany the
veins. The arrangement of the veins is very important for classifica-
tion. The exact investigation of the courses of the veins and their

FIG. 317.— A, Larva. B, female imago of Capnia nigra (Perlid) (after Pictet).

development, and especially the observation of rudimentary veins or
veins in the act of disappearing, have led to the result that the wings of
the various Hexapodan orders must be traced back not from one to the
other, but to a common form of wing. Thus the examination of wings
confirms the assumption that all orders of winged insects are derived
from a common winged racial group.

The 2 pairs of wings are appendages of the meso- and meta-thorax
of the Insects. There are never more than 2 pairs. Their narrowed
basal portions are articulated with the dorso-lateral parts of the meso-
and meta-thorax. Strongly developed wing muscles serve to move
them (see section on musculature).

The problem of the phylogenetic origin of the wings of insects is extremely
difficult, and as yet by no means solved. The rise of such organs is not explained
by saying that they are integurnental folds, which gradually increased in size,
stood out from and eventually articulated with the body. The wings must in all
stages of their phylogenetic development have performed definite functions. It is
impossible that they were originally organs of flight. "What function it was they

VI

ANTENNA TA— WINGS

457

performed before they became exclusively organs of flight is, however, entirely a
matter of conjecture. The following view is at present the most acceptable. (1)
The ancestors of the Hexapoda were, like the now living Apterygota, wingless land
animals breathing through tracheae. (2) The Aptery 'goto, -like ancestors of the
Ptcrygotan racial group became adapted to living in water. Dorsal integumental
folds served for breathing in the water. The rise of such respiratory folds offers no
difficulty, since every increase of surface, small or large, is of service. (3) The
respiratory appendages (into which tracheae were continued) became movable and
may perhaps have assisted in locomotion (swimming). This assumption also offers
no difficulty, since the gills of many aquatic animals are movable, and their power
of moving is an advantage on account of the exchange of water thus caused. (4)
In a new gradual change to land life the respiratory function became less important
and the locomotory function came to the front. Here, however, lies the greatest
difficulty. It may, however, be assumed, that the animals while still living in water
were capable of gliding over the surface of
the water by the swinging of their branchial
leaves, just as flying fish do by means of
their thoracic fins.

The limitation of the wings to the 2
pairs />f the meso- and metathorax must be
explained mechanically, as more suited
for the propulsion of the body in flight.
We still see among living insects an un-
doubted tendency to the stronger develop-
ment of one of the pairs of wings.

The so called tracheal gills of the larvse
of the Phryganidce, Sialidce, and Ephemeridce
may serve as an example for this conjectural
formation of integumental folds serving for
breathing in Avater. The Phryganid larvae
live in the water in tubes of their own con-
struction, and possess on their soft-skinned
abdomens thread-like appendages into which
tracheal branches enter. Such appendages
are called tracheal gills. Similar append-
ages are found on the abdomens of the Sialid
larvae. In the Ephemerid larvse, which live
free in water, there are found, on the seg-
ments of the hind-body, 6 or 7 pairs of lateral, movable, tracheal gills (Figs. 318,
342, 343), which are sometimes tufted, sometimes leaf-shaped, sometimes thread-
like. An anterior pair may even be developed as a sort of branchial cover for the
posterior pairs. All these tracheal gills are evidently integumental folds and re-
spiratory organs which have arisen as adaptations to aquatic life. When they are
leaf-shaped, the tracheae which enter them branch more or less richly. They begin
to form in a manner altogether similar to ordinary wings, and persist in the later
larval stages together with the wing rudiments (Fig. 318).

Unsuccessful attempts have been made to trace back the wings of Insects to other
organs in other more or less remote animals, e.g. to the dorsal gills of the Chcetopoda
or to the dorsal folds of the Crustacea. If there is any such connection the
rudiments of the wings as primitive organs ought to appear in Insect embryos ; this,
however, is not the case.

Some information has already been given in the Systematic Review about the
special form and arrangement of the wings in the various orders of the Insecta.

FIG. 318. — Thorax and anterior ab-
dominal Segments of the Larva of Cloeon
dimidiatum (Ephemerid), with tracheal gills
(tki, tkz, tks) and the rudiments of the fore-
wings (VF) and hind-wing (HF). tl, Tracheal
longitudinal trunks (after v. Graber).

458 COMPARATIVE ANATOMY CHAP.

II. The Integument and Glands.

The integument is of the same type as that of the Crustacea and
all other Arthropoda, the body being covered by a chitinous cuticle
forming an exoskeleton. This cuticle is of varying flexibility and
thickness, and shows many and great modifications, not only in the
various parts of the body and limbs of the same animal, but among the
different members of the class. A knowledge of the constitution of
the exoskeleton in the various parts of the body is of great importance
in classification, as is also a knowledge of the setae, hairs, scales, etc.,
which belong to the category of cuticular formations. The epithelium
which secretes the chitinous cuticle is here also called the hypodermis.

During eedysis, which accompanies the metamorphoses of the
Antennata and the growth of the larvse, the whole exoskeleton is
thrown off, together with the chitinous intima of that part of the
intestine thus lined, the chitinous intima of the tracheae, and the ducts
of the glands. The chitinous integuments thrown off are known as
exuvia.

Dermal glands are widely spread in the Antennata ; they appear
in a great number of modified forms, emerge at the most various points
of the body, and form secretions differing greatly in constitution. A
comparative study of these on a wide basis is urgently needed.
Investigation is especially needed as to which glands in the Antennata
correspond with the coxal and spinning glands of Peripatus, and which
glands, if any, are to be considered as transformed nephridia. At
present the observations in comparative anatomy and ontogeny neces-
sary to enable us to give a definite answer to these questions are
wanting.

Among the glands emerging on the outer integument we may first mention the
salivary glands, which open in the immediate neighbourhood of the mouth. They
are everywhere found in the Antennata, generally lying at the side of the fore-gut,
in the head or thorax, and occur in 1-3 pairs. They are either simple or much lobed
acinose glands, or else fall on each side into 2 or more glandular sacs. The ducts of
the glands of each side, however, nearly always unite in a common duct, which
finally unites with the duct from the other side to form an unpaired canal, which
usually emerges externally on the lower lip or the hypopharynx, but in any case in
the immediate neighbourhood of the mouth. Not infrequently a vesicular appendage
(saliva reservoir) is found on each side of the canal. It sometimes happens that
there are two separate apertures for a single pair of salivary glands, the unpaired
terminal portion being absent. Where there are several pairs of glands, their ducts
may also emerge separately ; usually, however, the ducts from each pair unite to form
a common terminal portion.

The salivary glands are, as far as their development is known, invaginations of
the oral edge of the stomodaeum. According to some observations, it appears that
the unpaired duct forms secondarily, the 2 salivary glands proceeding from paired
rudiments.

Spinning glands (sericteries) occur in many Insect larvse, and are specially strongly
developed in those which pass through a pupa stage (e.g. caterpillars of Lepidoptera

vi ANTENNATA— INTEGUMENT AND GLANDS 459

and larvse of Tenthredinidce). The thread-like secretion of these glands, which
hardens when exposed to the air, forms the web of which the pupal envelope consists,
but it may also serve for other purposes. The sericteries are glandular tubes which
are paired, elongated, and coiled, often running through the whole length of the body ;
the glandular cells of their epithelium often attain to an enormous size, and have cell
nuclei in the form of branched networks. The 2 ducts unite, like those of the
salivary glands, to form an unpaired terminal duct, whose aperture also lies near the
mouth. Accessory glands may also open into the ducts of the sericteries. The ducts
of these and other dermal glands have a chitinous intima, which, like the tracheal
intima, may become thickened in a close spiral line.

The spinning glands of Scolopendrella have already been mentioned when describ-
ing the rudimentary abdominal limbs. The glands there mentioned as emerging
through pores at the points of stylets on the anal segment, and through the pleural
pores of the last leg-bearing segment, are also said to be spinning glands.

Recalling the Protracheata, we are led to suppose that the salivary glands of the
Antennata are transformed nephridia, and that the spinning glands belong to the
same category as the coxal glands of Peripatus and the parapodial setiparous glands
of the Chaetopoda. Compare also on this subject the section on the rudiments of
abdominal limbs in the Hexapoda.

The morphological worth of the other numerous dermal glands which have been
observed in the Antennata cannot at present be rightly estimated. We can only
name a few of them.

The Myriapoda (Diplopoda) have stink glands for protection, which emerge
through the " foramina repugnatoria " on the dorsal side of a varying number of trunk
segments. These foramina are either paired, in which case they lie laterally, or
unpaired in the middle line. In Paradesmus gracilis the secretion of the protective
glands contains prussic acid. These protective glands, of which only 1 pair occurs
in a double segment, have been regarded as modified nephridia.

The glands of the Geophilidce among the Chilopoda, which emerge through
unpaired median ventral pores, may perhaps belong to the category of protective
glands.

Stink glands, which yield a strongly smelling secretion evidently serving as a
protection to the animal, are also found in many Insects (especially Neteroptera,
Coleoptera, and Orthoptera). They are sometimes paired, sometimes unpaired, and
emerge at different points of the body. In many Coleoptera these organs are appen-
dages of the rectum. An enumeration of the recorded observations, however, could
at present yield nothing of special interest to the comparative anatomist.

Poison glands. — The maxillipedes of the Chilopoda contain poison glands whose
outer aperture lies in the terminal claws. In the female of many Hymenoptera a
poison gland occurs, which pours its secretion into a stinging apparatus of complicated
structure placed at the posterior end of the body. The poison gland itself consists
of 2 simple or branched glandular tubes, which enter a poison vesicle or reservoir by
means of a common unpaired terminal piece (cf. Fig. 347, A, p. 487).

Leg glands are found in many Insects on the terminal joints of the thoracic legs.

Wax glands occur in many Rhynchota (Aphides, Coccidce). They lie either on
the back in cross rows, or near the anus, and secrete filaments, plates, etc., of wax ;
these are used either for forming a dorsal shield or a down which covers the body, or
for enveloping the excrement.

Rectal glands are papillae or thickenings with glandular epithelium, very
commonly found in the rectum of Insects.

In Mantis glands enter the coxee of the 1st pair of legs (coxal glands ?).

460 COMPARATIVE ANATOMY CHAP.

III. The Musculature.

The arrangement of the muscles in the body and their relation to
the exoskeleton is the same in the Antennata as in the Crustacea (cf.
Crustacea, p. 331). The musculature seems to be broken up into a
very great number of single muscles, which are arranged in a definite
manner suitable for moving the segments, the regions of the body, the
limbs and their separate joints, the mouth parts, the ovipositors, stings,
etc. The greater part of the muscles of the body can be traced back
to a paired system of dorsal and ventral intersegmental longitudinal
muscles. While in the Myriapoda, in accordance with the homonomous
segmentation of the body, the musculature is repeated in all the trunk
segments, in the Hexapoda the musculature is very differently developed
in the head, thorax, and abdomen. The musculature of the thorax is
very strong, as might be expected from the fact that its 3 segments
carry the limbs and wings. The wing1 muscles generally take a dorso-
ventral course in the lateral portions of the thorax. The most import-
ant parts among them are played by the elevators and depressors.

The musculature is transversely striated.

IV. The Enteric Canal.

The mouth lies in the head between the mouth parts ; the anus
always in the terminal segment of the abdomen. The enteric canal,
in most Myriapoda and in the Apterygota, runs straight through the
body and thus is not longer than the body. In the Pterygota, on the
contrary, it generally forms more or less marked loops which are
wanting or not so strongly developed in the larva. It everywhere
falls into the 3 already known divisions : the fore-gut, which comes
from the ectodermal stomodseum; the endodermal mid -p gut and
the hind -gut, coming from the ectodermal proctodseum. These 3
divisions are generally distinct. Each of them can be further sub-
divided, especially in the Hexapoda, where special organs in the form
of diverticula are always to be found. Tubular and pouch -like
diverticula of the hind-gut (wanting only in a few Apterygota) are
especially characteristic of the Antennata. They appear in varying
numbers, function as excretory organs, and have received the name of
the Malpighian vessels. The salivary glands and the spinning glands
of the larvae, both of which emerge at or near the mouth, have already
been described.

Myriapoda. — The enteric canal runs straight through the body ; only in the
Glomeridce is it coiled in its posterior part. The mid-gut has numerous short hepatic
tubes. At the beginning of the hind-gut 1 or 2 pairs of long Malpighian vessels
enter ; these run along the gut, frequently winding round it.

Hexapoda. — Each of the 3 principal divisions of the enteric canal may present
various modifications, except in the Apterygota and the larvae of those Inlecta whose

VI

ANTENNATA—THE ENTERIC CANAL

461

straight enteric canals present no complications. The canal is most specialised in
carnivorous Insecta, while in those Insecta that feed on plants it is generally uniform,
but much coiled.

The fore-gut often has 3 divisions : (1) a pharynx or oral cavity, (2) a narrow
oesophagus passing through the cesophageal ring,
and (3) a variously shaped fore-stomach widened
out like a sac. The latter may be wanting as a
separate division. If the fore-stomach is provided
with a strong muscular wall it is called. a crop
(ingluvies). In the honey bee it is called the honey
stomach. In Insects with sucking mouth parts,
and especially in the Lepidoptera (Fig. 348, p. 488)
and Diptera, it is constricted off in the form of a
stalked sac, which opens into the posterior part
of the fore -gut and is unsuitably called sucking
stomach ; it is more correctly a receptacle for food.
Between the crop and the mid-gut in many carni-
vorous Insecta (many Coleoptera, Neuroptera, and
Orthoptera} a muscular masticatory stomach is
interposed ; the chitinous intima of this stomach is
much thickened, and in the form of spikes, spines,
ridges, teeth, etc. projects into' the lumen ; these
processes in transverse section form most beautiful
and ornamental patterns. A peculiar pumping
apparatus is in a few Rhynchota connected with
the pharynx. The fore-gut is internally provided
with a chitinous intima, the continuation of the
chitinous exoskeleton.

The mid-gut, which lies in the abdomen, is the
most important division of the enteric canal for
the assimilation of food ; its epithelium consists of
glandular cells and often projects into the lumen in
the form of folds or villi. It is -a somewhat wide
tube frequently forming loops, and in it we can
often distinguish an anterior wider portion, the phagus ;
chyle stomach, and a longer thinner posterior masticatory stomach ; cd, chyle stom-
J . , . ach covered with villi ; twijMalpighian

portion (small intestine). The chyle stomach in yessels. ^ hind.gut with rectum

the carnivorous Coleoptera is beset with short diver- (r) • ad, anal glands with muscular
ticula, as if with villi ; in the Orthoptera longer vesicular appendages ab.
diverticula enter its anterior portion.

The hind-gut is lined with a delicate chitinous intima and has a muscular wall
which, at the terminal portion ending in the anus, is of considerable thickness. Its
length varies, it is often very long and coiled.

The limit between the mid- and hind-guts is often difficult to define, since the
mid-gut also may have an intima, and its ontogenetic development is not sufficiently
worked out. It is assumed, somewhat arbitrarily, that the hind-gut begins at the
point where the Malpighian vessels enter. Although these are undoubtedly forma-
tions of the hind-gut, they do not necessarily always appear at its anterior end.
The hind-gut is often further subdivided. Its last division sometimes carries an
unpaired caecum. Paired anal glands (stink glands) may also enter it. The anal
papillce, etc. classed as glands have already been mentioned. The intestine of certain
Rhynchota, Psyllidce, and Cicadce is peculiarly constituted. The mid-gut and part

FIG. 319.— Digestive apparatus
of Carabus auratus (after Dufour).
k, Head with mouth parts ; oe, ceso-
in, crop (ingluvies) ; pv,

462

COMPARATIVE ANATOMY

CHAP.

of the hind-gut form a loop (Fig. 322). The 2 limbs of the loop grow together
for a certain distance, and wind round each other at this part.

In the larvse of some Hymenoptera, Neuroptera, Myrmeleon, and Diptera
(Pupipara) the mid-gut ends blindly and is not yet connected with the hind-gut, the
latter performing exclusively excretory functions (Fig. 321).

The Malpighian Vessels. — These are long, generally filamentous appendages,
which begin to form as invaginations of the proctodseum. Their large epithelial

FIG. 320.— Nervous, tracheal, and digestive systems of the Honey bee (after Leuckart).
The fine branchings of the tracheal system are not represented, the tracheal system on the right
side of the figure is only partly drawn, aw, Facet eye ; a, antenna ; 6j, 62, 63, the 3 pairs of legs ; tb,
part of the tracheal longitudinal trunks swollen into a large vesicle ; st, stigmata ; Tim, honey
stomach ; cm, chyle stomach • vm, Malpighian vessels ; rd, rectal glands ; ed, hind-gut.

cells (with nuclei often branched) contain coloured concretions in which uric acid
is found. The Malpighian vessels occasionally have no distinct lumen ; they then
consist of a few rows of cells. The number of the Malpighian vessels is as varied
as their manner of entering the hind-gut.

Apterygota. — Malpighian vessels are wanting in lapyx and the Collembola. In
Gampodea there are ca. 16, and they are here short ; in the other Thysanura they
are long and 4-8 in number. The tubules always unite in pairs before entering the
hind-gut.

Pterygota. — The Malpighian vessels are either very numerous and relatively short
or less numerous' (2-8) and long. They are more numerous in the Dermaptera (ca.

VI

ANTENNATA—THE ENTERIC CANAL

463

30), Ephemeridae (ca. 40), Odonata (50-60), Plecoptera (40-50), Orthoptera (30-50 or
more), and Hymenoptera (very numerous, often over 100, seldom below 12). On the
other hand few (i.e. 2-8) are found in the Corrodentia (4-6), Thysanoptera (4),
Rhynchota (2-4), Neuroptera (4-6), Panorpata (6), Trichoptera (6), Lepidoptera (6,
seldom 2 or 4), Diptera (4 or 5), Siphonaptera (4), and Coleoptera (4-6). They
generally enter the hind-gut separately, but occasionally the vessels of each side unite
into a common duct, and sometimes the ducts from the two sides also have a common
unpaired terminal piece. Here and there the vessels open into a paired or unpaired
urinary bladder attached to the hind-gut. In the Aphides there are on each side

oe

-el

FIG. 321.— Larva (maggot) of honey bee, anatomy
of the digestive and nervous systems (after R.
Leuckart). g, Brain ; 6m, ventral chord ; oe, oeso-
phagus ; sd, spinning glands ; cd, mid-gut, or chyle
stomach ; ed, hind-gut, not yet connected with the mid-
gut ; vm, Malpighian vessels ; an, anus ; st, stigmata.

FIG. 322. — Enteric canal of
Psyllopsis fraxinicola (after
Witlaczil). oe, (Esophagus ; md,
mid-gut ; ed, hind-gut ; vm, Mal-
pighian vessels ;]s, the coil formed
by the hind-gut and the most
anterior part of the mid-gut.

2 vessels which unite together before entering a common duct. In Aletia, Danais
(Lepidoptera), there are on each side 3 vessels with short common terminal pieces
(Fig. 348, p. 488). In Galleria (Lepidoptera) there is an unpaired terminal piece into
which 5 or 6 branched vessels enter. In Ephippigera and the Crryllidce (Orthoptera}
there are numerous vessels which, uniting into a tuft, enter the hind-gut through
a long common ductus excretorius. In Orthezi'a (Coccidce) there are on each side 2
vessels which unite. The 2 terminal ducts themselves enter an unpaired terminal
piece. The pupae of the Noctuina have 3 pairs of vessels, united in pairs, entering an
unpaired urinary bladder. Lygaeus (Hemiptera] has on each side 2 vessels entering a
urinary bladder.

464

COMPARATIVE ANATOMY

CHAP.

The number of Malpighian vessels is occasionally smaller in the larva than in the
adult. Thus the larva of the Honey bee (Fig. 321) has only 4 vessels. In the
Blattidce and Gryllidce the number increases during the gradual development. In
the Lepidoptera the larva usually possesses the same number as the adult. Among
the Termites only do the young forms possess more numerous Malpighian vessels
than the adults.

V. The Nervous System.

This appears in the form which is characteristic of the Arthropoda

and consists of the brain (supra -
oesophageal ganglion), the cesophageal
commissures, and the ventral* chord.
The brain, which lies in the head
above the oesophagus, often attains to a
high degree of development (especially
in the highly developed Hymenoptera),
and is distinguished by the formation
of lobes (ganglion opticum, olfactory
lobes, etc.) From it arise the nerves
for the sensory organs which lie in
the head, for the eyes, the antennae, and
the olfactory organs on the antennae.
We can always distinguish in the
ventral chord a cephalic and a trunk
portion. The former consists of the
infra -oesophageal ganglion, composed
of the fused ganglia of the oral limbs,
which in the embryo are often separate.
The trunk portion of the ventral chord
must originally have consisted of as
many double ganglia united by longi-
tudinal commissures, as there are trunk
segments, but the ganglia of some of
the last trunk segments are always
fused to form a terminal ganglion,
generally somewhat larger in size than
the rest. The ventral chord is found
in this unconcentrated form in the
Myriapoda, Apterygota, and many Ptery-
gota, and especially in the larvae of the
Hexapoda. We find, however, within
various orders of the Hexapoda more

a, Antenna* ; fc/, maxillipedes (poison feet); or less pronounced Concentration of

the ventral chord in a way similar to
that described in connection with the
Crustacea. This concentration takes place by the fusing of pairs of
ganglia • it may appear in the abdomen as well as in the thorax and

FIG. 323.— Lithobius forficulatus seen
from the ventral side (after R. Leuckart).

sd, salivary glands ; bm, ventral chord ; cp,
coxal pores.

ANTENNATA—THE NERVOUS SYSTEM

465

generally, not always, proceeds in both from
behind forward. A junction of the fused ganglia
of thorax and abdomen to form a large thoracic
ganglionic mass may even take place (as in the
Bmchyura and many Copepoda) ; such cases occur
in the Diptera and Rhijnclwta. Although the
larva? generally possess a less concentrated
nervous system than the imagines, so that the
progressive concentration can often be followed
ontogenetically in the same species, this is not
always the case, in fact the very reverse occasion-
ally occurs. The interesting relation between
the nervous system of the larva and that of the
imago will be again referred to.

From the ganglia of the ventral chord of the
trunk (thorax and abdomen) arise the nerves for
its integument, musculature, glands, and limbs.
The 2 ganglia of a double ganglion are always
closely contiguous, and appear as one mass con-
sisting of two halves ; the longitudinal com-
missures, however, which unite the consecutive
ganglia very often remain separate. A sym-
pathetic nervous system seems present in all
Antenriata.

Myriapoda (Fig. 323).

One ganglion is found in each trunk segment. The
ganglia are mostly united by distinctly separate longi-
tudinal commissures. In the Pauropoda and Symphyla,
however, the ventral chord is a median strand with con-
secutive swellings, corresponding with the ganglia, and in
this strand the longitudinal commissures are not separate.

The 2 anterior trunk ganglia (or in Symphyla only
the first) generally form with the sub-oesophageal ganglion
a single mass, in which, however, the original composition
can easily be made out. The limbless anal segment has
no separate ganglion, and the ganglia of the 2 or 3 pre-
ceding segments are fused together.

The double segments of the Diplopoda each have 2
ganglia.

Hexapoda.

Apterygota. — In this division we have very good
illustrations of the concentrated and non - concentrated
nervous systems. The Thysanura have a non-concentrated

FIG. 324.— Central nervous system of Machilis maritima
(after Oudemans). au, Eye ; lo, lobus opticus ; g, brain ; an,
antennal nerve ; oe, oesophagus passing between the cesophageal
commissures ; usg, infra - cesophageal ganglion ; I • III, thoracic
ganglia ; 1-8, abdominal ganglia, the last (8 a b c) consisting of three
fused ganglia ; s, sympathetic nervous system of the ventral chord.

VOL. I

• ce

2 H

466

COMPARATIVE ANATOMY

CHAP.

nervous system, consisting of the brain, ojsophageal commissures, infra-oesophageal
ganglion, 3 ganglia of the 3 thoracic segments and 8 (in Campodea 7) ganglia of the
abdomen (Fig. 324). The finer structure of the last and largest abdominal ganglion
and the number of nerves proceeding from it show that it consists of 3 fused ganglia.
The number of abdominal ganglia would according to this be 10, corresponding with
the number of segments. In the Collembola the number of abdominal segments is
reduced, and in accordance with this reduction of the body there is, as it appears,
only 1 abdominal ganglion. Sminthurus is said to have only 1 thoracic ganglion.
The 2 longitudinal commissures remain distinctly separate in the Thysanura.
From each ganglion 2 nerves are given off on each side, and the same number proceed
from the cesophageal commissures. In front of the infra-cesophageal ganglion and
behind the oesophagus a transverse commissure connects the cesophageal commissures.
Pterygota. — The nervous system of the winged Insecta shows very great
variety in its arrangement : it is impossible here to go into details — the Diptera

FIG. S25.—A-B, The nervous systems of 4 species of Diptera, to demonstrate their various
degrees of concentration. A, Non-concentrated nervous system of Chrionomus plumosus, with 3
thoracic and 6 abdominal ganglionic masses. B, Nervous system of Empis stercorea, with 2 thoracic
and 5 abdominal ganglionic masses. C, Nervous system of Tabanus bovinus, with one thoracic
ganglionic mass and the abdominal ganglia moved towards each other. D, Nervous system of
Sarcophaga carnaria. All the ganglia of the ventral chord except the infra-oesophageal ganglion,
which always remains separate, are here united into one single thoracic ganglion mass (after E.
Brand).

(Fig. 325, A~D) are particularly instructive. In no other natural order of insects
are the extremes so great, and yet connected by such numerous intermediate stages.
The series begins with the suborder of the Nemocera, the Culicidce, Culiciformes,
Tipulidce, Fungicolce (e.g. Chironomus, A), which have very slightly concentrated
nervous systems. The ventral chord here consists of an infra-cesophageal ganglion,
3 thoracic ganglia, and 5-6 abdominal ganglia. The last thoracic ganglion is not
simple, but at least 1 of the anterior abdominal ganglia is fused with it. The last
and largest abdominal ganglion is also not simple ; it consists of several (in Chirono-
mus probably 2) fused ganglia. The concentration of the nervous system among the

vi ANTENNATA—THE NERVOUS SYSTEM 467

Diptcra begins in the families of the Empidce, Asilidce, Therecidce, Xylophagidce,
Bibionidce (e.g. Empis, B), where the 2 anterior thoracic ganglia become fused, so
that there are only two thoracic ganglia. In this respect the Diptera form a con-
trast to other insects with only 2 thoracic ganglia, e.g. many Coleoptera, Lepidoptera,
and Hymenoptera, in these cases it is the posterior thoracic ganglion which consists
of the 2-fused posterior ganglia. Tabanus (Fig. 325, C] exhibits a nervous system
in which all the 3 thoracic ganglia are fused into 1 thoracic ganglionic mass. This
is the case in the families of the Syrpliidce, Stratyomidce, and Tabanidce. The
abdominal ganglia show a tendency to approach each other and to fuse. Finally
the highest degree of concentration among the Diptera is shown by the Muscidce,
(Estridce, and Piqnparce, where all the ganglia of the ventral chord, except the infra-
cesophageal ganglion, are fused into 1 large thoracic ganglionic mass (Fig. 325, D,
Sarcophaga). From this mass a median nerve then runs towards the end of the
abdomen, giving off nerves to the abdominal segments at regular intervals.

A series similar to the above occurs in the Coleoptera, but the concentration
here rarely goes so far as in the Diptera, since, though the abdominal ganglia may
be wanting (in the Lamellicornia), the 2 thoracic ganglionic masses always remain
separate. Wherever among insects separate abdominal ganglia are wanting these are
fused with the most posterior thoracic ganglion, from which the abdominal nerves
then often radiate backward, like the cauda equina in vertebrates. These abdominal
nerves, however, may be united on each side into an abdominal longitudinal bundle,
or these 2 longitudinal bundles may be fused to form 1 median abdominal strand.
The Rhynchota, the Mallophaga (Corrodentia), and the Thysanoptera possess a much
concentrated nervous system. In many Rhynchota all the thoracic ganglia, not
excluding the infra-cesophageal ganglion, may fuse into 1 ganglionic niass, as is
the case in the Coccidce and also to a lesser degree in the Aphides.

All other insects have a non-concentrated or else slightly concentrated nervous
system, with separate infra-cesophageal ganglion, at least 2 thoracic and several
abdominal ganglionic masses, at the most 8 and rarely only 1.

The full number of abdominal ganglia is not found in any insect larva or imago.
In insect-embryos, however, the rudiments of all the 10 abdominal ganglia have been
observed.

It is clear from the above that the arrangement of the nervous system can be as
little used as a criterion for the natural division of insects as the structure of any
other organic system by itself. It can at the most be used for limiting the sub-
divisions within the orders.

The relation of the larval Nervous System to that of the Imago. — (1) Where
the nervous system of the imago is not concentrated, it is generally not concentrated
in the larva ; this is evidently the original condition.

(2) Where single ganglia are fused in the imago, they are often separate in the
larva. The honey bee affords an illustration of this ; the bee larva (Fig. 321, p. 463)
possesses the fully segmented nervous system : brain, infra-cesophageal ganglion, 3
thoracic and 8 abdominal ganglia. The last abdominal ganglion comes from three
rudimentary ganglia, which are separate in the embryo. The adult bee (Fig. 320,
p. 462) possesses a brain, infra-cesophageal ganglion, 2 thoracic and 4 abdominal
ganglia. The posterior and larger thoracic ganglion consists of the 2d and 3d
thoracic ganglia fused together ; the composition of the last abdominal ganglion out
of 3 ganglia can still be clearly made out.

(3) "Where the nervous system in the imago is much concentrated, it is very often
(e.g. Muscidce} much concentrated in the larva also, and at the same time slightly
differentiated. We have here a case of the imaginal characteristics being shifted
back on to the larval stage.

(4) The nervous system in the larva is seldom much concentrated when not con-

468

COMPARATIVE ANATOMY

CHAP.

centrated in the imago. Myrmeleon is, however, a case of adaptation of the nervous
system to the short compressed form of body of the larva.

net JTT d <rf

i i *J

710

ff*

f&r

FIG. 326.— Median longitudinal section through the head of Blatta orientals. The nervous
system of the head is drawn entire, hyp, Hypopharynx ; os, oral cavity ; Ibr, upper lip ; gf, ganglion

frontale ; g, brain ; na, root of the antennal nerve ;
TIO, root of the optic nerve ; ga, anterior ; gp, pos-
terior ganglion of the paired visceral nervous system ;
oe, oesophagus ; c, cesophageal commissure ; usg,
infra-oesophageal ganglion ; cc, longitudinal commis-
sure between this and the first thoracic ganglion ;
sg, common duct of the salivary glands ; Ib, lower
lip = 2d pair of maxillae ; nr, nervus recurrens ; d,
nerve uniting the frontal ganglion with the oeso-
phageal commissure ; e, nerve from this commissure
to the upper lip ; /, nerve from the infra-oesophageal
ganglion to the mandible ; g, to the anterior max-
illae ; h, to the lower lip (after Bruno Hofer).

(5) In many cases where in the larva
there is an apparently concentrated ventral
chord, its ganglia are quite distinct although
lying very close together. In the imago
they separate and the longitudinal commis-
sures become distinct, so that a non- con-
centrated or less concentrated imaginal
nervous system is developed.

The sympathetic nervous system seems
to be present in all Antennata. It consists
of an unpaired and a paired system. In
Blatta, whose visceral nervous system (Figs.
326 and 327) has been the most investigated,
the unpaired portion shows the following
arrangement. In front of the brain on the
oesophagus there lies an unpaired ganglion
frontale gf, which gives off nerves to the
upper lip and the oesophagus. It is con-
nected by a nerve on each side with the
cesophageal commissure, from which another
nerve as well goes off to the oesophagus and

FIG. 327.— Anterior portion of the paired
and unpaired visceral nervous system of
Blatta orientalis, seen from above. The
outlines of the brain (g) and the roots of the
antennal nerve (na), which cover a portion of
the sympathetic nervous system, are given by
dotted lines. Lettering as in Fig. 326. nsd,
Nerve to salivary -gland. The nervus recurrens
(nr) enters an unpaired stomach ganglion
further back (after Bruno Hofer).

vi ANTENN ATA— SENSORY ORGANS 469

the upper lip. From the ganglion frontale an unpaired median nerve, the nervus
recurrens nr, runs back under the brain along the dorsal wall of the oesophagus,
and enters an unpaired stomach ganglion in front of the masticatory stomach. A
lateral pair of nerves arises from this, 2 small ganglia occurring in their course.

The paired visceral nervous system consists of two pairs of ganglia ga and gp,
which lie on the oasophagus, the anterior pair being covered by the brain. These
nerves are connected with each other, with the nervus recurrens, and with the brain
by anastomoses. The nervus recurrens and the paired ganglia give off nerve
branches to the oesophagus and to the salivary glands.

Besides the above nerves insects may have sympathetic thoracic and abdominal
ganglia either paired or unpaired. The paired portion of the visceral system may
also be wanting.

In the Lepid'optera, close above the abdominal portion of the ventral chord, is
found a longitudinal strand of connective tissue, which seems to be a formation of
the Neurilemma of the ventral chord. Muscles are attached to this which run to
the neighbouring ventral exoskeleton. This strand, whose significance is not yet
sufficiently explained, has been called the chorda supra spinale. It has nothing to
do with the vertebrate chorda.

VI. Sensory Organs.

A. Eyes.

We can distinguish single -lensed eyes or ocelli from compound
eyes (facet eyes). The Myriwpoda have ocelli generally in large
numbers grouped closely together on each side dor sally. On]y
Scutiyera has a compound eye on each side : this eye, however, differs
in structure in many respects from the compound eye of the Insecta.

Most adult Hexapoda have ocelli as well as facet eyes. The small
ocelli then generally lie in threes on the frontal region between the
two large facet eyes. The larvae have ocelli only, these often occurring
in great numbers. Ocelli are seldom found alone (i.e. without facet
eyes) in adult Hexapoda, but this is the case in the Collemlola among
the Apterygota, and in lice (Pediculidce) and fleas (Aphaniptera). Ocelli
are wanting in adult Dermaptera, and among the Orthoptera in the
Locustidce, among the Rhynclwta in the Hydrocora, among the Lepidoptem
in the Geometrina and Rhopalocera, which thus as adults possess only
facet eyes.

Structure of the Ocelli. — The simply constructed ocellus of the
Dytiscus larva is very instructive (Fig. 328). Above the ocellus the
chitin thickens into a lens. Below it the hypodermis is depressed in
the form of a pit. The hypodermis cells standing at the base of this
depression form the retina of the eye. Each retinal cell is connected
with a nerve fibre, contains pigment, and produced outward towards
the lens in the form of a rod. The cells at the edge of the depression
are free from pigment at their outer ends, and push their way in
between the retina and the lens, filling it up and forming a sort of
vitreous body. The ocelli of other Insecta and Myriapoda are similarly

470

COMPARATIVE ANATOMY

CHAP.

re

no

FIG. 328.— Section through the ocellus of a young
Dytiscus larva (after Grenacher). ct, Chitinous cuticle ;

constructed, with this distinction, that they are in most cases completely
demarcated vesicles, over which and below the lens the hypodermis

functions as a vitreous
body.

The structure of the
compound eyes agrees in
general with that of the
compound Crustacean eye
(cf. p. 353). There is al-
ways a bi- convex corneal
lens for each single eye.
Eueone and Aeone eyes
are distinguished, according
to whether crystal cones
are formed (as in the

I, cuticular lens ; gk, cells of the vitreous body ; liy, hypo- Crustacea) Or not Acone
clermis ; st, rods ; re, retinal cells ; no, optic nerve. / , . ' , ~ ,

eyes are found in the Coleo-

ptera (excluding the Pentamera), in the Heteroptera, the Tipularidce among
the Diptera, and the Dermaptera, In the acone type (e.g. Tipula) each
corneal lens is separated from its neighbour by a strongly pigmented
zone. Under each lens lies a conical group of 4 crystal cells, whose
proximal end is imbedded between two pigment cells. The retinula,
consisting of 7 cells (6 marginal cells and 1 axial cell), joins the above.
Each of the retinular cells contains in its proximal portion a nucleus,
and in its distal portion a rod (rhabdomere). The single eyes are
separated from each other by pigment cells.

