Quarterly journal of microscopical science.

FOR THE PEOPLE
FOR EDVCATION
FOR SCIENCE

LIBRARY

OF

THE AMERICAN MUSEUM

OF

NATURAL HISTORY

QUARTERLY JOURNAL

OF

MICROSCOPICAL SCIENCE:

EDITED BY

E. RAY LANKESTER, M.A., LL.D., F.R.S.,

Fellow of Exeter College, Oxford, and Jodrell Professor of Zoology in University

College, London ;

WITH THE CO-OPEBATION OF

E. KLEIN, M.D., F.R.S.,

Lecturer on General Anatomy and Physiology in the Medical School of
St. Bartholomew's Hospital, London ;

H. N. MOSELEY, M.A., LL.D., F.R.S.,

Linacre Professor of Human and Comparative Anatomy in the University of Oxford,

AND

ADAM SEDGWICK, M.A., F.R.S.,

Fellow and Assistant-Lecturer of Trinity College, Cambridge.

VOLUME XXIX. — New Seeies.
ffilUtl) <fi%grapl)it plates anb dSngrabings on ®oob.

LONDON:

J. & A. CHURCHILL, 11, NEW BURLINGTON STREET,

. i n 1 1 i :» m 1 i » v

CONTENTS

CONTENTS OF No. CXIII, N.S., JULY, 1888.

MEMOIRS : page

Haplodiscus piger; a new Pelagic Organism from the Bahamas.

By W. F. R. Weldon, M.A., Fellow of St.John’s College, Cam-
bridge, Lecturer on Invertebrate Morphology in the University.

(With Plate I) . . . . . . . 1

The True Teeth and the Horny Plates of Ornithorhynchus. By
Edward B. Poulton, M.A., F.L.S., of Jesus and Keble Col-
leges, Oxford. (With Plates II, III, and IV) . . .9

Note on the Fate of the Blastopore in Ran a temporaria. By
Harold Sidebotham, M.R.C.S. (With Plate V) .49

Morphological Studies. — I. The Parietal Eye of the Cyclostome
Fishes. By J. Beard, Ph.D., B.Sc. (With Plates VI and VII) 55

On Some Oigopsid Cuttle Fishes. By F. Ernest Weiss, F.L.S.,
from the Zoological Laboratory, University College, London.

(With Plates VIII, IX, and X) 75

The Organ of Verrill in Loligo. By Malcolm Laurie, B.Sc.,
from the Zoological Laboratory of University College, London.
(With Plate XI) 97

CONTENTS OF No. CX1V, N.S., OCTOBER, 1888.

MEMOIRS :

On the Structure of Three New Species of Earthworms, with
Remarks on Certain Points in the Morphology of the Oligochseta.

By Frank E. Beddard, M.A., Prosector of the Zoological
Society, Lecturer on Biology at Guy’s Hospital. (With Plates
XII and XIII) 101

IV

CONTENTS.

PAGE

The Development of the Fat-bodies in Ran a temporaria. A Con-
tribution to the History of the Pronephros. By Arthur E. Giles,
B.Sc.(Lond.), M.B., CH.B.(Vict.). Platt Physiological Scholar,
Owens College, Manchester; House Surgeon, Manchester Royal
Infirmary. (With Plate XIV) ..... 133

Two New Types of Actiniaria. By G. Herbert Fowler, B.A.,
Ph.D., Assistant to the Jodrell Professor of Zoology in Uni-
versity College, London. (With Plate XV) . . . 143

Morphological Studies. II. — The Development of the Peripheral
Nervous System of Vertebrates (Part I. Elasmobranchii and
Aves). By J. Beard, Ph.D., B.Sc., Assistant to the Professor
of Human and Comparative Anatomy in the University of
Freiburg i/B. (With Plates XVI, XVII, XVIII, XIX, XX, and
XXI) 153

CONTENTS OF No. CXV, N.S., DECEMBER, 1888.

MEMOIRS :

Note on a New Organ, and on the Structure of the Hypodermis,
in Periplaneta orient alis. By Edward A. Minchin,

Keble College, Oxford. (With Plate XXII) . . . 229

On Certain Points in the Structure of Urochseta, E. P., and Dicho-
gaster, nov. gen., with further Remarks on the Nephridia of
Earthworms. By Frank E. Beddard, M.A., Prosector to the
Zoological Society of London, and Lecturer on Biology at
Guy’s Hospital. (With Plates XXIII and XXIV) . . 235

On the Development of Peripatus Novse-Zealandise. By
Lilian Sheldon, Bathurst Student, Newnham College, Cam-
bridge. (With Plates XXV and XXVI) . . .283

Note on the Development of Amphibians, chiefly concerning the
Central Nervous System ; with Additional Observations on the
Hypophysis, Mouth, and the Appendages and Skeleton of the
Head. By Henry Orr, Ph.D., Princeton, New Jersey. (With
Plates XXVII, XXVIII, and XXIX) . . . .295

CONTENTS.

v

Studies on the Comparative Anatomy of Sponges. II. On the
Anatomy and Histology of Stelospongus flabelliformis,
Carter; with Notes on the Development. By Arthur Dendy,
M.Sc., F.L.S., Demonstrator and Assistant Lecturer in Biology
in the University of Melbourne. (With Plates XXX, XXXI,

XXXII, and XXXIII)

On Some Points in the Natural History of Fungia. By J. J.
Lister, M.A. .......

CONTENTS OE No. CXVI, N.S., APRIL, 1889.

MEMOIRS:

Contributions to the Knowledge of Ampbioxus lanceolatus,
Yarrell. By E. Ray Lankester, M.A., LL.D., F.R.S., Pro-
fessor in University College, London. (With Plates XXXIV,
XXXV, XXXVI, XXXVI A and XXXVI B) .

Studies in the Embryology of the Echinoderms. By H. Bury,
B.A., F.L.S., Fellow of Trinity College, Cambridge. (With
Plates XXXVII, XXXVIII and XXXIX)

On the Ancestral Development of the Respiratory Organs in the
Decapodous Crustacea. By Florence Buchanan. A Paper
read to the Biological Society of University College, London.
(With Plate XL)

PAGE

325

359

365

109

451

Haplodiscus Piger; a new Pelagic Organism
from the Bahamas.

By

W. F. It. Weldon, M.A.,

Fellow of St. John’s College, Cambridge, Lecturer on Invertebrate
Morphology in the University.

With Plate I.

I propose the name Haplodiscus for a small pelagic
organism occasionally found in the tow-net near the island of
New Providence, Bahamas.

The specimens found by me were collected between the
months of July and November, about fifteen specimens in all
having been obtained during this period. As I employed a
great part of my time during my visit to the Bahamas in
using a tow-net, the creature may fairly be called rare.

The general appearance of Haplodiscus, as seen under a
simple lens, is shown in fig. 1. The body is ellipsoidal in
outline, the antero-posterior diameter being the shortest. In
an average specimen the long diameter measured 1"3 mm., the
short 1*1 mm. The dorsal surface of the body is slightly
convex ; the ventral surface is flat when the animal is at rest,
but capable of becoming concave as a consequence of muscular
contraction. It is by producing a concavity on its ventral
surface that the animal slowly and sluggishly moves through
the water ; this mode of progression, together with a general
superficial likeness to a Protozoon, producing a strong re-
semblance to the Lcptodiscus medusoides of R. Hertwig.

The internal anatomy can only be properly made out by

VOL. XXIX PART 1. NEW SER. A

2

W. F. R. WELDON.

means of sections. In the living animal all that can be seen
is a series of three opacities, one at each end and one in the
middle of the antero-posterior axis. Of these, the anterior
indicates the position of the brain (fig. 1, Br .) ; the median
that of the alimentary tract and reproductive glands ; while
the posterior is due to the presence of the ductus ejaculatorius
and vesicula seminalis (fig. 1, V. S.). The relations of these
various organs can be easily seen in the diagrammatic longi-
tudinal section (fig. 10). Besides the position of these main
organs, the presence of large numbers of “yellow cells,”
scattered irregularly through the tissues, can be seen in entire
specimens, whether fresh or preserved.

The body wall is formed dorsally of two, ventrally of three
layers. In both cases the outer layer is a cuticle (figs. 2, 3,
4, and 10, Cu.), which again differs in structure on the two
surfaces of the body. Dorsally it is an apparently structure-
less or very finely granular layer about 5 yu in thickness, which
appears in section somewhat ragged at its outer edge, being
sharply marked off internally from the subjacent tissues. On
the ventral surface the cuticle (fig. 4) is divided into two
layers ; an outer, similar in all respects to the whole dorsal
cuticle, and an inner ( i . cu.), which appears in section as a very
narrow transversely striated band. Whether this striation
was due to the existence of fine pores or not could not be
determined.

A muscle-layer seems to be present on the ventral
surface only, and to lie immediately beneath the cuticle. In
a longitudinal section through the ventral body wall a
clear space, filled with some feebly - staining homogeneous
material, is seen to lie in this position, and in this space is a
row of rounded dots, the cross sections of transverse muscle-
fibres (fig. 4, m. tr.). In the region of the ductus ejacula-
torius some of these fibres can be seen passing inwards to form
part of the sheath of that organ, and here there can be no
doubt of their muscular nature (fig. 5, m. tr.). Occasionally,
but very rarely, a nucleus or two can be seen in sections lying
in the neighbourhood of the transverse fibres, but outside them

HAPLODISCUS PIGER.

3

(cf. fig. 4) ; but whether such nuclei belong properly to the
muscle-fibres, or whether they are the remains of an ecto-
dermal epithelium which has otherwise disappeared, I have
been unable to determine.

Beneath the layer of transverse fibres is a longitudinal layer,
which appears to be much less important, its fibres being
fewer and farther apart. These fibres seem in section to be
connected with irregular, nucleated protoplasmic elements
(fig. 4, m. ep.)} the distinction of which has been perhaps
exaggerated in the figure.

The only other muscles of the body are those round the
ductus ejaculatorius (figs. 5, 7, and 10), where their
structure is more easily seen. In this region every fibre
appears to consist of a thin, wavy, contractile portion, often
branched at the extremities, and connected near its middle
with a granular protoplasmic body, containing a distinct
nucleus. These fibres resemble those described in Taenia by
Roboz more than any others with which I am acquainted.

A protoplasmic tunic, perforated only by the ductus
ejaculatorius, forms the innermost layer of the body wall,
lying immediately beneath the cuticle dorsally, but separated
from that structure on the ventral side by the muscles. This
tunic (figs. 2, 3, 4, 10, P. t .) consists of an irregular layer
of granular protoplasm, in which nuclei are embedded at fre-
quent intervals, but which does not show any trace of division
into distinct cells. From the inner wall of this tunic numerous
processes are given off (figs. 2, 3, 4, 10, P. r.) which anasto-
mose with one another in the cavity of the body, forming a
reticulum which is either continuous with, or forms an invest-
ment for, the remaining organs of the animal.

Embedded in the protoplasmic tunic, and opening from it
through the cuticle to the exterior, are numerous mucous
glands (figs. 2, 3, gl.). These are irregular spaces in the
tunic, filled with a deeply-staining, probably mucous sub-
stance. The glands often contain, besides mucus, the remains
of nuclei.

The brain is a transverselyeiongated body, lying embedded

4

W. E. E. WELDON.

in the protoplasmic tunic at its anterior end of the body (figs.
1, 3, 10, Br.). It is composed of a mass of fibres, below which
is a layer of nerve-cells. From some of these cells processes
go downwards to the cuticle, which some, and probably all,
perforate. At each side of the brain is a special group of
these processes, which stain more deeply than those nearer
the middle line, though they seem not to differ from the latter
in any other respect. I unfortunately neglected to make
macerated preparations of the fresh Haplodiscus while I
was in the Bahamas, and I cannot therefore say more about
these processes. There can, however, be little doubt that they
are in some way sensory.

A nerve having precisely the structure of the brain goes on
each side for a short distance round the edge of the creature.

The alimentary tract occupies the centre of the body,
communicating with the exterior by a mouth (figs. 2, 10, M.),
which is simply a small perforation of the ventral cuticle,
round which the muscles and other tissues do not seem to have
undergone any special modification. The alimentary tract
itself consists of a large mass of protoplasm, continuous at
the sides of the mouth with the general tunic of the body, and
sending processes from every point to join the protoplasmic
reticulum. Nuclei seem to be absent, except occasionally at
the edges of the mass. Vacuoles are frequently found, con-
taining generally small crustaceans in various stages of disin-
tegration. In one series of sections the alimentary proto-
plasm protruded from the mouth as represented in fig. 10, and
it seems probable that during life it is capable of forming
pseudopodia for the capture of prey.

The reproductive glands consist of a single testis,
which lies on the dorsal side of the body, vertically over the
mouth, and a pair of ovaries, one on each side of the alimen-
tary mass.

The testis (figs. 2 and 10, Te.) is a mass of large, deeply-
staining cells, lying in a meshwork of processes of the general
reticulum, but not separated by any definite investing mem-
brane from surrounding structures. The cells which form

HAPLODISCUS PIGEE.

5

the organ vary in character (fig. 8, a — c). First are found
masses of large, finely-granular cells, the nuclei of which are
evidently about to divide, presenting the appearance shown in
fig. 8, a. Amongst these are masses, one of which is drawn
in fig. 8, b, which resemble sperm-morulae, being made up of
a number of narrow, elongated pieces of protoplasm, each
piece containing an elongated, deeply-staining nucleus, the
pieces being spirally grouped around what appears to represent
the part of the original cell which remains behind after the
formation of spermatozoa. The elements of the third kind
(fig. 8, c) are free spermatozoa, which lie loosely in a line
running from the testis itself to a kind of vesicula seminalis
at the posterior end of the body. The spermatozoa are elon-
gated and wedge-shaped, seeming not to be provided with
vibratile tails. Their nuclei are apparently always elongated
and thread-like, though in most preparations there are indi-
vidual examples in which no nucleus at all can be detected.

The vesicula seminalis is simply a space in the general
somatic reticulum, a little larger than usual, which is filled
with spermatozoa ; its size varies according to the sexual
condition of the animal to which it belongs, but it has not
seemed worth while to do more than indicate its position in
the diagram (fig. 10).

The ductus ej aculatorius appears to open into the
somatic cavity at a point just ventral to the seminal vesicle.
It is in the form of a tube, so curved that while its lower half
is vertical its upper portion and its internal opening look
directly forwards. Near its external opening, which is situated
posteriorly in the ventral middle line, the lumen of the duct
exhibits a considerable dilatation.

The structure of the walls of the ductus I have not elucidated
in a satisfactory manner. So far as I have been able to
determine, it is lined by a thick continuation of the ventral
cuticle, which, however, exhibits many additional striations
aud other complications, so as to leave some doubt as to its
real nature. Outside the cuticle is a layer of large cells, which
may be either an epithelium or more probably a kind of

6

W. F. R. WELDON.

prostate, and outside these is a thick sheath of loosely-arranged
muscular tissue, the circular and longitudinal fibres of which
appear to be irregularly mixed.

I have given in fig. 10 a diagram only of the structure
described, because in actual preparations the course of the
ductus is complicated by small secondary twists, perhaps pro-
duced by the contraction of the creature in dying, which so
complicate sections as to render many figures necessary if any
attempt were made to reproduce the appearance actually seen.

The ovaries lie, as has already been said, one on each side
of the mouth. Each contains a comparatively small number
(under twenty) of ova, which lie loosely near to one another,
but only connected as it were accidentally by the general
somatic reticulum.

Each ovum consists of a mass of protoplasm, which is
granular and deeply-staining in younger, spongy and coloured
faintly by heematoxylin in older specimens (cf. figs. 2 and 9).
The nucleus is large and vesicular, having a reticulum which
in most cases breaks up during the preparation of sections, so
that the nucleus appears partly filled with a mass of granular
detritus. The nucleolus is a remarkable rounded structure,
of considerable size, which appears to consist of a homogeneous
substance, with a more or less excentric vacuole. The ova are
surrounded, at any rate for a considerable time, by a delicate
follicular epithelium, distinct from the surrounding reticulum
(fig. 9).

No duct of any kind is observable in connection with the
ovary, and the only way of escape which suggests itself for the
ripe ova is the mouth.

In one specimen an ovum was found in the condition shown
in fig. 9, with a large and conspicuous nuclear spindle, and
at one end something which might conceivably be a polar
body. Whether the dividing nucleus was in this case a pre-
paration for the extrusion of a second polar body or for seg-
mentation could, of course, not be determined, but this
observation points to the existence of some method of internal
fertilisation as at least probable.

HAPLODISCOS PIGER.

7

The yellow cells are, as has already been said, scattered
quite irregularly throughout the body. Iu the protoplasmic
tuuic they are numerous, lying generally freely iu a space
which separates them from the protoplasm of the tunic itself.
This relation is well seen in horizontal sections through the
body wall, such as that represented in fig. 6. It is, of course,
probable that the space surrounding each cell is a post-
mortem effect produced by the action of reagents on the proto-
plasm. In any case the appearance in sections is constant
and characteristic.

There is generally a considerable group of yellow cells
above the brain (fig. 2).

No distinct cell wall is discernible in the cells themselves,
which appear to consist of a mass of protoplasm, sometimes
solid and finely granular (fig. 6), more often vacuolated as in
fig. 2. A rounded concretion was often observed in some part
of the protoplasm, as iu the cell marked y, fig. 4. The
nucleus is always situated close to one end of the cell, and is
in sections somewhat coarsely granular.

The systematic position of Haplodiscus is not easily
determined. I regret that my limited opportunities of ex-
amining fresh specimens did not permit me to form an opinion
as to the presence or absence of an excretory system. But if such
a system be present, it may fairly be assumed, from its absence
iu sections and from the general character of the animal, that
it is built up on the ordinary Platyelminth type. And,
neglecting the excretory system, the other characters of Haplo-
discus seem exactly such as might be looked for in a free-
living Cestode, which, owing to the absence of a nutrient
fluid in which to bathe the surface of the body, and from
which to absorb food, had either retained or acquired a mouth.

At the same time it seems easily conceivable that a Cestode
or Trematode larva might, either normally or as the result of
exceptional surrounding conditions, acquire reproductive glands
of a simple type, and such a process would introduce into the
life-history of the species in which it occurred a form which
might easily present the characters of the animal before us.

8

W. F. R. WELDON.

EXPLANATION OF PLATE I,

Illustrating Mr. Weldon’s paper ou “Haplodiscus piger.”

List of Reference Letters.

Al. Alimentary mass. Br. Brain. Cu. Cuticle. I). E. Ductus ejacula-
torius. F. Food particle in food vacuole, fo. Follicle cells of ovary. Gl.
Gland-cell. i. cu. Inner layer of ventral cuticle, m. 1. Lougitudiual muscles.
m. 1. ep. Epithelial portion of muscle-fibres, m. t. Transverse muscles. M.
Mouth. Oo. Ovary. P. r. Protoplasmic reticulum. Sp. Sensory processes
in connection with brain. Te. Testis. V. S. Vesicula seminalis. Y. Yellow
cells.

Fig. 1. — View of Haplodiscus piger, under a simple lens, (x about
40 diam.)

Fig. 2. — Transverse section through the middle of the body, showing the
relations of the mouth and alimentary system, the ovaries, and the testis.

Fig. 3. — Transverse section through the anterior end of the body, showing
the brain.

Fig. 4. — Small portion of longitudinal section of the ventral surface, show-
ing the structure of the body wall.

Fig. 5. — From a longitudinal section which cut tangentially the sheath of
the ductus ejaculatorius, showing the transition between the muscles of
the sheath of that organ and those of the body wall.

Fig. 6. — Small portion of a horizontal section through the peripheral proto-
plasmic tunic, showing the absence of cell-outlines, and the relations of the
glands and yellow cells.

Fig. 7. — Two muscle-fibres, from the neighbourhood of the ductus ejacu-
latorius.

Fig. 8,«. — Young sperm-cells from the testis, lying in the general reticulum.

Fig. 8,6. — A sperm-morula from the testis.

Fig. 8, c.~ Ripe spermatozoa.

Fig. 9. — Ovum observed in one specimen, with nuclear spindle and perhaps
a polar body.

Fig. 10. — Diagram of median longitudinal section. For the sake of clear-
ness the yellow cells are omitted.

Mwr XZfW.S$i. /

HO « »^0‘ ^OC0WO<5C<, <> ©

S3H

*/.C£6

Weldon dol

P Huth,LithT KdinT

TEUE TEETH AND HOENY PLATES OF OENITHOEHYNOHUS. 9

The True Teeth and the Horny Plates of
Ornithorhynchus.

By

Edward U. Poulton, M.A., F.L.S.,

Of Jesus aud Keble Colleges, Oxford.

With Plates II, III and IV.

Part I. — The True Teeth.

Historical — Number of the Teeth — Position of the Teeth — Form of the Teeth
— Structure of the Teeth : 1 . Tooth-papilla or Dentine Germ ; 2. Dentine ;
3. Euamel ; 4. Inner Epithelium of the Enamel Organ ; 5. The Stratum
Intermedium of Hannover; G. The Middle Membrane of the Enamel
Organ ; 7. The Outer Membrane of the Enamel Organ — The Less De-
veloped Fourth Tooth — Conclusions — Future Investigations — Conclusions
of Other Writers.

Historical. — The mature Ornithorhynchus has always been
described as without true teeth. It is well known to possess eight
horny plates, two upon each side of each jaw. The true teeth
are developed at an early stage beneath the posterior horny
plates, so that some connection between the fate of these latter
and that of the true teeth will very probably be found to exist,
when material can be obtained.

Although no direct observations have been hitherto made
upon the subject of this part of the present paper, several writers
have argued that the ancestors of the MonoHemes must at one
time have possessed teeth. All who follow Hertwig’s identifi-
cation of the mammalian tooth with the placoid scale must
believe by implication that this was the case, for the Mam-
malia must have received their true teeth through those an-

10

EDWARD B. POULTON.

cestral forms from which the Monotremes, as well as all the
higher mammals arose. My friend Professor Howes has called
my attention to a paper by Professor Huxley Proc. Roy.
Soc./ No. 194, 1879, p. 405), in which the writer, after speaking
of the edentulous condition of the Monotremes, expresses the
opinion that among the higher Vertebrata there is strong reason
to believe that edentulous auimals are always modifications of
toothed forms. Again, Mr. Oldfield Thomas, in an interesting
paper on the teeth of Dasyuridae and the evolution of mam-
malian teeth (‘Phil. Trans./ vol. 178 (1887), b. pp. 443 — 462),
quite takes it for granted that the ancestors of Mammalia pos-
sessed teeth, and he even attempts to reconstruct the characters
of their dentition as far as number, form, and arrangement are
concerned. The actual proof of the existence of true mam-
malian teeth in the specialised descendants of the ancestral
mammals is a most satisfactory confirmation of the acute
predictions of the writers above-named, and adds another to
the numerous proofs of the high degree of probability with
which biological speculation may be attended, when based
upon the firm ground afforded by the careful consideration and
comparison of all available facts.

In the investigation of the epidermic structures of Orn-
ithorhynchus I was greatly aided by Professor Howes, who
informed Dr. W. K. Parker of my researches, and induced him
to send me a specimen of the young form of this species.
Wishing, however, to examine some sections of the bill, which
was absent in the specimen forwarded to me, I communicated
with Dr. Parker, who most kindly placed the whole of his
material at my disposal. With other things, there was a
series of transverse vertical sections through the head of
another young specimen, which was 8-3 centimetres long in
the curled-up attitude in which it had been received, and which
was fixed by the spirit. The larger hairs had alone appeared
above the skin. Examining these sections on the following day
I found that typical mammalian teeth were developing in the
upper jaw, the lower jaw being unrepresented in the section.
I at once communicated with Dr. Parker, who most generously

TECJE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 11

urged me to publish the fact in the f Proceedings ’ of the Royal
Society, and greatly assisted me with more material. I
received a cleaned skull of the same age, a cleaned left lower
maxilla, and the partially cleaned posterior halves of right and
left lower maxillae, and also found in the same bottle a piece
of epithelium with the subjacent tissues attached, in which I
rightly conjectured that teeth might be embedded. This had
probably been removed from the cleaned left lower maxilla.
The investigation of this material has afforded the means for
this part of the present paper.

A brief account of the structure and mode of occurrence of
the teeth was read before the Royal Society, Feb. 9th, and has
been printed in the ‘ Proceedings ’ (vol. 43, p. 353).

I quote from this paper a passage which insufficiently
expresses the extent to which I am indebted to Dr. Pax’ker.
“ When it is remembered that Dr. Parker had put the sections
aside for a time in consequence of the press of other work,
intending soon to make use of them for the investigation of
the skull, it will be seen at once that my association with this
discovery is purely accidental, and that I have been treated in
an extremely generous spirit.”

Number of the Teeth. — There are certainly three con-
siderably developed and large teeth in each upper maxilla.
That this is the case is proved by the comparison of Dr.
Parker’s consecutive sections, of which the most characteristic
are figured in PI. 11, figs. 1 — 15 x 14‘5. It is also most
probable that three teeth occur in each lower maxilla, but I
can only be absolutely certain of the existence of two, corres-
ponding to the posterior two of the upper jaw. These two
lower teeth are figured iu PI. II, fig. 16 x 9, as they
appeared in a dissected preparation of the posterior part of
the right lower maxilla. Dr. Parker has kindly consented to
add this preparation to the odontological series of the British
Museum, where it will shortly be placed. The fragment of
maxilla came to an end immediately in front of the anterior
tooth, so that it was impossible to ascertain whether a tooth
corresponding to the anterior upper tooth was present. The

12

EDWARD J3. POULTON.

part of the opposite maxilla, aud the detached fragment of
tissue, were cut into consecutive transverse sections; hut al-
though some appearances seemed to point to the existence of
such a tooth, the condition of the specimens prevented any
certainty on the point. Thus the teeth had been greatly in-
jured by the partial cleaning of one specimen, and the other
was by no means complete. These conditions did not, however,
affect the histological part of the investigation.

Since the account was sent to the Royal Society I have
been greatly interested to find an additional tooth, in a very
early stage, immediately behind and to the inner side of the
posterior tooth, as previously described, in both upper and
lower jaws (see PI. Ill, fig. 7 x 50). Hence there are
traces of four teeth in the upper jaw and probably the same
number below.

Position of the Teeth. — These teeth are placed in an
antero-posteriorly directed row, exactly as Tomes describes
in the development of the typical mammalian tooth, “ in a
widely open gutter of hone,” and the condition of my material
indicated that “ if at this stage the gum be stripped off from
the jaws the developing tooth capsules are torn off with the
gum” ('Dental Anatomy,’ 187G, p. 134). In the lower jaw
no bone had been developed between the teeth and the very
large inferior dental nerve which therefore passes along the
bottom of the dental furrow ; and the same fact holds as
regards the superior maxillary division of the fifth nerve aud
the upper teeth. The posterior upper teeth are similarly un-
separated by bone from the closely adjacent muscular tissue
lying between the zygoma and the skull. The teeth of
both jaws lie in the groove which subsequently holds the
posterior horny plates which subserve the function of mastica-
tion. At first, when I had only examined Dr. Parker’s sections
of the skull, I did not think, for reasons which will be given
below, that the teeth exactly corresponded to the future site
of the plates ; but this became certain when I carefully com-
pared the cleaned skull and inferior maxilla with the sections
of both upper and lower teeth, and with the dissected pre-

TRUE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 13

paration. The superficial epithelium shows very little trace of
its subsequent differentiation into the plates ; it is somewhat
thicker than elsewhere, but there are extremely few isolated
papillary elevations instead of the very numerous papillae
which are so characteristic in a vertical section of the horny
plate. In the upper jaw, the superficial epithelium just
external to the anterior teeth is abruptly raised to a some-
what higher level than the rest, suggesting the appearance
of the plates ; and for a few sections, anterior to the most
anterior tooth, this differentiation is continued, the epithelial
ridge becoming more pronounced but narrower. In the
majority of sections containing teeth there is, however, no
marked alteration of level in the epithelium and only a greater
thickness. It is possible that the anterior ridge represents the
front part of the plates, differentiating especially early ; but
however this may be, it is quite certain, from the relations to
the skull, that the latter correspond to the epithelium covering
the posterior as well as the anterior teeth. In many of the
twelve anterior sections which contained teeth, there was
present beneath this external epithelial ridge, an epithelial
tube invaginated from the side, which in transverse section
much resembled a very rudimentary enamel germ. The ap-
pearance is probably deceptive, for longitudinal and oblique
sections showed the existence of a tube, aud the transverse
sections indicated a distinct, although partially obliterated,
lumen between the invaginated and the other walls. It is pro -
bably a gland duct, but it is unfortunate that the sections are
often incomplete and unsuited for examination in this locality.
In the lower maxillae the epithelium only remained over the
teeth, and no raised ridge could be seen, while the relative
position of the teeth to the entrance and course of the inferior
dental nerve, as compared with the cleaned structure and with
the adult jaw, clearly showed that here also the teeth cer-
tainly develop in a wide, distinct alveolar furrow, which is
subsequently occupied by the posterior horny plate.

Form of the Teeth. — It is only possible to indicate
the form of the three anterior teeth, for the fourth is in

14

EDWAED B. POULTON.

far too early a stage for any attempt at such description.
Successive vertical, transverse sections through the first or
anterior upper teeth, beginning anteriorly, are shown in
PI. II a, figs. 1 — 3. In the lower jaw, I have already stated
that I cannot be certain of the presence of this tooth. The
sections indicate a long, narrow, very completely calcified
tooth, directed downwards and somewhat inwards, the apex
being very nearly in contact with the lower surface of the oral
epithelium. There is one chief cusp, and apparently a second
smaller one, externally placed (shown at PI. II a, fig. 3, o. c.) ;
but I cannot feel very sure about the latter, for the sections of
this tooth were not so satisfactory as those of the others. It
is quite clear that the tooth is far smaller than the second and
third, which lie behind it. These latter are shown for both
sides of the upper jaw in figures of a series of vertical trans-
verse sections (PI. II, b and c, figs. 4 — 15), and those of the
right lower maxilla are seen in a figure (see fig. 16) of a dis-
sected preparation, b being the anterior and c the posterior of
the two teeth, seen from within and above. The comparison
of b and c in the sections and in fig. 16, at once shows that
the anterior or second tooth is a larger tooth than the third *
It is also obvious from the figures that each of the second teeth
possesses two large calcified cusps (coloured red in the figures),
which are placed respectively on the anterior and posterior end of
the inner side of the upper teeth, and of the outer side of the
lower teeth, and which are therefore adapted for interlocking
in mastication. The rest of the tooth is uncalcified. The
surface (shown in b, fig. 16) is smooth and mammillated, shal-
low furrows separating the low rounded elevations and ridges.
The sections (b, figs. 6 — 11, o. c.) indicate that there are many
(probably four or five) small, uncalcified outer cusps in the
upper teeth, while the corresponding inner part of the lower
tooth (shown in fig. 16) has been accidentally cut away; but
there is no doubt that its appearance is, in this respect, very
similar to that of the third tooth (fig. 16 c), viz. that its border
is crenulated from the presence of small, soft inner cusps, of
which only the anterior now remains.

TRUE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 15

The third tooth, both above and below, differs in the fact
that the anterior large cusp is alone calcified, although the
posterior cusp is present. Furthermore, the anterior cusp is
not so large or so thickly calcified as those of the second teeth.
Figure 16, c, shows that five inner cusps are present in the third
lower tooth, the central one being very minute. The existence
of the same cusps on the outer side of the upper teeth is shown
in c, figs. 12 — 15, o. c. Further details of the form of the
upper teeth can be learnt from figs. 1 — 15 on PI. II, and in
the description of the plate. The shape of these teeth is
characteristically mammalian, and, together with their posi-
tion, points to correspondence with some part of the molar
series of other Mammalia.

Structure of the Teeth. — The structure is also charac-
teristic of mammalia. The tissues of the three anterior teeth
will be considered from within outwards.

1. Tooth-papilla or Dentine Germ. — This is indicated
at p in many of the figures in Plates II and III ; its structure
being entirely normal, and the same may be said of the layer of
odontoblasts (o.) which form its superficial part wherever
dentine is developed. These are well seen in PI. Ill, figs. 1
and 2, and in the latter figure the papilla has shrunk, and has
therefore drawn the dentinal fibres (o./.) out of their tubes in
the dentine. Some of the fibres ( o' .f '.) remaining fixed in the
latter tissue have become stretched to far more than their
normal length. I could not determine whether the fibres are
processes of the superficial odontoblasts or of deeper cells (as
stated by Klein), but the appearance of a conical process
with its apex continuous with the fibre, seemed to support the
former more common view, for, at any rate, the majority of the
fibres.

2. Dentine. — In most cases the tissue appeared homogen-
eous, but this was a result of decalcification, for sections of
tissue which had not been so treated gave the usual appearance
(PI. Ill, fig. 1, d). I have sometimes noticed the same
homogeneity in the decalcified teeth of higher mammals.
Although the dentinal tubes did not appear to be very mime-

16

EDWARD B. POULTON.

rous in any of the sections, the number is probably normal, as
judged from tbe retracted fibres seen in fig. 2. The staining
of the inner layer of recently formed dentine is peculiar
(PI. Ill, fig. 1, d'), for as a rule this part of the tissue stains
less in reagents. But this is also probably an accidental result,
for I noticed that some of the decalcified teeth were normal
in this respect. The reticulate outer margin of the recently
formed tissue shown in fig. 1 is also peculiar ; but here again
other sections which had been treated differently showed a
margin nearly parallel to the dentine surfaces, such as is usually
described. It is noteworthy that the former appearance in the
less altered tissue shows a margin which extends irregularly
into the calcified dentine, of which parts constricted off and
remaining uncalcified would produce the appearances known
as “ interglobular spaces.” The faint oblique lines on the den-
tine in fig. 1 are probably produced by cracks in the brittle
tissue caused by the razor. The decalcified tissue stains deeply
(fig. 2) ; the older part of the more normal tissue remains un-
stained (fig. 1). Interglobular spaces are represented as black
marks.

3. Enamel. — This tissue is of very normal appearance.
The prisms are seen in transverse section in PI. Ill, fig. 3,
and in longitudinal section in fig. 1, e. The line between it
and the dentine is very smooth and continuous, while that be-
tween it and the enamel cells (fig. 1, e. c. ) is very irregular.
Thetisssue stains faintly round the projections of the cells into
it, probably on account of less complete calcification. The
prisms are oblique to the enamel cells (fig. 1). I could not
determine whether each cell exactly corresponds to a prism,
but this is doubtless the case. The layer is finely striated
parallel to the surface, probably due to tbe transverse striations
of each prism. The prisms must vary in size, for figs. 3 and 1
are magnified to an equal extent.

The layer of enamel is especially thick at the apex of the
teeth. It is thinnest on the third tooth.

4. Inner Epithelium of the Enamel Organ — the
Enamel Cells. — These are normal, long and thin where

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 17

enamel is being formed, much shorter elsewhere (Tomes, 1. c.,
p. 112), They are shown bordering the enamel in PI. Ill,
fig. 1, e. c. Tomes’s processes are seen projecting from the
inner ends of the cells when torn away from the enamel. The
layer is somewhat sharply marked off from the stratum inter-
medium. In preparing the dissection shown in PI. II, fig. 16
it was noticed that these cells adhered firmly to the calcified
part of the tooth, although they were easily separated else-
where. The same fact is indicated in many of the figures (1
to 15 on PI. II).

5. The Stratum Intermedium of Hannover. — Of
entirely normal structure and appearance (see PI. Ill, fig.
1, s. i.). I could not detect capillaries in the layer, as
affirmed by Lionel Beale, although they are certainly present
in the stellate reticulum, and are sometimes seen very near
this layer. Injected specimens would be necessary in order to
be quite certain of their absence.

6. The Middle Membrane of the Enamel Organ — the
Stellate Beticulum. — This layer is largely developed, and
gives to the young teeth a very characteristically mammalian
appearance. The extent of the layer, and in fact the relative
distribution and thickness of nearly all the layers, is best seen
in figs. 1 to 15 (PI. II). The details of the layer are shown
in PI. Ill, figs. 1, 4, 5, and 6 ( m . m. in all figures). It is quite
certain that blood-vessels are present in this layer, and that
they extend into all parts of it. The presence of blood-vessels
in the mammalian enamel organ has been affirmed and denied
(Tomes, 1. c., p. 127). Klein also states that blood-vessels
are not present in the middle membrane (‘Atlas of Histology/
p. 185). I have, however, examined some beautiful sections
of developing teeth in the rat kindly lent me by Professor
Howes, and there is certainly no doubt about the presence of
abundant blood-vessels in this layer, in which they had been
previously noticed by Professor Howes. In many cases altered
blood-corpuscles remained in the lumen of vessels in very large
numbers. It is very extraordinary that the existence of such
obvious vascular channels should have been denied. I propose

VOL. XXIX, PART 1. NEW SER.

B

18

EDWARD B. POOLTON.

to study the distribution of these vessels iu the higher Mam-
malia by means of injected specimens. In Ornithorhynchus
the vessels can be seen entering through the outer layer from
the surrounding subepithelial tissues ; the same vessel can be
traced from the outside into the stellate reticulum (see PI. Ill,
fig. 4, c. and c'.). Within the latter the vascular channels are
unlike ordinary capillaries, having the appearance of cords of
fusiform cells with a very small lumen, which is often difficult
to detect (figs. 4 and 5, c'.). It seems probable that this abnor-
mality is due to shrinkage, which is in some way connected
with the extremely soft and delicate tissue in which the chan-
nels are embedded. Thus the change may have followed from
the post mortem drying up or absorption of the fluid in the
meshes of the stellate reticulum, before the animal was placed
in spirit, or, again, it may be connected with the action of the
spirit upon vascular walls traversing a tissue which yields them
so little support. That the peculiarity is connected with the
surrounding tissues seems clear from the normal character of
the capillaries in the tooth papilla (PI. Ill, fig. 1, c.), and
in the subepithelial tissues (fig. 4, c .) ; and yet continuity
between the channels in the stellate reticulum and the ex-
ternal normal capillaries (fig. 4) clearly indicates the vascular
nature of the former. Many appearances seemed to show
that channels such as have been represented in the figures —
although very numerous — are only the main vessels of the
layer, and that a much finer network of smaller vessels is also
present. A thin layer of subepithelial (mucosa and submucosa)
connective tissue appears to be invaginated with the former
system of vessels (PI. Ill, fig. 4, to'., which is seen to be
continuous with m. ; also see fig. 5, m'.) Such vessels often
penetrate in a radiate manner to a great depth, nearly reaching
the stratum intermedium. Thus fig. 5 represents such a deep-
seated position. It seems probable that the chief vascular inva-
ginations represent a further specialization of the papilliform
processes which are well known to invaginate the outer epi-
thelium into the stellate reticulum of the enamel organ (Tomes,
1. c., p. 134).

TRUE TEETH AND HORNY PLATES OP ORN1THORHYNCHUS. 19

One peculiarity of this layer is the presence of an epithelial
nodule situated just beneath the outer layer of the enamel
organ, almost immediately over the apex of each calcified
cusp of the second and third tooth (see PI. II, n, in figs. 4,
5, 11, and 12). Nothing of the kind could be made out in the
case of the first upper tooth. In thin sections of the lower
teeth, prepared for histological examination, the nodule was
repeated in many sections, although only shown once or twice
in the consecutive sections of the upper teeth represented on
PI. II (for these were prepared for morphological rather than
histological woi’k). In some cases there was the appearance
of an epithelial cylinder extending from the nodule towards
and perhaps reaching the stratum intermedium or enamel cells
over the apex of the cusp. It seems clear that the nodule is
in some way associated with the chief cusp, for there was always
a nodule above each of the latter, while they were never found
elsewhere. Further material and probably other stages will
be necessary in order to make out the significance of these
structures. The minute details are shown in PI. II, fig. 6, n,
where it is seen that the inner cells appear to be corneous and
collected into a dense central mass, between which and the
outer fusiform cells is a space containing loosely-packed cells
resembling the former in character. The position at the
extreme edge of the stellate reticulum is also shown.

7. The Outer Membrane of the Enamel Organ or
External Epithelium. — This appears to be of normal cha-
racter ; it is a highly irregular and apparently discontinuous
layer. Its structure and appearance is sufficiently indicated
in PI. Ill, figs. 4 and G, o. m. I could not detect the presence
of a persistent “neck” continuous from this layer into the
oral epithelium but the condition of the less developed tooth,
to be described below, indicated that such a neck had
existed. It is possible that the process of the superficial epi-
thelium shown in PI. II, figs. 1, 2, 3, d. p., is a remnant of
the “ neck.” Sec description of Plate.

The Less Developed Fourth Tooth. — The appearance
is shown in PI. Ill, fig. 7 x 50, the enamel germ being

20

EDWARD E. POTTLTON.

represented diagramraatically. All the four layers are very
distinet ; the enamel cells ( e . c.) are of the normal columnar
type, the stratum intermedium (s. i.) is very thick, and the
stellate reticulum (in. in.) is typical although of no great
thickness ; the cells of the outer epithelium ( o . m.) have
already lost their primitive columnar appearance, and are
somewhat flattened. There is a distinct and typical “ neck ”
(not shown in the figure), continuous with the oral epithelium.
The tooth-sac and papilla are also normal, and in fact the
whole structure is in every way characteristic of an early stage
in the development of a mammalian tooth.

Conclusions. — It has been already stated that the teeth
of Ornithorhynchus are typically mammalian. The two chief
and largest teeth seem to me to resemble closely the mnlti-
tuberculate molar teeth of Myrmecobius. In the lower jaw
the resemblance is very striking, nearly all the lower molars
of this animal having four small internal cusps and two ex-
ternal cusps, the only difference being in the fact that the
internal cusps are the higher, while the outer are higher in
Ornithorhynchus. In the upper jaw nearly all the molars of
Myrmecobius also have fewer (2 — 3) cusps on the internal
edge, and more numerous (4 — 6) cusps on the outer edge, and
the relative height is also the reverse of that found in Ornitho-
rhynchus.

In addition to the confirmation of the predictions quoted at
the beginning of this paper, the typically mammalian character
of these teeth confirms in the most striking manner an opinion
expressed by Dr. Parker and Mr. Oldfield Thomas as to the
ancestry of the Edentata. Thus the latter writes (1. c. p., 458) :
*‘In the Edentata on the other hand, we find, as is well known,
characteristics wholly at variance with those of all other
mammals. In fact a study of the teeth of this order soon
induces a belief that the variance is so great as to preclude the
possibility of the Edentates lying within the same lines of
development as other mammals, a belief that tallies exactly
with the conclusions of Professor Parker (‘Phil. Trans.,’ 1885,
p. 116, ‘Mammalian Descent/ p. 97, 1885), drawn from the

TRUE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 21

embryology of the group/’ Mr. Thomas, in Plate xxviii,
accompanying his paper, indicates the same argument in a
diagram, which shows the Edentate dentition as a side off-
shoot arising low down from the generalised Prototherian level
of the main Proto-meta-eutherian stem.

In the same connection, the present paper bears in an
important manner upon Tomes’s discovery of an enamel oi’gan
in the developing teeth of toothed Edentata (in Armadillo,
1. c. 128, 282, ‘ Quart. Journ. Micro. Sci.,’ 1874, and ‘ Phil.
Trans./ 1876).

Although Tomes did not consider that the presence of the
enamel organ proved that enamel had been previously formed
in the ancestors of the group, the observation is manifestly
consistent with such an interpretation, which is further sup-
ported by the conclusions of Oldfield Thomas ; and now that
an enamel organ and enamel have been proved to exist
in a living representative of those ancient mammals from
which the Edentates arose, there appears to be little doubt
about the significance of Tomes’s important discovery. In this
statement I am only referring to the existence of an enamel-
organ in Edentata, and express no opinion as to the universal
presence of this structure independently of the later produc-
tion of enamel. In order to come to safe conclusions upon
this latter fact, it will be necessary to study the development
of teeth more widely than has yet been possible throughout
the Vertebrate sub-kingdom.

Again, the structure and development of the rudimentary
teeth of Oruithorhynchus strongly confirm the opinions of the
many writers who hold that teeth are in a more ancestral
condition than perhaps any other structure possessed by the
adult mammal. While the other higher mammalian organs
and structures represented in the Monotremes are profoundly
modified in the latter, the teeth remain practically identical in
form, structure, and development. We have only to compare
the structure of the skeleton or ovary of a Monotreme with
that of any other mammal in order to realise how' much the
identity of the dental structures proves for the excessively

22

EDWARD B. POULTON.

ancestral condition of the mammalian organs of mastication.

I shall shortly have occasion to show that the very ancient
hairy covering of Mammalia is also greatly modified in the
Monotremes. As above stated the facts here set forth strongly
confirm the identification of the mammalian tooth with the
placoid scale.

Again, I have been enabled to suggest a possible explanation
as to the meaning of the largely developed middle layer of the
enamel organ — the stellate reticulum — which is so character-
istic of Mammalia. The condition of these structures in
Ornithorhynchus clearly indicates that the association of such
a peculiar tissue with teeth of a mammalian form must be very
aucieut. Tomes (1. c. p. 125, 126) in describing the tissue
states, “ It has been supposed to have no more important
function than to fill up the space subsequently taken up by
the growing tooth. ” I think that a little consideration will
show that such a function may be extremely important. It is
clear from the method of tooth formation, in which the oldest
dentine is the superficial crust, and all additions are upon the
inside, that the shape of a mammalian tooth, so far as it is
represented by the contour of the dentine, must be modelled
beforehand iu the soft tissue of the papilla or dentine-germ.
This is well seen in PI. II, fig. 16, c, in which the shape of
the tooth is obvious, although only a very small part of the
surface is calcified. When the subsequently formed tooth is
to be merely conical or of some other simple shape such as is
found in Vertebrata other than Mammalia, there is no reason to
suppose that the dentine-germ would encounter any difficulty
in assuming such a shape, although subject to the resistance of
the dense subepithelial tissues. But the case is different when
the soft papilla is compelled, as in Mammalia, to assume a
complex tuberculate outline; and hence I believe arose the
necessity for the existence of a superincumbent tissue of gelati-
nous consistence, which would exert a pressure only a little
greater than that which is necessai’y to keep the enamel cells
iu contact with the growing papilla. As a test of the value of
this suggestion, it will be of interest to compare the size

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHTJS. 23

attained by this layer in the enamel organs of the simple and
complex teeth of the same species of animal. At the same
time there is as yet no proof that the simple teeth of Mammalia
have not possessed a more complex form at some stage in
their developmental history.

In Ornithorhynchus the layer is much less developed in the
simpler first teeth than in the more complex posterior ones
(compare a, figs. 1 — 3 with n and c, PI. II). On the
other hand, the former are more developed and the layer
may have been somewhat reduced in size. It nevertheless
contains abundant blood-vessels.

Finally, the existence of such highly developed teeth in the
posterior part of the jaws, and the absence of any traces
anteriorly, at any rate in specimens of the age examined,
seem to clearly indicate that the bill of Ornithorhynchus is a
very ancient structure, if not in its present form, at least as
some kind of horny beak which could take the place of anterior
teeth. I think, however, that it is very probable that the
rudiments of teeth may be found anteriorly at a much earlier
stage, when the bill is less developed than was the case in Dr.
Parker’s specimens.

Future Investigations. — In addition to the last-men-
tioned point, other questions which require investigation are
the epithelial nodule in the stellate reticulum, the extent of
the capillary network in the latter, the possibility of any
further development of the teeth in later stages and their
relation to the horny plates, and the presence of an anterior
tooth in the lower maxilla. I am now working upon the
young stages of Echidna and of the toothless Edentata, and
hope to shortly publish an account of the results.

Conclusions or Other Writers.

Since the appearance of the preliminary note in the ‘ Proc.
Roy. Soc.’ three writers have published their opinions as to
the conclusions which may be drawn from the presence of true
teeth in Ornithorhynchus.

24

EDWARD B. POULTON.

Dr. St. George Mivart (c Proc.Roy. Soc.,’ vol. xliii, p. 372)
is led to reconsider the structural relations obtaining between
the Monotremes and all other Mammalia, and between both
these groups and the Sauropsida and Amphibia. He concludes
that the Monotremes arose from Sauropsidan ancestors, and
the higher mammals from Amphibia-like root forms ; and
that the resemblances which now exist between the higher
and lower mammals, including tooth structure, are induced
resemblances. In the first place, the existence of true teeth
in Monotremes — teeth which Dr. Mivart rightly asserts to
be mammalian and non-reptilian in form, and, I may add,
in the presence of a strongly-developed stellate reticulum
— can hardly be urged in support of this conclusion, for such
identity of dental structures strongly favours the converse and
more usual theory of a single instead of a dual origin for the
Mammalia. In support of his conclusion Dr. Mivart argues
for the independent origin of similar structures, and he in-
stances a number of single characters, most of which must be
admitted to be truly homoplastic. But many researches of the
last few years, leading us to miuimise or perhaps to disallow
altogether the importance of acquired characters in species
construction, tend very strongly against the relative importance
of homoplastic as compared with liomogenic characters ; and
the numerous resemblances between the Monotremes and other
Mammalia seem to me totally inexplicable on any theory which
supposes them to be induced, and the results of a compara-
tively recent convergence between groups which are funda-
mentally and in origin distinct.

Especially supporting the usual theory of mammalian origin,
is the most important fact that these two groups of mammals
bear a constant and definite relation to each other in respect to
so many structures represented in both, the relation being such
that the structures in question are always primitive, viz. nearer
to the lower vertebrates, in the Monotremes, and advanced, viz.
further from the lower vertebrates in all other mammals. Any
such constant relationship is entirely inexplicable on Dr.
Mivart’s theory of a dual mammalian origin. Whether the

TRUE TEETH AND HORNY PLATES OF ORN1THORHYNCHUS. 25

Monotremes are the descendants of the ancestral Mammalia or
not, it is quite certain that the higher mammals must at one
time have passed through a condition such as now exists in
the Monotremes, in nearly all parts of their organisation; and
many powerful arguments can be brought against the as-
sumption that the same stage has been reached independently,
and at widely separated periods, in the course of organic evo-
lution. Almost all recent work has strongly supported this
argument, the only exception being Gegenbaur’s reseaches
upon the mammary gland. I have already alluded to my own
unpublished work upon the hairs of Ornithorhynchus, which
will be found to enforce the argument in a most striking
manner. It would, however, be inappropriate to give further
supporting details on this occasion. It is sufficient for the
purposes of the present paper to again point out that the
presence of true mammalian teeth in Ornithorhynchus is, as
far as it goes, evidence for the single origin of Mammalia, and
against the theory suggested by Dr. Mivart.

Professor II. G. Seeley (f Proc. Roy. Soc.,’ vol. xliv, No. 267,
p. 129) has suggested that the horny plates of the adult Oruitho-
rhynchus are degenerate true teeth. This statement has ren-
dered necessary the addition of a second part to the present paper,
in which the structure of the horny plates is described in detail
(see below). There is, however, one part of Professor Seeley’s
paper which is better considered here. The writer enumerates
various characters by which mammalian are usually distin-
guished from reptilian teeth, and shows that there are many
instances in which these characters fail. He applies this argu-
ment to the horny plates and to the true teeth of Ornitho-
rhynchus. I shall presently show that the former have struc-
turally nothing whatever to do with true teeth, so that any
argument based on such a supposition falls to the ground. I
will now shortly prove that the true teeth exhibit all the typi-
cally mammalian characters which could be possessed at the
stage of development they have reached. Three of Professor
Seeley’s characters — the presence of distinct sockets, the wear
of the crown, and the method of replacement — cannot of course

26

EDWARD B. POBLTON.

be applied as tests to teeth in such early stages of development.
The existence of different kinds of teeth has been proved in
the upper jaw, and the presence of many distinct cusps has
been abundantly shown. In my short preliminary account
alluded to by Professor Seeley (1. c.. p. 354, footnote) there is
the statement: “ The two posterior (upper) teeth have many
cusps.” In the present paper I think it has been proved that
these teeth are not behind those of many mammals which, as
Professor Seeley admits, possess “ a specialisation which is un-
paralleled among reptiles.” Furthermore, Professor Seeley’s
suggestion that “ there is a certain relation .... between
the complexity of the crown and the complexity of the fangs ”
is extremely probable, and leads us to conclude that the de-
veloped teeth of Ornithorhyuchus must have possessed many
fangs. If we finally add the important test of the presence
of a highly-developed middle layer in the enamel organ, I
think we cannot escape the conclusion that, whatever tran-
sitional states may be met with iu certain characters of cer-
tain teeth in other mammals, these teeth, in the most
primitive mammal, show no indications of any such transi-
tion, hut are essentially and typically mammalian. Of course
I entirely agree with Professor Seeley as to the ultimate
origin of mammalian teeth from the simpler reptilian type, and
I should also agree in considering the differences as compara-
tively unimportant ; and this latter consideration renders it all
the more easy to understand how it is that the gap from rep-
tilian to mammalian tooth-structure was crossed before the
appearance of Monotreme life at its present level.

Professor E. D. Cope, in ‘The American Naturalist’ for
March, 1888 (p. 259), quotes the description of the form of
the teeth from the abstract of my preliminary paper, printed in
‘ Nature,’ February 16th, 1888, p. 383. He considers the
subject of great importance in relation to the secondary mam-
mals with multituberculate teeth. He states : “ The descrip-
tion reads like that of the dentition of the Plagiaulacid genus
Ptilodus. It renders it extremely probable that the

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 27

Multituberculata are Monotremata, and not Marsu-
pialia, as has been supposed.” Professor H. F. Osborn has
also written to me upon the subject, enclosing a proof plate of
his ‘ Memoir upon the Mesozoic Mammalia,’ to be published
shortly. The second and third teeth of Ornithorhynchus bear
considerable resemblance to his figures of the second (lower)
molar of Plagiaulax minor and of Ptilodus, except that the
two chief cusps of these are on the internal border of the tooth
instead of on the external border, as in the lower teeth of
Ornithorhynchus. There is also some considerable resemblance
to his figures of the teeth of Microlestes and of Bolodon.
Certainly, as Professor Cope implies, the character of the teeth
of Ornithorhynchus entirely negatives the argument that cer-
tain secondary Mammalia must have been specialised rela-
tively to the Monotremata, because of their multituberculate
teeth.

Part II. — The Horny Plates of Ornithorhynchus
Paradoxus.

Historical — Form aud Position of the llorny Plates — Structure of the
Horny Plates — Probable Relation of the Posterior Horny Plates to
the True Teeth.

Historical. — Home (‘Phil. Trans.,’ 1802, p. 71) correctly
describes these horny plates as differing “from common teeth
very materially, having neither enamel nor bone, but being
composed of a horny substance only embedded in the gum, to
which they are connected by an irregular surface in place of
fangs.” He describes the “ internal structure ” as “ fibrous,
like nail ; the direction of the fibres is from the crown down-
wards.” In this description he evidently alludes to the
papillae and columns of softer cells above them, erroneously
considering that they represent the hard part of the
plate. He also says, “ In the smaller specimens before
examined each of these large teeth appeared to be made up of
two smaller ones, distinct from each other.” In this irn-

28

EDWARD B. POULIiON.

portaut observation he is alluding to the posterior plates. He
wrongly concluded that, in order to attain the adult form of
plate, the animal must shed these structures. He found that each
cheek-pouch in the female Platypus contained a “ concreted
substance the size of a very small nut,” shown by the micro-
scope to be made up of “ broken crystals.” The substance
was evidently sand ; and I have found an even larger quantity
in the pouches of a specimen kindly lent me by Professor
Moseley. These observations bear in an important manner
upon the wear to which the teeth must be subject.

Heusinger ( ‘ Histologie,’ 1822) wrongly describes Home’s
“fibres” as “hollow tubes,” evidently relying on ground
sections of the dried plates, in which the papillae below and the
soft cells above have dried up. This has been a most prolific
source of error in the description of these structures, just as
it was originally in the case of bone.

Cuvier described the form and position of the horny plates.

Sir Richard Owen (f Odontography,’ 1840 — 1845, vol. i, pp.
309—311) gives an historical account, to which I am indebted
for reference to the authorities quoted above. He figures the
position of the plates, and somewhat roughly indicates their
shape. He describes the form of the teeth, but omits re-
ference to the small third concave surface of the posterior
plates in each jaw. He confirms Home’s important statement
that each of the posterior plates is made up of two separate
tubercles in the young animal. “ The subsequent conversion
of this apparently double into a bituberculate single grinder is
produced by the progressive extension and confluence of the
bases of the tubercles, not by a process of shedding and the
formation of a new tooth, as Home conjectured.” He
wrongly supports Heusinger’s description of hollow tubes, and
figures a horizontal (evidently dried and ground) section (vol.
ii, pi. lxxvi, fig. 3), which is described as “ showing the
concentric walls of the canals of the principal tubes, and the
minute pores or cells of the denser cementing fibrous sub-
stance.” The “ concentric walls ” are the epithelial cells con-
centrically arranged round the column of soft cells above a

TRUE TEETH AND HORNY PLATES OE ORNITHORHYNCHUS. 29

papilla; the latter cells, dried up, constitute the “ canals of the
principal tubes,” while the " minute pores or cells ” are the
dried-up softer centres of the corneous epithelial cells which
make up most of the horny plates.

Professor H. G. Seeley, as I have already mentioned, con-
siders that the horny plates are degenerate true teeth. He
considers that each of the posterior plates consists of "three
teeth on each side closely united together into one long ovate
mass.” Sir Richard Owen quotes a French analysis showing
that the plates consist almost entirely of horny substance.
Professor Seeley considers that this " can hardly apply to
the posterior teeth,” but he gives no evidence in support of
such an opinion. He observes that the central concavities of
each of the posterior plates is opaque, while the peripheral
border is translucent and horny. This appearance is merely
due to the fact that the former contains abundant papillae and
columns of soft cells, which dry up and contain air, while
these structures are only faintly represented in the latter
locality. But the author suggests that the border represents
“ the uncalcified enamel of the tooth, while the central portion
corresponds to the dentine.” He gives no evidence, histo-
logical or developmental, for the actual occurrence of a struc-
tural change so unique as the conversion of uncalcified enamel
into dense, translucent horn. Furthermore, I have already
shown that the enamel of the true teeth is calcified and hard,
and that it has reached a condition which a widespread ex-
perience of vertebrate dental tissues proves to be the climax of
histological differentiation. Professor Seeley supports his
identification of the central parts of the plates with dentine,
by describing certain appearances seen in sections which he
interprets as due to the presence of bony tissue. If the inter-
pretation were correct it would hardly support the writer’s
conclusions, for I have shown that the dentine of the true
teeth is as typical as their enamel, and has reached a stage of
differentiation at which its conversion into bone would be as
improbable as the change of enamel into horn.

But Professor Seeley’s statement that bony tissue is present

30

EDWARD B. POULTON.

in the central parts of the horny plates is only an extreme form
of the error which has followed from the exclusive examination
of dried and ground -down sections. In PI. IV, fig. 9, I
have shown the appearance of a horizontal section prepared in
this way, and it will be seen that there is some superficial resem-
blance to bone. Sir Richard Owen’s “canals of the principal
tubes” are Professor Seeley’s “large Haversian canals,” while
the “ minute pores or cells ” correspond to the “ canaliculi ”
described by Professor Seeley, which radiate from apparent
lacunae. If the more highly magnified fig. 8 be compared with
fig. 9, it will be at once seen that the deceptive resemblance to
bone is entirely due to the presence of air in the spaces left
by the dried-up softer parts, and it will be clear that the whole
tissue is typically epithelial. Different sections will include
more or less air according to slight differences in preparation,
and hence the supposed bone could only be detected “ in some
specimens.” Rut Professor Seeley can only have looked at
horizontal sections of the plates, for the first glance at a vertical
section, however prepared (see figs. 4 and 5), would prove that
the resemblance to bone was merely delusive. Since the above
was written I have been enabled, through the kindness of
Professor Stuart, to examine the sections in the Museum of the
Royal College of Surgeons, upon which Professor Seeley bases his
opinion. They entirely confirm the interpretation at which I
had previously arrived, and of which an account is given
above. There are several dried and ground-down sections,
containing different amounts of air, and those which contain
most air are fairly represented by fig. 9. Among them are
vertical and oblique sections and horizontal sections with the
air displaced. It is difficult to understand how Professor
Seeley can have reconciled the appearance of these latter with
his interpretation of those horizontal sections which still re-
tain a considerable quantity of air.

Professor Seeley’s suggestion that the true teeth may,
perhaps, be looked upon as successional to the horny plates
may be at once dismissed, for the former not only appear long
before the latter can be identified with any certainty, but have

TRUE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 31

reached a very high degree of morphological and histological
complexity, as I have shown in this paper and also in the pre-
liminary account in the ‘ Proc. Roy. Soc.’ If, as Professor
Seeley supposes, the horny plates are degenerate true teeth,
every consideration points to the conclusion that the latter
must be identified with the structures which I have described.

Lastly, Professor Seeley considers that the anterior horny
plates are still more degenerate and horny “ dental ridges,”
which have become “ dental layers formed of vertical parallel
plates of horn in which there is no division into separate
teeth.” I shall presently show that the minute structure of
the anterior plates cannot be described in these terms, and
that they are neither more nor less horny than the posterior
plates, but in every way identical with the latter in structure.
Since Professor Seeley’s conclusions as to the structure and
significance of the horny plates cannot be accepted, it is
unnecessary to allude to any of the arguments founded upon
such conclusions.

Five years ago I prepared sections of the posterior horny
plates, making use of some material kindly given me by Pro-
fessor Moseley. On making cut and stained as well as ground-
down sections, I saw at once that the structures were entirely
epithelial, and that previous descriptions of the miuute struc-
ture had been erroneous. I left the subject until last winter,
when I again examined and figured some of the sections,
intending to publish them with an account of the other
epidermic structures of Ornithorhynchus. At Professor

Lankester’s suggestion, however, I have added the description
and figures of these structures to the present paper, because
of Professor Seeley’s suggestions concerning them, and also
because the posterior plates are evidently connected in some
way with the fate of the true teeth, although structurally
distinct from the latter. I have also prepared many more
sections and have investigated the structure of the anterior
plates. I am indebted to Dr. Hickson — Professor Moseley’s
deputy, and to Mr. C. Robertson for kind assistance in the
loan of specimens. The most valuable material, however, was

32

EDWARD B. POULTON.

kindly provided by Professor Moseley, the horny plates having
been treated with chromic acid in Australia, in 1874. All the
other specimens made use of had been preserved in spirit or
dried.

Form and Position of the Horny Plates. — The form
and relative position of the plates of the upper jaw are seen in
PI. IV, fig. 1, and further details of the upper and lower
posterior plates are given in figs. 2 b, and 3 b. Transverse
vertical sections of the upper anterior and posterior plates are
represented in figs. 4 and 5 respectively. These figures being
fully explained in the description of plates it is unnecessary
to enter into further details here, especially as the most
important points must be again alluded to in considering the
probable relation between the true teeth and the plates. The
lower anterior plates have not been figured, for they are very
similar to those of the upper jaw, the longitudinal ridge being
also placed towards the outer margin and the furrow lying
between it and the inner margin.

Structure of the Horny Plates. — All the plates possess
the same histological structure. They are simply thickenings
of the oral epithelium, penetrated in many places by long
slender papillae, each of which sends up from its summit a
column of soft, deeply-staining cells, into the stratum corneum.
The thickening which forms the plates take place in the
stratum corneum, the stratum Malpighii being of normal
thickness. The plates are of course continuous on all sides
with the oral epithelium. These facts are at once apparent in
cut and stained sections, but when the dry teeth are ground
down, as in the usual method of preparation, the papillae and
columns of cells dry up and cease to be distinguishable, for
each papilla and column forms a single vertical tube full of
air, which may be more or less displaced by the mounting
fluid, so that the appearances differ greatly in different sec-
tions and in different parts of the same section. Such dried
and ground-down sections are represented in fig. 5 (vertical) and
in fig. 9 (horizontal). The deceptive resemblance of the hori-
zontal sections to bone, is chiefly due to the presence of air in

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNOHUS. 33

the dried-up, softer, irregularly-shaped centres of the corneous
cells which surround the columns. Such minute ramified air
spaces bear considerable likeness to lacunae with branching
canaliculi, and their arrangement relatively to the larger air
spaces in the dried-up columns of soft cells is also somewhat
suggestive of bone (see fig. 9, in which, however, the air of the
large central spaces has been almost entirely displaced by the
balsam). A vertical section, similarly prepared, at once dis-
poses of the resemblance to bone (compare fig. 5). The true
structure is, however, only seen in the cut and stained sections
(see figs. 4, 6, 7, 8, 10, 11, and 12, together with their descrip-
tion). The columns of soft cells above the papillse rise to the
surface of the plates, and doubtless assist in promoting the
wear of certain parts. Thus the papillse and columns are very
minute or absent in the central ridge of the anterior plates
(figs. 4 and 6), and in the raised border of the posterior plates
(fig. 5). Conversely, the papillse, &c., are large and numerous
in the concavities of all the plates, and they are seen in fig. 6
to be especially large in the furrow of the anterior plate. The
result of their presence is to render the corneous tissue friable,
so that its surface becomes irregular as compared with that
of the other parts. On examining the vertical sections it
was found that the true surface was generally preserved where
the papillae were small or absent, but that it was rarely intact
over more than a very short length where these structures
were large and numerous. Similar columns of soft cells, also
rising from the apex of long narrow papillae, occur in other
dense horny structures of Ornithorhynchus. Thus I have
described them in the horny teeth of the tongue ( ‘ Quart.
Journ. Micro. Sci./ vol. xxiii, N. S., PI. XXXII, fig. 7, l. s.).
The deep surface of the posterior plates is in close proximity
to the bone of their alveoli, being separated by a relatively
thin layer of connective tissue representing mucosa, sub-
mucosa, and periosteum (see fig. II). The character of the
epithelial cells of various parts of the plates may be seen in
figs. 8, 11, and 12 (see also description). In some sections of
plates which had been softened in an alcoholic solution of

VOL. XXIX, PART 1. NEW SER. C

34

EDWARD B. POULTON.

caustic potash, after the bone had been softened in dilute
nitric acid, the papillae had been accidentally drawn out of
their tubes in the epithelium (by the partial separation of the
plates), so that their shape was peculiarly distinct. In many
parts of the sections these papillae formed a fringe along the
surface of the subepithelial tissues.

Probable Relation of the Posterior Horny Plates
to the True Teeth. — The anterior plates are omitted from
this consideration because there is as yet no evidence of the
occurrence of teeth beneath them. No epithelial thickening or
any other indication of their presence could be made out in Dr.
Parker’s sections. Certain facts, however, seeem to prove that
there is some relation between the posterior plates and the true
teeth. These facts are, (1) the lodgment of the plates in the
alveolar cavity in which the true teeth appear at an earlier
stage ; (2) the existence of a certain correspondence between
the divisions of the plates, the compartments of the alveoli,
and the number of the teeth ; (3) the evidence that the plates
are developed as at least two separate tubercles, apparently
corresponding to the two chief true teeth situated beneath
them ; (4) the rough correspondence between the shape of the
plates and teeth, the chief and higher cusps being internal
above and external below, while the chief and higher lateral
borders of the upper and lower plates have the same position
respectively. On the other hand, the following facts point in
an opposite direction : — (1) The possible rudiments of the
upper plates in Dr. Parker’s sections as epithelial thickenings
which do not correspond with the position of the true teeth, but
are anterior and external to the latter ; (2) the occurrence of
the small third concavity at opposite ends of the upper and
lower plates, when considered in relation to the true teeth of
both jaws.

The first objection may be met by the undoubted fact that
the position of the upper plates in the adult corresponds to
the position of the true teeth in the young, and not
to that of the epithelial thickening. The position of the
thickening has already been briefly referred to ; it could be

TRUE TEETH AND HORNY PLATES OE ORNITHORHYNCHUS. 35

distinctly traced on the twelve anterior sections containing
teeth, and it gradually disappeared in a few sections posterior
to the twelfth. Anteriorly to the teeth it became narrower,
but more defined, becoming most distinct in the sixth, seventh,
and eighth sections in front of the first section, which ex-
hibited traces of a tooth (see PL II, fig. 1). In front of the
eighth section it rapidly disappeared. The epithelium of the
right ridge in the above-mentioned seventh section is shown
in vertical transverse section in PI. IY, fig. 13, and the ap-
pearance strongly suggests an early form of the plate. Traces
of papillae are visible beneath the ridge, and the greater thick-
ness of the epithelium is very noticeable.

I believe that the following account will be found to ex-
plain the relation between the teeth and plates, and to re-
concile the apparently conflicting observations. The calcified
true teeth of Ornithorhynchus became unsuited to the needs
of the animal when it adopted a mode of life in which large
quantities of sand were necessarily taken into the mouth with
the food, when in fact it first fed upon insect larvse, &c., which
it dug with its bill out of the mud and sand at the bottom of
streams. The fact that large quantities of sand are introduced
with the food has been already proved, and I have noticed that
the concavities of the posterior plates are sometimes filled with
mud, sand, and the debris of food. At the same time the
presence of sand may be valuable in assisting to grind down
the food, and it is possible that a store is kept in the cheek-
pouches for this purpose, and is intentionally added during
mastication. Under such circumstances two things might
happen : the true teeth might be protected from the effects of
wear by continuous growth from persistent pulps or by a con-
tinued succession ; or a constantly growing horny plate might
be developed from the oral epithelium, and might be substi-
tuted for the true teeth. While a corresponding difficulty has
been met by the first method almost universally among Mam-
malia, we must remember that there is no a priori reason why
this should be the case. Natural selection only demands rela-
tive success and feasibility, and the means by which such

36

EDWARD B. POULTON.

success is attained is entirely determined by the character of
the variations which appeared at the critical time ; so that
there is no difficulty whatever in believing that the case has
been met by continuous growth in many instances, and by the
substitution of another continuously growing tissue in other
instances. Ornithorhynchus is not the only example of this
method among Mammalia. A similar difficulty, doubtless
also caused by the presence of sand and mud in the food, has
been met in the same manner in the case of the lower incisors
of the Sirenia, which are completely functionless and covered
by a horny plate. In the Manatee these true teeth are absorbed
early, as in Ornithorhynchus, but in the Dugong they persist
until old age, thus proving the entirely independent origin of
the horny plate.

We must assume that the chief dental area has been made
use of continuously throughout the whole period of change,
for the plates are found to occupy the exact position of the
teeth. In this manner the muscular and other arrangements
upon which the movements of mastication depend would also
remain unchanged. We may suppose that the rapidly-wearing
true teeth were at first reinforced by an adjacent corneous epithe-
lial thickening in the position of that described in Dr. Parker’s
sections (Plate IY, fig. 13), and that the thickening gradually
extended over the young true teeth, so that these, instead of
piercing the epithelium, merely conferred the shape of their
crowns upon the latter. Each true tooth was in fact protected
by an additional indurated layer external to the enamel. At
first the teeth may have been thus protected during the earlier
part of the animal’s life, coming eventually to the surface.
This would take place at successively later periods until they
ceased to appear altogether. In strong support of this inter-
pretation is the fact, already quoted, that the two chief con-
cavities of the plates arise separately and fuse at a later
period. Each of these separate tubercles would, according to
this theory, correspond to one of the two chief teeth in each
jaw. A section across the specimen described by Sir Richard
Owen would probably settle the question, The anterior small

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNOHUS. 37

concavity of the upper plate may be similarly formed over the
small anterior tooth, or it may represent the epithelial ridge in
front of the true teeth. But it must be remembered that the
latter is external in position while the concavity is internal.
Furthermore, the concavity has its special alveolar compart-
ment, which seems to indicate the former existence of a true
tooth. On the other hand, the posterior small concavity of
the lower plate is a difficulty on this latter interpretation, for
it possesses its compartment, and yet it arises in a position
where no small true tooth developes ; for it is unlikely that
the very rudimentary tooth-germ attains any degree of spe-
cialisation ; furthermore, there is a corresponding germ in the
upper jaw, and yet no concavity. It seems on the whole pro-
bable that the alveolar compartments of the small concavities
of both upper and lower plates are simply parts of the alveoli
for the second and third true teeth respectively, and that the
simple anterior tooth does not impress itself on the plate, or, if
so, does not produce any effect which can be distinguished from
that of the large tooth behind it. Again, the small concavities
may be due to subsequent differentiation of the plates. It is
quite clear that we cannot be sure as to the correct interpretation
of details, although they will be settled with certainty when
more material is obtainable. Thus it is certainly possible that
the epithelial ridge shown in PI. IV, fig. 13 may have nothing to
do with the plates, and that the latter originally arose over the
true teeth only, in the manner described above. But, under
any circumstances, the subsequent history appears to be toler-
ably clear. The true teeth, after ceasing to come to the sur-
face, would be absorbed at successively earlier stages, thus
permitting the horny plates to gradually intrude into their
alveoli, so that in the adult animal the bone and the under
surface of the epithelium are everywhere in close proximity.
Many sections in various directions through both upper and
lower plates in their sockets failed to reveal any traces of the
true teeth, so that absorption is probably complete. The
contour of the surface of the plates, originally determined by
the underlying teeth, would still be maintained as far as

38

EDWAllD B. POULTON.

general proportion and arrangement is concerned, because the
shape was most favorable for the movements of mastication,
which on this theory are supposed to have persisted with little
change.

This theory seems to account for all the important facts.
Few things would give me greater pleasure than to have the
opportunity of testing it, and of being able to produce an
exact account of what actually takes place.

DESCRIPTION OF PLATES II, III & IV,

Illustrating Mr. Poulton’s paper on “ The True Teeth and
Horny Plates of O r n i t h o r h y n c h u s paradoxus.”

PLATE II.

The upper teeth of both sides are figured iu the series of vertical transverse
sections forming Pigs. 2 — 15 ; while in Fig. 1 the section, being slightly
oblique, did not pass through a tooth on the left side. The figures in compart-
ment a represent sections through the anterior tooth, iu b through the middle
tooth, and in c through the posterior tooth. In all cases the sections are
arranged consecutively, Fig. 1 being the most anterior, and Fig. 15 the most
posterior section; but many sections are omitted. The sections figured were
selected because they were iu the best condition, and because they were
suited to show the form aud structure of the various parts of the teeth. All
the teeth are shown in the natural position, with their apices directed down-
wards, but of course with their inner sides far more closely approximated
than iu the natural condition. The drawings were made from Dr. Parker’s
consecutive sections. All these figures are magnified 1P5 diameters. The
references are as follows in all the 15 figures.

e.p. Oral epithelium, d. p. Process of the epithelium passing towards the
enamel organ aud perhaps the remnant of the neck, which at an earlier date
connected the latter with the oral epithelium. Ou the other hand there are
gland-tubes iu close proximity, and many appearances render it probable that
these may be connected with it. It is seen in sections of the anterior tooth.
1. Inner layer of enamel organ, the columnar enamel cells, aud the stratum
intermedium of Hannover. 2. Middle layer of enamel organ. 3. Outer
layer of enamel organ. D. Dentine, coloured red iu all the figures in which
it is present. E. Enamel, indicated as a white liue external to the dentine.

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 39

P. Tooth papilla. o. Layer of odontoblasts, i. c. Inner chief cusp. o. c.
Smaller outer cusp. N. Epithelial nodule in the middle layer of the enamel
organ, almost immediately over the chief cusps of the two posterior teeth.
The references have not been unnecessarily repeated on all the figures.

A—

Fig. 1. — This was the first section in which teeth appeared, and here only
on the right side. The shape is long and narrow, and the apex is nearly
in contact with the lower surface of the oral epithelium. The direction
of the tooth is obliquely downwards and inwards. The dentine is thick, and
a thin layer of enamel is present : the former is cut tangentially towards the
base of the tooth, aud therefore it appears irregularly interrupted by the
tissue of the papilla and odontoblasts.

Fig. 2 represents the condition of the next, the second section. The
anterior tooth is now seen on the left side, and much resembles the section
just described, except that the papilla is only fairly reached at the apex of the
tooth. On the right side the tooth is broader, aud indicates a tendency
towards the formation of two cusps. The entrance of the papilla is seen,
although the outer layer of the enamel organ crosses this part of the section
at a different level, and can be seen by focussing. Much of the dentine is cut
tangentially. The very complete investment of the papilla by dentine is
noteworthy, especially when compared with the two posterior teeth. In this
and the next figure the dentine is seen to be formed round the very base of
the tooth, so that the entering papilla is encircled by it.

Fig. 3 represents the third section ; the existence of two cusps seems
especially clear on the right tooth. The fourth section is not figured : on the
right side the epithelial process ( d . p.) aud the tangentially cut enamel organ
are all that can be seen ; on the left side the base of the tooth is cut tangen-
tially. The fifth and sixth sections only show the diminishing epithelial process
on both sides. The seventh aud eighth sections show the tangentially cut
enamel organ of the second tooth.

B—

Fig. 4 represents the appearance in the ninth section from the most anterior
in which a tooth was seen. The dentine is cut tangentially towards the base
on both sides, the apex represents one of the chief inner cusps ; ou the left
side this is cut through at its highest point, aud on the right side very nearly
so. An epithelial nodule (N.) is seeu in section above the apex ; it may be
connected with the inner layer of the enamel organ above the summit of the
tooth. A thin layer of enamel is present.

F'ig. 5. — The tenth section : the shape of the large inner cusps is well seen
on both sides. The layer of odoutoblasts ( o .) is distinct. The small outer
cusps are indicated by the direction of the inner layer of the eDamel organ (1)
on the outer side of the papilla (P.). In this vertical section the anterior
part of both eyes arc also cut through for the first time, and it is thus seeu

40

EDWARD B. POULTON.

that the section is very nearly at right angles to the long axis of the head, but
a little further posterior on the left side, as is also indicated by comparison
of this figure with the last.

Fig. C. — The eleventh section. On the right side the entrance of the
papilla is seen, although discontinuous in this section from the rest of the
papilla. On the left side the entrance is seen, and the outer layer of the
enamel organ at a different level. In this latter section the small outer cusp
(o. c.) is shown ; no dentine is developed upon it. The large inner cusp is
cut through on its posterior slope in the left tooth, and higher up on the slope
near the apex in the right tooth.

Fig. 7. — The twelfth section. The two teeth are very uniform, the small
outer cusp being seen in both, and the large inner cusp cut low down on its
posterior slope. The entrance of the papilla is also distinct. The dentine is
cut somewhat obliquely, as in the left tooth of the previous figure. The
thirteenth section is not figured ; it is incomplete on the left side, so that the
tooth is absent ; while on the right the tooth is similar to that of Fig. 7,
except that the inner cusp is cut at a still lower level.

Fig. 8. — The fourteenth section. The entrance of the papilla is seen to
have shifted towards the inner side of the base of the teeth ; the inner cusp
is now cut through close to its base, while most of its contour is concealed by
the enamel organ. An outer cusp is seen in this and in all the remaining
sections of this tooth.

Fig. 9. — The fifteenth section. The teeth are now seen in section between
the two inner cusps, and no dentine is formed upon any part of the surface.
The contour of the enamel organ and oral epithelium on the left side could
not be completed.

Fig. 10. — The sixteenth section. In both teeth the anterior slope of the
second or posterior large inner cusp can be dimly seen beneath the enamel
organ.

Fig. 11. — The seventeenth section. The large cusps are now cut vertically
through their apices in both teeth ; above the apices the epithelial nodule is
seen on both sides (W.). In the eighteenth section, not figured, the posterior
cusps are cut through on their posterior slopes, rather below the apices.
In the nineteenth section traces of the posterior margin of the teeth is seen,
and in the twentieth section either the extreme posterior margin of these teeth
or the extreme anterior margin of the third teeth.

These sections through the middle tooth (b) should be compared with the
surface view of the corresponding tooth in the lower jaw (Fig. 16, b), bearing
in mind that the large cusps are external in the latter. Except for this differ-
ence, the sections indicate a great general resemblance between the teeth.
The number of sections in which a small outer cusp is seen proves that there
must be four or five of these structures corresponding to the appearance of
the inner side of the posterior tooth (c), shown in Fig. 16.

TRUE TEETH AND HORNY PLATES OP ORNITHORHYNCHUS. 41

C—

Pig. 12. — The twenty-first section. The figure indicates that the section
passes through the apex of the large anterior inner cusps on both sides. This,
together with many previous figures, proves that the anterior slope of the
large cusps is much steeper than their posterior slope, for many of the latter
are seen in the sections, while at the first, or at most the second section
through the anterior slope the apex of the cusp is reached. On the right side
the papilla is not continuous. The small outer cusp is seen on both sides in
its special compartment of the enamel organ, and the entrance of the papilla
is shown in both. These inner cusps are much smaller than those of the
second tooth (b), and their dentine is not nearly so thick, and the enamel is
very thin indeed and is not represented in the figures. The twenty-second
section is incomplete, so that neither of the teeth can be seen.

Figs. 13 and 14. — The twenty-third and twenty-fourth sections. These
show the appearance of the posterior slope of the large inner cusps cut
through at two levels ; the small outer cusps are very distinct in both.