The elements of the optic ganglia (ganglion opticum and retinal
ganglion) are arranged in a very complicated manner in the Insecta.

The Ocellus and the Facet Eye. — Attempts have been made to connect these two
forms of eyes. According to the view most widely approved, the two eyes are to be
derived from a primitive eye resembling the single eye (ommatidium) of the acone
eye of Tipula. An increase of the elements of this primitive eye led to the
formation of the ocellus ; an increase in number of the primitive eyes and their
approximation led to the formation of the compound facet eye. For confirmation
of this view we are referred to the groups of closely contiguous single eyes o.f the
Myriapoda, considered in connection with the compound eye of Scutigera. But it is
difficult to reconcile with this view several facts in the ontogeny of the eye, and
especially the structure of the middle eye of the Scorpion. Such a scorpion eye on
the one hand contains only a single lens, while on the other the retinal elements are
grouped into retinulse.

The last w'ord is certainly not yet spoken with regard to the finer structure and
the morphological significance of the Arthropodan eye, and we must here briefly
allude to an entirely new view. According to this view the compound eye consists
of two layers : (1) of a hypodermis layer which yields the single corneal lenses, and (2)
of a subjacent layer of single eyes. The latter is said to be a single layer, the
elements of the single eye known as crystal cells, retinular 'cells, and pigment cells
running with their processes through its whole thickness (Fig. 329, A}. The
rhabdomes and the rhabdomeres forming them are not secreted products of the
retinula, but belong, like the crystalline cone, to the crystalline cone cells, which are

ANTENNATA— SENSORY ORGANS

471

called retinophorse. The crystalline cones are the elements actually sensitive to light.
The retinophor?e are surrounded by pigment cells. A proximal (inner) ring of these
cells corresponds with the so-called retinulse, but these cells are also continued in the
form of fine processes as far as the hypodermis. Fig. 329, A, illustrates this view.
The layer of single eyes is said to answer-
to the posterior wall of a vesicular eye,
which first forms as an ectodermal de-
pression and later becomes constricted off.
The whole compound eye would according
to this be a modified and differentiated
ocellus, in which the epithelial cells of the
posterior proximal wall of the eye -pit
differentiate into retinophone and pigment
cells, grouped together as ommatidia (as in
the middle eye of the Scorpion). The
hypodermis, further, which spreads over
the vesicular eye, instead of forming one
cuticular lens, forms many such lenses,
which as corneal facets correspond in
number and position with the ommatidia.
The ocelli themselves may also be compared
with the similarly constructed visual organs
in the Annelida and Mollusca which arise
as ectodermal invaginations.

If this new view of the facet eye is
established, i.e. if the so-called crystalline
cones are not simply refractive bodies, but
the actual terminal apparatus of the optic
nerve, sensitive to light, it would lead to
considerable modification of the theory of
sight by means of the compound eye.

B. Auditory Organs.

FIG. 329.— The structure of an omma-
tidium (single eye) of the facet eye. A,
According to Patten's view. B, according
to Grenacher's view, cl, Cuticular corneal

nuclei of the same ; k, crystalline cone ; p,
pigment cells ; ret, retinulae ; rh, rhabdome ;
71, nerve. According to Patten (A) the
ommatidium is, apart from the corneal
hypodermis, of one layer, all its elements

In the most Various parts Of the lens ; Jiy, hypodermis cells of the corneal
bodies of the Insecta peculiar nerve lenses; r^retinophora = crystal cells; nr,

endings occur which are evidently
sensory ; their structure is as follows :
A peripheral nerve fibre enters a

1 . • •••• !_'!,* * "j-J.

ganghomc cell, which is in its turn passing by means of thin pr0cesses through

Connected with the integument by its whole thickness from the base to the

means of a long slender stretched Z^' S^fSSfZ
tube. An axial thread irom the consists of two layers.
ganglionic cell enters the tube, and

there ends in a terminal rod. The tube which conceals this rod is
called the seolopophore, and that part of it which becomes attached
to the integument is the terminal tube. Scolopophores with their
ganglia are rarely found singly, they are usually united in groups of
varying size (Figs. 330, 331, and 332). If the number of scolopo-
phores is small their terminal tubes are gathered into a bundle, but if

472

COMPARATIVE ANATOMY

CHAP.

great, the terminal tubes are generally isolated, running either irregu-
larly or radially, or in the shape of a fan, etc., to the integument.
These peculiar sensory organs are called ehordotonal organs, and
are considered to be auditory. A ehordotonal organ may be either
simply connected with the integument by a prolongation of its nerve,
or bent at right angles to its nerve and run parallel to the integument
(Fig. 330). The latter is the case when the terminal tubes of the

FIG. 330.— Right half of the 8th trunk segment of an older larva of Corethra plumicornis ;
nervous system and sensory organs (after v. Graber). g, Ganglion of the ventral chord ; Im, longi-
tudinal muscles ; en, ehordotonal nerve ; d, ehordotonal ligament ; eg, ehordotonal ganglion ; cs, rod
of the ehordotonal organ ; cst, terminal strand ; tb, tactile setse ; hn, outgoing fibres of the integu-
mental nerves.

ehordotonal organ unite into one tube, and then the bent portion is
connected with the integument by a second tube, the ehordotonal
ligament. The ehordotonal organ, together with the ehordotonal
ligament, then form a chord stretched between two parts of one and
the same segment.

The nerve forms a terminal ganglion cell at the point where it joins
the chord.

If the exoskeleton is affected by vibrations of sound, this apparatus
with its terminal rods is set in vibration, and a sensation of sound may
thus be produced.

Chordotonal organs, which offer great varieties of structure, occur in all orders
of the Pterygota, and are found in the most various parts of the body, in the trunk,
in the legs, the wings, the mouth parts, and the antennae. They may appear in
different forms in various parts of the body of the same animal.

ANTENNATA— SENSORY ORGANS

473

In the wings, but especially in the halteres or balancers of the Diptera, which are
minute transformed hind wings, they are connected with areas of the exoskeleton
provided with peculiar pores or papillre. They always lie superficially. A chordo-
tonal organ is, further, never stretched between points of the integument of two con-

FIG. 331.— The Chordotonal
organ of Fig. 330, strongly mag-
nified, el, Chordotonal ligament ;
en, chordotonal nerve ; eg, chord-
otonal ganglion ; cst, chordotonal
rod ; es, terminal tube.

FIG. 332.— So-called sub-genual Chordotonal organ
in the tibia of the middle leg of Isopteryx apicalis
(Perlid) (after v. Graber). tr, Trachea; bk, blood
corpuscles ; gz, nerve cells ; sc, scolopophores with
their rods ; es, terminal fibrous strands, attached to the
integument (c).

secutive segments of the body, or limbs, movable on one another, but always runs

within one of the same joint or segment, and is thus not affected by the movements

of the animal. The chordotonal nerves always arise out of the

ganglion belonging to their own segment. Chordotonal sensory

organs have till now been chiefly observed in the larvae of the

Insecta.

The tympanal organs of the Saltatoria, which have been long
known, agree in the finer structure of their nerve endings with
the chordotonal organs. It was this agreement which led to the
assumption that the so-called chordotonal organs were also
auditory organs, for the tympanal organs of the Saltatoria have
long been universally regarded as auditory organs, although the
animals continue to hear after their removal. This last fact
favoured the view of the acoustic properties of the chordotonal
organs, since the latter occur together with the tympanal organs.

Scolopophores are very numerous in the Saltatoria (over 100).
In the Acridiidce their terminal tubes are attached to points of
the inner surfaces of the hypodermis of special parts of the
exoskeleton, which, as compared with the surrounding exoskeleton,
are thinned away like a membrane and are called tympana
(Figs. 333 and 334). Such a tympanum may be stretched between
thickened portions of the skeleton, which form a frame for it ;
and, as in an outer ear, the integument may grow round it as a
covering fold for its protection. A tracheal trunk widens to form a cavity under the
tympanum, which may be compared with the hollow of a drum. Between the cavity

FIG. 333. — Tibia
of the fore leg of
Locusta viridis-
sima. td, Cover of
the drum ; tr, fissure
between the drum
and its cover (after
v. Graber).

474

COMPARATIVE ANATOMY

CHAP.

and the tympanum lies the terminal nerve apparatus, the so-called "Miiller's" gan-
glion with the scolopophores, whose terminal tubes attach themselves by peculiar inner
processes to the middle of the tympanum. The tympanum can be stretched by special
muscles. In the Gryllidce and Locustidce the terminal tubes of the scolopophores are

FIG. 334.— Side view of Acridium tartaricum. sj, Stigma of the mesothorax ; s2, stigma of the
metathorax ; sa, stigmata of the abdomen ; t, drum of the tympanal auditory organ (after Fischer).

not attached to the tympanum itself, but above it to the integument. The Locustidce
possess, besides the Miiller's ganglion with its scolopophores, another series of some-
what different scolopophores, lying on a hollow
formed by the trachea, which acts as a resonator.
The tympanal auditory organs of the Saltatoria
lie either (Acridiidce, Fig. 334) in a single pair
at the sides of the first abdominal segment, or
(Gryllidce, Locustidce, Fig. 333) on the tibiae of the
fore-legs. In the last case there are generally on
each tibia 2 tympana lying opposite one another.
Between the basal portions of the maxillse of
Scutigera there lies on each side a pouch, from
the base of which small closely -packed plates and
hairs rise into the cavity. These plates and hairs
stand on equally closely arranged folds of the
cuticle. It has been conjectured that these forma-
tions, which need further examination, may prove
to be auditory organs.

FIG. 335.— Sensory organs con-
sidered to be olfactory at the end
of the Antennae of lulus Sabulosus,
longitudinal section through the an-
tenna, k, Sensory cones ; z, sensory
points; gk, ganglia of the sensory
cones ; gz, ganglia of the sensory
points; n, nerve; grz, large cells in
the ganglia of the sensory cone (after
v. Rath).

C. Olfactory Organs — Gustatory Organs.

The seat of the sense of smell is, as in
the Crustacea, to be sought in the antennae
(and perhaps in the maxillar palps as well).
The antennae are in the Insecta very
often more strongly developed in the
males than in the females. The olfactory
organs are short processes in the shape of
knobs or cones, and generally open at
the point; they are found either free
or at the base of pits in the antennae.

vi ANTENNATA—THE CIRCULATORY SYSTEM 475

Beneath each olfactory process there is a pore in the cuticle, through
which the fibrillae of a hypodermal ganglion pass (Fig. 335). Similar
terminal apparati, which have been observed at the base and at the
point of the tongue, and on the lower side of the maxillae of Hymeno-
ptera, on the inner surface of the labellum of the proboscis in the fly,
and in the gnathochilarium of the Diplopoda, may perhaps be con-
sidered as gustatory organs.

Specific organs of touch are found in the long, simple, or feathered
setae (Fig. 330, p. 472) which occur all over the body surface, and
especially on the feelers and maxillar palps of the Antennata, on the
labellum of the proboscis of the fly, etc. An axial thread enters the
sensitive hair from a hypodermal ganglion cell, and runs through it.

VII. The Circulatory System.

This is very simple in the Antennata. The colourless or light-
yellow or light-green blood, containing amoeboid blood corpuscles, flows
in definite directions in a lacunar system (body cavity). The circula-
tion of the blood is maintained by the contractions of a dorsal vessel
(dorsal heart), which in the extremely reduced condition of the
arterial vascular system is almost the only part which has walls of its
own. The dorsal vessel is a delicate tube running longitudinally above
the intestine, and covered externally and lined internally by membranes
probably elastic. Between the two membranes there runs a system of
delicate muscle fibres, which generally have a circular course and
sometimes cross each other. The dorsal vessel falls into successive
segmental chambers, separated by valvular arrangements, which prevent
the streaming back of the blood from the anterior into the posterior
chambers during the progressive contraction of the vessel from behind
forward. The dorsal vessel is provided with paired lateral ostia,
which, as it appears, are mostly placed intersegmentally, and maintain
an open communication between the surrounding parts of the body
cavity and the interior of the heart. The heart may be connected in
various ways by muscular fibres with neighbouring parts of the body,
i.e. with the intestine, with the dorsal integument, etc. Paired so-
called alary muscles are particularly constant in their occurrence ;
they are almost triangular, and become attached by their narrowed
ends to the latero-dorsal body wall, while their broad ends are fastened
to the chambers of the heart. These alary muscles together form an
incomplete horizontal partition wall above the intestine, marking off
a dorsal sinus in which the heart lies. This sinus may be called the
perieardial sinus. This partition wall is arched upwards, and if
the alary muscles which form it contract, it straightens out and
becomes flat ; the pericardium thus becomes more spacious, and the
blood streams into it from the rest of the body cavity. It was
formerly thought that the alary muscles served to expand the heart.

476

COMPARATIVE ANATOMY

CHAP.

O —

This expansion, however, seems to be merely a result of the elasticity

of its own walls. The whole mechanism, however, requires fresh

investigation.

The heart ends blindly behind, but is continued anteriorly into an

aorta, which empties the blood into the lacunar system of the body.
We can occasionally distinguish with special
clearness a ventral sinus surrounding the ventral
chord (in Myriapoda, Orthoptera, and perhaps also
in certain Apterygota) • in this sinus the blood
flows from before backward. A nerve supplying
the heart like that in Peripatus has here and
there been observed.

Myriapoda (Fig. 336). — The heart runs
through the whole body, and has as many
chambers and pairs of alary muscles as there
are trunk segments. From the most anterior
chamber an aorta arises, which divides into three
branches in the head. In the lulidce and
Scolopendridce lateral arteries are said to diverge
in the neighbourhood of the ostia.

Hexapoda, Apterygota. — The heart of the
Tliysanura consists of 9 chambers, and has 9
pairs of ostia and 9 pairs of weakly developed
alary muscles ; this is the highest number reached
among the Hexapoda. It runs forward as far as
the last thoracic segment, or even to the second ;
this is an important fact, as in all Pterygota,
with the exception of a few Orthoptem (Blatta),
the heart is exclusively confined to the abdomen.
The aorta is short in the Thysanura, as the heart
extends so far forward. The number of cardial
chambers and pairs of ostia is reduced in the

Collembola (5 in Macrotoma).

Pterygota. — With the exception of a few Orthoptera, the heart is

restricted to the abdomen, and has 8 pairs of ostia at the most. The

number of pairs of ostia, of alary muscles, and of cardial chambers

may be reduced like the number of the ganglia of the ventral chord.

The aorta runs through the thorax and may be traced as far as into

the head.

In the Lepidoptera, the aorta bends up dorsally in the thorax, widens considerably
and forms a loop (Fig. 348, ac, p. 488). In a few cases the aorta has been found
branched in the head. In the larva of the Ephe,meridce, blood vessels with walls of
their own pass from the last cardial chamber into the three caudal setse. The valves
between the last cardial chamber and the one before it are here placed in such a way
that they prevent the passage of the blood from the former into the latter. When
the last chamber contracts, therefore, the blood is driven into the three arteries of
the caudal setfe.

FIG. 336.— Anterior end
of the heart of Scolopen-
dra (after Newport), ac,
Arteria cephalica ; ab, art-
erial rings ; al, lateral art-
eries ; hk, cardial chamber ;
o, ostia of the heart; fm,
alary muscles of the same.

vi ANTENNATA—THE RESPIRATORY ORGANS 477

/

The want of a developed arterial vascular system is compensated for as well as
conditioned by the extremely profuse branching of the respiratory organs (tracheae),
the oxygen being thus conveyed to the blood in all parts of the body. Thus,
whilst, as a rule in the higher animals, the blood in its closed channels generally
seeks out the localised seat of the respiratory processes, here, on the contrary,
the respiratory organs seek out the blood and the blood tissue in the most remote
parts of the body.

VIII. Fat Bodies — Luminous Bodies.

In the body cavity lies a mass of large cells filled with small drops
of fat, and forming together the so-called fat body. This is variously
shaped, and covers inner organs which appear enveloped in it ; it also
forms a layer under the integument, etc. It is specially strongly
developed in the larvae, and forms a reserve of nutrition, which is
drawn upon during metamorphosis, during the formation and ripening
of the sexual products, etc. The metabolism which goes on in the
fat body is very active, as is proved by the fact that its cells often
contain concretions of uric acid. In some cases it has been proved
that the fat body in the larva is rich in fat and poor in concretions of
uric acid, while in the imago it is poor in fat and rich in concretions
of uric acid.

There are Coleoptera which possess either on the abdomen (Lampy-
ridce) or on the thorax (a few Elateridce, Pyrophorus) intensely luminous
areas. The seat of light is a luminous organ which, morphologically,
must be considered as a specially differentiated portion of the fat body.
The cells of this luminous organ secrete, under the control of the
nervous system, a substance which is burnt during the appearance of
the light ; this combustion takes place by means of the oxygen con-
veyed to the cells of the luminous body by the tracheae, which branch
profusely in it and break up into capillaries. A weakly luminous
dorsal layer of the luminous organ, which lies ventrally in the pen-
ultimate and antepenultimate abdominal segments of the Lampyridce,
contains very numerous concretions of uric acid.

Other cell elements occurring in the body cavity (e.g. the pericardial
cells lying on the alary muscles of the heart, and occasionally containing
fat) cannot here be more closely considered. They form, together
with the blood corpuscles and the fat body, the so-called blood tissue.

IX. The Respiratory Organs.
A. The Tracheal System.

The respiratory organs of the Antennata are air-conducting canals
(traehese) which, on the one hand, communicate with the outer world
by means of paired, strictly segmentally arranged outer apertures
(stigmata), and on the other spread all over the body, penetrating

478

COMPARATIVE ANATOMY

CHAP.

between the various organs and even the component parts of organs,
breaking up into the finest capillaries. The essential structure of the
tracheae is everywhere the same, whether we examine a principal trunk
or a fine terminal branch. The tracheal tube (Fig. 337) is lined inter-
nally with a chitinous intima, which is a continuation of the chitinous
cuticle of the body, and, like the latter, is
shed during ecdysis. The intima is crenu-
lated spirally. This chitinous spiral chiefly
serves for holding the tracheae open. The
intima is clothed with an outer layer of cells,

I kg the boundaries of which are often difficult

to distinguish. This layer represents the
epithelium which secretes the intima, and
is a continuation of the outer hypodermis.
The outer apertures of the tracheal system,
the so-called stigmata, are wonderfully
varied in their arrangement. They some-
times lie openly on the surface, sometimes
are partly concealed by cuticular folds, or
again are covered by the elytra (Coleoptera).
Setae often project over the aperture from
its edges, and these may be elegantly

. bra»ched o^ feathered' so as to represent a

the epithelium and chitinous intima kind of fish-trap apparatus, which prevents
removed in order to show the the entrance of foreign bodies, dust, etc.,
frcS £S±?_*£K ™th the air. The outer aperture is often
the chitinous intima (cc), in which long and slit-like ; in other cases, on the con-
can be seen the spiral thickening trary ft ig large an(j tas a sieve-like mem-
(spiral thread). t* * i. j -± T> ±1,

brane stretched over it. Beneath each

stigma, at the beginning of each principal tracheal trunk, there is a
closing apparatus which we cannot here more minutely describe ; by
means of this apparatus, into which nerves enter, the trachea can be
completely closed towards the exterior. The air in the tracheae is
chiefly renewed by the respiratory movements of the abdomen (in the
Hempoda); these movements are caused by the contractions of the
muscular fibres which run dorso-ventrally. By the contraction of the
abdomen (exspiration) the trachea are compressed and the air driven
out through the stigmata, or, if the stigmata are closed, forced into
the tracheal system of the thorax, head, and extremities. By the
expansion of the abdomen (inspiration) new air from without again
enters the tracheae, the elastic spiral of the tracheal tube playing an
important part in its expansion.

We are justified in assuming, that the tracheal system of the
Antennata originally consisted of as many pairs of isolated tracheal
bundles and stigmata as there were body segments (excluding the
anal segment). Eeduction, however, has everywhere taken place, first
of all at the anterier and posterior ends of the body, so that in some

ANTENNATA—THE RESPIRATORY ORGANS

479

cases only a single pair of stigmata is left ; the tracheal system under-
goes corresponding modifications, the most important of which is the
connecting together of the originally separate bundles of tracheae by
means of transverse and longitudinal anastomoses.

With reference to the scattered and irregular apertures of the trachea in the
Protracheata, we must point out that the arrangement of the tracheal system in all
Antennata indicates a strictly segmental order of tracheal apertures in the racial form.
There is never more than one pair of tracheae in one segment in the Antennata.

Myriapoda.

The most primitive arrangement is found in the Diplopoda, where one pair of
stigmata and one pair of tracheal bundles occurs in each trunk segment. Each double
segment also has 2 pairs of stigmata and 2 pairs of tracheal bundles. The separate

FIG. 338.— Tracheal mass of a dorsal plate of Scutigera coleoptera. A, from above ; 13, in
transverse section through the inter-segmental fold of the dorsal plate ; diagrammatic (after Haase).
The tracheae, which enter the air cavity (ca) from both sides, are marked white, vs, Anterior ; hs,
posterior stomatic aperture ; as (A) and se (B), outer ; ms (A) and si (B), inner stomatic slit.

tracheal bundles are not connected by anastomoses. Each stigma leads into a
tracheal sac, whose base is produced in the form of pointed horns, into which the
numerous but unbranched trachea enter. Branched tracheae seem to occur only in
the Glomeridce.

The tracheal system of the Chilopoda is seen to be a secondary development from
the fact that the tracheae branch profusely, and that the tracheae belonging to each
stigma anastomose with one another transversely and longitudinally. Only in
the Gteophilidce and Plutonium one pair of stigmata is retained on each leg-bearing
segment, except the first and last. In the species Lithobius and in the Scolopendridce
(except Plutonium], beginning with the third trunk segment, a pair of stigmata is
found with considerable regularity on every second segment. The last leg-bearing
segment is here also without stigmata. In Scutigera and Henicops the stigmata

480 COMPARATIVE ANATOMY CHAP.

belong to the trunk segments 1, 3, 5, 8, 10, 12 (and 14 when there are 15 trunk
segments in all).

The tracheal system of Scutigera (Fig. 338) is distinguished by many peculi-
arities. The stigmata here are unpaired and lie in the dorsal middle line. Each
stigma leads into an air sac, into which on each side about 300 closely packed,
radially placed, branched tracheal tubes enter. This form of trachea is specially
interesting, as perhaps helping to explain the origin of the so-called book-leaf trachea?
of the Arachnoidea, which will be described later.

The respiratory system of the Symphyla, which in other respects may be reckoned
as very primitive Myriapodan forms, is much reduced. There are only 2 stigmata,
and these are on the under side of the head under the feelers. This is the only case
in the whole division of the Antennata in which a pair of stigmata has been retained
in the head. The branched trachea? do not extend either into the legs and feelers,
or into the posterior region of the body at all.

In the Pauropoda the tracheal system seems to be entirely degenerated.

Hexapoda.

Apterygota. — The tracheal system in a few Thysanura exhibits the original
condition in so far that the longitudinal and transverse anastomoses are wanting (in
a few species of Machilis, and in the Campodea, Fig. 339). In other Thysanura,
however, they are present (lapyx, Nicoletia, Lepisma, and a few species of Machilis}.
The most frequent number of pairs of stigmata seems to be 10 (Nicoletia, Lepisma,
Lepismina] ; in this case two occur on the thorax and 8 on the abdomen. Machilis
maritima has only 9 pairs of stigmata, 2 thoracic, and 7 abdominal. lapyx is said
to possess 11 pairs, viz. 7 on the abdomen, and — an arrangement standing quite
alone among the Antennata — 4 on the thorax. In Campodea the number is reduced
to 3 pairs belonging to the thorax (the last pair possibly belonging to the first
abdominal segment).

Pterygota (Fig. 340). — Transverse and longitudinal anastomoses appear to be
everywhere present. Among the longitudinal anastomoses on each side one develops
more than the rest, and at first sight it appears as if these tracheal trunks formed
the central portion of the tracheal system, from which, besides numerous lateral
branches penetrating the body, branches run to the stigmata, and there open out-
wardly. In the Diptera, Hymenoptera, and among the Coleoptera in the Hydro-
philidce and Lamellicornea, the trachea? may widen out in some places into large
tracheal sacs without spiral crenulations, known as air sacs (Fig. 320, p. 462).
The number and arrangement of the stigmata vary within wide limits. In the
same insect, at the same time, there are never more than 10 pairs of stigmata. Of
these 10 pairs there is one pair on the mesothorax, one on the meta thorax, and one
on each of the first 8 abdominal segments.

The imagines of by far the greater number of insects are holopneustic, i.e. they
possess many pairs of open stigmata, though the number of stigmata may be reduced,
especially in the abdomen. The Aphaniptera alone have more than 2 pairs of
stigmata in the thorax ; in them there is one prothoracic pair, not met with in other
adult Insecta. The tracheal system of the larvae of insects exhibits interesting
peculiarities, which are of great morphological importance.

1. The most primitive condition is found in those insects which undergo gradual
metamorphosis, and whose larva? live during all stages, like the imagines, on land
(Orthoptera}. These larva? are holopneustic, and their tracheal system simply passes
into that of the imago. Holopneustic larvae may also occur in insects with complete
metamorphosis, as in many Coleoptera (Malacodermata], which thus appear to be a
primitive group.

VI

ANTENNATA—THE RESPIRATORY ORGANS

481

2. In many Insects with incomplete metamorphosis the holopneustic condition
of the tracheal system is much altered in the larvae by the adaptation to aquatic
life. In the aquatic larvae of the ffphcmeridce, Odonata, Plccoptera, for instance,
there are no open stigmata, the tracheal system is completely closed, i.e. it is
apneustic. The rudiments of the tracheal branches running from the stigmata to
the longitudinal trunks are, however, present, but they are empty of air, and appear

FIG. 339. — Machilis maritima,
representing the tracheal system of
the right side (after Oudemans). Ic,
Head ; I, II, III, thoracic segments ;
1-10, abdominal segments ; s, stig-
mata.

FIG. 340. — Half -developed larva of an
agamous unwinged female of Aphis Pelar-
gonii. The tracheal system seen from above
(after Witlaczil). 7c, Head ; a, antenna ; I, II,
III, segments of the thorax ; IV-XII, segments
of the abdomen ; &j, &2> &3> the 3 pairs of legs ;
s, the stigmata of the tracheal system.

in the form of strands (Fig. 343, ^/) ; these play an important part in the shedding
of the larval tracheae, and then for a time open outwardly. When the last larval
stage passes into the imaginal stage these strands become hollow, and the tracheal
system becomes holopneustic. All such larvse breathe by means of tracheal gills.

3. In the majority of the larvse of insects which undergo complete metamorphosis,
the stigmata of those segments which in the imago carry wings (meso- and meta-
thorax) are closed, but on the other hand one pair of prothoracic stigmata wanting
in the imago is usually found. We evidently have here a delayed differentiation of
the tracheae which supply the wings and their musculature, referable to the absence
VOL. I 2 I

482

COMPARATIVE ANATOMY

CHAP.

t-e --

of wings in the larvse. The stigmatic strands belonging to these parts are present
as rudiments. The tracheal system of these larvse is peripneustic.

4. The tracheal system of peripneustic larvse may be modified in various ways by
adaptation to different modes of life : (a) it may become apneustic in larvse inhabit-
ing water, as in the larvse of the Phryganidce and Sialidce, which breathe through
tracheal gills, (b) By adaptation to life in water or parasitic life all the stigmata
may remain closed in the larvse except the last pair. . The tracheal
system is then called metapneustic. The larvse then obtain air
at the surface of the water or of the host, by means of this
posteriorly placed pair of stigmata, which is often elongated
like a siphon, or provided with other suitable structural adapta-
tions. The larvse of the water beetle and of many Diptera,
which are aquatic or parasitic, are metapneustic. (c) There is
occasionally besides the posterior an anterior prothoracic open
pair of stigmata (Fig. 341). This amphipneustic tracheal system
is found in many parasitic or half - parasitic Diptera larvse
(Oestridce, Asilidce), which stretch only their anterior and
posterior ends beyond the medium which surrounds the rest of
the body. The larval stigmata of the meta- and amphipneustic
fly larvse disappear during metamorphosis.

In all cases where the larva is not holopneustic, the stigmatic
branches of the tracheal system are present as rudiments. "We
must distinguish between such first rudiments remaining latent
during the larval period, and those rudiments of stigmatic
branches which are found in the imagines of the various Insecta.
The latter are the remains of stigmata which have disappeared.
Several pairs of such stigmata are often found in the abdomen.

The peculiar arrangements of the tracheal system in insect
larvse show very clearly to what an extent special conditions of
existence may influence the organisation of free-living larvse.

FIG. 341. -Right B. The Tracheal Gills (Figs. 342 and 343).

side of the tracheal

system of a fly mag- Something has already been said about these
got, seen from the respiratory organs of aquatic insect larvse in the
stigma ;!'*, interior section on " wings." Tracheal gills, i.e. delicate mem-
stigma • ti, longitud- branous processes of the body into which trachese
inai tracheal trunks. exten(jj are found not only in the larvse of the
E ' [jhemeridce, Trichoptera, and Sialidce there mentioned, but also in the
larvse of the Plecoptera (Perlidce), Odonata, and the aquatic larvse of a
few species of Diptera, Hymenopt&ra, Lepidoptera, and Coleoptera. The
tracheal gills of the Odonata are either external (Agrion) in the form of
3 branchial leaves on the last abdominal ring, or they are internal
(Libellula, jEschna) in the form of folds in the rectum. In the latter
case water is alternately drawn in and expelled through the anus.
The tracheal gills of the larvse of the Perlidce are very variously
formed; they are pouch -shaped or tufted, etc., and occur at very
different parts of the body. The same is the case with the tracheal
gills which occur singly in the larvse of Diptera, Hymenoptera, Lepidop-
tera, and Coleoptera. Larvse which are provided with tracheal gills are

vi ANTENNATA— SOUND-PRODUCING APPARATUS

483

apneustic. The tracheal gills, according to all observers, are respiratory
organs, which have arisen independently of each other in various
insect orders as adaptations to aquatic life. They are thrown off

...M

FIG. 342. — Ephemerid larva with two
tracheal gills on each side of each ab-
dominal segment, and with 3 caudal pro-
cesses (cerci) (after R. Leuckart).

Fio. 343. — Right half of the middle
abdominal segment of the larva of Baetis
(Chloe) binoculatus with tracheal gills (after
Palmen). trl, Longitudinal tracheal trunks ;
vf, strand like threads of attachment of the
longitudinal trunks to the integument (stig-
matic strands) ; ktr, gill tracheae ; trie, tracheal
gills.

during the transition to the imaginal form in the Ephemeridce, Agrionidce,
and Diptera, but are retained in the imagines of the Perlidce.,
Sialidce, Lepidoptera, and Coleoptera.

X. Sound -Producing Apparatus.

It is well known that many insects can produce sounds. These
sounds, which play many different parts in insect life, are principally
produced —

1. By rapid vibrations of the wings (Hymenoptem, Diptem), and
by the vibration of the halteres against the alulae (Diptera).

2. By the vibration of leaf -like appendages in the tracheae.
These often lie in great numbers near the stigmata, and are made to
vibrate during the respiratory movements of the animal by the inward
and outward streaming of the air (Hymenoptera, Diptera).

3. By the rubbing together of rough uneven portions of the
integument. The Acrididce stridulate by scraping their posterior femur

484 COMPARATIVE ANATOMY CHAP.

like the bow of a fiddle over the projecting ribs of the upper wing.
In the Locustidce and Gryllidce only the males stridulate, by rubbing
the rough basal portions of their wing cases against each other. In
other insects sounds are produced by the rubbing together of other
parts of the body.

4. In the singing Cicada also only the male can produce the well-
known shrill sounds. The sound apparatus is here somewhat com-
plicated. It consists of a pair of drum-skins (thin elastic extensions
of the cuticular skeleton) on the first abdominal segment, and of
strong muscles moving these skins. The abdomen filled with air acts
as a resounding apparatus. As a protection to the delicate drum-skins
folds of the thoracic and abdominal cuticular skeleton arch over them
from before and behind.

The sounds produced" by male insects are calls for attracting the
females.

XL Sexual Organs.

The sexes are separate in all Antennata. A comparative study of
the sexual organs justifies us in giving the following general plan of
the sexual apparatus. It consists of a pair of germ glands (ovaries
in the female, testes in the male) which pass into paired duets, the
latter opening separately. The sexual glands and ducts appear,
as far as their ontogeny is known, to proceed from a paired mesoder-
mal genital rudiment. Ectodermal invaginations of the cuticle are
often connected with the ends of the ducts.

Since the ducts of the sexual organs in the Protracheata are trans-
formed nephridia, we may perhaps infer the same for those of the
Antennata. There is, however, a considerable difference in the two
cases, as the greater part of the ducts in Peripatus arise out of the
ectoderm, while in the Antennata, on the contrary, they come from
the mesoderm. But it must not be forgotten that in the Annulata the
greater part of the nephridium (nephridial duct) is of mesodermal origin

Sexual organs, which are paired throughout (as in the plan above
sketched), are only found in reality in the Ephemeridce (Fig. 344, A).
In all other Antennata there are unpaired portions of the sexual appar-
atus arising in various ways.

1. The two germ glands may fuse to form 1 unpaired germ gland,
while the ducts remain separate either throughout their whole length
or at any rate towards their ends, and always open externally through
separate paired apertures. Such cases are found in the Diplopoda
among the Myriapoda.

2. The germ glands remain paired. The ducts remain paired for
the greater part of their course, only uniting to form a common
terminal portion. This is the case in all Antennata except the
Ephemeridce and the Diplopoda.

VI

ANTENNATA— SEXUAL ORGANS

485

In rare cases (Scolopendra) we find an unpaired germ gland and an unpaired duct,
into whose end, however, paired accessory organs (glands, receptacula seminis, vesiculae
seminales) enter.

The unpaired portion may arise in very various ways.

A. In the male of certain Forficulidce (Dermaptera) the ducts unite
at one part of their course to form an unpaired sperm vesicle, from
which the two ducts again as vasa deferentia separately run towards
the two male genital apertures. One of these ducts, however, becomes
reduced (Fig. 344, B\ while the other, as an unpaired ductus ejacula-
torius, runs from the sperm vesicle to the outer aperture.

B. An unpaired invagination of the integument grows from without
to meet the two ducts, so that these open externally as through an

FIG. 344.— A-F, Diagrammatic representation of the sexual apparatus of various Insects.
A-E, Male organs. F, Female apparatus. The parts proceeding from invagination of the outer
integument are indicated by thick black lines. A, Ephemerid. B, Forficula auricularia. C,
Larva of Orthoptera. D, (Edipoda (belonging to the Acridiidce). E, Cetonia aurata (Coleoptera).
F, ^schna (Libdlulid) (after Palmen).

unpaired terminal portion lined with a chitinous cuticle (Fig. 344,
C-F). The 2 ducts may open into the unpaired terminal section either
by two separate apertures or by a single aperture. This arrangement
is found in the Apterygota, also in the Libellulidce, Plecoptera, Orthoptera,
fihynchota, and perhaps in other orders as well. Paired or unpaired
accessory structures may appear in the terminal portion through
secondary invaginations.

C. A second unpaired section is sometimes added to the unpaired
terminal portion described under B ; this section arises by the fusing
of the two sperm ducts or oviducts throughout tracts of varying length
to form one unpaired duct (ductus ejaculatorius, uterus, vagina, etc.),
this unpaired duct enters the ectodermal invagination. This is
probably the case in all the so-called higher insects (insects with com-
plete metamorphosis).

The position of the outer genital apertures is very various.

In the Chilopoda among the Myriapoda the unpaired genital aperture
lies in the penultimate body segment, i.e. in the genital segment, whose
limbs may be transformed into genital appendages.

486

COMPARATIVE ANATOMY

CHAP.

Iii the Diplopoda and Pauropoihi the two genital apertures lie behind
the 2d pair of legs, generally at the boundary between the 2d and 3d
trunk segments.

In the Syinphyla (Scolopendrella) the unpaired genital aperture lies
on the -4th trunk segment between the le^s of this segment.

Fn;. 34t'..— .1, Female. ]'>, Male sexual
organs of Glomeris marginata (after
Fabre). os, Opened ovarial sac, in which
Hie two ovaries (uv) are seen; od, oviduct;
/, trstes; tji-rf, common vas deferens; pa,
paired ducts.

Fie. ?,\->.— Inner male sexual organs. A. Of Molophagus ovinus. J:, Of Acheta campestris.
C. Of Melolontha vulgaris (after Carus and Gegenbaur). /, Testes ; rd, vas deferens ; o1, si-niiiial
vesicle; <l<', duct us ejaculatorius ; ;//, accessory Clauds.

Iii iill J/t'.i'djiu(ht the genital apei'tui-es lie at the end of the abdomen,
in the male almost always behind the Oth, in the female behind the 8th
(in the Kphemeridw 7th) abdominal segment.

Accessory organs are almost always found connected with the
ducts of the male and female sexual apparatus of all Antcnnata ; the

VI

ANTENNATA— SEXUAL ORGANS

487

special arrangement of these is very varied. Single or paired sperm
vesicles (vesieulse seminalis), serving as sperm reservoirs, are often
found in the male sex, either as invaginations of the ductus
ejaculatorius or of the vas deferens. Accessory glands enter
either the ductus or further back enter the vas deferens and mix a
secretion with the sperm. Such glands occasionally yield a hardening
secretion which encloses small masses of sperm in the form of capsules

FIG. 347.—;!, Female. B, Male sexual apparatus of the Honey-bee (queen and drone) (after
R. Leuckart). ov, Ovaries, consisting of numerous chambered ovarian tubes ; vd, oviducts ; rs, re-
ceptaculum seminis ; va, vagina ; nva, accessory sac of the same ; 7cs, bulb of the stinging apparatus ;
md, rectum twisted back and cut off; sd, colleterial gland ; gd, poison glands ; gb, poison vesicle ;
t, testes ; vd, vas deferens ; e, wider portion of the same ; de, common ductus ejaculatorius ; ad,
accessory glands ; p, penis.

(spermatophores). The terminal section of the male sexual apparatus
is often protrusible as a penis.

Special invaginations from the vagina serve as bursse copulatriees
for the reception of the penis during copulation, and as reeeptaeula
seminis for the reception and the preservation of the semen. In the
Lepidoptera (Fig. 348) the bursa copulatrix opens outward separately
from the vagina, but is connected with the receptaculum seminis by a
duct. The sexual apparatus often enters, close to the anus, the base of
a common depression (cloaca). As in the male, so in the female, outer
organs are formed by the integument of the last abdominal segments, and
are brought into the service of the sexual apparatus as ovipositors, etc.

Colleterial or cement glands for attaching the eggs to foreign
objects enter the vagina.

In most Diplopoda among the Myriapoda the legs of the 7th trunk
ring are transformed into copulatory appendages.

The germ glands of the Hexapoda have still to be specially considered. The
testes are, as has been already mentioned, almost always paired ; in the Lepidoptera

488

COMPARATIVE ANATOMY

CHAP.

alone they may fuse to form an apparently
unpaired organ. Each testis consists of a
smaller or greater number of blind tubes
or testicle follicles; these, which are
sometimes short, sometimes long and
coiled, usually lie embedded in a common
envelope. It may be accepted as a general
rule, that where the testicle tubes are
very long, they are few in number, and
vice versa. The Diptera and Orthoptera
may serve as examples of the two ex-
tremes. In the former (Fig. 345, A)
there is on each side only one very long
coiled testicle tube, in the latter there
are often many hundred short tubes or
follicles united into one mass (6). The
testis on each side also may fall into
separate tubes or into separate groups of
tubes (c). Only one vas deferens is,
however, found on each side, into this
enter the testicle tubes united into one
testis, or the separate testicle tubes, or
the ducts of several groups of such tubes.
Just as the testes consist of a varying
number of long or short testicle tubes,
so each of the ovaries consists of a vary-
ing number of ovarian tubes, which
together enter an oviduct. In each of
these ovarian tubes (Fig. 349) we can
distinguish three parts : (1) the terminal
filament, (2) the terminal chamber, and
(3) the actual ovarian tube, divided into
chambers, this last is the largest.

FIG. 348.— Danais archippus (Lepidoptera)
Female. Showing the anatomy after removal
of the right half of the body. Lettering of the
head : «, Antenna ; ph, pharynx ; pi, labial
palps ; r, proboscis ; g, brain ; usg, infra ceso-
phageal ganglion. Lettering of the thorax:
I, II, III, thoracic segments; bi, 62» &3> the
coxal joints of the 3 pairs of legs ; bm, mus-
culature ; ac, aorta cephalica with its swelling ;
oe, oesophagus ; l>g, thoracic ganglia of the
ventral chord ; sd, salivary glands of one side,
those of the other side cut off near their entrance
into the common salivary duct. Lettering of
the abdomen : 1-9, abdominal segments ; h,
heart; sm, so-called sucking stomach (food re-
servoir); cm, chtlific stomach (mid -gut); a#,
abdominal ganglia ; ed, hind-gut with colon (c)
and rectum (r); vm, Malpighian vessels; ov,
ovarial tubes, those of the right side cut off;
ove, terminal filaments of the ovaries ; be, bursa
copulatrix ; obc, its outer aperture ; od, oviduct ; vag, vagina ; wo, its outer aperture : ad, glandular
appendages of the vagina partly cut away ; vkt connective canal between vagina and bursa copulatrix
with swelling (receptaculum seminis) ; an, anus (after Surges).

wo

CUV

VI

ANTENNATA— SEXUAL ORGANS

489

A

The thin terminal filaments of the egg tubes are generally attached to or near
to the dorsal vessel, and thus form a sort of suspender. The elements of these are
the same as those of the terminal chamber. The latter contains undifferentiated
cell elements as remains of the ovarial rudiments, from which proceed (either in
the embryo or the larva), on the one hand, the follicle epithelium of the ovarian
tubes, and on the other, the ripening eggs and nutritive cells contained in these
tubes. In the terminal chamber these
cell elements remain undifferentiated,
excepting when required for the renewal
of the follicle epithelium, eggs, and
nutritive cells in the adult insect. The
third section which leads into the oviduct
has usually the form of a string of beads.
It contains the ripening eggs. The
youngest and smallest lie nearest to the
terminal chamber, the oldest and largest
near the entrance into the oviduct. Two
sorts of egg tubes have been distin-
guished : those without nutritive cells
and those with nutritive cells. The
most simple are the ovarian tubes with-
out nutritive cells (Fig. 349, A), which
are found e.g. in the Orihoptera and
Apterygota (excluding Campodea). Here
in each tube there is a simple row of
eggs from the terminal chamber to the
oviduct. Between these consecutive eggs
the tube appears constricted, and this
causes the beaded appearance. Those
parts of the egg tube which lie between
two constrictions, each of which contains
one egg, are called ovarian chambers.
In the ovarian tube with nutritive cells
we must again distinguish two different
types. In the one type (Fig. 349, £)
there is a regular alternation of egg
chambers and nutritive chambers, each
of the latter containing one or more nutri-
tive cells, which serve for the nourishment
of the ripening egg contained in the
neighbouring chamber. The ovarian and
nutritive chambers may be distinctly separated externally by constrictions (Hymen-
optera and many Coleoptera}, or one nutritive and one egg chamber may lie in each
section of the ovarian tube, which is externally visible as a swelling (Lepidoptera,
Diptera). In the second type with nutritive cells, the actual tube consists (Fig. 349,
C) of ovarian chambers only, the nutritive cells here remain massed together in the
large terminal chamber. The single eggs in the tube are united with the terminal
chamber by connective strands, which convey the nutritive material to the eggs.

Egg cells, nutritive cells, and the cells of the follicle epithelium (epithelium of
the chambers of the ovarian tubes) are, according to their origin, similar elements,
like the egg and yolk cells of the Platodes ; division of labour leads to their later
differentiation. Only a few of the numerous egg germs develop into eggs, the rest
serving as envelopes and food for these few.

FIG. 349.— Various types of ovarian tubes,
diagrammatic. A, Ovarian tube without nutri-
tive cells. B, Egg tube with alternating nutritive
and egg compartments. C, Ovarian tubes in
which the terminal chamber (ek) is developed
into a nutritive chamber, with which the develop-
ing eggs remain connected by means of strands
(ds) ; ef, terminal filament ; ek, terminal chamber ;
efa, egg compartments or chambers ; fe, follicle
epithelium ; df, yolk chambers.

490 COMPARATIVE ANATOMY CHAP.

In a few Thysanura (Machilis, Lepisma, and especially lapyx) the ovarial tubes
(5-7 on each side) are placed in the abdomen in more or less strict segmental arrange-
ment. Each independently enters one of the two oviducts, which run through the
abdomen as straight canals. The two oviducts open externally by a short unpaired
terminal piece ; this common piece is said to be wanting in Machilis, only the outer
aperture of the two oviducts being in this case common to both. In Campodea and
the Collembola the ovaries and testes on each side are simple tubes.

XII. Dimorphism — Polymorphism.

In all insects the males and females differ, not only in the arrangement of their
sexual organs, but also in various details of their outer organisation. This sexual
dimorphism is in some cases very remarkable, and is generally connected with the
absence of wings in the female, e.g. in the scale insects (Coccidce), in the luminous
Goleoptera (Lampyridce), and in a few Bombycina (Psyche, Orgyia). In the parasitic
Strepsiptera the females are legless, wingless, eyeless, and feelerless, and are thus
like maggots. They are viviparous, and remain, as long as they live, enclosed in
their pupal envelopes, inside the host, i.e. in the abdomen of various Hymenoptera.

In colonial insects polymorphism arises in consequence of division of labour
between the individuals of the colony. In many colonial Hymenoptera (bees and
ants) only a few of the females (queens) are sexually mature and capable of re-
production. The great majority of the other females (workers) have reduced sexual
organs, and, among the ants, are wingless. Among the ants different forms of
workers may appear (soldiers and workers proper). In the colonial Termites, and also
among the Corrodentia, there are, besides the winged reproductive males and females,
unwinged males and females with rudimentary sexual organs ; these are again
divided into castes (workers, fighters), and vary in form accordingly.

XIII. Development and Life-History.
A. The Metamorphoses of Insects.

The greater part of the developmental processes, by means of which the adult
insect is produced from the fertilised egg, take place within the egg -envelope.
The time passed within the egg is the period of embryonic development.
The organism which escapes from the egg-envelope, or in other words, is hatched
from the egg, is always already very highly developed, and, apart from the
fact that it has no wings, .no mature sexual organs, and no compound eyes, 'is
provided with all the typical insect organs in functional condition. It possesses a
completely segmented body, antennae, mouth parts, thoracic limbs, developed nervous,
digestive, and tracheal systems, the dorsal vessel, the body musculature, etc. It
moves and feeds itself independently. It is called a larva. The larvae of the insects
are thus, when hatched, far more highly developed than the larvae of most other
Invertebrate.

The changes which an insect undergoes before it becomes an adult sexually
mature animal (imago) are extraordinarily various, and are conditioned by a whole
series of co-operating factors, among which the most important are : (1) the degree
of deviation of the imaginal form from the racial form ; (2) the different modes of
life and of places of habitation of the larvae and imagines.

I. The Apterygota (Thysanura and Collembola) are considered to have retained
their original wingless condition, and in other ways also appear to stand nearest to

vi HEXAPODA— METAMORPHOSES 491

the common racial form of the Insccta. The distinction between larva and imago is
here wanting. The young animal hatched from the egg resembles in all points the
sexually mature form, which it reaches by simple growth accompanied by ecdyses and
by complete development of the sexual organs. Both young and adult animals live
on land and lead the same sort of life. Development without metamorphosis
(Ametabole).

II. The adult insect is, apart from the complete development of the sexual
organs, principally distinguished by the possession of wings. In the simplest cases,
in the Orthoptera, Corrodentia, Thysanoptera, and most Rhynchota, the larvae lead
the same kind of life as the imagines. They change gradually into the imaginal
form, growing slowly through numerous ecdyses, while the wings arise and become
more strongly developed each time the integument is shed. Gradual meta-
morphosis.

III. In the Cicada the modes of life of the imagines and the larvse differ. The
former live on trees and shrubs, the latter underground on roots, and for this purpose
possess strong fore-feet adapted for digging. The transition from the last larval stage
to the imago must here be accompanied by a transformation of the fore-legs. Since
an intermediate life between that on trees and on the earth is not easily conceivable,
and since, consequently, any intermediate form between ordinary feet and digging
feet would be purposeless, the transition from the larva to the imago has become
direct. The last larval stage is then what is called quiescent, i.e. the organisation
of the imago develops within the chrysalis at the expense of the accumulated reserve
material. Gradual metamorphosis with pupal stage.

IV. The modes of life of the larvse and imagines of the Ephemeridce, Odonata,
and Pleeoptera are very different. The imagines live on land, the larvse have
become adapted to aquatic life. In the transition to the imaginal form the tracheal
gills are generally thrown off (Ephemeridce and many Libellulidce), the stigmata
break through, and the tracheal system becomes holopneustic. In other forms the
metamorphosis and the growth of the larvse occur gradually (in Chloe by means of
more than 20 ecdyses). Incomplete metamorphosis (Hemimetabole).

In cases II. III. and IV. the transformation of the larva into the imago is, as a
rule, accomplished very gradually. The still wingless larva hatched from the egg
shows an external organisation like that which the Apteryyota possess throughout life,
even in their sexually mature condition.

V. Some Rhynchota are wingless (the Pediculidce, many bugs and the females of
the plant-lice), and so are some- Corrodentia (the Mallophaga), Orthoptera (various
genera and species of Blattidoe and Phasmidaz), and Dermaptera. The wingless
condition in these forms is derived, as opposed to that of the Apterygota ;
they are descended from wing -bearing Rhynchota, Corrodentia, Orthoptera, and
Dermaptera, but their wings have been reduced in consequence of parasitism or of
other habits of life. The larva hatched from the egg has not, therefore, to develop
wings. This process is suppressed, and with it the metamorphosis. The young
form becomes, by simple growth, the sexually mature imago. In contrast to the
original ametabole of the Apterygota we have here an acquired ametabole.

VI. In contradistinction to the insects as yet mentioned, all others, i.e. the
Neuroptera, Panorpata, Trichoptera, Lepidoptera, Diptera, Siphonaptera, Coleoptera,
and Hymenoptera, are distinguished by so-called complete^metamorphosis (holometa-
bole). A wingless larva is hatched from the egg, which indeed grows and moults,
but, nevertheless, always retains the same organisation and undergoes no trans-
formation during the larval stage. At the end of larval life, however, when the fat
body has become strongly developed by rich nourishment, the larva moults and
passes over into the differently formed pupal stage. The pupse are very variously
shaped, often distinctly segmented with rudimentary extremities and wings, often

492 COMPARATIVE ANATOMY CHAP.

with concealed extremities. They are generally quiescent, immovable, and not
capable of taking food, and are often surrounded by special protective envelopes —
cocoons. The best known cocoons are those of the Lepidoptera, which are spun by
the caterpillars from the secretion of their spinning glands. At the end of pupal
life the envelope is opened and the imaginal insect (beetle, butterfly, fly, etc.) issues
from it.

This complete metamo'rphosis evidently proceeds from an incomplete or gradual
metamorphosis. The wingless larva adapts itself to various conditions of existence,
or is provided by its parents with excessive nourishment. It accumulates so much
reserve material in its fat body that the further larval stages do not need to feed
independently. By the suppression of numerous ecdyses these successive larval
stages are abbreviated into one stage, — the stage of the outwardly quiescent pupa,
within which the imago attains development at the expense of the reserve nourish-
ment.

The larvae of insects with complete metamorphosis vary much in their organi-
sation, each being adapted to its own special surroundings. Two principal groups
can, however, be distinguished : (1) with feet, e.g. the larvae of the Neuroptera, the
"caterpillars" of the Lepidoptera, the larvae of the Coleoptera and Trichoptera ; (2)
without feet, maggot -like larvae of the Diptera, larvae of most Hymenoptera and
Siphonaptera. The former by possessing legs are the least removed from the
Thysanura-like larval form of other winged insects ; they move freely, and with few
exceptions feed independently. Many are carnivorous, living either on land or in
water ; many feed on plants, on the leaves (caterpillars) or roots (cockchafer grubs).
Among the Hymenoptera the larvae of the Tenthredinidce are vermiform, and like
the butterfly caterpillars possess parapodia - like appendages on several abdominal
segments in addition to the thoracic feet.

The modes of life of the footless larvae of the Diptera and most Hymenoptera are
very varied. They sometimes live free and are carnivorous, generally living then in
water ; sometimes they are parasitic in the bodies of other animals or in plant tissue ;
sometimes in putrefying matter, dung, etc. ; sometimes inside cases or cells which are
filled with nutritive material ; sometime they are fed by the adults, etc. etc. Head-
less maggots without feelers or ocelli and with reduced mouth parts are distinguished
from larvae which have heads with these organs.

The larvae of insects with complete metamorphosis are all originally peripneustic.
By adaptation to aquatic or parasitic life they may become amphipneustic, meta-
pneustic, or even apneustic, and in the last case may develop trachaeal gills. The
mouth parts of the larva may differ greatly from those of the imago. This
difference is best known and most striking in the Lepidoptera, whose larvae have
masticating mouth parts, while their imagines have sucking mouth parts.

The more specialised the larva on the one hand and the imago on the other, and
the greater the difference in organisation between them, the more far-reaching natu-
rally are the transformations by means of which during the pupal period the larval
organisation becomes that of the imago. For instance, in the bee, the larva does
not pass direct into the pupal stage, but first into the pre-pupal stage.

In certain Coleoptera several larval stages differing very much from one another are
met with. The Coleopteran genus Sitaris (Fam. Meloidea) lives parasitically on a
bee (Anthophora). The larvae of this beetle, which are hatched from the egg, are
active little animals with thoracic legs. They lurk in flowers in order to spring
upon the bees coming to gather honey. They are thus carried to the hive, where
they seize upon the eggs of the bee as soon as these are laid in the honey of the
cells, and devour them. They afterwards moult and appear, after ecdysis, as meta-
pneustic maggot-like larvae with reduced feet, floating on the surface of the honey,
the mouth placed below, and the posterior end on the surface. When the honey is

vi HEXAPODA— EMBRYONIC DEVELOPMENT 493

exhausted they pass into a pupa-like stage, out of which, however, not the imago,
but a new larva emerges. Still further pupa-like stages of development then follow,
till at last the final real pupa stage occurs. Here we can very clearly recognise,
especially in the first two larval stages, the influence of various modes of life on the
larvae of one and the same animal.

Many Pteromalidce (Hymenoptera) pass through a series of peculiarly shaped larval
stages, which are as yet by no means explained. The larvse live parasitically in the
eggs, larvse, and pupse of other insects, in which the Pteromalidce lay their eggs by
means of an ovipositor. It is remarkable that the youngest larvse possess far less
highly developed inner organs than are usually found in the larvse of other insects.

The above is naturally but a very incomplete description of this most interesting
subject.

B. The Embryonic Development of Insects.

Hydrophilus, the water beetle, affords us a good illustration of this development.

The egg is a long oval, with pointed anterior and blunt posterior pole. The
segmentation is that typical of the centrolecithal eggs, and leads to the formation of
a blastosphere. In this blastosphere we can distinguish a single superficial layer of
small cells, the blastoderm, and, enveloped by it, the nutritive yolk with scattered
nuclei.

The formation of the embryo proceeds from one side of the blastosphere only, i. e.
from the future ventral side, on which the blastoderm cells are higher than elsewhere.
We may call this portion of the blastoderm the embryonic rudiment. At an early
stage we can distinguish the boundaries of the segments, appearing externally as
transverse streaks or lines. Anteriorly and posteriorly two longitudinal furrows
appear, grow towards each other and unite, so as to mark off on the embryonic
rudiment a peripheral portion, the lateral plates, from a central portion, the middle
plate. The middle plate sinks below the surface, and so forms the floor of a channel-
like medio-ventral invagination, whose edges grow towards each other on each side.
This invagination is represented in transverse section in Fig. 350, A- Its edge is to
be considered as the edge of the blastopore. How the lateral edges of this blastopore
approach each other in the middle line and finally fuse with each other is illustrated
by Fig. 352, A, B, C. After the closing of the blastopore the invagination becomes
a medio-ventral longitudinal tube, over which the blastoderm of the former lateral
plates spreads. From this invaginated tube proceed the mesoderm and perhaps also
the endoderm (epithelium of the mid-gut).

Even before the closing of the blastopore the rudiments of the embryonic en-
velopes so common among insects, viz. the amnion and the serous envelope, appear.
They first appear as a fold of the blastoderm at the edge of the embryonic rudiment.
This fold grows more and more from all sides over the embryonic rudiment,
finally covering it. The embryonic rudiment thus comes to lie in the base of a cavity
whose mouth, wide open at first, grows smaller and smaller by the growing and final
closing together of the amnion folds ; the final closing takes place over the anterior
end of the embryonic rudiment. The tranverse section B, Fig. 350, shows the rising
amnion folds, the transverse section C shows them grown over the embryonic
rudiment so as to form a continuous cover. In the surface views of Fig. 352 the
folds are denoted by af and of".

The cavity which is formed by the amnion is called the amnion cavity. Its roof,
in correspondence with its origin, consists of 2 epithelial lamellae, an inner one, which
at the edge of the embryonic rudiment is continued into its blastoderm and represents
the actual amnion, and an outer one, which at the edge of the embryonic rudiment
is continued into the blastoderm of the whole remaining surface of the egg, and

494

COMPARATIVE ANATOMY

CHAP.

3

alo « .-i.s

S M * £ <§ 3

o g ^ * g ^ •&

^illsll

•z-Z^
Kl

VI

HEXAPODA— EMBRYONIC DEVELOPMENT

495

together with this represents an unbroken epithelial membrane, the serous envelope ;
this latter surrounds the whole egg with its embryonic rudiment and the amnion on
all sides.

The amnion and the serous envelope have no share in the building up of the
embryo. The latter develops exclusively out of the blastoderm of the embryonic
rudiment and the invaginated tube, which we will call the germ streak. The blasto-
derm of the embryonic rudiment grows further and further dorsally at its peripheral
edges, so that at last it envelops the embryo on all sides as ectoderm. Though
somewhat out of its strict order, for the sake of clearness, this process as well as the
fate of the amnion and serosa may be here illustrated by means of the following dia-
gramatio transverse sections. Fig. 351, A, represents the same stage as Fig. 350, F,

FIG. 351.— Formation of the dorsal tube (process of involution of the embryonic integuments)
in Hydrophilus (after Graber and Kowalevsky). A, Transverse section through an egg, whose em-
bryonic rudiment is still covered by the amnion (a) and serosa (s). B, Amnion and serosa having
grown together in the middle line have now torn open, drawing back on each side to form a fold.
C, The fold, by the contraction of the serosa becomes more dorsal. D, The contracted serosa (dorsal
plate) is being grown over by the fold. E, The dorsal tube has become closed by the growing
together of the folds. F, The mid-gut is closed dorsally and has enclosed the dorsal tube (s).
a, Amnion ; d, nutritive yolk ; h, heart ; I, ccelome ; m, rudiments of mid-gut ; n, nervous system ;
s, serosa and the structures developed out of it, i.e. the dorsal plate and dorsal tube ; tr, principal
tracheal trunk ; ec, ectoderm.

but the position of the embryonic rudiment is reversed. In Fig. 351, B, we see the
amnion and the serous envelope torn in the ventral middle line, after they had
previously grown together. The amnion and the serous envelope thus form on each
side a fold projecting ventrally. In 0 the serosa has contracted and has become the
so-called dorsal plate, which now consists of high cylindrical epithelium. At the
same time the ectoderm of the embryonic rudiments on both sides has extended

496 COMPARATIVE ANATOMY CHAP.

further dorsally. In D the fold (which proceeded from the amnion and a part
of the serous envelope) has bent back and the dorsal plate has contracted still
more. In E the folds have overgrown the dorsal plate and their edges have grown
together in the middle line. By this process the so-called dorsal tube comes into
existence, and then sinks into the yolk. The ectoderm then joins over the dorsal
side of the embryo. The tube of the mid-gut enclosing the yolk, together with the
dorsal tube, is formed by the complete surrounding of the yolk by the endoderm.
The dorsal tube is then broken up with the yolk and reabsorbed within the mid-gut.
In other insects these processes take the same course. The most important
difference is caused by the fact that from the first the yolk penetrates between the
amnion and the serosa, so that the embryo with the amnion seems to be imbedded
deep in the yolk. The amnion remains connected with the serosa at one place only.
At the place where the two membranes adhere a rent arises later, through which the
embryo together with the amnion are everted. In the Lepidoptera no dorsal tube is
formed. The embryonic integuments are here simply constricted off from the embryo
and serve as the first food of the young grub. This is perhaps also the case in the
Diptera and Hymenoptera.

To return now to the developmental processes in the embryonic rudiment itself,
we must go back to Fig. 350, B. The blastopore here appears closed. The invagin-
ated tube (germ streak) is compressed dorso-ventrally and has a slit-like lumen. The
transverse section C of a still later stage shows us the germ-streak spread out flat
under the ectoderm. On each side of the middle line the latter is thickened and
bilaminar. The thickened parts are transverse sections of longitudinal thickenings
(primitive thickenings), between which there is a shallow median longitudinal groove
(primitive groove). The deeper cells of the longitudinal thickenings form the two
lateral strands, from which come the paired portions of the ventral chord. The so-
called yolk furrowing has taken place at this stage, the yolk belonging to each
nucleus being marked off, so that the whole is divided into irregular masses.

In the transverse section D we see the germ streak divided into two lateral halves,
a cavity appearing in each half ; these cavities become those of the primitive
segments. These appear more clearly demarcated in the transverse section E of
an older stage. They are the mesoderm cavities, which are repeated segmentally on
each side. The remaining mesoderm of the germ streak has again united in the
middle line. On each side near the primitive thickenings the ectoderm becomes in-
vaginated to form a trachea. These tracheal rudiments appear segmentally in pairs,
as is shown in the ventral surface view, Fig. 353, where their outer apertures
(stigmata) are seen.

In the transverse section F of a later stage the germ streak has drawn back some-
what from the surface of the yolk, and so gives rise to a cavity, which becomes the
definitive coelome and later joins the segmental cavities. The longitudinal trunks of
the tracheae are already formed, as shown in transverse section. The lateral strands
of the nervous system have separated from the ectoderm (hypodermis), and the
primitive groove between them has deepened. The fusing of its base with the lateral
strands yields the transverse commissures of the ventral chord. We find in each
side between the yolk and the segmental cavities a newly formed layer of cells. This
cell layer represents the endoderm. Gradually extending all over the surface of the
yolk, it becomes the epithelium of the mid-gut which encloses the yolk. The yolk
is gradually absorbed later. The wall of the mesoderm cavities which is in contact
with the endoderm follows the latter, as it grows over the yolk and yields, as the
visceral layer of the mesoderm, the muscular wall of the mid-gut.

Long before all these processes have taken place the stomodseum has formed at
the anterior (head) end of the embryonic rudiment and at the posterior end (in the
terminal segment) the proctodaeum, both being ectodermal invaginations which

vi HEXAPODA— EMBRYONIC DEVELOPMENT 497

become connected with the mid-gut at a later stage. The Malpighian vessels arise
as imaginations from the ectodermal proctodseum.

The limbs (Fig. 353) appear as paired bud-like outgrowths of the ectoderm and
the subjacent mesoderm ; they appear between the primitive thickenings of the nerve
chord and the stigmata.

It is an important fact that besides the rudiments of the head and thoracic limbs,
which alone are present in the adult animals, there occur at a certain stage distinct
rudiments of abdominal limbs as well, those of the first abdominal segment being
particularly clear (cp. especially Fig. 353, J5). These rudiments degenerate later.

The origin of the two endodermal cell layers, which we saw appear in the
transverse section (Fig. 350, F), is difficult to explain. At the stage at which this
section is taken, the whole endodermal rudiment consists of two lateral cell streaks

A ^

I • %

*Jt

p* /

af

Fig. 352.— A-E, Ventral view of 5 stages in the development of Hydrophilus (after Heider).
The anterior end is directed upward, a and b, Points at which the blastopore first closes ; af, edge of
the amnion fold ; af, caudal fold ; af", paired cephalic fold ; an, antenna ; es, terminal segment ;
g, pit-like invagination (rudiment of the amnion cavity) ; fc, procephalic lobes ; r, groove-like
medio-ventral invagination ; s, germ-streak covered by the amnion.

between the yolk and the two rows of mesoderm cavities. Anteriorly at the
stomodseum and posteriorly at the proctodseum the two streaks pass into one
another. At earlier stages the endoderm consists of an anterior and a posterior
U-shaped double streak. The limbs of the U, which are very short to begin with,
are at first directed backward in the anterior double streak, and forward in the
posterior streak ; they gradually elongate till at last the two anterior limbs meet the
two posterior, and thus give rise to the two above-mentioned endodermal streaks.
The first origin of the two U-shaped double streaks must probably be sought in the
anterior and posterior ends of the median invagination of the blastoderm of the
embryonic rudiments, which elsewhere forms only mesoderm. The whole invagina-
tion would thus be the rudiment of the mesoderm plus the endoderm, and might
be regarded as the gastrula invagination. In a transverse section of the still open
anterior part of the invagination its base would represent the formative zone of the
U-shaped endodermal streaks, while the lateral somewhat bulged-out walls yield
mesoderm. The mesodermal germ streak would thus be continued anteriorly in two
lateral bulgings of a gastrula invagination, arid we are therefore inclined to regard
the mesodermal formation of insects as a modification of that manner of forming the
mesoderm in which it proceeds from paired invaginations of the archenteron.

According to another view, the yolk with its nuclei represents the endoderm, and
VOL. I 2 K

498

COMPARATIVE ANATOMY

CHAP.

yields the epithelium of the mid-gut in a way which has not been more closely
observed. Other observers, again, maintain that the whole enteric epithelium is
formed from the proctodseum and the stomodseum.

It would perhaps be worth while to investigate whether the endoderm does not
arise from the yolk cells by a kind of micromere formation, in a way similar to that
in which the ectoderm is produced in many animals whose eggs contain much
nutritive yolk (cp. p. 124, etc.)

The heart arises out of two lateral originally widely separated rows of mesoderni

B

Fig. 353.— A and J5, Hydrophilus embryos with the rudiments of the extremities (after
Holder). In the somewhat older embryo, J5, the rudiments of abdominal feet, which disappear
later, are distinctly seen, a, Anal aperture ; an, antenna ; g, rudiment of the ventral ganglionic
chain ; m, oral aperture ; md, mandible ; mx\, first maxilla ; mx-2, second maxilla (rudiment of the
lower lip) ; pi, p*, jps, thoracic legs ; p±, p^y p?, p§, rudiments of the extremities of the first, second,
fourth, and sixth abdominal segments ; st, stigmata ; vk, procephalon.

cells. Each row, by sending out muscular processes, forms the side of a groove. The
two sides move towards each other, uniting later in the middle line to form the
cardial tube.

The brain arises, like the ventral chord, as two lateral ectodermal thickenings
(neural plates), which remain separate from each other for a considerable time, but
from their first appearance are continuous with the lateral strands of the ventral
chord. Besides this a second ectodermal invagination on each side takes part in its
formation.

Opinions as to the development of the eyes in insects still differ very much.
According to recent observations it appears probable that both the compound and the
simple eyes proceed from invaginations of the ectoderm, which become constricted off
as vesicles, and only secondarily become connected with the optic portion of the brain
(ganglion opticum).

The first appearance of the sexual organs is not yet sufficiently investigated.

vi HEXAPODA— METAMORPHOSES OF INSECTS 499

C. The Inner Processes in the Metamorphoses of Insects.

In insects without metamorphosis or with gradual or incomplete metamorphosis
the organs of the larva simply pass into those of the imago. There is no breaking up
and disappearance of the larval organs, and no new formation of the imaginal organs,
if we leave out of account the throwing off of tracheal gills and the formation of
wings, compound eyes, and so on.

In insects with complete metamorphosis the case is different. The larval organisa-
tion has here been adapted, independently of the imago, to special conditions of
existence. A gradual continuous transformation of the larval organs into the often
quite differently constructed organs of the imago, during which the different stages
would feed independently, is inconceivable, since organs undergoing such complete
transformations could hardly be capable of functioning. There are, further, numerous
phenomena in the most various divisions of the animal kingdom which prove that
organs which have functioned actively during larval life are only slightly capable
of development and metamorphosis. They are more often reabsorbed or thrown off
in the further course of development. We thus see why in insects with complete
metamorphosis the transition from the larva to the imago almost necessarily takes
place during a pupal stage, the pupa changing into the imago partly at the expense
of the reserve nourishment accumulated by the larva, it being unable to acquire
food for itself.

So as to understand the inner processes in the metamorphosis of the holometabo-
listic insect, we will take as an example the larva of Corethra plumicornis (Diptera,
Tipularia). Most of the larval organs here simply pass
during metamorphosis into those of the pupa and of the
imago. The larva, however, is footless and wingless.
The rudiments of feet and wings form shortly before the
pupal stage. Three pairs of ventral and 3 pairs of dorsal
invaginations of the hypodermis appear and are called
imaginal discs. In the bases of these invaginations out-
growths appear and grow continuously longer, while the
invagination in which they lie deepens (Fig. 354). The
outgrowths in the 6 ventral invaginations are the rudi-
ments of the thoracic limbs, the outgrowths of the 2
posterior dorsal pairs of imaginal discs are the rudiments
of the wings and halteres which thus lie hidden within Fig. 354. — Rudiments 01
the body, till they are protruded and attain development. *ne imaginal discs in the

The muscles of the wings are already rudimentarily larva of Corethra, diagram-

, ,, i i - .LI i matte. Invaginations (fe and

present m the embryo as cell strands, but they only ^ of the la°rval hypjdermis

begin to differentiate at the end of larval life. (ihy), in whose bases the rudi-

The complete metamorphosis of some insects, especi- ments of wings (fa) and legs
ally of the Muscidce (e.g. Musca vomitoria), is accom- <fca) arise5 ^, chitinous in-
panied by far more thorough transformations.

It must, first of all, be pointed out that the distinction between larval, pupal, and
imaginal stages rests upon external phenomena. In the inner organisation the series
of alterations is continuous ; the larva on the one hand already possesses the rudi-
ments of the imaginal organs, and in the pupa on the other, the larval organs only
gradually disappear. Speaking generally the inner metamorphosis is such that the
imaginal organs proceed out of parts of corresponding larval organs, which remain
during larval life in an undifferentiated embryonic condition (formative centres) ; the
portions which function during larval life gradually disappear during metamorphosis in
proportion as the imaginal parts attain development. The amoeboid blood corpuscles

C< LMPA If A T1VE A NA TOM Y

CHAP.

I'lay an important part in the breaking up and disappearance of the larval organs ;
since they, as phagocytes (leucocytes) seize upon the elements of the larval organs,
and like anni'luv take them into their protoplasm. The phagocytes thus laden with
the wrecks of the larval organs and floating in the hody cavity are themselves later
turned into food for the developing imaginal portions, especially for the epithelia,
into which they immigrate and break up.