Fig. 15. — The twenty-fifth section, the last figure. The posterior inner
cusp is faintly seen through the enamel organ. The section passes between
the two large inner cusps (as in Fig. 9 for the second tooth), but the anterior
slope of the posterior cusps are seen from the surface. These are not covered
with dentine (compare the posterior outer cusp of the corresponding tooth in
the lower jaw, Fig. 10, c, from which dentine is also absent). In the five
remaining sections (twenty-sixth to thirtieth) in which traces of the teeth
appear, the posterior parts of the last teeth are seen ; but nothing is gained
by figuring them. The twenty-ninth is the last section in which the eyes
appear. It is clear that the third tooth is considerably smaller than the
second (compare Fig. 16).

Fig. 16. — X 9. Two teeth in the lower jaw, corresponding to the second
and third (b and c) upper teeth, shown in the above-described sections.
The piece of jaw from which the preparation was made ended abruptly an-
teriorly (in the direction of the arrow), so that the presence of an anterior
tooth (corresponding to a in the sections) could not be ascertained. How-
ever, some sections of the opposite inferior maxilla render it probable, although
not certain, that the tooth is present. The superficial structures (epithelium,
mucosa, and enamel organ) were dissected away so as to expose the upper
surface of the teeth. The inner side of the anterior tooth (b) had been
previously cut so that all the inner cusps, except the anterior one, are removed,
and the tooth papilla, enamel organ, and sub-epithelial tissues are seen in section.
The enamel organ was easily removed from the surface of the teeth, except at
the upper parts of the calcified cusps to which it strongly adhered, doubtless due
to the formation of enamel in this region, and the consequent adherence of the
inner layer of the enamel organ. The teeth are drawn from above, and from
the inner side. The anterior tooth (b) is much the larger. Its large outer
cusps (a. o. c. and p. o . c.) are calcified and hard over the region indicated by

42

EDWAED B. POULTON.

the red colouring, the calcification terminating below in a sharp line of demar-
cation rendered especially distinct from the fact that the jaw had been faintly
stained as a whole in carmine. The tooth is seen in vertical section at P.,
the reference mark being placed upon the entrance of the tooth papilla. The
enamel organ and sub-epithelial tissues are also seen in vertical section at e'. o'.
and s'. ml. respectively, and the same tissues are also seen in horizontal
section at e. o. and s. m. The line l indicates the boundary between enamel
organ and sub-epithelial tissue. Between the two teeth (b and c) at the
point x. the two enamel organs appear to become fused. I could not, however,
feel sure upon this point. The entire crown of the smaller posterior tooth is
shown, and it is seen that there are four small inner cusps (the reference being
to the anterior one) besides a very minute fifth cusp. The large anterior outer
cusp {a. o. c.) is calcified, but the smaller posterior outer cusp ( p . o. c.)
remains soft like the inner cusps. Part of the inferior maxilla is shown at B.
The appearance of these teeth strongly confirms the conclusions as to relative
size, shape, and structure, drawn from the sections of corresponding teeth in
the upper jaw (Figs. 4 — 15). The relative position of the large and small
cusps on the upper and lower teeth respectively is an obvious remnant of the
time when the surfaces of the upper and lower teeth fitted together for the
performance of mastication.

PLATE III.

Fig. 1. — X 188. A portion of a vertical section through one of the
developing teeth of Ornithorhynchus. The tissue had not been decalcified,
and hence the structure of dentine and enamel is better shown than in other
cases. The teeth of which the structure is shown were contained in an
isolated piece of tissue, probably removed from the lower jaw. The relative
position of the tooth from which the section was taken is uncertain, but the
histological details are evidently quite typical, m. m. The middle membrane
of the enamel organ, made up of a honey-combed reticulum of cell-plates.
Capillaries are present in it, but could not be seen in the part of the section
figured, s. i. Stratum intermedium of Hannover : the outlines of the small
spherical cells are not indicated, e. c. The enamel cells : long and columnar,
bounded by a sharp and almost straight line of demarcation from the last-
mentioned layer, and separated by a very irregular line from the enamel prisms.
E. The enamel prisms : the layer is faintly marked by fine closely placed lines,
running parallel to the surface. The axes of enamel prisms are seen' to form
an obtuse angle with the axes of the enamel cells, perhaps due to shrinkage.
I could not determine whether each prism exactly corresponds to an enamel
cell, although this is probably the case. The prisms are faintly stained round
the projections of the last-mentioned layer, probably due to incomplete

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 43

calcification. D. The dentine : probably entirely typical, for the fact that the
number of tubules appears to be smaller than usual doubtless follows from
methods of preparation or preservation. The dentinal fibres are seen iu Fig-.

2 to be very numerous. The faint oblique striation iu certain parts is
probably due to parallel cracks in the brittle calcified tissue when cut by the
razor. The black spots represent the uucalcified “ iuterglobular spaces,”
D'. The inner part of the dentine which stains deeply, probably on account of
its recent formation. The boundary between this and the completely formed
dentine {!).) is highly irregular, the outer margin of recently formed tissue
being reticulate. In mammalian teeth of a similar degree of development this
line of demarcation is usually parallel with the inner and outer surfaces of the
dentine, and the younger tissue stains less deeply than the older, o. The
layer of odontoblasts : the appearance is entirely normal. I could not
determine whether the dentinal fibres arise from the superficial or from the
deeper cells. P. The tooth papilla of normal structure and appearance, con-
taining capillaries at c.

Fig. 2. — X 405. A small part of a vertical section through one of the
lower teeth of uncertain position. Iu this case the tissue had been decalcified
in dilute nitric acid, the thin layer of dentine {!).) is stained equally deeply
throughout and exhibits no trace of dentinal tubes. The fact that such tubes
exist in normal number is, however, conclusively proved by the numerous
dentinal fibres (o.f.) which have been drawn out of them, doubtless because
of the shrinkage of the internal softer tissues. The fact that some of these
fibres {o'./'.) are much longer than the thickness of the dentine must be due
to the stretching of the former to far beyond their normal length, probably
because their distal ends remained fixed iu the dentine during the process of
contraction. Each fibre appears to arise from the apex of a minute conical
process, projecting from the surface of the layer of odontoblasts (o.). This
appears to support the opinion that the fibres arc at any rate in many cases
derived from the superficial cells. The tissue of the papilla is seen at P.

Fig. 3. — X 188. Transverse sections of the enamel prisms, seen in a section
from the same tissue as that from which Fig. 1 was taken. The shape is seen
to be irregularly polyhedral ; but the variations iu this respect and in size
are considerable. The irregularity in size is further proved by the fact that
the prisms drawn in Fig. 1 appear to be much more slender than those which
are here represented.

Fig. 4. — X 188. A vertical section through a portion of the upper sur-
face of the enamel organ, viz. the surface which is nearest to the oral epi-
thelium. The figure was drawn from one of Dr. Parker’s sections of the
posterior upper tooth on the right side. The whole tooth is shown ( X 14'5)
in Fig. 14 on Plate II. The mucosa forming the tooth-sac is represented
at m„ and the highly irregular outer membrane of the enamel organ at o. m.;
the cell-outlines are not indicated. The middle membrane of the enamel

44

EDWARD B. POULTON.

organ is seen at m. m. It is of normal structure, except that it certainly
contains abundant blood-vessels. The vascularity of the enamel organ has
been frequently affirmed and denied by various observers, but there can be
no doubt about the question in Ornithorhyuchus. A normal capillary is seen
in the tooth-sac at c., and it can be traced deeply into the “ stellate reticu-
lum ” of the enamel organ at c'., accompanied by a small amount of connective
tissue (»<'.) from the tooth-sac. The appearance of the deep section of the
blood-vessel is somewhat peculiar, but continuity with an undoubted capillary
outside the organ, in this and in many other cases, leaves no doubt as to the
true nature. It is possible that the appearance of a thick-walled or even
solid cylinder of fusiform cells may be due to the shrinkage of a relatively
large thin-walled vascular channel, following from the peculiarly delicate and
fluid condition of the surrounding tissues. The absence of this latter cause
may account for the fact that the external capillary at c. and the capillaries in
the tooth papilla (Fig. l,c) possess an entirely normal structure and appearance.

Fig. 5. — x 188. A portion of a section of the middle membrane of the
enamel organ, showing the structure of its deeper part close to the stratum
intermedium. The section was taken from the same tissue as that from
which Fig. 1 was drawn. The figure indicates that the vascular channels [d .)
penetrate the layer to a great depth, carrying a small amount of connective
tissue (»/.) with them. The structure of the channel resembles that of the
deeper part of the blood-vessel, described in the last figure (e'.). An appa-
rent lumen is shown at l. It is probable that the vessels shown in this and
the last figure are the main vascular channels, and that smaller branches
form a network in the middle membrane of the enamel organ. Such a con-
clusion was suggested by many of the sections.

Fig. G. — X 188. From the same tissue as the last section, showing the
structure of the epithelial nodule in the most superficial part of the middle
membrane of the enamel organ, immediately over the apex of each chief cusp
of the large broad posterior teeth. The relative position of cusp, nodule
and oral epithelium is shown in many of the figures in PI. II. The nodule
(W.) is seen to lie in the superficial part of the middle membrane of the
enamel organ ( m . m.), and immediately below the irregular and apparently
discontinuous outer membrane (o. m.). The tooth-sac is seen at m. A space
(c. sp.) in the middle membrane probably contained a vascular channel. The
nodule is seen to be made up of a dense outer tissue, composed of fusiform
deeply-staining cells surrounding a concentric space, in which are scattered
thin yellowish cells, with a central dense mass made up of similar cells. The
latter are not indicated by outlines, but by the presence of minute traces of a
nucleus. In some sections an obliquely-cut cylinder of similar structure
appears to extend from the nodule towards the apex of the cusp, and is
perhaps continuous with the inner layer of the enamel organ or the stratum
intermedium in this locality.

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 45

Eig. 7. — X 50. A section from the same lower jaw as that from which
Fig. 2 was drawn. Behind and interior to the comparatively highly-developed
posterior tooth there was a much earlier rudiment of a third or fourth tooth,
also present in the upper jaw. The whole appearance exactly resembles the
corresponding stage in the development of teeth in the higher mammalia.
The oral epithelium is shown at e. p. Beneath this the young enamel organ is
also coloured red, and its constituent layers are diagrammatically represented.
The enamel cells ( e . c.) of its inner layer are of the normal columnar type ;
externally to these the very thick stratum intermedium ( s . i.) is shown, and
then again the beginning of a middle membrane at m. m., covered by an
outer membrane (o. m.) of cells which have already lost their primitively
typical columnar shape and have become somewhat flattened. The tooth-sac
(«.) is clearly marked off from the surrounding mucosa and submucosa (*».),
and continuous with it is the well-marked tooth papilla (P.), which ascends
into the space formed by the invagination of the enamel organ. The “ neck ”
of the enamel organ is not seen in the figure ; it is continuous with the oral
epithelium,

PLATE IV.

Fig. 1. — Natural size. The right side of upper bill and palate seen from
below, showing the relative position and form of the horny plates. The longi-
tudinal ridge of the anterior plate is nearer to its outer margin, and between
it and the inner margin there is a shallow furrow. The three concave surfaces
of the posterior plate are plainly seen. The animal had not attained the full
size, so that both plates arc smaller than usual, or perhaps the smaller size
may be due to sex. Anteriorly, in the middle line, is an oblique furrow con-
taining a canal which leads to the nasal passages, and on the inner side of its
opening into the furrow a small but distinct tubercle is seen. Posterior to
this are many curved corneous ridges.

Fig. 2. — Natural size. A. The socket of the right upper posterior plate,
seen from below. The concavity of the alveolus is seen to very roughly cor-
respond to the two posterior concavities of the plate, while it possesses a small
but distinct and well-separated anterior and internal compartment for the
corresponding concavity of the plate. The bony wall is very thin over most
of the alveolar surface, and it is seen to be perforated by numerous foramina
through which vessels, &c., pass to the base of the horny plate. B. The
surface of the right upper posterior horny plate as seen from below. The
internal border is very thick, and reaches a much higher level than any other
part ; and the anterior and posterior borders are much lower than the others.
The small anterior and internal concavity is at a somewhat lower level, and

46

EDWARD B. POULTON.

is excavated to a muck greater depth than the others. The divisions between
the compartments are lower than the borders of the plate.

Fig. 3. — Natural size. A. The socket of the right lower posterior plate>
seen from above. The concavity is in this case somewhat distinctly compart-
mented to correspond with the two chief parts of the plate. Each compart-
ment is roughly divided into four small concavities. There is also a very
distinct posterior and internal small compartment for the corresponding part
of the plate. The bone is pierced by numerous foramina, leading into the
very large canal for the inferior dental nerve. B. The surface of the right
lower posterior horny plate, as seen from above. The outer border is here
thicker and somewhat higher than the inner, but the highest part is the
anterior border, and especially its inner part ; the posterior border is also higher
than the lateral borders. The third small concavity is here posterior in position,
and it is not excavated so deeply as the others, and its edge is somewhat
higher than the adjacent borders. The divisions between the compartments
are lower than the borders of the plate. A comparison of the relative heights
of the borders and other parts of these plates will show that they are very
well adapted for interlocking in mastication, an antero-posterior motion being
especially favoured.

Fig. 4. — X 24-5. A transverse vertical section through the left upper
anterior horny plate. The section was taken towards the posterior end at
the maximum breadth of the plate. The ridge and furrow are seen in section.
The structure is obviously entirely epithelial, and passes into the oral epitke-
thelium on both sides. The section was cut and stained, and the stratum
Malpighii is clearly shown in the lower part, and the stratum corneum in the
upper. Numerous fine papilla; enter the former, and each of them sends up
from its summit a long column of soft, deeply-staining cells into the stratum
corneum. These columns reach the surface, and doubtless largely determine
the relative wear of the plate. Thus they are absent or very minute in the
ridge, and especially large in the furrow.

Fig. 5. — x 9. A similar section through one of the upper posterior plates ;
the slight elevation in the concavity being doubtless an oblique section of the
low ridge which separates the two concavities. This section had been ground
down and dried, so that the papilla; and columns of cells have dried up, their
place being occupied by air which causes the dark appearance. The raised
borders have only very minute papilla;, &c., in some places. The continuity
with the oral epithelium is also seen.

Fig. 6. — X 24-5. A horizontal section through the stratum corneum of
the plate shown in Fig. 4. The section was cut and stained, and shows the
columns of soft cells in transverse section. The position of the ridge and the
furrow can be determined by the size of the columns.

Fig. 7. — X 50. A horizontal section through the stratum corneum of the
concavity of the plate shown in Fig. 5. The section was cut and stained,

TRUE TEETH AND HORNY PLATES OF ORNITHORHYNCHUS. 47

and shows that the columns of soft cells occur isolated, and also arranged in
small groups. The concentric arrangement of cells round the columns is
indicated, and the corneous cells which make up the hard part of the structure
are represented by their darker central portions (better shown in the next
figure).

Fig. 8. — X 405. A single column of moderate size from the inner part
of the section drawn in Fig. 6. The column is seen in transverse section
surrounded by concentric cells, and these again by the matrix of corneous
cells. The column itself has stained deeply, especially the central cell, while
the concentric cells stain faintly and the corneous cells remain unstained.
Each of the latter contains a central mass of granular appearance, and con-
taining minute pigment granules. It probably represents the remains of the
nucleus, together with some of the granular material which occupies a much
larger space in many softer cells (compare the concentric cells of this section
and various cells in Fig. 11). This central portion remains comparatively
soft and dries up in ground sections, being replaced by air.

Fig. 9. — X 188. A horizontal section through the stratum corneum of the
concavity of one of the posterior horny plates. The section was ground
down, dried, and mounted in balsam, and the latter medium has displaced the
air from most of the larger spaces caused by the shrinkage of the columns
and the majority of concentric cells. Some of these latter, however, retain
abundant air and appear dark, and the same is true of the centres of the
corneous cells. In other cases the air may remain in the shrunken columns,
so that the resemblance to the Haversian systems of bone would be even more
striking than in the figure.

Fig. 10. — X 188. A part of Fig. 4, more highly magnified. The figure
represents a vertical section through the superficial stratum corneum of the
outer slope of the plate. Two columns of soft cells arc seen rising to the
surface through the matrix of corneous cells.

Fig. 11. — X 188. A vertical section through the lower part of one of the
posterior upper plates, including the bone. The space between the plate and
the bone is seen to be very narrow, when the magnification is taken into
account. This appeared to be the case in all the sections of the posterior
plates. One of the long thin papillae is seen together with the base of
another. The stratum Malpighii is of normal appearance : at its upper part
the cells become granular, and higher still become corneous peripherally.
Above this we enter the stratum corneum, where the cells are more flattened
and become almost entirely cornified. Nevertheless many cells occur,
especially in the lower part of this layer, in which the thickened border is
alone corneous, while the central part remains granular and is coloured by
carmine. At a higher level than that shown in the figure the nucleus ceases
to be distinct, but a central granular pigmented mass remains (compare
Fig. 8). The soft cells of the column which rises from the apex of the papillae

48

EDWARD B. POULTON.

are generally granular, although many of them retain the characters of the
cells of the deeper layers of the rete Malpighii.

Fig. 12. — X 405. A vertical section through a few granular cells which
are becoming corneous peripherally.

Fig. 13. — X 50. A transverse vertical section through the oral epithelium
of the right side anterior to the upper true teeth in the young animal. The
section was the seventh in front of the most anterior which contained teeth,
viz. that which is figured in Plate IT, Fig. 1. The thickened epithelial ridge,
which perhaps represents the anterior part of the posterior horny plates, is
shown in section.

Jtcvr. /aurw. WXA'S.ih.//

E B PouUon del

JS.

F Huth.Lith? EdinT

Fig. 16.

B s.rro Poc x 9. a.o.c. C e0 p.oc

s!ni'. , .

e.o

JLcrr. Jfou/nu'. /VZo. XXIX. NS. 9*1 III.

Fi£. 7 x so

Fi£. 6 x 188

m 77b.

E B Poulton del

Y Huth,Lithr Edin3:

E B Poulton del.

.Mm^otmo%i.WXjA IV.

Fig.

7 X 50

Fig. 11 X 188

o

F Huth.LitV EdinT

FATE OF THE BLASTOPORE IN RANA TEMPORARIA. 49

Note on the Fate of the Blastopore in Rana
temporaria.

By

Harold Sidebotlinm, M.R.C.S.

With Plate V.

In the description given in Balfour’s ‘ Comparative Em-
bryology ’ of the development of the Amphibia, that portion
of it which specially relates to the anourous Amphibia is
mainly founded on the researches of Gotte on Bombinator. It
is there stated that1 “ at first the mesenteron freely communi-
cates with the exterior by the opening of the blastopore. The
lips of the blastopore gradually approximate, and form a pas-
sage, on the dorsal side of which the neural tube opens ....
The external opening of this passage finally becomes oblite-
rated, and the passage itself is left as a narrow diverticulum,
leading from the hind end of the mesenteron into the neural
canal. It forms the post-anal gut and gradually narrows, and
finally atrophies. At its front border, on the ventral side,
there may be seen a slight ventrallv directed diverticulum of
the alimentary tract, which first becomes visible at a somewhat
earlier stage. This diverticulum becomes longer and meets an
invagination of the skin, which arises in Rana temporaria
at a somewhat earlier period than represented by Gotte for
Bombinator. This epiblastic invagination is the proctodseum,
and an anal perforation eventually appears at its upper
extremity.”

1 Balfour, * Comp. Embryology,’ 2nd edit., vol. ii, p. 130.

VOL. XXIX, PART ]. NEW SER.

D

50

HAROLD SIDEBOTHAM.

In a paper1 “On some Points in the Early Development of
Uana temp or aria” Professor Spencer states that he is unable
to find any trace of the inclusion of the blastopore by the
neural folds, or any trace of the closure of the blastopore itself.
He also states that soon after the stage has been reached in
which the neural folds have met, the hinder part of the neural
tube becomes closed, though the blastopore itself remains
open.

Miss Johnson and Miss Sheldon have also published some
“Notes on the Development of the Newt,”2 in which they
make some remarks with regard to the Frog. They apparently
agree with Professor Spencer in the opinion that the blastopore
persists as the permanent anus.

Mr. Herbert Durham3 has also stated that he fully agrees
with Professor Spencer in regard to the fate of the blastopore.

Being interested in this question I have cut a large number of
series of sections of Rana temporaria. A careful study of
these sections leads me to come forward and express my
opinion, and that with some degree of confidence, as more than
sixty embryos have been examined, that the history of the
blastopore more resembles the account of it given by Balfour
than that given by any of the other authors above alluded to.

After the formation of the neural folds, and while they are
still widely separated, the mesenteron opens to the exterior by
means of the blastopore, which is situated at the extreme
posterior end of the embryo. Fig. 1 shows a median, vertical,
longitudinal section, taken at this stage ; in it can be seen a
well-marked diverticulum from the hind end of the mesen-
teron, dipping down towards a distinct pit in the epiblast
below the blastopore and quite separate from it. This is the
rectum forming, and advancing to meet a true proctodseal pit.
This embryo has three mesoblastic somites. As the neural
folds grow up to meet each other they do not enclose the blas-
topore, but reach as far as its dorsal rim.

1 ‘Quart,. Journ. Micr. Sci.,’ vol. xxv, Supplement, 1885, p. 123.

2 ‘Quart. Journ. Micr. Sci.,’ vol. xxvi, 1886, p. 573.

3 ‘Quart. Journ. Micr. Sci.,’ vol. xxvi, 1886, p. 508.

FATE OF TEE BLASTOPORE IN RANA TEMPORARIA. 51

Fig. 2 shows the next stage in an embryo with five meso-
blastic somites. The neural folds have met and form a tube,
bending over the posterior end of the embryo, and opening
through the dorsal region of the blastopore, which has become
narrower and longer. Below the blastopore the proctodaeal
invagination may be seen to have increased considerably in
depth as also has the rectal diverticulum from the mesenterou.

In fig. 3 the same parts can be recognised, but the closure of
the blastopore has proceeded further. This stage has six meso-
blastic somites. The septum dividing the rectal diverticulum
from the proctodaeal invagination has become perforate. The
section of which fig. 3 is a drawing is not quite in the right
plane for showing the neurenteric canal.

Fig. 4 shows the next stage in an embryo with eight meso-
blastic somites. The blastopore is completely occluded, and is
represented by a column of epiblastic cells ( ep .) which may
be recognised by their containing much pigment. The anus
communicates freely with the alimentary canal, and the
neurenteric canal is still well marked. There is a slight
dilatation of this canal at the junction of its dorsal and ventral
limbs, which dilatation is the post-anal vesicle.

In the next stage, with nine mesoblastic somites, represented
by fig. 5, the tail has just commenced to bud out. The rem-
nant of the blastopore is represented only by a little heaping
together of the epiblast cells, where the blastopore opened to
the surface. The neurenteric canal has become occluded, but
still can be traced up as a solid rod of cells, representing the
post-anal gut, commencing from the normal position. There
is a diverticulum from the proctodseum just before it opens to
the surface which is the rudiment of the allantoic bladder.

Thus the history differs from that given by Balfour, in that
the neural folds do not enclose the blastopore, the closure of
the blastopore being effected subsequently to the meeting of
the neural folds. My conclusion differs essentially from the
description given by Professor Spencer, inasmuch as I find
that the anus is not derived from a persistent blastopore, but
is formed from an independent proctodmal invagination.

52

HAROLD SIDEBOTHAM.

Miss Johnson and Miss Sheldon, writing with reference to
the Newt,1 incline to the conclusion that the tail, as well as
the post-anal gut, is a secondary structure developed after the
permanent anus. Of course this view would be equally ap-
plicable to the same structures in the Frog, if Professor
Spencer is correct as to the fate of the blastopore.

But I think that the condition which obtains in the Frog at
the stage when the blastopore is still just open represents
an extremely primitive condition, for Balfour, in his paragraph
with reference to the post-anal gut and neurenteric canal,2
comes to the conclusion that the neural and alimentary canals
must have had a common opening, probably into a dilated
vesicle, before going directly to the exterior. This is exactly
what is found in the Frog just before the rectal diverticulum
becomes perforate.

Professor Spencer only figures one longitudinal section, and
this is from a stage subsequent to that in which the closure of
the neurenteric canal takes place, a stage, moreover, at which
all trace of the blastopore has gone.

During the time the blastopore is open it always runs in a
line with the mesenteron and opens posteriorly, while the rectal
diverticulum always runs ventrally, and keeps at about the
same angle to the mesenteron, even after the tail has become
fairly well developed.

In Professor Spencer’s fig. 15 the canal, the extremity of
which he marks (an.) and describes as the blastopore, I think
must be the rectal diverticulum, as I find just the same condi-
tion as there represented in my sections taken from embryos
at a corresponding age.

His figs. 5 and 6 are taken from an earlier stage than his
fig. 15, and show just the same features as do mine, but the
appearances they present may easily be interpreted in accord-
ance with my view, the opening he marks bl. in both figures being
the proctodacal invagination. In his fig. 5 the blastopore is

1 Loc. cit.

' 2 * Comp. Embryology,’ vol. ii, chap. xii.

FATE OF THE BLASTOPORE IN RANA TEMPORARIA. 53

continued posteriorly from the point ne. My figs. 6 and 7 also
show this.

Mr. Durhanr’s drawings also are all taken from a later stage,
but are not intended to illustrate the history of the blastopore
but the fact of there being a neurenteric canal. In the earliest
stage which he figures, the rectum is the ventrally inclined
diverticulum, and is shown exactly as I find it. The blasto-
pore is occluded previously to this stage.

Only one series out of over sixty has failed to show this
history. It was abnormal iu other respects than the blasto-
pore, and was pi’obably pathological altogether.

In conclusion, I wish to express my sincere thanks to Pro-
fessor Milnes Marshall for looking at my specimens, and
also for much kind advice in reference to the subject of this
paper.

54

HAROLD S1DEBOTHAM.

DESCRIPTION OF PLATE V,

Illustrating Mr. Harold SidebothanPs “Note on the Fate of
the Blastopore in Ran a temporaria.”

Reference Letters.

at. Alimentary canal, all. Allantois, bl. Blastopore, ep. Epiblast. m.
Mesoblast. n. c. Neureuleric canal. not. Notochord, n. y. Neural groove.
p.a.g. Post- anal gut. pr. Proetodseum. n. c. Neural canal, r. d. Rectal
diverticulum, sp. c. Spinal cord. x. Yolk.

Figs. 1 — 5. — Median vertical lougitudinal sections.

Fig. 1. From an embryo with three mesoblastic somites. The neural
groove is still open, except at the anterior end ; the blastopore is widely open ;
and the rectal diverticulum well marked.

Fig. 2. From an embryo with five mesoblastic somites.

Fig. 3. From an embryo with six mesoblastic somites.

Fig. 4. From an embryo with seven mesoblastic somites.

Fig. 5. F’rom an embryo with eight to nine mesoblastic somites.

Figs. G and 7. — Transverse sections from an embryo slightly older than
Fig. 1.

Fig. 6. Taken along line a — a in Fig. 1.

F’ig. 7. Taken along line b — b in Fig. 1.

F. Huth, LithT Fdinl

H Sid«Uothom del

MORPHOLOGICAL STUDIES

55

Morphological Studies.

No. I.— The Parietal Eye of the Cyclostome
Fishes.

By

J. Heard. Ph.D., B.Sc.

With Plates VI and VII.

Introduction.

The discovery of the parietal eye by de Graaf (No. 7), and
the beautiful account of its structure in a great many genera of
Sauria by Spencer (No. 14) are of so very recent occurrence,
and excited so much interest among zoologists, that I can
refrain from an historical account of their work and of previous
researches on the pineal body, with which this sense organ is
identical. Spencer has given a full statement of all that was
known of this organ in Lizards, and 1 shall therefore only pre-
face the following account of its structure in the Cyclostomata
by a brief notice of Ahlborn’s work on the subject. In his
paper on the brain of Ammoccetes Wiedersheim (No. 16)
had described the presence of a greyish-white pigment in the
pineal body, and this discovery was confirmed by Alilhorn
(No. 1, p. 233). Ahlborn has, in addition, given some account
of the minute structure of the organ in both Ammocoetes and
Petromyzon, and, indeed, as the result of his studies he pub-
lished in a separate essay (No. 2) some views on the nature of
the pineal body — independently of, but identical with, those
enumerated by Rabl.-Riickhard (No. 15) some time before.

56

J. BEARD.

The circumstances which led both these observers to the hypo-
thesis, afterwards converted into a fact by de Graaf and
Spencer, did not favour Ahlborn with the discovery of black
pigment in the pineal body of Ammocoetes. Had he chanced
to obtain sections such as I figure in figs. 1, 3 and 8 of
PI. YI, there can be little doubt that he would rightly have
regarded his idea of the rudimentary eye-nature of the pineal
body as more a fact than an hypothesis. I can, from my own
researches, easily understand Wiedersheim’s failure to find a
coloured pigment, black or otherwise, in the pineal body of
Ammocoetes, for I have only seen it in the three Ammocoetes
to be afterwards described, and such pigment appears to
occur very rarely in the parietal eye in the Ammocoetes stage.
But for a long time Ahlborn’s failure was to me an enigma
which I could only explain on the supposition that he had
never had fully adult Petromvzon in his hands, for I was
fortunate enough to find the black pigment in the first full-
grown Petromyzon examined, and it was some time before a
Petromyzon in which the pigment was absent came into my
hands.

Wiedersheim’s and Ahlborn's negative results regarding
the presence of black pigment are easily explicable, as will
afterwards be seen. Not all Petromyzon, and still less Ammo-
coetes, possess the black pigment in the parietal eye.

Since Spencer’s researches appeared, the only accounts of
actual work published on the parietal eye are the preliminary
notice of my own discoveries (No. 3, p. 246) and Beraueck’s
account of its development in Lacerta and Anguis (No. 4).

I shall have occasion further on to refer to Beraneck’s paper.
Here, be it remarked, that in his account of the development
there is little or nothing that was not already known.

I have also worked the development in these two forms, and
if I refrain from publishing the results it is only because I agree
with another observer, who also has investigated the matter,
that there is little or nothing to make known which is not
already common knowledge.

The Cyclostomata were chosen for the following research on

HOBPHOLOGICAL STUDIES.

57

account of their, in many points, exceedingly primitive charac-
ters. I hoped from their investigation to get insight into the
phylogeny of the parietal eye, for of that we know nothing.
Spencer’s otherwise beautiful researches do not appear to throw
any reasonable light on the former history of that organ, and
the only point about its ancestry that one can accept as the
outcome of Spencer’s work is his conclusion (No. 14, p. 233)
that “ the pineal eye may probably be most rightly considered
as peculiarly a sense organ of pre-Tertiary periods.”

As it was hardly to be expected that the organ had originally
developed in the group of Reptiles, the fishes were naturally
turned to as affording the most probability of finding the solu-
tion of the problems it presents. The result of an examination
of the Cyclostomata has been to alter Spencer’s statement, true
as it was when he wrote it, that “ there is not sufficient evi-
dence to prove or disprove the existence of the organ within
the group Pisces ” (No. 14, p, 233).

In addition to Myxiuoid fishes I have examined several
Teleostei, among them Callicthys, but with negative results
as to the presence of an eye-like structure.

Callicthys was brought under my notice by both Professors
Wiedersheim and Howes on account of the curious "parietal
foramen” it presents in the median line above the snout, but
in front of the brain.

This membranous portion of the skin has, however, nothing
to do with the epiphysis ; what its meaning is I am unable to
say, for I have only investigated its possible relations to a
parietal eye. In Myxiue, some Ammoccetes, and nearly all
adult Petromyzou examined by me the presence of a fairly
well-organised parietal eye could be determined.

I begin with the account of the structure of the parietal
eye in

The Ammoc(etes of Petkomyzon planeri.

1 have had at my disposal a very large number of Ammo-
ccetes obtained in Preiburg, and near Kirnhalde, in the Schwarz-
wald. In addition, Herr Schwarz, a pupil of Professor
Weismann’s, lent me sections of three Ammocoetes, to be

58

J. BEARD.

described shortly, and he has allowed me to figure them in
figs. 3, 5, 6, and 10. The three Ammocoetes of Herr Schwarz
probably came from the same brook, and are remarkable when
compared with every other Ammocoetes I have examined, in
that there is a very deep deposit of black pigment in
the parietal eye.

Ahlborn (No. 1, p. 230) has described very correctly the
topographical relationships of the pineal body in both Ammo-
coetes and Petromyzon, and all I need remind the reader of
here is that it is in both still connected with the brain, and
not pinched off from the pineal stalk, as in Anguis, and that
very early in development the pineal body is divided into two
vesicles, a dorsal one, the parietal eye, and a more ventrally
situated one, which never possesses black pigment, and never
presents any resemblance to an eye.

In the following account of the minute structure we are
solely concerned with the dorsal vesicle. I shall not describe
the structure of the ventral vesicle, which is represented as
seen in Petromyzon (longitudinal vertical section), in figs. 8
and 9, v. v.

The dorsal vesicle or parietal eye ( P.E .) lies some distance
below the surface of the body and within the skull, which dor-
sally is only membranous (figs. 3 and 5, PI. VI).

As pigment is at this period of the animal’s life but sparsely
scattered in the skin there is not such a marked pigment-free
spot above the eye indicating its position, as in Petromyzon.

The eye is almost hemispherical in appearance, and has the
anterior wall flattened. The anterior and posterior walls are
separated by a narrow space (figs. 3, 5, PI. VI), which is filled
with an albuminous coagulable fluid (fig. 9, cf.). The coagu-
lation of this fluid led Ahlborn (No. 1, p. 233) to the conclusion
that the cells of the two walls were connected by threads. This
is not the case, and Spencer has already suggested the expla-
nation given above (No. 14, p. 222, foot-note).

The anterior wall occupies the position but lacks the struc-
ture of a lens, as described by Spencer in Hatteria, &c., and by
de Graaf in Anguis. It is, however, thicker in the centre than at

MORPHOLOGICAL STUDIES.

59

the sides (figs. 3, 5, and 10, /.), and might perhaps be compared
to a slightly convex lens. The only thing that can be made
out in its structure is a fairly large number of rounded nuclei
lying in a certain amount of protoplasm. Cell boundaries
are here not to be made out in the Ammoccetes.

The posterior wall (figs. 3, 5, and 10, re.) presents structures
which one may compare with a retina, such as that described
by Spencer in Hatteria or Varanus (No. 14, p. 177).

This portion is much thicker than the anterior wall, and its
widest part is in the middle. It is figured in Schwarz’s three
specimens (in figs. 3, 5, and 10, re.). It presents from within
outwards a layer of longish rods ( rd .), the free ends of which line
the cavity of the vesicle. Without the rods is a layer of nuclei
(«., figs. 3, 4, and 10), and beyond these a second more scanty
layer of scattered nuclei («2.)is met with. The rods are elongated
cells, whose nuclei lie near their bases. In the three Ammo-
coetes mentioned the rods are more or less enveloped in a deep
black pigment, which extends to their bases, and even slightly
into the layer of nuclei beyond. The internal row of nuclei
are, like the fewer external nuclei, rounded, and could not be
traced into connection with the rods, though probably such
connections exist. The protoplasm in which the outer nuclei
rest is granular and fibrillar in appearance. Thus the retina
of the parietal eye of Ammocoetes presents practically the same
structure as that of Hatteria or Varanus figured by Spencer
(No. 14, PI. XIV, figs. 3 and 6). If the reader will com-
pare these figures with my figs. 3, 5, and 10, he will, I think,
be convinced of the agreement.

I have previously stated that more usually the parietal eye
retina of Ammocoetes presents no pigment. I have figured a
longitudinal section (fig. 4) of such an unpigmented parietal
eye, and this is typical of most Ammocoetes. While it presents
in other respects the same characters as the three specimens
described above pigment is very nearly but not quite absent ;
there are a very few minute dots, which are figured at p. s.}
fig. 4.

Where the pigment is absent it is not possible in specimens

60

J. BEARD.

prepared according to the usual methods, viz. sublimate or even
chromic-osmic-acetic acid, to make certain of the connection
of the rod-like retinal elements with the inner and outer nuclei.
As the demonstration of such connection is more the work of
the histologist than of the morphologist, and as I am concerned
here more with the morphological aspect of the matter, I am
content to have shown (1) the presence of black pigment in the
parietal eyes of some Ammocoetes, and (2) that the arrange-
ment of the rods and nuclei aud cells of the retina of the
Ammocoetes parietal eye is essentially that of the same elements
iu the more perfect organs of Hatteria and Varanus, as described
by Spencer.

As iu the adult the parietal eye of Ammocoetes is a variable
organ — a point which is naturally of importance in connection
with the question of its degeneration.

The Parietal Eye in Adult Petromyzon.

In Petromyzon planeri I have been able to make a fairly
exhaustive investigation of the organ iu question. Petro-
myzon marinus only came into my hands in the shape of one
specimen in a bad condition for histological work. For this
specimen, aud for an example of Bdellostoma, to he afterwards
mentioned, I have to thank Professor Howes.

Externally the position of the organ is marked by a large
whitish spot on the skin behind the olfactory hypophysial open-
ing. Iu Petromyzon marinus it is especially large, and, as
in Petromyzon fluviatilis, this spot is due to the absence
of black pigment over that portion of the skin. If this white
spot is any criterion for the existence of the parietal eye in a
fully-developed condition, as I believe is the case, then Petro-
myzon marinus and Mordacia mordax, as figured by
Gunther (No. 9, p. G93, fig. 318), must both possess the
parietal eye in a fair state of development. The latter form,
which is very rare, has not been at my disposal, and owing to
the condition of preservation of the specimen of Petromyzon
marinus I can give little information of the state of the

MORPHOLOGICAL STUDIES.

61

organ in question. The only two points I could observe were
that there is a large white spot, as stated above, and that a deep
depression in the cranium just beneath it is readily made out.
Now, as this depression in P. fluviatilis is always associated
with a fair development of the parietal eye in the individual,
I think we run little danger in assuming that the organ will
probably be found to be well developed in the marine form,
all the more as the marine form is certainly less degenerated
than the fresh-water one.

In adult fresh-water Petromyzon one finds the same varia-
tion in the presence or absence of pigment which we met with
in the Ammoccetes, a fact which partially accounts for the
non-finding of black pigment by earlier observers and especi-
ally by Ahlborn.

Relatively to the brain the organ in the adult lies further
forwards (fig. 1 ,P.E.), and is connected throughout life with the
brain by a somewhat long stalk. Its position and relations to
the left ganglion habenulee have been already described by
Ahlborn (No. 1, p. 233), and he has also recorded its division
as in the Ammocoetes into an upper and a lower vesicle, dorsal
and ventral.

As in the Ammocoetes the dorsal one alone concerns us
directly, for the ventral vesicle never presents any advance on
the development as described in the Ammocoetes.

The parietal eye in the adult usually lies in a deep depression
of the skull (figs. 1 and 8, s.f.), but if no pigment be present
in the eye, that is if the organ be ill developed, as happens
in some individuals, the corresponding depression in the skull
is also almost or entirely absent (fig. 9). This is a very
curious fact. The pigment in the skin (p. s.) does not reach
over the eye. And in longitudinal vertical section of the head
one sees that the pigment stops short (fig. 8), some distance
before and behind the organ.

Further, the amount of pigment deposited in the eye varies
in different specimens. In some the pigment is so thick as
entirely to conceal the structure of the retina (fig. 1 ,p.). In
others it is more modei’atcly developed (fig. 8,^.), and allows an

62

J. BEARD.

insight into the structural elements of the retina. In others
no traces of pigment are to be found (fig. 9).

This again, as in the Ammocoetes, is an indication of the
variability of the organ.

When I here speak of pigment I mean black pigment.
Ahlborn, following Wiedersheim’s discovery of grey-white pig-
ment in Ammocoetes, found in all his Petromyzon only white
pigment (No. 1, p. 233). I do not dispute these discoveries,
which appear to me well authenticated. As I have never
investigated perfectly fresh Petromyzon or Ammocoetes I have
never seen this white pigment, but as Ahlborn always found
grey-white pigment and no black, and asin my sections
if pigment could be recognised at all it was always
black, I think we may safely assume that in all those cases
of Petromyzon or Ammocoetes where I have found no pigment,
the grey-white pigment of Wiedersheim was originally present,
but was dissolved out in the process of preparing the sections.

The front wall of the vesicle iu adult Petromyzon is very
little different from what we saw in Ammocoetes. It contains
no pigment and is usually somewhat folded (figs. 1, 8, and 9).
It is composed of long cylindrical celis, and can hardly be said
to form a lens.

As in Ammocoetes the cavity of the vesicle is filled by a
coagulable fluid (fig. 9, c.f.) which in sections is drawn into
threads which appear to connect the two walls of the vesicle.
Of course the importance which Ahlborn attached to these
connections is negatived by their nature.

With low powers the posterior wall of the vesicle or retina
is seen to be made up of three layers : an inner layer of rods
(fig. 9, r .) which also contains the pigment ; following this a
layer of “ nuclei ” (fig. 9, n1.), and outside this a somewhat
granular striated layer, which contains a few ganglion-cells
(fig. 9, n 2.). Outside of all is the connective-tissue investment
of the eye.

The retina as figured in figs. 1, 8, and 9 would be compar-
able to the retina of Varanus or Ilatteria as described by
Spencer. However, it is interesting and important to examine

MORPHOLOGICAL STUDIES.

63

such a pigmentless retina of Petromyzon under very high
power.

Fig. 7 is taken from exceedingly good sections of such
parietal eye, and is drawn under Zeiss’s objective F.

The end elements are shown to be of two kinds, comparable
as it seems to me to those in the retina of the ordinary eyes.
By far the most numerous are the long rods ( rd .), but in
addition and between the latter one finds a few cones ( cn .).
The mode of connection of these end elements with the
“nuclear layer” is also figured, as well as their termination
in an outer ganglion cell layer {rftgl.).

The Parietal Eye in Myxine.

Although I have examined many Myxine in a better or
worse state of preservation, I have only found one in which
the structure of the organ in question could be well made out.
In fig. 12 I have drawn the general appearance and relation-
ships of the organ as seen in sagittal section under low magni-
fication. The eye is a large flattened organ lying within the
skull and connected to the thalemencephalon by a short, thick
solid stalk (fig. 12, st.). In the specimen under description
it contained no pigment, the dark portion of the vesicle
in fig. 12 being only the optical appearance of deeply stained
nuclei in a thick section. The epiphysis is here undivided
into two vesicles, in which respect it differs from the corre-
sponding organ in Petromyzon. Both anterior and posterior
walls have practically the same structure, although the
posterior wall shows the elements in a slightly better developed
condition. A piece of the retina (posterior wall) under high
magnification is figured in fig. 11. It shows that the struc-
ture is made up of a row of rod-like nucleated cells which
taper towards their bases. The tapering bases probably end
in some of the cells which form a scanty outer layer to the
retina. The nuclei of the rod-cells lie not very far away from
the cavity of the vesicle. Through the vesicle a number of
longitudinal fibres or striae pass (fig. 11). One may compare
the retina here described to that of Cyclodus as figured by

64

J. BEARD.

Spencer (No. 14, PI. XVI, figs. 18 and 19). It appears to be
more degenerate than that of Petromyzon, though hearing in
mind the variability of the organ in the latter form we must
not shut out the possibility that some Myxine may present a
much better developed eye than that under description.

In the one specimen of Bdellostoma at my disposal I could
make nothing out of the structure of the organ ; however, as
the brain of this form very closely resembles that of Myxine,
as Johannes Muller 1 first showed, it is not at all unlikely that
the resemblance will extend to the structnre of the parietal
eye. The two forms, Myxine and Bdellostoma, are certainly
very closely allied, and in other points of great importance,
such as the structure of their teeth, they closely resemble each
other and form a contrast to the Petromyzontidie, which while
in some respects less degenerate than the former are in others
less primitive.

General Considerations and Conclusions.

It was with the hope of getting at the phylogenv of this
remarkable sense organ that I began researches on its develop-
ment and its distribution in the group of Fishes. But indeed
the result was only to find that the development explains very
little. It is peculiarly one of those cases in which, as Dohrn
so often insists, “ niclit Anfangs und Endpunkt das wahre
phylogenetische Problem bilden, sondern der unbekannte Weg,
der sie verbindet.”

Leydig, whose misfortune it was to discover the organ in
Reptiles years before the modern perfected methods of re-
search enabled de Graaf and Spencer to convert Rabl-Riick-
hard’s (No. 15) and Ahlborn’s theoretical conclusion into
proved facts (No. 13, p. 535), has recently made use of
his undoubted right to an opinion on the question, and
declares his conviction 2 that the organ belongs to the system

1 ‘ Vergleichcnde Anatomie der Myxinoiden,’ p. 176.

2 From a former (erroneous) discovery of his that the organ in Batrachia
is enncrvated by the trigeminus. There can be little doubt that this is
wrong.

MORPHOLOGICAL STUDIES.

65

of sense organs of the lateral line, and that it finds a parallel in
the luminous organs of certain Fishes described by him (No. 13).

It is with regret that one must insist how impossible this
suggestion is, and how little likelihood there is that any
zoologist will adopt it.

I should be the last person in the world not to agree that
the system of lateral sense organs is a very remarkable one,
and one from which the so-called higher sense organs, except
the paired eyes, will be proved, if the proof is not to every-
body’s satisfaction yet complete, to have phylogenetically been
derived. And although the parietal eye may present resem-
blances in its structure to Leydig’s luminous organs, in all
other respects the proposed homology cannot be maintained;
and as soon as one attempts to compare the parietal eye in any
way with the lateral sense organs all possibility of their
homology vanishes. All the lateral sense organs develop
apart from the central system, and in connection with cranial
nerves and ganglia; while we have no facts as yet which show
that the parietal eye is otherwise than a portion of the central
nervous system, in which respect it agrees with the paired
eyes.

As things at present are, I see no advantage in a further
discussion of this matter, and beg to refer the reader who
wishes more light on the relations of the central to the peri-
pheral nervous system to the first part of my work on the
latter, which may see the light before this paper.

I may pass over Beraneck’s recent paper (No. 4) on the
development of the organ in Anguis and Lacerta, seeing that
it contains practically no facts of importance which were not
already known from de Graaf’s and Hoffmann’s researches
(No. 11). Of the latter, Beraneck appears to have been en-
tirely ignorant, although they cover a good deal of the ground
of his paper. Only one statement in Beraneck’s work calls for
notice, and that is his agreement with Spencer (No. 14) that
the lens passes continuously over into the retina in Anguis. From
his figures I do not suppose M. Beraneck’s specimens were in
a very good state of preservation, and I must undoubtedly

VOL. XXIX, TART 4. NEW SER. E

66

J. BEARD.

insist against Spencer and Beraneck that de Graaf (No. 8) was
right in his assertions.

In PI. II, fig. 13, I have given a figure of the eye in longi-
tudinal section ; it is taken from an advanced embryo of Anguis,
and shows very distinctly that the lens is well marked off from
the retina by a sharp line of division. I could, if I chose, give
similar sections through adult specimens, showing the same
fact.

I should not refer to this apparently trifling circumstance
were it not of great importance for some considerations to be
developed further on.

Regarding Spencer’s speculations on the origin of the
parietal eye from the larval Tunicate eye, I think I need say
little more thau I have already said in my paper in 'Nature’
(No. 3, p. 246). It reads: "With Wiedersheim and Car-
riere I consider that Spencer has placed the eye of the larval
Tunicate at the wrong end of the series — if it should come
in at all; for, as experience has abundantly shown, it is very
easy to compare organs of the higher Vertebrates with what
are supposed to be homologous organs in Amphioxus and the
Tunicata, and at the same time to be entirely in error. I need
hardly refer the reader to the instances in which such com-
parisons have been shown by Dohrn, in his well-known
‘ Studien/ to have been entirely wrong.” One might suppose
the degenerate nature of the Tunicata had been sufficiently
proved, and it is impossible to look with any favour on Spencer’s
attempt to re-establish that group in the position of ancestors
of Vertebrates, or, what is practically the same thing, near
allies of such ancestors. One thing more : in Spencer’s dia-
grammatic plate (No. 14, PI. XX), illustrating "the rise and
fall” of the parietal eye, he begins with a slight evagination of
the brain (larval Tunicate, fig. 1), which shows one layer of
cells, whose inner ends, their bases, are evenly pigmented. The
next two stages (PI. XX, figs, ii and in, Bufo) the pigment is
more confined to the centrally situated cells of the evagination,
that is, in those cells which, if the thing developed into a parietal
eye like that of Hatteria, would form the lens. This would

MORPHOLOGICAL STUDIES.

67

be inconvenient — very much so. So in PI. XX, fig. 4 we have a
figure of an early stage of a higher chordate, in which the pig-
ment has all disappeared ; and when in further development
(PI. XX, figs. 7 and 8, Reptiles) we find it again, it is confined to
its proper place in the retina, and the lens contains no pigment ;
while if it had developed from the structures in figs, i, xi, and hi
— if it could perform the physically impossible task of develop-
ing— it would be loaded with pigment. Now, these diagrams are
not untrue to nature, and all my criticism aims at proving is that
Spencer’s arrangement of them is artificial and misleading. It
would be a misfortune if these diagrams got into the text-books
in the order in which Spencer gives them. Even if placed
in a less artificial order they do not show the phylo-
genetic development of the organ — that is unknown.
All they show, if placed in a different order, is certain stages
of the ontogeny and certain stages of the degeneration.1 The
ontogeny is shown in figs. 4, 5, 6, and 7, and the degeneration
in figs. 6, 8, 10, 11, 12, and 9. These latter figures are a very
heterogeneous assemblage, and only show the state of degenera-
tion in a series of forms, and not the phylogenetic degeneration.
Seeing that most of the epiblastic cells of Anura contain pig-
ment, I do not see any advantage in placing figs. 2 and 3 in
the series at all, while if fig. x has any place in the plate it
ought to be last of all.

The phylogeny of the parietal eye is a very difficult problem,
and in spite of my former remarks (No. 3, p. 248) I do not think
the question can be yet fully solved. Spencer (No. 14, p. 230)
has compared its development with that of the paired eyes,
which he believes originated as secondary differentiations from
the brain — as secondary evaginations. This mode of regarding
the problem is easily disposed of, for if the paired and unpaired
eyes originated in that way, then in both cases the lens must be
the same ; and indeed, on physical principles, it is easy to un-
derstand that in the paired eyes the lens must of necessity be
formed as it is from the lateral epiblast. The anterior wall of

1 It is simply a fallacy to suppose that an organ in its degeneration passes
through the stages, or even some of them, of its phylogeny.

68

J. BEARD.

the optic vesicle, as it grows towards the surface of the body,
must catch the light, and this surface is obliged to remain as
retina. If the phylogeny of the unpaired eye were the same as
that of the paired eyes, the retina would be of the so-called
Vertebrate type.

Biitsclili (No. 5, p 178), in dealing with the problems pre-
sented by the eye of Pecten, sees the solution in the nature of
the lens. It seems to me that here, as in the Vertebrate eye, it is
the form of the retina — a closed cup — which gives rise to the
cellular lens and the inverted retina.

The appearances that one meets in Pecten are carried still
further in the eye of Onchidium. I have, through Professor
Howes’ and Dr. Gunther’s kindness, been able to study this
peculiar eye, though, as there were not many eyes on the two
specimens at my disposal, I could not follow the development.
In spite of Patten’s off-hand criticism in his paper on “Eyes
of Molluscs and Arthropods,” Semper was right in his state-
ments that the eye is pierced by the optic nerve, and that thus
an eye of the so-called Vertebrate type is formed.

An interesting point in my specimens is that the nerve is
double, and enters the optic cup at two points. This, I think,
throws light on the way in which the Onichidium eye has
developed from an eye of the Pecten type. The nerve-fibres
must originally, as in Pecten, have gone round the front
wall of the cup to their destination, and their piercing the
hinder wall is only a shorter way of getting to their destina-
tion.

After all, I think the development does show that the parietal
eye is a slightly later development than the paired eyes, but I
still hold to the view that the organ has developed in connec-
tion with the paired eyes. For this conclusion the two sorts of
end elements, rods, and cones, described by me in Petromyzon,
are of importance, as is also the fact that fibres have been traced
from the thalami optici to the epiphysis.

Most of us now accept the view of Balfour, Carriere (No. 6),
and others, that the eyes were once structures opening dorsally
on the surface of the unclosed neural plate, somewhat in the

MORPHOLOGICAL STUDIES.

69

fashion of figs. 3 and 16, PI. XI, of Heape's memoir of the
Mole (No. 10).

The parietal eye did not then exist (PI. VI, fig. a) . On the
closure of the neural plate the eyes of course got shut in, and
in order that no lens of the so-called Vertebrate type should
be formed from the epiblast of the median neural line above the
eyes, one must suppose that the median suture of the brain
was not composed of nervous sensory epithelium like that of
the retina (PI. VI, fig. b). A retinal epithelium of the median
dorsal line could not degenerate to form a lens like that of
Hatteria. It would be excited by the light, and a lens, if
formed, would arise from the indifferent epiblast. A piece of
ordinary nervous tissue, on the other hand, would degenerate
into an epithelial structure. We have instances of that in the
pallium of Teleostei and Ganoids ; and such a piece of tissue
must be postulated in the median suture of the brain above
the paired eyes.

If this be granted, the development of the parietal eye as an
apparently unpaired organ is easily explicable. After the
closure, according to Balfour, Wiedersheim and others, in their
phylogenetic development the paired eyes would receive light
from two sources, through the skin of the lateral surface of
the body, and through the suture of closure. As they grow
towards the surface a portion of the retina of each of them
still receives light through the suture, and it is this portion
which forms the retina of the parietal eye.1 Its lens is formed
by the epithelium of the suture which we assume is not sensory.

The process of this hypothetical development I have figured
in the three diagrams in PI. VI. The way in which this
subsidiary eye could be thus developed from part of the
sensory epithelium of the paired eyes is strikingly exemplified
in the actual facts of the development of Jacobson's organ in
Reptiles from a portion of the olfactory epithelium. Of this
development of Jacobson's organ I am preparing a memoir
which will soon follow these lines.

1 Thus, if no trace of the parietal eye now existed one could arrive at the
conclusion that such must once have been the case by induction.

70

J. BEAED.

And another instance is shown in the ontogenetic and
phylogenetic development of all the complicated parts of the
auditory organ from one bit of sensory epithelium. This
development of subsidiary sense organs from one piece of
parent sensory epithelium is a most remarkable fact of
embryology to which I hope to draw the attention it deserves.

Distribution of the Parietal Eye.

These researches show that the parietal eye was developed
in the group of Fishes, and still has the characteristics of an
eye in the very primitive group of Cyclostomata. It is not
impossible that in other fishes it may still present a good
development, though it is not very likely that such will be
found to be the case in any existing Elasmobranchii and
Ganoids.

In this connection a figure in ZitteFs ‘ Paheontologie ’ 1 seems
to me very interesting. It represents the bony skeleton of a
Placoderm Ganoid, Asterolepis ornatus, from the Old Red
Sandstone, and on the dorsal surface in the centre of a bone,
marked os dubium, there is something which looks suspiciously
like a parietal foramen.

Why the eye has degenerated can hardly yet be determined.
No doubt it has suffered in its competition with the paired
eyes. Apparently, too, it was worse fitted out with accessory
structures such as muscles, &c., than these.

I shall not attempt to discuss the question of whether it
is still functional or not in Cyclostomata. As no lens is
developed there it can be of little use as an organ of vision,
while I think Wiedersheim (No. 17, p. 149) has made out a
good case for its functional use in such forms as Hatteria. In
Cyclostomata it has all the characteristics of a degenerate
organ, one especially in a very high degree, viz. its variability

1 Zittel, * Handbuch der Palseontologie.’ Abtheilung I. “ Palseozoologie,”
Bd. iii, Heft. 1, tig. 161, p. 155.

MORPHOLOGICAL STUDIES.

71

in different individuals. I have only seen black pigment on it
in three Ammocoetes, while the majority of the adult Petro-
myzon I examined had such pigment in a greater or less
degree. I am hence forced to abandon as unlikely the idea
that black pigment is formed in the larva, that it then gives
place to white, and again in the adult a reversion to black
occurs. This seems to me now unlikely, and the only conclu-
sion I can draw is that the pigment is very variable, but that
as a rule sooner or later black pigment is formed in the
parietal eye of the Cyclostomata.

List of Writers Cited.

1. Ahlborn, F. — “Untersuchungen fiber das Gehirn der Petromyzonten,”

‘ Zeitschr. f. wiss. Zool.,’ Bd. xxxix, Heft 2, p. 191, 1883.

2. Aiilborn, F. — “Ueber die Bedeutung der Zirbeldriise,” ‘Zeitschr. f.

wiss. Zool.,’ Bd. xl, Heft 2, p. 331, 1881.

3. Beard, J. — “ The Parietal Eye in Fishes,” * Nature,’ No. 921, July 11th,

1887, p. 216; August 11th, No. 928.

1. Beraneck, E. — “Ueber das Parietal Auge der lleptilien,” ‘Jenaische
Zeitschr. f. Naturwiss.,’ Bd. xxi, Hefte 3 and 1, p. 371.

5. Butsculi, 0. — “ Zur Morphologie des Auges der Muscheln,” ‘ Verhandl.

der Nat. -Med. Vereins zu Heidelberg,’ 1880, p. 175.

6. Carriere, J. — ‘Die Sehorgane der Thiere,’ Muuchen u. Leipzig, 1885.

7. De Graaf, H. W. — ‘ Bijdrage tot de Kennis van der Bouw eu de

Ontwickeliug der Epiphyse bij Amphibien en Reptilicu,’ Leiden, 1886.

8. De Graaf, H. W. — “ Zur Anatomie u. Entwickelung der Epiphyse bei

Amphibien u. Reptilien,” ‘ Zool. Anz.,’ March 29th, 1886.

9. Gunther, A. — ‘The Study of Fishes,’ 1880.

10. Heape, W. — “The Development of the Mole,” ‘Quart. Journ. Micr.

Sci.,’ Oct., 1886, p. 123.

11. Hoffmann, C. K. — “ Weitere Untersuchungen zur Entwickelungsgesch.

der Reptilien,” ‘ Morphol. Jahrb.,’ Bd. xi, 1885.

12. Julin, Ch. — “ De la signification morphologique dc l’epiphyse des Verte-

xes” (a compilation), ‘Bull. Scieutif. du Departement du Nord,’
Paris, 2me Serie, Xme Annee, 1887.

13. Leydig, F. von.— “ Das Parietalorgan der Wirbclthiere,” ‘Zool. Anz.,’

No. 202, Oct. 10th, 1887, p. 531.

72

J. BEARD.

14. Spencer, W. B. — “ On the Presence and Structure of the Pineal Eye in

Lacertilia,” ‘Quart. Journ. Micr. Sci.,’ Oct., 1886, vol. xxvii, part 2,
p. 239.

15. Rabl-Ruckhard. — “Zur Deutuug u. Entwickelung des Gehirus der

Kuochenfische,” ‘ Arcliiv f. Anat. u. Physiol.,’ 1882, p. 111.

16. Wiedersheim, R. — “ Das Gehirn von Ammocoetes u. Petromyzon

Planeri,” ‘ Morphol. Studien,’ I.

17. Wiedersueim, R. — “ Ueber das Parietal Auge der Saurier,” ‘ Anat.

Anz.,’ Jahrgaug I, No. 6, 1886, p. 148.

DESCRIPTION OF PLATES VI & VII,

Illustrating Dr. Beard’s Memoir on “ The Parietal Eye of the
Cyclostome Fishes.”

Alphabetical List of Reference Letters.

ant. Anterior, bl. Blood, br. Brain, cn. Cones of retina, co. Coagu-
lated fluid, c. t. Connective tissue, ep. Epidermis, gl. trb. Ganglion tra-
beculoe. 1. Lens cells, md. br. Mid-brain. nl. Inner nuclear layer, m2.
Outer nuclear layer. n~. gl. Ganglion of outer layer. P. E. Parietal eye.
p. s. and p. Black pigment. r. Retina. rd. Rod elements of retina.
s.f Skull fossa for parietal eye. sic. Skull, s. Stalk, th. Thalamence-
phalon.

All the figures in Plates VI and VII are from camera lucida drawings,
except A, B, and C.

PLATE VI.

Eig. 1. — Longitudinal section through skull of adult Petromyzon
planeri, showing parietal eye ( P . E.) in situ. Zeiss C, oc. 2.

Eig. 2. — Transverse section of ordinary eye of Ammocoetes, drawn under
the same magnification as the following figure.

Fig. 3. — Transverse section of parietal eye, taken from the same animal
as preceding figure for comparison with it. Figs. 2 and 3 under same magni-
fication, viz. Z. D, oc. 2. These two figures are from one of Herr Schwarz’s
preparations.

Fig. 4. — Parietal eye of adult Petromyzon planeri in longitudinal
section. The only traces of black pigment are present at ps. Zeiss F, oc. 2.

MORPHOLOGICAL STUDIES. 73

Fig. 5. — Parietal eye of Ammoccetes in transverse section. From the
second of Herr Schwarz’s preparations. Zeiss D, oc. 2.

Fig. 6. — Portion of the retina (/•.) and lens (/.) of the parietal eye of the
preceding preparation under high power. Zeiss F, oc. 2.

Figs. A, B, C. Diagrams showing three hypothetical stages in the develop-
ment of the parietal eye.

PLATE VII.

Fig. 7. — Retinal elements of parietal of an adult Petromyzon planeri
under high power. There was no black pigment in the specimen. Zeiss F, oc. 2.

Fig. 8. — Parietal eye of adult Petromyzon planeri, showing position
of parietal eye, the skull fossa ( s.f. ), and the distribution of the pigment in
the skin over the eye. Longitudinal vertical section. Zeiss A, oc. 2.

Fig. 9. — Parietal eye of an adult Petromyzon planeri, in which there
was no pigment on the retina, showing the absence of pigment and of the
skull fossa of the preceding figure. Longitudinal vertical section. Zeiss
C, oc. 2.

Fig. 10. — Retina of the parietal eye of Herr Schwarz’s third preparation.
Transverse section. Zeiss D, oc. 2.

Fig. 11. — Retinal elements of the parietal eye of Myxine glutinosa
under higher power. Zeiss F, oc. 2.

Fig. 12. — Longitudinal vertical section of brain and parietal eye of
Myxine glutinosa. No pigment in the retina. Zeiss A, oc. 2.

Fig. 13. — Longitudinal vertical section through the parietal eye of an
advanced embryo of Anguis fragilis under high power, showing distinct
boundary at B between lens and retina. Zeiss F, oc. 2.

V/

Fiq. 3.

*'\V

8 a O'1'*; ' ^ «K‘

i xi a A * v ' ’

PETROMYZON

/©3®& m

AMMOCITES

AM MOCCETES

AM M OCCETES

AM MOCCETES

J Beard del

F Huth, L'.thf Edin'

MtorJourm 'doi.WXJ.sM V//.

»s*r^88e6.

MBn.

ON SOME OIGOPSID CUTTLE FISHES.

75

On Some Oigopsid Cuttle Fishes.

By

F. Ernest Weiss, F.E.S.,

From the Zoological Laboratory, University College, London.

With Plates VIII, IX, and X,

At the instance of Professor Lankester I have undertaken
a careful examination of some of the Cephalopoda forming
part of the valuable collection he has gradually acquired for
the Museum of Comparative Anatomy at University College,
London,

All the species dealt with in this paper belong to the group
of the Oigopsida, of whose anatomy and general structure our
knowledge is at present still very scanty. This may be said to
be especially the case with regard to some of the rarer forms,
such as Chiroteuthis and Doratopsis, which I have been enabled
to study. Besides these two my paper deals with some points
in the anatomy of Iiistioteuthis Riippelli, Tracheloteu-
this Behnii (Strp.), and Verauia sicula (Krohn).

Chiroteuthis, Histioteuthis, and Loligopsis (including the
form now called Doratopsis) were all united originally by
D’Orbiguy1 (1) in the family of the Loligopsidte, one of the
most important and distinguishing characters of which was the
absence of a valve in the siphon.

Brock, in 1880, 2 divided the Oigopsidae into two groups, the
Ommastrephes group and the Loligopsis group. To both he
attributed the general Oigopsid characters, which included,

1 D’Orbigny, ‘ Cephalopodes acetabuliferes,’ 1835.

2 Brock, ‘ Morphologisches Jakrbuch,’ 1880.

76

F. ERNEST WEISS.

according to him, the slit-like renal openings, loss of accessory
nidimental glands, and a uniting commissure between the
stellate ganglia. The Loligopsis group he characterised further
by the absence of a siphonal valve and the loss of one of the
oviducts (the right one in Chiroteuthis Veranyi). This
group included Chiroteuthis, Loligopsis, Owenia, and
possibly Hi stioteuthi s and Verauia, as also devoid of a
valve in the funnel.

But the name Loligopsis has now been restricted to the
genus formed by Lamarck from a specimen described by Peron
and Lesueur, which resembles Sepiola, with the exception of
having a rhomboidal fin and only eight arms ; and we can
therefore no longer adopt the name of Loligopsidse for a family
containing Histioteuthis and Chiroteuthis. Hoyle,1 in 1886, in
his ‘ Report on the Cephalopoda/ adopts a classification which
places Chiroteuthis, Histioteuthis, and Doratopsis, together
with other genera (Histiopsis, Brachioteutliis, and Calliteuthis),
in a family of Taonoteuthidse (Steenstrup, 1861), with the
single subfamily of Chiroteuthidm (Gray, 1849). Speaking of
the general characters of the family, he says, “ There seems
to be some uncertainty as regards the presence of a valve ; for
though the older observers affirm its absence, Verrill, in a
species of this genus (Chiroteuthis) of the Northern Atlantic,
distinctly affirms that a valve is present ; and Professor Lan-
kester informs me that in a Chiroteuthis V eranyi, in Uni-
versity College Museum, London, there is a very small, in fact
a rudimentary, valve, just a transverse fold, not projecting
much, and that he has acquired a Histioteuthis with a well-
developed valve in its funnel.”

I am able to confirm these observations of Professor Lan-
kester, and to add several new points to what little was known
before of these very interesting forms. The careful examina-
tion of these forms leads me to uphold the uniting of Chiroteu-
this, Doratopsis, and Histioteuthis into one family, as is done by
Hoyle, and was done also by D’Orbiguy, though the classifica-
tion of the latter was based on many erroneous observations; but
1 W. E. Hoyle, ‘ lteport of H. M. S. Challenger,’ “ Zoology,” 16, 1886.

ON SOME OIGOPSID CUTTLE FISHES.

77

the new facts I have been able to make out will place the Chi-
roteuthidae in a different relation to the remaining families
of the Oigopsida, and will necessitate a slight change in the
classification of this group, which I shall suggest at the close
of this paper.

I shall begin with the consideration of the type genus of
the Chiroteuthidae, the species being the one named after
Verany, and figured both by D’Orbigny (1) and by Verany.1

Chiroteuthis Veranyi (Ferussac), D’Orbigny.

The specimen I examined was slightly smaller than the one
of which Verany gives measurements in his description. It was
purchased at Nice by Professor Lankester for the Museum of
University College. The lengths of this specimen are as follows:

Body without arms, but including (in

11-2

cm.

Fin (length)

3-

>>

„ (breadth) .

3-5

>>

Breadth of body

2-

99

1st arm ....

9-

99

2nd „ .

11-2

99

3rd „ .

12-

„ with a median (in.

4th

17-3

„ with lateral external Cn.

Tentacular arms

8-

99

The order of the arms in descending order of the lengths is
4, 3, 2, 1, which order is also given by Verany, though, owing
probably to a printer’s error, the measurement for the third
arm is given as the smallest. Each pair of arms is provided
with two rows of suckers, which are well figured by D’Orbigny,
the fourth pair of arms being, besides, provided with a row of
deeply-pigmented warts, somewhat like those which occur all
over the body of Iiistioteuthis, and are possibly phosphorescent
organs.

The tentacular arms are exceedingly long, and carry modified
suckers at varying intervals along their whole length. The
club has modified suckers, as figured by D’Orbigny, and is
provided with two lateral-fluted webs (Shutzmembranen).

1 Verany, ‘ Cephalopodes de la Mediterranee,’ 1851.

78

F. ERNEST WEISS.

The buccal membrane is large, drawn out into seven points,
and is devoid of suckers.

The nuchal cartilage is flattened, and of what von Ihering1
(5) calls the Sepia type.

The fastening of the mantle to the base of the siphon (PI.
VIII, fig. 4) is by two depressions on the funnel, complicated
by a lateral tooth, and by triangular cartilages on the mantle.

The interior of the funnel is provided with a valve (PI. VIII,
fig. 5) near its apex, which is drawn somewhat to a point in
the centre. It is small, but not rudimentary, though it seemed
smaller on first observation, as it adhered to the dorsal wall of
the funnel.

Midway between the eyes and the funnel on the ventral
(infundibular) side of the body are two spoon-shaped organs,
like those figured for Cliiroteuthis lacertosa by Verrill2 (6),
but which have neither been described nor figured for Chiro-
teuthis Veranyi by D’Orbigny nor by Brock. They possess
a small ganglion at their enlarged base, which seems to be
supplied by a nerve from the cerebral ganglia, and I regard
them as at all events originally olfactory in fuuction. I will,
later on, point out their homology to the olfactory organ of
other Cephalopoda, and with regard to the other groups of
Mollusca, Professor Lankester has suggested to me that they
might be homologous to the Gasteropod tentacle.

The body of Chiroteuthis is elongate and very small com-
pared to the size of the head and arms.

On opening the mantle cavity we find the rectum with anal
appendages not reaching up to the base of the siphon.

The ink-sac is short, triangular in shape, and bears on its
surface two glandular organs, which in position especially
greatly resemble the accessory nidimental glands in Loligo,
which resemblance is increased by the relation to these organs
of the branches of the visceral nerves (PI. VIII, fig. 5). These
organs are visible in Verany’s figures by transparency (PI. 39).

Ventrally and slightly anterior to the base of the gills are

1 von Ihering, * Zeitschrift fiir wissenschaftliche Zool.,’ 1881.

2 Verrill, ‘ Transactions Connecticut Academy,’ 1882.

ON SOME OIGOPSID CUTTLE FISHES.

79

two well-developed renal papillae (PI. VIII, fig. 4), which pa-
pillae are continued and expand beneath the wall of the renal
chamber (PI. VIII, fig. 5).

At the side of the gills, coming up behind them, are the ovi-
ducts without terminal oviducal glands. These glands are
halfway up the duct (PI. VIII, figs. 6 and 7), and below the
glands the oviduct shows an annular marking and narrows
down to a small opening into the coelom (PI. VIII, fig. 7). As
I mentioned before, Brock (2) asserts that the right oviduct is
absent in Chiroteuthis V eranyi, which I find, in this speci-
men at least, is not the case. Chiroteuthis does not, therefore,
differ from the general Oigopsid type in this character.

Far back in the body are two nidimental glands (PI. VIII,
fig. 4), the existence of which is denied by Brock (2). They
are narrow and long, but show distinctly the structure of a
nidimental gland (PI. VIII, fig. 8).

On opening the renal chamber we find it to be a single one,
like that figured by Vigelius1 (7) for Ommastrephes, and which
Grobben2 (8) takes to be typical for Oigopsida. There is no
true dorsal renal chamber, but a more anteriorly situated di-
vision of the single chamber (fig. 7).

The vena cava passes down the right side of the ink-sac,
and divides close to its entrance into the renal chamber into
two renal veins, which receive two branches from the mantle.

Two distinct viscero-pericardial apertures lead from the
renal chamber to the coelom (PI. VIII, figs. 5, 6 and 7). The
pericardial portion of the coelom contains the heart, which is
large and broad, especially in the ventricular part. The pos-
terior portion of the coelom contains the stomach, which
reaches far back, the caecum and the ovary.

The branchial hearts lie in diverticula of the coelom. The
ovary partly overlies the stomach and is attached at two points
anteriorly to the stomach and posteriorly to the wall of the
coelom, as Brock has laid down for all Oigopsida3 (9) (fig. 7).

1 W. T. Vigelius, ‘ Niederlandisches Arcliiv fiir Zoologie,’ 1880.

- C. Grobben, ‘ Arbeiten des zool. Instituts zu Wien,’ 1884.

3 Brock, ‘ Zeitsclirift fiir wissenscliaftliche Zoologie,’ 1882.

80

F. ERNEST WEISS.

I was not able to make out any commissure connecting the
stellate ganglia, which was, perhaps, owing to my not being
able to dissect to the uttermost this valuable specimen which
was to be remounted for the Museum.

The pallial nerve is given off from the stellate ganglion itself.

Doratopsis vermicularis (Riippell), de Rochebrune,
[Loligopsis vermicularis (Verany)].

Plate IX.

The specimen studied by me was obtained at Messina, and
was presented to Professor Lankester two years ago by Pro-
fessor Kleinenberg, of the University of Messina, together with
several rare Cephalopoda.

From the excellent drawings of this specimen made by Miss
Stone, several interesting points can be made out which are
not given by Verany in his PI. 28, a, b.

The general proportions of the specimen I examined, which
was considerably smaller than the one described by Verany,
differ considerably from those of the latter.

Compared with Miss Stone's drawings, which are very accu-
rate, Verany figures the fourth pair of arms too short and
stout, and the three other pairs too long. The region between
the base of the arms and the eyes is too long, the neck too
thin, and the spine at the end of the body too long in fig. a.
In fig. b the spine is too broad, its length being about right.

In neither figures does he show what I shall presently
describe as the stellate organs and the olfactory organs
respectively.

The order of the arms is as in Chiroteuthis, 4, 3, 2, 1, and
the tentacles exceed the fourth pair in length. The arms,
1, 2, and 3 are very short compared to the size of the body,
with no dorsal webs or fins, and bear two rows of small
sessile suckers.

The fourth pair of arms are enormously large compared to
the others, more transparent and thicker than the tentacular
arms. They possess only one row of suckers disposed at
greater intervals.

ON SOME OIGOPSID CUTTLE EISHES.

81

Pfeffer1 (10), however, says that they seem to have a double
row of suckers, but here he is mistaken.

The fourth arm, indeed, has a broad lateral expansion
similar to that of Chiroteuthis, and this expansion bears small
thickenings which correspond to the ridge-like projections on
the lateral membrane of the club of Chiroteuthis.

Verrill, in Leptoteuthis diaphana, which is really a
Doratopsis, figures two rows of suckers and pigment spots,
like those of Chiroteuthis, on this fourth pair of arms.

The tentacular arms are long and terminate gradually in
club-like expansions, bearing circular sessile suckers in four
rows. These extend some way down the arm. The club is
provided laterally with a protecting membrane (fig. 7).

The mouth is surrounded by a thick papillate lip and a
narrow buccal membrane (PI. IX, fig. 4).

The eyes are large, but not pedunculate, and their opening
has no lachrymal sinus.

Somewhat below the eyes, on the ventral side of the body,
project two small organs (fig. 3, olf. ory.), which seem to be
supplied by nerves, and must, I think, be taken as homologous
with the spoon-shaped organs of Chiroteuthis, and olfactory in
function.

On the dorsal side of the cerebral nervous mass two reddish
spots are noticeable, the nature of which I was not able to
ascertain.

Behind the cephalic mass we find a very much elongated
and almost perfectly transparent neck region. This portion
of the body presents an appearance like that of segmentation,
by a series (8 in this specimen) of plates with radiating margin
lying along the median dorsal line. These organs, shown
enlarged in fig. G, overlie directly the two strangs of visceral
nerves, and seem supplied by branches from the aorta. I shall
call them the stellate organs.

The alimentary canal passes along the side of the aorta,
the vena cava along the ventral side of the neck.

1 Pfeffer, ‘ Abhandlungen des Naturwissenschaftliclien Vereins,’ Hamburg,
1884.

VOL. XXIX, PART 1. NEW SER.

F

82

F. ERNEST WEISS.

The mantle is attached to the head or rather neck by a
cervical plate of cartilage, somewhat narrower in proportion
than that of Chiroteuthis.

The mantle fastening by tooth and socket on the siphon is
almost identical with that of Chiroteuthis (fig. 5).

The apical portion of the funnel is bent almost at right
angles to the body, aud when opened a valve is seen just
below the bend. The valve is proportionately stronger than
in Chiroteuthis. On the ventral side of the funnel are two
glandular pads, which I can only compare with those found in
Yerrill’s organ (fig. 3).

On opening the mantle two very short papillae may be seen
leading into the renal chamber, which is short and broad
notwithstanding the elongation of the body (fig. 5).

The specimen examined was sexually not well differentiated
and probably still young. There were no nidimental glands
present, and there seemed to be only one genital duct on the
left side. I took it to be a male specimen. The female, I
think, would have shown two oviducts and nidimental glands
considering its general concordance with Chiroteuthis.

The pallial nerve came direct from the stellate ganglion.

Dimensions of Body.

Entire length of body without arms

52

cm.

Length of fin

IT

Breadth „

1-

Spine beyond fin

•8

Edge of mantle to beginning of fin

1-8

>>

„ „ to eyes

1-2

Arm 1

T2

»

„ 2

•23

»

„ 3

•28

„ 4

1-5

Tentacular arm

2 T

The spine at the end of body bears several swellings which
seem to be of a glandular nature.

ON SOME OIGOPSID CUTTLE FISHES.

83

Histioteuthis Ruppelli (Verany). PI. X, figs. 8 — 12.

The specimen examined by me was purchased at Nice in
1886, by Professor Laukester, for the museum of University
College.

The general external features are noted by Verany. The
suckers of the short arms are pedunculate, globular, and show
four large teeth on the upper margin of the chitinous ring.
The suckers of the tentacular clubs are sessile, and have teeth
all round the chitinous ring.

The tentacular club is badly represented by Verany, and re-
sembles more that figured for Calliteuthis by Hoyle (3).
Its extremity is drawn out into a narrow strip with suckers
in two or three rows. As the club expands we get four
rows, two of which have much larger suckers than the other
two. The club has a median dorsal fin, extending along
the distal half, and an external lateral fin along the proximal
half of the club. On the internal margin of the club are some
small suckers and cushions or pads, alternating, and forming
the fixing apparatus of the tentacles. These suckers and pads,
thirteen to fourteen in number, extend, with gradually en-
larging intervals, about halfway along the arm (PI. X, fig. 12).

The eyes are large, but show no lacrymal sinus. Between
the eye and funnel, and situated ventro-laterally, are two
lappets supported by a strong nerve, lappets which are homo-
logous with the spoon-shaped organs of Chiroteuthis, and the
similar projections in Doratopsis (PI. X, fig. 5).

The neck shows slight transverse and short longitudinal
ridges, corresponding in position, but very slightly in develop-
ment, to those of Thysauoteuthis (tc. and Ic. figs. 5 and 7, PI. X).

The small longitudinal ridges have been called olfactory
crests by Verrill (6) in Ommastrephes, where they are pro-
minent, and divide the neck region into separate areas or
facets; but only in the most ventral of these, i.e. the one
nearest the siphon, is found, in Thysauoteuthis, a small sepa-
rate crest, which is the homologue of the olfactory lappets of

84

F. ERNEST WEISS.

HistioteutLiis. In connection with this ridge, probably along
its base, as being there protected, will be found the sensitive
epithelium. In Ommastrephes the ridge in the ventral facet
is very small, and seems to be partly iutroverted into the skin.
Here I found a sensitive, or at least a highly modified, epithe-
lium, which I will describe later on.

The nuchal cartilage is more elevated and narrower than in
Chiroteuthis, and more of the Ommastrephes type.

The depressions at the base of the siphon are elongate,
broader aborally, and tending to separate into two fossae. The
cartilage on the mantle wall is elongate, pear-shaped, and less
definite and prominent aborally (PI. X, fig. 11).

The siphon is provided with a strong muscular valve, as
has been already stated by Professor Lankester (PI. X, fig. 10).

The gills are very powerfully developed. Two strong mus-
cular renal papillae are situated somewhat anteriorly to the base
of the gills and on either side of the rectum (PI. X, fig. 11).

Two large nidimental glands, on the same level as the
base of gills, but median to them, project freely into the
mantle cavity (PI. X, fig. 11).

Two oviducts, with terminal glands, open dorsad of the gills
between their base and the renal papillae. At the end of the
oviducal gland proper ( gl .) there is a further glandular struc-
ture, corresponding minutely in structure with the nidimental
gland, but about twice the size, and, indeed, almost as large as
the remainder of the oviduct. The external opening of the
oviduct extends about halfway along this gland. The oviducts
correspond so closely with those of Thysanoteuthis rhom-
bus, figured by Brock,1 that they might stand as a drawing of
those of Histioteuthis. The internal openings are slit-like and
situated near the lower end of the coelomic cavity, thus differ-
ing from the oviducts of Chiroteuthis, which are short, and open
far up into the ccelom. The beginning of the oviduct shows a
glandular passage.

On opening the renal chamber it is found to be compara-
tively long, containing the vena cava and renal veins, with
1 Brock, ‘ Zeitschrift f. wiss. Zoologie,” 1882.

ON SOME OIGOPSID CUTTLE FISHES,

85

their renal covering, and the hepatic ducts, with renal tissue
in bunches at intervals. Two very distinct membranous
funnels lead from the renal into the coelomic cavity. The
coelom is not distinctly divided into two parts. The heart lies
far back, and the ovary passes underneath the heart to its
anterior attachment on the stomach. Posteriorly the ovary
has another point of attachment, here to the coelomic wall.

The genital artery is given off from the posterior aorta, and
passes over the front of the heart. The coelom communicates
freely with the space containing the branchial hearts.

Most points in the anatomy of Histioteuthis lead, as I will
point out afterwards, to a close association of this form with
Thysanoteuthis.

Traciieloteuthts Behnii (Steenstrup).

The specimen examined (one of those captured at Messina,
and presented by Professor Kleinenberg to Professor Lankester)
agreed very closely with the description given by Pfeffer (10)
of Verilliola nymph a, and therefore in all probability
Hoyle is right in identifying the two genera. It certainly
agrees with Steenstrup’s1 (12) account of this species. The
order of the arms in descending order of lengths was 2, 3, 4, 1.

The second and third arm are very nearly equal, as may be
seen from the actual measurement appended. The suckers of
these arms, too, are much larger than those of the other arms.
They are provided with slight membranous fins, as is also the
fourth pair of arms (PI. X, fig. 1).

The tentacular arms are relatively long, with a distinct club
at the extremity. The suckers near the distal end are large
and in four rows, but proximally pass over into the stalk in
ight or ten rows of very minute suckers (PI. X, fig. 3).

The club is provided with a lateral fin-like expansion, as in
Doratopsis.

The eyes are large, but not very prominent.

On the back of the head are two pairs of large pigmented
1 Steenstrup, ‘ Vid. Meddel nat. Foren Kjobenhavn,’ 1.883 .

86

F. ERNEST WEISS.

patches, such as Pfeffer describes. On the ventral side of the
head, close beneath the eyes, are two small organs, probably
with sensory function. These organs agree most closely with
those of Sepiola, which lie exactly in the same position, and are
represented by a small elliptical ring, which probably protects
the sensitive epithelium at its centre. An olfactory crest, homo-
logous with that of Sepia, is absent in Sepiola and Trachelo-
teuthis. A fine nerve seems to pass over the ocular region to
this sense organ (PI. X, fig. 2, n.).

The neck of Tracheloteuthis is very long and thin, and the
mantle edge seems therefore loose and wide. The cartilaginous
fastening of the mantle at the base of the funnel is simple, and
like what we find in Architeuthis and in Loligo.

The interior of the funnel seems at first devoid of valve,
though, according to Steeustrup (12), it possesses one (i. e. in
Tracheloteuthis Riseii). But the funnel has what Ycrrill
has described in Desmoteutliis, and what Hoyle has called
V errilks organ. At the upper part of this organ, however, a
distinct though small valve is present, partly overlapped by
the median portion of Verrill’s organ (PI. X, fig. 4).

Whatever may be, therefore, the exact nature and function of
Verrill’s organ, whether mucoid or sensitive, though mechani-
cally it aids in closing the funnel, it does not preclude the
presence of a valve.

Besides the median portion of Verrill’s organ, which seems
composed of several parts, there are two lateral cushions on the
anterior wall of the siphon (fig. 4, Ip.).

These lateral pads, though not so strongly developed, are
seen in Doratopsis (PI. IX, fig. 8), and in Ilistioteuthis (PI. X,
fig. 10). Both lateral and median portions of YerrilPs organ
are absent in Cliiroteuthis.

On opening the mantle cavity we notice the characteristic
position of the viscera at the extremity of the body, and the
consequent development of the depressores infundibuli muscles,
in the middle of which runs the vena cava. The gills, too,
have been drawn out from back to front, and are thin and
loosely branched. The viscera in general aspect resemble those

ON SOME OIGOPSID CUTTLE FISHES.

87

of Doratopsis, The apertures I was unable to make out on
account of the smallness of the specimen. The same is the
case with regard to the sex of this specimen.

Its measurements were :

Lengtli of body .

2-7 cm.

„ mantle

21 „

„ fin . .

•6 „

Breadth of fin

•8 „

Arm 1 ...

•32 „

„ 2 ...

•75 „

„ 3 ...

•6 „

„ 4 ...

•4 „

Tentacles .

1-35 „

Tracheloteuthis, though in

many points agreeing with

topsis, is, I think, better placed now by Hoyle among the
Ommastrephidse.

Verania sicula (Krohn). PI. VIII, figs. 1 — 3.

In its main features this species resembles Enoploteuthis,
e. g. shape of fin and suckers modified into hooks, and is
classified together with Enoploteuthis. The specimens studied
by me were sent from Messina by Professor Kleinenberg to
Professor Lankester.

The arms, according to length, are 2, 3, 1, 4. Each arm has
at its extremity a small swelling preceded by eight to ten pairs
of modified suckers, and according to Pfeffer (10) these are
hectocotylized portions, just as we find them on the fourth
pair of arms in Enoploteuthis.

The tentacular arms are very short and thin, and the club
bears only a few suckers, three of which are relatively very
large. These suckers are narrow, but do not bear hooks
(fig. 3).

The tentacular arms were not observed by Riippell or Krohn,
and Verany thought that they were regularly lost at a certain
stage.

The specimen I examined was very much smaller than those
previously described, and probably still quite young. Brock

88

F. ERNEST WEISS.

(2) thinks that such a form regularly losing its long arms
points to the way in which the Octopoda gradually developed
from the Decapoda.

The eyes are not pedunculate, nor indeed prominent.

Behind the eyes is a small patch representing the olfactory
sense organ, and connected by a nerve with the main cerebral
mass (fig. 2, olf. org.).

Sections taken through this organ show a cushion of many-
layered epithelium cells, some oval and some spindle shaped,
and sunk away from the surface, and supplied with nerves from
the ganglion, which lies beneath the cushion. They resemble
greatly the sensitive cells figured and described by Sochaczewer1
(13) in the pedal gland of the Snail. The cilia, if present, are
very short ; probably there are only sensitive hairs, and not cilia
proper, which would be unnecessary, as the surface is exposed
to the free play of the water. Mucous cells seem absent.

I take this to represent the simplest form of cephalic
olfactory organ in Cephalopoda, as in sections of embryos of
Loligo and Ommastrephes I have found this patch with
similar modified cells situated in about the same position
behind the eyes. Indeed, on a surface view of a young
Ommastrephes it forms a very conspicuous elongated knob
laterally and posteriorly to the eye.

In the nearly related form Enoploteuthis this organ is more
prominent than in Yerania.

In Onychoteuthis it is represented by a ridge, and when we
get strong cervical ridges developed as in Thysanoteuthis, we
find it, as mentioned before, as a small lappet in the corner of
the partition formed by these ridges (PI. X, fig. 7).

In Ilistioteuthis, where these ridges become much reduced,
the lappet seems relatively larger but occupies the same posi-
tion ; and this form leads on to Chiroteuthis with its spoon-
shaped organs, and Doratopsis with its stalked and club-
shaped processes.

Unfortunately sections across the processes in Histiotcuthis
and Chiroteuthis do not reveal any modified epithelium cells
1 Sochaczewer, ‘ Zeitschrift fur wiss. Zoologie,’ 1881.

ON SOME OIGOPSID CUTTLE EISHES.

89

though a strong nerve supply exists in these organs. Possibly
they have changed their function and become tactile in nature,
in Chiroteuthis at least.

I hope, however, at some future date to be able to give
some further account of these organs.

Taking another set of forms we can pass from the olfactory
ridges on the embryo of Ommastrephes to its adult stage,
where there is still a specialised mass of cells similar to those
of Verania at the base of the ridge, as I was enabled to find in
sections of the ridge.

Then in Loligo we get the olfactory organ partly invagi-
nated but keeping the same relation to the ridges on the
neck, which become now specialised into an auricular, or better,
a protective crest (PI. X, fig. 6).

Verrill (6) speaking of the auditory pore of Omma-
strephes illecebrosa, must surely mean the structure which
is generally looked upon in Loligo and Sepia as an olfactory
pit, or, at least, as a sense organ, which is of the nature of
an olfactory or gustatory organ.

In Sepia sections of this pit show a similar structure to the
modified epithelium before mentioned, and the presence of
ciliated cells in large numbers indicate that a current must be
constantly kept up in the interior of the pit, bringing olfactory
particles to the sensory cells.

In Octopus and Eledone the pit is not protected by a crest,
but situated in an equally well-protected spot at the junction
of the mantle with the neck. In Octopus the pit is lined with
epithelium exactly like that of Sepia, so I think this olfactory
organ may be traced successively in the different groups of
Cephalopoda from the spoon-shaped organs of Chiroteuthis
to the invagination of Sepia and Octopus, both being the
extreme developments of the olfactory patches or cushions
seen in Verania. The series reminds one of the transition of
the olfactory organs of fishes from external processes to pits,
described by Professor Wiedersheim last year in his paper
before the British Association at Manchester.

The fastening of the mantle in Verania consists of a simple

90

P. ERNEST WEISS.

pyriform groove at the base of the funnel with a corresponding
ridge on the mantle (fig. 2).

The siphon is provided with a modification of Verrill’s
organ, but has also a valve relatively near its base and con-
nected with the median portion of Yerr ill’s organ. There are
also two large lateral cushions on the anterior wall of the
siphon, as noticed in Doratopsis and Tracheloteuthis. Perhaps
these structures are relatively large in these forms owing to
their being all still very young, and they may perhaps dis-
appear at a later stage, since they have not been described for
the larger specimens examined by Verany, Brock, and others.

The viscera of Verania showed no features worth special
notice; I was unable to make out much owing to the smallness
of the specimen.

Measurements.

Length of body

1-4

cm.

„ fin

'7

55

Breadth of fin

1-4

J»

Length of mantle .

'7

it

Arm 1 .

-7

if

„ 2 . . .

1-2

if

„ 3 . . .

1-

it

„ 4 . . .

-6

a

Tentacular arms .

-7

a

Concluding Remarks.

It will be seen from the above investigation of several of the
members of the Chiroteuthida?, that differing though they do on
some points and especially in general appearance, we may justly
unite them in a single family, though not on the grounds
formerly given for their separation from the other Oigopsida,
namely, the absence of siphonal valve, loss of accessory nidi-
mental glands and of one of their oviducts.

The concordance of Chiroteuthis with Doratopsis is very
complete indeed, as regards, for example, the fastenings of
mantle, the relative length of arms and tentacles, the projecting
olfactory processes, &c. Their main difference is in the relative
form of body and in some points of detail, as, for example, the

ON SOME OICtOPSID CUTTLE FISHES.

91

stellate organs and pigment spots, and the absence of Verrill’s
organ in Chiroteuthis which is present in Doratopsis.

Some of the main points of agreement between Histioteuthis
and Chiroteuthis, besides shape and proportion of body, are the
pigmented (phosphorescent ?) organs on the body and arms,
the suckers (modified) on the tentacular arms, and, perhaps
above all, the presence of renal papillae in both Chiroteuthis,
and Histioteuthis, which Vigelius,1 (14) in describing those
of Thy sail oteuthi s rhombus, believed to occur only in
that member of the group of Oigopsida.

Besides this they agree in the course of the vena cava, the
single renal chamber, and the extent of the coelom. On the
other hand, now that the great differences which separated the
Chiroteuthidae from the other Oigopsidae, notably the absence
of siphonal valve, single oviduct, &c., have been disproved, I
see no reason why we should not place the subfamily of Chiro-
teuthidae in the family of Ommastrephini side by side with the
subfamilies of Thysanoteuthidae, Ommastrephidae, and Masti-
goteuthidae, and thus abolish the family of Taonoteuthidae, the
name of which was not distinctive nor descriptive of any of its
genera. The Chiroteuthidae have some points in common
with all the subfamilies of the Ommastrephini, but especially
many and indeed important ones with the Thysanoteuthidae ;
and I should place the Chiroteuthidae next to this subfamily
on account of the following points of agreement.

The short arms of both Thysanoteuthis and Histioteuthis
are not so unequal as are those of Chiroteuthis. In Thysano-
teuthis2 they are protected by two large, fluted, membranous
folds, which might easily be developed into the connecting
umbrella of Histioteuthis.

The long arms in Histioteuthis and Thysanoteuthis agree
strikingly. The club has in both four rows of suckers, two of

1 Vigelius, ‘ Mittheilungen der zool. Station Neapel,’ 1831.

5 Two specimens of Thysanoteuthis rhombus are preserved in the
museum of University College, and have been placed at my disposition for
study. They were obtained by Professor Lankester from the Zoological
Station of Naples.

92

F. ERNEST WEISS.

which are much larger than the other two, and all along the
arm we have an almost identical row of alternating suckers
and pads or fixing cushions, which Steenstrup 1 (15) gives as
a main character of the whole family of Ommastrephini.

In speaking of the olfactory organ I spoke of the agreement
of the cervical ridges of Histioteuthis with those of Thysano-
teuthis, which are only reduced in prominence in Histioteuthis,
and are more strongly developed than in Tracheloteuthis, and
other forms which are actually included in the Ommastre-
phidae.

The mantle fastening in Histioteuthis is more simple than
that figured for Thysanoteuthis by Troschel,2 (16) from which
it only differs by the reduction of the tooth which projects
over the longitudinal depression at the base of the funnel.

Another point on which great stress has been laid by
Vigelius (14) is the presence of the renal papillae, usually
absent in Oigopsida but occurring in Thysanoteuthis, Histio-
teuthis, and Chiroteuthis.

Of the striking agreement of the oviducts in Histioteuthis
and Thysanoteuthis I have spoken before in treating of His-
tioteuthis. The same is the case with the large nidimental
glands which project and hang freely into the mantle cavity.
In both, too, the ovary reaches far forward in the coelom, and
is attached by two points and not along its entire length.

In fact Thysanoteuthis agrees more nearly with Histioteu-
this than with Ommastrephes with regard to its anatomy.
The stellate ganglia in Histioteuthis are invested by a very
strong commissure, which does not exist in Thysanoteuthis
according to Vigelius (14), but which brings Histioteuthis in
accord with the other Ommastrephini. In Chiroteuthis
I was unable to discover such a commissure.

Chiroteuthis has several points of agreement with Ommas-
trephes, viz. the complicated fastening of the mantle, the
position and structure of the oviducts and nidimental glands.

Then Chiroteuthis agrees in many points with Architeuthis,

1 Steenstrup, ‘ Oversigt Kong. Danske Vidk. Skels. Forh.,’ 1880.

2 Troschel, ‘ Arcliiv fur Naturgeschichte,’ 1857.

ON SOME OIGOPSID CUTTLE PISHES.

93

also a member of the Ommastrepbidm, and the same may be
said of Doratopsis, which Pfeffer has classified so closely with
Tracheloteuthis (his Yerrilliola).

Even Mastigoteuthis has affinity for one of the Chiroteu-
thidae, for Verrill (6) mentions an olfactory lappet which I
find is like the one I have described for Histioteuthis, though
his mention of an auditory pore is puzzling.

Taking all these points into consideration, I think we may
safely dispense with the family name of Taonoteuthidae, and
place the Chiroteuthidae as a subfamily under the Ommas-
trephini.

Index to Literature.

1. D’Orbigny et Ferussac. — “ Jlistoire naturelle generale et particuliere

des Cepkalopodes acetabuliferes, Vivants et Fossiles,” Paris, 1835-48.

2. J. Brock. — “ Versucli einer Phylogenie der dibranchiaten Cephalopoden,”

‘ Morph. Jahrb.,’ 1880.

3. W. E. Hoyle. — “ ltcport on the Cephalopoda,” ‘ Report of H. M. S.

Challenger,’ vol. xvi, 1880.

4. Verany. — ‘ Cephalopodes de la Mediterranee,’ Genes, 1851.

5. von Ihering.— ‘ Zeitschrift fiir wiss. Zool.,’ 1881.

G. A. E. Verrill. — “The Cephalopods of the N. E. Coast of America,”
‘Trans. Connect. Acad.,’ 1880, 1881, 1882.

7. W. T. V igelius. — “ Ueber das Excretionssystem der Cephalopoden,”

‘ Nied. Archiv fiir Zoologie,’ 1880.

8. C. Grobben. — “ Morphologische Studien liber den Ham uud Geschlects

apparat sowie die Leibeshoehle der Cephalopoden,” ‘Arb. d. zool.
Instituts zu Wien,’ 1884.

9. J. Brock. — “ Zur Anatomie und Systematik der Cephalopoden,” * Zeitsch.

fiir wiss. Zoologie,’ 1882.

10. G. Pfeffer. — “ Die Cephalopoden des Hamburger Naturhistorischen

Museums,” ‘ Abh. des Naturwiss. Vereins, Hamburg,’ 1884.

11. J. Brock. — = No. 9.

12. T. Steenstrup. — ‘VidMeddel nat. Foren Kopenhagen,’ 1881.

13. Sochaczewer. — ‘ Zeitschrift fiir wiss. Zool.,’ 1881.

14. W. T. V igelius. — “ Untersuchungen an Thysanoteuthis rhombus,”

‘Mitth. d. zool. Station Neapel,’ 1881.

15. S. Steenstrup. — “ De Ommatostrephagtige Blaecksprutters indbyrdes

Forkold,” ‘ Oversigt Kong Dansk. videnskabernes Skelskab Forhandl.,’
1880.

10. Troschel.— ‘ Archiv fiir Naturgeschichte,’ 1857.

94

P. ERNEST WEISS.

DESCRIPTION OF PLATES VIII, IX, & X,

Illustrating Mr. F. Ernest Weiss’s Paper, “On Some Oigopsid
Cuttle Fishes.”

PLATE VIII.

Figs. 1 — 3. — Verania sicula.

Fig. 1. Dorsal view of the entire animal, h. liectoeotylized end of the
arms. t. The short tentacular arms. s. g. Stellate ganglia.

Fig. 2. Ventral view after opening the mantle cavity, and the funnel
along the median line, showing olf. org., the olfactory organ, the valve
with the median portion of Verrill’s organ and its lateral pads (l. p.)
and central pads (c.y;.). /-.Kidneys, ov. Gonad, br. heart. Branchial
heart.

Fig. 3. a. The tentacular club, b aud c. Side and front view of one of
the large suckers.

Figs. 4 — 8. — Chiroteuthis Veranyi.

Fig. 4. Ventral view after opening the mantle cavity, olf. org. Spoon-
shaped organ (olfactory). n. gland. Nidimental gland. a. n. gl.
Accessory nidimental gland, red. Rectum, r. pap. Renal papilla.

Fig. 5. After opening the renal sac and the funnel, r. p. Renal papilla.
v. p. a. Viscero-pericardial aperture, b. v. n. Branch of visceral nerve.
a.n.gl. Accessory nidimental gland, nid.gl. Nidimental gland. Car-
tilaginous socket and tooth of mantle fastening, br. heart. Branchial
heart.

Fig. G. The pericardium ( p . c.) laid open. v. p. a. Viscero-pericardial
aperture, br. v. Branchial vein. br. h. Branchial heart (in its portion
of the pericardium), ov. Oviduct, ventr. Ventricle.

Fig. 7. The perigonadial portion of the coelom laid open, as also the an-
terior renal chamber ( ar .) containing its renal mass. br. v. Branchial
vessels (vein and artery), ext.ap. External aperture of oviduct, o. gl.
Oviducal gland, int. ap. Internal aperture of oviduct, ccc. Caecum to
intestine, lig. Ligament forming posterior attachment of ovary to the
wall of coelom, the position of the left oviduct indicated by dotted lines.
a. nid.gl. Accessory nidimental glaud. v.p.a. Viscero-pericardial
aperture, br. v. Entrance of branchial vein.

Fig. 8. Nidimental gland of Chiroteuthis.

ON SOME OIGOPSID CUTTLE FISHES.

95

PLATE IX.

Figs. I — 9. — Doratopsis vermicularis .

Figs. 1 — 3. Drawn by Miss Stone. Ventral (1), dorsal (2), and lateral
(3) views, lent. Tentacular arms. dell. org. Stellate organs, olf.
org. Olfactory organ, nuchal cart. Nuchal cartilage.

Fig. 4. View of oral surface, p. lip. Papillate lip. b. mernb. Buccal
membrane, t. Tentacular arm. iv. Fourth pair of arms with single
row of suckers.

Fig. 5. View of interior of mantle cavity. m.f. Mantle fastening.
r. p. Renal papilla. p. vein. Posterior renal vein. g. gl. Genital
gland.

Fig. 6. Portion of neck region (dorsal), showing the stellate organs,
si. org. v. n. visceral nerves, aorta, and a. c. alimentary canal.

Fig. 7. Tentacular club.

Fig. 8. Funnel opened along median line, showing valve, lateral pads of
Verrill’s organ, vena cava (v. c.), ink sac, and rectum.

Fig. 9. Ventral view of head region showing the olfactory organs, iv.
Fourth pair of arms. tent. Tentacular arms.

PLATE X.

Figs. 1 — 4. — Tracheioteuthis Behnii.

Fig. 1. Dorsal view. p. p. Pigmented patches, n, in. Second and third
arms, with larger suckers.

Fig. 2. Ventral view after opening the mantle, o. o. Olfactory organ
supplied by nerve («.). m.f. Mantle fastening. m. d. i. Musculi
depressores infundibuli. v. c. Vena cava. r. Kidneys, red. Rectum.
c. Ridge, cartilaginous ridge for mantle fastening.

Fig. 3. Tentacular club, with lateral protective membrane ( l . m.).

Fig. 4. Funnel, opened to show valve and median papilla (m. p.) and
lateral pads (/. p.) of Verrill’s organ, c. groove. Cartilaginous groove
for mantle fastening.

Figs. 5 — 7 show modifications of the olfactory orgau (o. o.) — Histioteuthis,
Loligo, and Thysanoteuthis. t. c. Anterior transverse crest, t. d . Posterior
transverse crest. 1. c. Longitudinal crest, e. Eye.

Figs. 8 — 12. — Histioteuthis Ruppelli.

Fig. 8. Renal chamber opened, showing v. c., vena cava, and h. d., hepa-
tic ducts with renal appendages. 1. m. Lateral renal masses on the
renal veins, ov. gl. Oviducal gland, red. Rectum, br. heart. Branchial
heart, r. p. Renal papillae, v. p. s. Viscero-pericardial aperture.

96

P. ERNEST WEISS.

Fig. 9. Coelom opened (pericardium and perigonadium). v.p. a. Viscero-
pericardial aperture. i. a. Internal aperture of renal papilla, st.
Stomach, g. ar. Genital artery. ovd. Oviduct. ini. ap. Internal
aperture of oviduct, rt. ovd. Right oviduct, red. Rectum, hr. art.
Branchial artery. Or. heart. Brauchial heart.

Fig. 10. Funnel opened, showiug valve and lateral pads of Verrill’s
organ, m.f Mantle fastening.

Fig. 11. Mantle cavity laid open. m. d. i. Musculi depressores infundi-
buli. ov. gl1. Accessory oviducal gland, gl. Oviducal gland proper.
nid. gl. Nidimeutal gland, c. llidge, cartilaginous ridge fitting into
the groove (m.f) forming the mantle fastening.

Fig. 12. Tentacular arm, with pad and suckers forming the fixing
apparatus.

Wr fam.xs.h. m

au nici. al

vpa.

aorta

v.p.a,.

br v. • -

FCq. /

post. aorta.

part of stomach.

ovary

stomach

vena.

cava

C. socket

C.tootfo

ovrrtur/.

oiferp

oviduct

nrtaits

mantle vein.

hr heart.

vena

'OM?

rariunrrf

b>- heart

post aorta

stomach

T Huth.Li(hr Zi’.n*

F £ Weiss del

, V E R A N 1 A

SICU LA.

Fi^ 4-8, CHIROTEUTHIS

Fig* L - 3

V E R A N Y

Mor.Jou/m.%6. XXIX, XSM. /X.

fitg. /. fiig. Z.

Stone & \Vcis» del

F Hath. Lith’ Zinf

DORATOPSIS VERMlCUL-ARis. (Rochebrune)

.

Afax&wm. V06XX/US. Ifox

Valve

oviduct

hr art

heart

hr heart

br heart

ovary

- KT!

mt. an.

val i-

nctum *

F.E.Weitt dll. H 1ST I OT E UT H I S LOLIGO.

Fig* 1-4.TR ACHE10TEUTH1S BEHN1I.

THYSAN)TEUTHIS

post CLorta,

Figs 8 12, HiSTlOTEUTHIS RUPPELLI.

Y Huth,LithT Edm'

1

THE ORGAN OF VERRILL IN LOLIGO.

97

The Organ of Verrill in Loligo.

By

Malcolm Laurie, B.Sc.

From the Zoological Laboratory of University College, London.

With Plate XL

In a paper on North American Cephalopoda1 Verrill describes
a valve-like organ at the base of the siphon in Desmoteuthis
and Taonia, in both of which genera the true valve is absent.
His drawing of this structure in Desmoteuthis tenera is re-
produced in fig. 1. It is there seen to be composed of a median
portion, ( m .) lying on the dorsal side of the siphon, and a pair
of lateral cushions (n. n' .). The median portion is raised into
three papillae, one (i.) median and a pair ( i '.) more posterior in
position.

This structure to which the name of Verrill’s organ has
been given by Hoyle2 has been observed in a few other species,
but is by no means common. It is figured by Mr. Weiss in
the present number of this Journal for several Oigopsidae (see
PI. VIII, fig. 2 ; PI. IX, fig. 8 ; PI. X, fig. 10). I was there-
fore much interested to discover, in sections of a young
Loligo about G mm. in length, made in Professor Lankester’s
laboratory at University College, London, a structure in the
siphon which I think there is no doubt is Verrill’s organ.
The general appearance of this structure is shown in fig. 2.
It consists of a median dorsal cushion, which is prolonged

1 ‘ Trans. Conn. Acad.,’ vol. v, part 2.

2 ‘ Challenger Report,’ vol. xvi.

VOL. XXIX, PART 1. NEW SER.

O

98

MALCOLM LAUEIE.

backwards into two large processes, and a pair of lateral
cushions on the ventral wall of the siphon. The dorsal
cushion is produced forwards into a papilla which is best seen
in section (fig. 6, i.). Figs. 3 — G show sections through the
siphon and Verrill’s organ. The valve (figs. 2 and 6, v.) is
well developed, and quite distinct from Verriirs organ.

An examination of the histological structure shows the
organ to be glandular. It is composed (fig. 7) of columnar
goblet cells almost entirely filled with a clear transparent
substance which stains very darkly with hsematoxylin. The
nuclei are at the bottom of the cells, and are surrounded by a
small quantity of granular protoplasm. On the surface of the
organ, more especially in the older specimens, there is a large
quantity of mucus-like substance which has apparently been
excreted from the cells. The organ of Verrill thus appears to
be a mucus gland. It may, as Verrill suggests, function as a
valve in those forms in which a true valve is wanting, but such
a function would be secondary.

It is well developed in Ommastrephes in specimens about
8 mm. long. I can find no trace of it in the adults of either
Loligo or Ommastrephes.

As regards its use to the organism or its homologies outside
the Cephalopoda, I can say nothing. I think that Verrill has
no foundation for his suggestion1 that the dorsal cushion is “ a
true homologue of the foot of Gasteropods.” Its presence, so
largely developed in the young, seems to indicate that it is an
archaic structure in the group of Cephalopoda, but there are
no grounds for identifying it with any particular structure
existing in Gastropods.

1 Loc. cit., footnote p. 432.

THE ORGAN OF VERRILL IN LOLIGO.

99

EXPLANATION OF PLATE XI,

Illustrating Mr. Malcolm Laurie’s Paper on the “ Organ of
Verrill in Loligo.”

List of Reference Letters.

s. Sipliou. m. Mediodorsal cushion of Verrill’s organ, n. n'. Lateral
cushions, i. Central papilla, i' . Lateral papillae, p. Mantle, v. c. s. Superior
vena cava. v. Yalve.

Fig. 1. — Siphon of Desmoteuthis tenera. Natural size. After Verrill.

Fig. 2. — Organ of Verrill in young Loligo. Dorsal cushion.

Fig. 2 a. — Ventral wall of Siphon, showing paired cushions.

Fig. 3. — Transverse section of young Loligo. x -f- .

Fig. 4. — Transverse section of young Loligo further down than Fig. 3.
X

Fig. 5. — Longitudinal sagittal section of young Loligo to one side of the
middle line. X -2^-.

Fig. G. — Longitudinal sagittal section of youug Loligo in middle line,
x

Fig. 7. — Section of part of the organ of Verrill in a youug Loligo. f X
n. Nuclei, mu. Mucus in the goblet cells, mu'. Mucus on surface of organ.

3.

vcs.

mw

M Laurie del

F. Huth, Lithr Edin?

ADDITIONAL LIST OF ERRATA IN MR. SEDGWICK’S MONO-
GRAPH OF THE GENUS PERIPATUS IN YOL. XXVIII OF
THIS JOURNAL.

On page 439, twelfth line from bottom, for “ PI. I ” read “ PI. XXXIV.”

On page 441, fourteenth line from top, for “ PI. I ” read “ PI. XXXIV.”

On page 448, fifth line from bottom, for “ 6 mm.” read “ -6 mm.”

On page 455, ninth and tenth lines from bottom, for “ legs ” read “ pairs

of legs.”

On page 464, eleventh line from bottom, for “ discernable ” read “ dis-
cernible.”

On page 407, ninth line from bottom, for “ twenty-nine and ” read “ twenty-
nine pairs and.” On the next line, for “ legs ” read “ pairs of legs.”

On page 468, sixth and tenth lines from top, for “legs” read “pairs
of legs.”

On page 468, twelfth line from top, for “ twenty-three ” read “ thirty-
three.”

On page 468, tenth, eighth, and sixth lines from bottom, and on the last line,
for “ legs ” read “ pairs of legs.”

On page 471, seventeenth, nineteenth, and twenty-first lines from top,
for “ legs ” read “ pairs of legs.”

On page 474, eighteenth line from top, for “higher” read “lighter.”

On page 476, ninth line from top, for “ fig. 27 ” read “ fig. 29.”

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 101

On the Structure of Three New Species of Earth-
worms, with Remarks on Certain Points in
the Morphology of the Oligochseta.

By

Frank E, Beddard, M.A.,

Prosector of the Zoological Society, Lecturer on Biology at Guy’s Hospital.

With Plates XII and XIII.

This paper contains (1) an anatomical description of three
new species of Earthworms, of which two are from New Zea-
land ; (2) a discussion of certain points in the structure of
the Oligochseta generally, which have been suggested by the
study of these forms, but which involve the partial description
of other Earthworms not systematically treated here.

The New Zealand Earthworms 1 owe to the kindness of
Mr. W. W. Smith, of Ashburton, New Zealand, who sent
me a large number of well-preserved examples from different
localities, as well as a few living specimens. I also desire to
express my thanks to Sir Walter Buller, K.C.M.G., who
asked Mr. Smith to procure me some specimens.

Typhseus Gammii comes from Darjeeling; a number of
examples were kindly collected for me by Mr. G. A. Gammie,
a member of the Chinchona Cultivation Staff. They were
transmitted to me with a number of Perichieta from the
neighbourhood of Calcutta, by Drs. Kiug and Bain, of the
Seebpore Botanical Gardens, whom I desire to take this
opportunity of thanking. My thanks are also tendered to
Dr. John Anderson, F.R.S., at whose request the specimens
were collected and forwarded to me.

VOL. XXIX, PART 2. NEW SElt.

h

102

PRANK E. EEDDARD.

Acanthodrilus annectens, n. sp.

This species combines to a certain degree the characters of
two other New Zealand Acanthodrili recently described by
me (1), viz. A. multiporus and A. no vse-zeal andise ; not,
however, to so marked an extent as might lead one to infer
the possibility of its being a hybrid.

It is a comparatively small worm, measuring about 3
inches in length. The colour of the living worm is “ pink, or
white and pink.”

External Characters.

The setse are paired, the individual setae being at some
little distance from each other.

The Clitellum occupies segments 13 — 20 (inclusive) ; the
glandular modification of the epidermis is not developed on the
ventral surface, as is usual in this genus.

The anterior end of the body is somewhat swollen, and the
segments here are somewhat difficult to map owing to the
division of the segments into numerous annuli. In these
particulars the present species agrees very closely with A.
multiporus.

The atrial pores are upon the seventeenth and nine-
teenth segments, and correspond in position to the outermost
of the ventral pair of setse. They are placed upon the summits
of prominent papillae; the two pores of each side are connected
by a groove. In one specimen the pore of the vasa deferentia
was visible upon the eighteenth segment. As a rule these
pores are invisible.

The oviducal pores are paired, and lie upon the four-
teenth segment ; each is placed in front of, and a little to the
inside of, the ventralmost seta of the ventral pair.

The spermathecal pores are in the furrows separating
segments 7 — 8 and 8 — 9; they correspond in position to the
atrial pores.

The neph ri diopores are visible in most of the segments of
the body ; they lie in front of the outermost seta of the outer pair.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 103

Internal Anatomy.

Reproductive Organs. — The most remarkable fact about
the reproductive organs of this species is illustrated in PI. XII,
fig. 13; that is, that the testes (t.) and ovaries (ov.) instead of
being situated on the anterior wall of their respective segments
are placed upon the posterior wall in close proximity to the
funnels (/. ov.). I should have been disposed to regard this
arrangement as abnormal had it not been for the fact that it
occurred in all of the two or three specimens studied by me.

The vesiculae seminales of this species are like those of
other Acanthodrilus in their racemose character, and in the
fact that they do not envelop the funnels of the vasa deferentia.
It may easily be seen in longitudinal sections of the worm that
the vesiculae, although so different in outward appearance
from those of Lumbricus, only differ really in being branched
instead of simple outgrowths (see Bergh 8, fig. 13, v. s *) of the
septa.

The atria, as is always the case with Acanthodrilus, are
two pairs situated in the seventeenth and nineteenth segments.
The vasa deferentia, as also appears to be the rule in this
genus, open quite independently of the atria upon the eigh-
teenth segment (PI. XIII, fig. 12). The two vasa deferentia
unite just before their external orifice (c?), which is situated
just on the boundary line between the seventeenth and eigh-
teenth segments ; the pores are also situated in a groove which
connects the two atrial pores of each side, and the presence of
which is highly characteristic of the genus Acanthodrilus
as also of Deinodrilus (see PI. XIII, fig. 3). The two
vasa deferentia run side by side and obliquely, through the
muscular layers of the integument to the external pores, crossing
on their way the duct of the atrium of the seventeenth segment
( p •). In longitudinal sections I traced the vasa deferentia
back to the thirteenth segment, running in the longitudinal
muscular layer and at some distance from the surface, nearly
midway between the two surfaces of the longitudinal muscular
layer ; after this they gradually approach the peritoneal face

106

FRANK E. BEDDARD.

External Characters.

The length of the largest specimen is about five inches.
The arrangement of the seta is shown in fig. 9 of PI. XIII.
A prostomium is present (PI. XIII, fig. 4), but does not
completely divide the circumoral segment. The clitellum
is well developed in one of the two specimens which I examined ;
it occupies segments 14, 15, and 16, having therefore precisely
the range which characterises so many species of Perichajta;
as in that genus the glandular modification of the epidermis of
the clitellar segments is continuous right round the body,
being equally well developed upon the ventral and upon the
dorsal surface (PI. XIII, fig. 3, cl.).

The only apertures visible upon the outside of the body (see
PI. XIII, fig. 3) are the dorsal pores, the apertures of the
male and female reproductive ducts, and of the spermathecse.
No nephridiopores could be made out. The dorsal pores com-
mence between the 11th and 12th segments. The oviducal pores
are upon the fourteenth segment ; they are paired and situated
a little in front and to the inside of the ventralmost seta.

The apertures of the atria are, as in Acanthodrilus, two
pairs : one pair are upon the seventeenth, the other upon the
nineteenth segment ; they correspond in position to the outer
seta of the ventral pair. The spermathecal pores are close to
the anterior border of segments 8 and 9 ; they correspond in
their relation to the setse with the male pores.

Internal Anatomy.

Integument. — PI. XIII, fig. 10, illustrates a section
through the body wall, in the hinder region of the worm.
The section has been drawn with the aid of a camera lucida,
and therefore indicates correctly the relative thickness of the
different layers, which together constitute the body wall. As
appears to be almost always the case, the longitudinal muscles
are much thicker than the circular ; in this worm they are
about six times as thick.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 107

The longitudinal muscles have their fibres arranged in that
remarkable bipinnate fashion which is found in many species
of Lumbricus and Allolobophor a, but is comparatively rare
elsewhere.

With regard to the vascular system, the only facts which
I am able to record are, (1) the condition of the dorsal vessel ;
(2) the number and connections of the “hearts.”

The dorsal vessel is a completely double tube, with the
exception of that portion which lies in the first four or five
segments. It resembles the dorsal vessel of Acanth odrilus
multiporus in the fact that the two tubes are perfectly separate
throughout, except where they become permanently fused at
the anterior extremity of the body. The somewhat contracted
condition of the worm frequently caused the two halves of the
dorsal vessel to become widely separate in the middle of each
segment, while at the mesenteries they come into close rela-
tion ; there is, however, no fusion of the two tubes at these
points, such as occurs in A. novae zealandise and Micro-
chaeta. I observed six pairs of lateral “hearts,” the last
pair being in segment 13 : the last four pairs are specially
large, and are connected with the supra-intestinal as well as
with the dorsal vessels. The anterior two pairs (there are pro-
bably one or two pairs in addition to those which I have
mentioned) are much more slender and only connected above
with the dorsal vessels.

Septa. — The septa separating segments 8 — 9, 9 — 10,
10 — 11, 11 — 12, 12 — 13, are thicker than the rest, but not to
so marked a degree as is often met with in Earthworms.

Alimentary Tube. — The pharynx has the usual charac-
ters. The gizzard lies in segments 6 and 7 ; the oesophagus
is thick walled and highly vascular, but there appeared to
be no distinct calciferous glands.

The intestine has a typhlosole (PI. XIII, fig. 8).

The nephridia are not obvious on dissection except in
segments 2, 3, and 4 ; in each of these segments is a tuft of
nephridial tubules of considerable size; in the posterior seg-
ments nephridia are present, and open on to the exterior by

108

EfiANK E. BEDDARD.

several pores in each segment. The nephridial system of this
worm is in fact like that of Acanthodrilus multiporus. I
have not as yet worked it out in detail; this I hope to do later.

Coelom. — It is the rule among the higher Oligochseta that
the coelom is a spacious cavity divided into a series of cham-
bers by the transverse septa ; only in the first few segments of
the body is this arrangement interfered with by the develop-
ment of strands of muscular fibres uniting the pharynx with
the parietes. In these segments the coelom forms an irregular
system of lacunae. Furthermore, the saccular outgrowths of
the septa in the genital segments which envelope the testicular
products, and sometimes also include the testes, the vasa
deferentia funnels, and part of the ventral blood-vessel and
nerve-cord, may be looked upon as specialised parts of the
coelom. Lastly, in Eudrilus there is a “perigonadial” space
surrounding the ovary.

In Deinodrilus the dorsal blood-vessel is surrounded by a
special coelomic space in a way that is, at present, unique
among Earthworms. This space does not appear to exist in
the first fifteen segments ; after this point the two dorsal blood-
vessels are not as plainly visible on a dissection of the worm as
they are anteriorly ; the red colour of the blood is masked by
the whitish colour of the tissues which form the walls of the
perihsemal space. The fact that the blood-vessels are so clearly
seen on dissection in the anterior segments, leads me to infer
that here there is no perihsemal coelomic space ; but I am
unable to support this view by a description of the micro-
scopical appearance of the dorsal vessels in this region of the
body, which I have not investigated by sections.

PI. XIII, fig. 7, is a longitudinal section through one of the
two dorsal vessels ; and PI. XIII, fig. 6, is a transverse section
of the dorsal vessels more highly magnified.

It will be seen from these figures that the blood-vessels are
surrounded by a widish tube which is further divided into two,
one for each of the paired vessels.

The walls of this perihaemal space consist of a thin layer of
fibres which are covered on both sides by peritoneal epithe-

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 109

lium. The outermost epithelium consists of delicate flattened
cells (fig. 6, p.) ; the perihsemal space is lined by rounded cells
which are aggregated here and there into clumps (fig. 6, p.).
I was at first inclined to regard these cells as free corpuscles
which had become adherent to the walls of the perihsemal
space ; the fact, however, that these cells were invariably,
so far as my experience goes, attached to the periphery
of the perihsemal space, not to the periphery of the con-
tained blood-vessel, seemed to show that they are the peri-
toneal lining of this section of the coelom. It seems certain
that the cells are in a state of energetic proliferation, and it is
possible that in this perihsemal space — and from its lining
cells — is carried on the formation of the coelomic corpuscles.
The blood-vessels themselves are covered by a single layer of
large cells ( p ".) filled with yellowish-brown granules. Some of
these cells appeared to be multinucleate, and there is some
variation in size.

The enclosure of the dorsal blood-vessel in a special coelomic
sac suggests of course the pericardium of higher types, and in
any case it may be compared with the condition of the coelom
in the Hirudinea, where the principal blood-vessels as well as
other organs are often included in separate coelomic spaces.
Among the Chaetopoda also a commencement of a secondary
subdivision of the coelom is to be seen. In the Capitellidse a
series of longitudinal chambers enclose the nephridia and other
organs; but I am not aware that hitherto anything of the kind
has been described in the Oligochseta.

Reproductive Organs. — The vesiculse seminales oc-
cupy segments 11 and 12; they are racemose organs like those
of Acanthodrilus. The testes I have not seen.

The vasa deferentia open by funnels in segments 10 and
11; the funnels of segment 11 are quite independent of the
vesicula. I could not trace the course of the vasa deferentia ;
but in all probability they open, as in Acanthodrilus, upon
the eighteenth segment.

The atria are in segments 17 and 19; the external aper-
tures of these organs have been already mentioned. The atria

110

FRANK E. BEDDARD.

themselves are so exactly like those of Acanthodril us that
no further description is necessary.

The ovaries are situated on the anterior wall of segment
13; they are digitate bodies like the ovaries of Acantho-
dril us.

The oviducts open by funnels which are placed near to
each other aud on either side of the nerve-cord, on the pos-
terior wall of segment 13. The external pores, as already
stated, are upon the 14th segment.

The spermathecas have a very characteristic form, which
is illustrated in PI. XIII, fig. 5. The spermatheca is a somewhat
oval pouch, which suddenly narrows into a slender duct, opening
close to the anterior margin of the segment. At the junction
of the pouch with the duct are three diverticula, two on one
side and one on the other; the diverticula are very much
smaller than the pouch, and of a regular oval form.

The following table indicates the principal points in which
Deinodrilus agrees with Acanthodrilus orPerichseta:

Acanthodrilus.

Deinodrilus.

Perichaeta.

Clitellum .

Segments 12 — 19, or
thereabouts ; unde-
veloped between the
atrial pores and the
corresponding area
on the other seg-
ments.

Segments 14 — 16
(inclusive) ; con-
tinuous all round
the body.

Usually segments 14
— 16 (inclusive) ;
continuous all
round the body.

Setae . .

8 per segment.

12 per segment.

20 — 100 per segment.

Atria . .

Two pairs of convo-
luted tubes opening
on to segments 17
and 19.

Two pairs of con-
voluted tubes
opening on to
segments 17 and
19.

Usually represented
by a single pair of
branched glands
opening on to eigh-
teenth segment.1

1 In Bourne’s P. Stuarti (9) the atria appear to be like those of Acau-
thodrilus.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. Ill

Typhseus Gammii, n. sp.

The largest specimen measured about 10 inches in length
(it is considerably contracted), and between a quarter and half
an inch in thickness at the head end.

External Characters.

As in T. orientalis there is no prostomium, the mouth is
therefore precisely terminal in position.

The setae are disposed in pairs; the dorsal and ventral pair
of one side are nearer together than the two ventral pairs ; the
interval which separates the latter is about one fourth to one
fifth of the space which lies between the dorsal pairs of setae.

The segmentation of the body is a little difficult to make
out, owing to the fact that there are numerous furrows in
addition to those which mark the limits of segments. The
accompanying drawing (PI. XII, fig. 7) illustrates the anterior
segments of the body viewed from the ventral aspect.

The peristomial segment is occupied by numerous short,
longitudinal creases, often of a zigzag form ; these cease to
exist some little way in front of the posterior end of the seg-
ment ; there is, however, a fairly well-marked furrow, dividing
this segment into two unequal halves. I am inclined to think
that the two halves really correspond to two segments ; the
only objection to this is that there were no setae discoverable
upon the supposed second segment, which — at any rate in all
other Earthworms — is the first seta-bearing segment. The
assumption, however, that this is really a segment brings other
organs of the body into positions more in accord with what is
found in other Lumbricidse. It will, therefore, be assumed
that the area occupied by the longitudinal creases equals two
segments.

The next two segments increase gradually in length, the
last being marked with a faint transverse furrow. The number
of furrows upon each of the following segments and their
arrangement can be understood by the figure (PI. XII, fig. 7).
The segments of the clitellum possess no secondary furrows.

112

FRANK E. BEDDARD.

The clitellum occupies four segments — 14 to 17 inclusive,
and a portion of segment 13. In my earlier paper on the
genus (2) an exactly similar condition of the clitellum is noted.
Dorsal pores are present on all segments after, and including,
the tenth. The pores of the spermathecse are very con-
spicuous between segments 7 and 8 ; they correspond to the
interval between the dorsal and ventral pairs of setae (fig. 7,

c-P •)•

The male generative orifices are upon segment 17, and
correspond to the ventral pair of setae.

Between segments 19 — 20, 20 — 21, and occupying the
whole of the space corresponding to the interval between the
two ventral pairs of setae, is a single large papilla (see fig. 3).
The arrangement of the genital papillae in this species is there-
fore apparently different from that which characterises T.
orientalis. The number and position of the papilla agrees
with Pontodrilus (Perrier (16), p. 177, pi. xiii, fig. 1, b.')}
with which genus, however, the present has but few other
points in common.

Internal Anatomy.

Body Cavity. — Under this head I refer to the condition of
the intersegmental septa in the anterior region of the body.

As in other Earthworms this species is seen to have a
number of these septa thickened and hypertrophied. The first
of these septa lies between the fourth and fifth segments; the
next in the succeeding segment. Tlie two following seg-
ments, which are occupied by the gizzard, are not divided by
a mesentery at all unless the muscular bands which bind
the anterior region of the gizzard to the body wall can be
regarded as the remains of the septum dividing segment 6
from 7. Farther back are three thickened septa which lie
between segments 8, 9, and 10. These latter are remarkable
from the fact that they do not divide the body cavity into
segments precisely equivalent to those indicated by the external
characters.

The first of these mesenteries corresponds to the first

STEUCTUEE OP THEEE NEW SPECIES OF EAETHWOEMS. 113

furrow upon segment 9 ; the second is situated a little an-
terior to the boundary line between this and segment 10 ;
the third is placed a little behind the first furrow of segment
10. It seems to me probable that these septa are those which
should separate segments 8, 9, 10 ; but so little do they corre-
spond to the external divisions of the segments in question,
that the space enclosed by the two last septa, which should
correspond to segment 10, actually has no setae. The setae of
this segment occur behind the septum, and therefore, so
far as the septa are concerued, in segment 11. The presence
of a pair of transverse vascular trunks between each of these
mesenteries is, however, a conclusive proof that they enclose two
segments (see description of vesicula seminale, p. 114)".

§ Vascular System. — As in the majority of Lumbricidae,
there is a dorsal vessel, a supra-intestinal, a ventral vessel, and
two lateral trunks. The dorsal and ventral vessels communicate
in segments 8, 9, and 10, by a pair of transverse vessels, a
pair to each segment. In the two following segments are two
pairs of stouter transverse vessels, which also communicate
with the supra-intestinal vessel (PI. XII, fig. 6).

The lateral trunks are very conspicuous in the gizzard seg-
ments. At each end of the gizzard they give off a system of
branches, which supply it with blood ; behind the gizzard the
two lateral trunks run beneath the intestine, and each ap-
proaches very closely its fellow. I am unable to state how
the lateral trunks originate.

§ Nephridia. — The nephridia of this species consist of in-
numerable delicate tubules, which are chiefly developed in the
anterior segments of the body ; they are at any rate more con-
spicuous here than elsewhere. As in T. orientalis, there is a
special mass of these tubules in the first and second segments
of the body. The characters of the nephridia in this genus
resemble those of Perichaeta, Acanthodrilusmultiporus,
Trigaster, &c., so far as the naked-eye appearances are con-
cerued. I have ascertained by cutting sections of a portion of
the integument in the region of, and including the orifice of a
spermatheca, that in this part of the body, at any rate, there is

114

FRANK E. BEDDARD.

more than a single pair of nephridial orifices to each segment.
It is probably also the case in other parts of the body, but I
am not able to give any accurate description of the arrange-
ment of the external pores.

I am inclined to think that in all Earthworms when the
nephridia have the characters recorded in this species, that is
to say, where they consist of abundant scattered tufts of
minute tubules, it will be discovered that the external apertures
agree with those of Acanthodrilus multiporus, Dicho-
gaster, and the present species (Beddard 5).

§ Alimentary Tract. — The gizzard is situated in the
sixth and seventh segments. As is generally the case when this
organ occupies two segments, the mesentery that should
separate these segments is absent, or at most represented by
rudiments. In the present species, as has been already said
(p. 112), there are two muscular bands of a strap-like form by
which the gizzard is attached to the body wall.

This species has a single pair of calciferous glands, which
are situated in the twelfth segment (fig. 8, ca.).

The alimentary canal presents only one other feature of
interest, and that is the presence of intestinal glands already
recorded in T. orientalis.

The glands are, however, not confined to this genus, since
they exist in much greater numbers in Megascolex, and
have also been described by Horst in Acanthodrilus, and by
myself in Eudrilus.

In Typhseus the glands agree in their minute structure
(PI. XII, fig. 2) with those of Megascolex, but differ
anatomically in the fact that the two glands of each pair
become fused together on the middle dorsal line of the intes-
tine, and also in the fact that the glands of consecutive
segments are connected. The minute structure bears a very
close resemblance to that of the calciferous glands.

§ Reproductive System. — This worm differs from the
greater number of Lumbricidae in the possession of only a
single pair of testes and a single pair of vesiculae seminales
corresponding to them. The vesiculae seminales (which are of

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 115

course the “testes” of my former paper on Typhaeus) are
long and tongue shaped, and extend back on either side of the
oesophagus as far as the male pore, i. e. to the seventeenth
segment ; they commence in the tenth segment, and therefore
occupy seven segments. The surface of the vesiculae is not
plain and smooth, but projects into numerous irregular
rounded clusters. At the anterior extremity the vesiculae
become attached to the last thick septum, and just below their
attachment is a small cavity, which contains the testes and the
funnels of the vasa deferentia (PI. XII, fig. 1 ,a.). This is to
be regarded, I imagine, as a median unpaired portion of the
vesiculae which so often occurs in Earthworms.

This compartment contained a mass of spermatozoa; it is not
divided up by trabeculae, as are the paired portions of the vesi-
culae, except for two fibrous bands which pass up to the mesen-
tery. The innermost pair of setae of the tenth segment (see
above, p. 112) are enclosed within this compartment.

The testes are contained within this compartment; they are
a pair of round bodies (fig. 1, t.), which have very much the
appearance of a woollen button.

The single pair of vasa deferentia funnels (/.) are also
contained within this compartment ; each is situated exactly
opposite to its own testis.

The vas deferens of either side passes down to the seven-
teenth segment, where it opens on to the exterior near to the
atrium and a bundle of penial setae, as in T. orientalis (PI.
XII, fig. 1).

Perhaps the most remarkable fact in the anatomy of this
worm is the peculiar relation that exists between the atrium
and the vas deferens. These two structures in other Earth-
worms open together by a common duct. In Typhaeus, how-
ever (fig. 1), the vas deferens, which becomes a little wider
at its termination, enters the body wall independently of
the atrium and behind it. A series of transverse sections
through this part of the body show that the vas deferens does
ultimately join the atrium, though only just beneath the epi-
dermis. The vas deferens is ciliated up to the point where it

116

FRANK E. BEDDARD.

perforates the tissues of the body wall. After this the epi-
thelial cells which line the vas deferens lose their cilia. The
sac containing the penial setae (PI. XIII, fig. 1) is a diver-
ticulum of the atrium (PI. XII, fig. 5, g.), just before the
junction of the latter with the vas deferens. These facts are
not only of interest as being unique among Earthworms, and
as forming a distinguishing feature of the genus Typhaeus,
but also from the point of view of a comparison with another
Oligochaet — Ocnerodrilus (see p. 125).

The ovary is situated in the thirteenth segment (fig. 9, ov.).

The oviduct is also similar in structure and position to that
of other Lumbricidae ; it opens on to the exterior in front of
the veutralmost seta of the ventral pair (PI. XII, fig. 9, od.).

The spermathecae are situated in the eighth segment, and
open, as already said, on the boundary line between this segment
and the one in front. Each spermatheca consists of a large
thin-walled pouch, and a small diverticulum on each side,
which is composed of a number of separate diverticula united
within a common muscular sheath.

The above account of the anatomy of Typhaeus seems to
indicate a general resemblance in structure to T. oriental is,
coupled with certain differences which appear to me to be on
the whole sufficient to warrant the specific separation of the
two forms.

The genital papillae are more numerous in T. orientalis
than in the present species ; it is true that this character has
to be used with caution in the discrimination of species, but
in the species under discussion which is represented by fully
mature individuals, the differences are so great that I cannot but
regard them as of specific value. The genital seta (cf. PI.
XIII, figs. 1, 2) are distinctively different in the two species.
The vas deferens in T. Gam mi i enters the body wall inde-
pendently of the atrium ; in T. orientalis, as in other Earth-
worms, the vas deferens joins the muscular portion of the
atrium.1

1 I mention this supposed difference with some hesitation, not having the
specimens of T. orientalis at hand to refer to.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 117

Typhous.

Generic Definition. — Setae paired aud confined to the
ventral half of the body ; dorsal pores present ; clitellum
developed upon segments 13 — 17. Male genital pores (in-
traclitellian) upon seventeenth segment corresponding with
ventral pair of setae. A single pair of spermathecae, each
furnished with two trifid diverticula, opening between segments
7 and 8 on a line with interspace between dorsal and ventral
pairs of setae. A single pair of testes in segment 10 ; a single
pair of vesiculae seminales enclosing testes and funnels of vasa
deferentia, and reaching back for three or four segments. A
single pair of vasa deferentia, each opening in common with,
or close to, a coiled tubular atrium like that of Acantho-
drilus ; penial setae present. Ovaries and oviducts occupying
the usual position in the thirteenth and fourteenth segments ;
gizzard single, intestine furnished with six or seven pairs of
glands on the dorsal surface. Nephridia forming incon-
spicuous tufts, nephridiopores of each segment numerous.

T. Gammii. — Penial seta with wavy ridges round distal
portion. Genital papillae two, on boundary line between seg-
ments 19 — 20, 20 — 21.

T. orientalis. — Penial setae with distal extremity flattened
and furnished with chevron-shaped ridges. Genital papillae
six or seven pairs between the several segments, immediately
following and preceding the seventeenth.

On the Structure and Homologies of the so called
Prostate Glands in the Oligoch^ta.

In certain Earthworks the vasa deferentia are unprovided
at their external orifice with any glands; this is the case, for
example, with Lumbricus, Urochaeta, and Microchaeta.
In other genera glands are present, which either pour their
secretion into the terminal region of the vas deferens, or else
open on to the exterior independently, but in the immediate
neighbourhood of the male generative pores; Perichaeta and
Pontodrilus are instances of the former condition, while in

VOL. XXIX, FART 2. NEW SLR.

I

118

FKANK E. BEDDAKD.

Acanthodril us there are large tubular glands opening close
to, but quite independently of, the pores of the male reproduc-
tive ducts.

Of these glands there appear to be two different forms. In
Perichseta, with the exception of P. Stuarti referred to
above (p. 110, footnote), Perionyx, Megascolex, and in
many of the Australian Lumbricids lately described by Mr.
Fletcher (11), these glands, which have received from their
position the name of “ prostate ” glands, are irregularly-shaped,
lobate bodies ; they communicate with the exterior by means of
a thick-walled, muscular duct, which receives at its upper ex-
tremity the vasa deferentia. In Acanthodrilus,1 Ponto-
drilus, and in some other genera, the prostate glands are
somewhat different in form ; they consist of a compact tubular
gland, which is frequently coiled, but which, like the prostate
gland of Perichseta, opens into a thick-walled muscular tube,
which in its turn opens on to the exterior. With the upper ex-
tremity of the latter, in Pontodrilus, is connected the vas
deferens; in some other genera, on the contrary (Acantho-
drilus) the gland preserves the same general appearance, and
the same histological structure, but is unconnected with the
vas deferens.

In Eudrilus the apparent homologues of these glands are
very different in their general appearance from those of any
other Earthworm, so much so, in fact, that Perrier, their
original describer, was inclined to doubt their homology with
the prostate glands of other Earthworms. The glands in
question are much larger than those of Acanthodrilus, and
are straight instead of being coiled. Furthermore, they have
a nacreous appearance, which is due to the presence of abun-

1 Perrier’s figure of the prostate gland iu Acanthodrilus ungulatus
(‘ Nouvelles Arch. d. Mus.,’ 1872, pi. ii, fig. 18, pr.) would seem to indicate
that iu this species alone the prostates have the racemose characters of those
of Perichseta. If, however, Horst (‘Notes from the Leyden Museum,’
vol. ix, p. 252) be right in assuming that my A. Layardi (‘ Proc. Zool. Soc.,’
1886) is really the same species, I can state most positively that they are like
those of other Acanthodrilus.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 119

dant muscular fibres. I have, however, myself been able to
show, in a paper recently communicated to the Zoological
Society (6), that these differences only mask a fundamental
similarity, and that the minute structure of the glands in
Eudrilus closely corresponds to that of the prostate glands in
Acanthodrilus. It can be hardly doubted that the “sausage-
shaped glands” of Eudrilus are the real homologues of the
prostate glands in Acanthodrilus and Pontodrilus.

In Criodrilus (Rosa, Benham) and Allurus (Beddard)
the termination of the vas deferens is furnished with a glan-
dular structure, which is not only different in structure from
the glands that have been already referred to, but is also unlike
in general aspect.

Finally, in Moniligaster (Beddard, 4) the male efferent
duct opens into a minute pouch, larger in M. Deshayesi
(Perrier 15) and M. Houteni (Horst 14) than in M. Barwelli,
which bears a certain resemblance to the prostate of Acan-
thodrilus, but which, as will be seen hereafter, differs in
certain important structural features.

The questions which I shall attempt to answer are : (1) Are
these various glandular bodies appended to the vasa deferentia
homologous with each other ? (2) What relation do they

bear to analogous structures in the aquatic Oligochseta?

Moniligaster exhibits a condition of the efferent ducts,
which is remarkably different from that of all other Earth-
worms. In a species (M. Barwelli) recently described by
myself (3) the vasa deferentia, as in many Limicolse, only
occupy two segments; there is only a single vas deferens
on either side, the internal funnel of which is situated in one
segment, and the external aperture on the following segment.
The vas deferens opens on to the exterior in common with a
glandular structure, which I have called a “prostate” in my
account of the anatomy of this worm, and compared with the
prostates of other Earthworms.

The structure in question is seen by an examination of
transverse sections to contain a wide cavity which opens on to
the exterior ; the cavity is lined with a layer of large glandular-

120

FRANK E. BEDDARD.

looking cells ; outside these again are abundant muscles fol-
lowed by a layer of glandular tissue. This latter has a very
remarkable structure, which is illustrated in PL XII, fig. 11.
This figure is a general view of a transverse section of the
whole organ, showing the vas deferens ( v . d .) just at its
point of entrance; the external covering is composed of large
granular cells, which are separated into groups by partitions.
Each cell is prolonged into a fine process, which extends at
least as far as the muscular wall ; indeed, it is difficult to
believe that the cells do not in some way or other reach the
lumen of the atrium, and there discharge their glandular secre-
tion.

There is evidently a very close resemblance between these
groups of cells and the “prostate” of Rhynchelmis,1 the
prostate of that worm consists of cells with fine long prolonga-
tions arranged in groups.

The structure of the organ, in fact, is exactly comparable to
that of the atrium in many Limicolse where the lining epi-
thelium is glandular and of considerable thickness as compared
with the surrounding muscular layer.

The atrium of Moniligaster differs from that of Rhyn-
chelmis in the presence of a muscular layer. Vejdovskv
does not record the presence of a muscular layer in that worm,
nor do his figures show any indication of it. It furthermore
differs in the absence of cilia from the cells of the lining epi-
thelium. In Stylaria lacustris2 amuscular layer is present,
and the lining epithelium does not appear to be ciliated. The
outer covering of cells is not segregated into groups as in
Moniligaster and Rhynchelmis; it consists of a single
layer of large glandular cells.

With its single vas deferens occupying only two
segments, and opening into an atrium of the charac-
ter just described, Moniligaster is more like certain
Limicolse than any other Lumbricid.

, In my former paper on Moniligaster I was unable to

1 Vejdovsky (18), p. 332, pi. xxiv, figs. 1, 3.

2 Vejdovsky (19), pi. iv, fig. 10.

STRUCT ORE OF THREE NEW SPECIES OF EARTHWORMS. 121

figure or describe the funnel of the vas deferens ; I could only
ascertain that it became continuous with the seminal reservoir
(erroneously termed “ testis5’). I have since discovered the
funnel by means of transverse and longitudinal sections.

The seminal reservoir appears to perforate the mesentery
which divides segment 8 from 9, and to lie in both of these
segments (see diagram, PI. XII, fig. 12). I am inclined to
believe that this appearance is produced by a bulging of the
mesentery, which is thin and delicate, and that the seminal
reservoir really lies in segment 8 attached to the posterior wall
of that segment. Its cavity is not divided up by anastomosing
trabeculae as is the case with Lumbricus, &c. The funnel
of the vas deferens opens into the interior of the seminal
reservoir, and it is important to observe that the funnel
is very simple in form as in many Limicolae, and is
not folded and plaited as is usually the case in Earth-
wo rms.

There is some discrepancy between my account of the
anatomy of Moniligaster Barwelli and M. Perrier’s
description of M. Deshayesi; this discrepancy is indeed too
great to be explained away on the grounds that the species
investigated were different.

In a recent paper Dr. Horst has described a third species,
M. Houteni; a study of this interesting paper gives me more
confidence in restating my own results, which I have every
reason now to believe are substantially correct.

Moniligaster, therefore, in respect of its efferent ducts, is
nearer to such Limicolse as Stylaria than any other Earth-
worm.

Having shown that the male reproductive ducts and the
accessory organs of the aquatic Oligochgeta are repeated down
to the most minute detail in Moniligaster, it remains to be
seen how far they are represented in other Earthworms.

Yejdovsky has pointed out in his great work on the
anatomy of the Oligocliseta (19) that the prostate gland of
Pontodrilus is probably the homologue of that of Eudrilus.
On the other hand, he regards the prostates of Perichseta,

122

PRANK E. BEDDARD.

Acanthodrilus, Digaster, &c., as equivalent to the pros-
tates (“ Cementdriiseu ”) of the Tubificidae, and therefore by
implication different from the analogous glands in Eudrilus
and Pontodrilus, which may possibly represent the atrium of
the Tubificidae. These views are naturally put forward with
some little hesitation.

I am disposed partly to agree and partly to disagree with
Vejdovsky's conclusions.

I entirely agree with his opinion that the so-called prostates
in Eudrilus and Pontodrilus are the homologues of the
atrium in the Tubificidae; I shall, however, bring forward
reasons for believing that the prostates in Acanthod rilus,
Perichaeta, &c., are the homologues of those of Eudrilus,
and therefore also of the atrium in the Tubificidae and other
families of the “ Limicolae.”

In Eudrilus I have been able to show (6) that the vasa
deferentia open into the interior of the large glandular body of
the seventeenth segment. The relation therefore of the vasa
deferentia to this body is precisely that of the vasa deferentia
to the atrium in the aquatic forms. It is true that the vasa
deferentia are not connected with the extremity of the sup-
posed atrium as in Monili gaster, Stylaria, &c. ; but in the
Lumbriculidae the vasa deferentia also communicate with the
atrium about half way down.

The atrium consists of two regions — of a glandular portion
and of a muscular tube prolonged into a penis. This differen-
tiation of the atrium has its counterpart in the Tubificidae,
and, moreover, the invaginated penis sheath of the latter is the
equivalent of the “bursa copulatrix” of Eudrilus. The atrium
in both consists of an epithelial lining and a muscular layer.
The epithelial lining is more complicated in Eudrilus than in
Tubifex ; in Tubifex and apparently in the Limicolous forms
generally the lining epithelium of the atrium is a single layer of
ciliated cells : this condition, minus the cilia, is retained in
Moniligaster. In Eudrilus the lining epithelium of the
atrium has the complicated structure which I have already
described. There is undoubtedly a close agreement in struc-

STRUCTURE OE THREE NEW SPECIES OP EARTHWORMS. 123

ture between the glandular cells which compose the greater
part of the atrial epithelium in Eudrilus and the cells which
cover the atrium in the Lumbriculidae and in Mouili-
gaster (PI. XII, fig. 11). Outside this muscular layer, which
covers the glandular lining, are faint traces of a peritoneal
investment, and it is this which is the homologue of the
glandular sheath in Rhynchelmis, Moni ligaster, &c.
The explanation of the difference in the structure of the atrial
epithelium is, as it appears to me, quite another one.

It has been conclusively proved by Yejdovsky that the atrium
in the Tubificidse is formed by an ectodermic involution just
as are the spermathecae and the “vesicle” of the nephridia,
and as a consequence it retains the structure of the integu-
mental layers. The ciliation of the lining epithelium is par-
ticularly interesting in this connection, because it often hap-
pens that ectodermic involutions, owing to the protection which
they afford, retain the ciliated condition which is lost on the
general body surface. Moreover, the cells in the distal glan-
dular part of the atrium in the adult have more completely
retained the characters of the epidermis than in the proximal
region, where it has undergone secondary modifications in
connection with the formation of the penis.

In Eudrilus the male reproductive pores are intraclitelline.
It is a fair assumption to suppose that the atrium is invaginated
from the ectoderm, and it will therefore retain to a certain
extent the structure of the body wall as it does in Tubifex.
The ectodermic cells in the young embryo, at the point where
the atrium is invaginated, have the potential capacity of de-
veloping into the complicated clitellar epidermis ; it is there-
fore not surprising to find that the invaginated cells also retain
this capacity, and ultimately form an epithelium nearly identical
in structure with the clitellum. The objection that those cells
which are nearest to the point of invagination are most unlike
the clitellar epidermis is to be met by reference to Tubifex.
The absence of cilia may be reasonably accounted for on the
supposition that Eudrilus is farther removed from the
ancestral ciliated condition than Tubifex.

124

FRANK E. BEDDARD.

In Acanthodrilus, Pontodrilus, and Typhaeus the
muscular layer of the atrium has been lost, and only a delicate
peritoneal layer1 remains ; otherwise the structure of the atrium
is the same ; it is therefore probable that in all these forms
(even Pontodrilus, which is now post-clitelline) the atria have
been invaginated from the clitellar area. On the other hand,
in Moniligaster the simple epithelium of the atrium may
perhaps indicate that it has not been formed as an ingrowth
from the clitellar area. This supposition is supported by the
forward position of the atria in this worm.

Professor Bourne has, however, lately (8, p. 662) described a
species of Moniligaster where a clitellum is present in the
neighbourhood of the male pores.

I therefore make the above suggestion, which indicates a
possible confirmation of Perrier’s classification of Earthworms,
with considerable hesitation.

In Pontodrilus the atria have acquired the character
of an appendage of the vas deferens, and the penis is
absent.

In Typhaeus the independence of the atria and the vasa
deferentia is more marked. The two organs appear to open
on to the exterior of the body independently, but in reality
they unite just below the epidermis.

In Acanthodrilus the vasa deferentia open on to the
exterior, quite independently of either of the two tubular
glands of the seventeenth and nineteenth segments.

In my paper on Acanthodrilus (1) I have wrongly
stated that the two vasa deferentia of each side communicate
with the two tubular glands. I have since found that in A.
multiporus and A. dissimilis this is not the case. The
vasa deferentia open on to the exterior on the eighteenth
segment, close to the ventral pair of setae; they open by a
single pore, and only unite just before the external pore. It
is probable, therefore, that in Acanthodrilus generally there

1 In Trigaster Benliam lias made the interesting observation (‘Quart.
Journ. Micr. Sci.,’ vol. xxvii) that the muscular layer of the atria is partially
retained.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 125

is only a single male pore situated on the eighteenth segment
(see above, p. 102).

Ocnerodrilus (Eisen 12) offers an interesting parallel. In
this Annelid there is a saccular body opening in common with
the vasa deferentia, which is probably, as Yejdovsky has sug-
gested, the atrium.

There seems, indeed, to be little doubt that the so-called
prostates in these types are (i) homologous with each other,
and (ii) are homologous with the atria of the aquatic forms.
A gradual series of transitions unites Eudrilus, which is the
least modified, with Acanthodril us, which stands perhaps
at the other extreme. It is possible that the division of the
atrium in Eudrilus (Beddard 6) bears some relation to the
double atria of Acanthodrilus, but I have not yet
thoroughly investigated this point. The term “prostate”
must therefore be no longer applied to these glands.

The racemose glands of Perichseta, &c., now remain for
consideration, and the question which must be answered is :
Are these glands the homologues of the prostates of the
Tubificidse, or do they correspond to the atria of Eudrilus
and Acanthodrilus?

The structure of these glands is as follows: — They consist of
a series of branching ducts, lined with a single non-ciliated
cubical epithelium ; the ducts appear to end blindly, but groups
of glandular cells are attached to them here and there, and
doubtless void their secretion into the ducts. The ducts unite
into a main duct, which opens in common with the vas deferens
into a thick-walled muscular tube, which, at least in Peri-
chaeta Houlleti, can be evaginated, and probably serves as a
penis. The glandular cells are exactly similar in their struc-
ture to the cells of the prostate in the Tubificidse ; they also
resemble the glandular cells of the atrium in Pontodrilus
(cf. PI. XIII, figs. 12, 13). At first sight, therefore, three
hypotheses seem to be possible: either the whole structure
corresponds to the atrium of Acanthodrilus, differing only
in the branching of the cavity and in the segregation into
groups of the glandular cells, or the ducts alone are collec-

126

FRANK E. BEDDARD.

tively the homologues of the glandular region of the atrium in
Acantliodrilus, and the groups of glandular cells are the
homologues either of the Cementdriisen of Tubifex or of the
glandular covering of the atrium in Rhyne helm is and
Moniligaster. The crucial fact, however, which to my
mind settles the latter homology, is the presence of
a delicate peritoneal layer surrounding the whole
organ. In Moniligaster and in Rhynchelmis, &c., there
is no peritoneal layer surrounding the prostates, for the very
sufficient reason that the prostates are themselves the modified
peritoneal cells. All the structures, therefore, which lie within
the peritoneal layer must belong to the atrium; the so-called
prostate of Perichseta is therefore not the homologue
of the glandular investment of the atrium of Rhyn-
chelmis and Moniligaster.

There are some difficulties in the way of a comparison
between the prostate of Tubifex and that of Perichseta. Itis
true that there is a very considerable superficial similarity. The
origin of the prostate in the Tubificidae from the epithelium of
the atrium has been followed by both Yejdovskv and Eisen;
these facts therefore are in favour of the comparison. On the
other hand, it seems on a priori grounds likely that the
prostate of Perichseta is the homologue of the atrium in
Acanthodrilus ; in this case, as already pointed out, the
glandular cells must correspond in both genera ; in the
Tubificidae the atrium is lined by a single layer of cells, some
of which become modified into the prostate gland; the super-
added glandular layer of the atrium is altogether wanting. I
should be inclined, therefore, for the present to regard the
prostate of Tubifex as not strictly homologous with
the so-called prostate of Perichseta.

Although the structure of the “ prostate gland ” in the
majority of Perichetse is like that of P. Houlleti described
above, and presents therefore considerable resemblances to the
prostate glands of the Tubificidae, this is not always the case.
In a species (P. Newcombei) which I have recently de-
scribed from Australia, and which is probably identical with

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 127

one of those described by Fletcher (11)., the prostate glands
(PL XIII, fig. 15) have a tubular form like those of Acan-
thodrilus; this fact is also noted by Fletcher. I find,
however, that these glands are really branched like those of
other Perichetae (PI. XIII, fig. 13), but there is only a faint
indication of the division of the gland into lobules. The
structure of the gland in this species is therefore intermediate
between the ordinary Perichetae and Acanthodrilus, &c.
It has been said that the structure of the atrium in Acantho-
drilus is identical with that of the corresponding organ in
Perichaeta, allowing only for the branched character of
the supposed atrium in Perichaeta. In Perichaeta, how-
ever, the lining epithelium (PL XIII, fig. 13) is distinctly
columnar and not glandular ; it is very sharply marked off
from the surrounding glandular layers. In Acanthodrilus
(PL XIII, fig. 14) and Deinodrilus (PL XIII, fig. 16)
the lining epithelium is loaded with granules, and is on that
account rather difficult to distinguish from the glandular
layers surrounding it; in any case, this difference does not
appear to me to be one of importance; but if it were then
Pontodrilus could hardly be referred to either category. The
layers of cells which form the atrium (PL XIII, fig. 12) are
like those of Perichaeta in the obvious difference between the
innermost layer of cells and those which surround them ;
the cavity of the atrium is, however, unbranched like that
of Acanthodrilus. This series of facts leads me to
believe that the so-called prostates of Perichaeta
are equivalent to the atria of Acanthodrilus and
Pontodrilus, &c.

In Criodrilus Rosa (17) has recorded the presence of a
gland surrounding the external orifice of the vas deferens.
This he has termed the atrium. The investigations of Benham
have shown that this supposed atrium is nothing more than a
group of cells continuous with the clitellum. I have found
the same thing in Allurus.

The vas deferens passes through the glandular body and
opens on to the exterior ; it undergoes no changes in its

128

FRANK E. BEDDARD.

character (in Allurus), and the terminal section is not in any
way widened out to form a chamber which might be compared
with an atrium. The term prostate, in the sense in which it
has been used in the foregoing pages, is not applicable to the
mass of glandular cells which surround the end of the vas
deferens in Allurus; the structure in question is more rightly
to be compared to such a group of glandular cells as that
which surrounds the termination of the vas deferens in the
Enchytraeidse (Yejdovsky 19).

Summary.

The most important facts described in the present paper
are :

(1) The independence of the vasa deferentia and atria in
Acanthodrilus (PI. XIII, fig. 12) ; the two vasa deferentia
of each side unite just before their opening on the eighteenth
segment. The atria (= “prostates”) open separately upon the
seventeenth and nineteenth segments.

(2) The independence of the single vas deferens and its
atrium in Typhaeus; they open near together on the same
segment — the seventeenth (PL XII, fig. 1).

(3) The occurrence of six pairs of setae in each (setigerous)
somite of Deinodrilus (PI. XIII, fig. 9).

(4) The completely double dorsal blood-vessel of Acantho-
drilus annectens and of Deinodrilus Benhami.

(5) The enclosure of each half of the dorsal vessel of Dei-
nodrilus in a separate ccelomic space (PI. XIII, figs. 6, 7).

(6) The presence in Moniligaster Barwelli of an atrium
consisting of a thick glandular covering of peritoneum of a
layer of muscular fibres, and finally, of a single layer of
columnar epithelium (PL XII, fig. 11). The atrium is similar
in structure to that of Rhynchelmis.

STRUCTURE OF THREE NEW SPECIES OF EARTHWORMS. 129

List of Papers Referred to.

1. Beddard, F. E. — “On the Specific Characters and Structure of certain

New Zealand Earthworms,” ‘ Proc. Zool. Soc.,’ 1885.

2. Beddard, F. E. — “Note on Some Earthworms from India,” ‘Ann. and

Mag. Nat. Hist.,’ Oct., 1883.

3. Beddard, F. E. — “ Notes on Some Earthworms from Ceylon and the

Philippine Islands,” ‘ Ann. and Mag. Nat. Hist.,’ Feb., 1886.

4. Beddard, F. E. — “ Note on the Reproductive Organs of Moniligaster,”

‘ Zool. Anz.,’ 1887.

5. Beddard, F. E. — “On the Occurrence of Numerous Nephridia, &c.,”

‘ Quart. Journ. Micr. Sci.,’ Jan. 1888.

6. Beddard, F. E. — “ Contributions to the Anatomy of Earthworms,”

‘ Proc. Zool. Soc.,’ 1887.

7. Benham, W. B. — “ Studies in Earthworms,” Nos. 1, 2, 3, ‘ Quart.

Journ. Micr. Sci.,’ 1886-87.

8. Bergh, R. S. — “ Geschlechtsorgane der Regenwiirmer,” * Zeitschr. wiss.

Zool.,’ 1886.

9. Bourne, A. G. — "On Indian Earthworms,” pt. i, ‘Proc. Zool. Soc.,’

1886.

10. Bourne, A. G. — ■“ Anatomy of the Hirudinea,” * Quart. Journ. Micr.

Sci.,’ 1885.

11. Fletcher, J. J. — “ Notes on Australian Earthworms,” ‘ Proc. Linn. Soc.

N. S. W.,’ 1886-88.

12. Eisen, G. — “ Ocnerodrilus,” ‘Act. reg. Soc. Upsal,’ 1878.

13. Eisen, G. — “ Oligochretological Researches,” ‘ Report of the Commiss.

for Fish and Fisheries for 1883,’ Washington.

14. Horst, R. — “ Descriptions of Earthworms,” ‘ Notes from the Leyden

Museum,’ vol. ix, p. 97.

15. Perrier, E. — “ Recherches pour servir a l’histoire des Lombriciens

terrestres,” ‘ Nouv. Arch. Mus.,’ 1872.

16. Perrier, E. — “Pontodrilus,” ‘Arch. Zool. Exp.,’ t. ix, 1881.

17. Rosa, D. — “ Sul Criodrilus lacuum,” ‘ Mem. R. Acc. Torino,’ 1887.

18. Vejdovsky, F. — “Rhynchelmis li mosel la,” ‘ Zeitschr. wiss. Zool.,’

Bd. xxvii, 1876.

19. Vejdovsky, F. — ‘ System und Morphologie der Oligochaeten,’ Prag, 1884.

130

FRANK E. BEDDARD.

EXPLANATION OF PLATES XII, XIII,

Illustrating Mr. Frank E. Beddard’s Paper “ On the Structure

of Three New Species of Earthworms, with Remarks on

Certain Points in the Morphology of the 01igoch^eta.,,

PLATE XII.

Figs. 1 — 9. — Typhaeus Gammii.

Fig. 1. Male reproductive organs. T. Testis. M. Septum, a. Part of
seminal vesicle, f. Fuiinel of vas deferens, v. d. Yas deferens, s.
Setae of ventral pair. gl. Glandular part of atrium, s'. Sac contain-
ing penial setae, m. Muscular part of atrium.

Fig. 2. Transverse section through a portion of one of the intestinal
glands.

Fig. 3. Segments in the immediate neighbourhood of the male reproduc-
tive pores, to show copulatory papillae.

Fig. 4. Supra-cesophageal ganglia, v. Visceral nerves, c. Circum-ceso-
phageal commissure.

Fig. 5. Section through junction of atrium and setae sac. s. Penial seta.

Fig. 6. Principal vascular trunks, d. Dorsal vessel, v. Ventral, si.
Supra-intestinal. e. Epidermis.

Fig. 7. Anterior segments, to illustrate the number of annuli in each
segment, cp. Orifices of spermathecae.

Fig. 8. (Esophagus in the region of calciferous glands ( ca .).

Fig. 9. Dissection of the thirteenth and fourteenth segments, v. d. Vas
deferens, ov. Ovary, od. Oviduct.

Figs. 10 — 12. — Moniligaster Barwelli.

Fig. 10. Section through a body in the tenth segment, which is probably
the oviduct.

Fig. 11. Transverse section through atrium, v. d. Vas deferens, pr.
Glandular peritoneal cells, bl. Blood-vessels.

Fig. 12. Diagrammatic longitudinal section through anterior region of
body. s.o. Supra-cesophageal ganglion, s. Salivary glands, sp. Sep-
tum. at. Atrial pore. v. s. Seminal vesicle, od. Oviduct, cp. Sper-
matheca.

Fig. 13. — Acanthodrilus annectens. Dissection of genital segments.
t. Testes, ov. Ovary, od. Oviduct, f. Funnel of vas deferens.

STRUCTCJRE OF THREE NEW SPECIES OF EARTHWORMS. 131

PLATE XIII.

Eig. 1. — Typhaeus Gammii. Penial seta.

Pig. 2. — Typhaeus orientalis. Penial seta.

Pigs. 3 — 10. — Deinodrilus Benhami.

Pig. 3. Ventral view of genital segments, cl. Clitellum.

Fig. 4. Pirst three segments of the body.

Pig. 5. Spermatheca.