At the end of larval life there are found in the thorax imaginal discs similar to
those in the Cur> thru larva. They here lie. however (Fig. 3f>">), much deeper in the body,
and are connected with the hypodermis liy means of long cell strands which are hollow
only in the neighbourhood ot'the imaginal discs. The thoracic limbs and wings begin

Fn;. 3.0'j. — .!, />', c, D, Diagrammatic representation of the development of the wings, legs, and
the imaginal hypodermis of the Muscidae from the imaginal discs of the larva during meta-
morphosis, diagrammatic transverse sections. ///, Chit incus integument of the larva, from which the
subjacent hypodermis (//<//) has withdrawn ; iid. imaginal discs ot'the wings ; uv, of the thoracic legs ;
•is, the strands connecting them with the hypod'ermis ; fl, wing rudiments ; l>, leg rudiments ; ihy,
imaginal hypodermis, spreading out in 1) from the imaginal discs. The imaginal rudiments of the
hvpodenais are indicated by thick black outlines, the larval hypodermis by two thin parallel
lines.

to form in just the same wav as in C'o/v7//m as outgrowths within the imaginal sacs.
At a later stage the processes of the imaginal discs which are connected with the
hypodermis shorten and become hollow. The larval hypodermis then opens over the
imaginal discs, which have moved outwards, and the feet and wings come freely
to the surface. A new hypodermis layer spreads out from the imaginal discs over
the thorax : this is the rudiment of the imaginal hypodermis, while in proportion as
the imaginal hypodermis spreads, the larval falls to pieces and disappears, in such a
way that the larval and imaginal hypodermis taken together at all stages of the
metamorphosis form a continuous cover round the body. AVhile in the thorax the
formation of the imaginal hypodermis proceeds from the imaginal discs, in the
abdomen it proceeds (later than in the thorax) from the formative centres, the so-
called islands, in the hypodermis. In each abdominal segment there are four larger

vi HEXAPODA— PARTHENOGENESIS— P^DOGENESIS 501

FIG. 356. — A and B, Diagrammatic representa-
tion of the formation of the imaginal hypodermis
in the abdomen of the Muscidae, proceeding from
the centres of the imaginal hypodermis (islands)
(hi). Ih, Larval hypodermis.

In trying to understand these metamorphosic processes we must always keep in
view that ecdysis only affects the chitinous cuticle of the body, from which the sub-
jacent hypodermis withdraws, secreting a new chitinous integument under the old one.

The formation of the head is very peculiar. It is (as oesophagus and optic vesicles
connected with the oesophagus) in-

vaginated into the thorax, and is later A

evaginated anteriorly out of the thorax
during the pupal stage. The anterior
part of the oesophagus becomes in this
process the neck, which after the
evagination of the head connects it
with the thorax.

Of the inner organs, the heart and
the rudiments of the sexual organs
seem to pass 'direct into the corre-
sponding organs of the pupa.

The whole musculature of the larva
except a few muscles of the second
thoracic segment disappears. The
imaginal connective tissue and the
greater part of the imaginal musculature
are formed anew from the mesoderm

elements, which early appear on the inner side of the imaginal discs. Perhaps the
imaginal discs themselves yield, besides the thoracic hypodermis, the mesodermal
layers belonging to them as well ; this point, however, needs further investigation.
Certain dorsal muscles of the second thoracic segment of the larva do not disappear,
but are transformed during a temporary loss of their transverse striation into the
wing muscles of the imago.

A large part of the larval tracheal system disappears. The imaginal tracheal
system seems to be regenerated out of scattered cells and cell groups of the larval
tracheal hypodermis. In the digestive tract the greater part of the mid-gut dis-
appears. The imaginal mid-gut forms anew out of persistent epithelial islands of the
larval mid -gut. Parts of the fore- and hind-guts proceed direct from the fore- and
hind-guts of the larva, while other parts arise out of circular islands or formative
centres, the so-called imaginal rings, of the larval fore- and hind-guts.

The central portion of the nervous system (brain and ventral chord) and prob-
ably also the beginnings of the larger peripheral nerves proceed by means of peculiar
alterations and transformations from the corresponding parts of the larva.

The salivary glands of the larva fall to pieces and disappear, falling victims to
the leucocytes. The imaginal salivary glands are regenerated out of imaginal rings
of the larval glands. The larval fat body is gradually devoured by the leucocytes.

As already pointed out, the disintegration of the larval organs and the new
formation of the imaginal organs do not belong to two distinct periods. Both pro-
cesses go on side by side, so that in general there is'no discontinuity either in the out-
ward form or in the structure of the organs. Physiological discontinuity prevails
only in this sense, that the organs cannot, during transformation, perform their
respective functions.

Parthenogenesis— Cyclic Reproduction— Paedogenesis.

Parthenogenesis occurs in many insects, and especially frequently in the plant
lice (Rhynchota) and in many Hymenoptera, though here also, most probably,
it is not the only method of reproduction, but merely alternates with reproduction

502 COMPARATIVE ANATOMY

by means of fertilised eggs. In the colonial J/t///«>/i(>pfsra only males come from
unfertilised eggs. In the Apliidex, in summer, several generations of partheno-
geneticaily reproducing, viviparous, generally wingless females succeed one another.
The last viviparous summer generation, however, produces winged males, and either
winged or wingless females, Avhose fertilised eggs remain through the winter.
From these latter the first summer generation of parthenogenetically reproducing
females is again produced. The reproductive cycle of F/it/Uoscra is similar, with
this distinction only, that all the generations are wingless, except that one out of
whose eggs the sexual (male and female) generation is produced. Phylloxera is
not viviparous.

The reproductive arrangements of Cltcnnrs are very peculiar ; the males of this
fir louse were, till recently, altogether unknown.

A wingless generation (I.) <rf Chermes ulieti* hibernates on the fir-tree, and in the
spring lays unfertilised eggs, out of which a second winged generation (II.) of females
is produced. Some of these females migrate from the fir to the larch. From their
unfertilised eggs a third wingless generation (111.) of females comes, which winter in
the larch, and in the spring of the second year lay unfertilised eggs, from which a
fourth winged generation (IV.) of females comes. These ily hack to the fir tree, and
from their unfertilised eggs a fifth generation (V.) of males and females is developed.
The sixth generation (VI.) which comes from the fertilised eggs of this generation,
again corresponds with the first hihernating generation with which we started.
Some of the second winged generation (II.) of females, however, remained on the lir-
tree. From the unfertilised eggs of these females which remained arises an unwinged
generation of females, and these, again, are succeeded hy a winged summer generation,
and so on. In this second parallel series of generations of Chcrmcs individuals
which remained on the lir-tree, two generations of females appear yearly, alternately
winged and wingless, Loth reproducing parthenogenetically. ISTow it is probable
that in this parallel series the generations do not thus reproduce (parthenogenetically)
oil infinitum, hut rather that, sooner or later, the parallel series re-enters the original
series, so that then a generation of males and females again appears. The different
generations dilier considerably from each other in form, even apart from the alternat-
ing ahsence or presence of wings.

The sijJu'fli'S allbrd an example of a kind of cyclic reproduction (heterogeny) in
which the parthenogenetically reproducing females are viviparous. The unfertilised
eggs here develop within the mother hody. A similar phenomenon occurs in the
JJi'l>tcr<i (Cccidomyia] also ; here, however, in the generation of females which repro-
duces parthenogenetically, the germarium which corresponds with the ovary becomes
mature very early, i.e.. in the larval stage. The unfertilised eggs are here developed
within the larval hodv ; thus, in the cycle of reproduction of Cccidomyia, an imaginal
generation reproducing sexually by means of fertilised eggs alternates with several
generations of parthenogenetically reproducing viviparous larva1. This special kind
of heterogeny is called Paedogenesis. In one species of ( '/u'ronomus also the pupa
may occasionally lay eggs, which develop just in the same way as the fertilised eggs
of the imago.

1). Development of the Myriapoda.

The embryonic development of the Miir!«p/>ilu. as far as it is known, does not
greatly dilier from that of Jn»<'<-hi. Kmbryoiiic envelopes, however, do not appear
to form.

AVhen the young Myrlapoda are hatched, they are either provided with the
definitive number of segments and legs, as is the case in the Hcohipr.iulrido' and
6Vo/,A/Y/V/</; ((_'/i f'/<>po>?(i j. or they possess a smaller number, to which the missing

VI

ANTENNATA—PHYLOGENY

503

ones are gradually added posteriorly during the many moults undergone by the

animal. The young of the Scutigeridce and Lithobiidce (Chilopodd) have 7 pairs

of legs. The number then increases to 15. The young Diplopoda ( Fig. 357), on

the contrary, have only 3 pairs of feet on the 3 anterior trunk segments, and a few

posterior segments still limbless. They

thus recall the type of the Insect larvte.

New segments gradually appear posteriorly

and the number of legs increases. After

each moult the number of rings is greater ;

the increase generally takes place very

irregularly, so that (e.g. Polydesmidce)

stages with 7, 9, 12, 15, 17, 18, 19, and

finally 20 rings, succeed one another.

From the above we see, firstly, that a sort

of metamorphosis takes place in many

Myriapoda, and, secondly, that the body ^O-^-L^-^ ^-V* — £~">-T — b,

there differentiates from before backward,

a point which can no longer be made out

in the Insecta.

XIY. Phylogeny.

Of the Antennata now living, the
Symphyla, perhaps, stand nearest the
common racial form. Yet even they are
one-sidedly developed, and many of their
organs, and above all, the tracheal system,
by no means show a primitive arrange-
ment. From the common racial form of
the Antennata, the Myriapoda branched
off to the one side, and the Hexapoda to
the other. The different orders of Myria-
poda perhaps developed polyphyletically,
while for all Hexapoda we can assume one
common racial form, resembling the now
living Apterygota, and especially the Thysanura. There is thus no special reason
for considering the Apterygota as originally winged insects, which became sexually
mature at progressively earlier stages of development, and finally at a larval stage.
At least one reason against such. a supposition is the occurrence of the protrusible
vesicles in the abdomen of the Thysanura, which is present in the Myriapoda
(Lysiopctalidce and Symphyla}, but almost entirely wanting in the Pterygota,
only one pair appearing temporarily on the foremost abdominal segment in the
embryo.

The racial form of the Pterygota is to be derived from the Apterygota-like racial
form of all Hexapoda, from which the various orders of insects have been produced.
These have of course developed independently of one another. Those orders, how-
ever, whose members undergo a gradual or incomplete metamorphosis, have retained
the original characteristics to a larger extent than the rest. Of the remaining
orders, again, it is the Lepidoptera, Hymenoptera, and Diptera which are furthest
removed from the racial form, and which reach the highest development among
the Insecta.

Regarding the relation of the Antennata to the Protracheata, there can be no

o/n

FIG. 357. — Larva of Polydesmus com-
planatus, just hatched (after v. Rath). Ibr,
Upper lip ; a, antenna ; bt, sides of the head
(cheeks) ; gcJi, gnathochilarium ; 6j, Z>2> 63, the
three pairs of legs of the larva ; sd, glands
(saftdriisen) ; an, anus.

504 COMPARATIVE ANATOMY CHAP.

doubt that the two classes are racially connected, and that Peripatus has, far more
than any member of the whole class of the Antennata, retained the original Annulatan
characteristics.

Review of the most important Literature.

I. Myriapoda.
Anatomy.

J. Bode. Polyxenus lagurus de Geer. Ein Beitrag zur Anatomic, Morphologic und

Entwickelungsgeschichte der Chilognathen. Diss. und in : Zeitschrift f. d.

gesammt. Naturwissenschaften. 1877.
Leon Dufour. Recherches anatomiques sur le Lithobiusforficatus et la Scutigera lineata.

Ann. sciences nat. Tome II. 1824.
Hugo Eisig. Monographic der Capitelliden des Golfes von Neapel. Berlin, 1887.

(Contains observations on the morphology of the coxal, spinning, and salivary

glands, and of the nephridia of the Myriapoda, Hexapoda, and Protracheata.)
L. Fabre. Recherches sur T anatomic des organes reproducteurs ct sur le developpem^nt

des Myriapodcs. Annales sciences nat. 4°. Zool. III. 1855.
H. Grenacher. Ueber die Augen einiger Myriapodcn. Arch. f. mikr. Anatomic.

18 Bd. 1880.
Erich Haase. Das Respirationssystem der Symphylcn und Chilopodcn. Zool. Bei-

trage von A. Schneider. 1 Bd. 1884.
Robert Latzel. Die Myriapodcn der osterreichisch-ungarischen Monarchic. Erste

und zweite Hdlfte. Wien. 1880-1884.

George Newport. On the organs of reproduction and the development of the Myria-
poda. Philos. Transact. Roy. Soc., London, 1841 (cf. also Ann. Magaz. Nat.

History. 1°. VIII. 1842).
The same. On the structure, relations, and development of the nervous and circulatory

systems. Ibid. 1843.
Otto von Rath. Beitrdge zur Kenntniss der Chilognathen. Dissertation. Bonn,

1886.
Friedr. Stein. Ueber die Gcschlechtsvcrhdltnisse der Myriapoden, etc., in Mullers

Archiv. Jahrg. 1842.
See also works of Newport, Stein, Plateau, Voges, Sograf, Humbert, Chatin, Kar-

linski, Meinert, Haase, Brandt, Koch, Gervais, MacLeod, Packard, Ryder,

Scudder, Savigny, Tbmosvary, "Wood-Mason, Heathcote, Brandt.

Ontogeny.

E. Metschnikoff. Embryologie der doppelfiissigen Myriapoden (Chilognatha}.

Zeitschr. f. wiss. Zool. 24 Bd. 1874.
The same. Embryologisches uber Geophilus. Ibid. 25 Bd. 1875.

II. Hexapoda.

Anatomy.

G. Ernst Adolph. Ueber InseTctenflugel. Nova Ada K. Leop. -Carol. Deutsch. Akad.

Naturforscher. 41 Bd. 1880.
Blanchard. " Insectes " in Regne animal de Cuvier.

vi ANTENNATA— LITERATURE 505

F. Brauer. Systematisch-zoologische Studien. Sitz.-Ber. math.-naturwiss. Klasse k.

Akad. Wiss. 91 Bd. 1 Abth. Wien, 1885.
Eduard Brandt. Numerous treatises (in German) on the nervous systems of

different insects in : Horce societatis entomological rossicce. Vols. 14 and 15

1879.

E. Burgess. Contributions to the anatomy of the milk-weed butterfly (Danais Archip-

pus}. Anniv. Memoirs Boston Soc. Nat. Hist. 1880.

Burmeister. Handbuch der Entomologie. Berlin, I. 1832 ; II. 1838, 1839.
Justus Carriere. Die Sehorgane der Thiere, vergleichend-anatomisch dargestellt.

Miinich und Leipzig. 1885.
Carl Chun. Ueber den Ban, die Entwickelung und physiologische Bedeutung der

Rectaldrusen bei den Insekten. Abhandl. Senkenb. Naturf. Gesellsch. Frankfurt

a. M. 10 Bd. 1875.
Leon Dufour. Recherches sur les Htmipteres, les Orthopteres, les Hymenopteres, les

Nevropteres et les Dipteres. Mem. Acad. de sciences. Paris, Tome IV., 1833 ;

VII., 1841; XL, 1851. See also numerous monographs in the Annales des

sciences naturelles.
V. Graber. Ueber den propulsatorischen Apparat der Insekten. Arch. f. mikr.

Anat. 9 Bd. 1873.

The same. Die tympanalen Sinnesapparate der Orthopteren. Denkschr. math.-
naturwiss. Klasse Akad. Wissench. 36 Bd. Wien, 1875.
The same. "Die Insekten." 2 Theile, in Naturkrafte. 21 and 22 Bd. Munich,

1877.
The same. Die chordotonalen Sinnesorgane und das Gehb'r der Insekten. Arch. f.

mikr. Anatomie. 20 Bd. 1882.
Battista Grassi. I progenitors dei Miriapodi e degli insetti. 7 treatises from 1886-

1888 in the publications of various Italian Acadamies and Societies, relating

chiefly to the Apterygota and Scolopendrella.
H. Grenacher. Untersuchungen iiber das Sehorgan der Arthropoden. Gottingen,

1879.

F. Grosse. Beitrdge zur Kenntniss der Mallophagen. Zeitschr. f. wissensch. Zoologie.

42 Bd. 1885.

E. Haase. Ueber Adominalanhdnge bei Hexapoden. Sitz.-Ber. d. Gesellsch. naturf.

Freunde in Berlin. 1889.

G. Hauser. Physiologische und histiologische Untersuchungen iiber das Geruchsorgan

der Insekten. Zeitschr. f. wissensch. Zoologie. 34 Bd. 1880.

F. E. Helm. Ueber die Spinndrusen der Lepidopteren. Zeitschr. f. wissensch. Zoologie.

26 Bd. 1875.
Bruno Hofer. Untersuchungen iiber den Bau der Speicheldriisen und des dazu

gehorenden Nervenapparates von Blatta. Nova Acta K. Leop.-Carol. Akad.

Naturf or scher. 51 Bd. 1887.
K. Jordan. Anatomie und Biologic der Physapoda. Zeitschr. f. wiss. Zoologie.

47 Bd. 1888.
Fr. Leydig. Zum fcinern Bau der Arthropoden. Mailer's Archiv. 1855. Besides

the text -books of Histology, many other classical works on the histology

of the Tracheata and on their sensory organs.
Joseph Heinrich List. Orthezia cataphracta. A monograph. Zeitschr. f. wissensch.

Zoologie. 45 Bd. 1887.
John Lubbock. Monograph of the Collembola and Thysanura. Ray Society, London.

1873.

The same. Origin and Metamorphosis of Insects. Macmillan. 1890.
P. Lyonet. Traite" anatomique de la chenille, qui ronge le bois de saule. La Haye.

1762.

506 'COMPARATIVE ANATOMY CHAP.

J. MacLeod. La structure des trachtes et la circulation piritrachienne. Brussels,
1880.

Meinert. Anatomia Forficularum. Dissert. I. Copenhagen, 1863.

H. Michels. Beschreibung des Nervensystems von Oryctes nasicornis im Larven-
Puppen-und Kdferzustande. Zeitschr. f. wissensch. Zoologie. 34 Bd.

Newport. " Insecta," in Cyclopaedia of anatomy and physiology. Vol. II. 1839.

J. T. Oudemans. Beitrdge zur Kenntniss der Thysanura und Collembola. Berlin,
1888. Holldndische Ausgabc : Amsterdam, 1887.

J. A. Palmen. Zur Morphologic, des Tracheensystems. Helsingfors und Leipzig,
1877.

The same. Ueber paarige Ausfiihrungsgdnge der Geschlechtsorgane bei InseTcten.
Helsingfors, 1884.

W. Patten. Eyes of Molluscs and Arthropods. Mitth. Zool. Station zu Neapel.
6 Bd. 1886.

F. J. Pictet. Recherches pour sermr a Thistoire et a V anatomic des Phryganides.
Geneva, 1834.

Reaumur. Memoires pour sermr a Vhistoire des Insectes. Paris. 12 vols. 1734-
1742.

J. C. Savigny. Memoires sur les animaux sans vertebres. 1 Partie. 1 fascicule.
Paris, 1816.

Emil Schindler. Beitrdge zur Kenntniss der Malpighi'schen Gefdsse der Insekten.
Zeitschr. f. wissensch. Zoologie. 30 Bd. 1878.

A. Sommer. Ueber Macrotoma plumbea. Beitrdge zur Anatomic der Poduriden.
Zeitschr. f. wissensch. Zool. 41 Bd. 1885.

Job. Swammerdam. Historia insectorum generalis. Utrecht, 1669.

The same. Bijbel der natuure. Lugd. Bat. 1737-1738. Bibel der Natur. 1752.

Strauss -Diirkheim. Considerations g£nerales sur Vanatomie comparee des animaux
articuUs et anatomie descriptive du Melolontha vulgaris. Paris, with Atlas,
1828.

E. Witlaczil. Zur Anatomie der Aphiden. Arb. Zool. Instit. zu Wien. 4 Bd.
1882.

The same. Zur Morphologic und Anatomie der Cocciden. Zeitschr. f. wissensch.
Zoologie. 43 Bd. 1885-1886.

The same. Die Anatomie der Psylliden. Zeitschr. f. wiss. Zoologie. 42 Bd. 1885.

Comprehensive, but chiefly either systematic or biological works of Bonnet, Rb'sel
von Rosenhof, Ch. de Geer, Kirby and Spence, Ratzeburg, 0. Heer, Taschenberg,
Jager, Westwood, and others.

Names of other authors, given without arrangement : Plateau, Gerstacker, Landois,
Kirbach, Langhoffer, Haase, Graber, Breitenbach, Walter, Lienard, "Wood-
Mason, Paul Mayer, Leon, Geise, Osc. Schmidt, Will, Leuckart, F. Dahl, H.
von Wielowiejski, Emery, Forel, Engelmann, Simmermacher, Koestler, Spich-
ardt, Packard, Kraepelin, Berger, Flogel, Dietl, Zimmermann, Cholodkovsky,
Dewitz, Korschelt, Faussek, Landois, Glaus, v. Siebold, Pagenstecher, Hanin,
A. Dohrn, F. Miiller, Adler, Cattie, Krancher, Grenacher, Hickson, Ticho-
miroff, Vayssiere, Riley, Meinert, Scudder, Leydig, Schiemenz, Grassi, Beaure-
gard, Wedde, Balbiani, Schneider, Redtenbacher, Miall and Denny, Audouin,
Westwood, Wagner, Lacaze-Duthiers, Stein, Nicolet, Gerstfeld, Brulle, Lubbock,
Semper, Kolliker, Claparede, Rathke, Hensen, Lespes, Rabl-Riickhard, Grobben,
Plateau, Carlet, Dimmock, Bolles Lee, Kniippel, Ruland, Malpighi, Suckow,
Loew, Targioni-Tozzetti, Owsjanikoff, Joh. Miiller, Brandt, Dareste, Serville,
T. de Charpentier, L. H. Fischer, Pictet, Hagen, Kirby, Curtis, Kaltenbach,
Lichtenstein, Becher, 0. Taschenberg, Herold, P. and Fr. Huber, de Saussure
Brunner v. Wattenwyl.

vi ANTENNATA— LITERATURE 507

Reproduction and Development.

H. Ayers. On the development of Oecanthus niveus and its parasite Teleas. Mem.

Boston Soc. Nat. Hist. Vol. III. 1884.
F. Brauer. Betrachtungen iiber die Verwandlungen der Insekten im Sinne der

Descendenz-Theorie. Verh. d. k. k. zool.-botan. Gesellsch. in Wien. 19 Bd.

1869.
N. Bobretzky. Ueber die Bildung des Blastoderms und der KeimUdtter lei den

InseUen. Zeitschr. f. wissensch . Zoologie. 31 Bd. 1878.
L. Dreyfus. Ueber Phylloxerinen. Dissertation. Wiesbaden. 1889.
M. Fabre. L'hypermttamorphose et les moeurs des Meloides. Ann. sciences natur. 4°.

Vol. VII. 1857.
Ganin. Beitrdge zur Kenntniss der Entivickelungsgeschichte der Insekten. Zeitschr.

f. wissensch. Zoologie. 19 Bd. 1869.
V. Graber. Ueber die Polypodie der Insektenembryonen. Morph. Jahrb. 13 Bd.

1888.
B. Grassi. Studi sugli Artropodi. Intorno allo sviluppo delle api nelVuovo. Atti

Acad. Scienze nat. Catania. 3°. Vol. XVIII. 1884.
Berthold Hatschek. Beitrdge zur Entwickelungsgeschichte der Lepidopteren. Jenaische

Zeitschr. 11 Bd. 1877.
K. Heider. Ueber die Anlage der Keimbldtter von Hydrophilus piceus L. Abhandl.

der Preuss. Akad. d. Wissensch. Berlin, 1885-1886.
0. and R. Herfrwig. Die Colomtheorie. Jena, 1881.
A. Korotneff. Die Embryologie der Gryllotalpa. Zeitschr. f. wissensch. Zoologie.

41 Bd. 1885.
A. Kowalevsky. Emlryologische Studien an Wurmern und Arthropoden. Mem.

Acad. imper. Petersburg. 7°. Vol. XVI. 1871.
The same. Beitrdge zur Kenntniss der nachembryonalen Entwickelung der Musciden.

I. Zeitsch. f. wissensch. Zoologie. 45 Bd. 1887.
R. Leuckart. Die Fortpfianzung und Entwickelung der Pupiparen. Abhandl. d.

Naturf. Gesellsch. zu Halle. 4 Bd. 1858.
The same. Die ungeschlechtliche Fortpflanzung der Cecidomyialarven. Arch, fur

Natur gesch. 1865.
E. Metschnikoff. Einbryologische Studien an InseUen. Zeitschr. f. wissensch.

Zoologie. 16 Bd. 1866.

W. Patten. The development of Phryganids, with a preliminary note on the develop-
ment of Blatta germanica. Quart. Journ. Micr. Science. N.S. Vol. XXIV.

1884.
J. van Rees. Beitrdge zur Kenntniss der innern Metamorphose von Musca vomitoria.

Zool. JaJirb. von Spengel. Abth. f. Anat. und Ontog. 3 Bd. 1888.
Viallanes. Recherches sur Thistologie des Insektes et sur les phenomenes, qui accom-

pagnent le developpement postenibryonnaire de ces animaux, in Annal. Scienc.

natur. zool. 6°. Vol. XIV. 1882.
Alfred Voeltzkow. Entwickelung im Ei von Musca vomitoria und: Melolontha

vulgaris. Ein Beitrag zur Entwickelung im Ei bei Insekten. Arb. aus dem

zool.-zoot. Institute Wiirzburg. 9 Bd. 1 Heft. 1889.
N. Wagner. Beitrag zur Lehre von der Fortpflanzung der Insektenlarven. Zeitschr.

f. wissensch. Zoologie. 13 Bd. 1860.
Aug. Weismann. Die Entwickelung der Dipteren im Ei. Zeitschr. f. wissensch.

Zoologie. 13 Bd. 1863.
The same. Die nachenibryonale Entwickelung der Musciden. Zeitschr. f. wissensch.

Zoologie. 14 Bd. 1864.

508 COMPARATIVE ANATOMY CHAP.

The same. Die Metamorphose der Corethra plumicornis. Zeitschr. f. wissensch.

Zoologie. 16 Bd. 1866.
Emanuel Witlaczil. Entwickelungsgeschichte der Aphiden. Zeitschr. f. wissensch.

Zoologie. 40 Bd. 1884.
Other authors : Brauer, Ganin, Pagenstecher, MetschnikofF, Blochmann, Dreyfus,

Fabre, Balbiani, A. Brandt, Biitschli, Dewitz, Dohrn, Graber, 0. v. Grimm,

Kblliker, Leuckart, Melnikow, P. Mayer, A. S. Packard, Tichomiroff, Zaddach,

Rathke, Robin, Henking.

vi ARACHNOIDEA— SYSTEMATIC REVIEW 509

CLASS III. Arachnoidea sive Chelicerota — Spider-like Articulata.

Systematic Review.
Order 1. Scorpionidse (Fig. 359, p. 512).

The body is divided into a compact unsegmented cephalo-thorax and a long seg-
mented abdomen. In the abdomen again a thick and broad pre- abdomen of 7
segments is marked off from a long slender post-abdomen of 5 segments. The
terminal segment of the latter carries a poison sting. On the ventral side of the 2d
abdominal segment there is on each side a comb-like appendage. The jaw-feelers
( chelicerae) and jaw-palps (pedipalps) are provided with pincers. The pedipalps are
leg-like, with large pincers. There are 4 pairs of book - leaf tracheae (lungs), whose
stigmata lie on the ventral side of the 3d to the 6th abdominal segments. Euscorpius,
Buthus, Androdonus.

Order 2. Solpugidse (Fig. 358, p. 511).

Head separate. Thorax of 3 segments, hind body cylindrical, of 10 segments.
Chelicerse with pincers, pedipalps long and leg-like. Tubular tracheae. Stigmata in
pairs on the 1st thoracic, and 2d and 3d abdominal segments. Galeodes, Solpuga.

Order 3. Pseudoscorpionidse (Chernetidse) (Fig. 360, p. 513).

Cephalo-thorax unsegmented or with two transverse furrows, abdomen broad, flat,
of 11 segments. Neither poison stings nor caudal cerci present. Chelicerse and
pedipalps like those of the Scorpionidce. Tubular tracheae. Two pairs of stigmata,
on the 2d and 3d abdominal segments. With spinning glands. Small animals.
Chernes, Chelifer, Obisium, •Chthonius.

Order 4. Pedipalpi (Thelyphonidee) (Fig. 364, p. 519).

Cephalo-thorax unsegmented, distinctly marked off from the hind body. The
latter flattened, consisting of 11-12 segments. Chelicerse claw-shaped. Pedipalps
large, ending either in claws or pincers. First pair of legs with flagellate ends, like
feelers. Two pairs of book-leaf tracheae, whose stigmata lie on the ventral side of the
2d and 3d abdominal rings. Thelyphonus (last 3 abdominal segments form a truncated
portion clearly marked off from the rest of the abdomen and carrying a long jointed
caudal cercus). Phrynus.

Near the Pedipalpi are perhaps to be classed the small and insufficiently known
divisions of the Tartaridce and Microthelyphonidce.

OrderS. Phalangidse.

Cephalo-thorax unsegmented, hind body of 6 segments, compact, applied along
its whole breadth to the cephalo-thorax. Chelicerae pincer-like, pedipalps leg-
like. Legs often extraordinarily long and thin. Tubular tracheae with one pair of
stigmata which lies ventrally at the junction of the cephalo-thorax and abdomen.
"Without spinning glands. Phalangium, Leiobunum, Gonyleptus.

Order 6. Cyphophthalmidse (often placed as a family of Order 5).

Cephalo-thorax unsegmented, abdomen of 8 segments. Of the pseudoscorpionid

type. Chelicerae and pedipalps like those of the Phalangidce. Tubular tracheae.

Cyphophthcdmus (without spinning glands, with one pair of stigmata on the ventral

side of the 1st abdominal segment). Gibbocellum (spinning glands at the base of the

510 COMPARATIVE ANATOMY CHAP.

abdomen behind the sexual aperture ; 2 pairs of stigmata on the 2d and 3d abdominal
segments) (Fig. 372, p. 529).

Order 7. Araneidae— Spiders.

Both cephalo-thorax and abdomen unsegmented, the latter large and egg-shaped.
Abdomen joined to the cephalo-thorax by a short narrow stalk, 4 to 6 pairs of
spinning mammillae at the end of the abdomen. Chelicerse claw-like with poison
glands. Pedipalps leg-like, terminal joint in the male transformed in a peculiar manner
into an organ for transmitting the semen in copulation (Fig. 377, p. 535). Tracheae
either exclusively book-leaf, or book-leaf and tubular at the same time.

Sub-order 1. Tetrapneumones.

"With 4 book-leaf tracheae, without tubular tracheae. The 2 pairs of stigmata,
ventral, behind the base of the abdomen. Generally 4 (in Atypus 6) spinnerets.
Mygale (Amcularia, Fig. 374, p. 531), Cteniza, Atypus,

Sub-order 2. Dipneumones.

With 2 book-leaf tracheae whose stigmata lie at the base of the abdomen, and with
tubular tracheae emerging through an unpaired (less frequently paired, e.g. Dysderidce)
stigma behind those of the book-leaf tracheae. The unpaired stigma of the tubular
trachese is generally moved far back, so that it lies in front of the 6 spinnerets.
This sub-order' includes most of the web-spinning spiders. Fam. Dysderidce (2 stig-
mata for the tubular tracheae) : Dysdera, Segestria. Fam. Salti grada : Salticus,
Attus. Fam. Citigrada (Lycosidce) : Lycosa, Tarantula. Fam. Laterigrada: Microm-
mata, Philodromus, Xysticus. Fam. Tubitelaria : Dictyna, Tegenaria, Agelena,
Argyroneta, Drassus, Clubiona. Fam. Retitelaria : Linyphia, Theridium, Pholcus.
Fam. Orbitelaria : Epeira, Zilla, Meta.

Order 8. Acarina— Mites.

Abdomen fused with cephalo-thorax. Body unsegmented. Mouth parts adapted
for biting, or piercing and sucking. Respiratory organs (tracheae) either present or
absent. Many Mites are parasitic.

a. Mites provided with tracheae : — Fam. Trombidiidce, : Trombidium. Fam.
Tetranychidce : Tetramjchus. Fam. Hydrachnidce : Atax, Hydrachna, Hydrodoma.
The sub-families of the marine Halacaridce : Aletes, Halacarus Avithout tracheae.
Fam. Bddlidce: Bdella. Fam. Oribatidce : Oribata, Leiosoina. Fam. Gamasidce :
Gamasus (Fig. 361, p. 514), Uropoda. Fam. Ixodidce : Ixodes, Argas.

b. Mites without tracheae :— Fam. Tyroglyphidce (cheese Mites) : Tyroglyphus.
Fam. Dermaleichidce : Listrophorus, Analges. Fam. Sarcoptidoe : Sarcoptes. Fam.
Demoditidce: Demodex. Fam. Phytoptidce : Phytoptus.

Appendage to the Class of the Arachnoidea.

The Linguatulidse (Pentastomidae)

Body vermiform, generally flattened, and ringed externally. No oral appendages.
Two pairs of movable hooks near the mouth. Without sensory organs, tracheae,
Malpighian vessels, or diverticula of the mid-gut. Male aperture in front, behind the
mouth ; female aperture at the posterior end of the body. Parasites, Pentastomum,
P. taenioides (Fig. 378, p. 537). Parasitic in the nasal and throat cavities and in
the cephalic sinus of the dog and wolf. The embryos, enclosed in their egg envelopes,
reach the exterior with the nasal mucus. If they are taken into the intestine of a
rabbit or a hare (or of a few other mammals) the embryos become free, pierce their

VI

ARACHNOIDEA— OUTER ORGANISATION

511

way through the enteric wall and enter the liver or lungs. They here become
encysted and undergo a remarkable metamorphosis, passing through many moults ;
the final result is a larva which has been named Pentastoma denticulatum. This larva
finally breaks through its cyst and moves about. If it in any way (most often with
the flesh of its host) reaches the mouth and throat of the definitive host, it chooses
its future place of location, and through a series of moults develops into the adult
Pentastomum.

I. Outer Organisation.

A. The Body.

If we compare the body of the Arachnoidea with that of the
Antennata, the most important difference that strikes us is that in the

Fig. 358.— Galeodes Dastuguei 9 , natural size. 1-6, The six pairs of extremities ; 1, chelicera
2, pedipalps ; c, head ; fh, the thorax of 3 segments ; ab, the abdomen (after L. Dufour).

former no head distinctly marked off from the thorax, or, what is the
same thing, on thorax separated from the head, can be distinguished.

512

COMPARATIVE ANATOMY

CHAP.

In the body of the Arachnoidea we find almost always a number of
anterior segments, probably 7, fused to form a generally unsegmented
eephalo- thorax. Following upon this cephalo- thorax there is an
abdomen consisting of a varying number of separate or fused segments,
which may again fuse with the cephalo-thorax, as is the case in the
Acarina and Linguatulidce ($) and thus the body appears neither segmented
nor divided into regions. We thus, within the class of the Arachnoidea,
have not only fusing of segments and an obliteration of segmentation,
but also a progressive concentration of the whole body ; there can be
little doubt that the Arachnoidea are no exception to the rule that the
more richly and completely segmented the body is the better has it
retained the primitive arrangement. The segmentation is richest in

the Scorjrionida' and Solpngidce, but
is very different in the two groups.
The segmentation in both these
forms claims special attention.

The Solpugidic (Fig. 3 5 8) vividly
recall the Insecta in the metamerism
of the body. In opposition to all
other Arachnoidea, not only is the
anterior division of the body,
answering to the cephalo-thorax,
distinctly segmented, but it even
falls into two parts, an anterior
unsegmented head, which may be
compared with the head of the
Antennata, and a posterior thorax,
consisting of three segments, which
may be compared with the thorax
of the Inwta and with the three
anterior trunk segments of the
Mi/riapnda. An abdomen of ten
segments follows the thorax.

There are considerable difficulties in
the •way of comparing the segments of the
body of the Sulpiujida with the head and
a corresponding number of trunk segments
of the A-ntciinata ; these difficulties arise
chiefly in comparing the. extremities and
nervous system, and will lie discussed later
on. The ontogeny of the Solpugidm is

unfortunatclv almost unknown.

i_'. :;:/.'.— Scorpio africanus (;dlur Cuvier,
J;«JIH animal).