Fig. 6. Transverse section through dorsal blood-vessels, m. Muscular
layer of blood-vessel, p" . Peritoneal covering, p'. Peritoneal lining
of perihaemal space, p. Peritoneal covering of perihaemal space. E.
Intestinal epithelium, bl. Blood-space, t. Peritoneal cells covering
intestine.

Pig. 7. Longitudinal section of the same, lettering as above.

Pig. 8. Typhlosole in transverse section.

Pig. 9. Diagrammatic section of body, to illustrate arrangement of setae.
Pig. 10. Transverse section through body wall.

Pigs. 11 — 12. — Acanthodrilus annectens.

Pig. 11. Spermatheca.

Pig. 12. Transverse section through pore of vasa deferentia and apertures
of atria, v. d. Vas deferens, p. Atrium of seventeenth segment.

£ aperture of vas deferens, m. Intersegmental septa, at. Atrial
pore of nineteenth segment.

Pig. 13. — Transverse section through atrium of Perichaeta Newcombei.
a. Peritoneal covering, b. Glandular cells, c. Lining epithelium.

Fig. 14. — Transverse section through atrium of Acanthodrilus ; lettering
a3 above.

Fig. 15. — Atrium of Perichaeta Newcombei, drawn with camera lucida.
vd. Vas deferens.

Fig. 16. — Longitudinal section of part of atrium of Deinodrilus
Benhami; lettering as in Pig. 13.

*

jttcorjXourrv TtiiXXK .VSih-. X//.

j&cr.JlMm'- fan'. XXIX. MS, fa. A'/'/.

F E Beddiri del

F Hufh.Lsth? Zi\nT

DEVELOPMENT OE FAT-BODIES IN DANA TEMPORARIA. 133

Development of the Fat-bodies in Rana tem-
poraria. A Contribution to the History of
the Pronephros.

By

Arthur E. Giles, B.Sc.(Loml.), 91. B., Ch.B«(Tict«)>

PJatt Physiological Scholar, Owens College, Manchester ;

House Surgeon, Manchester Royal Infirmary.

With Plate XIV.

It has generally been held, since the researches of von
Wittich, that the fat-bodies, or corpora adiposa, in the Frog
and allied Amphibians, are derived from the genital organs by
a process of fatty degeneration in the anterior end of the
primitive genital ridge.

Yon Wittich himself says1 “they (the fat-bodies) have not
at any time any connection with the Wolffian bodies, nor with
the kidneys or their ducts.” And again, “the genital organs
become constricted into an anterior and a posterior part, of
which the anterior becomes the fat-body, and the posterior the
genital organ.”

This investigation, conducted in the biological laboratories
of Owens College, was begun with the intention of ascertain-
ing if these views of von Wittich were correct. But its pro-
gress showed that the mode of development of the fat-bodies
is very different from what von Wittich thought, and that the
changes which take place are of a very interesting nature.

On dissecting a tailed Frog, as represented in fig. 1, the fat-
bodies were seen already beginning to take on the lobose form

1 ‘Beitrage zur morphologischen und histologischeu Eutwickelung der
Ilarn und Geschleclits Werkzeuge der nackten Amphibien,” ‘ Zeits. fur wiss.
Zool.,’ 4te Band, 1853, pp. 148, 149.

VOL. XXIX, PART 2. NEW SER.

K

134

ARTHUR E. GILES.

which characterises them later on, and having in this case a
peculiar resemblance to the fingers of a hand, as represented
in the right half of fig. 2. Thinking that the light-coloured
bodies from which they sprang were the geuital organs, these
were removed with the surrounding parts, and cut horizontally
in successive sections. It was then found that what had been
taken macroscopically for genital organs (the real genital organs
not appearing so plainly as in fig. 1) showed microscopically
most typical kidney structure, whilst at the same time it was
quite continuous with the fat-body at the anterior end.

Two questions naturally arise : (1) How come the fat-bodies
to be in relation with the anterior end of the kidneys ? (2)

How does the transition from this condition to that found in
the adult take place ? These questions I propose to answer in
the following account.

Up to the age with which we are concerned, the generative
cells are found in the condition of primordial ova, as described
by Balfour ;x hence there is no differentiation into ovary
and testis. I shall therefore uniformly use the neutral term
“ genital organ.’*

The method adopted was to begin with very young tadpoles,
and cut series of sections at various stages and in various
planes, with the following results.

In a tadpole 8 mm. long, that is, soon after the first appear-
ance of the external gills, the three primitive openings of the
pronephros into the body-cavity can be seen. The tubules
forming the pronephros are actually of larger diameter than
they are somewhat later. One of them is represented in fig.
3, in which it is seen that the cells lining the tubules are
cubical or columnar, granular at the part nearest the lumen,
and showing a distinct radial striation peripherally. The
nucleus is central, and stains readily, as does also the nu-
cleolus. The genital organ at this stage is situated nearer the
median line than the pronephros, and anterior to it, and is
well defined both anteriorly and posteriorly.

The young tadpole at the stage we are considering has still a
1 ‘Comparative Embryology,’ vol. ii, p. 747.

DEVELOPMENT OE PAT-BODIES IN RANA TEMPORARIA . 135

plentiful supply of food-yolk, and is consequently independent
of nutrition obtained from without. But now, as it grows, the
absorption of the food-yolk proceeds more rapidly, while at the
same time certain changes are observed, notably the gradual
atrophy and subsequent disappearance of the external gills,
while the recently acquired internal gills take on more work.

If at this stage the pronephros be examined, it will be seen
that the tubules have a narrower diameter than those of the
younger tadpole, while the cells are not so clearly defined.
By this time the number of funnel-like openings into the
body-cavity has increased from three to five, and a new and
important structure has made its appearance, the mesonephros,
developed, as Sedgwick has shown,1 in the mesoblast inde-
pendently of the peritoneal epithelium.

The meso- and meta-nephros are not distinct from one another
in the tadpole ; they together form the kidney as found in the
adult, and it is in this sense that the word kidney will be used.

The mesonephric tubules extend gradually from behind
forwards till they come in contact with the pronephros. The
whole nephros then acquires a distinct capsule, becomes sepa-
rated from the muscular substance of the lateral mass, and
lies freely in the abdominal cavity on the ventral aspect of the
vertebral column, the peritoneum passing over it. Between
the two kidneys is the aorta (fig. 5). The genital organs,
which arise as two hollow ridges, also gradually separate from
the body wall, lying internal and ventral to the kidneys (fig. 5),
and are still perfectly well defined anteriorly, the proper
genital substance extending quite to the anterior end.

Concurrently with these changes of conformation, the
structure of the pronephros has been undergoing modification
of the nature of a fatty degeneration. At the time that the
hind limbs are just making their appearance, the degeneration
has gone on to the extent represented in fig. 4.

1 “ On the Early Development of the Anterior Part of the Wolffian Duct
and Body in the Chick, together with Some Remarks on the Excretory
System of the Vertebrata,” ‘ Quart. Journ. Micr. Sci./ vol. xxi, N. S.,
1881, p. 449.

136

ARTHUR E. GILES.

The way in which this conversion of kidney parenchyma into
fat takes place is a true fatty degeneration, and not simply a
fatty infiltration, though the latter occurs in the first stage.
The change is seen best in the cells lining the glomeruli
and renal tubules.' The clearly defined margins of the cells
become hazy, and the nuclei less distinct ; fatty droplets
appear at various parts of the cell and run together. The
cells do not, however, swell up as the fatty matter invades
them, but their protoplasm becomes replaced by it. At a later
stage the contents of the cell consist only of fatty granules and
granular detritus.

For a while the outlines of the convoluted tubules can still
be made out, as in fig. 6, the line of distinction between normal
and degenerated kidney being well marked. But soon all trace
of structure disappears, and there remains only a uniformly
granular-looking mass, as in fig. 2. This is the fat-body,
or corpus adiposum.

We have thus answered the first question that we proposed,
“ How come the fat-bodies to be in relation with the anterior
end of the kidneys ? ” There further remains to be considered
tbe question, “ How does the transition from this condition to
that found in the adult take place ? ”

When the hind limbs of tbe tadpole have appeared they
develop fairly rapidly, the fore limbs sprouting out somewhat
later. The condition of the “ tailed Frog ” is now attained.
While this is going on a change takes place in the urino-genital
organs, which, as regards the time at which it occurs, varies
somewhat in different tadpoles, but usually begins during the
period in which the tail is commencing to atrophy, and is for
the most part completed by the time the tail is quite absorbed.

This change is as follows : the anterior end of the nephros
grows ventrally, and becomes secondarily attached to the
anterior end of the genital organ, ovary or testis, as the case
may be. Fig. 7 shows the urino-genital organs during this
intermediate stage of transition, the fat-body (/.) being directly
continuous both with the kidney ( k .) and with the genital
organ ( g .).

DEVELOPMENT OF FAT-BODIES IN EANA TEMPORARIA. 137

This occurs at about the time that the mesonephric tubules
are growing out towards the genital organ, forming the future
vasa efferentia in the case of the male. Thus in some sections
the condition of this double outgrowth from the excretory to
the genital organs can be seen, the three parts of the nephros,
pro-, meso-, and meta- being quite continuous.

Ultimately the attachment of the fat-body to the kidney
gives way, and the former remains attached to the anterior end
of the genital organ, as it is in the adult (figs. 8 and 9). We
thus see that the fat-body is not the anterior end of the genital
organ, which has undergone fatty degeneration, as was thought
by von Wittich, but that its attachment to the genital organ
is secondary.

The fatty degeneration is always complete before the attach-
ment to the genital organ takes place ; almost any tailed Frog
that has not very long had its four limbs showing the fat-body
attached to the anterior end of the kidney. That it is, in
reality, fat-body is shown by its macroscopic and microscopic
characters, and by its staining with osmic acid. The part
marked (/.) in fig. 7 has exactly the same structure and
appearance as that similarly marked in fig. 8, the two specimens
having been stained and cut at the same time.

Having thus decided that the fat-bodies are derived not from
the genital organs but from excretory structures, we have to
consider what part of the nephros it is to which they owe their
origin. It can only be pro-, meso-, or meta-nephros, or their
ducts; the ducts can be at once put aside, because their
destination has been clearly and definitely made out. The
meso- and meta-nephros are also known to form together the
permanent kidney, as found in the adult.

There remains, therefore, only the pronephros, which, in the
Amphibians at least, has hitherto received but little attention,
though Sedgwick1 mentions that it undergoes atrophy in the
young Frog. “ Atrophy,” however, implies diminution in size,
or even total disappearance ; the pronephros of the tadpole,
on the contrary, not only persists but actually gets larger (in
1 Op. cit., p. 445.

138

ARTHUR E. GILES.

its modified form) as the Frog grows. Now, we saw that at an
early stage the pronephros undergoes a fatty degeneration ;
that the degenerated part remains for a time continuous with
the rest of the kidney (fig. 6), and then becomes secondarily
attached to the genital organ. Hence the fat-bodies
represent the persistent pronephros, profoundly
modified both in structure and in function.

If it be objected that it is a priori improbable that the fat-
bodies should consist of the anterior part of the nephros
detached and fastened on secondarily to the genital organ, it
will be sufficient to recall the fact that the vasa efferentia are
formed by a quite parallel growing out of kidney structure —
the mesonephric tubules ; the only difference being that the
process in the case of the fat-body goes a step farther, since
the primary connection with the kidney is lost, while the vasa
efferentia remain connected with both kidney and testis.

Again, the question may be asked, “ Why should only a
part of the kidney structure undergo fatty degeneration — why
should not the meso- and meta-nephros share in the change?”
The answer would be even more difficult to find if, on the
supposition that von Wittich was right, such a question were
asked concerning the genital organs, for they are of equal value
in all their parts, and when the metamorphosis occurs no
portion of them has had any reproductive activity. But in the
case of the pronephros it is different. It is true that the nature
and origin of the pronephros are still matters of discussion, but
it is at least evident that the pronephros is in many respects
different from the mesonephros ; that the former, in the case
of the Frog and of all animals with a larval stage, has a period
of activity before the mesonephros appears at all, and in most
cases disappears as the latter begins to take on active functions.
On the other hand, in Vertebrates possessing no larval stage,
the existence of the pronephros is only dimly shadowed forth
by rudimentary traces, the meso- and meta-nephros performing
all the excretory functions from the first.

The answer then to the question, “ Why this change should
occur resulting in the formation of the fat-body,” seems to be

DEVELOPMENT OF FAT-BODTES IN RANA TEMPORARIA. 139

this — that with the close of larval life the pronephros is no
longer needed, and in harmony with the pathological law that
atrophy follows disuse, it degenerates to the condition of fat-
body. Doubtless, however, this law is here so far modified
that the fat-body still serves some useful purpose in the organism,
though what that purpose is is not at all clear. It is in all
probability an example of “ change of function,” the later
function being in some way nutritive.

As to the distribution of fat-bodies — they are unknown out-
side the Amphibian group. According to Stannius, Hoffmann,
Wiedersheim, and others, they are present in all Amphibians.
We have very little knowledge of their function beyond that
they are concerned in all probability with nutrition, serving
as a reserve stock at certain times of the year. They are
differently placed in the several groups in which they occur,
and it is by no means certain whether they are homologous
structures in all cases.

The fate of the pronephros in the Frog, as above-described,
throws some light on the condition that obtains in other
groups of Vertebrates.

It was stated by Balfour1 that “ the pronephros atrophies
more or less completely in most types, though it probably
persists for life in the Teleostei and Ganoids.”

In a later paper,2 however, after working over the condition
of the kidneys in the sturgeon and in certain Teleostei, he
stated that “ the whole of the apparent kidney in front of the
ureter, including the whole of the so-called head-kidney, is
simply a great mass of lymphatic tissue, and does not contain
a single uriniferous tubule or Malpighian body,” from which
he concluded that both in Ganoids and in Teleostei the organ
usually held to be pronephros is actually nothing of the kind.
He therefore considered that Rosenberg3 was mistaken in

1 * Comparative Embryology,’ vol. ii, p. 729.

5 “On the Nature of the Organ in Adult Teleosteans and Ganoids, which
is usually regarded as the Pronephros or Head Kidney,” ‘ Quart. Journ. Micr.
Sci.,’ vol. xxii, N. S., 1882.

3 ‘ Untersuchungen iiber die Entwicklung der Teleostierniere,’ Dorpat,
1867.

140

ARTHUR E. GILES.

thinking that he had traced in the pike the larval organ into
the adult part of the kidney called by Hyrtl the pronephros ;
and his final conclusion was “ that the pronephros, though
found in the larvae or embryos of almost all the Ichthyopsida,
except the Elasmobranchii, is always a purely larval organ,
which never constitutes an active part of the excretory system
in the adult state.” But Balfour did not apparently regard it
as possible that the pronephros might continue in the Ich-
thyopsida in a modified condition, but thought that if it did
not persist Avith at least its original structure, if not its original
function, it must have disappeared altogether. He was, how-
ever, led to this conclusion by the study, not of their develop-
ment, but of their adult structure.

But it seems to me, from a consideration of the state of
things in the tadpole and young Frog as above described, that
it is not at all necessary that the pronephros, if it persist,
should retain its original structure any more than its original
function ; that it is quite possible that Rosenberg’s observa-
tions were correct, since the only argument adduced against
them is this alteration of structure, and that there is nothing
in Balfour’s observations on the Ganoids and Teleosteans to
contradict them. The fate of the pronephros in Teleosteans
and Ganoids is, from this standpoint, closely analogous to that
in the tadpole, except that in the latter it undergoes yet
further modification in becoming quite separated from the true
kidney and attached permanently to the genital organ.

The fact that the pronephros does persist in a modified form
seems to me in nowise to detract from but rather to add to
the probability of Gegenbauer’s views being correct, namely,
that the pronephros is the primitive excretory organ of the
Chordata, and that its substitute in existing Vertebrata, the
mesonephros, is phylogenetically a more recent organ.

I may sum up my conclusions as follows :

I. The fat-bodies in the Frog, and hence presumably in
allied Amphibians, are formed by a fatty degeneration, not
of the anterior end of the genital organs, but of original
kidney structure.

DEVELOPMENT OF FAT-BODIES IN BAN A TEMPORAEIA. 141

II. The part of the kidney which undergoes this con-
version into fat-body is the pronephros or head-kidney.

III. It seems very probable from analogy, and from the
researches of Rosenberg, that the structure in front of the
true kidney in Ganoids and Teleostei, described by Balfour as
lymphatic tissue, is the persistent but structurally and func-
tionally modified pronephros.

1Y. The fact that a part of the kidney undergoes such
a remarkable change, the rest remaining normal and func-
tional, is an additional argument in support of the view that
the pronephros has a different phylogenetic history from the
mesonephros, and that it is more ancestral.

It only remains for me to perform the pleasant duty of
expressing my warm thanks to Professor A. Milnes Marshall
for the uniform and stimulating kindness with which he has
helped me iu this short research by suggestions and criticisms ;
he has been good enough to go over my specimens with me,
and to discuss with me my results.

I desire also to express my obligations to my friend Dr.
G. Herbert Fowler for much valuable and practical assistance.

My thanks are further due to Professor Stirling, under
whose direction the work has been done.

142

ARTHUR E. GILES.

DESCRIPTION OE PLATE XIV,

Illustrating Mr. A. E. Giles’s paper on “ The Development of
the Fat-Bodies in Rana temporaria.”

The letters have the same significance in all the figures, b. Musoles of
body wall on the ventral aspect of the vertebral column, f Fat-body. g.
Genital organ (sex undifferentiated), k. Kidney, n. Notochord.

Fig. 1. — Tailed Frog, dissected so as to expose the uriuogenital organs.
The kidneys are seen lying against the vertebral column, and continuous
anteriorly with the fat-bodies. Anterior and internal to the kidneys are the
genital organs. X 4.

Fig. 2. — Anterior end of the urinogenital organs of the tailed Frog shown
in Fig. 1, enlarged. The right half of the figure shows the surface view, the
left half shows the appearance in horizontal section. X 50.

Fig. 3. — Normal pronephric tubule, from a tadpole still possessing external
gills. X 350.

Fig. 4. — Pronephric tubules showing fatty degeneration, from a tadpole
whose hind limbs were just appearing. X 350.

Fig. 5. — Transverse section through the lumbar region of a tailed Frog,
showing the mode of development of the genital organs and their relation to
the excretory organs at this stage. X 50.

Fig. 6. — Sagittal section through the lumbar region of a tadpole that had
recently acquired its fore-limbs, showing the anterior end of the nephros
partly degenerated. X 60.

Fig. 7. — Sagittal section through the lumbar region of a tailed Frog whose
tail had begun to be absorbed, showing the fat-body connected with both
kidney and genital organ. X 60.

Fig. 8. — Sagittal section through the lumbar region of a young Frog that
had just lost its tail, x 60.

Fig. 9. — A young Frog at the same stage as the preceding, dissected so as
to expose the urinogenital organs, which present the same condition as in the
adult. X 4.

Pig: 4.

x350

g-

k-

Fig 2
x 50

Fig. 3

x350

F!g,5

x 50

A t. Giles. del.

Fag 6

x60

„ <(ur>-Jbu rru %( WY,NSMM''

Fig. 8.

x 60

Fig. 7.

x60

.1th 6 Imp Scl Initio

TWO NEW TYPES OF ACTINIARIA.

143

Two New Types of Actiniaria.

By

G. Herbert Fowler, B.A., Ph.D.,

Assistant to the Jodrell Professor of Zoology in University College, London.

With Plate XV.

In a bottle of corals, which had been collected from the reefs
at Papeete during the expedition of H.M.S. “ Challenger,” and
sent to me by Mr. John Murray for investigation, I was fortu-
nate enough to meet with three small specimens of an Actin-
arian, which differs so markedly from all known types, that it
will apparently necessitate in the future the formation of a new
tribe of Actinaria, of equal value with the Hexactinise, Edward-
siae, Ceriantheae, &c. From the study of such an isolated
form, it is naturally impossible to deduce a satisfactory defini-
tion for either tribe, genus, or species. I will therefore leave
this omission to be filled up by future observers of allied forms,
and merely describe the anatomical characteristics in order. I
propose for the animal the name

Thaumactis medusoides, gen. sp. nn.

Of the three specimens at my disposal, the largest was about
2'5 — 3 0 ram. in diameter, the second about 2'0 — 2 5 mm.,
and the smallest 0'8 mm. ; and, while the two larger were in a
state of contraction (fig. I),1 the smallest (fig. 2) was fairly

1 Figs. 1 and 2, though carefully drawn with camera lucida under reflected
and transmitted light, were seen, on the study of sections, to be inaccurate in
some points, such as the exact number of the tentacles. In cases of dis-
crepancy between these two figures and the text, the latter is therefore to be
followed.

144

G. HERBERT FOWLER.

well expanded. The whitish-yellow colour of the specimens is
to be attributed merely to preservation in alcohol.

The animal is flattened in shape, and almost medusiform ; it
appears to be free-swimming (? crawling), for the aboral
ectoderm is entirely similar to that of the oral surface, and
shows no trace of attachment, past or future, to any foreign
body. From the biconvex shape, it follows that there is no
true body wall (mauerblatt, colonna), but the animal is divisible
into oral and aboral surfaces. Of these, the oral surface is
beset irregularly with what I shall term pseudo-tentacles, since
neither in number, position, nor structure can they be regarded
as homologous with true tentacles (fig. 1). In the expanded
specimen (fig. 2) fourteen true tentacles surround the stomo-
dseum, and peripherally to them are seen the earliest buds
of the pseudo-tentacles ; but in retracted specimens (fig. 3) the
true tentacles, together with the stomodaeum, are drawn
downwards and outwards into the coelenteron. From the
regularity and symmetry with which this is effected in both
cases, it is evidently the normal mode of retraction, and is not
due to death struggles or alcoholic contortion.

The aboral surface is covered by a single layer of columnar
ectodermal cells, which are shortest at the centre of the disc,
and lengthen towards the circumference, at which the two
surfaces meet in an acute angle. The oral surface is histo-
logically identical with the aboral, but bears the pseudo-tenta-
cles scattered irregularly over its surface to within a short
distance of the bases of the true tentacles. The point at
which the pseudo-tentacles cease marks the boundary of that
part of the oral disc which is drawn inwards and downwards
in retraction by the action of the sphincter muscle.

The pseudo-tentacles, three stages in the growth of which
are shown in fig. 4, arise each as a simple hollow outgrowth
from the coelenteron, in which all three body-layers take part
(fig. 9). The bud extends laterally over the surface into three
or four “ roots,’’ and is continued upwards as a free, finger-like
process (fig. 4). The cavity is nearly obliterated by the
presence of great numbers of zooxantliellae. The ectoderm on

TWO NEW TYPES OF ACTINIARIA.

145

the apices of the “ roots ” is generally well supplied with
nematocysts, that of the finger-like process is simple and
devoid of nematocysts when present, but in the older specimen
it has generally disappeared, leaving the mesogloea bare. On
the latter a slight musculature is generally recognisable, but is
not sufficiently strongly developed for a determination of its
origin and direction ; it probably agrees with that of the rest
of the animal. From the true tentacles these structures are to
be distinguished by their shape (the presence of the “ roots”),
by the absence of nematocysts on the motile finger-like
process, and by the fact that they are irregularly distributed,
bearing no relation to the mesenterial chambers either in
number or in position, but appearing in all stages of forma-
tion in the neighbourhood of a single mesenterial chamber
(fig. 4).

The true tentacles are set on the boundary between oral
disc and stomodaeum. In the largest specimen they were
twenty in number, i.e. one to every pair of mesenteries, with
one exception. In the smallest specimen fourteen were present
as against eleven pairs of mesenteries. They are perfectly
normal evaginatious of the intra-mesenterial chambers (ento-
coeles) ; their ectodermal layer is slightly marked off into
batteries of nematocysts, most obvious in longitudinal sec-
tions. The ectodermal longitudinal muscle is well developed,
the endodermal circular layer much weaker. No pore is
present at the tip of the tentacle.

The stomodseum, which follows immediately on the ten-
tacles, in the expanded specimen occupies the usual position,
but in the contracted examples is turned upwards and inwards
in the remarkable manner represented in fig. 3, st. It is
covered by a single layer of deeplv-staining columnar ectodermal
cells. No siphonoglyphe is recognisable on it at any point,
in either the expanded or contracted specimens.

The musculature of the general wall of the body is for the
most part very slightly developed, but may be recognised as
consisting of an endodermal concentric (circular) layer, and of
an ectodermal radial (longitudinal) layer. As Prof. Hertwig has

146

G. HERBERT FOWLER.

pointed out (f Chall. Rep. Zool. Actiniaria/ Supplt., p. 12), the
occurrence of the latter muscle-layer on the “ mauerblatt 55 is
confined to Corynactis and Cerianthus among the Anthozoa,
but is characteristic of both hydriform polyps and Scyphistomse
among the Hvdrozoa. As there is no reason to believe that
Thaumactis is derived from an Actinian-like ancestor with a
“ mauerblatt/5 the presence of this ectodermal longitudinal
muscle is of considerable phylogenetic interest. Both sets of
muscles are extremely weak, and consist merely of single
parallel fibrils, which produce a slight unevenness of the
mesogloea lamina.

In two regions, however, of the wall of the body, these
muscles attain to a more considerable development, namely, on
the indrawn part of the oral surface, and on the stomodseum.
On the ectodermal side of the mesoglosa in both these regions
are developed longitudinal muscle-fibres, adhering to pleatings
of the mesogloea (fig. 5), by which the expansion of the animal
is doubtless effected. On the endodermal side of the invaginated
part of the oral surface occurs a strong circular muscle, forming
a true sphincter of the “ diffuse 55 type ; in the contracted condi-
tion, this is continuous from the bases of the tentacles right on
to the horizontal surface, where it passes into the general circular
muscle; it is the chief muscle concerned in the invagination of
the disc during retraction. The endodermal circular muscula-
ture of the stomodseum is very slight, its function being
merely to close the entrance into the coelenteron during
digestion. The musculature of the tentacles has been noticed
in connection with them.

The mesenteries amounted in the largest polyp to twenty-
one pairs, of which one pair only were directive mesenteries.
Of the total number, six pairs (including the directive) are
“ primary/5 and are attached along the whole length of the
stomodseum; six are “ secondary/5 of which those four only
which lie nearest to the directive meet the uppermost (in the
expanded state) part of the stomodseum ; while the nine pairs
of tertiary mesenteries are developed chiefly in the neighbour-
hood of the directive pair, and do not touch the stomodseum

TWO NEW TYPES OP ACTINIARIA.

147

at any point. Their arrangement, beginning with the directive
pair is as follows : 1, 3, 2, 3, 1, 3, 2, 3, 1, 2, 1, 3*, 2, 1, 3, 2, 3,
1, 3, 2, 3. Over each pair of mesenteries is placed a tentacle,
with the exception of the pair marked above with an asterisk,
the position of which suggests that it was more recently deve-
loped than the rest. Muscles are present on both faces of the
mesenteries ; those on the outer (ectocoelic) aspect are the pro-
tractors, and are recognisable in transverse section only at the
upper part of the mesentery (fig. 5). Their function is to assist
the ectodermal longitudinal muscles of the oral surface and
stomodseum in the expansion of the animal. Those on the
inner (entoccelic) aspect are the retractors, by which a general
contraction is effected, and the stomodaeum pulled upwards
and inwards ; they are indicated in fig. 3 by faint lines on the
mesentery.

The free edge of the mesentery is not thrown into much
contortion, and for the most part bears the normal form of
filament (fig. 7). Besides this, however, occurs on most
mesenteries a structure, which I can neither describe nor
figure with any accuracy, owing to scantiness of material and
imperfect preservation. In transverse section of the polyp it
sometimes presents the appearance represented in fig. 8, but
more often appears as a solid swelling on the edge of, or in the
centre of, the mesentery. Above and below it the ordinary
form of filament often occurs. A number of these structures
were extremely obvious when the larger specimen was stained
and cleared (fig. 1), they then presented a gastrula-like
appearance.

In the smallest specimen, eleven pairs of mesenteries were
present, of which the rather larger six pairs are primary, the
remaining five secondary. No directive pair is present. The
number of tentacles (fourteen) would seem to indicate that, as
is often the case, the addition of new cycles of mesenteries is
to a certain extent preceded by multiplication of the tentacles.
The mesenteries essentially agree with those of the larger
specimen, except for the fact that they do not exhibit the
peculiar form of (?) filament. The contortion of the free edge

148

G. HERBERT FOWLER.

is proportionately much more considerable than in the larger
specimen.

No generative organs occurred in either of the three
examples.

It is a matter of regret that this new morphological type
does not throw any fresh light on the obscure phylogenetic
relationship of the various tribes of Actinaria to each other.
Doubtless the non-fixation (cf. the primitive Halcampae) and
persistent biconvex shape of the polyp indicate a condition
more or less ancestral, while from the ectodermal, longitudinal
(radial) muscle, which characterises hydroid-polyps and Scy-
phostomse, Prof. R. Hertwig would infer a very close relation
with the Hydrozoa, a conclusion which is certainly strengthened
by the shape of the body. A study of the structure of the
animal certainly does not suggest that it is a highly modified
form derived from representatives of the existing Hexactinian
type, but rather that it is collaterally descended from an ancestor
represented by the somewhat flattened larva of modern Antho-
zoan embryology, with no distant kinship to the Hydrozoa.
From existing forms, Thaumactis is marked off by the pseudo-
tentacles, the method of retraction, the abnormality of the
directive mesenteries, the biconvex shape, and the ectodermal
radial musculature ; and may possibly stand as type of a new
tribe, the Thaumactinise.

Phialactis neglecta, gen. sp. nn.

Of this new Actinian, two broken examples were found in the
same bottle with the Thaumactis described above, attached to
pieces of Millepora sp. from the Papeete Reefs. Its claim to
interest lies in the fact that it affords a further example of that
retrogression of the tentacles, of which the only known
examples are four genera of Hexactinise, and two Paractinia?,
all occurring among the deep-sea “ Challenger ” Actiniaria
described so ably by Prof. R. Hertwig. From these, how-
ever, this new genus differs in the fact that the tentacles are
replaced, not by stomidia — slight elevations of the oral disc,
surrounding a large opening which is homologous with the

TWO NEW TYPES OE ACTINIARIA.

149

pore at the tip of some normal Actiniarian tentacles — but by
what I will term sphseridia, i. e. ampullate diverticula of the
inter- or intra-mesenterial chambers, devoid of an opening to
the exterior, and homologous, therefore, with the imperforate
tentacles of many genera.

This difference appears to necessitate the formation of a new
family of Hexactiniae, the Phialactidse, which will rank
beside the Liponemidae, and may be defined as “ Hexactiniae,
in which the tentacles have degenerated into sphaeridia.” The
possibility, however, must be borne in mind that Phialactis may
belong, not to the Hexactiniae, but to the Monauleae (should
such a group prove to be natural), since in the one specimen
of which transverse sections were made, only one pair of direc-
tive mesenteries could be detected with certainty (cf. p. 150).
With the single form only at disposal, it is better to abstain
from even a provisional definition of genus and species.

The animal is goblet-shaped, the cup being represented by
an upward extension of the oral disc, the stem by the body of
the animal, and the foot by the limbus or base of attachment
(figs. 10, 11). The sphseridia are borne on the inside of the
cup only, and are especially numerous round the oral cone.
The latter lies at the bottom of the cup, and does not itself
bear sphseridia, in its centre lies the oval entrance to the
stomodaeum (fig. 12).

The external dimension of the most perfect specimen are
approximately as follows :

Total height

. 6

mm

Height from limbus to oral opening

. 2

a

Height from oral opening to upper edge of cup

. 4

it

Diameter of cup above ....

. 10

it

Diameter of cup below (outside) .

. 7

D

Diameter of body

. 6

))

Diameter of limbus

. 10

it

The general structure agrees with that of an ordinary
Actinian, the abnormal shape being produced merely by a
considerable upward growth at the point where body wall
(mauerblatt) passes into oral disc. The outer part of the

VOL. XXIX, PART 2. NEW SER. L

150

$

G. HERBERT FOWLER.

cup thus formed is to be regarded, therefore, as belonging to the
former, the inner side, which carries the spheeridia, to the latter.
The structure of the cup can be gathered from the schematic
figure 14 without further description. The thickness of the
mesogloea appears to be characteristic of the whole animal.

The sphaeridia, the degenerate representatives of the ten-
tacles, call for little remark ; they are hemispherical ampullae
(figs. 13, 14, sph.), scattered irregularly over the oral disc,
communicating by a passage narrower than their diameter,
with either inter- or intra-mesenterial chambers. No dis-
tinction into cycles is possible, no special musculature is
recognisable. In a specimen not figured they were rather
more numerous than in fig. 12, and set more regularly in
rows corresponding to particular mesenterial chambers. They
are covered by simple columnar epithelium, are devoid of
nematocvsts, and present no terminal pore.

The stomodseum exhibits a slight structural variation from
the normal type ; it is marked internally by a series of tongue-
like ridges produced by inward growth of the mesogloea and
ectoderm, the endoderm taking no part in their formation
(fig. 13). They do not correspond to mesenteries or mesen-
terial chambers. No siphonoglyphe is recognisable.

The mesenteries in the most perfect specimen amounted, to
twenty-three pairs, at and below the plane of the oral opening,
of which twelve were complete, and comprised the first two
cycles, while the remaining eleven pairs may be referred to an
incomplete tertiary cycle. Near the lip of the cup, at least
fifty pairs were present, so that in this, as in some other genera,
new mesenteries take origin just under the oral disc, and not
in the angle between body wall and pedal disc.

Only one pair of directive mesenteries could be determined
by transverse sections of the most perfect specimen ; a second
pair was perhaps present, but unrecognisable owing to the
slight development of muscle on many of the mesenteries.
While in many cases the mesogloea lamina of the mesenteries
is reduced to a thin refringent line, in others it forms, at the
plane of the contorted edge of the mesenterial filament, a

TWO NEW TYPES OF ACTTNIARTA.

151

stout plate, club-shaped in transverse section, and carrying
large muscle-fibres (fig. 15). To this, but marked off from it
by a sudden change in the thickness of the lamina, is attached
the contorted region of the mesentery, provided with more
muscle than is generally the case.

The muscle of the body wall and oral disc is endodermal
and circular, and is not differentiated into a sphincter at any
point.

As is so often the case, nematocysts of two kinds were
present, of which the larger measured as much as ‘14 mm. x
•044 mm., and were provided with unusually large cnidocils.

In conclusion, I desire to express my thanks to Mr. John
Murray, by whose courtesy I am permitted to present an
account of these two interesting forms.

Since writing the above description, I have had the great
advantage of submitting my drawings to Professor ft. Hertwig,
who inclines to the opinion that Phialactis should be asso-
ciated with the Corallimorphidm. While it is probable that a
parallel retrogression of the tentacles has taken place in more
than one family simultaneously, it will perhaps be best, till
the steps in the process are known, to allow the Phialactidse to
stand near the Liponemidse, although the genera in both
families may be eventually found to be merely degenerate
representatives of other existing families.

EXPLANATION OF PLATE XV,

Illustrating Dr. G. Herbert Fowler’s paper on “Two New
Types of Actiniaria.”

Figs. 1 — 9. — Thaumactis medusoides, gen. sp. nn.

Fig. 1. — The largest specimen, retracted ; from the oral aspect. While
the rest of the polyp has been drawn by transmitted light, the pseudo-
tentacles scattered over the surface are, for clearness’ sake, represented
as if under reflected light. Compare with this fig. 3. The clear space

152

G. HERBERT FOWLER.

in the centre is the opening to the coelenteron, left on invagination of
the oral disc; the dark ring surrounding it is produced by the inverted
stomodseum, &c. ; and from this the true tentacles radiate outwards,
among the mesenteries. X 30.

Fig. 2. — The smallest specimen, expanded; viewed from the oral surface.
Round the true tentacles (cf. note, p. 143) are seen the budding
pseudo-tentacles. X 50.

Fig. 3. — Diagram of a vertical section of the contracted polyp, from a
camera lucida drawing, te. The true tentacles, ps. t. The pseudo-
tentacles. st. The inverted stomodseum. mes. The mesenteries.
X 30.

Fig. 4. — Portion of the oral disc, showing three stages in the development
of the pseudo-tentacles. X 47.

Fig. 5. — Transverse section through the invaginated oral disc and stoma-
todseum at the base of a tentacle in a contracted specimen. The arrow
indicates the plane of section in Fig. 6. The pair of mesenteries ex-
hibit the protractor exocoelic muscles in transverse section (compare
their trend in Fig. 3). The laminated cuticle external to the ectoderm
is probably only a mucous secretion, x 210.

Fig. 6. — Vertical section through the invaginated oral disc and stomato-
dseum, i.e. an enlargement of part of Fig. 3. The arrow indicates the
plane of section in Fig. 5. X 210.

Fig. 7. — Transverse section of normal mesenterial filament, x 210.

Fig. 8. — Transverse section of (?) abnormal mesenterial filament, x 210.

Fig. 9. — Section through the root of a pseudo-tentacle. The endoderm
is completely obscured by zooxanthellse. The endodermal circular
muscle of the wall of the body is well seen, x 210.

Figs. 10 — 16. — Phialactis neglecta, gen. sp. nn.

Fig. 10. — Lateral view of the polyp. X 3.

Fig. 11. — Diagrammatic longitudinal section of the polyp, showing the
position of the oral opening at the bottom of the cup. x 3.

Fig. 12. — View of the cup from above. X 3.

Fig. 13. — Schematic transverse section through the base of the oral
cone. sph. Sphseridia. st. Stomodseum. m. Mesentery, x 32.

Fig. 14. — Schematic transverse section through a part of the cup.
sph. Sphseridium. or.d. Inner or oral disc surface. b.u>. Outer or
body- wall surface, m. Mesentery, x 47.

Fig. 15. — Transverse section of a mesentery (p. 150). X 62.

Fig. 16. — Nematocyst of the larger kind, with cnidocil. X 210.

i^inated oral surface

.UicrSoum T^..UjXXsF. J7.

/"'to S.

inva£mated oral surface

Fur. IJ.

oral surface

Fiq. 10.

•Umrt"4*’

sfomatod® um

Fur II

tTtTT»r»rf,r

c >i w : ff \ <

GH. Fowler del

F Huft, Lift7 Eiirl

MORPHOLOGICAL STUDIES.

153

Morphological Studies.

II.— The Development of the Peripheral Nervous
System of Vertebrates.

Part I. — ELASMOBRANCHII AND AVES.

By

«V. Beard, Pli.D., B.Sc.,

Assistant to the Professor of Human and Comparative Anatomy
in the University of Freiburg i/B.

With Plates XVI, XVII, XVIII, XIX, XX, XXI.

TABLE OF CONTENTS.

Introduction (pp. 155 — 163).

I. The Peripheral Nervous System of Elasmobranchii.

a. The Spinal Ganglia. — The Zwischenstrang of His. The neural
ridge of Marshall. Julin’s views of the morphology of the lateral nerve.
The permanent and only root of attachment. Mode of growth of the
connecting fibres. The trophic properties of the ganglia as evidence
(pp. 161—173).

b. The Cranial Ganglia of Elasmobranchii. — The neural
ganglia. The lateral ganglia and sense organs. The origin of the nerves
of the sense organs. The formation of the permanent root of the nerve.
Tlie visceral motor fibres of the head (pp. 173 — 181).

II. The Peripheral Nervous System of the Chick.

a. The Spinal Ganglia. — The Zwischenstrang of His (pp. 183 — 185).

b. The Cranial Ganglia in the Chick. — Marshall’s conclusions.
His’s results and the “ Zwischenrinne.” Ouodi’s researches, c. The
neural ganglia. d. The rudimentary sense organs and their ganglia
(pp. 186—192).

154

J. BEARD.

III. The Development of Anterior Roots in Elasmobranchii . —
His’s views. Parablast cells. Confirmation of Balfour’s statements (pp.
192—196).

IV. The Ganglionic Development in Different Classes of Ver-
tebrates.— Identical results in all forms examined. Spencer’s statements
on Amphibia partially erroneous. Goette’s views in 1875 not identical with
mine (pp. 196 — 198).

V. The Neural Ridge of Marshall. — Sagemehl’s researches on
spinal gauglia. Onodi’s and His’s results on Chick. “ Zwischenriune ” has
no existence. In head is also a “ Zwischenstrang.” No direct connection
with formation of ganglia. Balfour’s views of origin of ganglia as outgrowths
of central nervous system. Marshall’s position (pp. 199 — 207).

VI. The Independent Epiblastic Origin of the Peripheral
Nervous System. — Semper’s and Goette’s observations. Van Wijhe’s
researches. Eroriep’s discoveries in Mammals. My statements on the system
of lateral sense organs in Elasmobranchs. Spencer’s views of origin of nerves
in Amphibia. Onodi’s results on Chicks. Researches recorded in this paper
show entire peripheral sensory nerve-elements to be formed in-
dependently of central nervous system from epiblast (pp.
207—210).

VII. The Relations of Cranial to Spinal Ganglia, and the
Question of the Morphology of the Lateral Sense Organs. —
Dohrn’s hypothesis. Eroriep’s conclusions (pp. 210 — 212).

Eisig’s comparisons between Vertebrates and Annelids. —
Eisig’s conclusions mainly hypothetical. Position of the question.

The Homology of Parapodial and Spinal Ganglia proposed by
Kleinenberg. — The difficulties still in the way of a comparison of the
“ Seitenorgaue ” of Annelids and Vertebrates (pp. 212 — 216).

VIII. Dr. Gaskell and the Eunctional Distribution of the
Cranial Nerves (216 — 218).

Resume (pp. 218 — 219).

Literature cited (pp. 220 — 223).

The researches recorded in the following pages were under-
taken in consequence of a grant for the purpose made from
the Government fund by the Grant Committee of the Royal
Society.

MORPHOLOGICAL STUDIES.

155

Introduction.

Nearly three years ago I published (No. 6) iu the pages of the
'Quart. Journ. of Micr. Sci.,’ some researches on the morphology
and development of the so-called "organs of the lateral line,”
which appeared to me then, as now, to be in reality special
branchial sense organs. When those researches were first
undertaken in Professor Semper’s laboratory (No. 4), it was
far from my intention to investigate the development of the
cranial nerves and ganglia, but in the course of the work it
soon became obvious that the thorough study of those sense
organs could only be accomplished by including the cranial
nerves and ganglia in the sphere of the observations. And,
indeed, almost the first Elasmobranch embryo examined
showed unmistakeably that the cranial ganglia and the sense
organs of the lateral line are intimately associated in their
morphology and development.

The researches then published include also attempts to
liomologise the sensory portion of the nose and ear with the
sense organs of the lateral line. And it is partly with the
wish to get more light on this question that last year I under-
took the investigation of the first beginnings of the cranial and
spinal ganglia. Two papers — very different in their stimulat-
ing effects — urged me all the more to a thorough study of
these problems.

In a short notice Dr. van Wijhe (No. 61) — whose right to
an opinion on this matter is unquestionable — considered his
researches on Ray embryos entitled him to combat my pre-
viously published views of the nature of the nose. I give in
the following lines van Wijhe’s own words.

He says : “ Die Auffassung nach welcher der Olfactorius ein
8egmcntaler Nerv sei, ist ueulich wieder von Dr. Beard ver-
treten. Er griiudet dieselbe auf die Theilnahme der Epidermis
an der Bildung des lleichnerven und seines Ganglions, wie
dies auch bei den Nerven der Seitenorganen der Fall ist und
glaubt die Reichgruben seien deshalb den Seitenorganen und
der Olfactorius den Nerven dieser Orgaue homolog.

156

J. BEARD.

“Ich finde dass der Olfactorius zu Anfang von

Balfour’s Studium I noch nicht vorhanden ist; er tritt erst
zu Anfang der Periode J auf, wann die vierte Kiemeutasche
schon angelegt, aber noch keine nach aussen durchgebrochen
ist. Das Riecborgan und der Nerv enstehen beide aus dem
vorderen Neuroporus. Der Olfactorius entwickelt sicli nicht
aus der Nervenleiste, denn er tritt in einer Periode auf, wann
dieselbe im Kopfe schon langst geschwunden ist ; auch ist er
von Anfang an mit der Haut in Verbindung, und unterscheidet
sicli durch diese zwei Merkraale von alien iibrigen dorsalen
Nervenwurzeln.”

“ Dass nun die Zellen der Anlage des Kiecborganes

an der Bildung des Nerven und seines Ganglions theilnehmen
ist, wie mir scheint, bei einer solchen Enstehungsweise a priori
zu erwarten, und stimmt mit der Nervenbildung bei vielen
Wirbellosen iiberein.”

“ Wenn Beard jetzt, seiner friiheren Bebauptung

entgegen, den Olfactorius und die Seitennerven nebst ihren
Ganglien allein aus der Epidermis enstehen lasst, so kann er
dies wohl nie beweisen, weil der Stamm der Nerven sich
urspriinglich aus dem Medullarrohre entwickelt.”

The above statements relate to some of the most important
problems in the development of the peripheral nervous system.
And if the whole of them are to be maintained in van Wijhe’s
sense they present insuperable difficulties in the way of the
acceptation of my previous interpretations. To me the most
serious question then, and before then, was the nature of
Marshall’s neural ridge and its supposed origin as an out-
growth of the central nervous system. If I had left it
entirely untouched in my previous work, such a proceeding
can be easily excused. In the first place, my material did not
then appear to me sufficient to settle the matter, and the com-
plete study of the “ neural ridge,” &c., required more time than
I could then devote to it. Further, I could not without ample
justification declare Marshall’s account and that of Balfour to
be at the basis erroneous. And when Professor Gegenbaur, in
his recent work (No. 21, p. 42), makes me the reproach that I

MORPHOLOGICAL STUDIES.

157

never entertained the question of the epiblastic origin, apart
from the central nervous system, of the neural ridge, the
accusation is unjust; for I can assure Professor Gegenbaur that
such was far from being the case.

And, as it now turns out, van Wijhe’s objection that
the main root of the nerve arises as an outgrowth of
the central nervous system will not hold, for the origin of
what he calls the main root, in the case of the cranial and
spinal ganglia, is demonstrably in principle in accordance with
the account of the development of the olfactory nerve and
ganglion, as given by van Wijhe himself. In fact, in the
main, I accept gladly and gratefully van Wijhe’s
researches on the olfactory organ as supporting and
confirming my view of its homology. But for a fuller
discussion of this matter, I must ask the reader to wait till
the ground has been cleared by the detailed account of the
researches on the very first origin of the cranial ganglia in
nearly all classes of Vertebrates.

And now a few words on the second work, which was the
great stimulating agent in impelling the researches about to
be recorded. If the reader w'ill refer to the introduction of
my work on the branchial sense organs (No. 6) he may read that
“ at present wre are acquainted with no Invertebrate nervous
system which is built on the same plan as that of Vertebrates.”
This conclusion led me to take up an attitude of expectancy
rather than of negation towards the Annelidan theory of the
origin of Vertebrates. And while I felt compelled to doubt the
homology between the “ Seitenorgane ” of the Capitellidse
and the “ Seitenorgane ” of Vertebrates, so ably maintained
by Eisig, I was not quite without hopes that further researches
on Invertebrates might reveal facts on which a comparison of
the peripheral nervous system of Vertebrates, with some allied
stock of Invertebrates, probably Annelids, might be main-
tained.

In October, 1886, appeared Kleinenberg’s epoch-making
researches on the development of Lopadorhynchus (No. 41).
I shall find plenty of opportunity in this and some of the fol-

158

J. BEARD.

lowing studies for reference to this remarkable paper, and will
here only quote one passage, which may serve as a text for the
researches I am about to record.

He writes (p. 219) : “ Die grossten Schwierigkeiten bietet
der Vergleick des Centralnervensystems der Wirbelthiere mit
dem der Anneliden. Zunachst der Umstand, dass bei den
ersteren eine durchaus einheitliche Anlage fiir Riickenmark
und Gehirn vorhanden zu sein scheint. Dies konnte durch eine
allmahlich eiugetretene und schliesslich mit den ersten
Bildungsvorgangen zusammen fallende Verschmelzung der
umbrellaren und sub-umbrellaren Abschnitte des Annelideu-
systems erklart werden, dann bliebe immernoch zu bestimmen,
welcher Theil dem urspriinglichen Kopfganglion entspricht.
Das ganze Gehirn gewiss nicht. Seine bei weitem uberwie-
gende Masse starumt offenbar vom Rumpf her ; in ihm fliesst
eine Anzahl von Ganglieu zusammen, die wohl nur auf die
sub-umbrellare Anlage zu beziehen sind. Die HofFnung, auch
bei den Wirbelthieren eine gesouderte Anlage, welche dem
Kopfganglion der Anneliden gleichwerthig ware, zu finden, ist
etwas kiihn. Wenn nicht wahrscheinlieh bleibt immerhin
moglich, dass das urspriiugliche Kopfganglion gauz unterdriickt,
und von den vorderen Theile der Bauclikette substituirt
worden ist. Denn weniger resolut als Dohrn, Semper, und
andere, bin ich doch geneigt mit ihnen die Homologie des
Riickenmarks der Wirbelthiere und des Bauchmarks der
Anneliden anzunehmen. Schon vorhin sprach ich die Ver-
muthung aus, dass der Primitivstreifeu einern Theil der sub-
umbrellaren Neuromuskelanlage entspricht. Es erscheint mir
als eiu grosses Verdienst Kolliker’s, fast allein den ecto-
dermalen Ursprung des Mesoderms an dieser Stelle festge-
lialten zu haben, ohne sich von dem lauten Widerspruch
beirreu zu lassen. Natiirlich trete ich alien deneu entgegen,
welche in der Primitivrinne weiter nichts als eiu Uberbleibsal
des Blastopors sehen ; eine gewisse Beziehung zu demselben
vertragt sich aber auch ganz gut mit meiner Auffassung.

“Das bleibende Riickenmark wiirde dann vielleicht nur den
vorderen Abschuitt der urspriinglichen Anlage enthalten. Der

MORPHOLOGICAL STUDIES.

159

Weg den die parallelen seitlichen Strange des Bauchmarks
durclilaufen haben miissen, um zum mediauen Riicken-
marksrohre zu werdeu, scheint mir durcli die mitgetheilten
Thatsachen aus der Entwiekelungsgeschichte der Anneliden
selbst hinreichend klar vorgezeichnet.1 Noch mebr. Aucb
die Spinalganglien diirften ihre Horaologie bei den
Anneliden finden, und zwar in den Parapodialgan-
glien. Dreht man die tig. 47, Taf. xi,2 um, so wird die
Uebereinstimmung nicht entgeben. Der Unterschied liegt
nur darin, dass die Spinalganglien bei ihrem Auftreten dickt
am Riickenmarksrohr liegen oder in dasselbe eingezogen siud.
Die hintere Wurzel — das am besten gekannte Beispiel der
Enstehung eines Nerven bei den Wirbelthieren — bildet sich
gerade so wie der mediane Parapodial Nerv, und die vordere
Wurzel diirfte dem Muskelnerven, der sich mit jenem zu
einem Stamm verbindet, gleich zu setzen sein.”

So much for the present from Professor Kleineuberg. In
general terms the result of my researches is a confirmation of
his views and comparisons.

If any further excuse were needed for a reopening of the
question of the origin of the ganglia in Vertebrates, one would
not have to seek far for ample reasons for such a course.
Just as I was completing the first part of this work, three
publications appeared, all of which showed the state of uncer-
tainty and vagueness in which these questions at present exist.
Professor Gegenbaur (No. 21) has undertaken no investigations
on the matter, but feels himself entitled to quote as final the
observations of one or other of his pupils, those of Sagemehl
(No. 56) more especially appear to him to be far away above
suspicion. I shall later on have occasion to point out how

1 In a subsequent paper I intend to demonstrate that the central nervous
system is a paired structure which arises as two lateral plates of
neuro-epithelium separated by a median ciliated groove, just as
in Annelida.

a I have reproduced this figure in Plate XIX, fig. 64. In my copy the figure
has been turned through an angle of 180°, to bring it in the “Vertebrate
position.”

160

J. BEARD.

little claim Sageiuehl’s researches on the spinal ganglia really
have to pose as a solution of the prize problem they were
undertaken to solve, and will here content myself with the
assertion that Sagemehl never saw any of the very earliest
stages of development. Professor His (No. 34) in a paper,
which in spite of a vast number of differences of opinion as to
both facts and hypotheses, I cannot regard otherwise than as a
valuable contribution to the morphology of the cranial nerves,
has, among other things, endeavoured to establish without
further observation his celebrated “ Zwischenrinne ” or
“ Zwischenstrang ” theory, and he believes that all that is
necessary for its final triumph is its rebaptism under the name
of Ganglienrinne or strang. As this work also will occupy
our attention for some time at a later stage of the work, I will
only express my strong dissension with the following extract
(p. 380) with which Professor His opens his campaign against
“ die jiingeren vergleicheudmorphologischen Schulen.” It
reads thus : “ Bei genauerem Zusehen findet man eben dass die
Differenzen nicht in dem liegen, was der eine oder der andere
Beobachtungskreis an thatsachlichen Befunden ergiebt, sondern
in demjenigeu was die Vertreter der einen und der andern
Schule zwischen den Zeilen zu leseu sich bemiihen.”

It will be time enough to consider the lecture which Pro-
fessor His reads to us younger morphologists, when the facts of
development which form the very basis of the question are
placed beyond the reason of dispute. The principle of the
origin of the ganglia from the epiblast, apart from the central
nervous system, is one on which I can agree with Professor
His. Not so with the way in which this takes place; for,
paradoxical though it may sound, right as Professor His was
in principle, he is till now further from recognising the true
facts than any embryologist who has worked on the origin of
the peripheral nervous system. Sad to relate the Zwischen-
strang, &c., has as little direct connection with the origin of
the ganglia as it has with the urinogenital system, as Professor
His at first supposed.

Professor His is astonished to notice that his views on this

MORPHOLOGICAL STUDIES.

161

matter have been “vollig unbeachtet,” and finds, — though
this discovery is not likely to be accepted by anyone compe-
tent to judge the question, — that his original views are
practically identical with the generally accepted account of
Balfour.

Now, among those naturalists who have worked on the
development of the peripheral nervous system, Balfour stands
pre-eminent in the precise formulation of his conclusions. I
am bound to maintain that on many of the most fundamental
questions Balfour’s observations cannot be longer upheld,
while I am also sure that none would be more ready than he
to accept the facts I am about to record.1 Balfour says (No.
2, p. 369) : “ All the nerves are outgrowths of the central
nervous system.” How this statement can be reconciled with
his Zwischenriune hypothesis (for it is nothing more than an
hypothesis) it is for Professor His to determine. The matter
need not trouble us much, for, as I shall afterwards show, the
Zwischenstrang (there is no Zwischenriune !) is just that portion
of the epiblast or ectoderm which takes no part at all in the
ganglionic formation. All I here wish to do is to enter a
protest against the way in which Professor His attempts to
convert all previous work on the early development of the
ganglia into a mere confirmation of his own more or less

1 It is certain that Balfour had an idea of the true facts, for he closes his
account of the peripheral nerves on page 383 of the ‘ Comparative Embry-
ology,’ vol. ii, with this passage : “ Situation of the dorsal roots of the
cranial and spinal nerves. The probable explanation of the origin of nerves
from the neural crest has already been briefly given. It is that the neural
crest represents the original lateral borders of the nervous plate, and that, in
the mechanical folding of the nervous plate to form the cerebrospinal canal,
its two lateral borders have become approximated in the median dorsal line
to form 1 lie neural crest. The subsequent shifting of the nerves I am unable
to explain, and the meaning of the transient longitudinal commissure con-
necting the nerves is also unknown. The folding of the neural plate must
have extended to the region of the olfactory nerves, so that, as just stated,
there would be no special probability of the olfactory nerves belonging to the
same category as the other dorsal nerves, from the fact of their springing
from the neural crest.” The reader may compare the first sentences of this
passage with the results recorded in the following pages.

162

J. BEARD.

hypothetical views on this matter. That the Zwischenstrang
has any concern in the formation of the ganglia is a baseless
assumption.

In his ‘ Lehrbuch der Entwickelungsgeschichte,’ &c., Ed. ii,
Professor O. Hertwig has made an attempt to extract a little
light from the chaos which reigns over our knowledge of the
development of the peripheral nervous system. For Professor
Hertwig, the most important researches are those made by his
pupils and by Sagemehl. As he mainly relies upon these and
ignores for all practical purposes almost entirely the more
recent work on the matter, it is not unnatural that the chapter
on the peripheral nervous system is one of the most unsatis-
factory in the whole work. As an instance of Professor
Hertwig’s treatment of recent authors, I may mention that for
him our knowledge of the formation of the lateral nerve of
Amphibians and Silachians is confined to the older observa-
tions of Semper and Goette, and he only mentions incidentally
that van Wijhe has seen similar fusions of epiblast and sensory
nerves in the head of Elasmobranchs (No. 23, p. 338).

Professor Hertwig has thought fit to illustrate his account
with one or two figures from as yet unpublished researches of
Professor Rabl. If Professor Rabl is to pose as an authority
on the formation of spinal ganglia, one may at least ask for
tolerably correct figures in illustration of his work. The two
figures 175 and 171 given by Hertwig are among the most
incorrect that have been published till now on this matter.
As the climax to Professor Hertwig’s appreciation of work on
the peripheral nervous system, let me add that he is of
opinion that “ um auf dem schwierigen Gebiete vorwarts

zu kommen, muss man bei der Untersuchung

von Embrvonen nicht nur Schnittserien, sondern auch andere
histologische Methoden zu Rathe ziehen” (p. 337). It is to
be hoped that besides giving the advice Professor Hertwig
will also show us the way to use his “andere histologische
Methoden.” From these citations the reader will, I think,
be convinced that from the researches till now published,
we may form very different conceptions of the results

MORPHOLOGICAL STUDIES.

163

obtained according as one belongs to this that or the other
school of embryologists. For myself, for the better compre-
hension of my work on later stages, it was absolutely essential
that a clear, precise, and uncontradictory account of the very
first stages of the peripheral nervous system, and of the rela-
tions of the latter to the central nervous system, should be
worked out. It was necessary to attempt to do for the Verte-
brate nervous system what Kleineuberg has done for that of
the Annelid. Not that the following researches make any
pretence to being an account comparable in minuteness of
detail with Kleinenberg’s work, they are rather the beginnings
of work on the matter; for there is still much to he done in
the early development of both central and peripheral nervous
systems of Vertebrates.

According to Professor Wiedersheim’s opinion and my own,
the most lasting results were likely to be obtained by drawing
within the sphere of investigation as many types of Vertebrates
as possible, and hence, although my original intention regarding
this and other researches was to consider only Elasmobrancbs
and Ganoids, in aid of which researches the Government Grant
Committee of the Royal Society of London made a grant of
money from the fund at their disposal, I nevertheless thought
it in the interest of science to extend my observations to various
classes of Vertebrates. So far I have had at my disposal embryos
of (various genera) Elasmobranchs, Teleostei, of Amphibia,
Reptiles, Birds, and Mammals. Researches on Ganoids1 and one
or two other types as yet unattainable, I hope in the course
of the year to be able to carry out. For the moment in con-
sequence of the time necessary for the preparation of the
numerous indispensable drawings, I publish the results obtained
on Elasmobranchii and Birds.

1 The development in Ganoids conforms exactly to those in other types.

164

J. BEARD.

ELASM0BRANCHI1.

The researches on this group were made on embryos of Tor-
pedo ocellata, Pristiurus melanostomus, Acanthias
vulgaris, Mustelus lsevis, and Scyllium canicula. Of
the first-named form especially a very large and complete
series of stages was at my disposal. This genus, Torpedo, is,
in my experience, the best suited for researches on the early
development, for the cell elements are larger, and the appear-
ances presented in sections much clearer than those of any of
the other forms mentioned. Of the other forms a sufficient
number of stages was at my disposal to show that there is no
essential difference in the development. And, in fact, for
both cranial and spinal ganglia of all the Vertebrates which
have till now come into my hands, including Teleostei, Rana,
and even the Chick, I may with full confidence say that the
appearances presented are all easily reducible to one type — to
that of the Elasmobranchii.1 The differences observed in
different forms are in reality very slight, and are readily ex-
plicable as variations in the time of development. As in the
case of other organs, the development may be either retarded
or accelerated. As a striking example of the way in which, for
instance, the spinal ganglia agree in development in Torpedo
and the Chick, I may mention that in sketches of portions of
sections of the two forms drawn under high power it is often
difficult, if not impossible, to find any differences, even in
detail; and if the reader will compare figs. 37 a, 42, and 68,
69, he will, I think, find it impossible of his own knowledge to
say definitely that the former are figures of Torpedo sections,
the latter of Chick sections.

a. Spinal Ganglia of Elasmobranchii.

Balfour is mainly responsible for our knowledge of the
development of the spinal nerves and ganglia in Elasmo-
branchs (Nos. 1, 2, and 3).

1 The development in Ganoids conforms exactly to that in other types.

MORPHOLOGICAL STUDIES.

165

The stages of development which Balfour described as the
earliest are, however, by no means such, for I can demonstrate
the first traces of ganglia some time before the neural canal
closes. Neither Balfour norOnodi,1 nor any other observer, has
seen the stages which I figure in PI. XVI, and in figs. 37 — 42
of PI. XVIII.

Figs. 1 — 4 of PI. XVI are taken from various parts of one
embryo of Torpedo ocellata. Figs. 1 — 3 are the only ones
which at the moment concern us, for they are all three from
the trunk, and hence from the region of the spinal ganglia.

In Prof. His’s recent paper (No. 34, p. 445) the author
remarks : “ Der Zeitfolge nach entwickeln sich die periphe-
rischen Nerven spat. Am Rumpf treten sei spater auf als die
Urwirbel, am Kopf fallt die Zeit ihrer Bildung zum nahe
an diejenige des Visceralbogen, aber da geht die Gliederung
des End-gebietes dem Vordringen der Stamme voraus.”

I am not quite sure that Professor His means these remarks
also to apply to the ganglia. But however that may be, I will
at once assert that the “ Anlagen”2 of the spinal ganglia are
formed very much earlier than has hitherto been supposed,
and, indeed, that the first traces of them appear when only two
or three of the mesoblastic somites3 have been entirely seg-
mented off from the main raesoblast (figs. 2 and 3). Generally
speaking, the first differentiation of the spinal ganglia may be
said to occur at about the time of separation of the notochord
from the hypoblast. In earlier stages than this fig. 1 (here the

1 Onodi’s researches, so far as they relate to the posterior root-ganglia of
Elasmobranchs after exclusion of the sympathetic, contain no new results
And their author was entirely in the dark as to the relations of the ganglia
to the lateral sense organs. Though he must have seen the skin fusions he
entirely ignores them.

s I use the word Anlage or Anlagen (plural) throughout this paper instead
of our only term rudiment, which has a double meaning.

3 In agreement with van Wijhe, Wiedersheim, and others, I use here the
word somites, or body-somite, instead of the older and incorrect term proto-
vertebrae. In the same way I shall call the “ head-cavities,” with van Wijhe,
the head-somites. With Dr. Eisig 1 use the terms haemal and neural instead
of ventral and dorsal.

VOL. XXIX, PAHT 2. NEW SER.

M

166

J. BEAKD.

notochord is already partly separated off) the neural plate
is a broad, slightly-grooved, shallow structure, which presents
no appearance of differentiation. Very soon the invagination
process begins, and with it the Aulagen of the spinal ganglia
begin at once to be distinguishable from the rest of the neural
plate. At the extreme outer boundary of what is really the
“Anlage” of the spinal cord a commencing proliferation of
the deeper layers of the epiblast is observable (fig. 2). This
leads to the appearance of a bud-like outgrowth of cells at each
side of the spinal cord Anlage. On the one hand this out-
growth is very soon sharply distinguishable from the spinal
cord Anlage ; on the other hand, it begins to separate some-
what from the rest of the epiblast in the form of a somewhat
wedge-shaped mass of cells. In figs. 3, 5, 9, this process is
readily made out. From an inspection of these figures it will
be obvious that the whole thickness of the epiblast is not con-
cerned in this outgrowth. The outer layer of epiblast
is quite indifferent, and neither takes share in the
Anlage, nor presents any resemblance at all to a
sense epithelium, a point on which I shall have more
to say in connection with Dr. Eisig’s comparisons.
The epiblast in the region of the ganglionic Anlage, and for
some distance lateral of this, is composed of several cell layers
(fig. 1). Now, the way in which the ganglionic Anlage sepa-
rates from the rest of the epiblast is such that a triangular
wedge of epiblast is left as the limit of the ganglionic forma-
tion. The poiut of this wedge, which exists in much the same
form for a considerably later period of development, projects
towards the mesoblast. It is somewhat difficult to describe
these appearances in words ; a glance at figs. 2, 14, 38 suffice,
I think, to make clear the meaning of the Zwischenstrang of
His, for that is what this portion of indifferent epiblast really
is. Let us follow the lateral epiblast upwards from the side
of the trunk to the lips of the neural plate. At first it is for
some distance neuralwards only one layer thick ; soon this
changes, and it becomes gradually thicker ; and if we follow it
in such a section as is figured in figs. 14, 38, we see that at

MORPHOLOGICAL STUDIES.

167

some little distance from the infolding neural plate it attains
its maximum thickness (leaving the neural plate itself out of
question as part of the epiblast). Beyond this point it abruptly
becomes one- layered again, and remains one-layered till it ends
also abruptly in the neural plate. The region of this one-
layered epiblast is that from which the ganglionic Anlage has
been cut out. The point of maximum thickness is that portion
of epiblast which has just failed to take any share in the forma-
tion of the ganglion. This point was one which gave me a
good deal of trouble in the course of the researches, but the
explanation of it gave the key to the origin of the ganglionic
Anlagen. In fact the first rudiments of the ganglia are formed
from the deeper layers of the epiblast just outside the limits of
the neural plate.

These stages in the formation of the spinal ganglia have
never yet been seen or figured by any observer.

The involution of the neural plate now begins to take place
very rapidly (fig. 15). Along with it the ganglionic Anlagen
get carried upwards. It seems as though they had not time to
get out of the way of the infolding process, and in missing the
chance to get out before the involution begins they are bound,
on account of pure mechanical processes — the explanation and
description of which I leave over to others — to follow the
neural plate, and thus they come to a somewhat abnormal
position at and between the dorsal lips of the neural plate.
The steps of this process are shown successively in figs. 14, 15,
5, 9, 21, 32.

Some figures of the head region (Nos. 38 — 42) are given
under high magnification, and tell their own tale in justification
of my statements of the very early appearance of the ganglia as
epiblastic proliferations and their independence of the neural
plate. They lie close to the latter, but can no more be
regarded as outgrowths of it than any other two organs which
lie close to each other in development can be considered, for
that reason, as derivatives one of the other. I have nearly
always been able, after the first traces of the ganglia were
visible, to distinguish the lateral limits of the neural plate,

168

J. BEARD.

and I think the reader will also have no difficulty in doing this
in very many of the figures given in Pis. XVI, XVII, and
XVIII.

In such figures as figs. 21, 33 — 36, 49, 50, 52, one sees that
the lips of the neural plate are very sharply defined. This
appearance was one which struck me as remarkable in the very
beginning of the investigations, all the more as till now no ob-
server seemed to have noticed it, and, so far as I am aware,
there is only one figure of it in existence, pi. xvii, fig. 12, in Pro-
fessor His’s paper on the peripheral nervous system (No. 29).
This figure also is taken from the spinal region of an Elasmo-
branch embryo, and tallies almost exactly with my figs. 22
and 32. Professor His, though he long ago noticed the ap-
pearance, incorrectly interpreted it, and attached no particular
importance to it. I shall refer to it again in reviewing the
work of previous observers.

To me it was the key to the origin of the ganglionic Anlage,
for it showed me unmistakably that this Anlage was not, as all
authors except His had supposed, an outgrowth of the spinal
cord. The identification of this sharp line of division, however,
was by no means a solution of the problem, for it was now a
question of where the ganglionic Anlage really arises. The thin
one-layered epiblast above the lips of the neural canal when
contrasted with the many-layered epiblast in the region of His’s
Zwischenstrang, suggests at once a possible point of origin ; but
in Elasmobranchs at first no proof of this could be found, and so
I had to look further back in earlier developmental stages before
the neural plate is involuted. The results of this search are given
above, and indeed it bears out my statement that this thin-
layered epiblast above the neural lips is really the point from
which the ganglionic Anlage has taken its origin. There is no
need to demonstrate, by means of mathematical formulce, &c.,
that the one-layered epiblast has during the involution of the
neural plate undergone a good deal of tension, — a tension
which no doubt helps to separate the ganglionic Anlage on
each side from the epiblast.

We have now arrived at a stage such as is figured in figs.

MORPHOLOGICAL STUDIES.

169

32 — 36, 21, 13, 17, 43 — 45, where the Anlage of the ganglia
sits upon and between the dorsal lips of the neural plate and
prevents the closure and fusion of the latter.

The next step is the further proliferation and removal of the
Anlagen to the sides of the spinal cord. In the words of most
authors, we have now got to Marshall’s neural ridge or crest,
and the Anlagen “ begin to grow out of the spinal cord”
(Marshall, Nos. 45, 46).

I think that in the preceding lines I have shown clearly
enough that there is really no outgrowth from the spinal cord,
nor do I find myself in the position to support Marshall’s view
of the origin of the ganglia from a neural ridge. From the
time of their first formation the ganglionic Anlagen appear to me
to be segmented, and if the Anlage of one segment passes over
into that of another in this and some of the following stages, I can
see in this no reason for saying that the spinal ganglia arise from
an uusegmented ridge of cells. No one attaches great morpho-
logical importance to the origin of the muscle-somites from an
apparently unsegmented structure, an origin which is condi-
tioned by the mode of formation of the cell elements, and as
all traces of such a continuous structure soon disappear, the
whole of its elements passing over into the various spinal
ganglia without leaving any permanent “ commissure,” I
must hold that if we are to say that the spinal ganglia are
outgrowths of a neural ridge, we must not forget two things :
that the outgrowths begin as epiblastic buds long before the
neural ridge stage ; and secondly, that the whole of the neural
ridge is absorbed by the various spinal ganglia. This latter
point is all the more necessary seeing that at least one
observer has suggested as an hypothesis worthy of proof the
origin of the lateral nerve of fishes from the remains of the
neural ridge. M. Julin says (No. 39, p. 31), “ Dans mon idee
le nerf lateral, tel qu’il se trouve constitue cliez
l’Ammocoetes ne serait que le reste de la Crete
neurale, ce qui expliquerait ses rapports avec les
racines du vague et les branches dorsales des nerfs
spinaux dorsaux.” He is further of opinion that this

170

J. BEARD.

avowed hypothesis “meriterait d’etre soumise a un controle
rigoureux, par des recherches embryogeniques.” Iu face of
the known facts concerning the morphology of the lateral
nerve as detected by van Wijhe and myself, we may regard
with perfect indifference M. Julin’s researches in the direction
of the above hypothesis. Such researches will turn out to be
neither more nor less than a wild-goose chase.

In my statements that the whole of the so-called “ neural
ridge ” passes over into the ganglionic formations, I agree
completely with Sagemehl, Onodi, and His. Balfour and
Marshall held different opinions which now can no longer be
maintained.

The ganglionic Anlage — now that it lies on the lips of the
neural plate, seems often to possess an unpaired character, but
from its subsequent fate, and from the appearances presented
in such figures as my figs. 21 and 29, it must really be con-
sidered as a paired structure, a point of the truth of which the
lateral origin of the Anlagen is sufficient evidence. The
Anlagen now begin to grow out from their position above and
between the lips of the neural plate (figs. 22, 23, 33 — 36), and
wander to their permanent position at the lateral portion of
the spinal cord (figs. 8 and 30).

We are now treading on ground which has been fairly
worked, but there are none the less one or two points of extreme
importance to which a few words must be devoted. As we have
seen, until now the ganglionic Anlagen have really no attach-
ment to the spinal cord ; and during the growth downwards to
its lateral side they are, as Sagemehl (No. 56, p. 30) first showed,
quite independent of that structure (fig. 57). The difference
between Sagemehl’s account and my own is obvious when we
remember that he, along with most other observers, regarded
the ganglia as outgrowths of the spinal cord. I maintain — and
the figures prove the truth of my view — that neither now nor
before are the ganglia attached to the spinal cord. The first
and only attachment to the spinal cord is the permanent one
now soon to be acquired. Before the attachment takes place
the ganglionic Anlage of each side divides into two portions.

MORPHOLOGICAL STUDIES.

171

the definite spinal ganglion and the sympathetic ganglion
(fig. 59). The latter will not concern us here, though on this
system of ganglia I shall later on have more to say. The
development of the sympathetic has been worked by Balfour
(Nos. 1 and 2), and more fully by Onodi (No. 52).

Arrived at the lateral surface of the spinal cord (figs. 28, 31,
61) the attachment1 to the latter takes place. On this point,
which in my opinion it is difficult to decide for the spinal
ganglia by direct observation, differences of view also obtain.
Sagemehl (No. 56, p. 31) and others hold that the con-
necting fibres grow out from the spinal cord, while Professor
His (No. 34, p. 373), with great confidence, says, “Die sen-
sibeln Nerven, der N. acustieus, und die Ge-
schmacksnerven entspringen in dem Ganglion und
sie wachsen mit ihren centralen Wurzeln in das
Ruckenmark und in das Gehirn herein. Diesen Satz,
den ich friiher nur indirect zu stiitzen vermocht hatte, vermag
ich nun mit grosser Sicherheit zu beweisen.”

His was the first to apply the physiological laws of the
trophic properties of the ganglia to the solution of this
question. He says (No. 29, p. 477), “ Die Frage, ob die
hinteren Wurzeln vom Ruckenmark aus nach den Ganglien
hin wachsen, oder von den Ganglien aus nach dem Rucken-
mark, ist bis jetzt noch ziemlich unerortet geblieben. In
meinem Augen spricht das Uebergewicht der Griinde fur die
letztere Alternative. Als einen dieser Griinde betrachte ich die
durch Waller und durch Cl. Bernard (No. 12) nachgewiesene
Trophische Abhangigkeit der hinteren Wurzeln vom Ganglion.
Bei Durchschneidung der hinteren Wurzeln zwischen Rucken-
mark und Ganglion degenerirt nach den Ergebnissen jeuer
Forscher der mit dem Ruckenmark in Yerbindung stehende
Stumpf; der mit dem Ganglion verbundene bleibt intakt
(No. 12, Bd. I, p. 237). Jede Zelle eines Ganglions nimmt
namlich zunachst eine spindelformige Gestalt an, dann aber
wachst sie in zwei Fasern aus, die nach entgegen-gesetzten

1 Marshall (Nos. 46, 49) held this to be a secondary attachment. It is the
first and only connection with the central organ.

172

J. BEARD.

Richtungen vom Zellkorper abgehen, &c.” I am inclined,
and was so before reading these words, to agree with Professor
His in his conclusions that the growth is a centripetal one ;
indeed, as Dr. Hill (‘Three Lectures/ p. 3, No. 27) also has in-
sisted, the matter is one of which the physiologists have already
furnished the solution. He says, “ It appears probable that
the fibres of the posterior roots also grow from the cells of the
ganglion centralwards into the cord, instead of from the cord
to the ganglion as usually supposed. A consideration of the
effects of cutting nerves in such cases as have been hitherto
described leads me to formulate the law that nerve- fibres
die when cut off from the cells of which they are
processes, and from which they derive their nutrient
supply. It is well known that, when the posterior roots are
cut, the fibres which remain attached to the root-ganglia
live ; those entering the cord die.”

I will not cast any doubt on Professor His’s very positive
statements on this point, indeed, I believe they represent the
facts of the case, but I must again say that the question is
difficult, if not impossible, to decide for the spinal ganglia by
direct observation, and for myself, I must admit that I have
not been able to make more of it as yet.1

The mode in which the connection between the ganglia and
their peripheral end organs takes place, is one on which there is
also much dispute. For His (No. 33, p. 375) and Kolliker (Nos.
42, 44) hold, as against all other observers, that the nerves
are processes of the ganglionic cells, without any intervention of
ganglionic cells or ganglionic cell nuclei in the course of the
nerve. I shall have occasion to discuss this question more fully
in connection with the anterior roots of spinal and cranial
nerves, and content myself here with the remark that I believe
Professor His’s and Kolliker’s conclusions cannot be maintained,
and that the peripheral connection in the case of sensory or

1 As an absolute maxim I am only inclined to support this as regards the
spinal ganglia ; in the case of the cranial ganglia, as we shall see, there are
reasons for holding this view only with regard to the sensory part of the
root.

MORPHOLOGICAL STUDIES.

173

motor nerves is brought about rather by a chain of ganglionic
cells.

I, b. The Cranial Ganglia in Elasmobranchii.

Compared with the development of the cranial ganglia that
of the spinal ganglia previously described is simplicity itself.
For just as the head of Vertebrates presents, when compared
with the trunk, a complexity of problems, the solution of which,
in the opinion of morphologists like Huxley, Dohrn, Froriep,
and others, will take years of careful work, so also the cranial
ganglia present a number of problems, towards the solution
of which I only can hope to go a little way in the following
pages.

While there can be no sort of doubt that the spinal ganglia
are strictly segmental in their origin — indeed, that such is the
case is easily demonstrable — the reduction of the cranial
ganglia to segmental order is a task of great difficulty. It
has been objected by Dohrn and others that the setting up of
tables showing the segmental nature of the head nerves is a
proceeding which is to be deprecated, and that the true
problem is the reduction of the components of the head to
simpler Aunelidan structures. To which one may reply that,
according to Dohrn, such Annelid ancestors were segmented
animals, and no matter how complex the Vertebrate head may
now be, it is at its basis composed of a number of Annelid
metameres, and the unravelling and ordering of the existing
complex, as far as it is possible, is the real task of the
morphologist.

I shall not at the moment attempt to discuss again the
claims of the various cranial nerves to “segmental rank/’ a
proceeding which, to my mind, is entirely justifiable, for it has
its meaning in the sorting of the cranial nerves for morpho-
logical (and physiological) considerations. Still, in the follow-
ing account of the very first signs of the cranial ganglia, I
must insist on such points of development as support, for
instance, the comparison of the auditory and olfactory ganglia
and sense organs with those of, for example, such a typical

174

J. BEARD.

cranial nerve, its ganglion, &c., as the glossopharyngeus.
This is all the more in place, as Professor His, in his recent
note of warning against the speculations of us unfortunate
younger morphologists, does not hesitate to maintain as a
fact the derivation of the auditory and olfactory organs
from what he calls the “ ganglion Leiste,” which also gives
origin to such ganglia as facial, glossopharyngeus, &c. I
hope to show to Professor His’s satisfaction that this “ fact ”
is as little a fact as his derivation of the spinal ganglia from
the “ Zwischenstrang,” which is the continuation backwards
of the “ ganglion Leiste ” of the head.

A further complication is presented by the superaddition of
the sense organs of the head (and their ganglia), excepting the
eye, which all enter into relationships with those portions of
the head ganglia which appear morphologically to correspond
to the spinal ganglia. These complications will be more clearly
explained in the course of the work.

I have mentioned in a recent paper (No. 8) that the cranial
ganglia are made up of more form elements than the
spinal, and I observe that Professor Gegenbaur, without
investigating the development, comes to the same conclusion
(No. 21).

The first traces of the cranial ganglia Anlagen are formed
in exactly the same fashion as those of the spinal ganglia, and
it is much easier, on account of their greater distinctness, to
make out the earliest stages. In the embryo in which I
described the first traces of the spinal ganglia such Anlagen
can also be distinguished in the head region. As the meso-
blast has not yet divided up into the body-somites, or so-called
protovertebra?, the head-somites are also not formed, and so
we are entitled to say generally, the traces of the posterior
root ganglia of cranial and spinal nerves are formed very early
and long before the closure of the neural plate.

A figure through the head region of an embryo, as early as
the one depicted in fig. 4, has been given by Professor Marshall
in one of his papers (No. 48, fig. 1), but he gave no trace of
any ganglionic formation, and, indeed, it is quite possible that

MORPHOLOGICAL STUDIES.

175

such traces were not differentiated in the section from which
he figured. In his monograph of the development of Elasmo-
branch fishes, Balfour has also given, on PI. IX, many figures
of stages corresponding to those on my PI. XYI ; but here again
no trace of the ganglionic Anlage, which is seen in all my
figures, has been represented.

In fact, of the cranial, just as of the spinal ganglia, no
observer has hitherto seen the very first stages which I am
about to describe, and the last observer, Onodi (No. 51), who
has given no figures at all, has, judging from his description,
only seen the Anlagen in much later stages, and, as we shall
afterwards see, has not interpreted rightly or seen all that is
to be seen in fairly decent sections.

Returning to fig. 4, we find, on examination, the same
appearances {g. a.) as were met with in the developing spinal
ganglion. If we examined an earlier stage than this we should
meet with no trace of the Anlagen of the cranial ganglia. We
see now a central portion which represents the brain part of
the neural plate in section. At each side of this, but inde-
pendent of it, one notices the budding out and separation, so
far as the lateral epiblast is concerned, of a process which is,
as we shall see, the first trace of a cranial ganglion, or rather
of part of one. Soon after this phase the involution of the
neural plate begins just as in the case of the spinal cord, and
along with the involution the ganglionic Anlagen are also
carried upwards. I have figured these stages in figs. 6, 10 — 12,
16, 19, 20, 39 — 43, taken from various parts of the brain, in
order to show that this mode of development holds for portions
of the olfactory ganglion (figs. 19, 20), mesocephalic or gan-
glion of the ophthalmicus profundus (No. 7), trigeminus,
facialis, auditory, glossopharyngeus, and vagus.