Iii the Scorpionidw (Fig. 359)
the. eephalo-thorax is, in the adult
condition, unsegmented; in the

embryo, however, a segmentation into seven somites, including the
frontal lobes, may be recognised. The abdomen, on the contrary,

vi AEACHNOIDEA— OUTER ORGANISATION 513

is distinctly segmented and consists of twelve segments. In it, again,
we can distinguish two divisions, distinctly marked off from each other.
The anterior division, the broad pre-abdomen, consists of 7 segments ;
the posterior, slenderer, tail-like division, the
post-abdomen, of 5 segments. At the end of
the post-abdomen is found the poison sting1,
which is often included as one of the segments.
It ought, however, probably to be considered
as an articulated appendage of the last seg-
ment, the latter being recognised, as in all
Arthropoda, by the position of the anus.

Two small, insufficiently known groups of Arach-
iioidea, the Tartaridce and Microthelyphonidee, appear,
as far as the metamerism of the body is concerned,
to occupy in some respects an intermediate position
between the Solpugidce and the Scorpionidce, and in
others between the latter and the Thelyphonidce. In pjg. 360.— Chelifer Bravaisii
the Microthelyphonidce an anterior and a posterior (after Cuvier, Regne animal).
division can be distinguished in the cephalo-thorax, 2'6> Second to sixth pairs of
and again, the latter shows on its dorsal surface indis- e

tinct division into three parts which recalls the arrangement of the cephalo-thorax of
the Solpugidce. The abdomen consists of ten segments, the last three being much
narrower and smaller than the rest and representing a sort of post-abdomen, whose
terminal segment carries a long, thin, jointed caudal filament. In the Tartaridce, the
thorax is divided by a distinct circular constriction into an anterior and a posterior
division. The abdomen consists of seven or eight segments followed by a small,
short, truncated post-abdomen, formed of a few (four) segments and carrying a
variously -shaped caudal appendage.

In the Pedipalpi the cephalo-thorax is unsegmented. The abdomen
consists of 12 segments (Thdyphonus) or of 11 (Phrynus). In Thelyphonus
the last 3 segments are very small and narrow, and form a sort of
post-abdomen, which carries an anal filament.

The cephalo-thorax of the Chernetidce (Pseudoscorpionidce), which
recall the Scorpionidce in their general type, is unsegmented or else has
2 dorsal transverse furrows. The abdomen consists of 11 (less often
10) segments. A pre-abdomen and a post-abdomen cannot be dis-
tinguished, and a poison sting or a caudal or anal filament is wanting.

The cephalo-thorax of the Phalangidce (including the Cyplioplithal-
midce) is unsegmented. The abdomen, which is sometimes clearly,
sometimes indistinctly segmented, is applied to the cephalo-thorax along
its whole breadth. There is no separation of the abdomen into pre-
abdomen and post-abdomen, and no caudal filament.

In the Araneidce (the true spiders) the cephalo-thorax as well as
the abdomen is unsegmented. The two are separated by a deep
constriction.

In the Acaridce the segmentation of the body as well as its division
into regions is suppressed. It is rightly assumed that the un-
VOL. I 2 L

514

COMPARATIVE ANATOMY

CHAP.

segmented body has here come from the fusing of an unsegmented
cephalo-thorax with an unsegmented abdomen. It only rarely occurs
that the limbless abdomen is distinctly marked off from the limb-

Fig. 361.— Larva of Gamasus fucorum (after Winkler). 1-5, The 5 pairs of extremities of the
larva (the sixth still wanting); 1, the chelicerse = mandibles ; 2, the pedipalps = maxillae ; m,
muscles ; vm, Malpighian vessels ; g, brain = supra-oesophageal ganglion ; Is, diverticula of the mid-
gut (hepatic tubes) ; md, mid-gut ; h, heart ; ac, aorta-cephalica ; r, rectal vesicle.

bearing cephalo-thorax. An anterior portion of the body, often
distinguished as the "head," and carrying the oral aperture, can
certainly not be regarded as an original division of the body, i.e. a

vi ARAGHNOIDEA— OUTER ORGANISATION 515

division derived from ancestors. It is also very doubtful whether a
ringing of the body, which can here and there be recognised, has
anything whatever to do with a true segmentation.

The body of the parasitic Linguatulidce is elongated, vermiform, and
ringed. But this ringing has again nothing to do with a real seg-
mentation.

B. The Extremities.

The Araehnoidea are typically provided with six pairs of
extremities exclusively belonging to the eephalo-thorax. The

abdomen is everywhere limbless.

Of the 6 pairs of extremities the most anterior pair is known as the
ehelieerse (mandibles, jaw-feelers, claw-feelers, falees), the second
as the pedipalps (underjaws, maxillae). The other 4 pairs are
mostly similar in form and serve as ambulatory legs for locomotion.

The First Pair of Extremities — the ehelieerse — lie in front of and
above the mouth. They are either 2- or 3-jointed, and serve for
seizing, and often also for killing, prey. The terminal joint is claw-
like. The chelicerae are chelate, when the terminal claw is, as in the
chelate feet of Astacus, movable against a process of the preceding
joint ; they might be called claw-jaws when the terminal claw can
merely be bent round upon the preceding joint, as in the seizing feet
of the Stomatopoda.

The Second Pair of Extremities — the pedipalps or maxillae—
lie on the two sides of the mouth, and everywhere function as oral
appendages, being nearly always provided with masticatory ridges at
their bases. The masticatory ridges, which elsewhere can move freely
against one another, have in the Pedipalpi (Thelyphonidce), Cyplwphthal-
midce, and Acarina, grown together in the middle line as an adaptation
for sucking. As opposed to the masticatory ridge, the remaining part
of the second extremity is called the palp or feeler. The many-
jointed palp (originally 6-jointed) everywhere serves as an organ of
touch, but may perform very different functions as well, and in
correspondence with these functions may be very variously modified.
In the Scorpionidce, Chernetidce, and in many mites it ends in pincers
and functions as a seizing organ. In the Pedipalpi it ends as a claw
feeler with a movable claw. The feelers of the Phalangidce and of many
Acarina have a terminal claw. In the male Araneidce the terminal joint
of the feeler is transformed into a copulatory organ. The feelers of the
Solpugidce serve, like the 4 subsequent pairs of extremities, for locomotion,
and are formed much like the rest. The same is the case in the Micro-
thelyphonidce, where the second pair of extremities not only resembles
those which follow it, but is even devoid of the masticatory ridges.

The Third Pair of Extremities lies behind the mouth, and is in
most Araehnoidea more or less like the 3 following pairs, and serves,
like them, for locomotion. In the Scorpionidce and Phalangidce the
basal joint is provided with a masticatory ridge. The third pair of

516

COMPARATIVE ANATOMY

CHAP.

extremities is differently shaped in the Pedipalpi ; it is long and thin,
with long flagellate ringed terminal joints. Here it is principally or
exclusively used as an organ of touch.

The Fourth, Fifth and Sixth Pairs of Extremities are, as a rule,
similarly shaped — 6- jointed locomotory organs. In the Scorpionidce
the fourth pair of extremities also carries a masticatory ridge.

In the Linguatulidce, which are regarded as Arachnoidea degraded
by parasitism, the limbs are reduced in number and form. Only two
pairs of clinging hooks are found near the mouth. Definite data for a
comparison of these clinging hooks with any special pairs of the limbs
of typical Arachnoidea are, however, wanting.

The homologies of the Arachnoidean limbs with those of other Arthropoda are
difficult to establish. If we compare the Arachnoidea with the Antennata, and
especially the cephalo-thorax of the former with the head and 3 anterior trunk
segments (thorax) of the latter, we find that the Arachnoidea possess one pair of
extremities less than the Antennata in the corresponding regions.

In the Sotyugidce, in which the section of the body which corresponds with the
cephalo-thorax of other Arachnoidea is, as in the Antennata, segmented, the head
carries 3 pairs of extremities, viz. the chelicerae, pedipalps, and the pair of limbs
which follow these. Each of the thoracic segments following the head possesses a
pair of limbs. Since the Antennata carry typically 4 pairs of limbs on the head,
viz., the antennae, mandibles, anterior and posterior maxillae, it follows — presup-
posing that the head of the Solpugidce, really corresponds with that of the Antennata
— that the wanting limbs belong to the head. Various facts, chiefly ontogenetic,
make it probable that it is the antennae of the Antennata which are wanting in the
Solpugidce and in the Arachnoidea generally,1 while the other limbs correspond accord-
ing to their order of succession. In this way we reach the following homologies
between the Arachnoidean and Antennatan limbs. —

Head

Antennata Arachnoidea

I Antennae wanting 2

1 Mandibles = Chelicerse

1 Anterior Maxillae = Pedipalps

( Posterior Maxillse = 3d pair of limbs

Head
of the
Solpugidce

first "j Trunk foot = fourth ^j Pair "j 3 thoracic
second |- (thoracic legs = fifth Y of >- segments of
third J of Insecta) = sixth J Limbs J the Solpugidce

Cephalo-
thorax
of the
Arach-
noidea

3 Anterior

trunk
segments
= thorax
of Insecta

If these homologies are correct, then the chelicerse especially, but also the pedi-
palps and the 3d pair of extremities of the Arachnoidea, differ very greatly from the
corresponding cephalic limbs of the Antennata, i.e. the mandibles and anterior and
posterior maxillae. The mandibles of the Antennata are never jointed and the
maxillae never elongated like legs, as in the Arachnoidea. Now since it is not con-
ceivable that the 3 anterior pairs of much-jointed limbs of the Arachnoidea have
proceeded from the reduced and specialised oral appendages of the Antennata, we are
compelled to assume that if there is any near relationship between the two groups,

1 While these sheets are passing through the press it is announced that the embryo
of a large spider (Trochosa singoriensis, Laxm. ) shows distinct rudiments of antennae
which disappear later ; and further, that more than 4 pairs of rudimentary abdominal
limbs are visible, with traces of several pairs of stigmata. (Zool. Anzeiger, llth May
1891.) [TB.] 2 Ibid.

VI

ARACHNOIDEA— OUTER ORGANISATION

517

the Arachnoidea branched off from the common Tracheatan stem at a time when the
oral appendages were still much jointed, and elongated like legs.

In connection with the assumed complete absence of the Antennatan feelers in the
Arachnoidea or Chelicerota, it is a striking fact that no rudiments which can be
proved to be those of antennae appear, as far as we know, at any stage of development,
even temporarily.1 And yet we must assume, keeping Peripatus in mind, that the
ancestors of the Tracheata possessed well-developed antennae.

In recent times the near relationship of the Arachnoidea, and especially of the
Scorpionidce, with the fossil Gigantostraca and the Xiphosura has been zealously
maintained. It cannot be denied that the 6 pairs of limbs of the Scorpion show-
greater agreement with the 6 pairs of limbs of the cephalo-thorax of Limulus than
with the limbs of the Antennata. There are, however, other serious objections to the
assumption of a nearer relationship between the Arachnoidea 011 the one hand and
the Xiphosura and Gigantostraca on the other (see p. 541).

Rudiments of Abdominal Limbs in the Arachnoidea.

1. In various Arachnoidea rudiments of abdominal! limbs appear temporarily
during embryonic development ; 6 pairs on the 6 anterior abdominal segments in the
Scorpionidce (Fig. 379, p. 538), 4 pairs in the

Ghernetidce and 4 pairs in the Araneidce (Fig.
380, p. 539). Such rudiments of abdominal
limbs will probably also be found in the
embryonic stages of other Arachnoidea, whose
development has been hitherto not sufficiently
investigated.

2. The Scorpionidce possess in the adult
condition on each side of the second abdo-
minal segment ventrally a comb-like organ
(Fig. 362, k), whose function is not yet fully
known. These two "combs" are said to
come from embryonic rudiments of the limbs
of the second abdominal segment.

3. It is in the highest degree probable
that the spinning mammillae of the Araneidce,
which, 4 or 6 in number, rise on the hind
body, represent rudimentary abdominal limbs.
This is supported by the following facts : (a)
they are for the most part jointed ; (b) they,
as recently observed, develop from the em-
bryonic rudiments of abdominal limbs ; (c) the
fact that they are spinning mammillae, i.e.
that the spinning glands open 011 them.
These facts gather weight when we remember
the coxal and spinning glands of the Pro-
tracheata and Antennata, and particularly the
slime papillae of Peripatus and the spinnerets
of Scolopendrella. The assumption that the
spinning mammillae of the CypliopJiihcdmid

genus, Gibbocellum, also represent rudimentary limbs, is open to the objection that
the two pair of mammillae lie in one and the same, i.e. the 2d abdominal segment.

FIG. 362. — Buthus occitanus (Regne
animal). Cephalo-thorax, pre- abdomen
(pra), and the first segments of the post-
abdomen ( pa), from the ventral side. Limbs
(1-6) not fully drawn. 1, Chelicerse; 2,
pedipalps (jaw-feelers, chelate-feelers) ; g,
genital aperture ; s, stigmata ; k, combs.

1 See note on p. 516.

518 COMPARATIVE ANATOMY CHAP.

The occurrence of rudimentary abdominal limbs in the Araclmoidea proves that
the ancestors of these animals possessed extremities on the abdomen (at least on 6
abdominal segments).

II. The Nervous System.

The segmentation of the body is reflected in the segmentation of
the nervous system. The Scorpionidce, which of all Arachnoidea show
the richest segmentation of the body, also possess the greatest number
of ganglia in the ventral chord, while on the other hand in the
Araneidce and Acaridce concentration both of body segmentation and
of nerve chord reaches its highest point. As in other Arthropoda
concentration of the nerve chord is due to displacements, fusings,
and to reductions of originally separate segmentally-repeated pairs of
ganglia ; these processes may be directly observed during onto-
genetic development.

The brain is connected with the ventral chord by a short ceso-
phageal commissure. From the brain arise the optic nerves, and
also, in most eases, the nerves of the ehelieerse. The fact that the
chelicerae are innervated from the brain seems to oppose the assumption
that they are homologous with the mandibles of the Antennata, since the
latter always receive their nerves from the infra-cesophageal ganglion.
It has been found, however, that in the earlier stages of development
that portion of the brain from which the nerves of the chelicerse arise,
and which is often still distinctly separate in the adult animal, begins
to form in the embryo as the first post-oral pair of ganglia. These later
take part in the formation of the cesophageal commissures, or else even
fuse with the ganglionic rudiments of the segment of the frontal
lobes, i.e. with the rudiments of the actual brain. This process is evi-
dently similar to the fusing of the ganglia of the posterior antennae
with the brain in the Crustacea. In the Phalangidce, however, in op-
position to the other Arachnoidea, the nerves of the chelicerae are said
to arise out of the anterior part of the great thoracic ganglionic mass.
A similar observation has recently been made in the Acarina (Gama-
sidce), where "the mandible nerves arise out of two spherical gan-
glionic masses of the infra-oesophageal ganglion, and pass through the
supra-cesophageal ganglion."

Throughout the Arachnoidea, even in the ventral chord of the
most richly segmented Scorpionidce and Solpugidce, all the ganglia of
the cephalo-thorax, and a number of the anterior abdominal ganglia,
fuse to form one great thoracic ganglionie mass, from which
arise the nerves for the second to the sixth pairs of extremities
and for the anterior abdominal segments. In the abdomen there
may be several separate ganglia (Scorpionidce, Fig. 363), or only one
or two (Thelyphonidce, Fig. 364, Solpugidce, Chernetidce, Phalangidce,
Mygalidce among the Araneidce, Fig. 374, p. 531). In the Dipneumones
(Araneidce) and the Acaridce (Fig. 365) on the contrary, the whole
central nervous system, the brain and ventral chord, form a single mass

VI

ARACHNOIDEA— NERVOUS SYSTEM

519

FIG. 363. — Nervous system of the
Scorpion (after Newport). 1-6, Nerves of
the 6 pairs of limbs ; ma, middle eyes ; sa,
lateral eyes ; g, brain ; bg, large infra-ceso-
phageal ganglionic mass ; 01-07, ganglia of
the abdomen.

FIG. 364.— Nervous system of Thely-
phonus caudatus (after Blanchard).
1-6, First to last pairs of limbs with the
corresponding nerves from the thoracic
ganglionic mass ; au, eyes ; g, brain ; ug,
thoracic ganglionic mass ; ab, abdomen ;
ag, abdominal ganglion ; sa, jointed caudal
appendage.

520

COMPARATIVE ANATOMY

CHAP.

pierced by the oesophagus ; the greater part of this mass, which lies
behind the oesophagus, represents the fused ventral chord, from which
the nerves radiate.

The disappearance of a segmented abdominal ganglionic chain in the Arach-
noidea may have been brought about by various co-operating factors : (1) by a forward
displacement of the ganglionic masses, and the fusing of the same with the thoracic
mass ; (2) by the fusing of several abdominal ganglia to form one or two abdominal
masses ; (3) perhaps also by the running of the ganglia into the nerve trunks which,
paired or externally unpaired, run through the abdomen ; and (4) by the concentra-
tion of the whole ventral chord into one single thoracic ganglionic mass.

In the Scorpionidce, where the nervous system in the abdomen is still most richly
segmented, there are 7 abdominal ganglionic masses, 3 in the pre-abdomen, 3 in the

FIG. 365— Central nervous system (thor-
acic ganglionic mass) of Gamasus, diagram-
matic (after Winkler). g, Portion lying over
the oesophagus (o)= brain ; ug, portion lying
under the oesophagus (infra-cesophageal gang-
lionic mass) ; 1-6, nerves of the 6 pairs of limbs ;
1, of the chelicerse (mandibles) ; 2, of the pedi-
palps (maxillse) ; nz, nerve of the tongue ; ns,
visceral nerve ; tn, nerve of the maxillar palp.

FIG. 366.— Nervous system of Pentas-
tomum taenioides (after R. Leuckart). o,

oesophagus ; m, anterior portion of the chylific
stomach ; on, oesophageal nerves ; os, oeso-
phageal ring running over the oesophagus ;
ug, infra-oesophageal ganglionic mass.

post-abdomen, and 1 on the boundary between the two, which as yet cannot be
certainly assigned to the one or the other.

The longitudinal commissures of the Arachnoidean ventral chord are, almost
everywhere where they are distinguishable, fused in the middle line into an exter-
nally single median longitudinal strand.

The nervous system of the endo-parasitic Linguatulidce (Fig. 366) is much reduced.
It is restricted to one ganglionic mass lying beneath the oesophagus and an apparently
double commissure embracing the oesophagus, in which no special cerebral swelling
can be made out. This great reduction of the brain is chiefly due to the degeneration
of the eyes.

A sympathetic nervous system has been proved to exist in various Arachnoidea
(Scorpionidce, Araneidce, Acaridce), and consists of an unpaired nerve connected with
the brain by paired nerves and running along the oesophagus and stomach. Ganglia
connected with the ventral chord have also been described as belonging to the
sympathetic nervous system.

VI

ARACHNOIDEA—THE EYES

521

III. The Eyes.

Most Arachnoidea possess eyes. These are unicorneal and are,
except in the middle eye of the Scorpions, constructed on the same
general plan as the ocelli of the Antmnata. The hypodermis is nearly
always continued under the cuticular lens to form the so-called vitreous
body.

Number and Position of the Eyes. — The eyes of the Arachnoidea
are sessile and from 2 to 1 2 in number ; they lie symmetrically arranged
on the upper side of the cephalo-thorax.

Solpugidce : 2 large ocelli on one common prominence. Scorpionidce : 2-6
pairs of eyes, one pair of which, the great middle eyes, are placed close to the median
line, and the rest at the anterior edge of the cephalo-thorax. Chernetidce : 0, 1 or 2
pairs of eyes. Pcdipalpi : 4 pairs of eyes, the largest lying in the middle, the other 3

FIG. 367.— A section through a middle eye of Euscorpius italicus (after Carriere). c, Chitinous
carapace ; I, chitinous lens ; hy, hypodermis, continued as so-called vitreous body under the
chitinous lens ; p, pigment cells ; r, retinulae ; rlc, the proximal nucleated portions of the retinulse ;
no, optic nerves. B, A single retinular cell (r), with the rhabdomere (rh), and the nucleus (fc).
C, A retinula with the pigment cells plt p.2, pR, p4. (B and C, after Ray Lankester.)

on the anterior edge of the cephalo-thorax. Phalangidce : usually 1 pair of eyes in
the middle of the cephalo-thorax on a prominence. In Cyphophthalmus there is an
eye on each side on a prominence ; in Gibbocellum 2 eyes on each side at the edge of the
cephalo-thorax, each on a prominence. Araneidce : generally 8, less often 6 or fewer
eyes, symmetrically arranged, generally in 2 transverse rows on the cephalo-thorax.
The special arrangement is of value for classification. Acarina : eyes are wanting,
or present in 1 or 2 pairs. Linguatulidce : eyes wanting.

The Structure of the Middle Eye of the Seorpionidse (Fig. 367).
—The middle eye of the Scorpion takes, according to its structure, an
intermediate place between a simple eye (ocellus) and a compound or
facet eye. It agrees with the ocellus in possessing one single cuticular
corneal lens, and with the facet eye in its retinal cells (understanding
these cells in Grenacher's sense) which form groups, the so-called
retinulse.

522 COMPARATIVE ANATOMY CHAP.

Under the cuticular lens an epithelial layer lies as a continuation of the hypoder-
mis ; this represents the matrix of the lens and is called the vitreous body. Under
the vitreous body lies the layer of the retinulse. Each retinula is separated from its
neighbours by pigment cells and consists of 5 retinal cells. One rhabdomere belongs
to each retinal cell. The 5 rhabdomeres fuse in the axis of the retinula into one
rhabdome.

IV. Glands Opening on the Outer Integument.

These are very numerous in the Arachnoidea. Many of them are insufficiently
known, especially as far as their constitution and the physiological significance of
their secretions are concerned. We shall divide the different sorts of glands into two
principal groups : (1) such as open upon limbs, and (2) those whose ducts have no
apparent connection with limbs.

1. Glands opening on Limbs. — Among these we have in the first place the
spinning glands of the Araneidse, since the 2 or 3 pairs of spinning mammillae on
which they open are very probably rudimentary abdominal limbs. There are on each
side several variously constructed spinning glands, whose secretion, which hardens by
exposure to air, serves for forming the various sorts of webs. Among these different
pairs of glands there is one pair which only seems to occur in the female and to serve
for the spinning of the egg cocoon. Immediately in front of the anterior pair of the
spinning mammillae there is found in some Araneidce a paired glandular region, the
so-called Cribrellum, in which extremely numerous spinning glands open through
fine pores. The cribrellum perhaps also represents the last remains of another (4th)
abdominal pair of limbs.

The spinning glands of the Araneidce are rightly reckoned among those variously
developed integumental glands (coxal glands, spinning glands, protrusible sacs),
which must be finally traced back to the segmental setiparous glands on the parapodia
of the Annelida.

In Gibbocellum (Cyphophthalmidce) there are found on each side ventrally on the
2d abdominal segment 2 spinning mammillae, on which several spinning glands open.

Glands which open on the 4 pairs of ambulatory legs (either on one or on several)
have often been observed. One large gland is found on each side of the cephalo-
thorax of the Scorpionidce. It emerges, at least in the embryo and in young animals,
on the coxal joint of the 3d pair of ambulatory legs. On the 1st and 2d pair at
the place where the glandular apertures lie in the 3d pair, there are bulgings which
suggest that there were once glandular apertures here also. The apertures of these
coxal glands are usually not demonstrable in adult animals.

Similar glands having apertures on the coxae of the 3d pair of legs are found in
the Araneidce also, in the Tetrapneumones (Mygale, Atypus] as well as in some
Dipneumones. Here also it is often difficult to prove the existence of the outer
apertures in the adults, and here also slits may appear on other legs which correspond
in position with the glandular apertures of the 3d pair of legs.

The fact that the coxal glands of the Scorpionidce and Araneidce are unmistakably
similar to the coxal glands of the Xiphosura in position (on the 5th pair of extremi-
ties), in structure, and in manner of opening, has been used as a further argument in
favour of the relationship of these latter with the Arachnoidea.

In the Solpugidce and Phalangidce also coxal glands are said to occur, in these
cases on the bases of the last pair of legs. Their ducts have, however, not been
observed. The occurrence of coxal glands in the Acaridce has also been described.
In the Oribatidce, for example, they lie at the bases of the 2d pair of legs ; in the
Gamasidce, it appears, between the coxal muscles of all the legs. It is not yet known
if certain stigma-like pores near the bases of the 1st pair of legs of the Halacaridce

vi ARACHNOIDEA— INTESTINAL CANAL 523

belong to the category of coxal glands. In Troiribidium a gland with its opening lies
on the terminal joint of each leg.

Glands of the pedipalps (the 2d pair of extremities) have been observed in various
Arachnoidea (Atypus and other Araneidcc, Solpugidce, Scorpionidce, Phalangidce, and
Tetranychus among the Acaridce). They have been classed somewhat arbitrarily,
some as salivary glands, others (Galeodes) as poison glands, and others again as spinning
glands (Tetranychus).

Glands emerging on the chelicerse are also somewhat widely distributed. The
best known and most investigated are the poison glands of the Araneidse, which
mostly lie in the cephalo-thorax, but often partly project into the chelicene themselves
and always open outward on their terminal claws. In the Gamasidce also there are
glands at the bases of the chelicerse. According to recent observations, the webs pre-
pared by the Pseudoscorpionidce are said to be formed by glands lying in the
cephalo-thorax, whose ducts penetrate into the chelicerse and open on the terminal
joint. Earlier observers had asserted that the spinning glands and their apertures
were to be found on the ventral side of the first abdominal segment. In the
Linguatulidce there are glands emerging at the bases of the 4 clinging hooks.

We thus find in the Arachnoidea a striking number of limb glands. This number
will no doubt be still further increased on more thorough investigation, and it may
perhaps be established, that many of these glands, especially those emerging on the
coxal joints of the extremities, belong, like the spinning glands of the Araneidce, to the
category of segmental coxal glands homologous with the setiparous glands of the
Annelida.

2. Glands not emerging on the Limbs. — Here belong the integumental glands
emerging through pores in the chitinous cuticle at various parts of the surface of the
body ; these have been observed in different divisions, with special frequency, however,
in the Acaridce and Linguatulidce , and have been called oil glands, stigmatic glands,
stink glands, etc. The poison gland of the Scorpion also belongs to these. It is
paired, lies in the swollen terminal segment of the post-abdomen and emerges by 2
separate apertures at the point of the sting with which the tail is armed. In the
Phalangidce and Cyphophthalmidce (Gibbocellum) there is in the cephalo-thorax one pair
of glands (the so-called Krohn's glands), the 2 ducts of which are said to emerge
through 2 apertures on the dorsal side of the cephalo-thorax.

V. The Intestinal Canal.

This has as a rule a straight course through the body. We can
again distinguish in it the three well-known divisions, fore-gut, mid-
gut, and hind-gut.

The Fore-gut. — The mouth or buccal cavity is followed by the
muscular pharynx, which functions chiefly as a suction pump, as
it can be expanded by means of special groups of muscles attached
to it, and contracted by circular muscles. The pharynx passes into
the narrow oesophagus. This passes through the oesophageal ring and
enters the mid-gut. Before entering the latter it swells, in the Aran-
eidce, into a special sucking stomach.

The mid-gut forms by far the largest portion of the digestive tract.
In the Araehnoidea it shows in a very high degree the tendency to
form C03cal invaginations which surpass all the rest of the viscera
taken together in size and importance. Where the cephalo-thorax and

524

COMPARATIVE ANATOMY

CHAP.

abdomen are distinctly separate, these invaginations may be repeated
in each. The whole mass of the mid-gut with its invaginations repre-
sents the digesting chylific stomach, and has glandular
walls. The liquid nourishment reaches to the very
extremities of these diverticula, which have been
inaccurately called hepatic tubes.

Jl

da

FIG. 368. — Digestive
tract of the Scorpion
(after Newport), pli, Phar-
ynx ; sdt salivary glands ;
md, diverticula of the mid-
gut ; vm, Malpighian ves-
sels; ed, hind-gut.

FIG. 369.—^, Digestive apparatus of Mygale caementaria
(after Duges in Cuvier's Regne Animal). B, The abdominal
portion of the same, from the side. C, Digestive apparatus
of a Gamasus, diagrammatic (after Winkler). Lettering the
same in the 3 figures : g, brain ; dt, enteric diverticula of the
thorax ; da, enteric diverticula (liver) of the abdomen (a), only the
portions entering into the abdominal mid-gut drawn ; md, mid-
gut with diverticula (d) of Gamasus ; vm, Malpighian vessels ; rb,
rectal vesicle (cloaca) into which both the digestive tract and
the Malpighian vessels enter ; o, oesophagus.

The hind-gut is generally very short. It opens externally through
the anal aperture which is placed ventrally at the posterior end of the
body. Into the hind -gut enter tube-like excretory organs, corre-
sponding with the Malpighian vessels of the Antennata. There is
generally one pair of these, less frequently several pairs. In the
Acarina an unpaired excretory tube is often found.

The fact that the Arachnoidea like the Antennata have Malpighian vessels, while
these vessels are not found in the Crustacea and Xiphosura and Pycnogonidce, is of
great importance in deciding the question of their systematic position.

vi ARACHNOIDEA— INTESTINAL CANAL 525

The dorsal wall of the hind-gut is often bulged out in the form of
a muscular sac. It then looks as if the Malpighian vessels on the one
hand and the mid-gut on the other entered a common terminal sac, not
at its blind end but near the anal aperture. In the structure of its
walls this rectal vesicle, which is also often called the cloaca, agrees with
the Malpighian vessels, and not with the mid-gut. This favours the
view that the excretory tubes of the Arachnoidea, like the Malpighian
vessels of insects, are invaginations of the hind-gut, and consequently
ectodermal formations.

Salivary glands have often been described in the Arachnoidea, but our knowledge
of them, especially of their manner of emerging, is very inadequate. The glands
which open on the pedipalps are also often regarded as salivary glands. In certain
Acarina (Oribatidce) a pair of glands emerging at the boundary between the fore-gut
and the mid-gut has been observed. In various Arachnoidea there are groups of
glands in the upper lip.

The anatomy of the diverticula of the mid-gut varies greatly in the different
orders. In the Scorpionidce (Fig. 368) they form a 5-lobed mass on each side in the
pre - abdomen, this mass being connected with the mid-gut by means of 5 canals
(hepatic ducts). In the mid-gut of Solpuga (Galeodes) numerous branched diverticula
are said to enter both its anterior and its posterior ends. In the Pseudoscorpionidce
there are 3 diverticula of the mid-gut, 2 lateral, and 1 unpaired ventral. The two
lateral diverticula again subdivide at their outer edges into 8 lobes. The mid-gut
here forms a double loop. In the Microthelyphonidce 5 pairs of shallow bulgings have
been observed in the mid-gut. In the mid-gut of the Araneidce (Fig. 369, A, B] we
must distinguish a cephalo-thoracic and an abdominal division. The former often has

5 pairs of diverticula. The first two diverticula may anastomose with each other over
the sternal side of the thorax and so form a ring. The lateral diverticula often bend
round from the side towards the middle line of the body under the thoracic ganglion,
first, however, giving off a blind branch to the coxal joint of each limb (e.g. in Epeira
and many other Araneidce}. In Atypus the thoracic portion of the mid-gut has only
3 pairs of diverticula, the most anterior pair in this case not forming a ring.

In the anterior portion of the abdomen of the Araneidce, the mid-gut, which is
here somewhat expanded, forms a considerable number of diverticula varying in size
and much branched ; these are united by connective tissue to the mass which is
erroneously called the liver. The coloured secretions occurring in some of the cells
of these diverticula distinguish them from the non-coloured diverticula of the cephalo-
thoracic mid-gut.

The mid-gut of the Phalangidce is a tolerably spacious sac covered laterally and
dorsally by numerous (30) blind tubes. These blind tubes enter the mid-gut through

6 lateral and 1 anterior pair of apertures.

The mid-gut of the Acarina (Fig. 369, 0) also has longer or shorter bulgings, in-
vaginations, or ccecal diverticula, whose number varies. There are often 2 or 3 pairs.

The mid-gut of the Linguatulidce is a straight tube without diverticula.

The Malpighian Vessels. — In the Scorpionidce, 2 Malpighian vessels enter the
hind-gut. In one species (Sc. occitanus] 4 vessels are said to occur, 2 of them being
branched. In the Araneidce, the Malpighian vessels consist of numerous fine branched
and anastomosing tubes which finally unite on each side into two collecting ducts.
The two ducts of one side enter the rectal vesicle by a common terminal portion.

The tubes of the Phalangidce, formerly considered to be Malpighian vessels, are
said by more recent observers to emerge at the mouth. In this case they can of
course not be regarded as Malpighian vessels. They reojiicfi^fiy^her investigation.

526 COMPARATIVE ANATOMY CHAP.

In the Cyphophthalmidce, and especially in the genus Gibbocellum, 2 long Malpighian
vessels are found which enter the sac-like expanded rectum (cloaca). Each vessel
begins with a blind terminal tube, which breaks up into a plexus of fine tubules,
uniting again into a single vessel entering the rectum. Malpighian vessels have been
found in many Acarina. They are generally in the form of 2 long, occasionally coiled
tubes, entering the hind-gut. Sometimes the 2 tubes unite in a common duct which
enters the hind-gut, and they thus assume the form of the letter Y (Atax). In other
.Acarina the excretory organ is an unpaired tube lying on the mid-gut. In
Hydrodoma it emerges close behind the anus, but separate from it. In other cases
numerous Malpighian vessels are said to enter the hind-gut near the anus (Argas).
Here and there a rectal sac like that of the Araneidce is found, and into this enter
both the gilt and the Malpighian vessels (Gamasidce, Fig. 369, O, and Halarachnidce).
The arrangement of the long Malpighian vessels in the larvae and the first nymph
stage of the Gamasidce is interesting. They here (Fig. 361, p. 514) reach far forward
and form a loop at each leg which may reach into its third or fourth joint. The
blind ends of the two vessels reach far into the first pair of legs. In some Acaridce
and in the Linguatulidce no Malpighian vessels have as yet been found. Our know-
ledge of the Malpighian vessels of the Arachnoidea in general is exceedingly scanty.

VI. The Blood-vascular System.

Among the Arachnoidea this system shows very various stages of
development. It is most highly developed in the Scorpionidce and next
in the Araneidce. The blood nowhere flows entirely in blood vessels
separated from the body cavity, but rather for a larger or smaller
portion of its course enters blood sinuses and lacunae, which represent
the coelome. In the Arachnoidea also distinct relations between the
blood -vascular system and the respiratory organs can be established.
Where the respiratory organs are very strictly localised, as in the book-
leaf tracheae of the Scorpionidce and Araneidce, the vascular system with
walls of its own is most developed ; where the respiratory organs are
dispersed over the whole body, as they are in the Antennata, and ajso
where there are no special respiratory organs, the peripheral portion
of the vascular system is reduced, as in the Antennata, and even its
central organ, the heart, may disappear.

The central organ, the heart (Figs. 370, 371), shows, like that
of the Crustacea, various degrees of concentration, from the extended
many-chambered dorsal vessel provided with numerous pairs of ostia
(Scorpionidce), to the short, one-chambered cardial sac with one pair of
ostia (Acaridce). This progressive concentration is evidently closely
connected with the progressive" concentration of the whole body.

That the heart lies in a pericardium has only been with certainty
observed in a few cases. Muscles and strands of connective tissue,
which are attached on the one side to the heart or the pericardium
(the latter appears to be the case in the Araneidce), and on the other to
the integument, seem to occur pretty generally.

After the heart itself the most constant portion of the vascular
system is a median anterior vessel-like prolongation of the heart,
running on the dorsal side to the brain ; this may be called the aorta

VI

AEAGHNOIDEA— BLOOD-VASCULAR SYSTEM

527

eephaliea. It is perhaps the remains of an originally long tubular
heart which reached as far as the anterior region of the body, and
which ceased to develop ostia.

The heart lies in all Araehnoidea in the abdomen, or in that part
of the body which corresponds with the abdomen.

Scorpionidse (Fig. 370, A}. — The extended tubular heart of the Scorpion
lies in the pre-abdomen. It is 8 chambered, and has 8 pairs of lateral openings.
From the posterior end of each chamber a pair of lateral arteries diverges. The heart
is continued posteriorly into an aorta of the post-abdomen, and anteriorly into an

FIG. 370.— The hearts of various Araeh-
noidea. A, Scorpion (after Newport). B,
Araneid. C, Obisium silvaticum, juv.
(Pseudoscorpionid) (after Winkler). D,
Gamasus fucorum, larva (after Winkler).
E, Young Phalangid (after Winkler).