The involution of the neural plate, on its completion,
encloses the cranial ganglionic Anlage just as occurs in the
spinal cord. A number of figures of this stage are given in
figs. 1 8, 21 , 24, 25, 29, 45, 48, g. a.

The Anlage is now separated from the skin, and in the head
of Elasmobranchs no trace of a Zwischenstrang is left behind.

176

J. BEARD.

If in such stages it is difficult in the trunk to be always
quite certain of the sharp boundary line separating the gan-
glionic Anlagen from the closing neural plate, such is never the
case in the head. I cannot remember having seen a single Elas-
mobranch section in which for the head it was at all a difficult
matter to distinguish the limits of the two; and in spite of
this fact there are no figures in existence which show this
separation such as I depict it in figs. 44 — 48, 24 — 27, 29.
Here, as in the trunk, the position of the ganglionic Anlagen
between the lips of the neural tube (figs. 25, 44, 47) prevents
their complete closure. But soon the Anlagen begins to grow
downwards and outwards towards the lateral surface of the
body. This outward growth leads, as is well known from the
researches of recent years, to a difference in position between
the ganglia of the head and those of the trunk. For while
the latter lie between the muscle-plates and the spinal cord,
the former take up a position outside the mesobiast and close
to the skin.

The portion of the ganglionic Anlagen of the head derived
from the neural epiblast corresponds, in development at least,
with the Anlagen of the spinal ganglia, but the cranial ganglia
of (apparently) all Vertebrates acquire a further form-element
derived from the lateral epiblast above the gill-clefts, and at
about the level of the notochord. For the formation of this
element I have not in this paper given any figures, but I
think such figures can be here entirely dispensed with, seeing
that in a former paper (No. 6) treating of the branchial sense
organs and their ganglia I figured a great many stages of this
ganglionic formation, for, what I there called the branchial
ganglia make up this additional form-element of which I just
wrote. I believe I showed conclusively enough in that paper
that above the gill-cleft ganglionic elements were given off
into the main ganglion — indeed, it then seemed to me that
most, if not all, the ganglion was formed there. As even such
a severe critic as Professor Gegenbaur expresses himself
satisfied that such form-elements of the ganglion take their
origin above the gill-cleft, I may assume it to be unnecessary

MORPHOLOGICAL STUDIES.

177

to give a very detailed account of such formation in individual
cases. From Professor Gegenbaur one must apparently be
thankful for small mercies, and as this is the one thing in
my researches which he admits unreservedly that I have seen,

I quote his testimony in my favour. He says (No. 21, p. 41),
“ Die Beziehung des Ganglions zu dem Ektoderm ist von
Beard richtig erkannt worden : er sagt, ‘The proliferated cells
form a mass of actively dividing elements still connected with

the skin and fused with the dorsal root; for some

time the cells continue to be given off, and of those already
given off many show nuclear figures/ Die epitheliale Ver-
dickung hat also die Bedeutung einer Quelle der Ganglien-
bildung. Das geht auch aus den beziiglichen Figuren Beard’s
hervor, die zudem in der Anordnung der Elemente der
am Ganglion befindlichen Ektodermschichte gar
nichts aufweisen, was man auf ein liier sich bil-
dendes Sinnesorgan beziehen konnte. Wenn die That-
sachen, wie sie in Wirklichkeit bestehen, die Grundlage der
Forschung abgeben, so kann man hier nur sageu ; der Nerv
wachst vom Centralorgane aus unter dem Ektoderm bis zu
einer Stelle, an der ihm aus dem Ektoderm ein Zufluss von
Formelementen zu Theil wird.”

For the present moment I leave entirely alone Professor
Gegenbaur’s doubts about the sense organs. Such doubts are
entirely unjustifiable. To return to the ganglionic Anlagen
derived from the epiblast at the neural side of the head.
These Anlagen grow outwards and downwards towards the
lateral surface of the body. Just above the gill-cleft there
is here a small portion of neuro-epithelium (figs. 94, 95),
which is the Anlage of the branchial sense organs or lateral
sense organs. This neuro-epithelium has begun to extend its
growth before the ganglionic Aulage fuses with it.1 In figs.
94, a and 95, a, 1 have represented this. The growth has

1 Fig. 101 shows this growth for the auditory epithelium of a lizard. Just
as all the lateral sense-organs are formed from a certain limited number of
pieces of neuro-epithelium, so all the sensory cells of the ear arise from the
extension of one little bit of neuro-epithelium ( o . e.)

178

J. BEARD.

already extended behind the gill-cleft (fig. 95, 5), and also
in front of the gill-cleft (fig. 94, a). In connection with
the morphology of nose and ear this point is one of con-
siderable importance, and I shall have occasion to refer
to it again. The ganglionic Anlage now fuses with neuro-
epithelium at one point. At the point of fusion a prolifera-
tion of the cells of the neuro-epithelium takes place into the
ganglionic Anlage. The proliferated cells form a mass of
actively dividing elements still connected with the skin, and
fused with the rest of the ganglionic Anlage. Externally to
this thickening is situate what Professor Eroriep (No. 17) and
I regard as the primitive branchial sense organ of this segment.
Here again I refrain from discussing any of the questions con-
nected with the formation and morphology of these sense
organs.

One fact at least holds as the result of this skin fusion, and
this is that a number of form-elements are given off into the
ganglion. The recognition of this fact does not come to me
after reading Professor Gegenbaur’s paper quoted above. I
had long before seeing that written as a note, on p. 21 of my
paper (No. 8), on the “ Old Mouth and the New/’ “The
cranial ganglia of Vertebrates are far more complicated mor-
phologically than has hitherto been recognised. In addition
to parts which appear to correspond morphologically to the
posterior root ganglia of the spinal nerves plus the sym-
pathetic ganglia, they also contain the special ganglia which
are formed in connection with the gill sense organs.”

The ganglion complex soon begins to leave the skin,
and in doing so a number of cellular fibrous cords are left
behind connecting the sensory epithelium with the ganglion
complex.

The sensory epithelium has, briefly stated, usually grown
in three directions in front of the gill-cleft, behind the gill-
cleft, and above the gill-cleft, either in a neural, or a forward,
or a backward direction. The nerves connecting these various
sensory elements with the ganglion appear to me to be all
derived as splittings off from the inner layer of the sensory epi-

MORPHOLOGICAL STUDIES.

179

thelium. This view, which I formerly only mentioned as true
for the suprabrauchial nerves, I must now also extend to the
prsebranchial and the sensory part of each postbranchial nerve.

The ganglion complex has also to acquire its first and per-
manent connection with the central nervous system, and of the
mode in which this takes place there can be no sort of doubt.
The main trunk of the nerve and its connection with the cen-
tral nervous system are formed respectively by the formation
of nerve-chains from some of the ganglion-cells, and by the
growth of fibres into the central nervous system.

Here again, however, the cranial ganglia present us with
complications as compared with the spinal.

It is well known that the whole of the motor fibres of the
spinal nerves (those to the voluntary and the visceral muscles
derived respectively from the anterior and lateral horns) pass
out in the auterior roots. Now, there can be no doubt
that the whole of the anterior root of a spinal nerve is a
direct outgrowth from the central nervous system. All
observers are agreed on this point. Quite other conditions
obtain in the head. In the oculomotorius, trochlearis, and
abducens, the only nerves which are comparable at all to ante-
rior roots of spinal nerves, no fibres are derived from the homo-
logue in the head of the lateral horn of the spinal cord ; in
other words, the anterior roots of the head give no fibres to
visceral muscles, and — a fact which is well known — the fibres
to the visceral muscles of the head pass out with the posterior
roots of the cranial nerves. It appears also that these fibres
take their origin in the continuation of the lateral horn in
the head. This being so, and it being also true that all other
motor nerves, including those of the spinal cord and the three
eye-muscle nerves, certainly occur as outgrowths of ganglia1
situated within the central nervous system, it becomes a ques-
tion whether the motor fibres of the gill-cleft muscles are not
also direct outgrowths of the central nervous system. I must
confess that I have not as yet been able to settle this point by

* See fig. 100, which depicts the third and its central ganglionic origin as
seen in Lacerta agilis.

180

J. BEARD.

direct observation, but I do not hesitate for a moment in ex-
pressing the opinion that such is the case. If this be true it
follows that a typical posterior root of a cranial nerve, that is,
a root passing to a gill-cleft, is composed of elements derived
from at least three sources : there is, firstly, the portion which
corresponds to the true spinal ganglion1 in its derivation from
the epiblast just outside the neural plate (neural ganglion) ;
secondly, a portion formed in connection with the branchial
sense organs (lateral ganglion) ; and thirdly, a portion derived
from the continuation of the lateral column in the head.

In addition, a part must he added comparable to the sym-
pathetic ganglia of the trunk, and this portion is probably, as
occurs in the case of spinal ganglia, contained in the portion
of the ganglion derived from the epiblast, just outside the
neural plate.

Here we are faced by some interesting problems, which I
will afterwards discuss.

His (No. 34, p. 394) and others have raised objections to
the view of Balfour, that the cells derived from the neural
ridge or crust are the Anlage of the posterior roots of the
cranial nerves. Balfour’s phraseology has been used by Mar-
shall, Spencer, van Wijhe, myself, and others. No doubt
objections may be urged against the use of this phraseology as
accurately representing all the facts.

While I admit that these cells are more a ganglionic Anlage
than that of a nerve, there are two points which must be urged
in extenuation of the offence, if offence it be. In the first
place, Balfour, Marshall, van Wijhe, and all of us who have
used this phraseology, have done so, in the sense of the inclu-
sion in the term posterior root, both root and ganglion of the
nerve; and secondly, in the head at any rate, in addition to
the cell processes which grow from the ganglion into the cen-
tral organ, it can be demonstrated (fig. 103) that some of the
cells of the ganglionic Anlage pass over iuto the root of the
nerve, and take a direct share in its formation. This question

1 This portion of the cranial ganglion is possibly only morphologically
an equivalent to the sympathetic part of a spinal ganglion.

MORPHOLOGICAL STUDIES.

181

of the formation of nerves is one on which, along with most
observers, I am completely at variance with His (Nos. 29 and
34) and Kolliker1 (Nos. 42 and 44) ; and I refer the reader to
a fuller discussion of it in another part of this paper.

With this I close my account of the very earliest stages of
the cranial ganglia in Elasmobranchii. The mode of develop-
ment here described from at least two sources is characteristic
for the ganglia of all the branchial nerves, facial, glossopha-
ryngeus and vagus (fig. 24), and also for the trigeminus.
Mutatis mutandis it also holds for the olfactory (figs. 19,
20), mesocephalic, and auditory (figs. 25, 29) ganglia. Here
I will only emphasize this point, reserving to myself the
right to return to it on a subsequent occasion. To go further
into the matter here would lead to the discussion of a great
many disputed points, and for the moment I wish to lay more
stress on the absolute facts of the development which can be
demonstrated. The conclusions which I feel entitled to draw
from those facts can for the moment be postponed.

II. The Peripheral Nervous System of the Chick.

Our knowledge of the development of the peripheral
nervous system in Birds is almost entirely due to His
(No. 29), Marshall (No. 46), and Onodi (No. 51). Kastschenko
(No. 40) has also contributed his item, which, so far as
nerves or ganglia are concerned, is of no particular value,
for it contains no new facts and throws no new light on
the morphology of either cranial or spinal nerves and
ganglia.

For many reasons I was obliged to include the Chick in the
sphere of my observations. For one thing His’s remarkable
observations and hypotheses were mainly established for this
animal (Nos. 28, 31), and I could not feel satisfied until the
explanation of His’s Zwischenstrang was got at the bottom of.
The striking manner in which the epiblastic origin of the

1 Kdlliker lias upheld his views in several papers.

VOL. XXIX, PART 2. NEW SER.

N

182

J. BEAKD.

ganglia in the head of the Chick attracts the attention in good
sections was also a reason for fully investigating the develop-
ment in this animal; for the question naturally arises, Are these
appearances primitive, or is the development modified in some
way or other in the Chick? One could hardly hope to maintain,
as a strict morphologist would be almost bound to do, if he had
only Onodi’s researches to go upon, that the mode of develop-
ment of the cranial ganglia in the Chick is a more primitive
one than that in Sharks. The facts, which I had discovered
before seeing Onodi’s paper, were at first a great puzzle to me,
a puzzle to which Onodi’s researches have given no solution ;
for, according to him, and so far he agrees with Marshall
(No. 46), the cranial ganglia of the Chick differ entirely in
mode of development from the ganglia, cranial and spinal, of all
the other forms. Sharks, Lizards, Mammals, &c., which he had
examined. Indeed, he maintains — and I find this attitude a
surprising one in the man who had seen the true development
in the cranial ganglia of the Chick — he maintains that in all
other cases the ganglia, both cranial and spinal, are developed
as outgrowths of a ganglion ridge (neural ridge of Marshall),
and this in its turn owes its origin to the central nervous
system.

Seeing that my researches on the cranial ganglia of the
Chick are partly a confirmation of Onodi’s, it might be sup-
posed that there was no necessity for giving them in detail.
However, I am of a different opinion, for they do not agree
with Onodi’s results on all points, and on the fundamental
question whether the ganglionic Anlagen of the head are or
are not parts of the central nervous system, Onodi says nothing.
As he holds that in all other cases the ganglia, spinal and
cranial, are outgrowths of the central nervous system, his
position as a comparative embryologist is not a very logical
one. The first traces of the ganglia, both cranial and spinal, are
met with in the Chick between the twenty-second and twenty-
sixth hours of incubation. In such embryos there are on the
average from two to ten body-somites or protovertebrae, and
it is in such embryos that evidence of the epiblastic origin

MORPHOLOGICAL STUDIES.

183

of both cranial and spinal ganglia can nearly always be ob-
tained.

The mode of preparation, which in Sharks is not of such
importance, is here a very weighty factor. My embryos were
all prepared by immersion for from half an hour to two hours in
Flemming’s chromic-osmic-acetic acid mixture, and afterwards
stained with borax carmine or picro-borax carmine. Osmic
acid must be used here, and used very carefully, or otherwise
no guarantee can be given that all the appearances depicted by
me in Plates XIX, XX, XXI will be visible. Thin sections are
of course also of importance, and I must express the opinion
that the results obtained by Professor His ten years ago (No. 29)
are vitiated by improper treatment of the embryos and by the
thickness — at that time unavoidable — of the sections. My
sections are mostly mm- thick. As was the case in the
account of Elasmobranchs, I shall begin this part of the paper
also with the

II, a. Development of the Spinal Ganglia in the

Chick.

The appearances about to be described may be even seen some-
times in embryos in which no body-somites are as yet formed,
and, speaking generally, an embryo with about six body-somites
will show in different regions the appearances presented in
seven sections (figs. 70 — 76) taken from the spinal region of
such an embryo with six mesoblastic somites. It will be noticed
that the medullary canal is everywhere open, and, in fact, here,
a9 in Sharks, the first traces of the cranial and spinal ganglia
are formed long before the closure of the neural plate. The
first section is in the region of the primitive streak — and here
no trace of ganglionic Anlagen is to be seen (fig. 70). The
next section (fig. 71) is taken much farther forwards, and on
the left side of the section, at any rate, the commencement of
the ganglionic differentiation (fig. 72, g. a.) can be seen. The
third section (fig 72) passes through the middle of a meso-

184

J. BEARD.

blastic somite on the right — and here it is difficult to distinguish
a ganglionic Anlage — but on the left it has cut the segment
near its end and the ganglionic Anlagen is distinctly seen at
g. a. as a small plug of cells being cut out of the epiblast just
outside of the neural plate.

The fourth section (fig. 73) shows very distinctly on both
side the spinal ganglionic Anlagen (g. a.). The boundaries of the
infolding spinal cord are sharply marked off at o from the gan-
glionic Anlagen, which lies just outside them at g.a. This figure
shows that the spinal ganglia in the Chick take their origin
in exactly the same way as those of Elasmobranchii, and that
by the cutting out of the ganglionic Anlagen at g. the epiblast
to the outer sides of them is left as a somewhat triangular body,
which, as in Sharks, where it is not so distinct, represents the
“ Zwischenstrang” of His. The following three figures, taken
still farther forwards from the embryo, show the same appear-
ances even better. On examining, for example, the six figures
(fig. 75, 76, 80, 81, 85, 89), one sees the following things: — The
spinal cord is rapidly closing in, and its lips are sharply defined
from the ganglionic Anlagen ( g . a.) just outside them. With
the lateral epiblast (<?.) the Anlagen of the ganglia have lost all
connection, and only retain it with the epiblast at the re-entering
angle between spinal cord and skin. The epiblast at ie, of which
the Anlagen, as in figs. 70, 71, originally formed a part, is, ever
since the separation, only composed of a single layer of cells,
which, in consequence of the tensions arising in connection with
the infolding process, has become much lengthened. Outside
this thin-layered part the epiblast passes almost abruptly
into a much thicker-layered portion of the outer layer, which
has taken no share at all in the formation of the ganglia.
This thicker portion (fig. 68) forms a somewhat triangular
mass of cells, the apex of which is directed towards the meso-
blast. From a consideration of these figures (figs. 75, 76, 80,
81, 85, 89,), and of several other figures (figs. 98, 99, 102),
to be afterwards noticed more in detail, the conclusion is forced
upon any unbiassed observer that the triangular mass of cells
is identical with the so-called “ Zwischenstrang” of llis. The

MORPHOLOGICAL STUDIES.

185

consequences of this conclusion are fatal to the observations
which His recorded nearly ten years ago, and to which he has
been true for nearly twenty years.

The next stages in the formation of the spinal ganglia
which we need consider are represented in Pis. XVIII, XIX,
and XX, figs. 51, 55, 98, 99.

The separation of the neural tube from the epiblast has now
taken place, and the ganglionic Anlagen have also no longer
any connection with the epiblast. The spinal cord has not yet
really closed, for its lips have not met, and in all the figures
they are sharply defined from the fused ganglionic Anlagen
which lie between and above them. In all the figures, but
especially in 55, 98, 99, the Zwischenstrang of His ( z .) is a
prominent object. It is represented under very high power in
figs. 98 and 99. As in Elasmobranchs, the ganglionic Anlagen
now begin to grow down the sides of the cord, leaving their
position above the lips of the neural tube ; a stage of this
process, showing that they are still unconnected with the cord,
is given in fig. 102. When the Anlagen leave the lips of the
neural tube the latter close together, and all subsequent stages
go on exactly as described in Elasmobranchs. As I can here
record no new facts I leave the development of the spinal
ganglia of the Chick at this stage. I have proved at least three
things for the spinal ganglia of the Chick. (1) That they
are direct epiblastic proliferations formed very early outside
the limits of the spinal cord Anlagen ; (2) that there is no out-
growth of cells from the spinal cord to form them ; (3) that the
Zwischenstrang of His is that part of the epiblast which just
fails to play any part in the formation of the ganglia. A
fourth conclusion may be drawn, and this also holds for Elas-
mobranchs and other forms. It is that there is no form element
in the spinal ganglia corresponding to that portion of the
cranial ganglia which is derived from the sensory epithelium
of lateral sense organs.

] 86

J. BEARD.

II, b. The Cranial Ganglia in the Chick.

As we have seen, the development of the cranial ganglia in
the Chick was described by Marshall (No. 46) just ten years
ago. He says, p. 15, “ About the twenty-second hour a small
outgrowth of cells appears along the mid-brain on each side, at
the angle between the external epiblast and the neural canal —
the neural ridge. This rapidly extends both forwards and
backwards ; forwards as far as the anterior part of the optic
vesicles ; backwards, along the whole length of the brain, and a
certain distance down the spinal cord. Its first appearance
precedes the closure of the neural canal.” And on p. 12 (1),
“ The neural ridge appears before closure of the neural canal
is effected, so that the ridges of the two sides are primitively
independent of each other.” (2) “ The ridge is not developed
directly from the external epiblast or from the neural canal,
hut from the re-entering angle between the two.”

His (No. 29) has also given some partially correct figures of
the first origin of the cranial ganglia in the Chick (Taf. xvii,
fig. 3, a — /). The remaining figures of the series g and h are,
I think, not correct; and the interpretation put on the (only
partially correct) figures by His is one which, along with
Balfour, I cannot accept.

His himself (No. 29, pp. 464 — 465) summarises his conclu-
sions as follows :

“ Ich halte dafiir, dass das Gebilde, welches ich beim
Hiihnchen Zwisclienstrang genaunt babe, weder eine'AVuch-
erung ’ des Medullarrohres, noch einer solclien des Horn-
blattes seinen Ursprung verdankt, sondern einem zwischen
diesen gelegenen besonderen Substanzstreifen. Dieser Sub-
stanzstreifen grenzt sich schon vor Eintritt des Markschlusses
in mehr oder minder auffalliger Weise ab und so bildet er
eine Rhine, die von mir sogeu. Zwischenrinne.

“Nach meiner Ansicht gliedert sich demnach das obere
Grenzblatt oder Ectoderm in dreierlei Anlagen ; in die
Medullarplatte, in die beiden Zwischenstrang-platten und in

MORPHOLOGICAL STUDIES.

187

das Hornblatt. Nach volleudetem Riickeuschluss ist auch die
Zwischenstrang-Anlage voru Hornblatt uberdeckt,1 ihre
Stellung aber zur Medullardecke wechselt in den verschie-
denen Bezirken, indem sie in deni einen iiber, in den anderen
neben dieselben zu liegen kommt.

“ Beistehende schematisirte Abbildung, in welcher die
Zwischenstrang-masse weiss ausgespart ist, kann die Art des
Zustandekommens veranschaulichen. Es nimmt namlich der
abgeloste Zwischenstrang beim Hiibnchen nur am Kopfe und
zum Theil nur iiber dem Yorderhein und dem Mittelhirn die
iutermediare Stellung zwischen Medullarrohr und Hornblatte
ein, weiter hinten bildet er, so weit er nicht zu Anlage der
Gelibr-grube verwendet wird, zwei seitlicli von der Scbluss-
stelle liegende, auf dem Durclischnitt dreikautig erscheinende
Leisten. Yon diesen letzteren haben Balfour, Marshall, und
Kblliker referirt, ich sehe sie als ^uswiichse’ oder als
c Wucherungen ’ des Horublattes an, eine Ansicht, die ich in
keiuer Weise vertrete (!). Fiir mich faugt das Hornblatt erst
da an, wo die Ganglienaulage aufhdrt, und die Ganglienaulage
da, wo das Medullarrohr aufhbrt ; ich verwerfe uberdies, wie
ich schon an anderem Orte ansgesprochen habe (Briefe iiber
uusere Korperform, S. 67, u. f.) das von Manchen Embryologen
so freigebig benutzte Princip der lokalen Wucherungen.”

My account of the cranial ganglia of the Chick is in most
points in agreement with that given by Onodi (No. 51, p. 260),
to whom the reader may refer for a fuller account. In a
later paper (No. 52) Onodi himself describes his results in the
Chick briefly as follows (p. 553) : — “ Beim Hulin stammt das
Ganglion intervertebrale2 am Kopfe theils von der Zellenproli-
feration, theils von der gleichseitigen Abschniirung des der

1 The italics are mine.

2 When, as here and elsewhere in his papers on nerve development, Onodi
speaks of the cranial ganglia as “spinal ganglia of the head,” he is begging
entirely the question of the homology of the two sets of ganglia. There is
no discussion of the homology in any of his papers (see list at end of this
work), neither do his researches contain any facts which justify this baseless
assumption of their homology.

188

J. BEA11D.

Umbieguugsstelle iii das Gehirnrohr uaheliegendeu Ab-
schnittes des Ectoderma/’

As in Elasmobranclis the first traces of the cranial ganglia
Anlagen in the Chick are found in embryos in which as yet
no division of the mesoblast has taken place. Marshall’s
statement of the time of first appearance, viz. the twenty-
second hour, may be taken to be as nearly correct as one can
determine.

The neural tube is still quite widely open.

Figs. 86, 87, and 67 are sections through the head region of
three such embryos. In fig. 86 no trace of the ganglion
Anlagen is to be seen, but the epiblast is much thickened,
especially in the region of the future central nervous system
and ganglionic Anlagen.

In fig. 87 traces may be seen of the ganglionic Anlagen at
g. a., and one sees that they occupy practically the same posi-
tion in reference to the central nervous system as the spinal
ganglion Anlagen. In figs. 82, 83, and 84 are figured three
sections through different regions of the head of a Chick em-
bryo with seven body-somites. In all these sections the origin
of the ganglia is very easily distinguishable. In fig. 82, which
is through the region of the fore-brain, the limits of the neural
plate are already marked out, and outside this the ganglionic
Anlagen of each side are visible as special differentiations of
the epiblast of the re-entering angle between the neural canal
and the external epiblast, and also of a portion of the lateral
epiblast beyond this. One notices that here, as in the spinal
cord, the separation of the ganglionic Anlagen will cut out a
particular region of the inner epiblast, and will leave a struc-
ture at Z. which is comparable to, and, indeed, identical with,
the Zwischenstrang of His in the region of the spinal cord.
Fig. 83 is in the region of the mid- brain, while fig. 84 is
taken farther back from the hind-brain ; these two sections
present exactly the same appearances, and in a more marked
degree than fig. 82. Figs. 77 — 81 are taken through the hinder
head region of a Chick embryo with nine body-somites or so-
called protovertebrae. Here the development of the ganglia has

MORPHOLOGICAL STUDIES.

189

advanced a step farther. The boundaries of the central nervous
system are well marked in all the sections, and one can see
that the latter is not connected with the ganglionic Anlagen.
In this series of figures, which go backwards to the hind-brain,
the ganglia Aulagen are already separated from the epiblast
outside the re-entering angle between brain and epiblast, but
still connected with the epiblast of the re-entering angle. For
Professor His’s satisfaction it may be added that in all the
sections a Zwischenstrang is more or less developed. The next
stages are shown in figs. 97, 46, 63, 65, 66, which are taken
through embryos with, on the average, eight to ten mesoblast
somites. The central nervous system with the ganglionic Aula-
gen are now shut off from the outside by the meeting and
fusion of the external epiblast.

The lips of the neural tube itself have not yet closed,
because the ganglionic Anlagan, which in some of the figures
(figs. 97, 88, 46) have still a distinctly bilateral character,
still exist between and above them. The ganglionic Anlagen
soon become entirely separated from the epiblast, and in
doing this leave behind them the traces of the limits of their
epiblastic origin in the shape of a three-cornered ridge of
cells which is identical with His’s “Zwischenstrang”1 of the
spinal region. It is figured at Z. in figs. 88 and 89. By the
growth outwards from their position above and between the
lips of the neural canal, the latter is able to close, and does so
without having contributed a single cell to the formation of
the ganglia.

The ganglionic Anlagen now begin to grow downwards
towards the lateral surface of the brain to the point at which
they acquire their permanent and only roots of attachment.
The attachment takes place as in Elasmobranchs, and what I
said about these forms on this subject may be taken as holding
for the Chick also.

On the other hand, a portion of the “ Anlagen ” grows
towards the lateral epiblast at about the level of the noto-

1 There is no such structure in the head of the Chick or any other Verte-
brate I have as yet examined as His’s Zwischenrinne.

190

J. BEARD.

chord and fuses with it (figs. 90, 91, 92, 93). I have not
seen any reason for giving numerous figures of this point ;
those represented appear to me sufficient for the purpose. We
are now concerned with the rudiments of the branchial sense
organs. These structures were first described for Mammals
by Professor Froriep (No. 17), and very shortly after his dis-
covery my own preliminary researches independently announc-
ing almost the same facts for Elasmobranchs appeared (No. 5).
In the full account I afterwards gave of them I also mentioned
the finding of similar rudiments in three- days-Chick embryos
(figs. 90, 91, 92, 93). Beraueck (No. 10) afterwards con-
firmed the discovery, and Kastsclienko (No. 40) in the account
he more recently gave of them, with his characteristic failing,
conveniently forgot to mention that either Beraneck or I had
ever seen the structures which he described.1

However, the following conclusions regarding the importance
of the rudimentary branchial sense organs for the embryo
Chick are peculiarly my own.

We saw that in Elasmobranchii the fusion of the ganglionic
Anlagen with the neuro-epithelium which forms the Anlagen
of the branchial sense organs leads to a certain amount of
ganglion form-elements passing from the neuro-epithelium to
the main ganglion. Such portion is really the ganglion of the
branchial sense organs or sense organs of the lateral line
(figs. 92, 93) (lateral ganglion).

It can be demonstrated also for the Chick that such form-
elements pass by proliferation from the rudimentary sense
organ into the ganglion (figs. 92, 93). This being so, and it
being also capable of demonstration that the sense organs, with

1 As Herr Kastsclienko quotes my paper, I presume he was aware of the
discovery ; this conclusion is all the more justified as Professor Wiedersheim
also briefly mentions my discovery in the last edition of his ‘ Lehrbuch der
Vergleichenden Anatomie der Wirbelthiere ’ (p. 332). Professor Strasser
also recently accuses Herr Kastschenko of a similar absent-miudeduess in
connection with another matter (Strasser, “ Ueber die Methoden der Plas-
tischen Reconstructionen,” ‘ Zeitschr. f. Wiss. Mikros.,’ Bd. iv, 1887,
Hefte 2 and 3).

MORPHOLOGICAL STUDIES.

191

certain exceptions; connected with1 the cranial ganglia of Ver-
tebrates above Ichthyopsida, are rudimentary organs which only
present themselves during embryonic life (Froriep (No. 17),
Beard (No. 6), Beraneck (No. 10), and Kastschenko (No. 40).

The explanation so frequently given of such phenomena as
this, viz. that such organs reappear in the ontogeny as pleasing
reminiscences of the ancestral forms, if it has any claim to
pass as an explanation at all, is only a partial one. There are
many reasons for the reappearance of such rudimentary organs,
one of which is the part they play in contributing to the for-
mation of other organs. In fact, to come to the point, we
are here dealing with cases of Kleinenberg’s law of the de-
velopment of organs by substitution (No. 41). I will not
enter at length here into the application of Kleinenberg’s law to
the nervous system of Vertebrates. For a full comparison of
the phenomena presented in the development of the Vertebrate
nervous system with analogous and homologous phenomena in
that of Annelids (No. 41), our knowledge of the former is as
yet not sufficient. Some comparisons can even now be made,
but the time for their consideration had better be deferred.

The neuro-cpithelia of the rudiments of the branchial sense
organs appear in the ontogeny of the higher Vertebrates, be-
cause they contribute certain form elements to the cranial
ganglia, and very probably also to some, at least, of the sensory
cranial nerves.

In the Chick (figs. 90 — 93, 96) such sense-organ rudiments
are found in connection with the mesocephalic (figs. 90, 93, m.g.),
trigeminus (figs. 90, 92, 94, v), facial (90, vii), glossopharyngeus
(figs. 90, 91, ix), and vagus ganglia. In Mammals (sheep em-
bryos) Froriep (No. 17) has described them in connection with
the facial, glossopharyngeus, and vagus ganglia. In Mammals
they have not as yet been described for the mesocephalic and

1 Tbe exception here has reference to the nose and ear, for both of wbicb
organs evidence is accumulating for tbe views of tbeir homology with tbe
sense organs of tbe lateral line which I originally expressed (Nos. 4, 5, 6). I
believe the organs of taste also arise from such neuroepithelium and wander
through one, or in some cases perhaps two, gill-clefts on each side into the mouth
cavity. My evidence for this conclusion will be produced in another Study.

192

J. BEAltD.

trigeminus ganglia, but there can be little doubt that they also
exist for these at some stage or other.

After the fusion of the mass of each cranial ganglion with
the skin, form-elements are, as we have seen, given off into it.
The ganglion leaves the skin, aud, as in Sharks, almost certainly
leaves sensory nerve branches behind it. The sense-organ
rudiments afterwards disappear. I have not followed the steps
of this process in the Chick, but I cannot doubt the general
accuracy of Kastschenko’s account (No. 40, pp. 281 — 284), for
it agrees fairly well with Professor Froriep's earlier researches
(No. 17) on the fate of the rudiments in Mammals.

This finishes the general account of the first formation of
cranial ganglia in the Chick.

III. The Development of the Anterior Roots of
Spinal Nerves in Elasmobranchs.

In Balfour's account of the spinal nerves in Elasmobranchii
(Nos. 1 and 2), he described the anterior roots as direct cellular
outgrowths from the lateral ventral region of the spinal cord,
and in the second volume of the ‘ Comparative Embryology/
p. 372, he says : “ The anterior roots of the spinal nerves
appear somewhat later than the posterior roots, but while the
latter are still quite small each of them arises as a small
but distinct concise outgrowth from the ventral corner of the
spinal cord, before the latter has acquired its covering of white
matter. From the very first the rudiments of the anterior
roots have a somewhat fibrous appearance and an indefinite
form of peripheral termination, while the protoplasm of which
they are composed becomes attenuated towards its end. They
differ from the posterior roots in never shifting their point of
attachment to the spinal cord, in not being united to each
other by a commissure, aud in never developing a ganglion.5'
The anterior roots grow rapidly, aud soon form elongated
cords of spindle-shaped cells with wide attachments to the
spinal cord." And in a note at the foot of p. 372 : “ The
cellular structure of embryonic nerves is a point on which I

MOEPHOLOGICAL STUDIES.

193

should have anticipated that a difference of opinion was im-
possible, had it not been for the fact that His and Kolliker,
following Remak and the older embryologists, absolutely deny
the fact. I feel quite sure that no one studying the develop-
ment of the nerves in Elasmobranchii with well-preserved
specimens could for a moment be doubtful on this point. And
I can only explain His’s denial on the supposition that his
specimens were utterly unsuited to the investigation of the
nerves. I do not propose in this work entering into the histo-
genesis of nerves, but may say that for the earlier stages of
their growth, at any rate, my observations have led me in
many respects to the same results as Gotte (‘Entwickl. d.
Unke/ pp. 482 — 483), except that I hold that adequate proof
is supplied by my investigations to demonstrate that the nerves
are for their whole length originally formed as outgrowths of
the central nervous system. As the nerve-fibres become differ-
entiated from the primitive spindle-shaped cells, the nuclei
become relatively more sparse, and this fact has probably
misled Kolliker. Lowe, while admitting the existence of
nuclei in the nerves, states that they belong to mesoblastic
cells which have wandered into the nerves. This is a purely
gratuitous assumption, not supported by observation of the
development.”

I could have been content to leave this matter of the anterior
roots unnoticed but for two circumstances. In the first place
the figures which Balfour has given of their development in
the ‘ Comparative Embryology ’ (vol. ii, p. 371, fig. 267),
“Elasmobranch Fishes” (PI. X, fig. 7), and in the paper on the
spinal nerves (No. 1, PI. XYI, figs. Da. b. and c. PI. XVII,
figs. H ii, Iu, and E. b.) are very diagrammatic, and His would
be justified from his standpoint in objecting to their repre-
senting the true facts. On the other hand, I can raise the
same objection to the diagrammatic figure of the development
of anterior roots iu Pristiurus, which His represents on p. 393
(No. 34, fig. 1) of his recent work. Nay, I cannot help insisting
that if Balfour’s figures were not what one might expect, His’s
figure is incorrect to a far greater degree, and the “Ehreu-

194

J. BEARD.

Wache 33 of parablast cells which Professor His, in conformity
with his peculiar doctrines, gives the nerve, has, so far as I can
find out in Pristiurus and other Elasmobranchii, no real exist-
ence in the world of fact. I should have been quite satisfied but
for these considerations to merely echo Dr. van Wijhe’s recent
remark on this point (No. 63, p. 76, Anmerkung). He says,
“ In BetrefF der zelligen nicht faserigen Struktur der ventralen
Wurzeln bei ihrem ersten Auftreten muss ich Balfour voll-
standig Recht geben.”

There are three investigators who have regarded the motor-
nerve formation as entirely due to fibres alone, without the
inclusion of any nuclei. The view is one which can only
be accepted if rigidly proved, and this in my opinion has not
yet been done. His holds that the anterior root-fibres are pro-
longations of cells which lie in the anterior cornu of the spinal
cord. Thus, according to him, a motor nerve-fibre passing
from the spinal cord to a muscle in the foot would be
a direct prolongation of a gauglion-cell within the cord, and
no cell nuclei would at any time intervene in its course
(Nos. 32 and 33, p. 375).

He asserts (No. 29, p. 475), “ Mit Beginn des vierten
Bebriitungstages, sind vordere Wurzeln erkennbar als Biindel
feiner, vom Riickenmark aus in die Leibeswand tretenden
Faden. Vom ersten Moment an, da sie iiberhaupt sichtbar
sind, haben sie die angegebenen Eigenschaften und ihrem
Auftreten gehen keine zelligen Urgebilde voraus.”

And again, in the more recent work (No. 33, p. 375),
“ Die peripherisch auswachsenden Fasern, sowohl die motor-
ischen, als die sensibeln, sammeln sich als kurzen Stammen.
Jeder dieselben besteht aus einer Anzahl feiner kernloser
Faden, die in der Nahe des Ursprungs eine deutlich fibrillare
Streifung zeigen. Innerhalb eines Stammes zeigen die Fasern
theilweise verschrankten Verlauf. Parablasticshe (!) Zellen
zeigen sich beim mensclilichen Embryo Anfangs nur sparsam

zwischen den Nervenfasern zerstreut Je jiinger

ein Nervenstamm, um so kiirzer ist er, das Auswachsen
gescliieht allmahlich und es vergehen einige Wochen bis z. B.

MOEPHOLOGICAL STUDIES.

195

die letzten Enden der FiDger und der Zehen ihre Nerven
erhalten haben.”

To return to nay own observations. I have as yet only in-
vestigated Sharks and Lizards on this point. The results, so
far as they concern the latter group, will be given later along
with observations on the anterior roots of the cranial nerves.
Figs. 58, GO, 62, 53, 54, 56 and 61, on PI. XIX, are intended to
illustrate the development of anterior roots of spinal nerves in
Elasmobranchii. One of the very earliest stages in the deve-
lopment of an anterior root (a) is shown in fig. 58 (Pristiurus),
and it possibly corresponds to the stage figured by Professor
His in the paper quoted above (No. 34, fig. 1). There is no
possibility of recognising “ parablastic ” cells in this section,
and one sees that while the root is partially fibrous there is at
least one nucleus passing out of the spinal cord, either entirely
or partly after cell division. A slighter later stage is repre-
sented in figs. 60 and 61. The fibres of the nerve have reached
the muscle-plate,1 but there are also two nuclei visible in the
nerve-cord lying partly also in the cord. There are here also
plenty of mesoblast — pardon, “ parablast ” cells in the neigh-
bourhood. But they are not destined for the nerve, but are about
to enclose the notochord to form the body of the vertebra.

Later stages in the development are figured in figs. 54, 56.
Here, too, the fibrous nature of the nerve is very obvious, but
one also observes a vast number of nuclei within the nerve,
which one cannot regard, from their form and characters, as
otherwise than offsprings of the nuclei which have passed at
earlier stages, and even still continue to do so (figs. 60,54,61),
from the anterior cornu to the nerve. When His regards the
nuclei here present as mesoblastic or “ parablastic ” cells, his
view is just as much a gratuitous assumption as the whole
parablastic doctrine, as the Zwischenstrang ganglionic forma-
tion in the trunk, and as the identification of a certain ganglion

1 The end plates of muscles (and of the electric organ) are derived from
ganglionic cells, which wander in this way in these early stages from the
anterior horn to the muscle-plate. Several figures show this, and 1 shall treat
of the matter at length elsewhere.

196

J. BEARD.

to be mentioned elsewhere as the ciliary ganglion. A figure
such as Professor His gives in his recent work of two fibres
passing out from two nuclei in the anterior cornu of the head,
outside which they receive an “ Ehren-Wache ” of four
“ parablastic ” cells, two on each side, is one which, in spite of
much search, I have never seen. On the other hand, the
figures I give in PI. XIX could be multiplied by the dozen, and
figs. 53 and 58 are representations of an appearance which I have
often met, and which Dr. van Wijhe assures me he also has
very frequently seen. From these facts, and from facts regard-
ing the development of anterior roots of cranial nerves, and
nerves of the sense organs of the lateral line or branchial sense
organs, I do not for a moment hesitate to declare that the
facts of development are contradictory to (1) Professor His’s
view of the absence of nuclei in the anterior roots, and (2) his
assumption that when such nuclei are present they are of
“ parablastic ” origin. It is worthy of notice that in his
original assertions Professor His absolutely (vide supra)
denied the presence of nuclei in the anterior roots, and only
now that their presence in those structures in Elasmobranchii
is obvious to every observer, including Professor His, does it
occur to him to make use of that wonderful doctrine of para-
blast to explain their presence. Professor His’s attempt to get
out of a false position here is only a little more dignified than
his endeavour to explain away the meaning he attached to the
Zwischenstrang.

IV. The Ganglionic Development in Different
Grouts of Vertebrates.

Without anticipating the results of my researches on other
groups, which so far include Teleostei, Lizards, Frog, Newt,
and Rabbit, I may be at least allowed to say now that the above
mode of development of cranial and spinal ganglia holds, with
very slight and unimportant modifications, for all these forms
also. I might have left these forms undcscribed but that for the
certainty that some observer or other would by-and-by quote

MORPHOLOGICAL STUDIES.

197

their development according to his ideas, as opposed to the
facts I have here described for Elasmobranchii and Birds.
The conviction was very early in the research forced upon me
that the development of spinal and cranial ganglia in all
Vertebrates must take place after one type, and any differ-
ences found in different groups must be referable to variations
or changes rung on that type. And as an example, the inves-
tigation of the development of cranial ganglia in the Anurous
Amphibians was one I could not leave unnoticed on account
of Spencer’s notes on the matter (No. 59). All the more,
as in my paper on the branchial sense organs (No. 6) I
felt obliged, after the examination of some of Spencer’s pre-
parations, and of a few I made myself, to support his conclu-
sions. We were then both in error on one point — of that I
am now quite sure — and that is in reference to the deeper
layer of epiblast above the level of the lateral sense-organ
thickening, and which connects the latter in early stages with
the neural plate. We both believed it gave origin to the
trunk of the nerve. This is not so. That layer is indifferent
except at two points corresponding exactly to the two points
at which the ganglionic form elements arise in Elasmobranchs.
In fact, as a preliminary note I take the opportunity of saying
that the cranial ganglia of the Frog develop in exactly the
same way as those of Elasmobranchii. Among other forms
examined the Lizard is one of the most favorable for such in-
vestigations. It also agrees essentially in the mode of develop-
ment of cranial and spinal ganglia with Elasmobranchii.

The Newt has been mainly studied by Bedot (No. 9), and
Misses Johnson and Sheldon (No. 38).

In both of these works I shall have occasion to underline a
number of mistakes and false interpretations; here I will only
remark that I am somewhat surprised that none of these in-
vestigators have seen the epi'blastic origin of the spinal ganglia
in this animal. I know no animal in which such origin is
easier to identify. The criticisms with which the two latter
authors have seen fit to honour my work may also be here left
unnoticed. The only one whose justification I will acknow-

VOL. XXIX, rART 2. NEW SER.

o

198

J. BEARD.

ledge is their doubt of the accuracy of Spencer’s and my in-
vestigations on the origin of the root of a cranial nerve in the
Frog. I have admitted the error above, and need not here
mention the matter further. To one assumption of these two
authors (No. 38, p. 11) I must, however, be allowed here to
reply. They remark : “ More recently the theory of the de-
rivation of the whole or greater part of the cranial nerves from
the epiblast has been supported by Mr. Spencer and Mr. Beard.
This view is a revival of that held by Gotte.”

(1) The origin of a part of each of the cranial ganglia, and
of what I called the suprabranchial nerves, was no longer a
theory after the publication of my paper on the branchial sense
organs (No. 6). It was then demonstrated for certain parts
of the cranial ganglia1 and for certain nerves that they have
an epiblastic origin, and the matter could for these hardly be
called a “ theory.” I can now demonstrate that the whole of
the components of the various cranial ganglia are epiblastic in
origin, and not wholly or in part outgrowths of the central
nervous system.

(2) Gotte never held this view, whatever may now be the
case. I can only suppose that the two ladies never read the
passages in his work which bear upon the question. The fol-
lowing quotation from Gotte’s ‘Unke’ (No. 22, p. 719) gives
a clear statement of Gotte’s conclusions at that time : — Bei der
Untersuchung der Kopfnerven handelt es siclx zunachst urn
ihre Zugehorigkeit zu den ganzen hintereinander liegenden
segmentalen Abtheilungen des Kopfes ferner um ihre Unter-
scheidung nach dem Ursprunge aus dem inueren oder aiisseren
Segmente des mittleren Keimblattes oder aus andcren Era-
bryonalanlagen jeder Abtheilung. Zu den letzteren gehbren
der Sehnerv und die Seitennerven als Erzeugnisse des oberen
Keimblattes, die ubrigen Kopfnerven eustchen aus dem
mittleren Keimblatte.”

1 I was inclined then to regard the whole of the ganglion as arising from
the epiblastic sense thickening, and the cells derived from the “ neural crest ”
as forming the root of the nerve. The point is a very difficult one to decide,
and I refer the reader to a discussion of it in another part of these researches.

MORPHOLOGICAL STUDIES.

199

Y. The Neural Ridge.

The reader may have remarked in the preceding pages that
the terms neural ridge and neural crest have been banished
from my account of the development of the ganglia, both
cranial and spinal. The reasons for this may now be explained,
and hand in hand with this explanation one may compare the
origin of the ganglionic Anlagen as described here with the
accounts of previous observers.

Considering for a moment the neural ridge without prejudice
as to its origin, most authors, following Marshall (No. 46, p. 15),
regard the neural ridge as a continuous structure passing for-
wards from the mid-brain right away backwards through the
head and along the whole spinal cord as a continuous struc-
ture ; and from its continuity in all parts, of which in a certain
sense there can be no doubt, Balfour and Marshall were in-
clined to attach great morphological importance to it. The
continuity of the neural ridge is originally most marked in the
head, in which the ganglia show tendencies to concentration and
fusion, and where also the ganglionic Anlagen are very large.

In the spinal cord, on the other hand, where the ganglionic
Anlagen are not so massive, the continuity of the neural ridge
is by no means so evident as in the brain. Indeed, from the
neural-ridge stage onwards, aud even from the very first forma-
tion of the spinal ganglia Anlagen, the segmental nature of
the latter is one about which a careful investigator can make
no mistake. For this reason, and the additional one that all
the cell elements of the neural ridge in both head and trunk
undoubtedly, as His insists (No. 34, p. 393), pass over into the
ganglia, I can sec no particular advantage in the use of the
term. And when one comes to consider, as we shall presently
do, the origin of the neural crest, my objections to the term
as at present used are intensified. Marshall, from the ap-
parent fusion of the neural ridges of the two sides, gave to
the single structure thus formed the name of neural crest.
Ilcre, again, as the structure is certainly a bilateral one and
not unpaired, and as in many cases its bilateral structure is

200

J. BEARD.

very evident (figs. 21, 24, 46, 51), I confess I see no conve-
nience in the use of a name to which doubtful morphological
characteristics are attached.

We are now met by the question, Assuming that the ganglia
arise as outgrowths of the neural ridge, what is the ultimate
origin of the structure, and are the ganglia first visible in the
neural-ridge stage ?

The foregoing researches give the answer to this question,
and in anything like a complete and correct form they are the
first researches which can lay claim to decide the question. Six
years ago Sagemehl (No. 56), in a prize research, published
observations which he believed, and apparently the judges of
the competition also, to be a solution of the problem, so far as
the spinal ganglia are concerned. How little claim his
researches have to pass as a last word on the origin of the
ganglia will be evident to the reader of this paper, and if he
will take the additional trouble to compare the numerous
figures I have given here of Elasmobranchii and the Chick
with the nineteen figures of Sagemehl’s work, he will, I think,
admit the correctness of my conclusion, that Sagemehl never
saw any of the earliest stages of the formation of spinal ganglia.
Except for Marshall’s and Onodi’s researches on the cranial
ganglia of the Chick, this remark applies to all the observa-
tions of various investigators of the development of cranial
and spinal ganglia. His (No. 29) has also seen, but only
partially interpreted in a correct sense, some of the earliest
stages in the cranial nerves of the Chick. As His’s Zwischen-
rinne theory was one of the earliest on the development of
cranial nerves, we can at once consider his claims to having
furnished the solution to the above question in the wider sense
of the origin of the ganglia Anlagen. Remak’s (No. 54)
older observations, originally supported by Balfour and Foster,
may be here passed over, for no one now believes that the
ganglia arise as differentiations of the “ protovertebrae.” And
the same also holds for Hensen’s conclusions (No. 24), which
are more of a theoretical nature than results of actual investi-
gation ; still, as I shall elsewhere show, there is an element of

MORPHOLOGICAL STUDIES.

201

truth in Henseu’s suggestions, though not quite the same
Hensen thought.

It is perhaps unkind to remind Professor His that his
“ Zwischenstrang” was originally believed by him to be con-
cerned in the formation of the urogenital system. The
Zwischenstrang was afterwards converted in the basis of a
theory of the origin of the spinal ganglia. In spite of the
persistent way in which Professor His, without full and com-
plete investigation of the matter, holds to this Zwischenstrang
theory of the proved origin of spinal ganglia, a persistence
which leads him in his recent work (No. 34, pp. 391 and 416)
to identify it with what Balfour, Marshall, Sagemehl, and
others have regarded as the first stages in the formation of the
ganglia, and to rebaptize the structure, which undoubtedly
exists (figs. 97, 98, 99, z.), under the name of “ Ganglionstrang,”1
I do not see how Professor His can escape the fatal conse-
quences of the researches I now record.

I think I have demonstrated, even to Professor His’s satis-
faction, that the Zwischenstrang is just that part of the epiblast
which takes no part in the ganglionic formation, and that it
owes its formation to the cutting out of ganglionic Anlagen
between it and the neural plate. As the crowning proof that
the Zwischenstrang is not identical with the neural ridge or the

1 Professor His (No. 34, p. 417) states that both olfactory and auditory
organs of Vertebrates take their origin from parts of the “ Zwischenrinne ”
or “ Ganglienrinne ” which remain open. This is absolutely incorrect. The
views of the homology of both these organs with the lateral or branchial
sense organs, which I formerly advocated (Nos. 6, 5, 4), can be still main-
tained. Prom figures in my former work (No. 6) and figs. 25, 27, 46 of this
paper, it is obvious to any unprejudiced observer that the auditory organ
develops ganglionic elements from two sources, just as occurs in a typical
gill-bearing segment. The same holds for the olfactory organ. I postpone
for the time the further elucidation of my views of the homology of these
two sense organs, but only for a time, for I intend shortly to discuss the
problems they present more fully ; here I will only say that no one has as
yet urged unanswerable arguments against my views. Personally, I may
remark, I care nothing about the quondam existence of gill-clefts for ear and
nose ; the important points to me are those which make the nose and ear parts
of the system of lateral or branchial sense organs.

202

J. BEAliD.

ganglionic Anlagen, I may refer to figs. 97, 98, 99 and others,
more especially figs. 97 and 98, in which the “ Zwischenstraug ”
and the ganglionic Anlagen can he seen in the same figure, and
where they are entirely distinct and separate.

When we turn to Professor His’s researches on the cranial
ganglia of the Chick (Nos. 28 and 29), we find that he was a
little more fortunate in seeing some of the true facts. But
here again his theory influenced his interpretation of the facts.
The foldings of an elastic plate by which, as is well known.
Professor His explained all embryonic phenomena1 (No. 31),
must also find their application in the formation of the cranial
ganglia. It is not merely in the assumption of such a folding
in of the epiblast of the head to form the gangliouic Anlagen
in his “ Zwischenrinne " that His is in the wrong; he has
actually figured such a Zwischenrinne (No. 29, PI. XVII, figs.

3, by Cy d, (?, /).

I have made a very large number of sections through the head
region of Chick embryos (well preserved) in this stage, and as
the result I do not for a moment hesitate to say that the
Zwischenrinne of His has no existence. On the contrary, in
the head just as in the trunk, as the result of the separation
of the ganglionic Anlagen from the epiblast, a “ Zwischen-
straug ” may be formed (figs. 63, 97, 88); but this structure
also plays no part in the formation of the ganglia. If Pro-
fessor His had not assumed or believed in the existence of this
“ Zwischenrinne,” and if he had left the “ elastic plate ” out
of question and acknowledged the proliferation of a certain
portion of the inner epiblastic layers to form the ganglia, he,
who certainly was the first to see some of the true appearances
on the Chick, would also have been the first to ascribe their
true epiblastic origin to the cranial ganglia. But under the
dominance of his theory he believed he saw structures2 which

1 This “ Mechanische Auffassung ” lias unfortunately more influence on
Professor ilis’s results than his conception of the great value of comparative
embryology, to which he lays claim in p. 405 of his recent critical study.

2 One must bear in mind that the sections of those days were nothing like
as good as those a fair worker can now make.

MORPHOLOGICAL STUDIES.

203

have uo existence ; and he says in a passage which on another
page I have quoted in full, “ Ich verwerfe iiberdies, wie
ich schou an auderem Orte ausgesprochen liabe, das von
manchen Embryologen so freigebig benutzte Princip dcr lokalen
Wurcherungen ” (No. 34, p. 4G5). The reference to another
place in this passage is to the “ Briefe iiber unsere Korperform ”
p. 67, u. f. — a work in which the foldings, &c., of au elastic
plate are used to explain fully the development of all the organs
of a Vertebrate embryo. By this declaration Professor His
gives the coup de grace to any possibility of the acceptance
of his account of the cranial ganglia in the Chick as a solution
of their origin. The two diagrammatic figures which are re-
presented ou p. 465 of Professor His's paper oil the peripheral
nervous system have been referred to recently (No. 34, p. 394,
Aumerkung) by him as representing really the true facts, and
as agreeing essentially with the x-esults of other iuvestigators ;
but that 1 may not be accused of an unfairness, which is far
from my thoughts, I quote the passage : “ Wie jedes Schema,
so ist auch dieses iu Betriff absoluter Correctheit1 anfechtbar,
aber, dass die untere Lamelle des dort ausgebogenen Streifeus
mit der von Kolliker, Sagemehl, u. A. abgebildeten Ganglien-
anlage zusammeufallt, bedarf kaum eine Erlaiiterung.” As
these figures show au epiblastic invagination to form the
ganglionic Aulagen, iu conformity with the elastic plate theory,
— an appearance which has no existence iu fact, — it is difficult
to see how the lower layer of this structure can be identical
with the ganglionic Aulagen of Kolliker, Sagemehl, and others.
This is as near being the case as any fancy figure drawn in the
same position would be. The principle of the epiblastic origin
of the ganglia, apart from the central nervous system, is one on
which His has long been in the right; the mode in which he
believes this origin takes place is one in which he has been
further from the true facts than anyone else. I have quoted
before the following passage from Professor His’s recent paper
(p. 380), and as we now see that the facts are not so much

1 This “Schema” of His’s is not relatively correct, it is absolutely
incorrect !

204

J. BEARD.

matters of agreement as Professor His supposes, one may
quote it again with the request to Professor His to furnish
us with the evidence in which he bases his opinions on the
origin of the ganglia from the Zwischenstraug and Zwischen-
riune, and of the olfactory and auditory organs from parts of
the latter structure which l’emaiu open (No. 34, p. 417).
These are questions of facts whose accuracy I challenge. Nor
are they the only points of fact on which I (and many others)
disagree with Professor His. Of that more elsewhere.

The passage reads: “Bei genauerem Zusehen findet man
eben, dass die Differenzen nicht in dem liegen, was der eiue
und der andere Beobachtungskreis an thatsachlichen Befunden
ergiebt, sondern in demjenigen, was die Yertreter der einen
und der anderen Schule zwisclien die Zeilen zu lesen sich
bemiihen. Nun sind aber die jiiugeren vergleichend rnorpho-
logischen Schulen in der Lecture zwisclien den Zeilen iiberdie
Maasen weitgegangen, und ich halte es fureine Pflicht, meineu
Bedenken hiergegen offenen Ausdruck zu geben.”

However it may be with the hypotheses, &c., one thing is
certain, that some of Professor His’s most funda-
mental facts are no facts at all, and we may not un-
naturally ask whether the reproach intended for us
younger morphologists does not partially recoil on
Professor His himself?

All other observers, excepting Spencer for the cranial
ganglia of Amphibia, are agreed in referring the source of the
posterior roots and ganglia to the neural ridge of Marshall,
and nearly all agree with Balfour’s maxim of the origin of the
latter structure as an outgrowth from the central nervous
system.

On p. 369 of the ‘Comparative Embryology’ of Balfour,
vol. ii — a book which represents his latest views on the ques-
tion— we read : “ All the nerves are outgrowths of the central
nervous system;” and on p. 374, “The neural crest clearly
belongs to the brain, from the fact of its remaining connected
with the latter when the medullary tube separates from the
external epiblast.”

MORPHOLOGICAL STUDIES.

205

Marshall’s position is not quite so simple. The cranial
nerves (and ganglia) of the Chick Marshall (No. 46) refers to
the re-entering angle between the neural plate and external
epiblast, but nothing definite is stated as to the relations of this
portion of the epiblast to the external epiblast on the one hand
or to the brain on the other. In other words, if we are entitled
to conclude that Professor Marshall held the independent epi-
blastic origin of cranial nerves and ganglia, we miss in the
account the necessary denial of Balfour’s view as stated above.
If Marshall recognised the epiblastic origin of the neural ridge
he did not tell us whether or not he holds with Balfour that it
“ clearly belongs to the brain.” This is important, for taken
in connection with his acceptance of Balfour’s view of the origin
of spinal ganglia, it does not preclude the possibility of the
assumption that the neural ridge in the Chick arose from a por-
tion of the brain which has not got shut in. Professor Marshall
has indeed seen and described part of the true origin of the
cells which form the neural ridge in the Chick. The whole
of the source he has not identified, and he did to draw the
conclusions of the independent origin of the ganglia to which
he was entitled.

The part he had not seen is that portion of the cranial
ganglion Anlage which is formed from the external epiblast
outside the angle between epiblast and brain. This was first
seen by Onodi (No. 51).

Judging from the following passage, it would appear as
though Professor Marshall held the origin of the ganglia to be
the same in both brain and cord, and the difference to be only
as to the time of closure of the neural canal. He says (No. 46,
p. 16): “Its (the neural ridge) first appearance precedes the
closure of the neural canal, but after about the fortieth hour the
closure of the canal proceeds backwards more rapidly than the
growth of the neural ridge, so that in the greater part of the
length of the spinal cord the ridge is developed as an out-
growth from the summit of the cord itself, and never
has any connection with the external epiblast.”

In order to get a little nearer Marshall’s position 1 turned to

206

J. BEARD.

his latest statements on the development of nerves, and find
(No. 50, p. 9) that he quotes with approval Balfour’s views.
He says, “ Balfour showed that, contrary to the generally
accepted theory, the nerves are outgrowths from the
central nervous system, and therefore of epiblastic origin,
instead of being, as formerly supposed, structures arising
independently in the mesoblast and only acquiring a secondary
connection with the brain and cord.” Ilensen (No. 25),
Kolliker (No. 43, p. 621), Sagemelil (No. 56, p. 33), van
Wijhe (Nos. 60, p. 18), Bedot (No. 9, p. 186), Shipley (No.
58), Beraneck (Nos. 10, 11), and Misses Johnson and Sheldon
(No. 38), have practically accepted Balfour’s and Marshall’s
views; and van Wijhe (No. 61, p. 4) has used the conclusion
as an argument against my views of the epiblastic origin of
the sensory nerves of the branchial sense organs (Beard, No. 6,
p. 69). He remarks, “Wenn Beard jetzt, seiner friikeren
Behauptung entgegen, den Olfactorius und die Seitennerveu
nebst ihren Ganglien alleiu aus der Epidermis enstehen lasst,
so kann er dies wohl nei beweisen weil der Stamm der Nerven
sich urspriinglick aus dem Medullarrolire entwickelt.”

It is not difficult from the researches I have here recorded —
and others as yet unpublished — to conclude that all these
authors have been mistaken in describing the ganglia as out-
growths of the central nervous system. The figures I have
given demonstrate the justice of this criticism, and as a final
argument, which more especially negatives Balfour’s remark
(quoted earlier), that the neural crest clearly belongs to the
brain, I will point out that the limits of the two structures,
brain and ganglionic Anlagen, are very early sharply separated
off by a well-defined line (figs. 45, 51, 32 — 36 and others), and
only in those stages in which the neural plate is quite open, in
fact only during the primitive-streak period can one really, with
any pretence to accuracy, speak of a common Anlage for both
structures, of an encephalo-gangliouic Anlage. But this is a
stage at which the embryo is barely differentiated into the
three embryonic layers.

Onodi (No. 51) has shown the true source of origin of

HOKPHOLOGICAL STUDIES.

207

the main portion of each crauial ganglion iu the Chick, — of
that portion which is not derived from the remains of the bran-
chial sense organs. The rest of his researches, on the cranial
and spinal ganglia of Elasmobranchii, Teleostei, Lizards, and
Mammals, and on the spinal ganglia of the Chick, lead him to
the same results as Balfour, Marshall, and others. His
researches hence agree partially with my own for parts of the
cranial ganglia of the Chick, but for all other types he has
failed to see the true epiblastic origin of both cranial and
spinal ganglia.

Hoffmann (No. 36, pp. 45 — 49) while supporting Balfour’s
views of the outgrowth of spinal ganglia from the cord, con-
siders it probable that the posterior root ganglia of the cranial
nerves of Teleostei arise from the epiblast beyond the limits of
the neural plate, and before the closure of the latter. He did
not prove that such was the case.

In later researches (No. 37, p. 204) he again refers to the
neural ridge, but says nothing of its origin.

VI. The Growth of our Knowledge of the Independent
Epiblastic Origin of the Peripheral Nervous
System.

The first conclusions on this question were arrived at by
Gbtte (No. 22, p. 72) and Semper (No. 57, p. 256), both of
whom stated that the lateral nerve has an epiblastic origin and
arises pari passu with the growth of the lateral line as a
differentiation of the epiblast. Gotte (p. 719) extended this
mode of development to the nerves of the lateral sense organs
of the head. These statements, on which doubt was cast by
Balfour, were practically confirmed by van Wijlie (No. 60, p.
35) and Hoffmann (No. 36, p. 89,) for Teleostei. I (No. 4)
believed Balfour’s doubts to be well founded, but in two sub-
sequent publications I was able to prove, for Elasmobranchii
the accuracy of Semper’s account. Just before my paper on
the origin of the cranial ganglia (No. 5) appeared Professor

208

J. BEARD.

Froriep published his researches on the rudiments of sense
organs in connection with several cranial ganglia in Mammalia
(No. 17). Without committing himself very definitely to the
matter Professor Froriep did not think it impossible that the
ganglia derived form-elements from the epiblastic fusion
(No. 17, p. 40), and the cranial ganglia concerned were re-
garded by him as the remains of the ganglia of sense organs
which in the course of phylogenetic development had got lost.

He says (p. 45) : “An der drei Nerven iibereinstimmend
gehen aus der Kiemeuspaltenorganen keine definitiven Bil-
dungen hervor, was von ihnen iibrigbleibt, ist lediglich die
gangliose Anscliwellung des Nerven, welche urspriinglich
die nervose Unterlage des Sinnesepitheliums gewesen ist.
Diese Ganglien, Ggl. genicule, Ggl. petrorsum, und Ggl.
nodosum, sind demnach als rudimeutare Organe zu betrachten,
sie stellen die Ueberreste phylogenetisch verloreugegangener
Sinneswerkzeuge dar.”

Professor Froriep was undoubtedly the first in point of time
to describe this fusion of cranial ganglia with the epiblast, and
to draw the conclusion that the modified epiblast at the point
of fusion was the remains of a special branchial sense organ.
He hesitated (p. 35, et seq.) to homologise them with the
sense organs of the lateral line in Fishes, considering it possible
that they corresponded with rudiments of other sense organs
connected with the ventral branches in Fishes as in Mammalia,
and which, as in Mammalia, probably disappeared in later
development.

The identification of the ganglion fusion with the “Anlagen”
of the sense organs of the lateral line for head and trunk in
Elasmobranchii, was first made by me (No. 5) independently
of Professor Froriep, and at that time also — a point which I
afterwards developed more fully — I was quite aware of the
relations of the sense organs to the gill-clefts, for I homo-
logised the nose with such a ganglionic epiblastic fusion, and
called it “ the modified sense organ of a gill-cleft rather than
a gill-cleft itself and in my note-book there still stands the
notice from which I wrote that conclusion, which shows, I

MORPHOLOGICAL STUDIES.

209

think, very clearly that, contrary to Professor Froriep’s recent
criticism (No. 19, p. 821), I was then fully aware of a point to
which he attaches a very great deal of importance, viz. their
typical position over a gill-cleft. The note is, “The nose is not
a gill-slit but the sense organ which sits above a cleft.”

In my paper on the branchial sense organs (No. 6) I showed
that out of this epiblastic fusion, which (No. 5) I had described
independently of Froriep, the sense organs of the lateral line
or branchial sense organs take their origin. The sensory
epithelium grows in various directions by division of its cells,
and it pushes away the indifferent epiblast. From the sensory
epithelium arise both sense organs and the nerves which supply
them and connect them with the ganglia. The ganglia were
considered as mainly arising from the thickenings, the cells
derived from the neural ridge only forming the root of the
nerve. Whether the latter conclusion is true or not I cannot
say, certainly some of those cells do take part in the formation
of the nerve, and their nuclei may be found along the course
of the nerve. The suprabranchial nerves were distinguished
from the praebranchial and postbranchial, and a morpho-
logical importance was attached to the former. At the present
time I regard the nature and mode of origin of suprabranchial,
praebranchial, and postbranchial nerves, so far as the latter
innervate the sense organs (for, as is well known, they also
contain motor fibres to the muscles of the gill-cleft) as entirely
the same, and would now say all the nerves to the sense organs
of the lateral line or branchial sense organs are derived from
the neuro-epithelial “ Aulagen ” of the latter.

Nothing was said in my former paper of the origin of the
neural-ridge of the spinal nerves, which lay beyond the scope of
my researches at that time. Nose and ear were considered as
modified branchial sense organs and their ganglia (for, in spite
of Gegeubaur, the nose1 has a ganglion) as differentiations of
the sensory epithelium. Rudiments of such branchial sense
organs and their ganglionic fusion were described in three-
days’-Chick embryos. Spencer (No. 59) on Amphibia (Frog),

1 Sec No. IV of these Studies.

210

J. BEAKD.

derived the cranial ganglia from the epiblastic thickenings
which form the lateral sense organs, and the main roots of the
nerves from the inner epiblast connecting this thickening with
the neural plate. This latter conclusion, which I formerly
supported, is wrong.

Onodi (No. 51) extended Marshall’s (No. 46) description of
the origin of the cranial ganglia in the Chick from the angle
between the epiblast and the neural plate, in that he stated
that the epiblast outside this also shares in the formation.
Neither Onodi nor Marshall distinctly say whether they regard
this portion of epiblast as part of the central nervous system
or not. And, as we have seen for the cranial and spinal
ganglia of other forms, they supported Balfour’s views.

In a note which I quoted in the introduction, van Wijhe
(No. 61) mentions that the olfactory nerve arises from an epi-
blastic differentiation at the lips of the anterior neuropore.
The present research, taken in connection with my former
paper on the branchial sense organs, shows that the sensory
nerve-elements of the whole of the peripheral nervous system
arise as epiblastic differentiations independently of the central
nervous system.

VII. The Relations of Cranial to Spinal Ganglia and

OF TIIE “ SEITENORGANE ” OF ANNELIDS TO THE SENSE
Organs of Vertebrates.

It is far from my intention to enter here into the discussion
of morphological questions. My contribution to recent con-
troversy may fitly find a place in a special paper in which I
intend to analyse the recent critical studies of Professors
Gegenbaur and His on Vertebrate morphology, and especially
on the nervous system.

But still, the conclusions to which Froriep and I arrived at
regarding the fundamental differences which obtain between
the head and trunk regions of Vertebrates may be here
slightly reviewed, and, so far as I am concerned, revised in the
light of the facts recorded in the preceding pages. Gegenbaur

MORPHOLOGICAL STUDIES.

211

(No. 20), and in a certain sense Dohrn and others, regard
the head as a specially modified portion of the trunk, and,
as is well known, Gegenbaur (No. 20) considered that
certain of the cranial nerves could be reduced to spinal nerves.
His present position with regard to recent researches is defined
more or less clearly in his recent paper (No. 21). I cannot
now enter into a criticism of that — the limits of my space
forbid it, — and, as far as possible, I have endeavoured to shut
speculative matter out of this research.

Dohrn (No. 13, p. 471) has formulated his conclusions as
to the relations of the spinal and cranial nerves and ganglia in
the following passages : —

“ Die Hirnnerven haben diejenigen Leitungsbahnen ver-
loren, welche die Urwirbel und deren Derivate iunervirten ; sie
haben aber in Folge der ausserordentlichen Yergrosserung
und Complicationen der visceralen, i. e. ventralen Theile des
Kopfes um so mehr gewonnen und sind durch die vielfaclien
Verschiebungen der beziiglichen Theile in ihrem Yerlaufe sehr
verwickelt geworden.

Die Spinalnerven ihrerseits haben am Rumpfe in ihren
visceralen Verrichtungen Verschiedenes verandert( — auf welche
Weise soli spiiter dargestellt werden — ,) haben aber durch die
Entwickelung der Kbrper — und Extremitaten — Musculatur
im Umfang im Allgemeinen nicht vermindert, und sind in
gewissen Sinne weniger modificirt, als die Cranialnerven.
Am Schwanz dagegen haben sie durch die Einbusse der
gcsammten Yisceralpartien die starksten Verluste erlitten und
sind dort demgemass am wenigsten complicirt.”

While there are some points in the above statements with
which I can express my agreement, my standpoint is more on
Froriep’s side than on that of Dohrn. For a general survey of
Froriep’s views I must refer the reader to that investigator’s
recent utterances (No. 19, p. 833, et seq.).

I agree with Professor Froriep that at present we cannot
see much beyond the primitive separation of the Vertebrate
body into two sharply-defined regions, — a respiratory region
the head, and a locomotive (and digestive) region the trunk-

212

J. BEARD.

We have hardly begun to get any idea of the more primitive
structures from which these two regions are derived.

I have previously with Froriep, much to the disapproval of
Gegenbaur, His, Dohrn, and Eisig (No. 15), sharply contrasted
the cranial and spinal nerves and ganglia, and declared my
conviction (No. 6) that it is a very doubtful question
whether the two sets of organs ever had the same primitive
characters. The development of the branchial sense organs
and ganglia, in connection with the cranial ganglia, was my
main consideration for saying this. And the same considera-
tions appeared to Froriep (independently) to add strength to this
conclusion at which he had arrived some years ago (No. 16).

The question arises, How is the position altered by the
researches I now record ?

Eisig (No. 15, p. 542) had, perhaps rightly, urged against
my views that it was not impossible that the spinal ganglia of
Vertebrates represent the “ Seitenorganen ganglia ” of Capi-
tellidse. Without devoting here the time which a thorough
examination of Dr. Eisig’s comparisons entails, I cannot omit
a partial discussion of this point. The exact weighing of the
pros and contras of Dr. Eisig’s views must be left over for
another publication, in which we must examine more closely
the lateral sense organs of Vertebrates.

I quote the following passage from Eisig’s great work1 (No. 15,
p. 542), in spite of its length, because it touches upon the pro-
posed homology between the spinal and cranial ganglia on the
one hand, and the parapodial ganglia of Annelids on the other.
This homology, as I previously mentioned, was suggested by
Kleinenberg (No. 41, p. 220), and in a strict morphological
sense I think, as the result of my researches, it can be
accepted.

The passage runs thus (p. 542) : “ Es muss dagegen speciell
der Punkt von mir erortert werden auf den sich Beard zum
Behufe der Perhorrescirrung der Homologie von Gehirn uud
Spinalnerven sti'itzt: namlich, die Thatsache, dass die Spinal-

' I take this opportunity of expressing my gratitude to Dr. Eisig for the
generous gift of a copy of his immense monograph.

MORPHOLOGICAL STUDIES.

213

nerven nicht ebenso wie die Hirnnerven mit Hautsinnes-
organen (Seitenorganen), respective mit Ganglien solcher in
Verbindung traten. Um so mehr muss dieser Punkt ins Auge
gefasst werden, als ich davon iiberzeugt bin, dass die in ihm
enthaltenen Probleme auf dem Boden der Vertebraten-rnor-
phologie allein nicht gelost -werden konnen, indem es sich um
V erhaltnisse handelt, welche phylogenetisch so weit zuriickliegen,
dass uns nur die den vermuthlichen Ascendenten der Verte-
braten naher stebenden Wirbellosen noch Anhalts-punkte
fur den Ausgang und die Ricktung der beziiglichen Entwick-
elungen zu bieten vermogen.

“ Wenn die Spinalnerven gegenwiirtig nicht mehr ahnlich wie
die Hirnnerven mit Seitenorganen, respective mit Ganglien
solcher im Bereiche der Haut in Verbindung treten, so frage ich
zuniichst Beard, woher er denn weiss, dass dies auch friiher nie
der Fall gewesen sei, ferner frage ich ihn, ob er irgend einen
triftigen Einwand gegen die Vorstellung beigebracht bat oder
beibringen kann, dass die Ganglien der hinteren Spinalnerven
wurzeln moglicherweise den Seitenorganganglien der Hirn-
nerven entsprechen ? Wie berechtigt diese Frage ist, geht
daraus hervor, dass nicht etwa nur Thatsachen der Vertebraten
— sondern auch solche der Anneliden-Morphologie zu Gunsten
einer solchen Vorstellung oder Hypothese sich anfiihren
lassen.”

Then follows the citation of Kleinenberg’s views respecting
the homology of the parapodial ganglia of Annelids and the
spinal ganglia of Vertebrates, which I have already quoted in
the introduction of this paper.

Dr. Eisig continues (p. 542) : “ Wenn man erinnert, dass ich
ganz unabhangig von der vorliegenden Frage dazu gekommen
bin, die Seitenorganenganglien der Anneliden von den Parapo-
dialgauglien der Anneliden abzuleiten, so wird man einsehen,
dass unserem weiteren Schlussverfahren schon derWegvorge-
zeichnet ist. Es entsprechen namlich aller Wahrschein-
lichkeit nach im Vertebratenrumpfe die Spinalganglien den
Seitenorganganglien (Parapodialganglien) der Anneliden.

“ Und auch die Frage, warum denn erstere Ganglien bei den

VOL. XXIV, PART 2. NEW SER. P

214

J. BEARD.

Vertebraten nicht mehr so wie diejenigen der Hirnnerven zu
der Haut, respective den Seitenorganen ontogenetische Bezie-
hungen aufweisen, lasst sich beantworten. Derselbe durch
die Concentrirung des Kopfes oder Gehirnes liervorgerufene
Prozess, der an den iibrigen Bestandtheilen des Seitenorgan-
systemes so tiefgreifende Veranderungen hervorrief, namlich,
die Anbabnung einer einbeitlichen und directen (Gehirn-)
Leitung an Stelle der segmentalen, bat auch die urspriing-
lichen Hautbeziehungen der Seitenorganganglien (Spinal-
ganglien) allmahlig zum Schwinden gebracht. Nachdem
einmal die directe Leitung zwischen dera Gehirne und dem
Seitenorgansysteme des Rumpfes hergestellt, und die Inner-
vation durcb Spinalnerven zuriickgetreten war, so lag auch
keine Yeranlassung mehr fur Verbindungen zwischen Spinal-
nerven und Haut vor, und so konnen wir einsehen, dass die
nunmehr fiir ihre Sinnesorgane ebenfalls bedeutungslos gewor-
denen Seitenorganganglien des Vertebratenrumpfes immer
unabhangiger von den Seitenorganen und schliesslich den
Spinalnervenwurzeln, respective dem Riickenmarke, einverleibt
wei’den. Alles das ist zwar — es sei wiederliolt— vorlaiifig
noch durchaus hypothetisch, aber es gewanne schon in dem
Momente solideren Bodcn, wo in der Entwickelungsgeschichte
der Spinalganglien irgend eincs Vertebraten noch Anzeichen
von Hautverbindungen nachgewiesen wiirden, und wer mochte
behaupten, dass unsere Kenntnisse bereits hinreichen, um die
Existenz-moglichkeit einer derartigen Recapitulation a priori
verneinen zu konnen P Wie dem aber auch sei, diese auf
Thatsachen beruhende Hypothese zeigt, dass es angesichts
der so verwickelten Verhaltnisse doch nicht an Anhalts-
punkten fiir eine mogliclie Losung fehlt, und die Aussicht auf
eine mit Schwierigkeiten verbundene Losung ist doch erfreu-
licher, als die auf gar keine. Gar keiner Aussicht auf Losung
kommt aber die Auffassung Beard’s gleich, welche, da sich
zwischen Rumpf und Kopf zahlreiche Divergenzen ausgebildet
haben, die Vergleichbarkeit beider iiberhaupt n Erage stellt.”
The above extracts naturally fall into two divisions. In the
first place there is the question of the actual facts of develop-

MORPHOLOGICAL STUDIES.

215

ment which Dr. Eisig puts to me, and in the second place there
is the answer which Dr. Eisig from his standpoint gives to these
questions. With the latter I am here little concerned, for the
answer is purely hypothetical, as Dr. Eisig admits, and no one
can object to his right to establish as an admitted hypo-
thesis the view that the lateral sense organs were once con-
nected with spinal nerves. According to my ideas the evidence
is entirely wanting, and the quotations from three or four
authors1 which Dr. Eisig makes to show that even now spinal
nerves send branches to the sense organs situate in the trunk,
do not seem to me to affect the question ; for, as I shall else-
where show, they are all either vague or of a very doubtful
character, and as yet no one has figured these connections.

These remarks also answer his questions as to whence I know
that such connection was never the case. We know nothing
of such connection of spinal nerves with the sense organs of the
lateral line, either now or in the past, and any opinion one may
express in favour of such a view is only an assumption.

To the second question, whether the spinal ganglia are not
homologous with the sense-organ ganglia of the head, I think
the answer must be decidedly in the negative.

I regret to be compelled to this result, but I see no way out
of the conclusion that the spinal ganglia of the trunk are
homologous with those portions of the cranial ganglia which
take their origin in the similar position to the spinal, viz. just
outside the lips of the neural plate. I have never as yet seen
a trace of the sensory epithelium and ganglia of the sense organs
in the trunk region of a Vertebrate embryo. Here, of course,
I except the sense organs derived from the vagus which wander
into the trunk, as I have shown elsewhere (No. 6, p. 19), by
displacing the indifferent epiblast.

I have, moreover, never seen a trace of a sensory epithelium

1 The authors quoted are Julin (No. 39), Ransome and Thompson (No. 53),
and Ryder (No. 55). While this paper was passing through the press, the
supposed connection between spinal nerves and lateral nerve has been totally
refuted by Professor Dohrn (“ Studien, &c.,” No. xiii, ‘ Mittheil. a. d. Zool.
Station zu Ncapel,’ Bd. viii, Hft. ii).

216

J. BEARD.

in connection with the neural ganglia, i. e. in connection with
those ganglia in head and trunk which are formed just outside
the limits of the neural plate ; and, as Froriep and I have indi-
rectly shown, the lateral sense organ Anlagen in higher Ver-
tebrates show no disposition to leave their original home above
the gill-clefts, and to wander into the epiblastic Anlagen of the
neural ganglia, but force the latter, as it were, to come to them
to receive their contingent of nerve-cells.

Like Dr. Eisig I support, as the result of these researches,
Kleinenherg’s view of the homology of the spinal ganglia of
Vertebrates, and the parapodial ganglia of Annelids. But I go
further, and say that what in the sense given above may be
called the cranial neural ganglia of Vertebrates, are also mor-
phologically equivalent to parapodial ganglia of Annelids. I
also am fully prepared now to accept with Eisig the homology
of the branchial sense organs of Vertebrates with the Seiten-
organe of Annelida ; but from the nature of the case it will be
obvious that at present I cannot admit the unproved homology
of the “ Seitenorganen” ganglia of Annelids with the entire
parapodial ganglia of Annelids. To meet the conditions of the
Vertebrate head the parapodial ganglion must at some time or
other have divided into two parts, one remaining neural and
corresponding to the neural ganglia of Vertebrates, and one
becoming lateral above the gill-clefts (and connected with them),
which would correspond to the lateral sense-organ ganglia of
Vertebrates, and to the same ganglia of Annelids. At present
such a view would be merely speculative.

VIII. The Functional Distribution of the Cranial

Nerves.