FIG. 371.— Heart of a Spider (Pholcus
pJw.langoides) (after Schimkewitsch). ac,
Aorta eephaliea ; o, ostia of the heart ; vp,
origin of the vena pulmonalis ; aj, ag, 0,3,
lateral arteries of the heart; aa, aorta or
arteria abdominalis ; m, alary muscles,
attached to the pericardium ; pc, pericardium.

aorta eephaliea which runs through the cephalo-thorax. From the posterior aorta
several lateral pairs of arteries arise. Immediately in front of the most anterior pair
of ostia, and thus at the root of the aorta eephaliea, a lateral artery is likewise given
off on each side. A little further forward there are two more lateral arteries running
downwards, and embracing the oesophagus, so forming an cesophageal ring. From this
cesophageal ring arises a medio- ventral longitudinal vessel running backwards ; this
lies above the ventral chord, and is called the supraneural vessel. In front of the
vessels forming the cesophageal ring the aorta eephaliea gives off numerous other
arteries, principally to the 6 pairs of extremities.

All the arteries apparently open into a lacunar system of the body, through which
the blood flows in special currents, those of the pre-abdomen bathing the book-leaf

528 COMPARATIVE ANATOMY CHAP.

tracheae, again returning to the pericardium, and thence into the heart. This por-
tion of the anatomy of the circulatory system, however, requires fresh investigation.
The circulatory system of the Scorpionidce shows considerable similarity with that
of the Xiphosura, which is increased by the occurrence of an cesophageal ring and a
medio-ventral longitudinal vessel. It must, however, be particularly noticed that a
medio-ventral longitudinal vessel (subneural vessel) occurs also in many Crustacea
(Malacostraca), which further, in the Isopoda, is connected with the anterior end of the
cephalic aorta by means of an oesophageal ring. In Peripatus also a medio-ventral
vessel is said to have been observed.

Araneidse (Figs. 370 B, 371). — After the Scorpionidce the Araneidce possess, as
far as we know, the most richly developed vascular system. The heart, which runs
along the dorsal side of the abdomen, is enclosed in a sac-like pericardium, which is
itself again, as it appears, surrounded by a blood sinus. The heart has only 3 (in
MygaU 4 ?) pairs of openings, and is continued anteriorly into an aorta cephalica and
posteriorly into a short aorta or arteria posterior, and sends off laterally 3 pairs of in-
considerable arteries which soon open into the lacunar system of the body. The
arteria posterior opens into a blood sinus placed near the anus. The aorta cephalica
runs further forward into the cephalo-thorax, soon dividing into 2 lateral trunks,
which bend downwards, and after a short course break up into several arteries
which run to the eyes and extremities. All these arteries open into blood lacuna or
blood sinuses. In this definitely arranged system of lacunae and sinuses the blood
flows through the body in definite directions. The greater portion of it finally
collects on the ventral side in the anterior part of the abdomen, here takes an upward
direction, and thus round the book-leaf tracheae, and then again finally enters the
pericardium, whence it returns to the heart chiefly through the most anterior pair of
ostia. From the book-leaf tracheae (lungs) to the pericardium the blood flows through
a special vein, formed by a continuation of the pericardial wall. Since, however, the
pericardium itself is only a part of the ccelome, this vein also cannot be regarded
as a genuine blood vessel, but only as a more sharply demarcated canal-like part of
the ccelome, i.e. of the general lacunar system.

In the Pseudoscorpionidse (Fig. 370, C), Phalangidse (E), Cyphophthalmidse and
Acarina (D) the vascular system is reduced to the heart and the aorta cephalica.
The heart itself, placed in the anterior part of the abdomen, becomes shorter and
more compact. The number of its pairs of ostia diminishes, till finally there is
, only 1 pair (Acarina, and Obisium among the Pseudoscorpionidce). This reduction
probably is caused by the anterior part of the heart losing its ostia, becoming narrow
and passing into the aorta cephalica, while only the posterior cardial chambers with
their pairs of ostia remain as a sac-like organ of propulsion.

The heart of the Pseudoscorpionidce lies in the 3 or 4 anterior abdominal segments,
and in Obisium is said to have only 1 pair of ostia, in Chernes, however, 4 pairs.
The heart of the Phalangidce and Cyphophthalmidce has 2 pairs of ostia. Among the
Acarina a heart has so far been found only in the Gamasidce and in Ixodes. It is
probable, indeed almost certain, that many other Acarina have no heart, and in
general no special blood-vascular system. The same is the case in the Linguatulidce.

In the other Arachnoidea the blood-vascular system has either not yet been in-
vestigated or else not sufficiently investigated for a comparative study.

VII. The Respiratory Organs.

The respiratory organs of the Arachnoidea are traehese, whose 1
to 4 pairs of outer apertures or stigmata almost always lie ventrally

ARACHNOIDEA— RESPIRATORY ORGANS

529

and anteriorly in the abdomen. Two sharply distinguished forms of

tracheae occur : tubular traehese and book-leaf tracheae. The former

essentially agree with the tracheae already known to us in the Pro-

tracheata and Antennata. The

latter, which are also called

lungs, lung tracheae, lung

sacs, or leaf tracheae, have till

now only been met with in the

Arachnoidea.

Tubular tracheae appear
in three modified forms, between
which, however, intermediate
stages occur. (1) The principal
trunk arising from the stigma
is branched like a tree in the
body, as in the Inseda and most
Myriapoda. Separate tracheal
trees are connected together by
anastomoses. A spiral thread
becomes differentiated in the
chitinous cuticle of the trachea?.
Such branched tree-like trachea?
are found in the Solpugidce,
Cypliophthalmidce (Fig. 372, Sj),
Phalangidce, a few Pseudoscor-
pionidce, and a few Acarina
(Gamasidce, Ixodes).

(2) The principal trunk
arising from the stigma gener-
ally divides only once into 2
chief branches. On each of
these principal branches, at
irregular intervals, are attached
tufts of long finer unbranched
tracheal tubules. Only one
such tracheal tuft is sometimes
found lying at the end of the

principal trunk. Such trachea? are found in many Araneidce, many
Pseudoscorpionidce, and in most of those Acarina which are as a rule
provided with tracheae. This second tracheal form, and especially the
modification of it last mentioned, leads over to the third form.

(3) A common tracheal trunk, arising from the stigma, is wanting.
The separate tubules of the tracheal tuft branch directly from the
stigma. We are the more justified in tracing back this third form to
a shortening and later disappearance of the common tracheal trunk,
since the posterior tracheal tufts of a few Pseudoscorpionidse (Chernes
cimicoides) still rise from the end of a short tube. Such simple tufted

VOL. I 2 M

Fig. 372. — Diagrammatic representation of the
tracheal system of Gibbocellum Sudeticum (after
Stacker). 1-6, 1st to 6th pair of limbs, only the first
(chelicerse) drawn fully ; au, eyes ; go, genital aper-
ture ; Sj, anterior pair of stigmata (for the tree-like
tracheae) ; so, posterior pair of stigmata for the tufted
tracheae ; an, anus.

530

COMPARATIVE ANATOMY

CHAP.

v-e

tracheae are found in a few Pseudoscorpionidce and a few Cyphophthalmidce
(Gribbocellum, Fig. 372, s2). They show much similarity with the tracheae
of the Scutigera (p. 479). True spiral threads are not found in the
second and third forms of tubular tracheae.

Book-leaf tracheae (traeheal lungs, lung sacs, Figs. 373

and 374). The stigma leads into
a sac filled with air, into which
there project from the anterior
wall numerous leaves arranged like
those of a book. They are, how-
ever, also attached by their side
edges to the lateral walls of the
sac, so that the latter may be com-
pared with a letter-case divided by
many partition walls into numerous
compartments ; the walls of the sac
are internally lined with a chitinous
cuticle, a continuation of the outer
chitinous integument of the body ;
this is also continued on to the
leaves, so that these consist of two
somewhat closely contiguous lamellae
connected by (muscular ?) trabeculae
or transverse supports. Between
the two lamellae of a leaf the blood
enters from the ccelome and the
respiratory process takes place
through the lamellae.

The most plausible view of tlie mor-
phological signification of these lung sacs
seems still to be that they are modified

FIG. 373. — Longitudinal section through a
book -leaf trachea of an Araneid, diagram-
matic, after MacLeod, v, Anterior ; h, pos-
terior ; ve, ventral side of the book-leaf trachea ;
d, dorsal side; fee, integument of the ventral
body wall of the abdomen ; st, stigmatic aper-
ture ; Ih, air- or traeheal cavity ; tr, the spaces
between the traeheal lamellae; p, transverse
supports between the tracheae.

traeheal tufts. If we imagine that in a traeheal tuft which opens outwardly
by means of a short traeheal trunk the separate tubules standing close together,
mutually flatten each other out into hollow plates, and that these hollow plates
become arranged in a row, we have before us a so-called book-leaf trachea or traeheal
lung. The separate very narrow spaces lying between the leaves of the air sac would
thus correspond with the lumina of the flattened tracheae.

Ribbon-like flattened tracheae are in fact to be found in the Araneidte. Compare
further the figures of the traeheal tufts of Scutigera, p. 479, which greatly facilitate
a comprehension of the view here given of the rise of book-leaf tracheae.

Another view as to the morphological significance of the book-leaf tracheae of the
Arachnoidea has been put forward by those who hold that the Arachnoidca and
especially the Scorpionidce are nearly related to the Xiphosura. According to this
view the leaves or partition walls which project into the lung sac answer to the
branchial leaves of the abdominal feet of Limulus, which have sunk below the
body surface. The 4 pairs of book-leaf tracheae in the Scorpion would thus represent
rudiments of the 4 < pairs of abdominal feet, i.e. of their branchial appendages. In
comparison with the view first given, this view seems to us artificial and unsupported
by comparative anatomy and ontogeny.

VI

ARACHNOIDEA— RESPIRATORY ORGANS

531

The Scorpionidce, Pedipalpi, and the tetrapneumonic Araneidce
(Mygalidce) have only book-leaf tracheae. In the dipneumonic Araneidce
book-leaf and tubular tracheae exist simultaneously.

Number and Position of the Stigmata.

The Scorpionidce possess 4 pairs of book-leaf tracheae and 4 pairs of stigmata,
lying laterally on the ventral, side of the 3d to the 8th abdominal segments
(Fig. 362, p. 517).

The Pedipalpi have 2 pairs of
book -leaf tracheae with 2 pairs of
stigmata on the ventral side of the
2d and .3d abdominal segments (Fig.
364, p. 519).

Among the Araneidce the ar-
rangements are different in the Tetra-
pneumones and the Dipneumones.
The Tetrapneumones (Mygalidce}
have 2 pairs of book -leaf tracheae
and 2 pairs of stigmata (Fig. 374,
ft, s) lying on the ventral side of
the base of the abdomen. The Di-
pneumones have only one pair of
book -leaf tracheae, corresponding
with the anterior pair of the Tetra-
pneumones. Besides these, however,

they have, as the equivalent of the - \s$^ ^ -'(iVj^LJ^^ffW - £

second pair of book-leaf tracheae
of the Tetrapneumones — tubular
tracheae, which generally open
through an unpaired stigma in the
shape of a transverse fissure placed
far back on the abdomen. This
unpaired stigma no doubt arises
from a posteriorly placed pair of
stigmata, which have united. This
supposition is supported by the fact
that in some Araneidce (Dysdera,
Segestria, Argyroneta} two separate
more anteriorly placed stigmata for
the tubular tracheae occur (behind
the pair for the book-leaf tracheae),
and also by the circumstance that
in a few cases (Dictyna] the tracheal
trunks which open through the un-
paired stigma are distinctly recog-
nisable as double.

The Solpugidce have tubular
tracheae with tree - like ramifica-

FIG. 374. — Mygale, from the ventral side. The
ventral wall of the cephalo-thorax removed to show the
large cephalo-thoracic ganglion (bg) and the 2d small
ganglion at the base of the abdomen. The ventral wall
of the abdomen on the left side opened out. m, Ventral
muscles of the abdomen ; I, lamellae of the book-leaf
trachete; ft, book-leaf _ tracheae; s, stigmata of the
same ; ov, ovary ; sp, spinning mammillae ; 1-6, 1st to
6th pairs of extremities. 2-6, not completely drawn
(Regne animal).

tions, opening through 3 pairs of
stigmata, the first pair lying in the first thoracic segment, the second and third pairs
in the second and third abdominal segments. The position of the first pair of stig-
mata in the thorax deserves to be specially noted.

532 COMPARATIVE ANATOMY CHAP.

The Pseudoscorpionidce have tubular tracheae with 2 pairs of stigmata, which lie in
the 2d and 3d abdominal segments. In Cheiridium there is only one pair of stigmata,
which has perhaps arisen by the fusing of the two pairs found in other Pseudo-
scorpionidce.

The ramified tracheae of the Phalangidce are said to open through a single pair of
stigmata, lying ventrally at the anterior end of the abdomen, which is closely applied
along its whole breadth to the cephalo-thorax.

Among the Cyphophthalmidce, Cyphophthalmus is said to have only one pair of
stigmata on the under side of the first abdominal segment. Gibbocellum (Fig. 372),
on the contrary, has 2 pairs of stigmata lying laterally and ventrally in the 2d
and 3d abdominal segments. The anterior pair of stigmata leads into richly branched
tracheae, whose two principal trunks unite into an unpaired median trunk in the
cephalo-thorax. The posterior pair of stigmata lead into tufted tracheae. Each
stigma is covered by a plate pierced like a sieve, and each pore in this plate represents
the aperture of a tracheal tubule.

In many Acarina, especially in the parasitic and marine Acarina, tracheae are
wanting. When they are present they open out through one pair of stigmata,
which are placed very unusually. This pair of stigmata generally lies near the coxal
joints of the last pair of extremities, but often much further forward. It occasionally
lies on the dorsal side, and sometimes above the base of the chelicerae. This arrange-
ment is not at present understood. In certain Acarina short tubes or sacs connected
with apertures in the outer chitinous integument have been considered as the rudi-
ments of tracheae.

The Linguatulidce are devoid of tracheae. The Microthelyphonidce also are said
to have no special respiratory organs. If this be established, it must not be considered
in the Acaridce the original arrangement.

The Tartaridce are said to possess lateral apertures supposed to be stigmatic in
the 2d, 3d, and 4th ventral rings, thus having 6 in all.

A review of the position of the respiratory organs and their apertures in the various
divisions of the Arachnoidea shows us that not only do several abdominal segments
possess stigmata, but that these may occur, as is shown by the example of the
Solpugidce, in the thorax also. Leaving out of consideration the anterior position
of the stigmata in certain Acarina, which are a very one-sidedly developed Arachnoid
group, and evidently, excepting the Linguatulidce, the furthest removed from the
racial form, we are justified in assuming that the, to us, unknown racial form of the
Arachnoidea possessed a larger number of stigmata x and of tracheae connected with
them than any Arachnoid form now living. This presupposes that the book-leaf
tracheae are modified tubular tracheae.

VIII. Sexual Organs.

In all Arachnoidea the sexes are separate. The sexual organs lie
in the abdomen. The testes and ovaries are either paired or single.
The paired condition must be the more primitive. The ovaries in
very many Arachnoidea appear as tubes beset with spherules or sacs,
and so have a grape-like appearance. The eggs arise only in the sacs,
which may be called egg -follicles, and they thence enter the ovarian
tube, which serves only as a duct. •

With very rare exceptions the ducts of the sexual organs are
paired. These unite in their terminal portion, and open externally

1 See footnote on page 516.

VI

ARACHNOIDE A— SEXUAL ORGANS

533

through an unpaired ventral genital aperture at the anterior end of
the abdomen. Several organs, chiefly accessory, are connected with
the terminal portion of the ducts, viz. receptacula seminis, vesiculse
seminales, glands, male and female copulatory organs. The anatomical
structure of the sexual apparatus in the different Arachnoidea varies
greatly in detail. The review which follows is incomplete, and only
takes into account the better known forms.

Scorpionidse. Female Apparatus (Fig. 375, A). — Three longitudinal tubes, beset
with spherical ovarian follicles, lie in the pre-abdomen, one median and two lateral.
The median tube is connected with the lateral by 5 transverse anastomoses, also beset

FIG. 375.— Female sexual apparatus of various Arachnoidea. Most of the figures are some-
what diagrammatic. A, Scorpio occitanus (after Blanchard). H, Galeodes barbarus (after L.
Dufour). C, Trichodactylus anonymus (Acarid) female sexual organs of the nymph (after
Nalepa). A An Araneid. E, Pentastoma taenioides (after R. Leuckart). F, Phalangium
opilio (after Gegenbauer). G. Cepheus tegeocranus (oribatid) (after Michael). H. Gamasus
crassipes (Acarid) (after Winkler). I, Trombidium fuliginosum (after Henking). ov, ovaries ;
od, oviduct ; go, genital aperture ; rs, receptaculum seminis ; or (in C), outer aperture of the same ;
va, vagina (in E also the uterus) ; op, ovipositor ; a, glandular appendages.

with ovarian follicles, so that the ovigerous portion of the sexual apparatus forms
a network of 8 meshes. From the anterior end of this there arises on each side
an oviduct, which at once swells into a rather long tube (receptaculum seminis, or
vagina?) The two tubes converge towards the ventral middle line, where they
emerge on the first abdominal ring, in front of the combs, through an aperture which
is covered by 2 valves. The Scorpionidce are viviparous. The embryos develop in
the ovarial tubes, which function as uteri.

Male Apparatus (Fig. 376, A).— The tubular testes are distinctly paired. There

534

COMPARATIVE ANATOMY

CHAP.

are on each side two testicle tubes connected together by anastomoses. These two
tubes unite anteriorly to form a sperm duct, which, joining the duct from the other
side, opens outwardly at the place where, in the female, the genital aperture lies.
Paired accessory organs. are connected with the ducts, viz. copulatory organs, seminal
vesicles, and glands.

Pseudoscorpionidse. — The ovary is an unpaired tube beset with follicles, which is
continued into two oviducts entering a short vagina. Numerous unicellular, and 2
long tangled tubular glands are connected with the vagina.

The testes in Chernes and Obisium recall in their form the ovaries of the Scor-
pionidce. In Chelifer, on the contrary, we find a single median testicle tube. There are
everywhere 2 sperm ducts, entering a common copulatory apparatus, with which are
connected glands similar to those in the female. The unpaired genital aperture lies
ventrally in both sexes on the boundaries between the 2d and 3d abdominal segments.

FIG. 376.— Male sexual apparatus of various Arachnoidea. Most of the figures are somewhat
diagrammatic. A, Scorpio occitanus (after Blanchard). B, Galeodes barbarus. C, Galeodes
nigripalpis (after Dufour). D, Philoica domestica (Araneid) (after Bertkau). E, Pentastoma
taenioides, only the anterior end of the testes drawn (after Leuckart). F, Uropoda (Acarid) (after
Winkler). G, Trombidium fuliginosum (Acarid) (after Henking). H, Phalangium opilio (after
Krohn). I, Gamasus crassipes (Acarid) (after Winkler). The letters in all cases signify : t, testes
(dotted) ; vd, vasa deferentia ; sb, seminal vesicle ; p, penis ; a, glandular appendages ; as, tubular
appendages ; ab, vesicular appendages ; be, bursa expulsatoria ; c, cirrus ; cb, cirrus pouch.

Solpugidse. — The female sexual apparatus (Fig. 375, B) consists of 2 long ovarial
tubes beset at their outer edges with numerous ovarian follicles, and placed in the
abdomen. From each ovary there arises an oviduct. The 2 oviducts unite, their
ends swelling. The external genital aperture is a longitudinal fissure on the ventral
side of the first abdominal segment.

The male apparatus (Fig. 376, B, C) consists of 2 thin and very long winding
testicle tubes on each side of the abdomen, and quite separate from each other. These
testicle tubes are continued into sperm ducts. The 2 sperm ducts of one side unite
after a longer or shorter course into one duct, which, joining that from the other
side, opens externally through a common aperture, which lies ventrally on the first
abdominal segment. The ducts have either 4 or 2 swellings, regarded as seminal

vi ARACHNOIDEA— SEXUAL ORGANS 535

vesicles ; if there are 4 they lie in the course of the 4 sperm ducts, if 2 in the course
of the 2 common efferent ducts.

Pedipalpi. — The ovaries and testes are paired with paired ducts, and a common
unpaired genital aperture on the ventral side of the first abdominal segment. The
Phrynidce are viviparous.

Microthelyphonidse. — The germ glands (the ovaries at least) are said to be un-
paired. There are probably 2 oviducts, which open outward by means of a common
terminal piece on the ventral side of the first abdominal segment.

Araneidse. Female Apparatus (Fig. 375, D). — There are in the abdomen 2 wide
tubes, beset with numerous ovarian follicles, and looking like a cluster of grapes.
The free ends of the ovaries sometimes fuse in such a way as to give rise to an un-
paired circular ovary. There are always two short oviducts, uniting to form a short
terminal portion (vagina), which emerges through the unpaired median genital
aperture at the base of the abdomen, on the ventral side, between or somewhat
behind the anterior pair of stigmata. All female Araneidce possess receptacula
seminis. There is either one receptaculum, or two lateral receptacula, less frequently
three, one median and two lateral. These receptacula, into wrhich, during copula-
tion, the semen is introduced, are entirely separate from the sexual apparatus in
many Araneidce, and have separate outer apertures near the female genital apertures.
In others they are accessory organs of the vagina. In Epeira each of the two
receptacula has 2 apertures — an outer one, placed on the genital plate near the
sexual aperture, and an inner one leading into the vagina.

Male Apparatus (Fig. 376, Z>). — Two testes lie as long tubes in the abdomen, and
are continued as 2 long thin and often much coiled sperm ducts, these' opening out-
ward by means of a short wide common duct through the male genital aperture, which
lies between the 2 anterior stigmata. The transition from the testes into the vasa
deferentia is often gradual, so that it is difficult to say where the former leave off and
the latter begin. Occasionally the blind ends of the 2 testes are united by connec-
tive tissue.

In the male sexual apparatus of the Araneidce a special copulatory organ is
wanting. The pedipalps of the male function as copulatory organs, their terminal
joints (Fig. 377) being transformed in a
peculiar manner. The inner side of this
terminal joint carries an outgrowth, through
which runs a spirally coiled canal emerging
at the pointed end. This canal is filled by
the male with sperm from the genital aper-
ture. When copulation takes place the
point of the outgrowth of the pedipalp is
introduced into the receptaculum seminis of

the female, and the semen discharged from /

the spiral canal into the receptaculum.

PViaioTKririca rn\n *7* P Ufa AM TT\ FlG- 3T7.-Last joint of the pedipalp of Fili-

Phalangida (Fig. 3/5, F, Fig. 3/6, S). gtata tegtacea Latr (after Bertkau).

— Both the ovaries and testes are here un-
paired. Each germ gland is a semicircular tube which, by analogy with the
arrangement described in the Araneidce, may well be considered to have arisen by the
fusing of the blind ends of originally paired germ glands. The ovarian tube is
superficially beset with ovarian follicles. The 2 ends of the germ glands are con-
tinued into 2 ducts (sperm ducts in the male, oviducts in the female), and these
unite to form a common duct which enters the copulatory apparatus. This, in the
male, is a rod-like penis, in the female, a long cylindrical ovipositor. Both the penis
and the ovipositor are enclosed in special sheaths, and they can, together with their
sheaths, be protruded and evaginated. The 2 vasa deferentia are very much coiled

536 COMPARATIVE AX ATOMY CHAP.

shortly before entering the common duet. IV fore the latter enters the penis, its
wall becomes strongly muscular. This muscular portion of the duct evidently serves
as a propelling organ for driving the semen out of the penis. A pair of accessory
glands enters the end of the penis sheath.

In the female the common efferent duct has 2 divisions, the proximal division
being enlarged into a uterus, which at maturity is lilled with eggs, the narrow and
long distal portion being a vagina which is continued into the ovipositor. The vagina
has 3 lateral sacs, which are regarded as receptacula seminis. Accessory glands enter
the end of the sheath of the ovipositor.

The genital aperture in both sexes lies ventrally, on the boundary between the
cephalo-thorax and the abdomen.

It often occurs in the male PhalangiclcG that eggs develop on the surface of the
testes ; these apparently do not leave the body, but are reabsorbed.

Cyphophthalmidae. — Here also the genital aperture lies ventrally at the base of
the abdomen (on the first abdominal segment). The male has a long penis, the
female a long ovipositor.

Acarina (Fig. 375, C, G, If, I, Fig. 376, F, G, /).— Great variety here prevails in
the structure of the sexual organs. The following are 2 extreme cases. In the
first there are 2 separate symmetrically-placed germ glands and 2 separate ducts,
opening outwards through a common unpaired copulatory organ. We here see the
more original arrangement. The other extreme is rare. It is found in the female of
the Gamasidce (Fig. 375, If), where a single unpaired ovarv is continued into a single
unpaired duct, which opens outwards through the copulatory organ. Transition
forms between these 2 extremes are very commonly found. The 2 germ glands fuse
in various ways to form one, which sometimes still shows traces of its originally
double character ; the ducts, however, remain separate to a greater or smaller extent.

Accessory organs, glands, receptacula seminis are often connected with the ducts.
The unpaired terminal portion of the ducts nearly always leads to the outer sexual
apparatus, which in the male is the penis, and in the female may be developed as an
ovipositor. There are often found near the genital apertures adaptations (c.y. suckers)
which assist in copulation. The sexual organs are by no means limited in position
to the posterior part of the body, on the contrary, the fact that they often run far
forward shows the extent to which the concentration of the whole body and the
obliteration of the boundary between the cephalo-thorax and the abdomen have
taken place. The aperture of these organs often lies far forward also, in some cases
as far forward as between the most anterior pair of legs. It has been observed in
the Tyroylyphu (Trichodactylus anonyinus] that the genital aperture which in the
adult female lies between the 2d pair of legs, in the last larval stage (before the
last moult) still lies between the last pair. This observation also throws light on
the anterior position of the stigmata in many mites, which must be attributed to
displacement.

As in the Arnncida^ so also in certain Acurida there occur in the females recepta-
cula seminis witli apertures separate from the rest of the sexual apparatus. Tricho-
dactylus thus has a receptaculum at the posterior end of the body, opening outward
through a post-anal aperture. The penis is introduced into this aperture during
copulation. The receptaculum is connected by 2 short tubes with the 2 ovaries.
Tin's arrangement and that found in Eprira recall to a certain extent the well-known
arrangement in the Trcniatoda and (Jvxl.uda, where the female sexual apparatus is
connected with the exterior not only by means of the usual genital aperture, but by
Laurer's duct as well. In Trichodactylus the receptaculum arises independently by
an invagination of the integument, and becomes connected with the ovaries only
secondarily.

Home Acarina are viviparous, others ovoviviparotis, i.e. the eggs develop to a

VI

ARACHNOIDEA— ONTOGENY

537

-oe

t S

certain extent within the mother body, so that the young is hatched soon after the
laying of the egg. Most Acarina, however, are oviparous. The eggs or embryos
collect often in great numbers in the expanded oviducts, which then function as
uteri.

Linguatulidse. Female Apparatus (Figs. 375 E, 378). — The ovary is a long un-
paired tube, beset with ovarian follicles, which runs through the body over the
intestine in a longitudinal direction. It is continued
anteriorly into 2 oviducts, which surround the oeso-
phagus, and under it enter the anterior end of the
unpaired vagina. This serves at the same time as
uterus, the first embryonic development of the eggs
taking place in it. The vagina is an unusually long
tube which runs backward with many windings,
accompanying the intestine, and is often filled with
several hundreds of thousands of eggs and embiyos ;
it opens outward through a female genital aperture
close to the -anus. The ducts of 2 long receptacula
seminis, which lie at the 2 sides of the mid-gut, enter
the most anterior end of the vagina at the point where
the oviducts join it.

Male Apparatus (Fig. 376, E}. — The testes are
paired or unpaired tubes placed like the ovary.
The testis or testes are continued anteriorly into
an unpaired efferent division, which has been
regarded as a seminal vesicle. This vesicula semi-
nalis divides anteriorly into 2 canals, the vasa
deferentia, which encircle the ossophagus. Each vas
deferens ends in a male copulatory apparatus. The
male genital aperture common to the 2 copulatory
apparati lies, in contradistinction to that of the
female, in the anterior portion of the body, between
the 2d pair of hooks. Corresponding with the recep-
tacula seminis of the female there are in the male 2
blind tubes running backward, these are apparently
organs for propelling the semen, and enter the 2 sperm
ducts. The end of each sperm duct enters a very long
chitinous cirrus, which at a time of rest is rolled up
in a special sac.

IX. Ontogeny.

FIG. 378.— Female of Pentas-
tomum taenioides at the time of
copulation, with the viscera (after
Leuckart). Ti, Hooks ; oe, oeso-
phagus; rs, receptacula seminis,
one of which is still empty ; d,
gut ; ov, ovaiy ; va, vagina.

We can only bring forward a few facts concerning
the ontogeny of the Arachnoidea, chiefly such as are
of most importance from the point of view of compara-
tive anatomy.

1. The segmentation is in the main that of the centre- or mesolecithal eggs. A
blastoderm is formed covering the yolk, in which, however, merocytes remain. The
formation of the germ layers and of the rudiments of the most important organs
proceeds, as in other Arthropoda, from a blastoderm plate, which may be called the
embryonic rudiment. In the Scorpionidce, however, the egg seems to be meroblastically
telolecithal, and the furrowing takes a corresponding course, so that no blastoderm
enveloping the yolk on all sides is formed, but a germ disc is developed at one pole
of the egg.

538

COMPARATIVE ANATOMY

CHAP.

\

2. Embryonic envelopes have till now only been found in the Scorpionidce. The
embryonic envelope here, as in the Insecta, consists of 2 membranes, the outer repre-
senting the serosa, the inner the amnion of the Hexa-
poda.

3. The formation of segments in the embryonic
rudiments takes place as a rule from before backward,
so that new segments are continually formed from the
terminal segment behind those already developed. Fre-

;V%f~~" ~yyjjil quently, however, the segment bearing the chelicerfe,

1 V •" ^ • ~ * and sometimes that bearing the pedipalps, appear only
after the formation of a few of the subsequent segments.

4. The rudiments of the extremities seem in various
Arachnoidea to have different orders of succession. The
permanent extremities, with the exception of the cheli-
cerfe, which begin to form later, often develop simul-
taneously. In the Pseudoscorpionidce the rudiments of
the extremities are even said to be recognisable before
the marking off of the segments on the embryonic
rudiments.

5. In all Arachnoidea, except the Linguatulidce, the
body is, in its embryonic condition, more richly seg-
mented than in the adult animal. The cephalo-
thoracic region especially shows embryonic metamerism.
This region consists at certain embryonic stages of a
cephalic or frontal lobe, in which the stomodseum and
the definitive oral aperture form, and of 6 subsequent
and thus post-oral segments, the 1st being that of the

ciba

FIG. 379.— Embryo of a Scor-
pion, spread out flat, from the
ventral side (after Metschni-
koff). U, Frontal lobes; 1,
chelicerse; 2, pedipalps; 3-6,
the 4 pairs of legs ; aba, rudi-
ments of the abdominal limbs ;
pa, post abdomen.

chelicerse, the 2d that of the pedipalps, while the 4 others are the segments of the
4 following pairs of extremities. In the abdomen also, even when there is no meta-
merism in the adult animal, .segmentation is to be recognised in the embryo. The
number of the embryonic abdominal segments in the various divisions of the Arach-
noidea, however, differs greatly.

It is a specially important fact that the segment bearing the chelicerse is
in the embryo post-oral. No extremities develop on the frontal lobe, where, in the
Crustacea, Protracheata, and Antennata, the antennae form.1 From this fact it follows
that there is no correspondence between the chelicerse of the Arachnoidea and the
antennae of the Antennata.

The appearance of rudimentary abdominal limbs in the embryos of many
Arachnoidea has already been mentioned ; some of these rudiments disappear later,
some, however, are retained (e.g. the combs of the Scorpion, the spinning mammillae of
the Araneidce).

6. "What has been said about the relation of the embryonic metamerism of the
body to the definitive metamerism of the same is also true of the nervous system.
Investigations show that in the Scorpionidce and Araneidce a pair of ganglia forms in
each embryonic segment. The embryonic pair of ganglia of the frontal lobe is the
rudiment of the supra-resophageal ganglion. In the first post-oral segment a special
ganglion for the chelicersB is developed, which only secondarily joins the supra-
cesophageal ganglion, forming with it the brain. In the Antennata and Protracheata,
on the contrary, the antennae are from the very first innervated from the preoral
supra-cesophageal ganglion. Each of the following embryonic segments except the
terminal segment in like manner possesses 1 pair of ganglia. The more or less

1 See footnote on page 516.

VI

A RA CHNOIDEA—PHYLOGENY

539

concentrated form of the nervous system of the adult animal arises in consequence of
the fusing of pairs of ganglia which were separate in the embryo. The whole central
nervous system arises in a manner similar to that in other Arthropoda.

7. The Mesoderm of the Araclmoidea at a certain embryonic stage is developed
just as in the Annulata, Protracheata, Antennata, and perhaps also the Crustacea,
in the form of 2 lateral segmented streaks or bands with segmental cavities.

8. The fore- and hind-guts develop in the well-known way as imaginations of
the ectoderm (stomodteum and proctodseum). Opinions still differ as to the manner
of formation of the tube of the mid-gut.

9. The first rudiments of the lungs (book-leaf tracheae) appear as imaginations of
the ectoderm, and thus in the same way as the tracheae in the Antennata.

10. Most Arachnoidea when born or hatched from the egg resemble the adult.
As far as we know, a post-embryonic metamorphosis occurs only in the Pseudo-
scorpionidce and Acarina. The former are hatched in a very imperfect condition, but

-fte

FIG. 380.— A, B, C, Embryos of Agelena labyrinthica at three stages of development ; in A and
B supposed to be spread out flat, in C in the natural form from the venti'al side ; Jd, frontal lobes ; st,
stomodseum ; 1-6, 1st to 6th pairs of extremities (of the cephalo-thorax) ; viz. 1, the chelicer* ; 2, the
pedipalps ; 3-6, legs ; a, rudiments of abdominal limbs ; aw in C spinning mammillae (after Balfour).

remain for some time (parasitically) attached to the body of the mother, who carries
the eggs about with her. The young larvae of the Acarina (Fig. 361, p. 514) are still
devoid of the last pair of extremities, i. e. of the 4th pair of legs. The metamorphosis
in the Acarina is often very complicated, and is accompanied by many moults.
Sometimes several pupal and larval stages occur. In such cases the metamorphosis
is accompanied by the same inner processes as in the Insecta with complete meta-
morphosis, i.e. by the breaking up and disappearance of larval organs, and by the forma-
tion of the definitive organs out of imaginal portions of the larva. The development
and life-history of the Linguatulidce were briefly described in the systematic review.
The ontogeny of this group yields no data for deciding their systematic position.

X. Phylogeny.

The Arachnoidea form a sharply demarcated natural division of the Arthropoda.
Leaving the Linyuatulidce out of consideration, there can be no doubt as to the

540 COMPARATIVE ANATOMY CHAP.

relationship of the various orders united in this class. There is, further, no doubt
that those Arachnoidea whose bodies are most richly segmented have best preserved
the original character. These are the Scorpionidce, on account of the rich segmenta-
tion of their abdomen and abdominal nervous system, and the Solpugidce, on account
of the segmentation of that part of the body which answers to the cephalo-thorax of
other Arachnoidea ; this cephalo-thorax consists of an anterior section and three sub-
sequent segments, which may be compared with the head and three anterior trunk
segments of the Antennata. For such a comparison, however, further data as to the
structure and development of the Solpugidce are needed.

The Acarina, with their highly concentrated organisation, are evidently the
furthest removed from the racial form of the Arachnoidea. The other divisions in
various points take intermediate positions, and may be used as examples of pro-
gressive concentration.

It is by no means proved that the Linguatulidce belong to the Arachnoidea. We
are not even in a position strictly to prove that they are Arthropoda. It is, in any
case, quite possible that the Linguatulidce are mite-like animals greatly modified by
parasitism, but they might just as well be degenerate descendants of other Arthropoda.
Definite data are wanting for deciding whether the want of a respiratory and blood-
vascular system is original, or has only arisen secondarily through parasitism. The
presence of 2 pairs of hooks does not prove that they are Arachnoidea or indeed Arthro-
poda at all. The history of their development also affords us no assistance. The
position of the female genital aperture at the posterior end of the body is unusual in
the Arachnoidea. The diverticula of the mid-gut so common among the Arachnoidea
are wanting here. The reduction of the central nervous system to an cesophagcal
ring with several cesophageal ganglia is probably connected with the reduction of
sensory organs, extremities (?), etc., brought about by parasitism, but this does not
help to decide the question whether they are descended from Arachnoidea and not
from other animals provided with such a nervous system. There finally remains only
the constitution of the ovarial tubes beset with ovarian follicles, which specially recall
the Arachnoid arrangement, and opinions may differ as to the value of this point of
agreement.

The question as to the systematic position of the Arachnoid class in the Arthro-
podan system is still a matter of discussion. There are two views on this subject.
According to one of these views the Arachnoidea are nearly related to the Xiphosura
and fossil Gigantostraca ; these three would then together form a third subphylum of the
Arthropoda, distinct both from the Crustacea and the Antennata. According to the
other view the Arachnoidea are racially connected with the Antennata, and form with
these and the Protracheata the subphylum of the Tracheata. At present we prefer the
latter view, and consider the Arachnoidea as Tracheata which have lost their antennae,1
while the first postoral pair of limbs, homologous with the mandibles of the Anten-
nata, has been pushed forward, so that in all adult Arachnoidea they are inserted in
front of the mouth. The cephalo-thorax of the Arachnoidea would then correspond
with the fused head and thorax (3 anterior trunk segments) of the Antennata, and
we thus perhaps find in the segmentation of the cephalo-thorax in the Solpugidce a
primitive arrangement. If this supposition should prove correct, then the compari-
son of the other organs presents no very great difficulties. In judging of the system-
atic position of the Arachnoidea, their relationship to the Antennata is strongly
supported by the facts that the Arachnoidea possess both Malpighian vessels and
tracheae, which are wanting in the Crustaceans as in the Xiphosura.

In the Antennata, the mandibles and the 2 pairs of limbs of the head which

1 See footnote on page 516.

vi ARACHNOIDEA—PHYLOGENY 541

follow them are either not at all or very little like legs, they are changed into mouth
parts. In the Arachnoidea, the appendages which probably correspond with them
(the chelicerfe, the pedipalps, and the 1st pair of legs) have preserved far better the
character of long-jointed extremities. Now, since we can find no justification in com-
parative anatomy or ontogeny for deriving long many-jointed extremities from
abbreviated and specialised mouth-parts adapted for chewing, sucking, etc., but are,
on the contrary, distinctly justified in assuming that the opposite process is the usual
one, we conclude that the racial form of the Arachnoidea branched off from the racial
form of the Antennata very early, at a time when the limbs lying directly behind
the mouth were not yet changed into specialised mouth-parts. The Arachnoidea
(Chelicerota) on the one hand, and the Antennata on the other, would thus represent
two branches diverging early from the Tracheate trunk. The Protracheata cannot,
it is true, be placed at the root of this trunk, but may still in many points of their
organisation much more faithfully retain the primitive condition than do the Arach-
noidea and Antennata, and may thus to a certain extent represent an offshoot from
the root.

The above statements must make the relationship of the Arachnoidea and especi-
ally of the Scorpionidce with the Xiphosura and Gigantostraca appear at present
doubtful. At the same time it cannot be denied that the limbs of the cephalo-
thorax in the Arachnoidea show a remarkable agreement with those of the Xiphosura
and Gigantostraca, a much greater agreement than with the corresponding limbs of
the Antennata. The want of preoral limbs comparable with the antennae is also a
point of agreement not to be underestimated.1 But we may possibly have here only
a phenomenon of convergence. The agreement in the rest of the organisation, leaving
out of account characteristics common to all Arthropoda, appears to us not so great
as to justify a nearer relationship based upon it. Even if the occurrence of rudi-
mentary abdominal limbs forces us to assume that the ancestors of the Arachnoidea
possessed abdominal limbs, the -'same is true', of the Hexapoda also, the Myriapoda
still possessing limbs on all the trunk segments.

The comparison of the book-leaf tracheae with the book-like gills of the Xiphosura
seems far-fetched compared with their derivation from the tufted tracheae. The
assumption that the tubular tracheae in the Arachnoida have arisen independently of
those of the Protracheata and Antennata can only be resorted to as a makeshift. Mal-
pighian vessels are wanting in the Xiphosura. The sexual organs may emerge at very
different regions of the body in the Antennata, as \vas seen as early as in the Myria-
poda, and therefore no very great weight should be attached to the circumstance that
their position is almost similar in the Arachnoidea and the Xiphosura. The presence of
coxal glands, which emerge in the Arachnoidea and Xiphosura on the third pair of
legs, does not bear much upon this question, since, on the one hand, coxal glands may
occur on other pairs of legs as well in the Arachnoidea, and on the other, these glands
are very widely distributed among the Protracheata and Antennata (especially
Myriapoda}, and apparently were originally found in all the pairs of legs, as is .still
the case in the Protracheata.

In any case further investigations as to the relations between the Arachnoidea and
the Xiphosura cannot but be fruitful, and may throw much light upon the as yet
by no means solved problem of the relationship of the two groups.

The Pantopoda (Pycnogonidce) are also often considered as related to the Arach-
noidea, a view which was arrived at originally in consequence of the great similarity
in appearance of the two groups. This view had, however, to be abandoned when
their organisations were more closely compared (cf. p. 424).

1 See footnote on page 516.

542 COMPARATIVE ANATOMY CHAP.

Review of the most important Literature.

Anatomy.

Ph. Bertkau. Ueber den Generationsapparat der Araneiden. Archivf. Naturgesch.

41 Jahrg.
The same. JSeitrdge zur Kenntniss der Sinnesorgane der Spinnen. Arch. f. mikr.

Anatomie. 27 Bd.
The same. Ueber die Respirationsorgane der Araneen. Arch. f. Naturg. 38 Bd.

1872.
The same. Ueber das Cribrellum und Calamistrum. Em Beitrag zur Histiologie,

Biologie und Systematik der Spinnen. Archiv fur Naturgeschichte. 48 Jahrg.

1882.
The same. Ueber den Bau und die Function der sogen. Leber bei den Spinnen.

Arch. f. mikr. Anat. 23 Bd. 1884.
The same. Ueber den Verdauungsapparat der Spinnen. Arch. f. mikr. Anatomie.

24 Bd. 1885.
Edouard Claparede. fitudes sur la circulation du sang chez les Arantes du genre

Lycose. Memoir es Soc. Physique et d'Histoire natur. Geneva. 17 Bd. 1863.
The same. Studien an Acariden. Zeitschr. f. wiss. Zool. 18 Bd. 1867-1868.
G. Cuvier. Le Regne animal. Nouv. Edition. Paris, 1849. Insectes, Arachnides,

Crustacees von Audouin, Blanchard, etc.
L. Dufour. Histoire anatomique et physiologique des Scorpions. Mem. Acad. Scienc.

Savants etrangers. XIV. 1856.
The same. Anatomie, Physiologic, et Histoire naturelle des Galeodes. Mtmoires de

I' Acad. d. Sciences. Paris. Savants etrangers. XVII. 1858.
Hugo Eisig. Monographic der Capitelliden des Golfes von Neapel. Berlin, 1887.

(Contains the morphological significance of the coxal glands, spinning glands,
etc. of the Arachnoid ea.)
Batt. Grassi. I. Progenitori dei Miriapodi e degli Insetti. V. Intorno ad un nuovo

aracnide artrogastro. Boll. Societa entomol. italiana. XVIII. 1886.
Hermann Henking. Beitrdge zur Anatomie, Entwicklungsgeschichte und Biologie

von Trombidium fuliginosum. Zeitschr. fur. wiss. Zool. 37 Bd. 1882.
W. E. Hoyle. On a new Species of Pentastomum (P. protelis), from the Mesentery of

Proteles cristatus. Transact. Roy. Society, Edinburgh. Vol. XXXII. 1883.
G. Joseph. Cyphophthalmus duricarius. Berliner Entom. Zeitschr. 1868.
E. Ray Lankester. Limulus an Arachnid. Quart. Journ. Microsc. Science. N.S.

Vol. XXI. 1881.
E. Ray Lankester and A. G. Bourne. The minute structure of the lateral and of the

central eyes of Scorpio and of Limulus. Quart. Journ. Microsc. Science. N.S.

Vol. XXIII. 1883.
Rud. Leuckart. Ueber den Bau und die Bedeutung der sog. Lungen bei den

Arachniden. Zeitschr. f. wiss. Zool. 1 Bd. 1849.

The same. Bau und Entwickelungsgeschichte der Pentastomen. Leipzig and Heidel-
berg. 1860.
J. MacLeod. La structure des trachees et la circulation peritrache'enne. Brussels.

1880.
The same. Recherches sur la structure et la signification de Vappareil respiratoire des

Arachnides. Arch. Biolog. Tome V. 1884.

A. Menge. Die Shecrenspinnen. Schriften der naturf. Gesellsch. zu Danzig. 1855.
Albert .D. Michael. British Oribatidce. Ray Society. London, 1884.
Nalepa. Die Anatomie der Tyroglyphen. I. Abth. Sitzber. math.-naturwiss. Classe.

vi AEAGHNOIDEA— LITERATURE 543

Akademie Wissensch. Wien. 90 Bd. 1885. II. Abth. Ibid. 92 Bd. I. Abth.

1886.
Newport. On the structure, relations, and development of the nervous and circulatory

systems in Myriapoda and macrourous Arachnida. Philos. Transact. I. 1843.
E. Rossler. Beitrdge zur Anatomic der Phalangiden. Zeitschr. f. wiss. Zool. 36

Bd. 1882.
Robert von Schaub. Ueber die Anatomic von Hydrodoma. Ein Beitrag zur Kennt-

niss der Hydrachniden. Sitzber. AJcad. Wiss. Wien. math-naturw. Classe. 97

Bd. 1888.
Wladimir Shimkewitsch. Etude sur I'anatomie de VEpeire. Annales Scienc. natur.

6°. Tome 17. 1884*
Anton Stecker. Anatomisches und Histiologisches iiber Gibbocellum, einc neue Arach-

nide. Arch. f. Naturgesch. 42 Jahrg. 1876.
Alfred Tulk. Upon the anatomy of Phalangium Opilio. Ann. Magaz. Nat. Hist.

Vol. XII. 1843.
Bernh. Weissenborn. Beitrdge zur Phylogenie der Arachniden. Jenaische Zeitschr.

f. Naturwissensch. 20 Bd. N. F. 13. 1885 (with Bibliography).
Willibald Winkler. Das Herz der Acarinen nebst vergleichenden BemerTcungen iiber

das Herz der Phalangiden und Chernetiden. Arb. Zool. Inst. Univers. Wien.

7 Bd. 1886.
The same. Anatomie der Gamasiden. Arbeit. Zool. Inst. Universitdt Wien. 7

Bd. 1888.
Other authors : De Graaf, Loman, Krohn, Henking, Horn, Dahl, MacLeod, Ehlers,

Karpelles, Stecker, Oeffinger, Croneberg, Pelseneer, Bertkau, Lohmann, Kramer,

Haller, Menge, Parker, Pagenstecher, 0. P. Cambridge, Ray Lankester, Gulland,

Meckel, Plateau, Abendroth, P. J. van Beneden, Blanchard, Brandt, Gervais,

Megrin, Grube, J. van der Hoeven (on Phrynus), Leydig, Nicolet Lucas, Kittary,

Duges, Treviranus, Hutton, Hasselt, Koch, Blanc.
On Acarina many works of Kramer and Haller.

Ontogeny.

F. M. Balfour. Notes mi the development of the Araneina. Quart. Journ. Micr.

Science. Vol. XX. 1880.
Edouard Claparede. Recherches sur Devolution des araignees. NatuurTc. Verhandl.

Provinciaal Utrechtsch gcnootschap van Kunsten en Wetensch. Deel I. Utrecht

1862.
William Locy. Observations on the development of Agelena ncevia. Bullet. Mus.

Comp. Zool. Havard Coll. Cambridge. Vol. XII. 1886.
El. Metschnikoff. Embryologie des Scorpions. Zeitschr. f. wisscnsch. Zoologie. 21

Bd. 1870.
The same. Entwicklungsgcschichte von Chelifer. Zeitschr. f. wissensch. Zoologie. 21

Bd. 1870.
W. Schimkewitsch. Etude sur le developpement des Araignees. Arch. Biol. Tome

6. 1887.
Other authors : Rathke, Balbiani, Barrois, Herold, Ludwig, P. J. van Benedeu,

Claparede, Henking, Kowalevsky and Schulgin, Moriu, Jaworowski (in reference

to note on p. 516).

544

COMPARATIVE ANATOMY

CHAP.

Appendage to the Race of the Arthropoda.
The Tardigrada, or Bear Animalcule.

The body of these small animals, which does not exceed 1 mm. in
length, is cylindrical or a long oval ; it is outwardly unsegmented, and
carries 4 pairs of short truncated appendages armed with claws,
and not marked off from the body by joints. The last pair of these
appendages lies at the posterior end of the body. The most anterior
portion of the body is either narrowed like a proboscis, or marked off

FIG. 381.— Macrobiotus Hufel-
andii. Outlines of the body and
ventral chord. The supra -oeso-
phageal ganglion is not repre-
sented, ug, Infra-oesophageal gan-
glion ; 0!, 02, 03, 04, the 4 following
ganglia (after Plate).

FIG. 382.— Head of Doyerla simplex, from
the ventral side. The nervous system not drawn.
pa, Oral papillae ; mh, oral cavity ; ev, hypoder-
mal thickenings near the mouth (glands ?) ; z,
teeth; mr, oral tube; 'ph, pharynx; oe, oeso-
phagus ; md, mid-gut or stomach intestine ; bv,
hypodermal thickenings (leg glands? coxal
glands?) in the feet (after Plate).

like a head. The body is covered by a probably chitinous cuticle,
which is thrown off from time to time (ecdysis). The mouth lies at
the anterior, and the anus at the posterior end of the body. In the
straight digestive tract which passes through the ccelome the 3 well-
known regions, the fore-gut, mid-gut, and hind-gut (rectum) may be
distinguished, The oral aperture, which is surrounded by papillae, and
in some cases by setae, also leads into an oral cavity, into which project
the pointed and sometimes calcified ends of 2 teeth. Two pear-shaped
or tube-like glands (salivary glands ? poison glands ?) enter the oral
cavity. The oral cavity is followed by a generally narrow oral tube,
which swells out at its posterior end into a muscular, spherical, or
egg-shaped oesophageal bulb (pharynx). Between the mid-gut (stomach)

VI

TARDIGRADA

545

and the oesophageal bulb an oesophagus is intercalated. Two blind
tubes enter the rectum, no doubt corresponding with the Malpighian
vessels of the Tracheata.

The sexes are separate ; the germ glands unpaired, and sac-
shaped. In both sexes they enter the hind -gut, which thus becomes
a cloaca. Special circulatory and respiratory organs are wanting.
The nervous system consists of a brain, an infra -oesophageal gan-
glion which is connected with the brain by two oesophageal commis-
sures, and 4 other ventral ganglia, which are connected by longitudi-
nal commissures placed far apart from one another. There are two
eye spots in the head, lying on two
small ganglia connected by nerves with
the brain. The musculature is richly
developed. Various dorsal, ventral,
and lateral longitudinal muscles run
under the integument. Special muscles
serve for moving the legs. All the
muscles are smooth.

The systematic position of the Tardigrada
is uncertain. The three related facts that
they possess accessory organs of the hind-gut
comparable with the Malpighian vessels, tube-
like oral glands, and truncated feet provided
with claws, make it not improbable that
they belong -to the Arthropoda, and especially
to the Tracheata. The want of oral limbs,
the structure of the nervous system, and the
manner of emergence of the sexual organs,
stand in the way of a comparison of the Tar-
digrada, with the Acarina. Although we
may agree with the view that they are some-
how related to the Tracheata, or to the ances-
tors of the Tracheata, we cannot in any case assume that they, in any way, resemble
the primitive arrangements. The want of a blood-vascular system, the unpaired
germ glands, the reduced and abbreviated condition of the whole body, the absence
of nephridia and of coxal glands (?), rather make the Tardigrada appear as a one-
sidedly developed lateral branch.

Most of the Tardigrada live among moss and lichens, a very few in fresh or salt
water. They can stand desiccation, and remain apparently dead for a long time,
reviving again when wetted. Echiniscus, Macrobiotus, Milnesium, Doyeria.

Fig. 383. — Posterior portion of the
body of Macrobiotus Hufelandii $ , from
the side, h, Testis ; ad, accessory gland of
the male sexual apparatus ; d, cloaca ; an,
anus ; vm, excretory tube (Malpighian
vessel) ; md, mid-gut (after Plate).

Literature.

Ludwig H. Plate. Beitrage zur NaturgescMchte der Tardigraden. Zool. Jahrbucher
von Spengel, Abth.fur Anatomic und Ontogenie. 3 Bd. 1888. In this treatis
there is a bibliography up to 1888.

VOL. I

2 N

INDEX

Numbers in Italics give Systematic Position,
refer to Figures.

Numbers in Black Type

ABDOMINALIA, 292

Abyla, 70

Acantharia, 6

Acanthia, 440

Acanthin, 16

Acanthocephala, 180

Acanthodrilus, 181

Acanthometra, 6

Acanthocystis, 6

Acarina, 510

Accessory intestine (Vermes), 206

Acheta campestris, sexual organs, 486

Acliromatin, 35

Achtheres, 291

Acidaspis, 415

Acineta, 11

Accela, 134

Acone eyes (Insecta), 470

Acontia, 83

Acotylea, 133

Acraspeda, 71

Acridiidce, 439

Acridium tartaricum, 474

Acronyctidce, 44%

Actinaria, 71

Actinia, 71

Actinocephahis, 9

Actinophrys, 6

Actinophrys sol, 6

Actinosplicerium, 6

Actinotrocha, 272

Aculeata, 44%

Adamsia, 71

Adradii, 74

dSga, 295

JSgineta, 69

jEginopsis, 69

JElosoma, 181

Forskalea, 68 ; auditory vesicles,

95

JEschna, 439 ; sexual organs, 485
Agalma, 70
Agalmopsis, 113

Agelena, 510

,, labyrinthica, embryo, 539,
Aglantha digitalis, 69
Agnostus, 415
Agrion, 439
Agrotidce, 442
Alary muscles, 475
Alcippidce, 292

Alcyonaria, 70; gemmation, 107
Alcyonella, 184
Alcyonium, 70
Alcyonidium, 184
Alcyopidce, 182
Aletes, 510
Aletia, 463

Alima type of larva, 393
Allantonema mirabile, 200
Alloiocoela, 134
Allolobophora, 181
Atturus, 181
Alma nilotica, 246
Alpheus, 298
Alternation of generations (Protozoa), 21 ;

(Cnidaria), 115 ; (Vermes), 267 ; (An-

tennata), 502
Alydus, 440
Amalthceidce, 90
Ametabole, 491
Ammothea, 4% 4
Amnion, 434, 493
Amoeba, 4

,, polypodia, 3, 12
Amcebina, 4
Amphiasters, 35
Amphifflene, 234
Amphilina, 134
Amphinomidce, 182

Amphipoda, 296 ; circulatory system, 362
Amphiporus, 178

Moseleyi, 220

Amphipneustic tracheal system, 482
Amphistoma, 134 '
Amphtthoe penicillata, thoracic limbs, 318

548

COMPARATIVE ANATOMY

Amphitrite, 182
Anachceta, 181
Analges, 510
Anceidce, 295
Anchorella, 291
Anchylostoma, 180
Androctonus, 509
Anemonia, 71
Anguillula aceti, 178
Anilocra, 295 ; pleopoda, 324

Anisospore, 20

Annulata (Annelida), 180 ; diagram of
pharynx, 202, 203 ; section, 237

Anomura, 298

Anonymus, 133

Antennata, 438-508

Anthea cereus, 50

Anthemodes, 70

Anthomedusce, 68

Anthophora, 492

Anthozoa, 70

Antipatharia, 71

Aphaniptera, 441

Aphanoneura, 181

ApMdce, 440

Aphis pelargonii, tracheal system, 481

Aphroditea, 182

Aphrophora, 441

Apidce, 442

Apis melijica, anatomy, 462, 463 ; sexual
organs, 487

Aplysina, 62

Apoda, 292

Apolemia, 70

Apneustic tracheal system, 481

Apseudes, 293 ; auditory hair, 356 ; thor-
acic limbs, 318

Apseudes Latreillii nervous system, 346

Apterygota, 439

Apus, 288 ; section, 315 ; larvae, 383
,, longicaudatus, trunk feet, 316
,, lucasanus, mandibles, 311

Arachnoidea, 509-543

Araneidce, 510 ; book-leaf tracheae, 630 ;
sexual organs, 533

Arcella, 4

„ vulgaris, 3

Archceostraca, 293

Archenteron, 37

Archiannelida, 181

Archigetes, 134

Arenicolidce, 182

Argas, 510

Argiope, 185 ; larva, 273

Argulidce, 291

Argulus, 291

„ Corregoni nervous system, 346
„ foliaceus, 291

Argyroneta, 510

Aridities, 182

Arthrobranchiae, 328

Arthropoda, 287-545

Arthrostraca, 293 ; thoracic limbs, 318

Articulata, 287

Asaphus, 415

Ascandra, 60

Ascaridce, 180 ; section, 194

Ascaris lumbricoides, 180 ; genital appara-
tus, 255
„ nigrovenosa, 179

A scones, 62

Asellidce, 295 ; thoracic limbs, 318

Asellus aquaticus, enteric canal, 340; geni-
tal apparatus, 374 ; nervous system, 346

Asilidce, 443

Aspidiotus, 440

Aspidogaster, 143

Asplancha, 185

Astacidce, 298

Astacus fluviatilis, 298, 298, 299 ; 2d
antennae, 310 ; antennules, 308 ; anterior
maxillae, 312 ; cephalo-thorax, 326 ; for-
ceps, 335 ; gills, 327 ; mandibles, 311 ;
muscles, 334 ; nervous system, 346 ;
ontogeny, 399-406 ; pleopoda, 324 ; pos-
terior maxillae, 313 ; sections, 338, 365 ;
segments, 333 ; sexual organs, 375 ;
thoracic feet, 323

Asterias glacialis, eggs, 31, 33

Asterope, 182

Astrcea, 71

Astroides, 71

Atax, 510

Atractonema, 200

Atractosoma, 438

Attus, 510

Atypus, 510

Aulactinium, 7

,, actinastrum, 7

Aulosphcera, 7

Aulostomum, 181

Auralia, 70

Aurelia aurita, 72 ; development of, 130

Auronecta, 70

Aurophore, 70

Autolytus, 182

Avicularia, 510

Avicularia, 267

j\.xolotl, 119

Axopodia, 14

BACILLUS, 439

Baetis binoculatus, segment, 483

Balanidce, 292

Balanoglossus, 119

Balantidium, 9

Balanus, 292

„ Hameri, 303

„ perforatus, trunk feet, 316

„ tintindbulum, 304
Bdellidce, 510
Belinurus, 417
Beroidce, 72
Bibionidce, 443
Biogenesis, 118

INDEX

549

Bipalium, 133
Birgus, 299

„ latro, sections, 328, 367
Bittacus, 441
Blastocoel, 123
Blastoderm, 126
Blastomeres, 121-131
Blastopore, 57, 123
Blastula, 36, 57
Blatta, mouth parts, 446

„ orientalis, nervous system, ?468
Blattidce, 439

Bombycina, 44%

Bombylliidce, 443

Bonellia, 182 ; egg segmentation, 124

Bopyridce, 295

Boreomysis scyphops, 321

Boreus, 441

Borlasia, 178

Bostrychidce, 441

Bothriocephalus latus, 134 '•> scolex, 164
,, punctatus, 154

Bougainvillea ramosa, 68, 67, 104

Brachionus, 185

Brachiopoda, 184, 263

Brachonidce, 44%

Brachycera, 44$

Brachydesmus, 4^8

Brachyura, 299 ; development, 396

Bract, 112

Branchellion, 181

Branchial formula, Astacus, 329 ; Cancer
pagurus, 329

Branchiata, 287

Branchiobdella, 181

Branchiomma, 247

Branchiopoda, 288

Branchiostegite, 307

Branchipus, 288 ; mandibles, 311 ; section
of eye, 354 ; tactile
hair, 356 ; ontogeny,
398
,, stagnalis, 191

Branchiura, 291

Braula, 443

Brood capsule, 316 ; care of, 379 ; cavity,
380 ; chamber, 434 ; lamella, 321, 378 ;
plate, 319 ; pouch, 318

Bronteus, 415

Bryozoa, 183

Bugula, 184

Bunodes, 411

Buprestidce, 44%

Buthus, 509

,, occinatus, anatomy, 517

C^NOGENESIS, 118

Calanella mediterranea, 353
Calappa, 299
Calcaria, 60
Calceoli, 356
Calcispongice, 60

Caligus, 291

Callianassa, 298

Callianira, aboral pole and sensory body,

97 ; development, 131
Calliaxis, gills of larva, 322
Callizona grubei, eye of, 230
Calopteryx, 439
Calyconecta, 70
Calymene, 415
Calymna, 6
Calyptopis larva, 389
Campanaria, 68
Campanularia geniculata, 68
Campanulina tennis, 68
Campodea, 439

,, staphylimis, 444
Campylaspis nodulosa, mandibles 311
Cancer, 300
Cancrion miser, 379
Cannopilus, 7
Cannorhiza, 72

,, connexa, 85

Cannostomce, 72
Cantharidce, 441
Canthocamptus, 291
Capitellidce, 181
Capitibranchiata, 182
Capnia nigra, larva and imago. 456
Caprella acutifrons, 295
Caprellidce, 296
Capsus, 440
Carabidce, 442

Carabus auratus, digestive apparatus, 461
Carcinus, 300
Carchesium, 11
Carididce, 298
Carina, 304
Carinella, 178
Carmarina hastata, 69
Caryophyllceus, 134
Caryophyllia, 71
Cassiopcea, 72
Catallacta, 11
Cathammal plates, 74, 78
Catometopa, 300
Cecidomyia, 502
Cellepora, 184
Cellulose, 16
Centrotus, 441
Cephalodiscus, 184
Cephalogaster, 341
Cephalo-thorax, 302
Cepheus tegiocranus, sexual organs, 533
Cerambycidce, 441
Ceraspongice, 62, 61
Ceratiocaridce, 293
Ceratium 8,

„ Tripus, 9
Ceratopsyllus, 441
Cercaria, 169
Cerebratulus, 178
Cerianthus, 71
Cestidce, 72

550

COMPARATIVE ANATOMY

Cestoda, 134

Cestoplana, 133

Cetochilus, 291 ; development of, 385

,, septentrionalis, 385 ; ontogeny,

398

Cetonia aurata, sexual organs, 485
Chcetogaster diaphanus, anterior part of

body, 221
Chcetogastridce, 181
Chcetognatha, 185
Chcetopoda, 181
Ch&tozone, 253
Chalaza, 30
Challengeria, 7
Charybdeidce, 72
Cheirurus, 415

,, Quenstedtii, 414
Chelicerae (Xiphosura), 417; (Arachnoidea),

515

Chelicerota, 509
Chelifer, 509

,, Bravaisii 513
Chermes, 440
Chernetidce, 509
Chilaria, 417
ChUodon, 9
Chilognatlia, 438
Chilomonas, 8

,, Paramcecium, 8

Chilopoda, 438
Chilostomata, 264

Chironomus plumosus, nervous system, 466
Chirodropus, 72
ChloS, 439 ; segment, 383
Chlorcemidce, 182
Chloragogen cells, 213
Choanqftagellata, 8
Chondracanthus, 291
Chondrosia, 61
Chordeumidce, 438
Chordotonal ligament, 472 ; organs, 472,

473

Chorion 28
Chromatin, 35
Chrysaora, 101
Chrysomelidce, 441
Chrysopa, 441
Chthonius, 509
Cicada, 441
Cicindelidce, 442
Ciliata, 9
Ciliofiagellata, 8
Cirolana, 295

„ spinipes, posterior maxillie, 313
Cirratulidce, 182
Cirri, 188
Cirripedia, 291
Cistelidce, 441
Citigrada, 510
Cladocera, 289
Cladocora, 71
Cladocoryne, 90
Cladonema radiatum, 68

Clathria, 61
Clathrulina, 6
Clausocalanus, 291

,, mastigophorus, 290

Clepsidrina, 9
Clepsine, 181
Clistomastus, 266
Clitellio, 181
Clitellum, 192

Cloen dimidiatum, segments of larva, 457
Clubiona, 510
Cnidaria, 66-132
Cnidoblast, 39
Coccidce, 440
Coccinellidce, 441
Ccdenterata, 58-132
Coeloblastula, 123
Coelogastrula, 123
Cceloplana Mecznikowi, 136
Coaloplanula, 126
Ccenurus, 172

,, cerebralis, 135
Coleoptera, 441
Collembola, 439
Collosendeis gigas, 4%4
Collosphcera, 6
Collozoum, 6
Columella, 99
Complementary males (Cirripedia), 305 ;

(Myzostomidce), 266
Conchoderma, 292
Conilera cylindracea, 361
Conocephalites, 415
Conopidce, 443

Convoluta, 134 > pharyngeal apparatus, 139
Copepoda, 290

Corals (Anthozoa), 70 ; section, 76
Cordilophora lacustris, 68 ; stinging cells,

81
Corethra, imaginal discs, 499

,, plumicornis, nervous system, 472
Coreus, 440
Corixa, 440
Corizus, 440

Cormidium, 70, 112, 113
Cornutella, 7

Corophium, 296

,, longicorne, thoracic limbs, 318

Corrodentia, 440
Cortina, 7

„ typus, 7
Corycceus, 291
Corydalis, 441
Cossidce, 442
Cotylea, 133
Cotylorhiza, 72
Crab's eyes (gastroliths), 337
Crambessa, 72
Crangon, 298
Crania, 185

Craspedota, 67, 69, 73 ; diagram, 73 ; for-
mation of, 105

INDEX

551

Crayfish, see Astacus fluviatilis

Crevettina, 296

Cribrellum, 522

Criodrilidce, 181

Cristatella, 184

Crustacea, 288-413

Cryptoniscidce, 296

Cryptopentamera, 441

Cryptophyalus, 292 •

Cryptops, 438

Cryptotetrama, 441

Ctenaria ctenophora, 68, 80

Cteniza, 510

Ctenodrilus, 181

,, monostylus, 267

„ pardalis, 267

Ctenopliora, 72 ; adhesive cells, 81 ; de-
velopment, 131 ; segmentation of egg,
125 ; sensory body, 79, 96 ; swimming
plates, 79, 81

Ctenoplana Kowalevskii, 136, 137

Cubomedusce, 72

Cuculliadce, 442

Culex, mouth parts, 450

Culicidce, 443

Culiciformes, 443

Citmacea, 297

Cunina, 69

,, lativentris, tentaculocysts, 95

Curculionidce, 441

Cyamidce, 296

Cyanea, 72

Cydometopa, 299

Cydoporus, 133

Cyclops, 291

,, coronatus, posterior maxillae, 313
,, serrulata, antennules, 308
,, signatus, 2d antennae, 310
,, tenuicornis, mandibles, 311

Cyclostomata, 264

Cydippidce, 72

Cylicobdella, 259

Cylindrostoma, 156

Cymothoa oestroides, sexual organs, 381

Cymothoidea, 295

Cynipidce, 44%

Cyphonantes, 271

Cyphophthalmidce, 509

Cypridina, 290

,, mediterranea, 289

,, messinensis, posterior maxillss,

313
,, stellifera, anterior maxillae, 312

Cypris, 290

Cypris larva, 387, 386, 388

Cyrtopia larva, 389

Cysticercus, 171

,, celluloses, 135, 172
,, fasciolaris, 135
,, pisiformis, 135

Cystoflagellata, 8

Cystonecta, 70

Cythera, 290

Cythera viridis, anterior maxillae, 312
Cytod, 2

DANAIS, 463

,, archippus, anatomy, 488
Daphnia, 290

,, pulex, antennules, 308

,, similis, 289 ; anterior maxillae,

312 ; trunk feet, 316
Daphnidce, 289
Dasybranchus, 181
Decapoda, 298
Decticus, 439
Delamination, 126
Demodicidce, 510
Dendrobcena, 181
Dendroccelum, 133
Dendrocometes, 11
Dendrometridce, 44%
Dermaleichidce, 510
Dermaptera, 439
Dero, 181
Desor's larva, 275
Deutomerit, 9, 16
Deutoplasm, 26
Diastylis, 297

,, stygia, 295 ; thoracic feet, 319

pleopoda, 324
Dictyastrum, 6
Dictyna, 510
Dicyemidce, 58 -
Dicyema, 59
Difflugia, 4

,, pyriformis, 3
Digenetic Trematoda, 134
Dimorphism, sexual (Vermes), 265; (Crust-
acea), 376 ; (Antennata), 490
Dinoflagellata, 8
Dinophilus, 185

,, apatris, 265

,, gyrociliatus, 246

,, vorticoides, 265

Diopatra, 182
Diphyes, 70
Diplopoda, 438
IHplozoon, 134
Dipneumones, 510
Diptera, 443 ; mouth parts, 450
Discalia, 70
Discocelis, 120
Discodrilidce, 181 ; segmentation of egg,

125

Discogastrula, 128
IHscomedusce, 72
Disconantha, 70
Disconecta, 70
Discophora, 180
Discoplanula, 128
Distoma, 134

, , divergens, water- vascular system,
153

,, hcematobium, 154

hevaticum 134 \ life-history 169

552

COMPARATIVE ANATOMY

Distoma isostomum, 134 > intestinal and

nervous system, 143
,, lanceolatum, 134 > sexual organs,

158

Dochmius duodenalis, 180
Dorippe, 299

Doropygus porcicauda, trunk feet, 316
Dorsdbranchiata, 182
Doyeria simplex, head, 544
Drassus, 510
Drepanophorus, 178

,, Lankasterii, nervous sys-

tem, 216
Dromia, 299
Dwarf males (Cirripedia), 305 ; (Crustacea

generally), 377
Dysderidce, 510
Dysmorphosa carnea, 68
Dytiscidce, 44%
Dytiscus, section through ocellus, 470

EARTHWORM, see Lumbricus
Ecardines, 185
Ecdysis, 458
Echinococcus, 172

,, veterinorum, 135

Echinodera, 186
Echinoderm, 119
Echinorhynchus, 180 ; genital organs, 257 ;

organisation, 258
,, gigas, 180

Echiuridce, 182

Echiurus, enteric canal, vascular system,
and nephridia, 207; nervous system,
223

Ectoparasitica, 134
Ectoplasm, 13
Ectoprocta, 184
Ectosarc, 13
Edriophthalmata, 293
Edwardsia, 71

Eggs, animal, 25 ; parthogenetic, 32 ;
segmentation, 121-131 ; summer (Crust-
acea), 376,'382 ; types of, 27-32 ; winter
(Crustacea), 376, 382
Elateridce, 442

Elytra (Vermes), 189 ; (Antennata), 441
Embidce, 439
Embole, 123
Empidce, 44$

Empis stercorea, nervous system, 466
Enchytrceidce, 181
Ency station, 16
Endite, 315
Endomychidce, 441
Endoparasitica, 134
Endoplasm, 13
Endopodite, 307
Endosarc, 13
Enoplidce, 178
Entomostraca, 288
Entoniscus, 296
Entoprocta, 184

Epeira, 510

Ephe?neridce, 439 ; sexual organs, 485 ;

larva, 483
Ephippigera, 463
Ephippium, 380
Ephyra, 72
Epibole, 124
Epimerit, 9, 16
Epipodite, 315
Epistylis, 11