The recent researches of Gaskell (No. 19 a, p. 58) lead him to
divide the anterior and posterior roots of each spinal nerve into
two sets of fibres, which are visceral and somatic respectively.

Somatic motor nerves are those fibres derived from the
anterior horn ; somatic sensory nerves are those derived from
the posterior horn ; while the motor visceral nerves arise in

MORPHOLOGICAL STUDIES.

217

the lateral horn and pass out with the other motor nerves in
the anterior root ; and the sensory visceral fibres take their
origin in Clark’s column and pass out with the posterior root.

Both sets of sensory fibres possess ganglia, the motor fibres
being unganglionated.

I do not propose to devote any great amount of space to the
examination of the bearings of Dr. Gaskell’s results on the
cranial nerves as given by himself, or as they appear to me ;
still, a few morphological conclusions can be drawn from those
researches just as my results may be of use to the physiologist.
The oculomotorius, trochlearis and abducens correspond mor-
phologically and physiologically, as van Wiihe (No. Gl), Hill
(2G), Gaskell (19), 1 and His (34) have insisted, to the motor
somatic roots of spinal nerves. They arise in the combination
of the anterior horn in the head, and they are distributed to
muscles of the somatic system. Thus one is faced at once by
the conclusion that the motor visceral fibres do not enter
anterior roots in the head, and, on the contrary, they pass
through the posterior roots, which are mainly sensory.

Now, these motor somatic fibres in the trunk develop as
direct outgrowths of the spinal cord, and as the ganglia which
form them lie in the cord they ought also to arise in the head
as direct outgrowths of cells in the brain, and in the homo-
logies of anterior root of spiual nerves. The latter is certainly
not the case, for they pass out with the posterior roots : and
the question arises. How do they develop in the head ? Either
the old course with anterior roots in the head never existed,
or it has been lost, and they have acquired new paths through
the afferent fibres of the posterior root.

Which of these things is really the case I cannot decide, for
as yet I have been unable to prove the first by the demon-
stration of an element of the posterior root of a cranial nerve
which develops as a direct outgrowth of cells from the brain.

1 Gaskell lias quite recently arrived at very different conclusions (‘Proc. Roy.
Soc.,’ Feb. 9th, 1888), which appear to be largely erroneous. I shall consider
them in the second part of this work, after Dr. Gaskell has published the com.
plete paper.

218

J. BEARD.

From what is known about the development of all other motor
nerves, we may expect that such is the case ; and I believe that
sooner or later it will be shown that these fibres, which are the
nerves to the muscles of the gill-clefts, do develop as direct
outgrowths of cells in the brain like the anterior roots of
spinal nerves.

When one also considers that to those four groups of nerves
distinguished by Gaskell there must be added a fifth ganglio-
nated sensory element connected with the lateral sense organs,
the exceedingly complicated nature of the problems presented
by the cranial nerves of any Vertebrate higher than Amphi-
oxus will be very evident.

Resume or Results.

The spinal ganglia of Vertebrates are formed as differentia-
tions of the inner layers of the epiblast just outside the limits of
the neural plate. As the result of the cutting out from the epi-
blast of these ganglionic elements an appearance is presented by
the epiblast which is left, to which Professor His gave the
name of “ Zwischenstrang.” This has no share in the formation
of the ganglia. “ The Zwisclienrinne ” of His has no existence,
but certain portions of the cranial ganglia, called here neural
ganglia, are developed from the epiblast before closure of the
neural tube in exactly the same way as the spinal gan-
glia. These portions of cranial ganglia are more or less
homologous with spinal ganglia, possibly only with the sympa-
thetic portion of the spinal ganglia Anlagen. After separa-
tion from the epiblast the neural cranial ganglia and the
spinal ganglia get carried up with the closing in of the neural
tube, and come to lie between its lips, but are quite distinct
from the central nervous system, and the line of boundary
between the two can always be distinguished. After the
closure of the epiblastic folds the Anlagen grow out of their
position between the lips of neural tube, which then also
closes. They grow downwards and to the sides of the neural
tube, and acquire their first and only connection with it by the

morphological studies.

219

probable growth of fibres from the gauglia into the central
nervous system. The neural cranial ganglia also grow to-
wards the lateral epiblast at the level of the notochord and
fuse with it. Here are the Anlagen of the lateral or branchial
sense organs of Froriep and myself. From this fusion in all
Vertebrates form-elements pass into the cranial gauglia; these
form-elements I distinguish as lateral ganglia. The parapodial
gauglia of Annelids appear to be homologous with the spinal
ganglia of Vertebrates, as Kleinenberg suggested, and also
more or less with the neural cranial ganglia.

The anterior roots of cranial and spinal nerves arise as out-
growths of ganglia situate in the central nervous system. To
form them cells leave the nervous system, and are distributed
in the nerve. All the anterior roots at first contain many
nuclei, which are of nervous and not parablastic origin. These
statements on the anterior roots are only a confirmation of
Balfour’s researches.

In addition to the four elements of the anterior and posterior
roots, two ganglionated and sensory, two motor and ungau-
glionated, distinguished by Gaskell, Hill, and partially by His,
the cranial nerves contain a fifth element, derived from the lateral
or branchial sense organs. Such are, in very brief form, the
main results of the researches recorded in the preceding paper.

It is with more than ordinary feelings that I desire to record
here my most heartfelt gratitude to Professor Wiedersheim, in
whose laboratory I carried out the above researches, for the
generosity and kindness with which he in many ways supported
my work. I owe him many thanks for his advice and criticism,
and for the use of his valuable library, and, not least, for the
gift of various material which was of great use to me.

220

J. BEAM).

List ot the Literature cited in this Paper.

1. Balfour, F. M. — “ On the Development of Spinal Nerves in Elasmo-

brancli Fishes,” ‘ Phil. Trans.,’ vol. 166, 1876.

2. Balfour, F. M. — ‘ A Treatise on Comparative Embryology,’ vol. ii, 1881.

3. Balfour, F. M. — ‘A Monograph of the Development of Elasmobranch

Fishes,’ 1878.

4. Beard, J. — “ On the Segmental Sense Organs of the Lateral Line, and

on the Morphology of the Vertebrate Auditory Organ,” ‘ Zool.
Anzeiger,’ 1884, Nos. 161, 162.

5. Beard, J. — “ On the Cranial Ganglia and Segmental Sense Organs,”

‘Zool. Anzeiger,’ No. 192, 1885.

6. Beard, J. — “The System of Branchial Sense Organs and their Asso-

ciated Ganglia in Ichthyopsida,” * Quart. Journ. Micr. Sci.,’ Nov., 1885.

7. Beard, J. — “ The Ciliary or Motoroculi Ganglion and the Ganglion of

the Ophthalmicus Profundus in Sharks,” ‘Anat. Anz.,’ Bd. ii, 1887,
Nos. 18 and 19.

8. Beard, J. — “ The Old Mouth and the New,” ‘ Anat. Anzeiger,’ No. 1,

1888, Bd. iii.

9. Bedot, M. — “ Recherches sur le developpement des nerfs Spinaux chez

les Tritons,” ‘ Recueil Zool. Suisse,’ 1884, t. i, p. 161.

10. Beraneck, E. — “ Etude sur les rcplis Medullaires du Poulet,” ‘ Recueil

Zool. Suisse,’ t. iv, 1887.

11. Beraneck, E. — ‘ Recherches sur le developpement des nerfs cranieus chez

les Lezards,’ t. i, 1884, p. 519.

12. Bernard, Cl. — ‘ Lemons sur le systeme nerveux,’ 1858.

13. Doiirn, A. — “Studien zur Urgeschichte des Wirbelthier Korpers,”

‘Mittheil. a. d. Zool. Station zu Neapel;’ X. ‘Zur Phylogenese des
Wirbelthierauges,’ Bd. vi. Heft 3.

14. Eisig, II. — “Die Seitenorgane und der beckerformigen Organeu der

Capitelliden,” ‘ Mitthiel. a. d. Zool. Station zu Neapel,’ Bd. i, 1879.

15. Eisig, II. — ‘ Die Capitelliden, nebst Untersuchungen zur Vergleickenden

Anatomic und Physiologie. Fauna und Flora des Golfes von Neapel,’
XVI. Monographic, 1887.

16. Froriep, A. — “Ueber ein Ganglion des Hypoglossus und Wirbelaulagen

in der Occipital Region,” ‘ Archiv fiir Anat. u. Entwickeluugs-
geschichte,’ 1882, p. 280.

17. Froriep, A. — “ Ueber Anlagen von Sinnesorganen am Facialis, Glosso-

pharyngeus, und Vagus, &c.,” ‘Archiv fiir Anal, und Entwickelungs-
gesch.,’ 1885, p. 1.

MORPHOLOGICAL STUDIES.

221

18. Froriep, A. — “ Bemerkungen zur Frage naclx der Wirbeltkeorie des

Kopfskelettes,” ‘ Anat. Anz.,’ Bd. ii, No. 27, 1887.

19. Gaskell, W. H. — “The Structure, Distribution, and Function of the

Nerves which Innervate the Visceral aud Vascular System,” ‘ Journ.
of Physiology,’ vol. vii. No. 1.

20. Gegenbaur, C. — “ Ueber die Kopfnerven von Hexanchus uud ilir Ver-

haltniss zu Wirbeltheorie des Schadels,” ‘ Jenaische Zeitschr. f. Med.
u. Naturwiss.,’ Bd. vi, 1871.

21. Gegenbatjr, C. — “Die Metamerie des Kopfes und die Wirbeltheorie des

Kopfskelettes,” ‘Morphol. Jahrb.,’ Bd. xiii, Heft 1.

22. Goette, A. — ‘ Entwickelungsgeschichte der Unke,’ Leipzig, 1875.

23. IIertwig, 0. — ‘Lehrbuch der Entwickelungsgeschichte des Menscheu

und der Wirbelthiere,’ Bd. ii, 1888.

24. IIensen, V. — “ Zur Entwickcluug des Nervensystems,” ‘ Virchow’s

Archiv,’ No. xxx.

25. IIensen, V. — “ Beobachtungen iiber die Befruchtung und Entwickelung

des Kaninchens und Meerschweinchens,” ‘ Zeit. f. Anat. u. Entwickl.,’
Bd. i, 1876, zwei Theile.

26. Hill, A. — ‘ The Plan of the Central Nervous System,’ Cambridge, 1885.

27. Hill, A. — ‘ Abstract of Three Lectures on the Brain Mechanism of Sight

and Smell,’ London, 1886.

28. His, W. — “Untersuchungen iiber die Entwickelung des Wirbelthier-

leibes,” * Die erste Entwickelung des Hiiknckens im Ei,’ Leipzig,
1868.

29. His, W. — “Ueber die Anfiinge des peripherischen Nervensystems,”

‘ Archiv fur Anat. u. Entwickelungsgesch.,’ 1879, and also 1882.

30. His, W. — ‘ Anatomie menschlichcr Embryonen,’ 1880-85.

31. His, W. — ‘ Briefe iiber unsere Korperform,’ Leipzig, 1875.

32. His, VV. — “ Zur Geschichte des menschlicben Iliickenmarkes uud der

Nervenwurzeln.” * Abhandln. der Math. Physk. Classe.der Kgl. sachs.
Gesellschaft der Wissenschafteu,’ Bd. xiii, No. 6, 1886.

33. Ills, W. — “Die Entwickelung der ersten Ncrvenbahnen beim rnensck-

lichen Embryo. Uebersichtliche Darstellung,” ‘ Archiv f. Anat. u. Ent-
wick.,’ 1888, Heft 1.

34. His, W. — “ Die morphologische Betrachtung der Kopfnerven. Eine

kritische Studie,” ‘Arch. f. Anat. u. Entwick.,’ 1888, Heft 1.

35. Hoffmann, C. K. — “ Zur Ontogenie der Knochenfische,” ‘ Archiv f.

Micro. Anat.,’ vol. xxiii.

36. Hoffmann, C. K. — “Zur Ontogenie der Knochenfische,” ‘Verhand.

Konenkl. Akad. van Wetenschappen,’ Bd. xxiii, 1882.

222

J. BEARD.

37. Hoffmann, C. K. — “ Weitere Untersuchungen zur Entwickelungs-

gescbichte der Reptilien,” ‘Morphol. Jahrbucb,’ Bd. xi, 1880.

38. Johnson and Sheldon. — “ Notes on the Development of tbe Newt.,5’

‘Quart. Jouru. Micr. Sci.,’ June, 1886.

39. Julin, C. — III. “ De la valeur morpbologique du nerf lateral du Petro-

myzon,” ‘ Acad. Hoy. de Belgique Extrait des Bulletins,’ 3me serie,
tome xiii, No. 3, 1887.

40. Kastschenko, N. — “ Das Scblundspaltengebiet des Hiibucbens,” ‘ Arcbiv

f. Anat. und Entwick.,’ Hefte 4 u. 5, 1887.

41. Kleinenberg, N. — “Die Enstekung des Annelids aus der Larve von

Lopadoryncbus. Nebst Bemerkungen iiber die Entwickelung anderer
Polycbaeten,” ‘ Zeitschr. f. wiss. Zoologie,’ Bd. xliv, Hefte 1 and 2.

42. Kolliker, v. A. — “ Zur Entwickelung des Auges und Gerucbsorganes

menscklicber Embryonen,” ‘ Festschrift f. d. Universitie Wurzberg,’
1883.

43. Kolliker, A. v. — ‘ Entwickelungsgescbicbte des Menscben und der

hohereu Tbiere,’ 2te Auflage, 1879.

44. Kolliker, A. v. — “ Note sur le developpemeut des tissus cliez les Batra-

cbiens,” ‘Ann. Sci. Nat.,’ 3 serie, t. 6, 1864.

45. Marshall, A. M. — “On tbe Early Stages of tbe Development of Nerves

in Birds,” ‘ Journ. of Anat. and Physiol.,’ vol. xi, 1877, p. 89.

46. Marshall, A. M. — “The Cranial Nerves in the Chick,” ‘ Quart. Jouru.

Micr. Sci.,’ vol. xvii, 1878.

47. Marshall, A. M. — “ The Morphology of tbe Vertebrate Olfactory Organ,”

‘ Quart. Journ. Micr. Sci.,’ vol. xix, 1879.

48. Marshall, A. M. — “ On the Head Cavities and Associated Nerves in

Elasmobranchs,” ‘Quart. Journ. Micr. Sci.,’ vol. xxi, 1881.

49. Marshall and Spencer. — “The Cranial Nerves of Scyllium,” ‘Quart.

Journ. Micr. Sci.,’ July, 1881.

50. Marshall, A. M. — “ The Segmental Value of the Cranial Nerves.”

Reprinted from the 1 Jouru. of Anat. and Physiol.,’ 1882.

51. Onodi, A. D.— “Ueber die Entwickelung der Spinalganglien und der

Nervenwurzeln,” ‘ Internat. Monatschrift f. Anat. u. Histologie,’ Bd. i.

52. Onodi, A. D. — “ Ueber die Entwickelung des sympath. Nervensystems,”

‘ Archiv f. mikrosk. Anat.,’ xxvi, 1886.

53. Ransome and Thompson. — “ On the Spinal and Visceral Nerves of

Cyclostomata,” ‘Zool. Anz.,’ 1886, No. 227-

54. Remak, R. — ‘ Untersuchungen iiber die Entwickelung der Wirbelthiere,’

Berlin, 1850-55.

MORPHOLOGICAL STUDIES.

223

5b. Ryder, John A. — “A Contribution to the Embryography of Osseous
Fishes,” ‘ Ann. Rep. Comm, of Fish and Fisheries,’ 1882, Washington,
p. 54.

50. Sagemehl, M. — “ Untersuchungen iiber die Eutwickelung der Spinal-
nerven. Gekronte Preisschrift,” Dorpat, 1882.

57. Semper, C. — “Das Urogenitalsystem der Plagiostomeu u. seine Bedeu-

tung fur die hoheren Wirbelthiere.” ‘ Arbeiten a. d. Zool.-Zoot.
Institut zu Wurzburg, ’ Bd. ii, 1875.

58. Shipley, A. E. — “Some Points in the Development of Petromyzon

fluviatilis,” ‘Quart. Journ. Micr. Sci.,’ Jan., 1887.

59. Spencer, W. B. — “Notes on the Early Development of Rana tem-

poraria,” ‘Quart. Journ. Micr. Sci.,’ Supplement, 1885.

60. van Wijhe, J. W. — ‘ Ueber die Mesodermsegmente und die Entwick-

elung der Nerven des Selachierkopfes,’ Amsterdam, 1882.

61. van Wijhe, J. W. — “ Ueber die Kopfsomite und die Pbylogenie des

Geruchorganes der Wirbelthiere,” ‘ Zool. Anz.,’ 1886, No. 236.

62. van Wijhe, J. W. — “ Ueber Somiten und Nerven im Kopfe von Vogel

und Reptilienembryonen,” ‘ Zool. Anz.,’ 1886, No. 237.

63. van Wijhe, J. W. — “ Ueber die Entwickelung des Exkretionssystems

und anderer Organe bei Selachiern,” ‘ Anat. Anz.,’ Bd. iii, 1888, pp.
74 — 76, Nos. 2 and 3, Jan. 18.

64. Wiedersiieim, R. — ‘ Lehrbuch der Vergleichenden Anatomie der Wir-

belthiere,’ 2te Auflage, 1886.

224

J. BEARD.

DESCRIPTION OF PLATES XVI— XXI,

Illustrating the Memoir by Dr. Beard on “The Development
of the Peripheral Nervous System of Vertebrates. Part I.”

List of References.

1. Hi, v, vii, &c. Olfactory, motoroculi, trigeminal, facial, &c., nerves.
a. Anterior root. al. c. Alimentary canal, au. o. Auditory organ. Or. Brain.
Or. gl. Branchial or lateral ganglion, cl. Cleft, e. Epiblast. f. Or. fore-brain.
f. gl. Facial ganglion, gl. Ganglion, gl. gl. Glossopharyngeal, g. A. Gan-
glionic Anlage. h. Or. Hind-brain. h. c. Head-cavity, or head-somite.
i. e. Indifferent epiblast. m. Mesocephalic ganglion, m. Or. Mid-brain, me.
Mesoblast. n. Notochord, n. s. Nervous system, olf Olfactory, p. g.
Parapodial ganglion, s. e. Neuroepithelium, sp. Spinal, sp. c. Spinal cord.

All figures, except Fig. 64, are drawn under Zeiss’s camera lucida. The
magnification is indicated by such letters as Z. D, oc. 2, which signifies Zeiss’s
objective D, ocular No. 2. The objectives used were those of Zeiss and Hart-
nack, and are distinguished as Z. and H. respectively. Except Figs. 90 — 93,
96, 100, 101, which are from longitudinal frontal sections, the figures represent
transverse sections.

All figures are reduced in the plates to two-thirds of their original size.
PLATE XVI.

Figs. 1 — 3. — Sections through the trunk of a Torpedo embryo. Z. D, oc. 2.
Fig. 4. — Section through the head region of Torpedo ocellata. Z. D,
oc. 2.

Fig. 5. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Figs. 6 and 7. — Sections, head region, T. ocellata. Z. D, oc. 2.

Figs. 8 and 9. — Sections, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 10. — Section, mid-brain region, T. ocellata. Z. D, oc. 2.

Fig. 11. — Section, head region, T. ocellata. Z. D, oc. 2.

Fig. 12. — Section, head region, T. ocellata. Z. D, oc. 2.

Fig. 13. — Section, trunk region, T. ocellata. H. 8, oc. 2.

Fig. 14. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 15. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 16. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 17. — Section, trunk region, T. ocellata. H. 8, oc. 2.

Fig. 18. — Section, head region, T. ocellata. Z. D, oc. 2.

Figs. 19 and 20. — Sections, brain region, of two Torpedo embryos.
Origin of olfactory neural ganglion. Z. D, oc. 2.

MORPHOLOGICAL STUDIES.

225

PLATE XVII.

Fig. 21. — Section through facial ganglion Anlage, T. ocellata. Z. D,
oc. 2.

Fig. 22. — Section, trunk region, Mustelus 1 sc vis. Z. D, oc. 2.

Fig. 23. — Section, trunk region, M. laevis. Z. F, oc. 2.

Fig. 24. — Section, vagus ganglion, T. ocellata. Z. D, oc. 2.

Fig. 25. — Section, auditory organ and ganglion, T. ocellata. Z. D, oc. 2.
Fig. 26. — Section, spinal cord region, Pristiurus. Z. F, oc. 2.

Fig. 27. — Section, auditory region, T. ocellata. Z. F, oc. 2.

Fig. 28. — Section, trunk region, Pristiurus. Z. D, oc. 2.

Fig. 29.— Section, vagus region of head, T. ocellata. Z. D, oc. 2.

Fig. 30. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 31. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Figs. 32 — 36. — Sections, trunk region, Pristiurus. Z. D, oc. 2.

The order from before backwards is 33, 34, 35, 36, 32.

PLATE XVIII.

Fig. 37. — Section, anterior head region, T. ocellata. Z. C, oc. 2.

Fig. 37 — The small figure marked out in preceding section under high
power, to show Anlage of a cranial ganglion. Z. F, oc. 2.

Fig. 38. — Part of a section of head region, T. ocellata. Z. F, oc. 2.
Fig. 39. — Section, head region, T. ocellata. Z. F, oc. 2.

Figs. 40 — 42. — Portions of sections through head region of three Torpedo
embryos. Z. F, oc. 2.

Fig. 43. — Section, trunk region, Mustelus. Z. D, oc. 2.

Fig. 44. — Portion of a section, trunk, T. ocellata. Z. F, oc. 2.

Fig. 45. — Portion of a section of Mustelus through vagus region. Z. F,
oc. 2.

Fig. 46.— Section through auditory region of a Chick embryo. H. 9, oc. 2.
Fig. 47. — Section, head region, T. ocellata. Z. F, oc. 2.

Fig. 48. — Section, mid-brain region, Mustelus. Z. F, oc. 2.

Fig. 49. — Section, region of anus, Mustelus. Z. D, oc. 2.

Fig. 50. — Section, region of head, Mustelus. Z. C, oc. 2.

Fig. 51. — Section, trunk region. Chick, eight somites. Z. F, oc. 2.

PLATE XIX.

Fig. 52. — Section, anus region, T. ocellata. Z. D, oc. 2.

Fig. 53. — Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 54.— Section, trunk region, T. ocellata. Z. D, oc. 2.

Fig. 55. — Section, trunk region, Chick, eight somites. Z. F, oc. 2.

226

J. BEARD.

Fig. 56. — Section, trank region, T. ocellata. Z. D, oc. 2.

Fig. 57. — Section, spinal cord region, Scyllium canicula. The epiblast
is not represented. Z. I), oc. 2.

Fig. 58. — Section through developing anterior root of a spinal nerve,
Mustelus. Z. F, oc. 2.

Fig. 59.— Section through tail region, T. ocellata. H. 8, oc. 2.

Fig. 60. — Section of developing anterior root (near anus), Mustelus.
Z. D, oc. 2.

Fig. 61. — Section through trunk region, T. ocellata. Z. D, oc. 2.

Fig. 62. — Section, head region, Chick with no somites. Z. D, oc. 2.

Fig. 63. — Section in region of infundibulum, Chick, nine somites. Z. D, oc. 2.

Fig. 64. — Copy of Kleiuenberg’s figure of developing parapodial ganglion
(yj. g.) of Lopadorynchus. The sketch has been turned through 180 degrees.

Figs. 65 and 66. — Section, head region, Chick embryo with nine somites.

Fig. 67. — Section, head region, Chick embryo with four somites. Z. D, oc. 2.

Figs. 68 and 69. — Two sections through trunk and head regions respectively
of a Chick embryo with four somites. Z. D, oc. 2.

PLATE XX.

All the figures on this Plate are from Chick embryos. All are under Zeiss’s
D, oc. 2.

Figs. 70 — 76. — Series of sections through trunk region, from behind for-
wards, of an embryo with six somites.

Figs. 77 — 79. — Series of sections from before backwards through brain
region of an embryo with nine somites.

Fig. 80. — From same embryo, but through first somite.

Fig. 81. — From same embryo, but through end of second somite.

Figs. 82 — 84. — Three sections through brain region of an embryo with
seven somites.

Fig. 85. — Section through spinal region of same embryo.

Figs. 86 and 87. — Two sections through brain region of two embryos with
no somites.

Fig. 88. — Section, hind-brain region of a Chick embryo with ten somites.

Fig. 89. — Section, trunk region of a Chick with eight somites.

PLATE XXI.

Figs. 90 and 91. — Two longitudinal frontal sections through the head of a
three-days’ Chick embryo, showing the rudiments of branchial sense organs.
Z. A, oc. 2.

m = Mesocephalic ganglion and sense organ.
v. Trigeminus „ „

wt. Facial ,, ,,

ix. Glossopharyngeal „ „

MORPHOLOGICAL STUDIES.

22 7

Fig. 92. — Trigeminus ganglion and sense organ from Fig. 91, highly
magnified. Z. F, oc. 2.

Fig. 93. — Mesocephalic ganglion and its sense organ from Fig. 91, highly
magnified. Z. F, oc. 2.

Fig. 91. — Section in front of a gill-cleft of T. ocellata. Z. A, oc. 2.

Fig. 94a. — The black portion of this section highly magnified to show growth
and extension of the lateral sense-organ epithelium.

Fig. 95. — Section behind a gill-cleft of T. ocellata. Z. A, oc. 2.

Fig. 95 a. — The blackened portion of this section highly magnified, to show
growth and extension of lateral sense-organ epithelium. Some cells wander
into mesoblast to form ganglion-cells.

Fig. 96. — Glossopharyngeal ganglion and its sense organ in three-days’
Chick, from Fig. 90. Z. C, oc. 2.

Fig. 97. — Section of hind-brain of a Chick embryo with nine somites,
showing “ Zwischenstrang” ( Z .) and its relation to ganglion Anlage,
Z. F, oc. 2.

Figs. 98 and 99. — Sections of trunk region of Chick embryo of second
day. Z. F, oc. 2.

Fig. 100. — Portion of longitudinal vertical section of mid-brain of a lizard
embryo (L. agilis), showing origin of oculomotorius and its “ ganglion ” in
brain. Z. D, oc. 2.

Fig. 101. — Portion of longitudinal vertical section of the auditory epithe-
lium of a lizard embryo (L. agilis). Z. F, oc. 2.

Fig. 102. — Section, trunk region of a Chick embryo of second day. Z. F, oc. 2.

Fig. 103. — Section through glossopharyngeus nerve and ganglion of an
advanced Torpedo ocellata embryo. Z. C, oc. 2.

.

’

Jkcr.Joa nc %6XX!X,NSM XW.

. ' -

«9 *»*

i-^,

TORPEDO.

OCELLATA.

F Hufb.Litb' “EdinT

a. a.

I

i.e

XV///.

•" . •• 4 - .

^9Msh

;:XrvV *&t,r

® kftss>

%. 4)1§V& i W®Js®

% &&, -

i\ /Sfc *•> ' / 49

®<«£

J Beard del

Fii5s 37-45 ELASMOBRANCHII.

Fig? 46-51 CHICK.

F Huth.Lith' ESm1

'tffierjfownu T^.WXJSMjT

mm.

r. • • •- * • 1 • .

M><-0

P Huth,LithT Ediif

J Beard del.

are from CHICK EMBRYOS

1/ot JJ/Z.VS,^ XX/

J. Beard del.

Figs 9 0-93,96-99, 102 CHICK.

Fig? 94-95* 103 T.OCELLATA. Fig? 100,101 LACUTA.

¥ HuOi,Lithir Edinr

Jdcr. ■fowrn'l/oi XXIX, KS&t. XXII

u ' • ' • *. ••

E. A Min chin del

F Hufti L.thT Edin*

Mu yrjfa

Fuf. /.

■.Tv'JSVi’!

Fia. 10.

F E.Bcddard del

F Huth, Lithr Edmr

PERIPLANETA ORIENTALIS.

229

Note on a New Organ, and on the Structure of
the Hypodermis, in Periplaneta Orientalis.

By

Edward A. Mincliin,

Keble College, Oxford.

With Plate XXII.

The organ which I have to describe, and of which I have
been unable to find an account in the latest works on the
anatomy of the Cockroach, consists of two pouch-like invagi-
nations of the cuticle lying close on each side of the middle
line, between the fifth and sixth terga of the dorsal surface of
the abdomen. In the normal condition these are covered by
the fifth tergum, but when this is removed they are seen as
two shallow pouches lying beneath the transparent intertergal
membrane and opening to the exterior by two slit-shaped
openings, which face backwards (fig. 1). They are lined by
a continuation of the chitinous cuticle, which forms within
the pouches numerous stiff, branched, finely-pointed hairs,
beneath which, i. e. on the side towards the body- cavity, are
numerous glandular epithelial cells. Fig. 2 shows a section
longitudinal to the body of the animal passing through one of
the pouches. The figs, v and vi are placed above the fifth and
sixth terga respectively ; d. denotes the upper dark layer, and
t. the lower transparent layer of the chitinous cuticle, and h.
the hypodermis of at least two layers, and resting on the
basement membrane (6. m.). At p., the most posterior point
of the fifth tergum, the transparent layer of the cuticle and the
uppermost layer of the cells of the hypodermis are continued

VOL. XXIX, PART 3. NEW SER. Q

230

EDWARD A. MINOHIN.

into the intertergal membrane ( i.m .), which again becomes
continuous with the tergum immediately behind, the whole
forming a continuous chitinous investment. The letter a., figs.
1 and 2, denotes the most anterior extremity of the sixth tergum,
to which the longitudinal tergal muscles (m. 1 and m. 2) are
attached ; r. is a ridge close behind this part, the space between
a. and r. being usually overlapped by the hinder part of the
fifth tergum. These parts are present between every pair of
terga, but between the fifth and sixth are found in addition the
peculiar glandular pouches (P., figs. 1 and 2) above mentioned.
As may be seen, they are lined by a cuticle continuous with
that of the intertergal membrane (c. 1.), which is produced into
numerous stiff hairs, which bend towards the opening of the
pouch. Below this cuticle is a layer of small flattened cells,
with somewhat irregularly-shaped nuclei ; this layer is evidently
continuous with the hypodermic layer beneath the intertergal
membrane (figs. 2 and 4, u. /.). These cells are crowded very
close together, and their outlines are very hard to make out.
Beneath this layer again comes a number of columnar cells with
very distinct outlines, forming one layer on the dorsal side,
and two or more on the ventral side of the pouch (/. /.). These
cells have large rounded nuclei, each with a distinct nucleolus,
and their cell contents are very granular. Outside all is a
basement membrane continuous with that of the hvpodermis.
The tergal muscles pass immediately under the pouch, but I
have satisfied myself, both by sections and by dissecting care-
fully, that the pouches have no special muscles. In fig. 3 are
shown some of the hairs highly magnified. They are hollow
and swollen at the base, and in a chloride of gold preparation
processes of the cells of the upper layer (u. I .) can be seen
running into them from below; sometimes nuclei are to be seen
inside the swollen base. These hairs exactly resemble those
found lining the abdominal stigmata, which are branched in
exactly the same manner. Their chitinous nature is evident
from the fact that if one of the pouches be macerated for a
sufficient time in caustic potash (30 per cent.) nothing is left
but the hairs and the cuticle on which they stand.

PEEIPLANETA ORIENTALIS.

231

It is clear from the above description that these pouches are
merely invaginations of the body wall between the terga, round
which the hypodermis has greatly proliferated, and this is
borne out by their development. Fig. 4 represents a section
from a larva of about 8 mm. in length; there is a slight de-
pression lined by a few chitinous hairs, which are branched
at the base, and the cells of the upper layer are somewhat
crowded. There are only two of the large columnar cells
to be seen ( l . 1.), which are here rounded, and probably become
columnar by flattening against one another. A series of
gradations can easily be made out between this early stage
and the adult stage shown in fig. 2. The fact of the hairs
being branched at the base, and the crowding of the upper
layer, seems to indicate that the hairs develop as simple foldings
of the cuticle. The lower columnar cells seem to be modified
hypodermic cells, but it should be noticed that they have a
distinct nucleolus, which I have been unable to make out in
other hypodermic cells ; in fact their nuclei exactly resemble
in shape, size, and appearance those of the fat-bodies.

As to the function of this organ, I have no doubt that it is
glandular, and it is probable that it is a stink-gland, though I
have not been able to satisfy myself of this. The interior of
each pouch is usually filled with granular matter, which stains
with haematoxylin but not with borax carmine. I have no
doubt this is the secretion. The function of the hairs, into
which the secretion must in the first place pass, may perhaps
be to act as ducts, and at the same time to diffuse the odour
by offering a larger surface for evaporation.

Structure of the Hypodermis. — Miall and Denny, in
their valuable work on ‘the Cockroach/ describe the hypo-
dermis as a single layer of cells, each cell corresponding to
a polygonal area of the chitinous cuticle, and resting on a
basement membrane. Below this are described here and there
large nerve-end cells which send up fine processes into sensory
hairs on the upper surface. I am unable, however, to agree
with this description as far as the dorsal sui’face is concerned.
From my sections (see figs. 2 and 4) I find the hypodermis, in

232

EDWARD A. MINCHIN.

the older stages at all events, to consist here of two layers
everywhere except where the cuticle is folded to form an arti-
culation ; in these parts only the upper layer remains. In
certain places the cells of the lower layer become giant-cells
(g. c.), which have large nuclei and very granular cell contents,
and are provided with processes, and are without doubt ganglion-
cells. They are scattered over each tergum, but are extremely
abundant in the fore part of the tergum, i. e. that part which
is overlapped by the tergum next in front, where they form an
almost continuous layer. Hence it does not seem possible
that they should always correspond with hairs. If a Cock-
roach be dissected in strong spirit, these cells can be seen
with a low power forming irregular brownish patches in the
fore part of each tergum, and with care they can be scraped
off and floated on to a slide and stained and mounted in situ,
when they present the appearance shown in fig. 5. They are
large irregular cells, more or less overlapping one another and
provided with processes, usually two or three. Between them
are seen ordinary nucleated cells of the lower layer of the
hypodermis, with which they are continuous. Fig. 6 shows a
single one of these giant-cells. Though most numerous in the
fore part of each tergum, they are also found scattered through-
out its whole extent.

Hence I believe the hypodermis (fig. 7) to have the following
structure : an upper regular layer of cells, which corresponds
to the polygonal areas of the cuticle, and is continuous through-
out the body, and is continued on to the tracheae ; and a
lower, very irregular layer, which occasionally seems to form
two layers, and is wanting where the cuticle is folded to form
an articulation, and the cells of which are in places directly
modified into nerve-end cells. These nerve-end cells are prob-
ably connected with a seta where the terga are exposed, but
where they are overlapped seem merely to be connected with
small papillae. Below these two layers is placed the basement
membrane ( b . to.), which is apparently a thin, homogeneous
membrane.

My investigations were carried on in the Morphological

PER1PLANETA ORIENTALIS.

233

Laboratory at Oxford, and in conclusion I have to express my
thanks to Dr. S. J. Hickson, Mr. G. C. Bourne, and my tutor,
Mr. E. B. Poulton, for much kind help and advice.

EXPLANATION OF PLATE XXII,

Illustrating Mr. Edward A. Minchin’s paper "Note on a New
Organ, and on the Structure of the Hypodermis, in Peri-
planeta orientalis.”

Letters of Reference.

I — X. The ten terga of the abdomen. P. The glandular pouches, a. The
anterior extremity of the sixth tergum. r. The transverse ridge across the
sixth tergum, marking the part of it overlapped by the fifth tergum. d. The
dark chitinous layer of the cuticle, t. The transparent ditto, h. The hypo-
dermis. b. m. The basement membrane of the hypodermis. p. The posterior
extremity of the fifth tergum. c. 1. Chitinous lining of the gland, u. 1.
Upper layer of small cells of ditto. 1. 1. The lower layer of columnar cells of
ditto, m. 1 and m. 2. Tergal muscles, g. c. Large ganglion-cells of the
lower layer of the hypodermis. i. m. Intertergal membrane.

Fig. 1. — A diagram of the dorsal surface of the abdomen of the £ Cock-
roach, the fifth tergum being removed to show the two glandular pouches
placed under the intertergal membrane between the fifth and sixth terga,
which is represented dotted.

Fig. 2. — A longitudinal section through the adjacent parts of the fifth and
sixth terga of an adult Cockroach, somewhat diagrammatic.

Fig. 3. — Some of the chitinous hairs lining the glands, highly magnified.

Fig. 4. — Longitudinal section through a gland of a larva of about 8 mm. in
length.

Fig. 5. — Ganglion-cells of lower layer of hypodermis, mounted in situ.

Fig. 6. — Single ganglion-cell of lower layer of hypodermis.

Fig. 7. — Ideal section of hypodermis and cuticle.

MwfowrrMt.mjSjh. II//

E A Minchm del

STRUCTURE OP UROCH/ETA AND DTCHOGASTER.

235

On Certain Points in the Structure of Urochseta,
E. P., and Dichogaster, nov. gen., with further
Remarks on the Nephridia of Earthworms.

By

Frank. E. Beddard, HI. .4.,

Prosector to the Zoological Society of Loudon, and Lecturer on Biology at
Guy’s Hospital.

With Plates XXIII and XXIY.

I. The Structure of Urochseta.

M. Perrier’s elaborate memoir (221) upon the structure of
this worm leaves little to he done in the way of general ana-
tomy. All the principal points which are of importance in
the systematic grouping of Earthworms are thoroughly de-
scribed and figured, with the sole exception of the female
reproductive apparatus, which was not present in the examples
studied by him. Perrier has also given a most detailed
description of the vascular system down to the minutest
ramifications, which forms one of the most complete accounts
extant of the Annelid circulatory organs. The method of
study adopted by M. Perrier was almost entirely that of
laborious dissection, and the results which he has obtained by
this means are undoubtedly striking. The elucidation of
many points in the anatomy of Earthworms demands, how-
ever, a recourse to the section-cutting method, which has been
adopted by myself in studying this Earthworm. I have,
therefore, been able to add some few facts to what is already
known, thanks to Perrier’s researches, of the anatomy and
histology of Urocheeta.

1 The numbers enclosed in brackets refer to the “ List of Memoirs ”
on pp. 279, 280.

236

FRANK E. BEDDARD.

§ Integument.

I have elsewhere (4) criticised Perrier’s figures of the epi-
dermis, and have now to make some remarks upon the pene-
tration of blood-capillaries into the epidermis.

The vascularity of the epidermis in Annelids was first made
known by Lankester (19) in Hirudo; subsequently Bourne
(12) showed that in all the Gnathobdellidae the epidermis was
traversed by blood-capillaries. Claparede (13), and later
Horst (17) and v. Mojsisovics (21), figured blood-capillaries in
the clitellum of Lumbricus, but did not find them in the
general epidermis. The first record of the presence of intra-
epidermic blood-capillaries in an Earthworm is by myself (5)
in Megascolex; subsequently (6) I found the same thing in
Perichaeta and Perionyx. The figures of Vejdovsky
(29), Rosa (25), and Benham (9, No. 3) show that the epi-
dermis of Criodrilus is also vascular. I have now to state
that in Urochaeta blood-capillaries penetrate between the
cells of the epidermis. In the Leeches and in Criodrilus
the blood-capillaries form loops in the epidermis, but in Uro-
chaeta I could never trace a returning limb of the capillary
which entered the epidermis. Judging from Yejdovsky’s
figures (pi. viii, figs. 16, 17) of Limnodrilus, the blood-
capillaries which enter the epidermis of that worm appear
to end abruptly in the same way.

Quite recently the brothers Sarasin (27) have described the
penetration of blood-capillaries into the epidermis of Peri-
cliaeta (without referring, I may remark, to my own record of
this fact, which may, however, have been inaccessible), which
they furthermore observed to open on to the surface
of the body, thus putting the blood-vascular system into
communication with the exterior. This, if true, is a most
remarkable fact. I cannot, however, pending the publication
of their more detailed account, accept it. The blood-
capillaries of Urochaeta reach to the very cuticle, but
there they stop. Furthermore, the following appears to be
an argument against the free communication of the integu-

STEUCTUEE OP UEOCH^TA AND DICHOGASTEE.

237

mental blood-capillaries with the surrounding medium. These
capillaries were perfectly obvious since they were gorged with
blood ; asuming for the present that they open by means of
pores, it would no doubt be the case, as the Sarasins suggest,
that capillary atti’action would prevent the blood from leaving
the body. But when the body of the worm is contracted by
the preservative fluid the blood would surely be driven out
through the pores. Nor can it safely be said that the con-
traction of the epidermic cells would be sufficient to occlude
the orifice of the blood-capillary; this would be forced open by
hydrostatic pressure induced by the far more powerful con-
tractions of the circular and longitudinal muscles. I cannot,
therefore, believe that — in Urochseta at any rate — there are
any pores which put the haemal system into communication
with the surrounding medium; and in the meantime their
resemblance to the integumental blood -capillaries of Limno-
drilus is worthy of note.

§ Excretory System.

The nephridial system of Urochseta has been partly de-
scribed by Perrier (22) ; according to his account it consists
of (1) a large gland, termed “ glande h mucosite,” occupying
the first few segments of the body ; (2) a series of “ segmental
organs/’ one pair to each segment ; (3) a series of remarkable
structures only developed in the posterior region of the body,
where they are present to the number of a pair to each seg-
ment; they coexist in these segments with the ordinary
nephridia. I shall consider severally these different organs,
which together constitute the excretory system.

Mucous Gland. — The mucous gland is figured and de-
scribed by Perrier as consisting of a tuft of long, much coiled
glandular tubules, which ultimately unite and open on to the
exterior by a long muscular duct. The orifices are situated
upon the first segment, and are each surrounded by a group
of muscular fibrils forming a sphincter. I have already (4)

238

PRANK E. BEDDARD.

pointed out the presence of this sphincter, which Perrier
was unable to definitely prove. Its presence is of course a
point of similarity between the mucous gland and the more
typical nephridia of the succeeding segments.

The structure of the glandular tubules is compared by
Perrier with that of nephridia, and he rightly points out their
resemblance, abandoning his earlier belief (24) that these
glands were a part of the alimentary system. At the same
time Perrier does not consider that their homology with
nephridia is definitely proved.

It will be obvious from an inspection of PI. XXIII, fig. 1, of
this paper that the structure of the mucous gland is identical
with that of the nephridia. It consists of rows of perforated
cells enveloped in a peritoneal sheath, which are so far abso-
lutely indistinguishable from the nephridia of the remaining
segments of the body.

There is one point, however, to which Perrier has not
directed attention in his memoir, and which conclusively
proves that these mucous glands are nephridial ; that is, the
presence of ccelomic funnels agreeing in their struc-
ture with the funnels of the nephridia in the other
segments of the body.

The “ mucous glands occupy the first six segments,
which contain no nephridia of the normal type ; these latter
do not commence until after. It is therefore a matter
of interest to inquire how far the “ mucous glands ” represent
the missing nephridia. Do they, in fact, simply represent
the hypertrophied first pair of nephridia, or are they formed
by a fusion of all the nephridia typically present in the
space which they occupy ? The fact that the external aper-
tures are single would seem to prove the truth of the former
supposition. On the other hand, the mucous gland does
not only differ from the typical nephridium by its
branched character, but also by the presence of
several ccelomic funnels.

In my preliminary notice (2) of the mucous gland I have,
I now believe erroneously, stated that each gland has four or

STRUCTURE OF UROCHiETA AND DICHOGASTER.

239

five funnels. In several instances I have good reason to
believe that there are only three present. The funnels are of
considerable size, and apparently of a somewhat horse-shoe-
shaped form ; hence in a continuous series of sections it is not
difficult to mistake one funnel for two contiguous but separate
funnels. The three funnels of each mucous gland are, how-
ever, so widely separated that no mistake of that kind could be
made. At the same time I do not wish to state positively that
there are only three present; and for the matter of that, the
exact number does not appear to me to be of great importance
unless it could be proved that each branch opens into the
coelom by a funnel. I shall presently show reasons for be-
lieving that this is not the case. Now, two of these funnels
are situated at the distal extremity of the gland and correspond
in their position to the fourth segment ; the third funnel is
more anterior in position and corresponds to the fifth
segment, so far as one can judge in the absence of definite
septa separating these segments. Although there is some
appearance of correspondence to the segments in the arrange-
ment of the ciliated funnels, yet it must be admitted that the
ciliated funnels are not arranged in a regular, meta-
meric fashion. Taking into consideration the facts (to be
referred to more at length below) that the nephridia of some
of the posterior segments are furnished with more than a
single funnel, and the extent of the first pair of nephridia
(“ tubiparous glands”) of certain Polycliseta sedentaria,
it is perhaps more likely that the mucous glands represent the
nephridia of the first segment alone ; on the other hand, there
is nothing in the facts, as I read them, which is contrary to
the supposition that the mucous gland represents the nephridia
of all the segments which it occupies, and that the primitive
condition is only shown, and that imperfectly, in the disposi-
tion of the ciliated funnels ; the concentration of this portion
of the nephridial system being due to its specialised function.

Then again, there is a third alternative. Supposing that the
mucous gland is the nephridium of the first segment alone, is
its branching to be considered as a remnant of what I have

240

FEANK E. BEDDAED.

elsewhere (1) urged is the primitive condition of the Anne-
lid nephridium, or is the branching, as Dr. Eisig would
argue (15), secondary ? This raises again the whole question
of the derivation of the Annelid excretory system, to which
Dr, Eisig’s recently published Monograph upon the Capi-
tellidse is a most weighty contribution.

In the latter part of this paper (p. 260) I discuss some
general questions relating to the nephridial system of Earth-
worms ; but it will be convenient to treat here of the argu-
ments which the structure of the mucous gland of U rochseta,
and of some other genera, furnish for the derivation of these
glands from a continuous network of tubules.

I have already stated that this gland in Urochseta com-
municates with the coelom by three funnels ; I am not quite
certain whether there is not a fourth. In any case there seems
to be no doubt that the number of branches is in ex-
cess of the number of ciliated funnels. Perrier’s
figure of the organ (22, pi. xvi, fig. 35) is, so far as I can
ascertain, accurate, in that it indicates the convergence of a
large number of nephridial tubules to form the long duct
of the gland. I have reason, however, to believe that in
some cases the tubules unite before their opening into the
muscular duct ; but this is not a matter of great importance.
One of two things must therefore follow : either the tubules
again unite before the ciliated funnels, thus forming a
network, or a large number (the greater number) of the
tubules end blindly without any coelomic apertures. I can
find no evidence of the truth of the first supposition, and
must therefore come to the conclusion that the mucous
gland is a branched nephridium, of which the greater
number of branches end blindly, while a few open
into the coelom by ciliated funnels.

These facts would seem to show that the gland is in some
respects degenerate ; that it primitively possessed a larger
number of ciliated funnels, the greater part of which have been
lost. So far this is merely an assumption, which at any rate
harmonises with the structure of the organ. Although the

STRUCTURE OF UROCH^TA AND DICHOGASTER.

241

nephridia of Earthworms are richly supplied with blood-capil-
laries, it seems nearly certain (particularly from the investiga-
tions of Kiihenthal) (20) that a good deal of the waste matter
that is excreted by them is not extracted from the blood-
capillaries by the cells of the nephridia, but is taken up by
the funnels ; the large granular peritoneal cells which clothe
the intestinal blood-vessels play an important part in this
process of elimination.

Now, the very differences between the mucous gland and the
other nephridia suggest that it plays a different part in the
economy of the animal. A suspicion that this was the case
led M. Perrier to term it “ glande h mucosite,” although he
had no evidence to bring forward of a positive nature ; this
supposition would account for the reduction of the ciliated
funnels ; the high development of the secreting part of the
organ, and the presence of a large vesicle for the storage of the
secretion, coupled with the reduction of the ccelomic apertures,
is clearly in favour of the view that this gland secretes a sub-
stance which is used for some definite purpose.

I describe below (p. 258) the structure and relations of the
anterior section of the nephridial system in Dichogaster.
This worm has an anteriorly situated gland which resembles
in many particulars the mucous gland of Urochaeta. It
consists of a tuft of highly convoluted tubules which have the
same structure as nephridia ; these tubules open by means of a
wider duct; the segments (Nos. 1 — 3) occupied by this gland
contain no other nephridia. The “mucous gland” of
Dichogaster differs from that of Urochaeta in cer-
tain important particulars; in the first place it has
no coelomic funnels; in the second place the duct
opens, not on to the exterior of the body, as in
Urochaeta, but into the buccal cavity; thirdly, it
appears to be formed by a single tube much coiled.
Apart from these points of difference, the similarity between
the two glands is so great that I cannot but regard them as
homologous. The fact that the mucous gland of Dichogaster
opens into the buccal cavity suggests that its function is

242

FRANK E. BEDDARD.

analogous to that of a salivary gland ; it may be at least
admitted that its function is probably different from that of
the nephridia in the remaining segments of the body. A
comparison between the structure of the mucous gland in the
two genera Dichogaster and Urochseta leads to the
inference, firstly, that they are homologous, and secondly,
that they present two stages in the evolution of the gland.
The primitive characters are more completely retained in the
mucous gland of Urochseta; it possesses funnels and opens
on to the exterior of the body on the first segment; the reduc-
tion in the number of the funnels, correlated with the changed
uses (?) of the gland, culminates in Dichogaster, where there
are no ciliated funnels ; at the same time the external aperture
comes to be situated in the buccal cavity.

I have elsewhere (7) described a similar gland in Acantho-
drilus novse-zealandise which, like that of Dichogaster,
opens into the buccal cavity. I could find no ciliated funnels.
In this case, as in that of Dichogaster, I discovered (see p.
259) the ciliated funnels of the nephridia elsewhere, and their
absence from the mucous gland rests upon observations which
are therefore more to be trusted.

Benham (9, No. 2) has recorded a gland in Diachseta which
occupies the same position and has the same general appear-
ance as the mucous glands of the types already referred to.
He states that it is not a branched gland, but consists only of
a single much contorted tube.

In Acantliodrilus annectens (Beddard 8) there are
a pair of anterior nephridia exactly like those of A. multi-
porus; and each opens in the same way into the buccal
cavity. I cannot discover very much evidence of this gland
being branched ; but fig. 14 appears to show that branching of
the tubules does occur, though apparently not to any great
extent,

There is nothing in the facts so far which is contrary to
Eisig’s supposition that the branching of the nephridium,
whether of the terminal (external apertures) or distal (coelomic
funnels) region, is secondary; on the other hand, these facts

STRUCTURE OP UROCHAETA AND DICHOGASTER. 243

may be equally well interpreted on the view that we have here
a rudiment of a primitive condition in which the nephridial
system formed a continuous network, with many funnels and
many external apertures in each segment.

I shall now bring forward further evidence of the truth of
this latter view.

Perrier has referred to the presence in Perichaeta of a
mass of glandular tubes in the anterior segments ; these were
figured by him in P. Houlleti, and were at first erroneously
regarded as connected with the alimentary canal. Later, they
were correctly referred to the excretory system. M. Perrier
remarks (22, p. 639) “that the segments (in Perichaeta) which
contain these glands are usually filled by a thick yellow secre-
tion, which the animal evacuates when annoyed.” This secre-
tion must be expelled, M. Perrier thinks, by the dorsal pores,
since he was unable to discover any excretory canal like that
of Urochaeta. Now, Urochaeta is an extremely small
worm, and an anatomist who lias proved himself sufficiently
skilful, as M. Perrier has done, to dissect out the minute duct
of the “mucous gland,” embedded as it is among the muscles
of the pharynx, would hardly fail to trace the same duct, if it
existed, in the comparatively large Perichaeta. By the study
of transverse and — which are perhaps better for this purpose —
longitudinal sections, I can quite confirm Perrier’s conclusion
as to the absence of an excretory canal like that of Urochaeta.
1 have, however, already (1) shown that the nephridia of these
segments open on to the exterior by numerous pores, and that
the nephridia of adjacent segments communicate through the
septa; this at any rate applies to P. aspergillum. In the
few first segments of the body of P. aspergillum (1)
the nephridial system is enormously developed ; all the
coelomic space available is closely packed with tubules. On
dissection this part of the excretory system has, comparatively
speaking, a solid appearance ; through the rest of the body the
nephridia are by no means so conspicuous, and, indeed, they
require a microscope for their demonstration.

The massing of the nephridia in a few of the anterior seg-

244

FEANK E. BEDDAED.

ments and their apparently different function from the nephridia
in other parts of the body (if one may so interpret M. Perrier's
experiments), renders plausible a comparison of this part of the
excretory system with the “ mucous gland” of Urochseta.
If this comparison be allowed the most important consequences
follow; it would seem, in fact, as if the specialisation of this
part of the nephridial system ultimately led to the concentration
of the numerous excretory pores into one long duct; that
in fact the branched mucous gland of Urochseta is
traceable to the specialised nephridial mass of the
anterior segments of Perichseta; the numerous ex-
ternal pores of the latter being replaced by the single
aperture of Urochseta.

I have in a previously published paper pointed out that if
the peculiar cutaneous glands of Urochseta correspond to
abortive setse, as they appear to do from a comparison with
similar glands in Anachseta (Yejdovsky, 29, pi. vii, fig. 1),
the eight setse per segment of Urochseta are brought about
by a reduction of a complete circle of setse such as exists in
Perichseta. On this hypothesis Perichseta is the primitive
form, Urochseta comes next, and finally Dichogaster and
Acanthodrilus, in which there is no trace of the missing setse,
complete the series. It will be noticed that the evolution of
the mucous gland, as I have traced it in the foregoing pages, is in
correspondence with this series of facts.

Nephridia. — All the segments of the body in Urochseta
from the fifth are furnished with a pair of nephridia.

The external apertures of these are perfectly plain on the ex-
terior of the body.

Perrier has already referred to the fact that the aperture of
the nephridium is surrounded by a peculiar cup-like structure,
which seems to be composed of radially arranged, short muscular
fibres. He has also figured the funnel. I find that with
respect to the funnel there is a remarkable difference between
the mucous gland and the nephridia of the anterior segments
on the one hand and the posterior nephridia. Perrier's
figure of the nephridial fuunel (22, pi. xvi, fig. 42) evi-

STRUCTURE OP UROCHJITA AND DICHOGASTER.

245

dently represents one of the latter. The funnels of the
mucous gland of the nephridia of the anterior segments are
in the first place much larger than those of the posterior
nephridia ; their structure also is different. The funnel itself
(PI. XXIII, fig. 5) is composed of the same columnar ciliated
cells with large nuclei, but it does not at once communicate
with the narrow tubule ; the latter is dilated into a wide cavity
of considerable length. This portion of the nephridium is not
to be confounded with the funnel although its lumen is of the
same size; its walls are tolerably thick and exhibit a faint
transverse striation, and contain oval nuclei embedded at
intervals. The structure of this part of the nephridium shows
that the lumen, although it is extremely wide, is nevertheless
intracellular ; it is simply a dilatation of the tubule.

This dilatation of the nephridial tubule recalls an analogous
dilatation which Bourne (12, figs. 51, 52, 53, 54), has
described and figured in Leeches, only in these animals the
lumen appears to be intercellular.

I have always observed this dilatation to be filled with what
are apparently degenerating corpuscles, the nuclei of which
were deeply stained by borax carmine. Bourne has observed
similar contents in the corresponding part of the nephridium
in Leeches.

In the genus Thamnodrilus (Beddard, 3) the funnels
of the anterior nephridia also differ from the funnels of the
posterior nephridia.

In a few segments I observed two nephridial funnels, but this
branching of the nephridium appears to be rare.

Perrier states that the nephridial funnel is contained in the
same segment as the nephridium itself. In a paper upon the
structure of an Australian species of Urochseta (4) I pointed
out that the funnel, as is usually the case among the Oligo-
clucta, was situated in the segment anterior to that which is
occupied by the rest of the nephridium. InUroclueta hystrix
I find a justification for Perrier’s statement; the nephridia are
sometimes entirely contained in one segment and sometimes
are not. In the second case the funnel is in the segment in

VOL. XXIX, PART 3. NEW SER.

R

246

FRANK B. BEDDAR.D.

front; the former arrangement seemed to be restricted to the
anterior nephridia. Finally, the anterior nephridia agree with
the mucous gland, and differ from the posterior nephridia in
the small calibre of the duct ; in this they agree with the
anterior nephridia of Perichaeta (see p. 262).

Ovaries and Oviducts. — Perrier’s memoir (22) pponUro-
chseta contains no description of the female reproductive
organs, except of the spermathecae. He remarks "that the
female reproductive apparatus seems to he fully developed
after the male reproductive organs.” This opinion is borne
out by my own experience. I have never found the two sets
of organs to be completely developed in the same individual.
All the specimens that I have examined were either “ males ”
or “ females.” This functional separation of the sexes, so
frequent in hermaphrodite animals, cannot be said to be general
among Earthworms. Benham, however, has found (9, No. 2)
that Urobenus, Diachaeta, and Trigaster agree with
Urochaeta in this particular ; in the two first genera he could
only discover the male organs, while in Trigaster the female
organs alone were fully developed.

I have investigated the minute structure and the position of
the generative organs, both by transverse and longitudinal
sections; their position and general relations could only be
properly determined by longitudinal sections, owing to the
arrangement of the septa in this region of the body. Perrier
has already recorded the fact that in the specially thickened
septa — the last of which bounds the tenth segment — the middle
region is very far behind the lateral margins which are attached
to the parietes. Each septum is therefore somewhat thimble-
shaped with the concavity forwards, and is largely enclosed by
the following septum. This does not only apply to the thick
septa, but to a large number of the excessively fine septa which
come behind. It is not in fact until the twentieth segment that
theintersegmental septum is disposed perpendicularly to the long
axis of the body. The ovaries and oviducts are situated anteriorly
where the delicate septa are hardly separable. The ovaries
and oviducts, as well as the funnels of the vasa deferentia, come

STRUCTURE OP UROCH^TA AND DICHOGASTER. 247

to lie opposite to the setae of segments which in reality are con-
siderably behind those which contain the several organs. In
correspondence with the arrangement of the septa the oviducts
run forward for some distance before opening on to the
exterior. Their position, however, is in reality perfectly
normal. The external apertures are upon the fourteenth
segments, and the funnels open into the thirteenth.

The vasa deferentia funnels open into the segment in front,
i. e. the twelfth.

In two specimens I found the female reproductive apparatus
fully developed, and the male organs, with the exception of
the vasa deferentia, not fully developed. The vesiculae semi-
nales in those individuals were very readily visible as out-
growths of the posterior side of the septum which separates seg-
ments 13 and 14 ; the vesicula was in the condition illustrated by
Bergh (11) in Lumbricus on pi. xxi, fig. 13, of his memoir. It
consisted for the most part of a solid mass of cells, with a narrow
lumen extending for a very short way into its thickness.

In these specimens (PI. XXIII, fig. 2) there were no testes,
but the twelfth segment as well as the thirteenth con-
tained a pair of ovaries. In another individual the gland
of the thirteenth segment contained ova in abundance. There
were also a few ova in the gland of the twelfth segment. I
figure (PI. XXIII, figs. 3, 4) a small fragment of the glands of
segments 12 and 13. In another specimen in which the vesiculae
seminales were in a further advanced condition, the genital
gland of the twelfth segment and that of the thirteenth segment
appeared to be a testis. These facts are, of course, a confir-
mation (though indeed a confirmation is hardly wanted) of the
accepted view that the ovaries and testes are serially homo-
logous structures. Prom this point of view the facts are of just
as great importance, even if it were shown that the individuals
were only abnormal. I am inclined to believe, however, that
they are not so, and that in Urochseta the same gland
may produce ova or spermatozoa.

In all the four individuals which I investigated by means of
longitudinal sections there were a number of bodies resembling

248

FRANK E. BEDDAR1).

mature ova lying in the body-cavity behind the thirteenth
segment quite detached from the reproductive glands of that
segment. They appeared to be contained in the fourteenth or
fifteenth segment, or even to occupy both of these segments.
In at any rate one instance these bodies appeared to be con-
tained in a thin-walled muscular sac, to the walls of which
were closely applied the transverse vascular trunks. In the
other cases they were grouped together, but I did not. observe
any structure resembling a muscular sac surrounding them.

The maturation of the ova1 of Urochseta outside the gland
in which they are developed is of some interest, even if the
supposed muscular sac enveloping them is nothing but a
partially detached (by the processes of embedding, &c.) portion
of the delicate intersegmental septa. Moreover, the ova them-
selves differ in some important particulars from the ova of the
majority of Earthworms.

Vejdovsky (29), as well as the earlier observers d'Udekem
and Claparede, dwells upon the fact that the ova of Earthworms
are small and numerous as compared with those of the majority
of the aquatic Oligochseta, which are large and few. The
greater size of the ova of the “ Limicolae ” is due to the fact that
they contain very much more abundant yolk. The greater
development of yolk in the ova of the “ Limicolae ” is, Vejdovsky
thinks, due to the different way in which they become mature.
In the aquatic Oligochaeta the ova detached from the ovary
are nourished by the perienteric fluid, while the ova of Earth-
worms remaining in the ovary are provided with special blood-
capillaries. The latter mode of nutrition, as the facts prove,
leads to the formation of numerous small ova, the former to the

1 I found these structures in two specimens of Urochseta, and occupying
the same position. I cannot, however, be certain that they are not Gregarines.
I am not aware that it is possible in preserved specimens to be absolutely
certain about such a point. All that can be said is that the bodies in question
arc closely similar to the ovarian ova of Phreoryctes, and that I only found
them in the situation mentioned. The fact of their not being surrounded by
smaller ovarian cells as arc the egg masses of llhynchclmis is not a
conclusive argument, since in Earthworms the ova in the receptacnlum are not
accompanied by sucli cells.

STRUCTURE OP UROCHA3TA AND DICHOGASTER. 249

increase in size of a few ova. Whatever may be the funda-
mental explanation of this structural dissimilarity, the fact
remains that there is a certain difference in the mode of
development of the ova in the aquatic and in the terrestrial
Oligochaeta. At the same time it has to be borne in mind that
in many Earthworms the ova when fully developed leave the
ovary and make their way to the interior of receptacula ovorum.
These chambers must at least be analogous to, if not homo-
logous with, the “ egg-sacs” of Stylaria, &c., in which ova
also undergo maturation. They differ, however, in being rela-
tively much smaller and thicker walled, and in having their
cavity divided up by trabeculae like the vesiculse seminales.
Yejdovsky does not give a detailed account of the development
of the egg-sacs (Eiersiicke) in Stylaria, and their homology
with the receptacula ovorum of Earthworms must be left for
the present undecided. The question as to homology does not,
however, affect the functional similarity of the two structures.
The receptacula ovorum of Earthworms are thicker walled,
and supplied with abundant blood-capillaries, which give them
a reddish appearance. The egg-sacs of Stylaria are thin
walled, and have no capillary network, but are supplied by
the hypertrophied vascular arch of their segment. This differ-
ence may perhaps be responsible for the unequal development
of the contained ova in the two cases. The whole question
requires further investigation.

Judging from Bergh’s (11) figures, the mature ova con-
tained in the ovary of Lumbricus hardly differ in size from
those contained in the receptacula ovorum. I have carefully
compared the relative sizes of the ovarian ova and those from
the receptaculum ovorum in Alluru s, and find that the latter
are rather larger; but the difference is not sufficiently striking
to lead me to the opinion that the ovum undergoes any im-
portant increase of bulk during its sojourn in the receptaculum.
Indeed, the observations of Dr. A. Collin (14) show that in
Criodrilus the ova contained in the receptaculum are
smaller than the largest ovarian ova ; but this is probably to
be explained by supposing that the smaller immature ova ripen

250

FRANK £. BEDDARD.

in the receptaculum, while the large mature ova pass from the
ovary directly into the oviduct.

The mature ova of Allurus and perhaps of Uro-
chaeta differ from those of the majority of Earth-
worms, and agree with those of the “ Limicolse,” in
the fact that they are of comparatively large size.

The only other Earthworm known to me, in which the ova
are of large size, is Allurus. Fig. 22 of PI. XXIY illustrates
the comparative size of the mature ova of a number of Earth-
worms and ofPhreoryctes. It will be seen from that figure
that the ova of Allurus are markedly larger than those of
Eudrilus, &c., though smaller than the supposed ova of Uro-
chaeta. This fact is of particular interest in relation to other
points in the structure and economy of Allurus. This worm,
although structurally nearer to Allolobop bora than to any
other Oligochset, is not terrestrial ; at least, not exclusively
terrestrial in its habits. I received some specimens from
Teneriffe which were collected in company with a number of
specimens of Lumbricus and Allolobophora in soil;
on the other hand, Mr. Martin Woodward was so good as to
forward me a specimen of Allurus, which he discovered in a
vessel containing Char a which had been collected for the
use of the botanical students at the School of Science, South
Kensington. There was no reason to believe that this indi-
vidual had accidentally found its way into the vessel ; it had
been in all probability collected in the stream which furnished
the Char a. Mr. Benham has lately contributed to ‘ Nature’
a note in which he points out that Allurus is largely aquatic
in its habits.1 It is interesting to find that this particular
genus approximates to the “ Limicolse ” in its habits ; indeed,
it is the only instance known to me of an aquatic Earthworm,
though of course many of the “ Limicolae” live in damp soil.

1 Since writing the above I find that Vejdovsky in his paper upon
Rhynchelmis (‘ Zeitschr. f. wiss. Zool.,’ 187G) has mentioned the occurrence
of Allurus in streams. During a recent visit to the Plymouth Station of the
Marine Biological Association 1 found Allurus in abundance among coarse
gravel in the lliver Plym, near Bickleigh.

STRUCTURE OF UROCILETA AND DIOHOGASTER. 251

Allurus, furthermore, resembles certain of the Limicolous
genera in the large size of its ova, and in the fact that the
female reproductive pores are behind the male. I have not
any evidence that Urochaeta can, like Allurus, lead an
aquatic life ; but the resemblance which it bears to the
“ Limicolae ” is shown in the possession of bifurcate setae as
well as (perhaps) in the large size of its ova. Perrier dis-
covered the former fact, and I have occasionally observed the
same in specimens from British Guiana.

II. Dichogaster Damonis, nov. gen. et sp.

The present section contains some account of the anatomy
of a species of Earthworm, which appears to be sufficiently
unlike any other type at present known to justify the creation
of a new genus for its reception.

I have examined two specimens which I acquired from Mr. K..
Damon, of Weymouth. One of these was dissected, the other
studied by means of transverse sections.

The worms formed a part of the Godeffroy collection,
recently purchased by Mr. Damon, and are labelled “ Hypo-
gaeon.” This name has been applied to several very different
species of Earthworms, and in Savigny’s original description
is characterised by the possession of a single median seta in
addition to the eight which are ordinarily found. In this
character Hypogaeon differs from the present species.

The species was collected in Fiji.

§ External Characters.

The setae are paired, and lie on the ventral side (fig. 8).
The dorsal and lateral pair of setae are separated from each
other by a rather greater interval than that which separates
the ventralmost pairs of setae. The clitellum extends from
segments 13 — 20 inclusive. It is not so markedly developed
on the ventral as on the dorsal side ; hence the number of
segments of which it is composed can be more easily reckoned

252

FRANK E. BEDDARD.

from the ventral side. The twentieth segment has the whole
ventral region enclosed between the lateral pairs of setae
entirely devoid of glandular epithelium, which is only developed
on the dorsal region of this segment. The more anterior
segments, in like manner, have no development of glan-
dular substance for the greater part of the ventral area.
The seventeenth segment bears the apertures of the vasa
deferentia, which do not correspond to the ventral setae, but
are more ventrally placed. The apertures are situated on a
tumid area which occupies the space lying between the setae.
On the two following segments there are similar areas, but
more distinctly marked off from the surrounding integument.
Dorsal pores are present, but I could not ascertain where
they commenced.

On the eighth segment are the apertures of the single pair
of spermathecae. These are closely approximated in the median
ventral line, and open near to the anterior margin, as is so
generally the case. The various layers which compose the
body wall appear to have much the same structure in this as
in other species of worms. Particularly noteworthy is the fact
that the longitudinal muscular layer shows the bipinnate
arrangement of its fibres which is so characteristic of some,
although not of all, species of Lumbricus, and is found also
occasionally in other genera. This is illustrated in fig. 6 of
Plate XXIII. In the anterior part of the body the fibres of
the longitudinal muscular coat do not show any such regularity
in their arrangement.

§ Alimentary Canal.

The most salient fact in the structure of the alimentary
canal of this Earthworm is the presence of two gizzards (fig.
21) ; these are situated close together in consecutive segments,
and are only separated by a very minute oesophageal portion,
the calibre of which is not far short of that of the gizzards
themselves; the segments occupied by the gizzards are 7 — 10,
the mesenteries separating these segments from each other are,
as is often the case, not obvious. It will be seen, therefore.

STRUCTURE OP UROCMTA AND DICHOGASTER.

253

that each gizzard occupies two segments. The presence of
more than a single gizzard is not new among Earthworms ;
Digaster, Perrier (24), and Didymogaster, Fletcher (16),
as their names imply, have two gizzards, but the present genus
cannot be confounded with any of these ; more than two
gizzards occur in other Lumhricidae, viz. Trigaster (Benham)
and Moniligaster (Perrier).

The oesophagus is furnished behind the gizzard with cal-
ciferous glands; of these there are three pairs, situated in
segments 15, 16, and 17 respectively (fig. 21) ; the two anterior
pairs of these glands are rather larger than the posterior
pair and in the specimen studied by me were full of cal-
careous particles, the product of their activity, which were
entirely absent from the smaller pair ; the oesophagus contained
a large quantity of the calcareous secretion of the calciferous
glands.

The posterior pair of calciferous glands is divided by longi-
tudinal furrows into four distinct lobes ; its blood supply is
derived direct from the dorsal vessel, there being apparently
no supra-intestinal trunk; the blood-vessel enters the gland
along the short pedicle, which unites it with the walls of the
oesophagus. The same appears to be the case with the two
anterior pairs, and in all the glands the vascular supply is
also in connection with the blood sinus of the oesophageal
walls.

§ Generative Organs.

Testes and Vesiculse Seminales. — I have only been able
to study these structures by means of transverse sections ; by
dissection I could not, owing to the friable condition of the
specimen, make out the exact relationship between the com-
ponent parts of the male generative organs.

The testes (fig. 15, t.) are two pairs of small glands situated
in segments 10 and 11. The organ is somewhat irregular in
shape, and furnished with numerous finger-shaped processes.
A dissection even of the immature example which I studied by
transverse sections would not have shown the testes, inasmuch

254

FRANK E. BEDDARD.

as they are completely overgrown and surrounded . by the
vesiculse seminales of their respective segments.

The fact that the testes are actually surrounded by the
vesiculse during the growth of the latter is, of course, a result
of numerous investigations. The adult structure of the
Earthworm at present under consideration would prove this
point, supposing, that is to say, that there was the least need
of proof.

The sac-like vesiculse seminales completely enclose the testes,
and in the case of the anterior pair, at any rate, enclose also a
tuft of nephridial tubules, which happen to be closely associated
with the testes. In other worms other organs of the body,
e. g. the ventral blood-vessel, are enclosed within the cavity of
the vesiculse.

The testes of Dichogaster have apparently the same
structure that characterises these organs in other Earthworms.
They are attached to the mesentery close to the ventral median
line on either side of the nerve-cord ; at the point where they
are attached the walls of the vesiculse come into contact, and
are fused with the mesentery.

The structure of the vesiculse seminales is curious and differs
in certain particulars from the vesiculse of other Earthworms.

In the example which I dissected the eleventh and twelfth
segments contained each a pair of racemose structures of small
size (fig. 15, r.'), appearing on each side of the gut. These
presented every resemblance to the vesiculse seminales of many
species of Acanthodrilus. In the tenth segment a mass of
developing spermatozoa occupied the ventral region of the seg-
ment, and partly obscured the fimbriated apertures of the vasa
deferentia. A study of the generative apparatus by means of
transverse sections showed that the structure in segment 10 is
not a loose mass of developing spermatozoa set free from the
vesiculse of segments 11 and 12, and ready to be extruded
through the open funnels of the vasa deferentia. It is really
a pair of vesiculse seminales (fig. 15, r .) with a delicate outer
wall, and presenting the usual structure. This vesicle, al-
though presumably originally a paired structure, does not show

STRUCTURE OF UROCHiETA AND DICHOGASTER. 255

much evidence of being a paired structure in the adult worm ;
the two halves of the vesicle are almost completely fused in
the ventral median line where they enclose the nerve-cord. The
ventral blood-vessel is not enclosed within the vesiculae, but is
suspended by a vertical mesentery some little way down between
the two vesiculae, which here become distinctly separate ; a
portion, however, of the transverse vessel of this segment, as
well as (necessarily) a branch on each side, which runs to the
testis, are enclosed by the vesiculae. The vesiculae send off a
narrow lateral band, which seems to become fused with its
fellow of the opposite side in the dorsal median line (see
fig. 15).

In the eleventh segment is another pair of vesiculae, for a
description of which the foregoing remarks will nearly suffice.
The same segment also contains (see fig. 15) the racemose
structures already referred to. These are composed of a large
number of small spherical acini, which contain bundles of de-
veloping spermatopliora. The whole structure is firmly attached
to the mesentery, which divides its segment from the one in
front. I have not been able to make out any connection
between this portion of the vesiculae and the undivided median
sac.

Finally, segment 12 contains another pair of these racemose
organs, which have apparently no connection with the vesiculae
of the preceding segment.

Yasa Deferentia. — There are two pairs of vasa deferentia
funnels situated in segments 10 and 11 ; they open into the
middle of the vesiculae seminales of these segments, on each
side of the nerve-cord and near to it. Their structure calls for
no special remark, neither does that of the vasa deferentia,
which open, in common with the glandular body, upon the
seventeenth segment of the body.

When the worm was opened in dissection the seventeenth,
eighteenth, and nineteenth segments were seen to be largely
occupied by three pairs of glands, a pair to each segment, of a
whitish colour, and meeting above the intestine. The anterior
pair of these is very much larger than those which follow, and

256

FRANK E. BEDDARD.

somewhat contorted ; the latter are narrower tubular organs
exactly resembling each other (fig. 7).

An examination of these glands by transverse sections shows
that they all open on to the exterior at a corresponding point
in the three segments ; the external apertures of these glands,
in fact, correspond in position to the innermost of the ventral
pair of setae. I find, however, that in these three segments,
viz. 17, 18, and 19, the ventral pair of setae are altogether
absent, although they are present in the neighbouring seg-
ments. The dorsal pair of setae are as well developed in
segments 17, 18, and 19 as in any others.

It is very common to find some modification of the setae in
the segments which bear the male generative pores, such as,
for example, the bundles of elongated setae in Acanthodrilus,
but I am not acquainted with any other instance (except Eu-
drilus) in which the setae entirely disappear on these segments
(figs. 16, 17).

The anterior pair of glands (figs. 7, \6.pr.), those which occupy
segment 17, and which are distinguished by their greater size
and greater opacity, are the real atria ; that is to say, it is these
glands alone which are connected with the vasa deferentia.
These glands have much the same structure as in other Earth-
worms; the very narrow lumen is surrounded by a layer of
columnar cells ; outside these is a mass of glandular cells, the
exact relations of which the condition of the material does not
enable me to state positively. Apparently these cells resemble
very closely the corresponding cells in the prostate of Eu-
drilus. Outside is a delicate sheath containing blood-vessels
which send off branches among the gland-cells. The atrium
comm unicates with the exterior by a slender but thick-walled
muse ular duct ; this duct is at first much contorted, but when
it enters the body wall is perfectly straight; its course through
the latter is oblique, the external orifice being placed nearer to
the ventral median line than the point where the tube enters
the body wall.

The tubular glands of segments 18 and 19 (figs. 7, 1 7,pr.')
are straight, and not contorted like the glands of segment 17,

STRUCTURE OF UROCH7ETA AND DICHOGASTER. 257

and their diameter is considerably less ; their minute structure,
however, seems to he identical, except that the glandular layer
is naturally less developed.

Ovaries. — These organs (fig. 15, o.) occupy the usual posi-
tion in segment 13 ; they are large and conspicuous.

Oviducts. — The oviducts (fig. 15, od.) open by a wide,
funnel-shaped orifice into the interior of segment 13 ; their
duct perforates the mesentery, dividing this from the succeeding
segment. Each opens separately on to the exterior ; the external
orifices are very closely approximated, and lie within the ven-
tralmost setse at the same level as the apertures of the atria.

Spermathecae. — There is only a single pair of these organs
present, which are situated in the eighth segment ; the external
aperture, as already stated, corresponds in position to the
ventral pair of setae. The spermatheca is divided into two
parts (fig. 8), a large sac lying posteriorly and opening on to
the exterior in common with a mulberry-like structure which
represents the diverticulum ; as in so many other species of
Earthworms, the diverticulum lies anterior to the pouch. The
minute structure of these two sections of the spermatheca
differs ; the pouch itself is lined (fig. 19) with a tall columnar
epithelium, which appears to resemble in every particular the
lining epithelium of the spermatheca of Lumbricus. Outside
this is a comparatively thin layer of muscular tissue permeated
by abundant blood-capillaries ; the muscular sheath gets much
thicker where the pouch narrows to its external opening, and
here the character of the lining epithelium alters slightly and
becomes indistinguishable from the epidermis of the body
surface. The structure of the numerous diverticula differs
somewhat; the presence of numerous small diverticula gives to
the region of the spermatheca its mulberry-like aspect. They
are all, however, enclosed within a common muscular sheath
(fig. 20), which is proportionately thicker than in the case of the
spermatheca itself, and abundantly vascular. The diverticula
are closely packed with bundles of spermatozoa, and the lining
epithelium differs from that of the spermatheca itself; the
epithelial cells arc low and cubical. I have called attention

258

FEANK E. BEDDAED.

elsewhere (7) to the fact that the diverticula of the spermathecse
in Lumbricidae are of different minute structure to the sper-
mathecas, and are usually occupied by the bundles of sperma-
tozoa which are absent from the spermathecse themselves.

Nephridia. — The excretory system of this worm, as of so
many others, differs in different regions of the body. Professor
Spencer has called attention, in a paper (28) to which I shall
have again to refer, to the fact that the nephridia of Megas-
colides are different in the anterior and in the posterior regions
and of the body ; this is also the case with Pericliseta (p. 262)
and Urochaeta (p. 246); and the same condition occurs in
many genera (e. g. Microchaeta and Thamnodrilus) in
which the nephridia consist of paired tubes, each with a single
coelomic funnel and external pore.

With regard to Megascolides, Spencer points out that the
nephridia of the anterior segments present more primitive
characters than those of the posterior segments, where they
first begin to be modified. This statement appears to hold
good (as I have already pointed out) in Acanthodrilus, and
the facts which I shall bring forward in the present paper
show that in Perichaeta the nephridial system of the pos-
terior segments is more modified than that of the anterior
segments. In Urochaeta it is only in the anterior segments
that a single nephridium has more than a single ciliated
funnel. With regard to such genera as Microchmta and
Thamnodrilus, it is difficult to say that the anterior ne-
phridia are in any way more primitive than those of the
posterior segments.

In Dichogaster the same generalisation with respect to the
nephridia appears to hold good. I am unable, however, to
give so complete an account of the nephridia as I could have
wished. The first five segments are occupied by a large
nephridium, which evidently corresponds to the large anterior
nephridium of Acanthodrilus multiporus and A. annec-
tens. I could not find the funnels of this organ (if they are
really present), nor could I find any very decided evidence of
its being a branched gland. I am rather inclined, however, to

STRUCTURE OF UROCH^ITA AND DICHOGASTER. 259

believe, from the analogy of Urochseta, that it is branched.
This nephridium terminates in a comparatively wide, thick-
walled tube, which becomes wider and thinner walled as it
approaches the external orifice, which is within the
buccal cavity, as in the two species of Acanthodrilus
mentioned above. In the segments of the body which follow
(I am uncertain how many), the nepliridial system is much
like that of Acanthodrilus multiporus; that is, it consists
of tufts of tubules which open by numerous apertures on the
surface of the body. These apertures have no regular
arrangement that I could observe; frequently they are
situated near to the setae, but as frequently they open near to
the anterior or posterior boundaries of the segment. The
apertures are extremely obvious, both in transverse and longi-
tudinal sections, on account of their large size. I have not
been able to observe any funnels connected with these ne-
phridia.

In the posterior region of the body the nephridia are
different, and, as already mentioned, are in certain respects
more modified than those of the anterior segments.

On a dissection of this region of the worm the nephridia
appeared to be separable into a number (about six) of pairs of
distinct nephridia. In transverse sections the nepliridial
system was seen to consist of scattered tufts of tubules aud of
a large pair of nephridia ; the arrangement being, in fact,
much like that of Megascolides. The calibre of the large
nephridia was many times greater than that of the small tufts,
or about equal to that of the nephridia of such types as Lum-
bricus. Each of these large nephridia is furnished with a
large ciliated funnel, which lies in the segment in front. I
have been quite unable to detect the external apertures of the
nephridia of these posterior segments.

The tufts of smaller tubules were not in all cases (if in any)
detached from the large nephridia ; their apparent distinctness,
when seen in a dissection of the worm, is due to the fact that
they are for the most part embedded in the centre of a mass
of peritoneal cells. These peritoneal cells, which form aggre-

2G0

FRANK E. REDDARD.

gations round certain parts of the nephridia, are exactly like
those which surround the nephridia of Pontodrilus, as well
as of Phreoryctes. Perrier was the first who drew attention
(23) to the resemblances in this particular between the nephridia
of Pontodrilus and those of the “ Limicolae ; ” and Dicho-
gaster is another instance of an Earthworm which so far ap-
proximates in the characters of its nephridia to the “ Limicolae/’

III. Further Remarks on the Nephridia of
Earthworms.

The Nephridia of Perichaeta aspergillum. — With re-
gard to the nephridia of Perichaeta aspergillum, I am able
to make some additions to my former papers upon this genus (1).
The most important point which I was then able to prove is
that the nephridiopores, instead of being present to the number
of only one pair in each segment, are extremely numerous.
I stated (1, p. 401) that there were often four or five nephridio-
pores lying between two setae, making, therefore, a total of
from one to two hundred in each segment. I have figured and
described these nephridiopores as forming a continuous row
round the middle of each segment. After discovering that in
Dichogaster the nephridiopores are not limited to the spaces
between the setae of a segment (v. suprh, p. 259), I carefully
re-examiued Perichaeta aspergillum with reference to this
point ; the result of this re-examination is to show that P.
aspergillum resembles Dichogaster. The nephridio-
pores are scattered irregularly over every part of
the body, and are not by any means confined to the
area lying between the setae of a given segment.

Ciliated Funnels. — Another fact of some little importance
which I am able to add to my former paper upon Perichaeta,
is the description of ciliated funnels. In the posterior region of
the body the funnels were extremely obvious although small;
the small size of the ciliated funnel corresponds to the small
calibre of the excretory tubules. These structures were ob-
vious, for the reason that, as a general rule but by no means

STRUCTURE OF UROCHiETA AND DIOHOGASTER. 261

always, they are borne at the extremity of a very straight
tubule (fig. 10). The structure of the funnels is illustrated in
fig. 10. There is nothing specially remarkable about them
except their small size. The presence of ciliated funnels has
been already described in the genus Pericheeta by Rosa (26),
who found in P. armata a pair of ciliated funnels in every
segment. Dr. Benham informs me that he has noticed in a
species of Pericheeta from the Philippines numerous funnels in
each segment, corresponding to the numerous nephridia, which
he has already briefly referred to (9, No. 1, p. 256) as existing
in that species (which has apparently not yet been identified).

In P. aspergillum I have satisfied myself that there are a
number of funnels in each segment; this, however, only applies
to the segments behind the clitellum. In the anterior seg-
ments, the nephridia of which alone were described in my
former paper, I am still unable, after a renewed search, to
discover any evidence of the presence of ciliated funnels.

It has been stated that the ciliated funnels are of small size,
but they are not all of the same size; some (fig. 10«) are
distinctly larger than others (fig. 10 b ). I shall have occasion
to point out directly that the nephridial tubules of these pos-
terior segments are partly of greater calibre than those of the
anterior segments ; it is possible in the posterior segments to
distinguish these wider tubules from the minute tubules which
resemble those of the anterior segments. This accounts for
the difference in size between the funnels. The larger funnels
are connected with the larger tubules. It occasionally happens
that the larger funnels are borne upon tubules, which imme-
diately perforate the septum and join the nephridial tufts of
the segment behind.

Comparison of the Nephridia of the Anterior with
those of the Posterior Segments.

In my paper already quoted upon the nephridia of Peri-
chseta aspergillum I have described the perforation of the
intersegmental septa by tubules which connect tbe nephridial
systems of adjacent segments. In some of the anterior seg-
VOL. XXIX, PART 3. NEW SER.

s

262

FRANK E. BEDDARD.

merits of this Pericliaeta, particularly those which contain
the spermathecse, the nephridial system consists of an enormous
mass of tubules which almost completely fills the available part
of the coelom. So closely are the excretory tubules packed
that I have found it impossible to distinguish a series of sepa-
rate nephridia corresponding to the numerous external pores.
This fact, together with the perforation of the septum by
tubules, led me to the impression that there must be in this
region of the body a continuous nephridial network
independent of the segments.

It occurred to me while making these observations, and it
has occurred to me lately after discovering the ciliated funnels
of P. aspergillum, that the supposed connection between the
nephridial system of two adjacent segments might be really
nothing more than the normal perforation of the septa by
tubules, terminating on the anterior side of the septum in
ciliated funnels.

This supposition, however, appears to be negatived by the
following considerations: In the first place I succeeded in
many cases in tracing a given tubule through the septum until
it became lost in the excretory mass of the segment in front.
Secondly, in the posterior region of the body the ciliated
funnels are usually not borne upon the anterior face of the
mesenteries in the way that is so general among Earthworms,
though this sometimes happens. In most cases the long
straight tube bearing the funnels rises up from a tuft of tubules,
and does not perforate the septum, but ends in the same seg-
ment. Thirdly, it occasionally happens, both in the anterior
and in the posterior region of the body, that a mesentery was
perforated at one spot by a number of tubules running close
together in irregular windings. Such masses of nephridial
tubules did not pass between the individual muscular fibres of
the septum, but the continuity of the tissues of the septum was
broken at the point where they traversed it. A conspicuous
gap was thus formed, which was entirely occupied by the
nephridial tubules and peritoneal cells coating them. In these
cases it appeared to me that the bundle of tubules passing

STRUCTURE OF UROCH^TA AND DIOHOGASTER. 263

through the septum was not formed by the coils of a single
tube, but that it really represents a number of separate tubes
running side by side. On the assumption that the perforation
of the intersegmental septa by nephridial tubes is not evidence
of an intercommunication of the nephridia of successive seg-
ments, one would have expected to find a number of funnels
dependent from the septum at this point. I could not,
however, detect these structures, and in the posterior region of
the body, as already stated, the funnels are rarely attached to
the septa.

The probability of my statements being correct is also largely
increased by the discovery of Professor Spencer (28), that in
Megascolides australis there is a continuous network of
nephridial tubules uninterrupted by the septa.

The observation of the nephridial tubules within the thick-
ness of the septum is not always easy. In some cases, how-
ever, they are accompanied by a tolerably thick coating of
peritoneal cells, when they can be readily detected. I have
found that in the anterior region of the body it is easier to
trace the tubules from segment to segment in transverse
sections. In the case of the larger tubules of the posterior
segments the branches connecting the tufts of adjacent seg-
ments are not difficult to make out.

In the posterior region of the body the nephridia are not so
well developed as they are anteriorly. The nephridial tubules
are, however, much like those of the anterior segments (unless
there are really no funnels in the anterior segments), but they
are closely attached to the body wall, and particularly to the
septa. They do not occupy a large portion of the body-cavity.
I have ascertained by sections, as well as by an examination
of stripped-off pieces of cuticle, that the nephridiopores have
the same irregular distribution that they have in the anterior
segments. Furthermore, there is, as has been mentioned,
an intercommunication between the nephridial tufts of suc-
cessive segments. I have observed frequently a connection,
by tubules traversing the septum, between two nephridia
adherent to opposite sides of the same septum. At the

264

FRANK E. BEDDARD.

same time it appears to be certain that in the nephridia of
these segments there is no longer an intimate connection
between all the nephridial tubules of the same segment. An
examination of a series of sections shows that there are tufts
of tubules which are quite isolated from neighbouring tufts.
On the other hand, there is — as has just been said — frequently
no break between nephridial tufts of adjacent segments. These
facts appear to me to be of some importance with regard to
the views which I have elsewhere (1) advanced as to the origin
of the Oligochset excretory system. We have here, as it appears
to me, a commencing separation of the continuous excretory
network into isolated nephridia. This breaking up has at first
no relation to the segmentation of the body. The nephridial
tufts have no regular arrangement within the segment, and
their apertures are dotted about irregularly over its surface,
and the separation into separate nephridia does not follow the
lines of the intersegmental septa. The excretory system,
in fact, appears to retain, longer than many other
organs of the body, traces of the primitive unseg-
mental condition.

For the most part the nephridia of the posterior segments
have the same appearance as those of the anterior segments,
that is to say, they consist of tufts of tubules having an
excessively fine bore. There are, however, tubules of greater
calibre which appear to be wanting in the anterior segments.
In this particular there is a resemblance between P. asper-
gillum and Megascolides (Spencer). In that genus the
posterior segments of the body contain nephridial tubules
which are much larger than others in the same segments, and
than all in the anterior segments of the body. There is also
the further resemblance that the tufts of larger tubules are
connected with funnels which project into the segment in front.
In Perichaeta, however, the smaller nephridial tufts also
possess funnels, which they apparently do not in Megascolides.
Until the publication of Professor Spencer’s illustrated account
of Megascolides it is impossible to say how far this resem-
blance in the specialisation of the nephridia goes. The

STRUCTURE OF UROCH2ETA AND DICHOGASTER. 265

difference iu size is not very marked iu P. aspergillum, not
nearly so much so as in P. armata.

In my paper upon Perichseta aspergillum I have not
figured the cuticular pores of the nephridia, and so I have
thought it worth while to introduce into the present paper
illustrations of their structures. Pig. 23 of Plate XXIV
represents a portion of the cuticle of P. aspergillum,
showing the cuticular ingrowths which surround the proximal
region of the seta (a), and the very delicate cuticular tube ( b )
which lines the extremity of the duct of the nephridium.
When these structures are viewed from above the aperture,
whether of the seta or of the nephridium, they appear to be sur-
rounded by a thickened layer of the cuticular membrane. This
is, I believe, only an optical effect due to the inturned edges of
the cuticle. It seems, however, to define very plainly the orifice.
The very great size of the seta orifice, as compared with that
of the nephridiopore, will be evident from an examination of
the figure cited. The cuticular pore of the nephridium is
further remarkable for the fact that its edges are usually
much crinkled, which is probably due to the contraction of
the epidermic cells by the preservative reagent. The cuticular
pores which lead into the seta sacs never show these crinkled
edges, probably for the reason that they remain distended by
the seta.

Having ascertained that these cuticular pores belong to the
nephridial system, I have examined the cuticle of another
species of Perichseta of which I possess examples not suf-
ficiently well preserved to show the modifications of the
epidermal cells round the nephridiopore. 1 find that they
are present in Perichseta Iloulleti, and I consider myself
therefore at liberty to infer that iu this species (and, indeed,
probably in all Perichseta in which the nephridia have a
“tufted” character) the structure of the nephridial system is
much the same as that of P. aspergillum.

In Acanthodrilus and Dichogaster the external orifices
of the nephridial system are larger than those of Perichseta
and (judging from Spencer’s description) of Mcgascolides;

266

FRANK E. BEDDARD.

their greater size renders them very plainly visible in transverse
and longitudinal sections of the body wall and upon fragments
of the cuticle. The cells surrounding the orifice are tall, thin
cells, not bulged like those of Perichseta and Megas-
colides.

The Nephridia of Perichseta armata (F. E. B.). —
I owe the material, upon the study of which the present descrip-
tion is based, to the kindness of Mr. W. L. Sclater, of the
Indian Museum, Calcutta.

The species was first described by myself, and has been
recently in some respects more fully characterised by Rosa (26).

There is, however, one point in which Rosa’s description
differs from my own. I stated that the nephridia, at least in
the anterior region of the body, consisted of numerous tufts
of tubules, resembling in this particular the nephridia of the
greater number of species of Perichseta. The characters of
the nephridia in the specimens examined by me was such that
I should have presumed — in the light of my own subsequent
investigations — that the number of nephridiopores in each
segment would be greater than two.

On the contrary, Rosa’s description of those organs shows
that he considers them to be like those of Lumbricus, i. e. a
single pair to each segment. He describes, and I can confirm
the accuracy of his description, the presence in each segment
of a pair of coiled nephridia, each of which opens into the
segment in front by a ciliated funnel. Rosa was unable to
find the external pores. So far I can fully bear out the state-
ments made by Rosa ; but this description of the nephridial
system of P. armata is not exhaustive. It consists also
of numerous tufts of minute tubules which are
scattered about irregularly in the segments. These
tubules are not obvious on a dissection of the worm, but they
are quite easily seen in transverse sections.

The nephridial system of Perichseta armata differs in
important particulars from the nephridial system of any
species of Perichseta; it differs from that of P. aspergillum
(see p. 265) and an undescribed species briefly referred to by

STRUCTURE OF UROCH^TA AND DICHOGASTER. 267

Benliara (9, No. 1) in the presence in each segment of a
pair of large nephridia, opening by a funnel into the segment
in front, in addition to the tufts of minute tubules present in
these types. In one or two species from Australia, described
by Mr. Fletcher (16), only the large pair of nephridia are
present. The minute tufts of tubules are unrepresented.

There is, however, a close resemblance between the nephridia
of P. armata and those of Megascolides australis, which
have been briefly described in a note published in ‘ Nature ’ of
June 28th, 1888, by Professor Baldwin Spencer. I have not
yet had the opportunity of seeing Professor Spencer's detailed
memoir upon this most interesting genus of Earthworms, but
the note referred to is an abstract of the more important results
of his investigation of the nephridial system.

It appears that in the anterior segments of Megascolides
there are abundant scattered tuft6 of minute nephridial tubules,
which are connected by a network lying within the peritoneum
and extending from segment to segment. In the posterior
segments of the body there are in addition a pair of coiled
nephridial tubes of a very much greater calibre than the minute
tubules. Each of these opens by a funnel into the segment in
front, and they are connected by a continuous longitudinal duct
which runs from segment to segment. These larger nephridia,
as well as the longitudinal duct, are also in connection with the
system of minute tubules ; the latter have no ciliated funnels
but open externally by numerous pores.

In P. armata I have not actually traced the nephridial
tubules through the body wall to their point of opening on to
the exterior. I have, however, found upon the cuticle the
nephridiopores, which were abundant in each segment, and
agreed in all particulars with those of P. aspergillum (see
p. 265), so that I cannot admit any doubt as to the resemblance
in this particular between the nephridia of P. armata and those
of P. aspergillum. I have also been unable to detect any
ciliated funnels except those belonging to the large pair of
nephridia. In all these points, therefore, there is au agreement
with Megascolides. But the nephridial tufts of P. armata

268

FRANK E. BEDDARD.

appear to be at any rate largely isolated from each other and
from the pair of large nepliridia ; and I have not found a
longitudinal duct passing from the large nephridia of successive
segments and connecting them. Neither can I discover evi-
dence of any nepliridial network uniting the tufts of minute
tubules of successive segments. In all these points the
nephridia of P. armata are different from those of Mega-
scolides. I shall refer again to the nephridia of P. armata
and to Professor Spencer’s description of Megascolides (see
below.

Comparison of the Nephridia of Perichaeta, Mega-
scolides, Acanthodrilus multiporus, Deinodrilus,
Dicliogaster.

Before attempting to draw any conclusions as to the path
of development of the excretory system in Earthworms, it will
be convenient to briefly review the facts already known con-
cerning the nephridia of those genera in which there is a
greater or less development of a network with numerous
external pores in each segment.

It appears to be possible to separate those genera into two
groups: the first group contains Perichaeta and Mega-
scolides; the second, the remaining genera enumerated above.
I am at present uncertain as to the relations of Typliaeus,
which has not yet been properly investigated.

The principal character which distinguishes the nephridia of
these two groups is the size of the tubules.

In Perichaeta, and apparently also in Megascolides, the
greater part of the nepliridial system (the whole of it in the
anterior segments of the body) is made of tubules having an
excessively fine lumen ; the entire diameter of the tubules is
not inconsiderable, but the perforation of the cells which form
the duct is much less than the thickness of its walls. Besides
the network of fine tubules, both these genera possess coils of
tubules of a much greater diameter which are more or less
closely connected with the network of fine tubules ; that is to

STRUCTURE OF UROCH^ITA AND DICHOGASTER. 269

say, they form a more or less independent nephridium opening
internally in Megascolides and Perichueta armata by a
single funnel.

In Deinodrilus, Acanthodrilus, and Dicliogaster,
the general nephridial network is made up of tubules, the
lumen of which is greater than in Perichseta; the diameter
of the cells is not greater, but the lumen occupies a greater
proportion of the cell. These tubules resemble in fact very
closely the finer portion of the nephridium of Lumbricus.
In Deinodrilus (at any rate in those segments of the body
which I have investigated — some of the more posterior ones)
the nephridial network appears to be entirely made up of
tubules of this kind. In the other two genera, however,
part of the nephridial network is composed of tubules of a
much greater calibre, equal in size to the larger tubules of P.
armata, or of such Earthworms (e. g. Allurus, Pontodrilus,
Eudrilus, Acanthodrilus no vse-zeal audite) as possess
but a single pair of nepliridia in each segment of the body.
In Acanthodrilus multiporus the larger tubules are
not independent of the smaller tubules, and the network opens
into the ccelom by numerous funnels, as in Perichseta asper-
gillum. In Dicliogaster, in the anterior segments, this
specialisation of the network is not seen; in the posterior
segments, on the other hand, there is not much beyond the
coil of large nephridial tubules, which have to a great extent
the characters of a single nephridium, such as that of Lum-
bricus, &c., and open into the coelom by a single funnel borne
at the end of a duct which traverses the intersegmental septum.

We have therefore a parallel series in the nepliridia of
these two groups which may be expressed in the following
Table :

270

FRANK E. BEDDARD.

A. Nepbridia forming a network,
consisting of excessively fine
canals, continuous from seg-
ment to segment.

(1) ?

(2) Nephridial network of pos-
terior segments, partly com-
posed of tubules of greater
calibre. Numerous ccelomic
funnels. Pericbseta asper-
gillum.

(3) Larger nephridial tubules in-
creased in size and forming a
nephridium nearly independent
of the finer tubes, and opening
by a single ccelomic funnel.
P. armata, Megascolides.

B. Nepbridia forming a network con-
sisting of wider canals, discon-
tinuous at the septa.

(1) No further specialisation.
Deinodrilus .'

(2) Nephridial network, partly
composed of tubules of greater
calibre. Numerous ccelomic
funnels. Acantbodrilus
multiporus.

(3) Nephridial network of pos-
terior segments, chiefly com-
posed of larger tubules, open-
ing by a single ccelomic funnel.
Dicbogaster.

The nephridia of Acantbodrilus multiporus, of Dicho-
gaster, and of Deinodrilus, are formed of tubules which,
as said, are on the whole of greater calibre than those of
Perichseta. The measurements may be approximately de-
termined by a comparison of figs. 11-14. At the same time
the nephridia of these types present other differences from
Pe rich set a. The network is much reduced in extent and
in two ways. First, only a limited area of each segment
is occupied by the nephridia. They are by no means so
abundantly developed as in Perichseta, not nearly so
abundantly developed as in the anterior segments ofPericheeta
aspergillum. Secondly, the intercommunication from
segment to segment has disappeared in Acantbodrilus aud
Dichogaster, and has almost disappeared in Deinodrilus.
In the last-mentioned genus the nephridia are attached to the
anterior wall of their segment, and are, for the most part,
entirely restricted to this situation. In one or two instances,
however, a small tuft of tubules was attached to the posterior
wall of a segment and in these cases (which are not at all

1 Tbe apparent absence of ccelomic funnels in this genus may perhaps be a
secondary modification.

STRUCTURE OF UROOH^ETA AND DICHOGASTER. 271

numerous) the tuft of tubules attached to the posterior wall
was in communication through the septum with the nepliri-
dium of the segment behind. This seems to me to indicate
that the nephridial system of Deinodrilus is in a more
archaic condition than that of either Acanthodrilus or
Dichogaster. In Deinodrilus the primitive disposition
of the excretory system ofPerichseta has been so far retained
that there is still an intersegmental communication here and
there. The metameric arrangement of the nephridial system
is not so complete as in Acanthodrilus and Dichogaster,
though, for the matter of that, neither of these forms have an
excretory system perfectly metameric in its disposition.

Another point of difference between the excretory system of
Perichgeta on the one hand, and that of Acanthodrilus,
Deinodrilus, and Dichogaster, is in the form of the ex-
ternal orifices.

Professor Spencer (28) described the external orifices of
the nephridia of Megascolides in the following words:
“ The external opening itself is formed of cells of the epidermis,
so modified as to present very much the external appearance
of a taste-bulb ; that is, they form a sphere with the cells
thicker in their middle parts, and the two ends attached to the
poles of the sphere, the duct passing right up through the
centre.”

This description applies very closely to the modified epi-
dermic cells which surround the nephridiopores ofPerichgeta.
When I first observed these cells in Perichseta I thought for
a moment that they really belonged to sense organs. The cells
are so much swollen in their middle parts that the duct which
forms up between them is of au excessively fine bore ; for this
reason it is not always easy to detect upon fragments of the
cuticle the actual orifice.

272

PRANK E. BEDDARD.

The Evolution of the Excretory Organs in Earth-
worms.

I shall now proceed to deduce, from the facts described in
the present paper and in Professor Spencer’s account of
Megascol ides, what I believe to have been the course of
development of the nephridial system of Earthworms.

In my paper upon Perichseta (1) I pointed out that the
facts therein described were in favour of the assumption
that the presence of a single pair of nephridia per segment
(e. g. in Lumbricus) was the last stage of a reduction of an
excretory system like that of Perichaeta; and that the ex-
cretory system of Perichaeta was distinctly comparable to
that of the Platyhelminths. With regard to the first point.
Professor Spencer’s observations are, as he has pointed out,
decidedly confirmatory of that view. Indeed, the nephridial
system of Megascolides appears to me to be hardly intel-
ligible on the hypothesis that Lumbricus represents the
primitive condition.

Dr. Hugo Eisig’s magnificent monograph of the Capitellidae
(15), which has just been published, contains a very detailed
discussion of the nephridial question. It must be confessed
that the structure of the nephridia in the Capitellidae might be
equally well explained on the hypothesis that the ancestral
condition of the Annelid nephridial system is represented by a
pair of distinct nephridia in each segment. And this is the
position which Dr. Eisig takes up. The branching, whether
of the distal or proximal end of the nephridium, and the
connection between nephridia of the same segment, as well as
the multiplication of the latter, he regards as secondary. It
appears to me that this position may be safely yielded without
affecting the strength of the converse view which is main-
tained in the present paper. I believe it to be unnecessary to
assume that the Oligochaeta and the Polychaeta have been
derived from the same Annelid stock: I hold that the ancestral
form from which they diverged was intermediate between the
Platyhelminths and Annelids. There is no difficulty in drawing

STRUCTURE OF UROCHiETA AND DICUOGASTER. 273

a sharp line of division between the Oligochseta and the Poly-
chaeta. The peculiarities of the reproductive system will be
the basis of this distinction. The investigations of Korschelt,
Meyer, and Weldon upon Dinophilus have gone a long way
towards demonstrating that this worm stands at the base of the
Polychset series. Now, the nephridia of Dinophilus are in
their minute structure comparable to those of the Platyhel-
minths ; in most species they form a single pair of branched
organs terminating in numerous “ flame-cells.” In D. gyroci-
liatus, according to Meyer, each single nephridium is broken
up into a series metamerically arranged, and each opening by a
separate external pore. This I believe to be the way in which
the Polychaet nephridia have arisen.

There is no known form which seems to me to represent an
intermediate stage between the Oligochseta and the Platyhel-
minths. On the whole, it must be admitted that certain of
the aquatic Oligochmta, such as the Naidomorpha, stand at
the base of the Oligochset series. The fact that the nephridia
of these Annelids are paired is a difficulty in regarding Peri-
chseta as representing in the structure of its nephridia an
ancestral form. It must be remembered, however, that our
knowledge of the aquatic Oligochaeta, though no doubt fairly
advanced as regards indigenous forms, is very small as regards
exotic genera. Also there are traces (in An a ch set a, Vej-
dovsky (29) (PI. VII, fig. 14) of what I believe to be the
primitive condition. It may be that the (presumed) reduction
of the nephridia in these aquatic forms has some relation to
their small size, and, in consequence, to the reduced size of the
coelomic cavities.

It will be of no advantage to endeavour to combat Dr.
Eisig’s arguments against regarding the nephridia of Acan-
thodrilus multi por us as representing an archaic condition,
principally for the reason that at the time when he wrote he
was able to say that only one or two genera exhibited the
dysmetameric condition, the vast majority having a metameric
condition of the nephridia.

We are now, however, acquainted with the following genera

274

FRANK E. BEDDARD.

in which the nephridia are often or always dysmetaraeric : —
Perichseta, Acanthodrilus, Typhseus, Deinodrilus,
Dichogaster, Megascolex (?), Megascolides, Notos-
colex, while traces of the same are to be seen in Urochseta.
The argument of the rarity of the occurrence of the dys-
metameric nephridia cannot any longer have any weight, and
his detailed criticisms, though powerful at the time, are now,
through the progress of discovery, of less weight. His
other arguments depend chiefly upon the fact that this con-
dition is only found among the Polychseta in the Capitellidse.
Regarding, as he does, the Capitellidse as nearly akin to the
Oligochseta, and in fact forming the intermediate link between
them and the Polychseta, this argument is a powerful one.
I find myself, however, unable to accept this position.
The peculiarities of the reproductive system in the Oligochaeta,
coupled with the entire absence of parapodia and external
gills, distinguish them from the Polychaeta. Dr. Eisig com-
pares the peculiar modification of the integument which
surrounds the genital pores with the clitellum of the Oligo-
chaeta. I would myself rather compare it with the modified
integument which surrounds the aperture of the vas deferens
in Allurus and Allolobophora ; but I do not think that
the possibility of this comparison is necessarily a mark of near
affinity. The comparison between the nephridia of the Capi-
tellidae and those of Acantliodrilus does not really show a
very close resemblance; the structure of these organs is so
peculiar, as Eisig has shown, that it renders a detailed com-
parison difficult, as does also the fact that they are often pre-
ceded by a provisional set. Indeed, I cannot help agreeing with
Dr. Eisig that their modifications in the Capitellidse are
secondary, though I would maintain that this is not the case
with Urochseta, Acanthodrilus, &c.

The nearest approach to the primitive condition of the
excretory system in the Oligochseta is, in my opinion, seen in
Perichseta aspergillum; in the anterior segments the
resemblance to the Platyhelminth excretory system is closest.
There is here a continuous network of tubules, with numerous

STRUCTURE OF UROCH^TA AND DICHOGASTER. 275

external pores. The network is not interrupted by the septa,
and the external pores are not in any way related to the seg-
mentation of the body. If funnels are really absent, as appears
to be the case, then the termination of the tubules in single
cells will be an additional point of resemblance to the Platy-
helminths ; if, on the contrary, funnels are really present,
they must be small and inconspicuous and not much advanced
beyond the single flame-cell.1

In the posterior segments part of the nephridial network
consists of tubules of a greater calibre, and these, as well as
the smaller tubules (which are exactly similar to those of the
anterior segments), are provided with funnels. The external
apertures are still extremely numerous, and irregularly distri-
buted over the surface of the body. The network of tubules
is beginning to break up into more or less isolated tufts ; but
the separation of the continuous network into isolated nephri-
dia has no discernible relation to the segmentation ; the tufts
of tubules have no regular arrangement within the segment,
and the septa do not as yet form barriers between the excretory
tubes of different segments.

In the posterior segments, therefore, the primitive characters
of the nephridial system are just beginning to disappear. If
the posterior segments resembled the anterior segments the
nephridial system of P. aspergillum would exhibit the pre-
sumed ancestral condition.

From this point the modification of the excretory system
has, as I think, proceeded along two slightly divergent paths;
the ultimate point reached, however— the reduction of the
nephridial system to a pair of isolated nephridia in each seg-
ment— is the same in both cases. The facts known appear to

1 I have already (1) discussed the “ funnel ” of the Annelid nephridium
and its relation to the Platyhclminth flame-cell. Since that paper was written
Yejdovsky has published (‘ Zool. Anzeiger,’ Bd. x) an account of the nephridia
of certain Oligochseta. The " provisional ” nephridia, which are preceded at
the anterior extremity of the body by a “ larval ” set, terminate in a flame-cell.
These nephridia entirely disappear in the first two or three segments; behind
this they become converted into the permanent nephridia; the flame-cell
divides and gives rise to a funnel.

276

FRANK E. BEDDARD.

me to necessitate this view of the gradual reduction of the
excretory system ; it is difficult to harmonise the facts with the
hypothesis of one continuous line of development.

It is obvious that any theory of the development of the
nephridia must allow for the reduction of the nephridial net-
work in Perichseta aspergillum to a single pair of ne-
phridia, such as is found in P. novae-zealandiae,1 and also in
the genus Perionyx, which is in all respects a very near ally
of Perichseta; and this reduction must not involve the
various stages represented by Deinodrilus, Acantho-
drilus, and Dichogaster, though these are intermediate
between P. aspergillum and P. novae-zealandiae.

The intermediate stage between P. aspergillum and P.
no vae-zealandiae is represented by P. armata. In this Peri-
chaeta the nephridia of the posterior segment are, as Spencer
pointed out in the case of Megascolides, separable into two
categories ; firstly, there are the tufts of minute tubules ;
secondly, a pair of convoluted nephridial tubes, with a ciliated
funnel borne upon the extremity of a tube which has traversed
the septum, and lying in the segment anterior to that which
contains the nephridium ; these latter are of the same calibre
as the nephridia of P. novae-zealandiae, and indeed of most
Earthworms in which there is but a single pair of nephridia
per segment. I believe that these have originated from the
somewhat larger nephridial tubules of such a form as P.
aspergillum; the minute nephridia form tufts which are
largely, if not entirely, isolated from each other and from the
large nephridia; they are comparatively inconspicuous, and
seem to be in course of disappearance. Megascolides offers
an analogous stage in the development of a single pair of
nephridia out of the nephridial network. I quite agree with
Spencer that the single pair of nephridia of certain
Earthworms (e. g. Perichaeta novae-zealandiae and
Perionyx) have arisen by a gradual increase in

1 This is an apparently new species of Perichseta, which I hope to
describe shortly ; it possesses a single pair of nephridia per somite, as in
Lumbricus.

STRUCTURE OF UROCHJ1TA AND DICHOGASTER. 277

calibre of a part of the nephridial network in each
segment to form a pair of nephridia, and by the
gradual disappearance of the rest.

The second way in which I conceive the gradual reduction
of the network to a single pair of nephridia to have been
brought about is as follows :

The network became arranged metamerically by the isola-
tion of the networks of successive segments at the septa ; at
the same time the tubules themselves acquired a greater
calibre. This stage is nearly reached in Deinodrilus,
where the nephridial system forms a continuous series of tufts
attached to the anterior wall of each segment ; but here and
there in Deinodrilus the nephridia are connected through
the septa with feebly-developed tufts of tubules lying on the
posterior side of the segment in front.

In Acanthodrilus multiporus this stage is exemplified ;
all trace of the intercommunication between the nephridial
systems of successive segments through the septa is lost, and
the tubules are uniformly of greater calibre than those of
Perichaeta; at the same time they are more decidedly re-
lated to the setae of their segments. From this point the paired
nephridia of other species of Earthworms have been derived
either by a great increase in the calibre of the tubules coupled
with the disappearance of part of the network and all the
external orifices, except a pair to each segment (Dichogaster
seems to be a stage further advanced than Acanthodrilus in
the direction of those worms with a single pair of nephridia in
each segment),1 or by the breaking up of the network into
separate nephridia. Brachydrilus (Bcnham, 10) offers an
intermediate condition in this reduction ; the nephridial net-
work has been broken up so as to form two separate pairs of
nephridia in each segment. One pair then disappears, and
the typical condition of the Earthworm excretory system is
arrived at.

I am disposed therefore to believe that the paired

1 I have elsewhere (8) called attention to other points in which
Deinodrilus is intermediate between Perichmta and Acanthodrilus.

VOL. XXIX, PART 3. NEW SER.

T

278

FRANK E. BEDDARD.

nephridia of certain other Earthworms (e. g. Acan-
thod rilus novse-zealandiae) have been derived

through the gradual increase in calibre of the
tubules forming the primitive network, which has
become isolated into metamerically disposed tufts
of tubules, corresponding more or less to the setse;
these separate nephridia have become ultimately
reduced to a pair in each segment.

In the first case, therefore, the single pair of nephridia have
been derived directly from a part of the primitive network ;
in the second case the primitive network has become converted
into a single pair of nephridia in each segment by a more
gradual series of changes.

The annexed scheme shows the relationship between certain
genera of Earthworms, as indicated by their excretory system.

Acanthodrilus

Dichogaster

Acanthodrilus multiporus1

This scheme, as will be seen, only refers to the genera which

1 It is possible that this species is really a distinct generic type. It has no
paired setse like the other species of the genus. If so, A. annectens
(Beddard, 8) should probably be referred to the same genus.

STRUCTURE OF UROOHjETA AND DICHOGASTER. 279

have been specially treated of in the present paper. I do not
feel able at present to extend the diagram so as to embrace all
the known genera, or even the greater number.

I would point out, however, that the above scheme, though
meant only to express the probable course of the development
of the excretory system, does not do violence to the relation-
ships in other structural characters between the different
genera.

List of Memoirs referred to.

1. Beddard, F. E. — “ On the Occurrence of Numerous Nephridia iu the

same Segment in certain Earthworms, &c.,” ‘ Quart. Journ. Micr.
Sci.,’ Jan., 1888.

2. Beddard, E. E. — “Note on the Mucous Gland of Urochceta,” 4 Zool.

Anz.,’ 1887.

3. Beddard, F. E. — “ On the Structure of a New Genus of Lumbricidae

(Thamnodrilus Gulielmi),” ‘Proc. Zool. Soc.,’ 1887.

4. Beddard, F. E. — “ Observations on the Structural Characters of certain

New or Little-known Earthworms,” ' Proc. Roy. Soc. Edinburgh,’
1887.

5. Beddard, F. E. — “The Anatomy and Histology of Pleurochaeta

Moseleyi,” ‘Trans. Roy. Soc. Edinb.,’ vol. xxx, part ii.

6. Beddard, F. E. — “The Structure of the Body-wall in certain Earth-

worms,” ‘ Proc. Roy. Phys. Soc.,’ 1884.

7. Beddard, F. E. — “New Zealand Earthworms,” 4 Proc. Zool. Soc.,’ 1885.

8. Beddard, F. E.— “ The Anatomy of Three New Species of Earthworms,

&c„” ‘Quart. Journ. Micr. Sci.,’ Oct., 1888.

9. Beniiam, W. B. — “ Studies in Earthworms,” Nos. 1, 2, 3, ‘ Quart.

Journ. Micr. Sci.,’ 1886.

10. Beniiam, W. B. — “ Brachydrilus,” ‘Zool. Anzeig.,’ Bd. x (1887).

11. Bergji, R. S. — “ Gcschlechtsorgane der Regenwiirmer,” ‘ Zeitschr. f.

wiss. Zool.,’ Bd. xliv (1886).

12. Bourne, A. G. — “ Anatomy of the Ilirudinete,” ‘ Quart. Journ. Micr.

Sci.,’ J 884.

13. Claparede, E. — “ llistologische Untcrsuchungen ueber den Regen -

wiirrn,” ‘ Zeitschr. f. wiss. Zool.,’ Bd. xix (1869).

14. Collin, A. — “ Criodrilus,” ‘ Zeitsch. wiss. Zool.,’ 1888.

280

PRANK E. BEDDARD.

15. Eisig, Hugo. — “Die Capitellideu,” ‘Fauna und Flora des Golfes von
Neapel.’

18. Fletcher, W. — “ Australian Earthworms,” ‘ Proc. Linn. Soc. N.S.W.,’
1886-7-8.

17. Horst, R. — “ Aanteckeningenop de Anatomie van Lumbricus,” ‘Tijd.

Nederl. Dierh Yer. Deel,’ iii, Afl. i.

18. Horst, R. — “ Notes on Earthworms,” ‘ Notes from the Leyden Museum,’

vol. viii.

19. Lankester, E. Ray. — “ Epidermis of the Leech,” ‘Quart. Journ. Micr.

Sci.,’ 1880, p. 303.

20. Kukenthal, W. — “Lymphoid Zellen der Anneliden,” ‘Jen. Zeitsch.,’

1885.

21. Mojsisovics, E. von. — “Die Lumbriciden Hypodermis,” ‘ S. B. Wien.

Akad.’

22. Perrier, E. — “ Urochaeta,” ‘Arch. Zool. Exp.,’ t. iii, 1874.

23. Perrier, E. — “ Pontodrilus,” ‘Arch. Zool. Exp.,’ t. ix, 1881.

24. Perrier. E. — “ Recherches pour servir a l’histoire des Lombriciens

terrestres,” ‘Nouv. Arch, du Mus.,’ t. viii, 1872.

25. Rosa, D. — “Criodrilus lacuum,” ‘ Mem. Ace. Torino,’ 1886.

26. Rosa, D. — “ Perichetidi di Birmania,” ‘ Ann. Mus. civ. Genova,’ vol. vi

(1888).

27. Sarasin. — ‘Arbeit. Zool. Zoot. Inst., Wiirzburg,’ 1885.

28. Spencer, W. B. — “The Nephridia of Earthworms,” ‘Nature,’ June,

1888.

29. Vejdovsky, F. — 1 System und Morphologic der Oligochaeten,’ Prag.,

1881.

STRUCTURE OE UROOH^ITA AND DICHOGASTER. 281

EXPLANATION OF PLATES XXIII & XXIV,

Illustrating Mr. Frank E. Beddard’s paper “ On Certain
Points in the Structure of Urochaeta, E. P., and Diclio-
g aster, n. g., with further Remarks on the Nephridia
of Earthworms.”

PLATE XXIII.

Fig. 1. — Semi-diagrammatic longitudinal section through anterior extremity
of Urochaeta corethrura. The aperture of the mucous gland at o is
correctly drawn as regards its position relative to the setae, but it should be
more ventral in position, n. Ventral nerve-cord, f Funnels of mucous
gland (3). g. Gizzard, s. Setae, c. Supra-cesophageal ganglion, al. Cavity
of anterior end of the alimentary tract.

Fig. 2. — Longitudinal section through genital segments of the same species.
The vesiculae seminales are not represented, t. Testis, o. Ovary, ov. Ovi-
duct pore. v. d. Vas deferens. The segments are numbered.

Figs. 3 and 4. — Contents of genital glands of the same specimen. Both
testes and ovaries have produced ova in this individual.

Fig. 5. — A funnel of the mucous gland of Urochaeta.

Fig. G. — Transverse section through body wall ofDichogaster Damonis.
e. Epidermis, m. Circular muscles. 1. Longitudinal muscles, p. Peritoneum.

Fig. 7. — Dichogaster. Segments in the neighbourhood of the male

reproductive pores, v. d. Vasa deferentia. pr. Atria, pr1 . Glands in 18th
and 19th segments, similar in structure to the atria, but unconnected with
the vasa deferentia.

Fig. 8. — Dichogaster. Ventral external view of segments in the neigh-
bourhood of the male reproductive pores, to show pores upon the 17th, 18th,
and 19th segments.

Fig. 9. — Fragment of nephridium of ditto, with glandular peritoneal
cells («).

Fig. 10. — Perichseta aspergillum. Nephridial funnels, u smaller, b
larger. In a one of the two funnels, that to the right, is seen in longitudinal
section.

Fig. 11. — Perichseta armata. Large nephridia of posterior segments.
a. From a glycerine preparation, which showed very clearly the boundaries
between the successive “drain-pipe” cells.

282

FRANK E. BEDDARD.

PLATE XXIV.

Fig. 12. — Perichscta aspergillum. Nephridial tube, a, with larger,
b, with smaller lumen.

Fig. 13. — Deinodrilus Benhami. Nephridial tubes from posterior
segment.

Fig. 14. — Acanthodrilus multiporus. Nephridial tubes, a, with
small lumen; b, with wider lumen ; c, represents the greatest size to which
the nephridial tubes of this species reach.

Figs. 11 — 14 are all carefully drawn to scale with camera lucida.

Fig. 15. — Diehogaster. General view of genital segments dissected.
The upper wall of the semiual reservoirs is removed on the left side to show
the funnels and testes, r. Seminal reservoirs, r'. Seminal reservoirs of a
racemose appearance. 1. Testes, f. Funnel of vasa deferentia. o. Ovary.
od. Oviduct.

Fig. 16. — Transverse section of body of the same worm at the line of the
atria ( pr .). v. d. Vasa deferentia joining the muscular portion of atria.

Fig. 17. — Corresponding section through nineteenth segment, pr1 . Glan-
dular body.

Fig. 18. — Spermatheca with appendix, a.

Fig. 19. — Transverse section through wall of spermatheca.

Fig. 20. — Transverse section through appendix of spermatheca.

Fig. 21.— Anterior region of alimentary canal, to show two gizzards,^., and
calciferous glands, Ca.

Fig. 22. — Ova of different species of Oligochacta, to illustrate their relative
sizes. Drawn to scale, a. Of Uroclueta corethrura, from ccelom. a1.
Largest ova from ovary, b. Of Phreoryctes Smithii, from ovary, c.
Of Allurus. c. From receptaculum ovorum. d. From ovary, d. Of
Eudrilus, from ovary, e. Of Acanthodrilus, from ovary.

Fig. 23. — Fragment of cuticle of Perichaeta aspergillum, a. Orifices
of setae, b. Nephridiopores.

Fig. 24. — Perichaeta aspergillum. Diagram to illustrate nephridia of
posterior segments, o. Nephridiopore. f Funnel.

A cint Wuxixjsxh.xni

F E.Beddard del.

F Hufh, bthT Edir/

Mxyr. JX1V

ng-n

F.E.Beddard del

..-CL,

Flo -22

Fig. 26.

r rrf'r,. Lifhr Edior

Fig. 16.

Fig. 13.

F'glS.

Fig. 21.

t Fig. 15.

DEVELOPMENT OF PERI PAT US NOViE -ZEALANDIiE. 283

On the Development of Peripatus
N ovse-Zealandise.

By

Lilian Sheldon,

Bathurst Student, Newnham College, Cambridge.

With Plates XXV and XXVI.

In January last, through the kindness of Mr. Sedgwick,
I received another supply of Peripatus novae- zealandise.

As before, they arrived in the living condition, and the eggs
were removed from the uterus immediately after the animal
had been killed with chloroform.

The proportion of males was considerably larger than on
previous occasions, being twenty-two out of a total of forty-
nine. There were nine smallish females which contained no
embryos ; and in the remainder, which varied in size from
about three-quarters to two inches in length, the uteri were
filled with embryos. The number of embryos in a single
female varied considerably, the maximum being eighteen and
the minimum seven.

Most of the embryos were preserved in corrosive sublimate
and glacial acetic acid used hot, but the best results were
obtained from some which were placed for six or seven hours
in a mixture consisting of equal parts of -5 per cent, chromic
acid and 2 per cent, acetic, and afterwards washed in alcohol.
In this method it is not necessary to prick the egg-shell before
the embryo is removed to alcohol. After this method of pre-
servation, which is that recommended by Hertwig for amphibian

284

LILIAN SHELDON.

eggs, the yolk is rendered much less brittle than after auy other
methods which I have tried; the protoplasm and nuclei are
well preserved, and also the egg-shell expands and lies at some
distance from the periphery of the ovum, and so can easily be
removed. The eggs were all stained with picro-carmine, and
passed through the various strengths of alcohol in which a
small quantity of picric acid was dissolved.

The embryos, with the exception of a few quite old ones,
were all of stages between those shown in figs. 10 and 15 in
my last paper (4), that is, they came in age between those
received in December and April respectively.

My material is again very incomplete, and the new stages
which I shall describe, though they throw some light on the
early development, are very few, and do not unfortunately by
any means fill up the gaps which were left in the account of
the development given in my last paper (4); but it seemed
advisable to publish my results, in the hope that they might
prove useful if anyone should have the opportunity of working
on the development of this interesting species with a better
supply of material than I have been able to obtain.

The ovum, which represented the latest segmentation stage,
described and figured (fig. 10) in my last paper (4), was one
which was taken out of the uterus in December. In it the
nuclei were present round slightly more than half the ovum,
lying in small masses of branched protoplasm. The central
one of these nuclei lay on the surface and showed signs of
karyokiuetic figures. There were also two or three proto-
plasmic masses in the central yolk. I have now (in the
January lot) several stages later than this, which show that the
nuclei in the centre of the surface of the ovum beneath which
they lie multiply with considerable speed and very much more
quickly than those over the rest of the ovum, a condition
which is shown in fig. 4, until by their repeated increase the
egg acquires the form shown in fig. 11 (4), which represented a
transverse section through an April ovum. In the ovum there
figured there is a specially-marked area of reticulate proto-
plasm, containing a large number of nuclei extending through

DEVELOPMENT OF PERIPATUS NOV/E-ZEALANDDE. 285

about one third of the length of the ovum, and having in
transverse section an irregular triangular shape, the base of
the triangle resting on the surface; nuclei are also present
throughout the yolk, more especially near the periphery.

As was said in (4) the protoplasmic area next becomes more
compact and flattened out against the side of the ovum, so as
to form a kind of plate of fairly dense protoplasm which is
closely packed with nuclei, and which lies on the surface at one
side of the ovum through about the middle third of its length :
this is shown in transverse section in fig. 13 (4), which is from
one of the ova taken out in April.

A good many of the January eggs were of stages which
came after this last. The protoplasmic area begins to grow
round the surface of the egg until, as is shown in fig. 1, it
covers nearly half the periphery. The nuclei in the central
region of this protoplasmic area are two or three deep, but
towards its edges the protoplasm thins gradually and the nuclei
form only one layer. The ovum is closely packed with yolk,
among which a few nuclei are irregularly scattered. The pro-
cess of epibolic growth of the protoplasmic area continues, so
that, as is shown in fig. 2, still more of the periphery of the
egg becomes covered with the thin layer of protoplasm which
may be called the blastoderm, until when the growth of the
latter ceases only a very small portion is left uncovered.

The blastoderm now consists of a thin layer of protoplasm,
in which a single layer of flat nuclei is present, extending
round the whole periphery of the egg with the exception of a
small space lying in the centre of one side of the egg in which
the yolk is left uncovered. This uncovered space corresponds
with the position of the future blastopore, and is, I believe,
situated on the ventral side of the embryo, thus agreeing with
other Arthropods, except the Scorpions, in which the blasto-
pore is situated dorsally.

A change now takes place in the blastoderm behind and at
the sides of the uncovered area. In the middle line behind
this area a proliferation of the nuclei takes place extending
some distance backwards, so that a keel-shaped mass of nuclei

286

LILIAN SHELDON.

embedded in a loose reticulum of protoplasm is present in
about the posterior half of the ovum ; this condition is shown
in transverse section in fig. 3. The nuclei at the sides of the
area also proliferate, as is shown in fig. 3 a , which represents
a transverse section through this region from the same ovum
as fig. 3, but is drawn under a higher power.

The proliferating mass of nuclei increases in size and
occupies a larger portion of the surface of the ovum, and
both nuclei and protoplasm are closer and more compact ; a
transverse section through the posterior half of such an ovum
is shown in fig. 8. At the same time the protoplasm at the
sides of the uncovered area become slightly inflected ; this is
shown in fig. 9, which is a transverse section through this
region from the same ovum as fig. 8, but drawn under a
higher power. The uncovered area thus forms a passage
lined by the cells of the blastoderm, which have become in-
flected, and leading into the yolk ; it may be spoken of as the
blastopore, as is the case in P. capensis it is traversed by
strands of protoplasm.

These two structures, i. e. the blastopore and the area of
proliferating cells lying posterior to it, soon acquire very close
resemblance to the blastopore and primitive streak of P.
capensis, with which they are probably homologous. The
blastopore increases in length and the protoplasm at its sides
shows a true invaginate character, and a groove is present
running from the posterior lip of the blastopore down the
centre of the primitive streak. Three transverse sections
through an egg of this stage are shown in figs. 5, 6, and 7.
Fig. 5 passes through the blastopore at about the middle point
of its length, where it is very clearly open ; at this time it is
about an eighth of the total length of the ovum. Fig. 6
passes through the region immediately behind the blastopore
through the primitive streak and groove. Immediately beneath
the primitive groove there is a small cavity bounded by the
protoplasm of the primitive streak, the nuclei round it being
arranged in a roughly columnar manner, and filled with very
small yolk-spheres, among which are one or two nuclei ; this

DEVELOPMENT OF PERIPATUS NOV2E-ZEALANDI2E. 287

cavity is marked c in the figure. It ends blindly in front and
behind, and probably is homologous with the area which Mr.
Sedgwick (1) calls the polar area in P. capensis, which, both
in position and structure, it closely resembles, with the excep-
tion of the fact of its being filled with yolk-spheres in the
New Zealand species. Fig. 7 passes through the posterior
end of the primitive streak, where it is thinning out, and the
groove is much shallower. These three sections bear a very
close resemblance to figs. 25 and 26 of Mr. Sedgwick’s second
paper on the Cape species (1). The peripheral nuclei in the
region of the blastopore and primitive streak have a more or
less columnar form instead of lying flat against the side as
they do over the rest of the ovum. The anterior part of the
egg is enveloped in a single layer of flat nuclei.

In an ovum of a slightly later stage the blastopore has
increased a little in length, the primitive streak is much
larger and more marked, the nuclei being very closely packed,
and the primitive groove is considerably deeper ; the so-called
polar area has disappeared. This stage is the latest which
was present among the January eggs, except some quite old
embryos which were almost ready for birth.

I have examined several series of sections of older embryos,
i. e. in which they were developed, but have not thought it
necessary to give an account of them, as the process of
development seems to be similar to that of P. capensis de-
scribed by Mr. Sedgwick (2).

The only point of interest in which it differs from that species
is that the first somite (i. e. that of the prseoral or antennal
segment) opens by a duct to the exterior in precisely the same
way and position as do those of the third to fifteenth segments,
so proving it to be the nephridium of the segment with the
same relations as those of the posterior ones. This is shown
in figs. 10 and 11 ; in the former the opening of the duct to the
exterior just outside the nerve-cord is shown, and in the latter,
which is separated from the former by three sections, its opening
into the somite. The probability of the nephridial nature of
this somite was pointed out by Mr. Sedgwick.

288

LILIAN SHELDON.

Summary of the Results of my Iuvestigatious on the
Development described in this paper and the
previous one (4) published in vol. xxviii, part 2,
of this Journal.

1. The ovum is heavily charged with food-yolk; the seg-
mentation is on the centrolecithal type; the protoplasm is
mainly at one pole of the egg, and in this protoplasm nuclei
arise, probably by the division of the segmentation nucleus.
The protoplasm forms a loose reticulum containing nuclei on
the surface of the egg, which first extends over only a small
area, but later spreads over the surface until, in the latest
stage which I have, it covers about half the periphery of
the egg.

2. In the latest segmenting ova there are small masses of
protoplasm in the centre of the egg, which masses sometimes
contain nuclei.

3. Shortly after the segmentation begins the yolk becomes
divided up into a number of rounded segments, which, however,
bear no relation to the true segmentation.

4. The central nuclei of those lying just beneath the peri-
phery multiply mnch more rapidly than those over the rest of
the ovum, thus coming to form a special area, which finally
extends along about the middle third of the ovum, and consists
of a loosely-reticulate mass of protoplasm containing a large
number of nuclei, and having in transverse section an irregular
triangular shape. Nuclei are present through the rest of the
ovum, being more numerous near the periphery than the
centre.

5. The triangular-shaped protoplasmic area becomes more
compact and flattens itself out, forming a plate-like mass of
protoplasm densely packed with nuclei on the surface of the
middle third of the ovum. This plate is the blastoderm. The
nuclei over the rest of the egg have undergone no change.

6. The blastoderm grows round the ovum till it covers about

DEVELOPMENT OF PERIPATUS NOY2E-ZEALANDI2E. 289

one half of its surface, at which time it is thickest in the centre
and thins gradually towards its edges.

7. The epibolic growth of the blastoderm continues until
only a very small space in the middle of the ventral face of
the ovum is left uncovered.

8. A proliferation of the nuclei behind the uncovered area
in the middle line takes place, forming a (in transverse section)
keel-shaped mass of nuclei extending along about the posterior
half of the ovum. The nuclei at the sides of the space also
proliferate.

9. The protoplasm round the space becomes inflected, and so
forms a blastopore. The proliferating mass of nuclei or primi-
tive streak increases in amount.

10. The blastopore increases in length considerably, and
becomes more open. The primitive streak also becomes wider
and deeper, and a groove — the primitive groove — appears along
its centre. Beneath the primitive groove a small cavity filled
with yolk and bounded by columnar nuclei, and apparently
homologous with the polar area of P. capensis, appears.

11. The blastopore and primitive streak and groove increase
and become more marked. The polar area disappears.

12. Up to this stage no trace of any cell-outlines is visible,
but the protoplasm forms a syncitium, in which nuclei are
irregularly scattered. At this point a large gap is present in
my investigations.

13. A layer of yolk is present outside the embryo. This
peripheral yolk becomes gradually absorbed, and various
changes are undergone by the embryo (descriptions of which
are given in (4), vide summary, p. 230) until it reaches the
stage at which the absorption is complete, when the appendages
begin to appear, &c.

14. The later development, i.e. after the appendages are
formed, is similar to that of P. capensis, the only interesting
point of difference being that the duct of the first somite opens
to the exterior.

290

LILIAN SHELDON.

General Considerations.

The investigations which I have made on the January
eggs of Peripatus novse-zealandise, although the stages
examined were few, nevertheless throw a good deal of light on
the subject of the early history of the development. In my
former paper (4) I remarked upon the strange dissimilarity
which existed from the segmentation stages up to quite late
ones between the three species of Peripatus whose develop-
mental history has been at all fully worked at. In the cases
of P. capensis and P. n ovse-zealandise at all events this
remark now requires modification. The developmental history
of the latter is iioav fairly complete as far as the gastrula
stage, and up to that point its resemblance to that of P.
capensis is very marked. As I pointed out before (4) the
segmentation is very similar, the main differences being easily
accounted for by the presence of the yolk in the one species,
and its almost total absence in the other. I have now shown
that in the New Zealand species the ectoderm, which at first
covers only a portion of the ovum, gradually grows round
until only a small space on its ventral side remains uncovered,
and at this spot an invagination takes place forming the
blastopore, behind which in the middle line the primitive
streak and groove are present. In all these stages the resem-
blance to the corresponding ones of P. capensis is very
striking, the main difference consisting, as in the segmenta-
tion stages, in the presence of the yolk. This similarity is
clearly seen on a comparison of the figures in Mr. Sedgwick’s
paper (1) and my own (4). In fact it seems somewhat strange
that the almost total loss of the yolk, which must almost cer-
tainly have been possessed originally by the Cape species,
should have apparently been accompanied by so few modifica-
tions in its development, since so important a change of con-
ditions might have been expected to exert a considerable
influence on the latter.

Unfortunately there are many stages wanting between the
gastrula stage and the next one which I have described in my

DEVELOPMENT OP PERIPATUS NOViE-ZEALANDIiE. 291

previous paper (4). In it one of the most remarkable features
was the presence of the yolk outside the embryo, between it
and the vitelline membrane. It might be more correct and
intelligible to consider this as ectodermic yolk. In P.
capensis protoplasmic strands are present, passing from the
ectoderm to the egg-shell, especially in the region of the dorsal
hump, and these very probably indicate that ectodermal yolk
was present at one time in this position. In his last paper (5)
on the development of the Cape species, Mr. Sedgwick states
that the ectoderm is much vacuolated and contains globules
which he believes to be yolk up to a comparatively late stage of
development. Thus it seems probable that both species were pos-
sessed of ectodermal yolk. In the gastrula stage in both species
there is no sign of any trace of yolk which probably therefore
arises later. In P. novse-zealandise this yolk is so thick that
it completely obscures the external characters, which cannot be
made out in surface view till the stage at which the appendages
are forming when the ectodermal yolk is almost completely
absorbed. As to the mode and time of origin of this ecto-
dermal yolk in P. novse-zealandise I am not able to make
any statement, as in the youngest egg in which it is present it
is already well formed, and constitutes a very thick layer; and
I do not know whether it is derived in some way from the
central yolk, or whether it arises as a fresh formation in the
ectoderm cells. However, the fact that in both species it is
absent in the gastrula stage and appears later seems to point
to its being an ancestral feature in the development. This, as
well as many other points of interest in the development of
this interesting species, will unfortunately have to remain un-
explained until someone shall be fortunate enough to obtain
embryos of the intermediate stages.

Summary of Dates of Embryos which are figured in
this paper and the previous one (4).

Previous paper (4) :

Figs. 1 — 10. — December. Figs. 11 — 20. — April. Figs.

292

LILIAN SHELDON.

21a, b, c. — July. Fig. 22. — July. Fig. 23. — December.

Fig. 24. — December. Fig. 25. — July. Fig. 26. — July.

This paper :

Figs. 1 — 9. — January. Fig. 10. — July. Fig. 11. — July.
Thus, speaking generally, the ages of the embryos received
in the various months are :

December. — Stages from unsegmented ova up to that
at which nuclei were present at intervals just beneath the
surface round about half the ovum.

January. — From late segmentation up to gastrula stage.
April. — Two embryos showing the beginning of the for-
mation of the blastoderm, and also several stages with ecto-
dermal yolk.

July. — Stages in which the appendages are being formed
up to embryos which were ready for birth. There were also
newly-born young.

Conclusion arrived at as to Time of Development.
Probably the ova pass from the ovary into the uterus in
December, and the young are born in July, the development
thus occupying a period of about eight months.

This, though apparently usually the case, cannot be universal
since in each lot there were one or two females which con-
tained embryos ready for birth, and also the embryos in one
female vary somewhat in age.

This statement as to the period of gestation has already
been made by Mr. Sedgwick (3).

List of Papers referred to.

(1) Sedgwick, A. — “ The Development of the Cape Species of Peripatus,”

Part II, ‘ Quart. Journ. Micr. Sci.,’ vol. xxvi.

(2) Sedgwick, A. — “ The Development of the Cape Species of Peripatus,”

Part III, ‘ Quart. Journ. Micr. Sci.,’ vol. xxvii.

(3) Sedgwick, A.— “A Monograph of the Species and Distribution of the

genus Peripatus,” ‘ Quart. Journ. Micr. Sci.,’ vol. xxviii.

(4) Sheldon, L.—“ On the Development of Peripatus novse-zealandiae,”

‘Quart. Journ. Micr. Sci.,’ vol. xxviii.

(5) Sedgwick, A. — “ The Development of the Cape Species of Peripatus,”

Part IV, ‘ Quart. Journ. Micr. Sci.,’ vol. xxviii.

DEVELOPMENT OF PEEIPATUS NOVAl-ZEALANDlAi. 293

EXPLANATION OF PLATES XXY & XXVI,

Illustrating Lilian Sheldon’s paper, “ On the Development of
Peripatus novse-zealandiae.”

List of Reference Letters.

Bl. Blastoderm. Btp. Blastopore, c. Cavity, corresponding to polar area
of P. capensis. Pm. A. Protoplasmic area. Pr. Gr. Primitive groove.
Pr. St. Primitive streak.

All the figures were drawn with Zeiss’s camera lucida ; Figs. 3 a and 9 were
drawn under Zeiss’s oc. 2, obj. cc; Fig. 4 under Zeiss’s oc. 2, obj. a; and
the rest under Zeiss’s oc. 2, obj. B.

Fig. 1. — Transverse section through an ovum, in which the blastoderm has
grown nearly half way round the yolk.

Fig. 2. — Transverse section through the centre of an ovum, in which the
yolk is nearly covered by the blastoderm.

Fig. 3. — Transverse section through an ovum, in which the primitive streak
is beginning to arise.

Fig. 3 a. — Transverse section through the portion of the same ovum which
is not covered by the blastoderm, drawn under a higher power.

Fig. 4. — Transverse section through the centre of a young ovum before the
formation of the blastoderm, showing the multiplication of the nuclei in one
region near the periphery.

Figs. 5, G, and 7 . — Three transverse sections through an ovum in which
the blastopore is well formed.

Fig. 5. Through the blastopore.

Fig. G. Just behind the blastopore.

Fig. 7. Near the posterior end of the primitive streak.

Fig. 8. — Transverse section through an ovum with a primitive streak.

Fig. 9. — Transverse section through a portion of the same ovum, to show
the invagination at the blastopore beginning at the anterior end of the primi-
tive streak. Brawn under a higher power than Fig. 8.

Fig. 10. — Transverse section through an embryo with appendages, to show
the duct of the first somite opeuing to the exterior.

Fig. 11. — Transverse section through the same embryo four sections pos-
terior to Fig. 10, to show the duct opening into the first somite.

VOL. XXIX, PART 3. NEW SER.

U

. -Xwr. S&ivrn. vcX. XXIX, XSffi. A.

prst.

PrGr

Y Hath.LiOj'Eajn'

F Huth, Lithr F.dinr

NOTE ON THE DEVELOPMENT OP AMPHIBIANS. 295

Note on the Development of Amphibians, chiefly-
concerning the Central Nervous System;
with Additional Observations on the Hypo-
physis, Mouth, and the Appendages and
Skeleton of the Head.

By

Henry Orr, Pli.D.,

Princeton, New Jersey.

With Plates XXVII, XXVIII, and XXIX.

The material for the investigations which are described in
this article was collected at Princeton, N.J., except a few
specimens of Triton alpinus kindly given me by Professor
H. F. Osborn. Through a failure to obtain adult specimens
at the time when I obtained the embryos, there remains room
for doubt as to the exact species of the embryos. The Ambly-
stoma embryos correspond exactly to Clarke’s1 description of
the embryos of Amblystoma punctatum, but there is one
difference in the appearance of the egg-membranes which leads
me to think that this may be a different species from that
described by Clarke. It is, perhaps, A. bicolor, for which
Jordan2 gives only the habitat New Jersey. The Frog em-
bryos are either Ran a halecina or R. palustris. I judge
them to be the former. In the stages of development with

1 S. F. Clarke, “Development of Amblystoma punctatum. Part I,
External,” ‘ Studies from the Biological Laboratory of the John Hopkins
University,” No. ii, 1880.

2 D. S. Jordan, ‘Manual of the Vertebrates of the Northern United
States, &c.,’ 187G.

296

HENRY OBR.

which my work has dealt there are probably no specific
differences in the embryos.

The developing eggs of this species of Amblvstoma seem to
present a remarkable case of symbiosis. The eggs, surrounded
by their gelatinous matrix, appear as a white mass floating on
the surface of the water. (I found them in a small swampy
pool on elevated ground.) In the first lot that I collected the
medullary plates were just forming, and the two membranes
surrounding each egg appeared perfectly homogeneous and
transparent. In the second lot, collected some days later, the
embryos were somewhat elongated, and the medullary canal
had apparently just closed. In this lot the internal membrane
of each egg was coloured a uniform light green by the presence
in the membrane of a large number of minute globular green
Algse. Neither in the spaces adjoining the internal membrane,
nor in the external membrane, nor in the matrix, was there
any colouration or trace of this Alga. The external membrane
was transparent and the matrix white and translucent as before.
In a third lot, collected when the balancers and gills had
appeared, these conditions were the same, except that the
Algje had increased in number and the colour was a much
darker green. I have not discovered how the Algse enter the
membrane, nor what physiological effect they have on the
respiration of the embryo, but it seems probable that in this
latter respect they may have an important influence.

Clarke has given a detailed description of the external
appearance and changes of the embryo of Amblystoma
punctatum, so that for this part I may refer the reader to his
work. As might have been expected, Amblystoma and Triton
present much similarity in their development, while both differ
in about the same degree from the Frog. The chief points of
which I shall treat are the central nervous system, the hypo-
physis, and some other parts and appendages of the head.
The comparison of the embryos of the different groups affords
some light in the attempt to explain the development of some
of the more complicated parts. In order to avoid repetitions
I have not separated the descriptions of parts which are inti-

NOTE ON THE DEVELOPMENT OP AMPHIBIANS. 297

mately related to each other in the process of their development.
Much of what I have written will have been readily inferred
by embryologists, though the embryology of the genus on
which most of my work is based has not hitherto been worked
out. But as some of my conclusions are different from those
previously expressed, I have thought best to give in detail a
description of the facts that the reader may thereby test my
conclusions.

The first differentiation of the central nervous system of
Amblystoma appears as figured in the sagittal and transverse
sections (figs. 1, 2 a, G a). The transverse section is cut through
the middle dorsal region. By the thickening of the dorsal
epiblast there are formed two broad epiblastic plates (M.P.),
connected with each other on the median line by a thinner
portion of epiblast. A slight longitudinal groove ( G .) is formed
by the sinking inward of this thinner median portion of epi-
blast. Directly underneath this groove there is a longitudinal
fold in the hypoblast, which causes a conspicuous median
groove in the roof of the alimentary tract. The dorsal part
of this hypoblastic fold touches the epiblast along the median
line; and the part adjacent to the epiblast is the part which
later forms the notochord. On each side of the hypoblastic
fold, and apparently fused with it, lie the two layers of the
mesoblast [So., Sp.). An examination of all my sections shows
that the condition of the mesoblast at this point is the same in
Amblystoma as Her twig discovered it to be in Triton.1
Some of my sections show a space between the two layers of
mesoblast continuous with the archenteron. I have not found
any trace, however, of mesoblast originating from any other
part of the hypoblast or from the yolk.

The sagittal section (fig. 1) passes through the epiblastic
groove (G.) and the dorsal groove in the hypoblast, thus cutting
only the thin median part of the epiblast aud the hypoblast.
The sections on either side of this median section pass through
the thicker part of the body wall, which contains also the

1 For a statement of Hertwig’s results, see A. C. Haddon, ‘ An Introduc-
tion to the Study of Embryology,’ 1887.

298

HENRY ORR.

mesoblast. In section, fig. 1, it may be seen that the thinner
epiblast in the median line does not extend to the extreme
anterior end of the rudiment of the nervous system, but that
it ceases in the head region, while anteriorly the lateral
medullary plates unite with each other undiminished in thick-
ness, thus forming an anterior medullary plate (a. M. P.).
The distinction between anterior and lateral plates is arbitrary
and adopted only for convenience. They might be described
as one thickened epiblastic plate, bent in such a manner that
the curved part lay in the head region, while the two straight
parallel ends lay one on each side of the dorsal median line.
The distal periphery of the anterior medullary plate is a curve.
Fig. 6 a represents a section through the anterior plate of the
same embryo from which fig. 2 a was taken. It may be seen
here that there is no sign of a bilateral division of the neural
rudiment at its anterior end.

The further development of the medullary plates in the
dorsal region is shown in figs. 3 b, 4 c, 5 d, PI. XXVII. The
lateral edges of the plates roll slightly upwards, forming the
folds ( M.F. ). At the same time the median edges become
pressed together, thus causing the floor of the median groove
(G.) to sink farther inwards. The space between the medullary
folds [M. F.) gradually decreases in size as the folds approach
the median line. There is no very marked increase in the
height of the folds. The originally dorsal surfaces of the
medullary plates, bending inward, become pressed together in
a vertical median plane under the groove, G. Across this line
of median vertical contact there is no fusion of the cells. A
heavy pigment marks this line (g., fig. 5 d) as continuous with
the pigmented surface. Beneath the lower extremity of the
line g. a small ridge of cells, continuous with the lateral halves
of the neural rudiment, marks the original median connection
of the medullary plates. By a comparison of the figures
illustrating this period of development it may be seen that the
cells of the neural rudiment gradually become smaller (owing
to division and loss of yolk), that the whole organ becomes
more compact and occupies much less space in the end than in

NOTE ON THE DEVELOPMENT OF AMPHIBIANS.

299

the beginning. The primarily broad space enclosed laterally
between the medullary folds (M. F.) diminishes in size uutil it
becomes a small rounded groove in the dorsal part of the
neural rudiment, as represented at fig. 5 d. In a section
through the cervical region of the same embryo (fig. 10 d) the
epiblast has met above the groove, thus forming a relatively
very small neural canal. As the epiblast of the two sides
fuses above the canal the lumen of the latter becomes in some
places suppressed, leaving as its only remnant a heavy accu-
mulation of pigment.

In the species of Frog examined by me the lumen of the
neural canal at this period of development becomes suppressed
throughout the posterior part, thus differing in this respect
from Goette’s account of Bombinator. Towards its posterior
extremity the neural rudiment of the Frog closely resembles
that of an osseous fish at the same period. This manner of
development of the neural rudiment in Amphibians presents a
stage intermediate to the condition of Elasmobranchs and
Reptiles on the one hand and the condition of Petromyzon and
the Teleosts on the other.

After the fusion of the epiblast dorsal to the neural rudiment
the neural canal opens from before backwards aloug the
pigmented line (</., fig. 5 n) which has previously been de-
scribed. The canal, at first somewhat irregular, becomes in
cross-section dorso-ventrally elongated. The walls of the
neural tube become bilaterally symmetrical, and are thickest
laterally. A transverse section through an older embryo of
Amblystoma (fig. 11) shows the result of these changes. In
the Frog embryo the appearance is fundamentally the same.

In the cephalic region the development of the neural rudi-
ment differs from that in the dorsal region. In order to clearly
understand this difference, it is necessary to bear in mind that
the anterior medullary plate (a. M. P., figs. 6 a, and 7 b) is
not a paired continuation of the dorsal medullary plates, but
is a transverse curved plate connecting the two dorsal plates.
It is also necessary to distinguish the modifications caused by
tlie cranial flexure, in order to recognise the homology between

300

HENRY ORR.

the parts in the dorsal and cephalic regions. It will be seen
in the sagittal section (fig. 1) and in the transverse section
(fig. 6 a) that at first the anterior medullary plate (a. M. P .) is
externally flattened. In fig. 7 b, where the dorsal medullary
folds have appeared (compare fig. 3 b), the lateral edges of
the anterior medullary plate turn slightly upward (where
the same letters are affixed to the numbers the
sections are from the same embryo). In fig. 8 c this
upward bending of the lateral edges has increased, and in
fig. 9 n the edges meet dorsally. During this process the
median part of the anterior medullary plate (A. F.) departs
from its original slanting position (fig. 1), and comes to lie
nearly parallel to the dorsal surface of the embryo, though at a
lower level. The floor of the dorsal medullary groove extends
forwards nearly horizontal as far as the region of the mid-
brain ; there it bends downwards almost at a right angle, and
joins the posterior edge of the medially horizontal anterior
plate. Thus is formed the primary cranial flexure before the
medullary folds have fused above the neural canal. In the
head this fusion takes place later from behind forwards, curving
down to the anterior edge of the anterior medullary plate
(. A . F., fig. 9 d). The cranial flexure therefore is not simply
a bend in the floor of the primitive neural tube, but is also a
bend in the dorsal surface. It involves the anterior part of
the neural tube in a bend about equal to a right angle. The
line of fusion of the medullary folds in the head is homologous
with the same fusion in the dorsal region. The morpho-
logically dorsal surface of the neural tube extends therefore
throughout the region of the fusion to the anterior edge of the
anterior medullary plate. Taking into account the anterior
bending of the axis of the neural tube, its morphologically
anterior surface would be represented by the anterior medul-
lary plate, which extends from the above-mentioned vertical
portion of the floor to the anterior end of the dorsal fusion.

The anterior medullary plate of Amblystoma is homologous
with the anterior medullary fold of the Lizard, and for the pur-
pose of indicating this homology I have marked it in the drawings

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 301

as the anterior medullary fold. In both cases it forms the
primitive morphologically anterior surface of the brain. There
is a marked difference between this anterior brain-surface in
the Lizard and the same part in Ambly stoma. In the Lizard
the anterior brain-surface comes to lie at a right angle to the
axis of the dorsal part of the neural tube, and faces posteriorly ;
in Amblystoma it lies parallel to the axis of the dorsal part of
the neural tube, and faces veutrally. This difference seems to
be due to the different methods according to which in the two
forms the medullary folds unite to form the medullary tube.
In Amblystoma the condition is caused in the following manner.
In the primitive neural rudiment there is a thinner median
portion of epiblast lying between the dorsal medullary plates
and behind the anterior medullary plate. As the distal lateral
edges of the neural rudiment approach each other different
effects are produced in the region of the thin median epiblast
and in the anterior plate. In the first-named region, as the
lateral edges of the medullary plates approach each other,
their median edges are compressed, and as the width of the
neural rudiment decreases its median thickness increases. In
the anterior plate there is no thin median portion and no
thickening resulting from compression, therefore as the lateral
edges approach each other the median portion must bend
downward. In this manner the median portion of the anterior
plate comes to lie at a much lower level than the floor of the
neural tube in the dorsal region. The cranial flexure is the
result of the presence of an anterior medullary plate, and, as I
have elsewhere pointed out, this seems to be the case also in
the Lizard.

In the Frog the anterior medullary plate forms a fold
directly comparable to the medullary folds in the dorsal region.
The anterior fold is, however, much more prominent than the
folds in the dorsal region. Fig. 19 represents a median sagit-
tal section of a Frog embryo at this stage. The lateral sections
of this embryo show that the anterior fold (A. F.) is laterally
and posteriorly continuous with the paired medullary folds,
thus enclosing anteriorly and laterally the anterior enlarge-

302

HENRY ORR.

meat of the neural groove ( F . B.). This anterior enlargement
is the first rudiment of the vesicle of the fore- brain. The
cranial flexure in this embryo is in process of formation ; when
the medullary folds in the head later meet dorsally, the cranial
flexure is complete. The presence of an elevated anterior fold
in the Frog, and its absence in Amblystoma, is not so much
due to absolute difference in the form of the neural rudiment
as to the relative growth of the surrounding parts. In Am-
blystoma the presence of the hypoblast and anterior end of the
alimentary tract beneath the anterior medullary plate (fig. 9 d)
prevents the latter from appearing as a fold raised above the
head surface. But at a later period the hypoblast disappears
from beneath the anterior plate, and the external surface of
the anterior plate is then covered only with epiblast (fig. 12 e).

The disappearance of the hypoblast and alimentary cavity
from beneath the anterior medullary plate, or rather the
(morphologically) anterior surface of the brain, is due to the
more rapid growth of the brain, especially an increase of
length, by which the fore-brain advances to a position in front
of the anterior end of the alimentary cavity. At a very early
stage the anterior end of the alimentary cavity is enclosed
only by hypoblast and epiblast (Ep., Hyp., fig. 1). A fusion
of these two layers soon takes place at this point, and indicates
the eventual position of the mouth-opening. As the fore-brain
is projected anterior to this mouth-fusion, the epiblast dorsal
to the fusion is brought into close contact with the anterior
surface of the bi’ain (fig. 12 e). Figs. 12 e and 13 e represent
two nearly sagittal sections of the same embryo, one section
passing through the oral fusion and hypophysis-rudiment,
the other passing through the notochord and pineal rudiment.
The age and general condition of development of this embryo
will be best understood by comparing these sagittal sections
with sections 14 f, 15 f, 16 f, which are horizontal and taken
from an embryo of the same age. The anterior part of the
alimentary canal is distended into a large pharyngeal branchial
cavity (fig. 12 e). The hypoblast of the anterior wall of this
cavity touches the nearly vertical floor of the fore-brain which

NOTE ON THE DEVELOPMENT OF AMPHIBIANS.

303

forms the wall of the infundibulum. The lower anterior wall
of the pharyngeal cavity is fused with the epiblast at M., form-
ing the oral fusion. A wedge-shaped mass of epiblast ( Hph .)
extends inward between the oral fusion and the wall of the
infundibulum ; this is the rudiment of the hypophysis. It is
not necessary to interpret this condition as an ingrowth of
the epiblast. I am inclined to think that the wedge-like
shape of the epiblastic mass is due to the pressure of the more
rapidly growing brain. It is evident from this section that at
this stage of the development of Amblystoma there is no ap-
pearance of a stomodaeum or epiblastic mouth-cavity. From
this time on the rudiment of the lower jaw begins to extend
forward, and grows beyond the oral fusion and hypophysis
toward the nasal tip of the head. The epiblast retains its
connection with the hypoblast, and also for a time with the
hypophysis ; thus the epiblast posterior to its point of fusion
with the hypoblast is pressed close against the epiblast anterior
to the point of fusion. These two united layers of epiblast
form an apparently solid mass extending from the hypoblast
to the surface of the head ( M .). This stage is illustrated in the
nearly median sagittal section, fig. 17 g. The hypophysis
[Hph.) has broken loose from the in-folded mass of epiblast,
and still remains adjacent to the posterior wall of the infun-
dibulum (In.). The point for the ultimate external opening
of the mouth ( M .) has been moved by the growth of the lower
jaw, forward to a position anterior to region of the optic
chiasma ( Ch .). The position of the perfected mouth-opening
is shown in fig. 18. The condition of the rudiment of the
mouth, as represented in fig. 17 g, is that which has been
described by other writers as a solid ingrowth of epiblast or a
stomodaeum ; but it is evident from the above-described manner
of development that the term ingrowth leads to a false con-
ception as to the origin of the part referred to. The primary
development of the hypophysis, and the growth forward of the
lower jaw, are fundamentally the same in Amblystoma as I
found them in the Lizard.

During the process above described, the parts of the brain

304

HENRY ORR.

and the hypophysis and notochord change their positions with
relation to each other. Fig. 13 e shows the anterior end of
the notochord, which in this embryo is at some distance from
the hypophysis, while the floor of the hind-brain ( H . B.) is
widely separated from the infundibulum. There is a median
thickening of the hypoblast extending from the anterior end
of the notochord down to the hypophysis. This thickening
seems to disappear very quickly after formation. It seems
possible that this median thickening may be homologous with
that foremost part of the notochord which in the Lizard and
in the Mole extends as far as the epiblast at the hypophysis.
In the Anura at an early stage there is a layer of mesoblast
extending across the median line between the anterior end of
the notochord and the hypophysis-rudiment. Why the meso-
blastic product of the hypoblast along the median line at this
region does not become differentiated into notochord in the
Amphibia, as it does in the Lizard and the Mole, may be ex-
plained perhaps by the changes which immediately succeed
this stage — changes which would be hindered by a developed
notochord in this region. The changes thus referred to are
exhibited in fig. 17 g. Here the secondary cranial flexure has
appeared in the hind-brain, and the floor of the hind- brain is
pushed against the infundibulum, causing the latter to be
slightly compressed. At the same time the bending floor of
the hind-brain has pushed the notochord downward, so that
the anterior end of the developed notochord touches the hypo-
physis. (These changes of position are of course to be under-
stood only in terms of relative topography as the absolute
changes of location cannot be ascertained. Thus, the changes
might be accounted for by supposing the secondary cranial
flexure to lift the anterior part of the brain and head upward ;
but the former view lends itself more readily to the explana-
tion of the facts, and admits of more extended homologies).

As nearly as can be judged from the more limited number
of my specimens of Triton, the method of development during
the above-described stages is exactly the same in Triton as in
Amblystoma; though I should add that my youngest stage of

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 305

Triton corresponds with the stage of Amblystoma represented
in figs. 12 e to 16 e. From this stage onward my series of the
embryos of the two genera run about parallel, and a great
similarity continues to exist throughout all the stages which I
have examined.

The development of the hypophysis and mouth in the Frog
differs in a marked manner from the development of the same
organs in Amblystoma and Triton. The same fundamental
principles seem to obtain in both methods of development, but
the difference is apparently due to a different proportional rate
of growth of the parts adjacent to each other. The development
of these parts in the Frog is illustrated in figs. 19 — 23. These
sections are sagittal, or nearly sagittal, and all meet the
median vertical plane in the centre of the mouth-fusion. In
fig. 19, between the lip of the anterior medullary fold (. A . F.)
and the mouth-fusion, lies the epiblast which is to form the
hypophysis. In this embryo the cranial flexure is not yet
complete, and the alimentary cavity extends forward beyond
the anterior fold. The rudiment of the hypophysis lies there-
fore immediately exterior to the anterior fold. In a somewhat
older embryo (fig. 20) the brain is enclosed, and has increased
so much in size that it projects forward anterior to the mouth-
fusion ( M .). The increase of the cranial flexure has caused a
change in the position of the anterior fold. In fig. 19 the
anterior fold occupies a vertical position, and in fig. 20 it
occupies a horizontal position (A. F.), forming in both cases
the morphologically anterior wall of the brain. In embryos
slightly younger than the one represented by fig. 20 serial
sections show that the dorsal linear opening of the central
nervous system extends as far as the horizontal anterior fold
to about the point indicated by o. g. in fig. 20. When this
opening becomes closed by the dorsal median fusion of the
lateral walls, the line of fusion remains marked by the accumu-
lated mass of cpidermoidal pigment. This pigmented line is
cut at o.g. in fig. 20, very near the end which indicates the
boundary of the anterior fold. It may be seen from the
figures that the change of position of the anterior fold is

306

HENRY ORR.

accompanied by a corresponding change in the position of the
hypophysis rudiment ( Hph .), so that the latter continues in the
same topographical relation to the anterior fold. The rudiment
of the hypophysis extends a short distance posterior to the
limit of the anterior fold (o. g.).