Equitidce, 442 •

Erichthus larva, 393, 394
Eriphia, 300
Errantia, 182
Erscea, 70, 113
Eschara, 184
Estheria, 288
Eucharis, 72
Eucone eyes, 470
Eucope campanulata, 68, 74, 102
Eucopepoda, 290

Eucopia australis, posterior maxillae, 313
Eudendrium ramosum, 68
Eudorina, 8
Eudoxia, 70, 113
Eudrilus, 181
Euglena, 8
Eulimnadia Agassizii, 2d antennae, 310

,, texana, 2d antennas, 310

Eunice limosa, 188
Eunicidce, 182
Eupagurus, 298
Euphausia pellucida, 2d antennae, 310 ;

anterior maxillae, 312 ; nervous system,

346 ; pleopoda, 325
Euphausidce, 298 ; larval history, 389 ;

larvae, 390
Euphrosyne, 182
Euplectella, 61
Euplotes, 9
Eupomatus uncinatus, egg segmentation,

123 ; larva, 269
Euprepiadce, 44%
Eurylepta, 133
Eurypteridce, 415
Euscorpius, 509

,, italicus, eye, 521
Euspongia qfficinalis, 62
Eustrongylas gigas, 180
Evadne, 290
Evaniadce, 442
Exogone, 182
Exopodite, 307
Exotheca, 99
Exumbrella, 73
Exuvia, 458
Eye-spots (Cnidaria), 75, 97 ;' (Protozoa),

18 ; x-shaped (Crustacea), 352 ; (Pla-

todes), 169

FABRICIA, 234
Facet-eyes, 470
Falces, 515

INDEX

553

Fat-body (Crustacea), 370 ; (Antennata),

477

Filaria medinensis, 179
Filariidce, 178

Filistata testacea Latr., pedipalp, 535
Finn, 135, 166, 171, 172
' ' Fish trap " apparatus (Pantopoda), 422 ;

(Antennata), 451
Flabellum, 71
Flagellata, 7
Flame-cell, 152
Floscularia, 185
Flustra, 184

Forficula auricularia, sexual organs, 485
Forficulidce, 439
Formicidce, 442
Forskalia, 70
Fossoria, 44%
Fredericella, 184
Freia, 9
Fulgora, 441
Fungia, 71
Fungicolce, 44$
Funiculus, 215
Furcilia larva, 389

GALATEID^;, 299

Galeodes, 509

,, barbarus, sexual organs, 533, 534

,, Dastaguei, 511

,, nigripalpis, sexual organs, 534
Galleria, 463
Oallicolce, 448
Gamasidce, 510
Gamasus, nervous system, 520 ; digestive

apparatus, 524

,, crassipes, sexual organs, 533, 534
,, fucorum, larva, 514 ; heart, 527
Gametes, 20
Gammarus, 296
Gasterostomum, 162
Gastrsea theory, 56
Gastrceidce, 58-60
Gastroblasta Ra/celii, 68
Gastroliths, 337
Gastrophysema, 58
Gastrotricha, 186
Gastrula, 37, 57, 118
Gastrulation, 120-131
Gebia, 298
Gecarcinus, 300
Gemmulae, 65
Geocores, 440
Geodesmus, 133
Geodia, 61
Geometrina, 442
Geonemertes palaensis, 199
Geophilidce, 438
Gephyrea, 190

Gerardia, 71 ; horn skeleton of, 99, 100
Germaria, 25, 155

Germinal disc, 128 ; spot, 26 ; vesicle, 25
Geryonia, segmentation of egg, 126

Geryonia proboscidalis, 69

Giant nerve tubes (Crustacea), 351 ; (Ver-

mes), 222
Gibbocellum, 509

,, Sudeticum, 529

Gigantostraca, 415, 416
Glands, 40

,, anal (Protr.), 431

,, autennal (Crust.), 368, 369

,, cement (Antenn.), 487; (Crust.),

369 ; (Pantop. ), 424
,, coxal (Arach.), 522; (Protr.),

432; (Xiph.), 420
,, granular, 160
,, green (Decapoda), 368
„ hook (Caprella), 331
., Krohn's (Arach.), 522
„ leg (Antenn.), 459 ; (Crust.), 330 ;

(Protr.), 432
,, Morren's (Oligochseta), 206

oil (Arach.), 522
,, pedipalp (Arach.), 522
,, pharyngeal (Oligochseta), 202
,, poison (Antenn.), 459 ; (Aran-

eidaa), 522 ; (Vermes), 199
,, rectal (Antenn.), 459 *
,, salivary, (Antenn.), 458 ; (Arach.),
525 ; (Crust), 336 ; (Platocles),
140 ; (Protr. ), 429 ; (Vermes),
202

,, septal (Oligochseta), 202
,, shell (Crust.), 368, 369; (Plat-
odes), 158
,, slime (Antenn.), 487 ; (Protr.),

432
,, spinning (Antenn.), '458 ; (Aran-

eidae), 522 ; (Vermes), 192
„ stigmatic (Arach. ), 522
,, stink (Antenn.), 459; (Arach.),

522

,, ventral (Crust.), 330
,, wax (Antenn.), 459
Globigerina, 4
Glomeridce, 4$$

Glomeris marginata, sexual organs. 486
Gflyeeridce, 182
Glyptonotus, 348
Gnathobdellidce, 1S1
Gnathochilarium, 447
Gnathophausia, 349
Gnathostomata, 311
Gonades, 102
Gonangia, 68
Gonochorism, 24
Gonophore, 105, 106, 111
Gonyleptus, 509
Gordiidce, ISO
Gordius, 256

,, aquaticus, 180
,, Preslii, 236
Gorgonia, 70
Graffilla, 134
Grapsus, 300

554

COMPARATIVE ANATOMY

Gregarina, 8
Gromia, 4

,, oviformis, 5
Gryllidce, 439
Gryllotalpa, 439
Gunda segtnentata, 133
Gyge, 296

Gynoecophorus Iwematdbius, 134
Gyrator, 157
Gyrodactylus, 134

H&MATOPINUS, 440
Hcemopis aulastomum, 181
Halacaridce, 510
Halesus, 441
Halichondrina, 61
ffaliphysema, 58
Haliscara, 61

,, lobularis, section, 65
Halistemma, 70
Holla, 182
Halobates, 440
Halocypris, 290
I Halycryptus, 183
Halteres, 473
Haplosyllis, 182

,, spongicola, 266

Harpes, j.15
Heliconiidoe, 442
Heliozoa, 6
Hemerobius, 441
Hemiaspidce, 416
Hemiraetabole, 491
Hemiptera, 440 ; mouth parts, 451
Henicops, 438
Hermaphroditism, 24

„ protandrous, 65

Hermellidce, 182
Hermione, 182
Hesionidce, 182
Hesperidce, 442
Heterogeny, 171
Heteromera, 441
Heteroptera, 440
Heterotricha, 9
Hexacorallia, 71
Hexactinellidce, 61
Hexapoda, 438
Himantarium, 438
Hippidce, 299
Hippolyte, 357
Hippopodius, 70

Hirudinea, 180 ; eye, 232 ; section, 249
Hirudo medicinalis, 181 ; genital organs,

258 ; intestinal canal, 206 ; nervous

system, 218. 219, 220
Histriobdella, 181
Holometabole, 491
Holopneustic tracheal system, 480
Holotricha, 9
Homarus, 298
Homoptera, 441
Honey-bee, see Apis

Honey-stomach, 461
Hoplonemertina, 178
Hormiphera, 72

,, plumosa, 79

Hornera, 184
ffyalonema, 61
Hyaloplasm, 26, 49
Hyalosphenia, 4

lata, 3
Hydatina, 185

,, senta, 245

Hydra, 67 ; neuromuscular cells, 46
ffydrachnidce, 510
Hydridce, 67
Hydracorallia, 67
Hydrocores, 440
Hydrodoma, 510
Hydromedusce, 67
ffydrometra, 440
Hydrophilidce, 442
Hydrophilus, embryos, 455 ; development,

494, 495, 497, 498
Hydropolyp, 66 ; diagram, 73
Hydropsyche, 441
ffydrozoa, 66
Hydrula, 66
Hymenocaris, 293
Hymenoptera, 422 ; mouth parts of larva,

449

Hyperina, 296
Hypotricha, 9

IAPIX, 439 \ ^

Ibla, 292

Ichneumonidce, 44%

Ichthydium, 186

Idotheidce, 295

Ilia, 299

Imago, 490

Imperforata, 4

Inachus, 299

Infusoria, 9

Ingluvies, 461

Insecta, 438

Interarticular (intersegmental) membranes,

300

Interradia, 74
Interseptal spaces, 88
Intima, 336
Intraseptal spaces, 88
lolanthe acanthronotus, 2d antennae, 310
Isis, 70

Isopoda, 294 ; circulatory system, 360
Isospore, 20

Isopteryx apicalis, chordotonal organ, 473
Isotoma, 439
lulidce, 438

lulus Sabulosus, sensory organs, 474
Ixodidce, 510

KARYOKINESIS, 35
Kentrogonidce, 292
Kentrogon larva 387, 386

INDEX

555

KopJwbelemnon, 70

,, Leuckartii, 71

LABIDURA, 439

Lacenularia, 185

Lacinia, externa, 312, 313 ; interna, 312,

313

Lambrus, 299
Lamellicornia, 44%
Lamodipoda, 296
Langia, 178
Lanice, 182

,, conchilega, 241
Laomedia Caliculata, 6S
Laterigrada, 510
Laurer's canal, 159
Lecanium, 440
Ledra, 441
Leiobunum, 509
Lemnisci, 248
Leiosoma, 510

Lepadidce, 292 ; organisation, 303
Lepas antifera, 303

, , fasciculata, Cypris-like larva,

388

,, pectinata, pupa, 388

Lepidoptera, 44%
Lepisma, 439
Leptidce, 443
Leptochelia, 293
Leptodiscus, 8
Leptodora, 290
Leptomedusce, 68 1 } 7
Leptoplana, 133 ; sexual organs, 156
Leptostraca, 292 ; pleopoda, 324
Lermcea, 291

„ branchialis, posterior maxillae,

313
Lernceascus, 291

,, nematoxys, 377

Lernceocera, 291

,, esocina, 380

Lernceopodidce, 316

Lernanthropus, 291 ; genital organs, 372
Leucandra, 60
Leucocytes, 500
Leucones, 63
Libellulidce, 439
Lichas, 415
Ligula, 134

Limnadia, 288 ; nervous system, 346
Limnceus truncatulus, 169
Limnobiidce, 443
LimnocytJiere incisa, posterior maxillae,

313 ; trunk feet, 316
Limnodrilus, 181
Limnophilus, 441

Limulidce, 4%1 > section of eye, 419
Limulus moluccanus, 421

,, polyphemus, J$l, 418 ; trilobite

stage, 421 ; section, 419
Linens, 178
Linguatulidce, 510

Lingula, 185 ; embryo, 273
Linyphia, 510
Liothetim, 440
Liparidce, 44%
Listrophorus, 510
Lithobiidce, 438
Lithdbius forficulatus, 464

,, validus, 447
Lithodes, 299
Lituola, 4
Lima, 440
Lizusa octocilia, 68
Lobatce, 72

Locusta viridissima, tibia of fore leg, 473
Locustidce, 439
Lopadorhynchus, 280
Lophogaster typicus, thoracic feet, 321
Lophogastridce, 298
Lophophore, 183

Lophopus, 184 J anterior body, 184
Loricate, 410
Loxosoma, 184
Lucernaria, 71
Lucifer, 298 ; mandibles, 311
Lumbricidce, 181
Lumbriculidce, 181

Lumbricus, 181 ; embryo, 276 ; germ
streaks, 278, 279 ; section,
250

,, agricola, genital organs, 260

,, terrestris, anterior body, 251

Lyccenidce, 442
Lycosidce, 510
Lygceus, 440
Lymph gills, 246
Lynceus, 290
Lysianassa, 296

,, producta, pleopoda, 324
,, umbo, posterior maxillae, 313

Lysopetalidce, 438
Lysiopetalum carinatum, gnathochilarium,

447
Lysiosquilla maculata, posterior maxillae,

313

MACHILIS, 439

,, maritima, nervous system, 465 ;
tracheal system, 481 ; ven-
tral shield, 454

Macrdbdella, sensory organ, 232

Macrobiotus Hufelandii, hinder body, 545 ;
outline, 544

Macrolepidoptera, mouth parts, 448

Macromeres, 122-131

Macronucleus, 9, 18

Macrorhynchus, 134

,, croceus, proboscis, 151

Macrospores, 20

Macrostoma, 134

Macrotoma, 439

Macrura, 298

Madreporaria, 71 ; diagram of structure,

556

COMPARATIVE ANATOMY

Moeandrina, 71
Magosphcera, 11 ; planula, 11
JUaja, 299 ; Zocea of, 396

,, squinado, nervous system, 346
Mala interna and exterua, 446
Malacobddlina, 178
Malacodermata, 44%
Malacostraca, 292 ; pleopopa, 324
Mallophaga, 440
Malpighian vessels, 460, 462 ; (Arachn. ),

525; (Tardigr.), 545
Mantidce, 439
Mantispa, 441
Margelis ramosa, 68, 67
Marginal lobes, 71, 92 ; vesicles, 75
Mastigophora, 7
Mastobranchus, 253
Mecostethus, 439
Medusce, 67, 71
Megalopa, larva, 396
Megaloptera, 441
Melicerta, 185
Meloidce, 441

Melolontha vulgaris, sexual organs, 486
Melophagus, 443

,, ovimus, sexual organs, 486

Mermecophila, 439
Mermis nigrescens, 179
Mermithidce, 179
Merocytes, 126
Merostomce, 1+15
Mesenterial filaments, 75, 83
Mesostoma, 134 > intestinal and nervous
system, 142 ; pharyngeal
apparatus, 139
,, Ehrenbergii, sexual {organs,

158

Mesothorax, 444
Meta, 510
Metagenesis, 115
Metanauplius larva, 389
Metapneustic traclieal system, 482
Metathorax, 444
Metazoa, 4 ; divisions of, 57
Microlepidoptera, 442
Micromeres, 122-131
Micrommata, 510
Micronucleus, 9, 18
Micropteryx, 442
Micropyle, 28
Microspores, 20
Microstoma, 134

„ linear e, 166

Microthdyphonidce, 509
Miliola, 4, 5
Millepora, 67
Miris, 440
Misidce, 297
Mitrocoma, 94
Moina, 290

„ rectirostris, ontogeny, 397
Monas, 8
Monera, 2, 4

Moniligaster, 181

Monocystidce, 9

Monogenetic Trematoda, 134

Monostomum, 134

Monothalamia, 16

Monotus, 134

Monozoa, 134

Mutter's larva, 168

Munnopsis typica, antennules, 308

Muscidce, 443 ; development, 500, 501

Mygale, 510 ; anatomy, 531

,, ccementaria, digestive apparatus,

524

Myoblasts, 49
Myophriscs, 15
Myrianidce, 182
Myriapoda, 438
Myrmeleon, 441

Mysidce, 297 ; antennal gland, 369
My sis flexuosa, thoracic feet, 321

,, relicta, 349

,, larva, 298, 392
Mysodopsis, 365
Mystacides, 441
Myxicola, 182
Myxilla, 61
Myxodiction, 4
Myxopodia, 14
Myzostoma, nervous system, 224 ; section,

263

,, cirriferum, organisation, 262
Myzostomidce, 182

NADINA, 134

Naidomorpha, 181

Nate, 181
,, barbata, 267

Narcomedusae, 69

Nassella, 7

Nassellaria, 7

JVaucoris, 440

Nauplius, 307, 386, 390

Nausithoii, 72, 11 ; sensory bodies, 96,
116

Nebalia, 293 ; antennules, 308 ; mandibles,
311 ; olfactory tubes, 356 ;
pleopoda, 324 ; thoracic foot,
317
,, Geo/royi, 293

Nebalidce, 293

Nebaliopsis, 292

Nectophore, 110, 112

Nemathelmia, 178

Nematocysts, 39, 81

Nematoda, 178 ; diagram of nervous sys-
tem, 218

Nemertes, 178

Nemertina, 178; anterior end of body,
217 ; diagram of proboscis, 198 ; in-
testinal canal and genital organs, 205 ;
larva, 274; nephridial and circulatory
system, 235 ; section through body, 192

Nemocera, 443

INDEX

557

Nemura, 440

Nepa, 440

Nephelis, 181 ; vascular system, 249

Nephridia, 235 ; as ducts for sexual pro-
ducts, 243 ; excretory organs, 243 ; per-
manent, 239 ; phylogenetic origin of,
247 ; provisional or embryonic, 237,
238

Nephroblasts, 279

Nephrops, 298

Nephthys, 182

,, scolopendroides, 251 ; section,
252

Nereidce, 182

Nereis cultrifera, 239

Nerocila, 295

Neural plate, 167

Neurilemma, 52

Neuroblasts, 278

Neurochord (Vermes), 222 ; (Crustacea),
351

Neuroptera, 441

Nicoletia, 439

Noctiluca, 8

,, miliaris, 9

Noctuiformes, 44$

Noctuina, 442

Non-Calcarea, 60

Notodelphys, 291 ; trunk feet, 316

„ agilis, anterior maxillae, 312

,, Almannii, mandibles, 311

Notodontidce, 442

Notoinastus, 181

Notommata, 185

Notonecta, 440

Notopoda, 299

Notopygos, 182

Nucleus, 1

Nuda, 72

Nycteribia, 443

Nymphalidoe, 44%

Nymphon, 4%4

,, hispidum, 423

OBELIA geniculata, 68
Obisium, 509

,, silvaticum, heart, 527
Ocellata, 97
Ocelli, 469

Ocneria, mouth parts, 448
Octocorallia, 70
Ocypoda, 300
Odonata, 439
Odontoblasts, 45
(Ecanthus, 454
CEdemeridce, 441

(Edipoda, 439 ; sexual organs, 485
(Estridce, 443
Olenus, 415

Oligochceta, 181 ; nephridia, 240
Oligodadus, 146
Olindias Mulleri, 69
Olynthus, 60, 62

Ommatidium, 353, 471
Ooecia, 264
Ootype, 159
Oniscidce, 295
Onychophora, 4%7
Operculum, 417
Opheliacea, 181

Ophrydium, 11
Opisthotrema, 162
Orbitelarice, 510
Orchestra, 296

,, cavimana, 294
Oribatidce, 510
Orthezia, 463
Orthonectidce, 58
Orthoptera, 4^9 ; sexual organs of larva

485
Oscarella, 61

,, lobularis, section, 65
Osculosa, 7
Osteoblasts, 45
Ostia, 358
Ostracoda, 290
Oxycephcdus, 296
Oxyrhyncha, 299
Oxystomata, 299
Oxytricha, 9
Oxyuris vermicularis, 180

PACHYDRILUS, 181
Psedogenesis, 502
Paguridce, 298
Palcemon, 298

,, Squilla, eyes, 354
Palceonemertina, 178
Palingenia, 439
Palingenesis, 118
Palinuridce, 298
Palinurus, 298

,, Phyllosoma of, 395
Pallene, 424
Palpons, 112
Paludicella, 184
Pandorina, 8
Panorpata, 441
Pantopoda, 422-425
Paradesmus gracilis, 459
Paradoxides, 415
Paragnatha, 315
Paramcecium, 9

,, aurelia, 17

Paranebalia, 293

,, longipes, anterior maxillae,

312 ; posterior maxillae,

313

Parapodia, 188
Parasitism, 172
Parthenogenesis (Vermes), 266 ; (Insecta),

501, 502

Parthenogonidia, 21
Pauropoda, 438
Pedicellina, 184 ', larva, 271

558

COMPARATIVE ANATOMY

Pediculidce, 440

Pedipalpi, 509

Pedipalps (Arachn.), 515

Pedunculata, 292

Pegantha, 69

Pelagia noctiluca, 72

Peltogaster, 292

Penaeus, 298 ; posterior maxillae of larva,

313 ; older larvae, 392
Penella, 291
Pennatula, 70
Pentamera, 441
Pentastomidce, 510
Pentastomum, 510

,, denticulatum, 511

, , tcenioides, 510 ; nervous sys-

tem, 520 ; sexual organs,
533, 534, 537

Pentatoma, 440 ; mouth parts, 451
Perforata, 4
Perichceta, 181

„ mirabilis, 240
Pericolpa, 72
Peridinium, 8
Peripatus, 437

,, capensis, 4^7 ; anatomy, 430 ;
anterior bo4y, 429

, , Edwardsii, 437 ', embryo, 435 ;
head, 427 ; nephridia, 431 ;
section of segment, 428 ;
sexual organs, 436

,, Leuckartii, 1$7

, , Novce Zealandice, J$7, 427
Periphylla, 72
Periplanata, 439
Peripneustic tracheal system, 482
Peritricha, 11
Perlaria, 440
Peromedusce, 71
Peronia, 69
Perradii, 74
Phacelli, 78, 85
Phacops, 415
Phceodaria, 7
Phaeodium, 7
Phagocytes, 500
Phalangidce, 509 ; heart, 527
Phalangium opilio, sexual organs, 533, 534
Phalansterium, 8
Phascolion, 244
Phascolosoma, 183
Phasmidce, 439
Phialidium varidbile, 68
Philodromus, 510

Philoica domestica, sexual organs, 534
Philopterus, 440
Pholcus, 510

,, phalangoides, heart, 527
Phoronidce, 183
Phoronis, 183 ; larva, 272
Phoxichilidium, 424
Phractaspis, 6

,, prototypus, 6

Phreatothrix, 240

Phreoryctidce, 181

Phronima, 296

Phryganea, 441

Phrynus, 509

Phthirius, 440

Phylactolcemata, 184

Phyllacanthidce, water- vascular system, 153

Phyllium, 439

Phyllobothrium, 134

Phyllobranchiaa, 328

Phyllodocidce, 182

Phyllopoda, 288

Phyllosoma larva, 395

Phylloxera, 440

Physalia, 70

Physemaria, 58

Physonecta, 70

Physophora, 70

Physopoda, 440

Phytometridoe, 442

Phytophthires, 440

Phytoptidce, 510

Pilema, 72

Pilidium larva, 274, 275

Pilumnus, 300

Pinnoteres, 300

Pisa, 299

Piscicola, 240

Plagiostoma, 134

Plakina, 61

Planaria, 133 ; ovary, 29

Planocera, 133 ; pharynx, 139 ; intestinal

and nervous system, 141
Planula, 129, 130 ^
Platodes, 133-172
Platydeis, 439
Platyscdus, 296
Plecoptera, 440
Pleon, 301
Pleopoda, 315, 322
Pleura, 414

Pleurobranchiae, 328, 327
Pleurochceta, 181
Plumatella, 184

,, repens, 208
Plumularia, 68
Plusiada, 442
Pneumatophore, 70, 108
Podobranchiae, 328, 327
Podoconus, 7
PodocoTgjne carnea, 68
Podopfrrya, 11

,, gemmipara, 11
Podophthalmata, 296
Podura, 439
Pcecilopoda, 417
Polar bodies, 31
Polia, 178
Pollicipes, 292
Polychceta, 181
Polydadidce, 133
Polycystidce, 9

INDEX

559

Polycyttaria, 6

Polydesmidce, 438

Polydesmus complanatus, anterior body,
452 ; larva, 503

Polygordius, 181

Polymnia, 182

Polymorphism (Insecta), 490

Polynce, 182

Polyophthalmus, 181

Polyphemus, 290

Polystomum, 134

Polythalamia, 16

Polyxenidce, 438

Polyzoa, 134

Polyzonidce, 438

Pontobdella, 181

Pontonia, 298

Porcellanidce, 299

Porcellio, 295

Porifera, 60-66

Porpalia, 70

,, prunella, 114

Porpita, 70

Portunus, 300, 378

,, mcenadis, embryo, 379
,, Megalopa, larva, 396

Pranizidce, 295

Pray a, 70

„ galea, 111, 113

Priapulidce, 183

Proboscidea, 139, 157

Proctodaeiim, 269

Proetus, 415

Proglottides, 164

Promesostoma, 157

Pronucleus, 32

Prorhynchus, 134

Prosopygia, 182

Prosthiostomum, 133 ; pharyngeal appar-
atus, 139

Protamceba, 4

Protella, 296

Proteolepadidce, 292

Prothorax, 444

Protista, 4

Proto, 296

Protodrilus, 181

Protomerit, 9, 16

Protomyxa, 4

Protoplasm, 1

Protopodite, 307

Protospongia, 8

,, Haeckelii, 8

Protozoa, 4-23

Protozocea larva, 390, 391, 392

Protracheata, 427-437

Protula, 182

Provortex, 157

Proxenetes, 156

Psammoryctes, 181

Pselaphidce, 442

Pseudocalanus elongatus, 2d antennae, 310

Pseudoceridce, 150

Pseudoneuroptera, 44$

Pseudophana, 441

Pseudopodia, 4, 14

Pseudoscorpionidce, 509

Psocidce, 440

Psychidoe, 442

Psygmobranchus, 280

Psyllidce, 440

Psyllopsisfraxinicola, enteric canal, 463

Pterobranchia, 183

Pteromalidce, 442

Pterophoridce, 442

Pterygota, 439

Pterygotus, 416

,, osiliensis, 416

Pulex, 441
Pupipara, 44$
Pycnogonidce, 422
Pygidium, 414
Pyralidce, 44%
Pyrrhocoris, 440 ; mouth parts, 451

QUADRULA, 4

,, symmetrica, 3

RADIOLARIA, 6

Ranatra, 440

Raphidia, 441

Rathke's organ, 342

Redia, 169

Reduviis, 440

Reniera, 61

Reticulum, 151

Retinacula, 200

Retinophorae, 471

Retitelaria, 510

Rhabdites, 138

Rhabditis nigrovenosa, 178

Rhabdocoelidce, 133

Rhabdom, 354

Rhabdomere, 470

Rhabdopleura, 184

Rhachis, 414

Rhipiphoridoe, 441

Rhizocephala, 292

Rhizophysa, 70

Rhizopoda, 4

Rhizostomce, 72

Rhodalia, 70

Rhopalia, 71, 95, 96

Rhopalocera, 44%

Rhopalonema velatum, 69 ; tentacles, 95

Rhopalura, 58

,, Giardii, 59

Rhynchelmis, 181
Rhynchobdellidce, 181
Rhynchocoela, 178
Rhynchodaeum, 197
Rhynchonella, 185

,, psittacea, 191

Rhynchota, 440
Rhynclwthorax, 424
Rotalia, 4

560

COMPARATIVE ANATOMY

Rotalia Freyeri, 5
Rotatoria, 185
Rotifer a, diagram, 185
Rugosa, 70

S A BELL A, 182
Sabellaria, 182

,, alveolata, nervous aiid nephri-

dial system, 221
Saccocirrns, 181
Sacculina, 292

, , carcini, 305 ; development, 386 ;
section, 373 ; larval stages, 386
,, externa, 387
,, internet,, 387
Sagitta, 185 ; development, 281 ; section,

197

,, bipunctata, head of, 227
hexaptera, 227 ; eye, 231

Saltatoria, 439
SaMicus, 510
Saltigrada, 510
Sao, 415
Sapphirina, 291

,, Edioardsii, nervous system,

346

Sarcodictyum, 15
Sarcodina, 4
Sarcolemma, 48
Sarcomatrix, 15
Sarcophaga carnaria, nervous system,

466

Sarcopsylla, 441
Sarcoptidce., 510
Sarcosepta, 99
Sarsia siphonophora, 110

,, tubulosa, 68
Saturnidce, 442
Satyridce, 44%
Scalpellum, 292
Schistocephalus, 134
Schizonemertina, 178
Schizoneura, 440
Schizopoda, 297 ; thoracic feet, 321'; larva,

392

Schultzia, 156
Scolex, 138, 164
Scolopendrella, 438 ; heart, 476 -

, , immaculata, 444 ; poster-

ior body, 453
Scolopendridce, 438
Scolopophore, 471

Scorpio, embryo, 638 ; nervous system,
519 ; digestive tract, 524 :
heart, 527

,, Africanus, 512

,, occitanus, sexual organs, 633,

534

Scorpionidce, 509
Scuta, 304

Scutigera coleoptera, trachse, 479
Scutigeridce, 438

Scyllarus, 298 ; development, 395

Scyphistoma, 72

Scyphomedusce, 71

Scyphopolyp or Scyphula, 70 ; develop-
ment, 130

Scyphozoa, 70

Sedentaria, 182

Segestria, 510

Segmental organs, see Nephridia

Segmentation, 120-129 ; nucleus, 33

Seison, 185

Semostomce, 72

Sensory body of Ctenophora, 79, 96

Sergestes, 298

Serictery, 458

Serpulidce, 182

Sertularia, 68

Sesiadce, 442

Sialidce, 441

Sida, 290

Silphidce, 442

Siphonantha, 70, 108-114, 109

Sipkonaptera, 441

Siphonophora, 69

Siphonostoma, 182

Sipunculacea, 183

Sipunculidce, 183

Sipunculus, 183 ; anatomy, 208 ; diagram
of muscles, 195 ; larva, 270

Siriella, 297

,, gracilis, pleopoda, 324

„ Thompsonii, pleopoda, 324

Sitaris, 492
Sklerosepta, 99
Sminthurus, 439
Solenopharynx, 157
Solmaris, 69
Solpugidce, 509
Spadella, 185
Sphceromidce, 295
Sphcerozoum, 6
Sphcerula bombi, 179
Sphingina, 44%
Spicules, 64
Spiculispongice, 61
Spionidce, 182
Spirifer, 185
SpirograpMs, 182
Spirostomum, 9
Spongelia, 62
Spongilla, 61
Spongin, 63
Spongioplasm, 26, 49
Sporocysts, 169
Spumellaria, 6
Squame, 309

Squilla, 295, 296 ; thoracic feet of larva,
320 ; circulatory system of
larva, 363

,, viantis, genital organs, 375
Staphylinidce, 442
Statoblasts, 264
Stauridium cladonema, 68

INDEX

561

Stauromedusce, 71

Stenobothrus, 439

Stenorhynchus, 299

Stenostoma, 134', water- vascular system, 153

Stentor, 9

,, Roeselii, 10
Stephalia, 70

,, corona, 110
Stephanoceros, 185
Stephanosphcera, 8
Sternaspidce, 182
Sterrogastrula, 124
Sterroplanula, 127
Stichocotyle, 143
Stigmata (Protozoa), 18 ; (Antennata),

426-478

Stomatopoda, 297 ; larvae, 393, 394
Stomodaeum, 167, 269
Stratiomyidce, 44$
Strepsiptera, 490

Strobila, Monodisc, 107 ; Polydisc, 107
Strobilation, 116
Stroinbidium, 11
Stylaria, 181
Stylaster, 67
Stylockoplana, 162
Stylochus, 162
Stylodrilus, 181
Stylonichia, 9

,, mytilus, 10

Stylorhynchus, 9

,, longicollis, 9

Stylostomum, 133
Suberites, 61
Subradii, 74
Subumbrella, 73
Suctoria, 11
Sycandra, 60

,, ciliata, 63
Sycones, 62, 66
Syllidce, 182
Syllis, 182

,, ramosa, 267
Sympathetic nervous system (Antenn.),

468; (Arachn.), 520; (Crust.), 350;

Protrach.), 430 ; (Vermes), 220
Symphyla, 438
Syncoryne Sarsii, 68
Syncytia, 41
Syrphidce, 443

TABANID&, 443
Tabanus, mouth parts, 450

,, bovinus, nervous system, 466
Tcenia, 134

ccenurus, 135
crassicollis, 135
cucumerina, 135
Echinococcus, 135
saginata, 134 5 proglottis, 159 ;
(mediocanellata) sexual organs,
159

scolices, 164
VOL. I

Tcenia serrata, 135 ; nervous system of

scolex, 147

, , solium, 135 ; proglottis, 159 ; sco-
lex, 164 ; Finn 172
Taenioles, 75, 78
Talitrus, 296
Tanais, 293
Tarantula, 510
Tardigrada, 544, 545
Tartaridce 509
Tegenaria, 510
Tdphusa, 299
Telson, 322
Tenebrionidce, 441
Tentaculata, 72
Tentaculocyst, 69, 95
Tenthredinidce, 442

Tenthredo, mouth parts of larva, 449
Terebellidce, 182
Terebrantia, 442

Terebratula, 185

,, vitrea, 226
Terga, 304

Termitidce, 440

Tessera, 71

Testicar dines, 185

Tethya, 61

Tetracorallia, 70

Tetranychidce, 510

Tetraphyllidce, 162

Tetrapneumones, 510

Tetrarhynchus, 134

Tetrastemma. 178

Tettigonia, 441

Tettix, 439

Thalassema, 182

Thalassicolla, 6

Thalassoplancta, 6

,, brevispicula, 6

Tkamnotrizon, 439

Theca, 68, 99

Thecidium, 185

Thelyphonidce, 509

Thelyphonus caudatus, nervous system,
519

Therevidce, 44%

Theridium, 510

Tharacostraca, 296

Thrips, 440

Tkysanoessa gregaria, thoracic feet, 321

Thysanopoda, 298

Thysanoptera, 440

Thysanozoon, 133

Thysanura, 439

Tineidce, 442

Tintinnus, 11

Tipularia, 44$

Tomopteridce, 182

Tortricidce, 442

Tracheae, 477 ; structure, 478
,, book-leaf, 530
,, tubular, 529

Tracheal gills, 457, 482, 483

2 o

562

COMPARATIVE ANATOMY

Tracheal lungs, 530

Tracheata, 287

Trachelius, 9

Trachomedusce, 69

Travisia, 181

Trebius caudatus, 2d antennse, 310

Trematoda, 134

Tricenophorus, 134

Trichina spiralis, 179, 179

Trichobranchise, 328

Trichocephahis dispar, 179

Trichocyst, 13, 17

Trichodactylus anonymus, sexual organs,

533

Trichodectes, 440
Trichodina, 11
Trichoplax adherens, 58, 60
Trichoptera, 441
Trichotrachelidce, 179
Tricladidce, 133; intestinal and nervous

systems, 142 ; sexual organs, 156
Trigonoporus, 133
Trilobita, 414, 415 ; restored trunk seg-

ment, 414
Trinudeus, 415
Tristomum, 134
Trochosa singoriensis, 516
Troctes, 440
Trombidiidce, 510
Trombidium fuliginosum, sexual organs,

533

Tuberella, 61
TuUcinella, 292
Tubificidce, 181
Tubipora, 70
Tubitelaria, 510
Tubularia larynx, 68
Turbellaria, 133 ; pharyngeal apparatus,

139 ; excretory cell, 152
Tylenchus scandens, 178
Tympanal organ, 473
Typhlosolis, 206
Tyroglyphidce, 510

Urochceta, 181

Uropoda, 510 ; sexual organs, 534
Uropoda, 389
Urospora, 9

,, Scenuridis, 9
Urostyla, 9
Uterus bell, 257

VACUOLES, contractile, 17 ; noncoutrac-

tile, 14
Velarium, 92
Velella, 70, 114
Velum, 75, 90
Vermes, 177-286
Versuridce, 101
Vesiculata, 94
Vespidce, 442
Vibracularia, 267
Vitellaria, 26, 155
Volvox, 8

,, globatur, 21
Vortex, 134
Vorticella, 11

,, microstoma, 10
Vorticeros, 134

WALDHEIMIA, 185

,, flavescens, 196

Water-vascular system, 136

XANTHO, 300

Xestoleberis aurantia, mandibles, 311

Xiphosura, 417-421

Xylophaga, 442

Xysticus, 510
YUNQIA, 133

ZlLLA, 510

Zocea.\aj:va, 298, 391, 392, 396

Zoanthus, 71
Zoophyta, 58-132
Zona radiata, 28

Zygote, 20

THE END

Printed by R. & R. CLARK, Edinburgh.

2 7 9 2 2

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED

BIOLOGY LIBRARY

TEL. NO. 642-2531

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

LD 21A-15m-2,'69
(J6057slO)476— A-32

General Library

University of California

Berkeley

U.C. BERKELEY LIBRARIES