The next three stages (figs. 21, 22, 23) illustrate the further
development of these parts. The most striking changes are
the increase of the cranial flexure and the growth of the dorsal
part of the fore-brain. (This latter is not so well shown in
fig. 21 owing to the obliquity of that section.) It is evident
that these changes would cause a relative change of position
of the point marked o.g. in fig. 20. In two of the sections
(figs. 22, 23) may be seen a slight groove (o.g.) in the mor-
phologically anterior surface of the brain. This groove lies at
first between the optic stalks, and ultimately just anterior (or
morphologically dorsal) to the chiasma. I have not been able
to absolutely demonstrate that the groove (o. g.) is developed
from the point o.g. in fig. 20, but the evidence in favour of the
view that such is the case seems to me so strong that I have
been forced for the present to accept that conclusion. In the
Lizard the primitive opening of the brain extends down the
anterior surface of the brain to a point between the optic stalks,
and in the Lizard there is also a similar groove at that point.
In the present case we have only to imagine that owing to the
increase of the cranial flexure and the growth of the fore-brain
the point o.g., fig. 20, has receded relatively in a posterior
direction, until it reached the point o. g., fig. 23. In figs. 22
and 23 such a relative posterior recession of the groove o. g. is
perfectly evident. This relative recession is due chiefly to the
greater growth in the region in front of the point o. g. It will
be seen that in all these five embryos (19 — 23) the posterior
end of the hypophysis-rudiment lies at about the same distance
behind the region of the point o.g., but the lower jaw advances
continually until it extends anteriorly beyond the posterior end
of the hypophysis and beyond the point o. g. This process of
growth is essentially the same in the Frog as iu Amblystoma
and Triton ; but in the Frog the growth of the dorsal part of

NOTE ON THE DEVELOPMENT OE AMPHIBIANS. 307

the fore-brain and the growth forward of the lower jaw take
place at the same time, and in nearly the same extent, thus
making the hypophysis appear as an ingrowth, whereas it is
simply that part of the epiblast which has retained its original
position with relation to the brain, and which has become sur-
rounded and embedded by the expansion of the adjacent parts.
There is another point of difference between the hypophysis of
the Frog and the hypophysis of the Urodele embryos. In
Amblystoma and Triton the hypophysis at the very beginning
of its differentiation lies immediately adjacent to the posterior
wall of the infundibulum, and later the anterior end of the
notochord touches its posterior side. In the Frog the hypo-
physis at first does not reach the posterior wall of the infundi-
bulum. As it begins to loose its connection with the epiblast
it gradually comes to lie nearer the posterior wall of the
infundibulum, and finally lies slightly ventral to the anterior
end of the notochord, the latter being pressed against the
infundibulum. Thus a nearly similar condition results from
two apparently different methods of development. In Ambly-
stoma the position of the hypophysis is the result (mechanically)
chiefly of a forward movement of the anterior part of the brain.
It seems most probable that the case is the same in the Frog,
but that the forward movement of the anterior part of the
brain takes place at a later date.

To GoetteV description of the other parts of the brain of
Anura during these stages I have nothing to add. There are
a few points, however, which may be mentioned for the sake of
orientation as to the stages of development of the embryos
here referred to. In a transverse section through the head of
an embryo at the stage of fig. 20 the lumen of the fore-brain
appears triangular, with one angle representing the dorsal
crest of the brain, and the side opposite that angle representing
the morphologically anterior wall of the brain. The lateral
angles of the lumen are the beginnings of the optic outgrowths.
In au embryo at the stage of fig. 21 the optic outgrowths are
somewhat prolonged, and the lumen is drawn out laterally in
1 Goette, ‘ Die Entwickelungsgescbichte der Unke,’ Leipzig, 1875.

308

HENRY ORE.

them. In the embi’yo of fig. 22 the optic outgrowths are bent
backwards and upwards, and in the embryo of fig. 23 the eye
has progressed so far that the lens has appeared.

In the Amblystoma embryo of series d, in which the fore-
brain is not yet enclosed, there is no trace of the optic vesicles.
The next older stage of Amblystoma among my specimens is
illustrated in figs. 12 e — 16 f. The condition of the cranial
flexure is shown at 12 e. In the anterior wall of the brain
may be seen the optic groove (o. g.), and behind the latter is
the anterior fold (A. F.). Immediately posterior to the anterior
fold is the rudiment of the infundibulum. At this stage the
primary triple division of the brain is not yet very pronounced,
and there is no trace of nerve-fibres in the brain. The position
of the rudiment of the epiphysis ( Eph ., fig. 13 e) indicates the
posterior extent of the primary fore-brain. In fig. 16 f the
fore-brain is represented in section parallel to its morphologi-
cally anterior surface very near the latter, and in the region of
the optic stalks {Eg.). In this section the lateral thickenings
of the brain wall in front of the optic stalks are the rudiments
of the corpora striata, which appear much earlier in Amblystoma
than in the Lizard. Fig. 15 f represents a horizontal section
of the embryo passing through the dorsal part of the pharyngeal
cavity and through the mid-brain above the region of the in-
fundibulum. This section shows the rudiments of the eyes
(-By.), which as yet possess no lens. Fig. 14 p represents a
horizontal section through the hind-brain and dorsal medulla.
This section shows the rudiments of the fifth, seventh and
eighth, ninth and tenth cranial nerves. In three places the
hind-brain shows a marked dilation of its lumen, and the
lateral walls of the brain pass around these dilated parts un-
diminished in thickness. Opposite these dilated parts of the
lumen arise the three chief nerve-roots of the hind-brain. The
most anterior dilation corresponds to the fifth nerve- root (n. V ).
The next dilation corresponds to the common root of the seventh
and eighth nerves [n. VIII and VII), and the posterior dilation
corresponds to the root of the tenth nerve {n. X). The rudi-
ment of the ear (B.) lies between the regions of the posterior

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 309

and middle dilations, and immediately behind the ear arises
the root of the ninth nerve ( n . IX). These dilated parts of
the hind-brain in Amblystoma resemble in some degree what
I have described as the neuromeres in the hind-brain of the
Lizard, except that in Amblystoma they are fewer in number,
and certain intermediate neuromeres appear to have been sup-
pressed. I am inclined to think that the large quantity of
yolk present in these parts in Amblystoma has considerably
changed their appearance and development. These dilations
of the hind-brain have disappeared in Amblystoma, as in the
Lizard, by the time the nerve-fibres of the brain have appeared.
It will be seen in fig. 14 f that the cranial nerves meet and
fuse with the epiblast. This fusion I think corresponds with
what has been described by Miss Johnson and Miss
Sheldon5 as the first or dorsal fusion of the cranial nerves
with the epiblast. These authors have described this fusion
for the fifth, seventh, and ninth nerves, and supposed it for
the vagus. My section shows the correctness of their suppo-
sition. The vagus retains for some time this fusion with the
epiblast, and from the point, of fusion there soon grows poste-
riorly a large linear thickening of the epiblast, which forms
the lateral nerve. This in its earlier stage is very conspicuous,
but soon becomes much smaller. I have not been able to trace
the different steps between what the above-named authors have
called the “ first (dorsal) fusion” and the “ second (ventral)
fusion.” One of my series of sections of Triton alpinus
shows the condition described by them as the “ second (ventral)
fusion.” In this series the distal ends of the two primary
branches of the fifth nerve touch the epiblast and appear to be
fused with the same.

The further development of the brain is shown in figs. 17 g
and 18. The irregular appearance of these sections is due to
the fact that they are neither exactly median nor exactly vertical ;
they cross the median vertical plane in a line drawn through
the epiphysis ( Eph .) and the region of the optic chiasma ( Ch .)

1 Johnson and Sheldon, “Notes on the Development of the Newt (Triton
cristatus),” ‘ Quart. Journ. Micr. Sci.,’ vol. xxvi, N. S., 18S6.

VOL. XXIX, PART 3. NEW SER.

X

310

HENRY ORR.

and hypophysis ( Hpli .). The morphologically anterior surface
of the brain has remained in about the same position that it
occupies in fig. 12 e, but the floor of the hind-brain is bent
downward and is pressed against the infundibulum. Just an-
terior to the epiphysis ( Eph .) is a deep fold, extending trans-
versely across the dorsal wall of the brain, and thus dividing
off the secondary fore-brain. There is another longitudinal
and median fold, extending from this transverse fold forward
to the anterior surface of the brain ; thus dividing the secon-
dary fore-brain into the two hemispheres. This longitudinal
fold is not so deep as the transverse fold. Fig. 35 represents a
section transverse to the long axis of an embryo of the same
stage as fig. 17 g. This section is behind the deepest extent
of the median longitudinal fold, but still shows the transverse
fold. The rudiments of the corpora striata, which are already
evident at the stage of fig. 13 e, st., are shown again in trans-
verse section in fig. 35. The corpora striata extend parallel to
each other on each side of the median line, along the morpho-
logically anterior surface of the brain, and are limited ventrally
by the optic groove (o. g., fig. 17g). Immediately ventral to
the optic groove is seen the remnant to the anterior fold, con-
taining a bundle of transverse nerve-fibres, of which a part
form the optic chiasma ( Ch .). In an exactly median vertical
section of the brain of an embryo at the stage of fig. 17 g,
this remnant of the anterior fold would be the thickest portion
of the brain wall, being about as thick as the lateral walls of
the medulla. The thickness of the floor of the hind-brain in
the median line is shown in fig. 24 G, w. H. B.

Before the embryo of Amblystoma has reached the stage of
development represented by fig. 17 g, the first development of
nerve-fibres has taken place in the central nei'vous system.
The arrangement of these nerve-fibres corresponds very closely
to the first arrangement of the nerve-fibres in the Lizard, and
the arrangement seems to be identically the same in Triton
and Rana. The nerve-fibres in the neural tube of the dorsal
region first appear as two flat bands of longitudinal fibres,
lying next the lateral surfaces of the tube. Fig. 34 shows a

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 311

section of the neural tube of Amblystoma in the anterior
dorsal region. The band of longitudinal fibres ( L . F.) extends
nearer to the ventral median than to the dorsal median surface
of the tube. Goette has described these fibres as originating
in the external halves of the peripheral cells throughout this
portion of the tube ; while the internal half of each cell, with
the nucleus, becomes one of the cells of the grey matter.
These points I have not been able to follow out with the
material at my command-. Shortly after the longitudinal fibres
have appeared another system of fibres arises — the transverse
fibres or ventral commissure (T. F.). These fibres appear as
polar outgrowths of the cells which lie internal to the longi-
tudinal band. They pass ventrally along the inner surface of
the longitudinal band, and cross transversely the ventral sur-
face of the neural tube immediately inside the cuticula. Both
of these systems of nerve-fibres develope later in the posterior
than in the anterior part of the central nervous system. The
transverse fibres extend as a continuous ventral commissure as
far forward as the point where the floor of the mid-brain bends
ventralwards into the posterior wall of the infundibulum.
This is shown in median vertical section in fig. 24 g. The
lateral bands of longitudinal fibres extend forward through the
hind- and mid-brain, showing the same relations as in the
dorsal region (fig. 34). On passing from the mid-brain to the
fore-brain the lateral bands follow the curve of the cranial
flexure ; and on reaching the morphologically anterior surface
of the brain, they cross it, blending with each other immedi-
ately ventral to the optic stalks. The lateral bands thus blend
into an anterior band, which is cut transversely into the median
vertical sections, 17 g and 18, at Ch. This anterior band com-
prises a bundle of fibres, which I would roughly estimate to be
about twenty times as large as the bundle of fibres which
appears shortly afterwards on each optic stalk. The course of
the lateral band (L. F.) in the mid- and fore-brain is shown in
the lateral vertical section fig. 32 g; the dotted line indi-
cates the lower median contour of the brain. Fig. 33 shows
the anterior band (A. F.) of the Frog just behind the optic

312

HENRY ORR.

stalks. This section is cut transverse to the long axis of the
embryo. No fibres appear in the region of the infundibulum
which lies between the anterior band and the anterior edge of
the above-described continuous ventral commissure. Of the
brain commissures (not including the anterior band) the pos-
terior commissure is the first to appear. It developes about
the time that the ventral commissural system appears. The
posterior commissure is shown at P. C., figs. 18, 32 g, and 35.
It crosses the dorsal surface of the brain immediately posterior
to the epiphysis. Its fibres seem to be not continuous with the
fibres of the lateral bands, but, as far as they can be traced,
they cross the course of the lateral bands ; losing themselves,
however, in the region of the latter. The anterior commissure
developes relatively much earlier in Amblystoma than in the
Lizard. It first arises as two lateral symmetrical bundles of
fibres, passing along the exterior surfaces of the corpora striata
and intersecting the lateral bands just posterior to tbe optic
stalks ( A . C., fig. 32 g). This section shows that these fibres
are not continuous with the fibres of the lateral bands. A
part of these bundles of fibres crosses the anterior surface of
the brain a short distance dorsal to the optic groove at the
point A. C. in figs. 18 and 30 h. The rest of these fibres con-
tinue on toward the roots of the olfactory nerves, n. 7, fig.
29 h. A short time after the anterior band has appeared, there
appears on the morphologically anterior surface of each optic
stalk a small growth of nerve-fibres, developing as far as can
be seen, in exactly the same manner as the development of the
fibres of the lateral longitudinal bands. These optic fibres
appear at the point n. II, in fig. 32 g (Amblystoma), and are
shown in fig. 33 (Frog), where they are cut nearly longitudinally.
The latter section shows that no fibres appear in the posterior
wall of the optic stalk (op.). Medianly, the optic fibres meet
and blend with the anterior band ; distally, they pass unbroken
into the inner surface of the eye-cup (fig. 33). I have not
followed the later growth of the optic nerve in the Amphibia,
but I judge from the close similarity between this stage and a
stage in the Lizard, that the development of the optic nerve

NOTE ON THE DEVELOPMENT OE AMPHIBIANS. 313

in the Amphibia is throughout about the same as I have
described it for the Lizard.1

Figs. 27 h — 30 h show four horizontal sections through the
head of an embryo of Amblystoma at an age corresponding to
that of fig. 18. These sections show the nerve-fibres of the
brain at a more advanced period than that above described.
Of these sections, 27 h is cut nearest the dorsal surface of the
head, and on the left side passes above the lateral band of
longitudinal fibres ( L . F .) in the region of the secondary cranial
flexure just in front of the ear. On the same side of the
section the lateral band in the hind-brain is seen to be con-
tinuous with the lateral band in the mid-brain ( L F.). In front
of the mid-brain is seen the posterior part of the cerebral
hemispheres. The next more ventral section (28 h) passes
through the*thalamencephalon and through the fold which
separates the infundibulum (In.) from the hind-brain. In the
hind-brain may be seen the transverse fibres of the ventral
commissure (T. F.). These are also visible ( T.F .) in section
29 h, the hind-brain in this section being cut tangentially to
its ventral convexity. In this same section may be seen on
the right hand side the connection between the lateral baud
and those fibres which run dorsally along the corpora striata.
One part of these fibres forms the anterior commissure as
above mentioned (fig. 30 h, A. C.) ; while the other part con-
tinues onward to the region of the olfactory nerve (n. I), and
here blends with a superficial layer of nerve-fibres, which
covers the lateral dorsal part of each hemisphere, and extends
so far upwards and backwards as to appear in section 27 h.
Fig. 30 u shows the brain in section very near its anterior
surface. At A. F. may be seen the fibres of the anterior band,
with the fibres of the optic nerve ( n . II) blending with its
dorsal edge. At A. C. may be seen the fibres of the anterior
commissure. Between the thickening of the anterior band
and the anterior commissure appears the optic groove ( o.g .).

1 Orr, "Contribution to the Embryology of the Lizard,” ‘Journal of
Morphology,’ vol. i, No. 2, 1887.

314

HENRY ORR.

Orientation as to the direction of this section through the
brain may be easily acquired by comparing it with fig. 18.
The section 30 h would be perfectly horizontal in the fig. 18.
Thus it enters the brain at the hinder edge of the anterior
baud and passes forward at an acute angle to the morpho-
logically anterior surface of the brain. In this way the fibres
passing from the region of the lateral bands to the anterior
commissure are cut obliquely. The relations of these fibres
to the lateral bands are shown in fig. 32 g. Here it appears
that they do not bend and run with the lateral bauds, but may
be traced for some distance, crossing the latter at right angles.
The anterior commissure is at first undivided and lies next to
the surface of the brain, but in the latest stage which I have
examined an internal part has become divided off from the
superficial part (fig. 18). This internal part I judge to be the
corpus callosum.

The growth of the hind-brain, together with its change of
form, has in this oldest stage brought the cranial nerves of
this region much nearer together. These conditions are
illustrated in figs. 27 h and 28 h. The nerve-roots which are
present form very large ganglia. The common ganglion of
the seventh and eighth nerves ( n . VIII, VII) lies relatively
much nearer the root of the fifth nerve ( n . V) than it did at
the time of its first appearance. The roots of the ninth and
tenth nerves appear to have fused in a common ganglion
( n . X, IX). This may be due to the great growth of the
auditory vesicle pushing the root of the ninth nerve backward.
I have been unable to find in these stages any traces of the
third, fourth, and sixth nerves. In the Lizard the third nerve
developes as soon as the other ventral roots of the nervous
system ; the sixth nerve developes somewhat later than the
other cranial nerves, except the fourth, which first appears at
a stage much later than the present stage of Amblystoma.
The olfactory nerve ( n . I) is shown in fig. 29 h, entering the
olfactory sac (N. a.). The course of this nerve from its origin
in the fore-brain is backwards and downwards. The fibres of
the optic nerve are also shown in fig. 30 h, entering the brain

NOTE ON THE DEVELOPMENT OP AMPHIBIANS. 315

at n. II, where they join the forwai’d or doi’sal edge of the
anterior band of fibres (A. F.).

Appendages and Skeleton of the Head.

The gill-clefts develope in Amblystoma after the usual
manner from before backward. The first or hyoid cleft (I)
does not break through, but forms like the others a laterally
extended hypoblastic pouch (figs. 15 f, and 26). In the case
of the hyoid this pouch extends in a venti’al and median direc-
tion, forming a groove which meets a similar groove from the
opposite side. The median portion of this groove is shown in
the longitudinal vertical section of fig. 18, Th. From com-
parisons with the work of other writers I suppose this part
marked Th. to be the rudiment of the thyroid gland, though in
this case I have traced the development no farther. Whether
this relation of the thyroid rudiment to the hyoid clefts can
be considered as an argument for the phylogenetic origin of
the thyroid gland from the ventral coalition of the hyoid clefts,
is, I think, doubtful. The ventral groove may be the result of
the early development of the tongue-rudiment. In the Lizard
the hyoid clefts are widely opeix to the outside, and the thyroid
rudiment appears between the transverse areas of the hyoid
and first branchial clefts. The thyroid rudiment in the Lizard
has no appai’ent connection with the hyoid clefts.

In the stage represented in fig. 15 f, the hyoid ( I) and the
first two bi’anchial cleft-rudiments (II, III ) have appeared ; in
the stage of fig. 26 five in all have appeared (I — V), but none
of them have as yet broken through. These stages show the
development of the head-cavities or mesoblastic somites of
the head. The anterior somite is the first to develope, and
appeal’s just behind the eye. The other somites are separated
off from this first one by the successive development of the
hyoid and branchial clefts. These somites of the head do not
attain a characteristic development as cavities as is the case
with Elasmobranchs and the Lizard. Nevertheless there is
here a tendency in that direction, and sometimes a slight cavity

316

HENRY OUR.

appears as iu fig. 26, H.C. Where this happens it is generally
in the most anterior somites.

Previous to the breaking open of the gill-clefts there appears
on each side of the mandibular arch a small thickeniug and
protrusion of the epiblast. These protuberances appear long
before any of the external gills of the other arches. They
become later rod-like structures, and are then easily recog-
nised as the organs which Clarke has called “ balancers.” An
examination of their structure and relations shows them to be
homologous with the external gills. They are supplied with
blood by the most anterior or mandibular fork of the ventral
aorta, and a branch of the fifth nerve may be traced down to
the neighbourhood of the base of each balancer. Fig. 25 shows
a vertical section passing longitudinal to the axis of the embryo
and nearly longitudinal to the balancer ( bl .). The balancer
consists of a cylinder of compact epiblastic tissue, growing
slightly thicker toward the distal end, where it forms a thick
epiblastic cap similar to the epiblastic cap generally observed
on the limb-rudiments of vertebrate embryos. Internally the
balancer is nearly hollow, but is generally divided longitudinally
through the greater part of its proximal length by a thin mem-
branous network of which the function is probably to separate
the courses of the arterial and venous blood. The large
amount of blood which passes through the balancer indicates
that it subserves in part a respiratory function. Balfour,
following the account of Goette, has stated that the mandi-
bular artery is never developed in Amphibians. In Amblystoma
I find the mandibular artery developed, though in a less
degree than the posterior branchial arteries. Owing to the
early disappearance of the balancers this artery probably
atrophies at an early date. Though my sections of Triton were
not so favorable to the observation of this point, yet I find
traces there of the existence of a mandibular artery.

At a stage when the branchial clefts have broken through
and the cartilaginous skeleton has appeared (fig. 30 h), the
relation of the balancer to the mandible becomes even more
pronounced. The quadrate cartilage sends out laterally a

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 317

crescent-shaped process immediately above the articulation
with the Meckelian cartilage. This process extends to the
base of the balancer separating the two blood-vessels which
pass to and from the balancer. This process appears crescent-
shaped only in vertical longitudinal section,, and the posterior
blood-vessel lies partly enclosed in the crescent. The process
is, shown at p. in the horizontal section, fig. 30h. Here also
may be seen a bundle of muscle-cells extending from the
pterygoid muscle (m.) into the base of the balancer ( bl .).
Another band of apparently undifferentiated muscular elements
passes from the end of the above-mentioned process down into
the balancer. Section 30h is cut through the base of the
balancer, the free end of which extends below the plane of the
section.

The balancers of Triton are of the same character as those
of Amblystoma, but in Triton they appear to be not quite so
highly developed.

Clarke observed the use of the balancers in the living
embryos, and came to the conclusion that the chief function
of the organs was as a means of support for the embryos to
prevent them from sinking into the slime on the bottom of the
pools in which they live. My own observations on the living
embryos have led me to the same conclusion. It seems there-
fore that we have in this case not only the peculiarity of a
homologue of the external gills arising from the mandibular
arch, but also a homologue of the external gills becoming
metamorphosed into an organ for the support of the body. It
is also noteworthy that the balancers drop off after the limbs
have appeared.

If we seek among the Anura for organs homologous with
these balancers of the Urodela, the only organs which we can
fix upon with any degree of probability are the suckers of the
tadpole. Balfour has stated that these suckers arise on the
hyoid arch, but in the embryos of Anura which I have examined
they appear immediately posterior to the mouth-fusion (fig. 20)
long before any trace of a division into visceral arches has
appeared. I think for this reason that the suckers cannot

318

HENRY ORB.

properly be assigned to the hyoid arch. Balfour, iu describing
the embryology of the Urodela(‘ Comp. Embry.’), says,“ Stalked
suckers of the same nature as the suckers of Anura are formed
on the ventral surface behind the mouth.” The balancers in
the two forms of Urodela which I have examined possess none
of the characteristics of suctorial organs. Yet the balaucers of
the Urodela and the suckers of the Anura serve ultimately the
same purpose, namely, to prevent the embryos from sinking into
the soft organic mud usually found iu the bottom of the pools
which they inhabit, immersion iu which would undoubtedly
prove fatal to large numbers. An examination of a larger
number of species may bring to light iutermediate forms of
these organs which would prove a more direct homology
between the balancers and suckers.

The condition of the branchial apparatus and the skeleton
of the head at a time when the branchial clefts have opened
and shortly after the cartilage has appeared, is shown in figs.
27h — 31, and also fig. 18. Fig. 31 is an approximate recon-
struction from drawings of sections in three planes at right
angles to each other; it represents the skeleton of the left side
of the head. The four branchial clefts (II — V) are situated
between the hyoid arch (I') and the posterior branchial arch
( V', fig. 30 h). Each of the posterior four cartilaginous arches
(II — V, fig. 28 h) supports an external gill. The carti-
laginous hyoid arch has no external gill, but supports an
opercular fold (o./., fig. 30 h) which extends transversely across
the ventral side of the head (fig. 18, o.f.) and a short distance
up the lateral sides, partly overlapping the external gills. The
cartilaginous bars of the visceral skeleton are of unequal
length. Only the hyoid and the first two branchial bars
extend to the median line, where they unite in a basi-hyo-
branchial plate of cartilage ( B . Hy., figs. 31 and 18). From
this basi-hyobranchial plate there extends in a ventral and
posterior direction a long curved process of which the flattened
end touches the pericardium. The posterior two branchial
bars (IV , V') each unite with the next preceding bar as
shown in fig. 31. The hyoid bar does not extend dorsalwards

NOTE ON THE DEVELOPMENT OP AMPHIBIANS. 319

more than half as far as the first branchial bar, so that in the
sections of series h it first appears in the section 30 h. Each
of these five bars is supported dorsally by a small muscle ; the
muscles are shown at in' . in! . in fig. 27 h. Dorsally the four
branchial bars are united by a continuous piece of cartilage,
c. b., fig. 27 h. None of these bars are articulated into different
pieces, but the hyoid and first two branchial bars show ventrally
a rudimentary beginning of an articulation.

The cranial skeleton of Amblystoma at this stage shows
certain peculiarities, the homologies of which I am unable to
determine in other forms. A general idea of the shape of the
skeleton may be derived from fig. 31 and series H. Each
lateral half of the cranial skeleton, together with the corre-
sponding quadrate, appears as one continuous piece of carti-
lage. Of this piece the parts corresponding to the trabecular
( tbr .) and parachordal ( prc .) cartilages are easily distinguished;
the former lying along the anterior surface of the brain, and
the latter lying along the floor of the hind- brain adjacent to
the notochord. The trabeculae do not meet anteriorly. From
the anterior end of the parachordal region there extends in a
dorso-lateral direction a small bar of cartilage (x, figs. 31 and
28 h). This is met by another bar of cartilage (y), which
extends upward and backward from the trabecula at the region
of the optic nerve. These two bars (x and y) form thus a
triangle, of which the base is the posterior part of the trabe-
cula. At the junction of the anterior bar ( y ) with the trabe-
cula there is a foramen through which passes the optic nerve
(n. II, fig. 31). The greater part of the bar marked y.
appears to pass dorsal to the optic nerve (fig. 31). This bar is
a relatively thin piece, and separates the eyeball from the
thalamencephalon. The posterior bar ( x ) of the triangle lies
in the lateral groove between the floor of the hind-brain and
the infundibulum, that is, lateral to the fold caused by the
primary cranial flexure. At the dorsal apex of this cartila-
ginous triangle the cartilage is continuous with the dorsal
proximal part of the quadrate cartilages (Q.). This is seen at
Q. x, y, in figs. 27 h and 28 h. The dorsal part of the quad-

320

HENRY ORR.

rate cartilage is rather thin, and lies transversely with its
lateral edge curved postei’iorly toward the otic cartilage ;
farther ventralwards its section is shown in figs. 29 H and 30 h.
The Meckelian cartilage shows no unusual peculiarities. An-
teriorly it is connected with the cartilage of the opposite side
by a short band of undifferentiated connective tissue.

What the significance of this manner of development of the
choudrocranium may be, or how much importance should be
attached to it, I am unable to say, as I have observed it only
at this one stage.

At this stage well-characterised rudiments of teeth have
appeared ( d ., figs. 30 h and 18). They are present in a semi-
circle above the Meckelian cartilages. They do not appear in
a single row, but in several irregular rows. In the same
manner they appear just ventral to the trabeculae cranii along
those parts of the trabeculaj which lie anterior to the optic
nerve. In a cross section of the several irregular rows of
teeth the teeth seem to radiate from the bar of cartilage on
which they rest. In this respect the trabeculae cranii and the
Meckelian cartilages present the same appearance.

General Conclusions.

The central nervous system of Amphibians first appears as
a transverse epiblastic thickening dorsal to the mouth-fusion,
and continuous with paired elongated epiblastic thickenings
lying dorsally on each side of the median line.

The primary cranial flexure is due to the presence of the
transverse epiblastic thickening (anterior medullary plate).

The transverse epiblastic thickening forms, when the brain
is enclosed, that part of the brain wall which lies between the
infundibulum and the optic groove (i. e. the depression just
dorsal to the chiasma of the optic nerves).

The first nerve-fibres which develope in the brain appear on
what was originally the internal surface of the primitive epi-
blastic thickenings running longitudinally in the dorsal region

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 321

and uniting continuously in the region of the primitive trans-
verse thickening.

A subsequent development of nerve-fibres gives rise to a con-
tinuous ventral commissure extending through the floor of the
mid-brain and hind-brain and spinal cord ; and to the anterior
and posterior commissures of the brain.

The fibres of the optic nerves are intimately connected with
and are developed in the same manner as the main bundle
of fibres in the region of the primitive transverse epiblastic
thickening.

The hypophysis of Amblystoma presents a form of develop-
ment intermediate to that of the Lizard and that of the Frog.

The balancers of Amblystoma may be considered as external
gills of the mandibular arch which have become metamor-
phosed into embryonic organs of support.

P.S. — In his work entitled ‘ Untersuchungen fiber die ver-
gleichende Anatomie des Gehirns/ Dr. Ludwig Edinger has
described a Commissur der basalen Vorderhirnbundel,
which he says appears in all classes of Vertebrates. The
position of this Commissur in the adult brain immediately
behind the optic chiasma is identical with that of the anterior
band of nerve-fibres (A. F.), which I have described in the
embryonic condition. The relatively large size and pronounced
character of the anterior band in both Reptilian and Amphi-
bian embryos lead me to think that it was once of primary
importance, and that the Commissur in the adult brain is
probably a rudiment of the same with changed relations and
functions.

322

HENRY ORR.

EXPLANATION OF PLATES XXVII, XXVIII, & XXIX,

Illustrating Mr. Henry Orr’s paper “Note on the Develop-
ment of Amphibians, chiefly concerning the Central
Nervous System ; with Additional Observations on the
Hypophysis, Mouth, and the Appendages and Skeleton of
the Head.”

Where a number of figures represent sections of the same individual embryo,
all those figures have the same letter affixed to their numbers.

All figures of sections have been drawn with the Abbey camera lucida and a
Zeiss’s microscope, so that in figures magnified to the same degree the size of
the parts may be directly compared. (Z. 2, A, means Zeiss’s ocular 2, and
objective A, &c.)

Index Letters.

A. C. Anterior commissure of fore-brain. A. F. Anterior medullary fold.
A. F'. Anterior band of nerve-fibres, continuous with the lateral bands, L. F.
a. M. P. Anterior medullary plate. B. Ey. Basi-hyobranchial plate of
cartilage. Bl. Region of the blastopore, hi. Balancers, c. b. Cartilaginous
bar connecting dorsally the cartilaginous gill-arches. Ch. Optic chiasma
united with the anterior band of nerve-fibres (A. F'.). d. Dental rudiments.
E. Ear. Ep. Epiblast. Eph. Epiphysis cerebri. Ey. Eye. F. B. Fore-
brain. G — g. See explanation of Eigs. 2 a — 5 d. H. B. Hind-brain. H. C.
Head cavity. Hph. Hypophysis. Et. Heart. Eyp. Hypoblast. In. In-
fundibulum. L. Lens of eye. L. F. Primary longitudinal fibres of central
nervous system ; L. F'. the same in the region of the thalamencephalon.
M. Mouth-fusion or mouth. m. and ml. Pterygoid and branchial muscles.

M. B. Mid-brain. Md. Medulla spinalis. Mes. Mesoblast. M. F. Lateral
medullary fold. mk. Meckelian cartilage. M. P. Lateral medullary plate.

N. Notochord or rudiment of the same. Na. Nasal sac. n. I , n. II — n. X.

Olfactory, optic, and succeeding cranial nerves, o.f. Opercular fold of the
hyoid arch. o.g. Optic groove, o.p". Posterior wall of the optic stalk.
p. Lateral cartilaginous process of the quadrate at the base of the balancer.
P. C. Posterior commissure of the brain, p. g. Rudiment of the pectoral
girdle. PE. Pharyngeal cavity. So. Somatopleure of mesoblast. Sp.
Splanchnoplcure of mesoblast. St. Corpora striata. Hr. Trabecula: cranii.
T. F. Transverse nerve-fibres forming a continuous ventral commissure. Th.
Rudiment of the thyroid gland, w. E. B. Floor of hind-brain, x — y. See

explanation in text, p. 319. Y. Yolk. I — V. Hyoid and branchial clefts.
V — V. Cartilaginous gill-arches.

NOTE ON THE DEVELOPMENT OF AMPHIBIANS. 323

PLATE XXVII.

Fig. 1. — Median-longitudinal and nearly vertical section of the egg of
Amblystoma, at a time when the medullary plates have first appeared. It
shows the anterior epiblastic thickening {a. M. P.), which unites anteriorly
the two dorsal medullary plates. Also the thinner median portion {g. g.)
between the dorsal medullary plates. Bl. Region of the blastopore. N. Un-
differentiated hypoblastic tissue of the notochord. (Z. 2, A A.)

Figs. 2 a, 3 b, 4 c, and 5 d. — Transverse sections through the anterior
dorsal region of embryos of Amblystoma, showing successive stages of develop-
ment. The first of the series (2 a) is at a stage corresponding with that of
Fig. 1. G. The thinner median portion of epiblast between the dorsal medul-
lary plates which becomes pushed downwards, so that the surfaces immediately
lateral to it become pressed together along the Hue g. (Z. 2, A A.)

Figs. 6 a, 7 b, 8 c, 9 d. — Transverse sections through the anterior region of
the head of the same embryos respectively as Figs. 2 a — 5 d. These sections
show the anterior medullary plate {a. M. P. or A. F.) which connects the
lateral dorsal medullary plates. A. L. Anterior end of the alimentary cavity.
(Z. 2, A A )

Fig. 10 d. — Transverse section through the posterior region of the head to
show the reduction of the lumen of the neural canal. (Z. 2, A A.)

Fig. 11. — Transverse section through the cervical region of an embryo of
Amblystoma, somewhat more advanced than that of series d, showing the
change of shape in the neural tube and canal. (Z. 2, A A.)

PLATE XXVIII.

Figs. 12 e and 13 e. — Longitudinal and nearly vertical sections of an em-
bryo of Amblystoma (at a stage represented by Clarke’s Fig. 14). Fig. 12 e
passes through the vertical plane at the hypophysis (Hp/i.) and the dorsal
notochord ; Fig. 13 e at the anterior end of the notochord and the epiphysis
(epL). (Z. 2, A A.)

Figs. 14 f, 15 f, and 16 f. — Horizontal sections of an embryo of Ambly-
stoma at the same stage as the preceding two figures. Fig. 14 f passes

through the hind-brain and part of the dorsal medulla. Fig. 15 f passes
through the mid-brain and the dorsal part of the alimentary cavity. Fig. 16 f
passes through the hypophysis and the anterior part of the fore-brain.
(Z. 2, A A.)

Figs. 17 g and 18. — Longitudinal median vertical sections of two embryos
of Amblystoma. 17 g is older than the stage of series e and f, and 18 is older
than 17 g. These sections, together with 12 e, show the development of the
lower jaw, the formation of the mouth, and the hypophysis. (Z. 2, A A.)

Figs. 19 — 23 ine. — Longitudinal median vertical sections of successive
stages of Frog-embryos, showing the formation of the fore-brain, the hypo-

324

HENEY OER.

physis, and the mouth-fusion, with their relative changes of position.
(Z. 2, A A.)

Fig. 24 g. — Part of a longitudinal median vertical section of an embryo of
Amblystoma, showing the floor of the hind-brain and the fold between the
hind-brain and infundibulum, which is caused by the primary cranial flexure.
(Z. 2, A.)

Fig. 25. — Taken from a series of longitudinal vertical sections of an embryo
of Amblystoma ; it shows the balancer in nearly longitudinal section, and also
a superficial portion of the mandibular arch. (Z. 2, A.)

PLATE XXIX.

Fig. 26. — Horizontal section of an embryo of Amblystoma at a stage
between the stages e — F and the stage G. This figure shows the first formation
of the optic lens, also the formation of the gill-clefts and the somites of the
head. (Z. 2, A A.)

Figs. 27 n — 30 h inch — Horizontal sections of an embryo of Amblystoma
at the same stage as the embryo of Fig. 18. Of these sections, 27 H is the
most dorsal and 30 h the most ventral, the others being intermediate in the
order of their numbers. The sections show the development of the nerve-
fibres in the brain, and the early development of the cartilaginous skeleton
and the branchial apparatus. (Z. 2, A A.)

Fig. 31. — Approximate reconstruction of an early stage of the cranial and
visceral skeleton of Amblystoma, made from drawings of series of sections
cut in three planes at right angles to each other. It shows the skeleton of
the head viewed from the left side, also the shape and relative position of the
left rudiment of the pectoral girdle (P. g.).

Fig. 32 g. — Lateral longitudinal vertical section of the brain of an embryo
of Amblystoma, cut through the left side, showing the course of the nerve-
fibres at the time of their first development in the brain. (Z. 2, A.)

Fig. 33. — Part of a transverse section through the fore-brain of a Frog-
embryo, showing the fibres of the lateral and anterior band ( L . F. and A. F.),
the latter crossing the anterior surface of the brain ; also the fibres developing
on the anterior wall of the optic stalk ( n . II). The posterior wall of the
optic stalk (o. y;".) is free from fibres. (Z. 4, A.)

Fig. 34. — Transverse section of the spinal cord in the dorsal region of an
embryo of Amblystoma at the stage of series n. It shows in cross-section
the longitudinal nerve-fibres of the lateral band (L. P.), also the transverse
fibres of the ventral commissure ( T . F.). (Z. 4, A.)

Fig. 35. — Transverse section passing through the mid brain and secondary
fore-brain of an embryo of Amblystoma at the same stage as the embryo of
series g. It shows the corpora striata (si.) and the transverse fold which
separates the secondary fore-brain from the thalamencephalon. P. C. shows
the position of the posterior commissure. (Z. 2, A A.)

Fig. /

G

Fig. 2. A.

G

MP

Mivr.tfowrQ/ol XXIXJS&. XXVII.

MF

MP MP 9

1 'f

Al

Fig.Jl.

MdL

Ott it!

r Huth, Lithr Edmr

jeumcfwrnftfa XX/X,Fsd%.XXV///.

Fig. J 4. F.

,\C'v\V;

Fig.IZ.F.

nVIlUVa

»<mnti<y

Fig. jy. g

%mrnm

-TF

Otr d»l.

F. H-uOi,Lit'nT Edmx

'fojyr /» O/oC. ////,// SM XXIX

*1

Fig.2J.H.

:v'^

Fig. 30. F

'■ v\ ’'rr >’’

Tl\W/J

Fig. 28. F.

,i ’ui "*•'

nVOI.V//

Orr del

F Huth,Lithr Edinr

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 325

Studies on the Comparative Anatomy of
Sponges.

II. On the Anatomy and Histology of Stelo-
spongus flabelliformis, Carter; with
Notes on the Development.

By

Arthur Dendy, M.Sc., F.L.8.,

Demonstrator and Assistant Lecturer in Biology in the University of
Melbourne.

With Plates XXX, XXXI, XXXII, & XXXIII.

The species upon the study of which the present paper is
based, was first described by Mr. H. J. Carter, F.R.S., in 1885
(6), under the name Stelospongus flabelliformis. The
first specimens were dredged by Mr. J. Bracebridge Wilson,
M.A., and forwarded by him to England as part of a large
collection entrusted to Mr. Carter for description. Mr. Carter's
description is unfortunately brief and unaccompanied by figures,
and he does not enter into any details concerning the anatomy
of the Sponge. He notes, however, in a subsequent paper (7),
that it is characterised by the presence of very large embryos,
and this remark first led me to the identification of my
specimens.

Daring the last Easter vacation it was my good fortune to
spend some days in dredging with Mr. Bracebridge Wilson in
the neighbourhood of Port Philip Heads, and amongst
numerous other interesting Sponges we happened to obtain a
horny Sponge which, when torn open, was found to contain a
large number of enormous spherical embryos, each as large as

VOL. XXIX, PART 3. NEW SER.

Y

326

ARTHUR DENDY.

a small pea, embedded in the choanosome. I picked out a
number of these embryos from the living Sponge and placed
them directly in strong spirit, and also preserved portions of
the mother Sponge in the same way. This simple method of
preservation, adopted at the time for want of other reagents,
subsequently proved to have been highly successful.

After cutting sections of the embryos and finding that they
presented certain very remarkable features, I determined to
make a complete study of the anatomy of the adult Sponge.
Fortunately, I found amongst a splendid collection of well-
preserved Sponges already sent up to the Melbourne University
by Mr. Wilson, several more specimens of Stelospongus
flabellifor mis, and I thus had abundance of material at
my disposal.

My preparations were stained with borax carmine and cut in
paraffin in the usual manner, and I found that in order to
ensure success it was very important not to allow the tem-
perature of the paraffin to rise above 128° F. Probably a
lower temperature would have been better still, but unfortu-
nately I had only hard paraffin to work with.

With regard to the figures, I may state that while some of
them are more or less diagrammatic, all are taken from actual
preparations, either individual or combined, with the exception
of figs. 9 and 10, which are pure diagrams.

I have very great pleasure in taking this opportunity of ex-
pressing my sincere thanks to Mr. Bracebridge Wilson, to
whom I am indebted for the whole of my material, and to
Professor W. B. Spencer, who, throughout the progress of the
work, has given me much valuable advice.

Anatomy and Histology.

External Characters.

An idea of the general appearance and size of the Sponge may
be gained from fig. 1, which represents a fairly typical specimen,
from which a portion has been cut off, drawn of the natural

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 327

size. The Sponge consists of a stout cylindrical stalk ter-
minating below in a basal expansion for attachment to the
substratum, and above in a broad, compressed, but thick frond.
Along the upper margin of the frond is a single row of large
oscula.

The character of the surface of the frond is subject to con-
siderable variations. On both sides it usually exhibits a
number of prominent, branching, and anastomosing ridges,
enclosing a number of concave depressions, as represented in
the figure. The entire surface of the Sponge is more or less
thickly encrusted with sand particles ; these are, however,
more abundant over the ridges than in the depressions. The
latter are also characterised by a minutely reticulate appearance
due to the presence in them of the inhalant pore-areas, each of
which bears a number of pores. The pores are enormously
abundant all over the depressed portions of the surface, but
appear to be absent from the ridges.

The incrustation of sand already referred to gives to the
surface of the Sponge a very hard, impenetrable character, and
must form an admirable protection against the attacks of the
numerous parasites to which Sponges are very subject. It
functionally replaces the special dermal skeleton of spicules
which exists in very many siliceous Sponges.

The Skeleton.

The skeleton (fig. 3) is composed of a rather irregular reti-
culation of cylindrical horny fibres, branching and anasto-
mosing freely. The fibres are rather slender, and the meshes
between them are wide, so that in thin sections the skeleton
scarcely appears at all. As in most horny and siliceous
sponges, it is easy to distinguish between two sets of fibres,
primary and secondary. The primary fibres (fig. 3 are
long and rather stouter than the secondaries, measuring about
O096 mm. in diameter. They radiate towards the surface of
the Sponge, sometimes branching in their course, and end in
the sandy incrustation. But they are most easily distin-
guished by the presence in them of numerous foreigD bodies,

328

ARTHUR DENDY.

grains of sand, broken sponge-spicules, &c., which form an
axial core, surrounded and held together by concentric layers
of spongin. The secondary fibres (fig. 3, s.f .) are short, and
contain no foreign bodies. They run in various planes, and
unite together adjacent primary fibres which they meet at
various angles; they may also branch and anastomose inter
se. They measure from about 0’048 to 0-08 mm. in diameter.

Thus the skeleton is thoroughly typical in structure and
arrangement, and essentially the same as that of the ordinary
bath-sponge, only much coarser. The distinction between the
primary and secondary fibres is well marked. There is good
reason to believe that primary fibres are phylogenetically of
earlier origin than secondaries, and really very distinct from
them. The primitive skeleton of the horny Sponges and of
the siliceous (excluding the Hexactinellida) was probably
composed simply of large, radiating spicules, or bands of spicules,
coming out from a common centre, the centre of the Sponge.
Such a condition occurs now very frequently amongst the
Clavulina, Tethyadse, and Tetractinellida, all of which
groups we have reason to believe are more primitive than the
Halichondrina and Ceratosa, which have reticulate skele-
tons. Amongst the Clavulina, Tethyadae, and Tetracti-
nellida there is usually little or no spongin present, but the
spicules are very large, and arranged side by side in dense tufts
radiating from a common centre. In the Halichondrina
spongin is almost invariably present in considerable quantities,
and the spicules gradually diminish in size as the amount of
spongin increases, so that the fibres of the skeleton, instead of
being stiff and rigid, become flexible and elastic. All stages
in the gradual replacement of spicules by spongin may be
traced in the large family of the Chalininse, as I have
endeavoured to show in a former paper (8). It is obvious
that while a skeleton, composed solely of radiating primary
fibres, would be very efficient so long as the fibres remained
stiff and rigid, yet when the fibres became soft and flexible
owing to the replacement of the spicules by spongin, such a
skeleton would be almost useless. Hence arose the necessity

STUDIES ON THE COMPARATIVE ANATOMY OP SPONGES. 329

for the formation of secondary fibres to connect the primaries
together. A reticulate skeleton must therefore be regarded
as derived from a radiate one by the development of secondary
fibres connecting the primaries. There can no longer be any
doubt that the majority, at any rate of the so-called “ horny ”
Sponges, are descended, probably along several lines, from the
Halichondrina, by the gradual loss of spicules and the
greater development of spongin in a reticulate skeleton.

The occurrence of grains of sand in the skeleton fibres is
not confined to the Ceratosa, for sometimes sand and
spicules are simultaneously present in the fibres, as in
Siphonochalina spiculosa (8). This replacement of
spicules or of spongin by sand is no doubt of great advantage
to the Sponge in saving material, and hence we not unfre-
quently get Sponges whose skeleton is entirely arenaceous
(e. g. Dy sidea).

In Stelospongus f labellifor mis the skeleton fibres
may sometimes be seen projecting freely from the surface of
the Sponge (fig. 5, /'.), but this does not appear to be at all a
constant character. It is difficult to understand how such a
condition can have arisen ; perhaps it is in some degree com-
parable to the projection of the spicules from the surface in
very many siliceous Sponges.

In the stalk the skeleton is more strongly developed than
elsewhere, and hence it acquires a tougher and denser charac-
ter than the remainder of the Sponge.

The Canai. System.

(a) The Pores.

The inhalant apertures, or pores, are thickly scattered all
over the depressed areas on the surface of the Sponge. Hence
these areas might with some justice be termed pore-areas,
in the sense defined by Ridley and Heudy in the Report
on the “ Challenger ” Monaxonida (14). But there is an
objection to the use of the term in this particular case in
that these larger areas are themselves subdivided into a great

330

ARTHUR DENDY.

number of smaller areas (fig. 4, p. a.), and as the latter are
much better defined and more constant in size and relations
than the former, it is better to apply the term pore-areas to
them. These smaller areas appear to be strictly comparable
to the pore-sieves of Phakellia ventilabrum, var. con-
nexiva, or the pore-areas of Myxilla nobilis (14). Each
one is an irregularly rounded or oval area, about 0T9 mm. in
diameter, overlying a subdermal cavity, and each contains
some five or six oval or rounded pores (fig. 4, p.) averaging
about 0-05 mm. in their longer diameter.

The most satisfactory way of studying the arrangement and
form of the pores in this and many other Sponges is to slice off
as thin a portion as possible of the surface, and stain and
mount in balsam in the usual way, without cutting sections.
Fig. 4 represents such a preparation seen from above as a
transparent object. The pores may also be seen in sections
taken at right angles to the surface (fig. 2, p.), but in the
present case it is rather difficult to obtain satisfactory sections
of this kind owing to the presence of the sand grains in the
ectosome.

(b) The Subdermal Cavities.

In this, as in my previous paper (9), I use the term sub-
dermal cavities in the sense defined in the Report on the
“Challenger” Monaxonida, i.e. to mean the spaces into which
the pores directly lead. Sollas, in his article on Sponges in the
‘ Encyclopedia Britannica5 (17), appeal’s to make use of the
term in a different sense, as synonymous with subcortical
crypts, while he applies the term chones to the structures
which I term subdermal cavities. The homologies of
these various structures are not at present sufficiently under-
stood to enable us to give them a really satisfactory nomen-
clature, and so I prefer to use a purely empirical one. The
term subcortical crypt is used by Sollas and myself in the
same sense, i. e. to mean the space underlying the cortex or
ectosome into which the subdermal cavities (chones of Sollas)
lead.

STUDIES ON THE COMPARATIVE ANATOMY OE SPONGES. 331

In Stelospongus flabelliformis each subdermal cavity
(fig. 2, s. c.) is a hollow space corresponding in size and form
to the pore-area which it underlies, and communicating with
the exterior by means of the pores in its roof. The different
subdermal cavities are separated from one another by anas-
tomosing vertical walls of tissue constituting the bulk of the
ectosome — represented as seen from above in fig. 4 ( cy .) and
in section in fig. 2 (cy.) ; each one communicates below with a
very much larger inhalant channel (fig. 2, i.l.). Thus each
subdermal cavity receives the stream of water directly from
the exterior through five or six distinct apertures in its roof,
and passes it on through a single aperture in its floor into a
relatively large inhalant channel. Just as a number of pores
lead into one and the same subdermal cavity, so also a number
of subdermal cavities lead into one and the same inhalant
channel.

(c) The Inhalant Canal System below the Sub-
dermal Cavities.

The large inhalant channels (fig. 2, i.l.) into which the
subdermal cavities directly lead are comparable to the sub-
cortical crypts described by Sollas (17) in the Tetractinellida,
but it is needless to apply a special name to them in Stelo-
spongus. They are merely the larger proximal portions of
the inhalant canal system, commencing immediately beneath
the ectosome and penetrating deep down into the choanosome.

These larger channels lead into an irregular system of much
smaller, more or less lacunar channels, whose ultimate rami-
fications open into the flagellated chambers (fig. 6) ; and
numerous flagellated chambers open out of one and the same
inhalant lacuna.

It must not be supposed that the inhalant canal system is
always constant in arrangement ; the above description applies
to what appears to be a fairly typical case, but there seems to
be a good deal of variation, especially with regard to the
subdermal cavities and the channels into which they lead.

332

ARTHUR DENDY.

(d) The Flagellated Chambers.

These are more or less spherical sacs (figs. 6 , 7, 10) about
004 mm. iu diameter, with two wide apertures of about equal
size placed at opposite poles, whereby they communicate on
the one hand with an ultimate inhalant lacuna and on the
other with an ultimate exhalant lacuna. In the same way
that several chambers open out of one and the same ultimate
inhalant lacuna, so also several may lead into one and the
same ultimate exhalant lacuna. Both the inhalant and
exhalant apertures of the chambers are usually drawn out into
short and relatively wide cameral canaliculi (fig. 10, i. c. c.,
e. c. c.), but I do not think too much importance must be
attached to this fact.

(e) The Exhalant Canal System.

The ultimate exhalant lacunae1 (fig. 6, e. L), into which the
flagellated chambers open usually through the medium of
distinct canaliculi, collect together and finally discharge their
contents into branches of the oscular tubes. Each oscular tube
is a perfectly definite tubular canal about 6 mm. wide, with
distinct walls of its own, and leads vertically upwards to a wide
osculum situated on the upper margin of the Sponge. The
oscular tube itself may readily be dissected out from the
surrounding choanosome, from which its walls are very easily
separable. Fig. 5 represents a dissection of the oscular tube
( o . t.) showing its relations to the osculum (o.) and the openings
into it of a number of larger and smaller branches.

The osculaj are wide, circular openings, about 6 mm. in
diameter, placed iu a row along the upper margin of the Sponge ;
their position is indicated in fig. 1 by the letters o. o. o., and in
the same figure a portion of an oscular tube (o. t.) is seen on
the cut surface.

In connection with the exhalant canal system I may here
mention certain spherical cavities (fig. 5, e. c .) lying in the

1 It is impossible to distinguish sharply between a canal and a lacuna;
either term might be applied in this case.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 333

neighbourhood of the oscular tube and containing each
a single large embryo (e.). These cavities appear to be
entirely closed and cut off from the remainder of the canal
system of the Sponge, but it seems possible that they are really
portions of the exhalant canal system, specially modified to
serve as receptacles in which the embryos are lodged during a
large portion of their development. There are two arguments
in favour of this view : (1) The position of the cavities in
question, in close proximity to the oscular tubes. (2) The
well-known fact that in many Sponges the embryos normally
escape from the parent through the exhalant cauals1.

(f) General Remarks on the Canal System.

From the foregoing account it will be evident that the canal
system of Stelospongus flabellifor mis approaches most
nearly to Dr. Vosmaer’s third type (18) ; although, however,
several chambers communicate with one and the same ultimate
inhalant or exhalant lacuna, the openings of the chambers are
provided with short and relatively wide cameral canaliculi.
Still the canal system differs little from the ordinary lacunar
type so characteristic of the large family Halichondrina
(14), and also found occasionally in the Clavulina (14) and
other groups.

With regard to other members of the Ceratosa, some have
been shown to possess the lacunar type of canal system and
some the canalicular type. Professor Schulze (15) has de-
scribed both these conditions as they exist in different members
of the group. Unfortunately, I am unable at present to obtain
access to his original papers, but I may quote the following
passage from Polejaeff's work (13) which will sufficiently indi-
cate the state of the case: — “ Schulze ascertained, in fact,
that while an Aplysina, and on the other hand a Eu-
spongia or Cacospongia, are characterised, in the or-
ganisation of their canal system, by comparatively small,
round, or pear-shaped flagellated chambers, each possessing

1 liidley and Dendy have figured an embryo escaping through an exhalant
canal in Esperella Murrayi (14).

334

ARTHUR DENDY.

its own narrow inhalant and exhalant canaliculi, and while
the ground-mass surrounding these flagellated chambers is
always opaque owing to the presence of small granules, the
forms like Spongelia and Aplysilla possess no special
cameral canaliculi, their large, pouch-shaped flagellated
chambers receiving the water from the subdermal cavities
directly by means of the pores in their walls, and expelling it
also immediately, without the help of any intermediate narrow
canals, into large exhalant cavities, the diameter of these latter
being usually far larger than that of the exhalant opening of
the corresponding flagellated chamber; and that in these latter
instances the parenchyma in the zone of the flagellated
chambers is devoid of any granules, being lucid and trans-
parent.”

Polejaeff further informs us that “the flagellated chambers
of the representatives of the genera Aplysina and Yerongia
are small, pear-shaped, or rather hemispherical, each provided
with one (?) inhalant and one exhalant narrow canaliculus ;
and again, the surrounding ground-mass is so very rich in
granules that the outlines of the cellular elements in the
neighbourhood of the flagellated chambers are scarcely distin-
guishable. On the other hand, the flagellated chambers of an
Aplysilla or Ianthella are large and either of regularly
elongated form (pouch-shaped) or of quite irregular outline;
no special cameral canaliculi are to be discerned ; the flagel
lated chambers receive the water from the subdermal cavities
by means of numerous pores in their walls, and expel it by
means of a large exhalant aperture; the surrounding groxind-
mass is clear and transparent. There are, however, amongst
the horny Sponges forms uniting these two extreme differences
in every direction.”

It would appear from my observations that Stelospongus
flabelliformis is one of these connecting forms, for although
the inhalant and exhalant canaliculi are only very feebly
developed, the ground substance, as we shall see later on, is
densely charged with minute granules.

Von Lendeufeld (11) has published a good illustration of the

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 335

lacunar type of canal system, as it occurs in his Euspongia
canaliculata, and this species appears to have the clear
transparent ground substance usually found in association
with eurypylous chambers.

I have endeavoured to show in a previous paper (9) that we
cannot draw any hard and fast line between flagellated
chambers with and flagellated chambers without special incur-
rent and excurrent canaliculi, and I believe with Polejaeff that
these two types graduate into one another. I have also stated
above that the chambers of Stelospougus usually have
short, relatively wide cameral canaliculi, but that we must not
lay too much stress upon this fact. Now, according to Pole-
jaeff, as we have just seen, Schulze describes the flagellated
chambers of Euspongia as being provided with special
cameral cananiculi and embedded in a granular ground sub-
stance, but to judge from Schulze’s figure, as copied by
Vosmaer (18), the exhalant canaliculi at any rate are only very
slightly develoved, and the arrangement of the chambers in
Euspongia agrees very closely indeed with that found in
Stelospongus.

Histology of the Soft Tissues.

(a) The Ectosome.

The ectosome (fig. 2, ect .) forms a relatively thin external
layer all over the body of the Sponge. Owing to the presence
in it of a large amount of sand, especially abundant in the
raised ridges, it is very hard and tough, and forms an excel-
lent protection against the attacks of parasitic crustaceans,
worms, & c., to which Sponges are subject.

The outermost portion of the ectosome is formed by an
extremely thin and delicate epidermis, which I have succeeded
in making out chiefly in the pore-areas, where the sand grains
are absent. In preparations such as that represented in
fig. 4, one can easily distinguish the nuclei of the epidermic
cells ( n . e. c.) scattered in the transparent, pore-bearing mem-
brane. These nuclei are small granular bodies, round or oval

336

ARTHUR DENDY.

in shape, about 0 0048 mm. in diameter, and in preparations
stained with borax carmine they stand out very sharply. I
have endeavoured to demonstrate the outlines of the epi-
dermic cells by means of silver nitrate staining, but possibly
owing to the fact that I had only spirit material to work with
without success. Doubtless this epithelium is continued in-
wards through the pores to line the subdermal cavities, but
I have not succeeded in detecting it here.

A very large proportion of the ectosome is occupied by the
sand grains above mentioned, but surrounding these is a con-
siderable quantity of mesodermal tissue.1 This is for the most
part made up of cystenchyme (fig. 13), but stellate mesodermal
cells (fig. 12) are also present.

The term cystenchyme has been applied by Sollas (17)
to a peculiar form of tissue not uncommonly met with in the
ectosome of Sponges. This tissue consists essentially of a
number of more or less spherical cells, each provided with a
distinct cell wall, and containing a very much vacuolated
protoplasm in the interior. The nucleus appears to be sus-
pended in the centre of the cell in a central protoplasmic mass
connected with the cell wall by radiating strands of proto-
plasm. The whole structure resembles very much an ordinary
vegetable parenchyma cell. The individual cells are packed
more or less closely together, and the spaces between them
are filled with a granular or sometimes fibrous substance,
which is probably chiefly of an intercellular nature.

In Stelospongus the cystenchyme cells (fig. 13) are oval
or subglobular in shape, measuring about 0'024 mm. in
diameter, and the nucleus is small and granular. The proto-
plasmic strands connecting the nucleus with the cell wall are
best seen in unstained preparations mounted in glycerine ;
they are in such seen to form a network, branching and anasto-
mosing inter se. Fig. 14 represents a single cell from such

1 I use the term mesoderm here and elsewhere because it is in such very
general use amongst spongologists, and not because I believe the tissues
thereby designated to be homologous with the mesodermal tissues of other
animals.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 337

a preparation ; the outlines of the protoplasmic strands are
probably much more hard and distinct than in life, owing to
the action of the reagents. In balsam preparations (fig. 13),
owing doubtless to the greater transparency, the protoplasmic
strands are not nearly so distinct.

I may here mention the fact that cystenchyme occurs also in
the choanosome of Stelospongus, but to this point I shall
recur later on.

This form of tissue, or very slight modifications of it, has
been observed in Sponges of very divers groups. Sollas (16,
17) has described it in Tetractinellida (Pachymatisma
Johnstoni), Polejaeff (13) in Ceratosa (Cacospongia
vesiculifera), and Ridley and Dendy (14) in Clavulina
(Latrunculia apicalis).

The stellate mesodermal cells of the ectosome appear to be
thoroughly typical. They may be seen investing the grains of
sand in a kind of delicate network, the individual cells being
mutually connected by long slender processes (fig. 12). The
body of the cell is somewhat granular and the nucleus is oval
and of moderate size.

I have not observed any definite arrangement of the cysten-
chymatous and stellate tissues with regard to one another, nor
have I any reason to suppose that such exists.

(b) The Choanosome.

I propose to consider the histological characters of the
choanosome under the following heads : (1) The walls of the
inhalant and exhalant canals. (2) The walls of the embryo-
containing cavities. (3) The walls of the flagellated chambers.
(4) The general mass of mesoderm in which the chambers and
canals are embedded. (5) The spongoblasts and other meso-
dermal cells surrounding the skeleton fibres. This arrangement
is a purely arbitrary one, and I have adopted it merely as a
matter of convenience, in view of the necessity of some
definite plan to go upon. It is a matter of no small difficulty
to classify satisfactorily the various forms of tissues and cell-
elements which occur in any given Sponge.

338

ARTHUR DENDY.

(1) The Walls of the Inhalant and Exhalant
Canals.

It ■will be convenient to give these the first consideration on
account of their close relationship to the ectosome, from which
they cannot be sharply separated.

The larger or proximal portions of the inhalant canal system
are provided with special walls of mesodermal tissue. The
true nature of this lining membrane is difficult to determine.
In sections it is seen to consist of a very much vacuolated
gelatinous tissue, composed more or less of cystenchyme, but
in parts becoming fibrous.

No doubt even the larger portions of the inhalant canal
system are also provided in life with a very delicate epithelial
lining, forming the outermost layer of their walls, but this I
have not been able to detect, possibly owing to the manner in
which the entire gelatinous lining membrane shrivels up in
spirit.

The ultimate inhalant lacunae, which open into the flagellated
chambers as above described, have no special mesodermal walls ;
but here the nuclei of a delicate, flattened, lining epithelium
can be easily detected in thin sections (fig. 6, n. e. c.), and
occasionally a large cystenchyme cell may be seen embedded
in the choanosome immediately beneath this epithelium.

The ultimate exhalant lacunae, into which the flagellated
chambers discharge their contents, have, like the ultimate
inhalant lacunae, no special mesodermal walls, but are lined by
a delicate flattened epithelium, whose nuclei can be detected
in thin sections. The larger exhalant channels, or oscular
tubes, are, however, provided with most distinct walls, which
can be dissected away from the underlying tissues with great
ease (fig. 5). These walls are membranous and fairly tough,
and they are continued from the oscular tube itself along its
various larger branches as a distinct lining membrane.

The wall of the oscular tube is seen in transverse section to
be made up of the following layers from within outwards :

(a) A thick rather irregular layer of very much vacuolated

STUDIES ON THE COMPARATIVE ANATOMY OP SPONGES. 339

gelatinous tissue, composed of anastomosing strands of trans-
parent jelly-like substance, containing small nuclei here and
there. It is doubtless owing to the presence of this layer of
very delicate tissue that the wall of the oscular tube can be so
readily peeled off from the underlying structures.

( b ) A much thinner layer of deeply-staining fibrous tissue,
in which the fibres are closely packed and arranged circularly
around the oscular tube. Judging from its position and the
arrangement of its component fibres, it seems probable that
this layer may be muscular, and its fibres myocytes (Sollas, 17),
which by their power of contraction serve to regulate the
diameter of the oscular tube. The wall of the oscular tube is
smooth on the interior, and devoid of diaphragms or special
circular sphincter muscles, such as occur in many Sponges.
(Diaphragms are well developed in the genus Spirastrella,
and circular sphincter muscles in Quasillina; both doubtless
serve the same function. The condition of Stelospongus in
this respect is comparable to that of Ridleia (cf. 9) ).

(c) A continuous layer, only about one cell thick, of cysten-
chyme. This layer may best be studied by peeling off portions
of the wall of the oscular tube and preparing and mounting
them without embedding, for cystenchyme appears to be a very
delicate tissue, which suffers greatly in the latter process.

In preparations stained with borax carmine, and mounted in
the usual way, the cystenchyme is seen to form a continuous
layer, about one cell thick, of closely packed cells which have
become somewhat polygonal from mutual pressure. Between
the cells a deeply-staining, granular, intercellular substance is
present, and the structure of the individual cells is the same as
that which I have already described in the case of the ectosomal
cystenchyme.

This layer is at first sight deceptively like a layer of large,
flattened, epithelial cells, and I at first mistook it for such ;
but the characters of the component cells made me doubt if this
could be so, and on cutting sections its real nature became
readily apparent. The individual cells measure about 0-03 mm.
in diameter.

340

ARTHUR DENDY.

I have no doubt that the wall of the oscular tube is completed
on the inside by a delicate flattened epithelium, but I have
not succeded in demonstrating its presence.

(2) The Walls of the Embryo-containing Cavities.

I have above stated the reasons for regarding the cavities in
which the embryos are lodged as specialised parts of the exha-
lant canal system. Whether this view be adopted or not — and
it is still an open question — the structure of their walls maybe
most conveniently treated of in this place, although I shall be
obliged to refer again to this portion of my subject when
speaking of the development.

The only ovum which I have observed previous to the com-
mencement of segmentation lies in a small cavity, about Cbl
mm. in diameter, situate in the innermost part of the gela-
tinous layer of the wall of an oscular tube. This cavity has a
special wall, about 0'0144 mm. thick, composed of fibrous
tissue with elongated nuclei, similar to that represented in
fig. 15, but not so strongly developed. I have not detected a
lining epithelium, but some of the nuclei which are observable
in the outermost part of the wall may possibly belong to a
delicate epithelial layer similar to that which lines the smaller
branches of the canal system, and which must almost certainly
occur here also.

The large embryo- containing capsules now to be described
are probably developed simply by growth of the small capsules
containing the ova. The walls of these large capsules are,
however, very much more highly differentiated than those of
the ovum-containing capsules, and consist of two very distinct
layers, (a) a fibrous layer, and (b) a lining epithelium. The
fibrous layer of the wall (fig. 15) is very dense next to the
lining epithelium, but further in it becomes looser and is
broken into by large lacunar spaces. It is composed of cir-
cularly arranged fibres each consisting of a greatly elongated
fusiform granular cell, with a deeply-staining oval nucleus in
the centre. The fibres are so densely packed in the outer part

STUDIES ON THE OOMPAEATIYE ANATOMY OP SPONGES. 341

of the layer next to the lining epithelium that the outlines of
the individual cells can no longer be distinguished, but further
in the cells lie farther apart and the tissue partakes more of
the nature of a compact stellate mesoderm.

The lining epithelium of the embryo capsule (fig. 16) is very
peculiar and, so far as I am aware, entirely different from any-
thing which has hitherto been described in Sponges. It is
composed of a single layer of enormous polygonal cells. These
cells, although flattened, are thick, each one measuring from
about 0072 mm. in diameter for the youngest embryo examined
up to 0T2 mm. for older ones, and about 0’024 mm. in thick-
ness. The body of the cell is finely granular, and each
contains in its centre a very large, flattened, oval nucleus
containing a number of deeply-staining granules. In the largest
cells the nucleus may be seen to be undergoing division,
doubtless preparatory to the division of the whole cell. Thus,
in fig. 1G the nucleus of one of the cells has acquired a horse-
shoe shape, the two arms of the horse-shoe being nearly
separated from one another, and in another cell the division is
complete and the cell contains two nuclei. I have observed
no karyokinetic figures.

In transverse sections the outer surfaces of the cells are
frequently, but by no means always, seen to be indented
(fig. 20). These indentations would appear to correspond in
some way to the upper portions of the cells of the outer layer
(ectoderm) of the embryo, which in life are closely connected,
as wc shall see later on, with the epithelial layer of the embryo
capsule.

In transverse sections also the body of the cell is seen to be
granular throughout, but the granules are very much finer
around the nucleus than towards the periphery of the cells
(figs. 17,21). The cell always has a definite bounding wall
on its outer, and sometimes also on its inner, surface ; but
frequently its inner surface, which in life is pressed against
the fibrous layer of the embrvo capsule, exhibits no such wall
(fig. 21).

In sections the nucleus sometimes appears solid (fig. 21),

VOL. XXIX, PART 3. NEW SER. Z

342

ARTHUR DENDY.

and sometimes as a hollow vesicle provided with a distinct wall
and enclosing a granular substance (fig. 17).

These epithelial cells very readily become detached from the
underlying fibrous layer of the capsule, and sometimes remain
adherent to the embryo when the latter is removed from the
parent Sponge (fig. 18). In sections, owing to the treatment
undergone, they very often appear entirely isolated, having
been torn both from the embryo and from the fibrous layer
of the capsule, or they may remain adherent to the embryo
while separated from the fibrous layer, or to the fibrous layer
while separated from the embryo. Owing to its relationship
to, and intimate connection with, the outer layer of cells of
the embryo (and for certain other reasons) I believe this peculiar
lining epithelium of the embryo capsule to be nutritive in
function, but to this point I shall return again when treating
of the development.

It must be borne in mind that there is no evidence actually
to prove that these large epithelial cells belong to the mother
Sponge and not to the embryo itself, but the latter hypothesis
seems to me so improbable that I shall not consider it any
further.

(3) The Walls of the Flagellated Chambers.

The walls of the flagellated chambers are, of course, com-
posed of collared cells, but these cells exhibit certain very
peculiar and interesting details in structure.

In his article on Sponges in the ‘ Encyclopaedia Bi’itannica ’
(17) Sollas has shown that in certain Sponges the collar of the
collared cells (or choanocytes, as he terms them) are united
together at their margins by a continuous membrane which
forms a kind of inner lining to the flagellated chamber. He
says, “ In Tctractinellida, and probably in many other
Sponges — certainly in some — the collars of contiguous choano-
cytes coalesce at their margins so as to produce a fenestrated
membrane, which forms a second inner lining to the flagel-
lated chamber. The presence of this membrane enables us
readily to distinguish the excurrent from the incurrent face of

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 343

the chamber, since its convex surface is always turned towards
the prosopyle/’1

This short passage, and a not very satisfactory woodcut accom-
panying it, comprise all the information which we as yet possess
concerning this very remarkable and important structure, but
doubtless further details will be given in Professor Sollas's
forthcoming report on the “ Challenger/’ Tetractinellid a.
Meantime it has been my good fortune to be able to demon-
strate, beyond the possibility of a doubt, the existence of this
connecting membrane, which I propose to call Sollas’s mem-
brane, in Stelospongus.

In Stelospongus the collared cells are arranged at about
equal distances all around the flagellated chamber, but they
are interrupted at the proximal pole by the inhalant, and at
the distal pole by the exhalant aperture (figs. 6, 10). They are
not all of the same size ; they are largest around the inha-
lant aperture, gradually diminishing towards the exhalant
aperture, around which they are smallest (figs. 6, 10). Each
cell (figs. 8, 9) consists of a cylindrical “ collum ” or neck,
with a large oval nucleus lying in its slightly expanded base
(the body of the cell). The collum projects freely into the
chamber, and gives support to the delicate membranous collar.
The collar is rather longer than the collum, and, though
necessarily of the same diameter as the latter at its base,
considerably wider at its summit. Thus, the whole cell, in-
cluding the collar, has somewhat the shape of a dice-box,
being narrower in the middle than at the two ends. I have
not been able to trace any definite outline to the body of the
cell, which is embedded in the highly granular ground sub-
stance, but the nuclei are always very conspicuous as relatively
large, deeply-staining, oval, granular bodies, sometimes appa-
rently with a nucleolus (fig. 8).

The flagella cannot be detected in my preparations, being
entirely shrivelled up, or possibly retracted, when the Sponge
was placed in spirit. Certain granular bodies, sometimes visible
on the collars and represented in fig. 8 (g.), may possibly
1 = inhalant aperture.

344

ARTHUR DENDY.

represent the shrivelled remnants of flagella, but it is ex-
tremely doubtful. The largest collared cells of a chamber
measure about 0‘0096 mm. in total height (including the
collar), and the nucleus is about 00032 mm. in its largest
diameter.

The margins of the collars are all connected together by a
continuous, very delicate membrane, Sollas’s membrane, which
lies in a plane at right angles to the long axis of the collared
cell. This membrane is seen in thin vertical sections as a
fine thread running from collar to collar, as shown in fig. 8,
which represents an actual preparation. If the section, how-
ever, instead of being taken at right angles to Sollas’s mem-
brane, happens to be taken in a plane more or less parallel to
it, then the membrane frequently appears as an irregular
network of delicate transparent strands, shrivelled up and
distorted by the action of the reagents, but easily recognisable
lying within the chamber. Fig. 7 represents such a section.

It might perhaps be thought that if Sollas’s reticulate
membrane exhibits its true form and relationships in vertical
sections it ought also to do so in horizontal sections ; but this
by no means follows, for in horizontal sections the membrane
is severed from the collars of the cells upon which it is naturally
supported, and being no longer kept in position by these is at
liberty to shrivel up, which it promptly does.

Fig. 9 is a diagram representing what I believe to be the
natural relationships of the parts under discussion.

From what has been said of the sizes and arrangement of
the collared cells in each chamber it will be seen that the
membrane uniting their margins, Sollas’s membrane, will not
run parallel to the wall of the chamber, but will be furthest
from it at the proximal or inhalant pole, and nearest to it at
the distal or exhalant pole. This is actually the case, for at
the proximal pole the membrane is widely separated from the
wall of the chamber, while at the distal pole the two become
confluent, as shown in the diagram, fig. 10. Hence the mem-
brane has the form of a hollow cup, whose concavity is turned
towards the exhalant aperture of the chamber.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 345

It is obvious that Sollas’s membrane divides the cavity of
the flagellated chamber into two portions, (1) a central portion
into which the flagella project, and (2) a peripheral portion
lying between the collared cells. It is farther clear that the
stream of water kept up by the movements of the flagella can
pass only through the central portion of the chamber. It may
give access to the peripheral space by means of gaps between
the collared cells around the inhalant aperture, but as it has
no means of egress this would be of no consequence. In
short, the membrane may serve to facilitate the flow of the
water by diminishing the friction, for it is evident that the
water will flow more easily over a smooth membrane than if it
had to run in and out between the collared cells. It has been
pointed out to me by Sir James Hector that Sollas’s membrane
may also play some part in determining the direction of the
current of water, but this is a mechanical problem which I am
at present unable to solve.

Now that attention has once been directed to its existence
this membrane will probably be found to be very generally
present in Sponges. I may state here that I have already
detected it in H alichondria panicea, of the minute anatomy
of which species I hope to be able to give an account before
very long, so that it is now known to exist in three groups,
Tetractinellida, Ceratosa, and Halichondrin a.

(4) The General Mass of Mesoderm in which the
Chambers and Canals are Embedded.

The flagellated chambers are pretty closely packed together
in the choanosome, and together with the various branches of
the canal system make up the greater part of its bulk. Hence
the amount of fundamental or ground tissue in which they
are embedded is not very great. What there is is packed full
of minute, highly refringent granules (figs. 6, 7, 8, 9, 10),
fairly evenly distributed through it, and resembling in size and
general appearance the intracellular granules of some of the
forms of tissue already described (e. g. the fibrous cells which
form the inner layer of the embryo capsules).

346

ARTHUR DENDY.

Embedded in this granular matrix may be observed, scattered
here and there, small nucleated cells of irregularly rounded
outline (figs. 6, 10, m. c.), doubtless the amoeboid cells of
authors. This ground tissue appears to agree thoroughly
with that which Schulze has described (15) as existing in
Euspongia.

(5) The Spongoblasts and other Mesodermal Cells
surrounding the Skeleton Fibres.

In most parts the skeleton fibres are surrounded by a sheath
of ordinary stellate and slightly fibi’ous connective tissue. In
some places, however, doubtless those in which growth of the
fibre is going on and active secretion is taking place, the stel-
late mesodermal cells are specially modified as spongoblasts
or glandular cells whose function it is to secrete the spongin of
which the horny fibre is composed. In Stelospongus
flabelliformis these spongoblasts have the form indicated
in fig. 11, and they form a layer one cell thick around the
fibre. Each spongoblast is a somewhat club-shaped, slender,
elongated, granular mesodermal cell, about 0 048 mm. in
length. One end is drawn out into a long, gradually tapering
neck, and the other broader end is usually rounded off (but
sometimes stellate), and contains a spherical nucleus. The
whole cell is frequently more or less bent or contorted; its long
axis, however, always lies approximately at right angles to the
surface of the fibre against which its narrow end abuts. There
is commonly, if not always, a layer of ordinary stellate meso-
derm outside the layer of spongoblasts, and it is easy to see
that the spongoblasts themselves are simply slight local modi-
fications of the ordinary stellate type of cell, their origin
being still sometimes plainly indicated by the stellate form of
the broad end (fig. 11).

The spongoblasts thus described are practically identical
with those observed and figured by Schulze in Euspongia
(15), with the exception that they are very much more elon-
gated.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 347

Notes on the Development.

My observations on the embryology of Stelospongus are
as yet necessarily very imperfect, for all the embryos which I
have yet found are in pretty much the same stage of develop-
ment. Of this particular stage there is, however, an abundant
supply, and it presents such very remarkable features that I do
not hesitate to give a detailed account of it in this place, hoping
at the same time to be able to extend my observations at a
later date.

Historical Account.

In connection with this portion of our subject it is necessary
to bear in mind in the first place some very remarkable ob-
servations of Mr. Carter’s (1, 2, 3, 4, 5), which, though published
many years ago by such a careful observer, appear to have
almost entirely sunk into oblivion. The gist of the observations
referred to is that in the developing gemmule (seed- like
body. Carter) of Spongilla the flagellated chambers (a m-
pullaceous sacs, Carter) are formed each from a siugle
large amoeboid mesodermal cell whose contents break up into
a number of small cells (germs or ovules, Carter) which
arrange themselves round a central cavity and develope into
collared cells.

A few quotations will suffice to illustrate this point. Mr.
Carter (4) says that Spongilla “is charged towards the base
with a number of seed-like bodies of a globular shape, each of
which consists of a coriaceous membrane enclosing a number of
delicate, transparent spherical cells, more or less filled with
ovules and granular matter. ... It has also been shown that
at an early period of development the spherical, which we shall
henceforth call ‘ ovibeariug,’ cells are polymorphic — identical
but for the ovules, with the ordinary sponge- cells — and
surrounded by a layer of peculiar cells equally polymorphic,
which I have conjectured to be the chief agents engaged in

constructing the capsule The seed-like body

presents a hole, which we shall call the ‘ hi! urn.’ ....

348

ARTHUR DENDY.

The contents issue through the liilum under the form of a
gelatinous mass, in which the ovibearing cells and their

contents appear to be embedded entire The

ovibearing cells are developed into spherical ampullaceous sacs,
communicating with the afferent canals The ovi-

bearing cells do not burst and allow their contents to become
indiscriminately scattered through the gelatinous mass in
which they are embedded, but each becomes developed
separately and entire in the following way, viz. the ovules and
granules of the ovibearing cell subside into a granular mass
by the former losing their defined shape and passing into small
monociliated and uniciliated sponge-cells ; this mass then
becomes spread over the interior surface of the ovibearing cell,
leaving a cavity in the centre into which the cilia of the mono-
ciliated sponge-cells dip and keep up an undulating motion ;
meanwhile an aperture becomes developed in one part of the
cell which communicates with the adjoining afferent canal,
and thus the ovibearing cell passes into an ampullaceous
spherical sac.”

It is astonishing how such a precise account, coming from
the pen of so careful an observer as Mr. Carter, has received so
small a share of attention from subsequent writers. I do not
see any reason to doubt the accuracy of Mr. Carter’s statements,
and I shall presently endeavour to show that the flagellated
chambers develope in precisely the same manner in the em-
bryos of Stelospongus.

Two other authors, viz. Metschnikoff and Goette, have
described a mode of development of the flagellated chambers
which appears to me to agree pretty closely with that observed
by Mr. Carter in the gemmules of Spongilla, and by myself
in the embryos of Stelospongus. Unfortunately, I am
unable to obtain access to the original papers of either of these
authors, and I am obliged therefore to content myself wTith the
very brief abstracts, fortunately accompanied in the first case
by figures, given by Vosmaer (18).

Metschnikoff (12) describes and figures the embryo of
Halisarca duj ardini at a certain stage as consisting of an

STUDIES ON THE COMPARATIVE ANATOMY OP SPONGES. 349

outer layer of columnar epithelium and a few “ rosette-cells ”
in the interior. The origin of these so-called “ rosette-cells ”
is unknown; from the figures, copied by Yosmaer, they appear
to consist each of a spherical ball composed of a few small cells
arranged around a small central cavity. The rosette-cells
increase until finally they fill the whole cavity of the embryo.
Metschnikoff regards the rosette-cells as mesoblastic structures,
aud he states also that amoeboid cells wander in amongst them
from the epiblast ; the canals commence as isolated spaces in
in the mesoblast. Vosmaer gives no further account, but from
what he says, aud from the figures, it seems very probable that
the so-called rosette-cells are the young flagellated chambers,
and that they have themselves been formed each from a single
large amoeboid cell derived from the epiblast.

Goette (10) gives a long account of the development of
Spongilla, and Yosmaer tells us in his abstract that “Die
Geisselkammern entwickeln sich getrennt von einander und
von anderen Hohlraumen ‘ vermittelst einer Knospenbildung
einzelner Zellen.’ M

Development of Stelospongus flabelliformis.

(a) The Ovum.

The ovum appears in section as a somewhat oval cell lying in
a fibrous capsule as described above. The body of the ovum
is granular and deeply staining. At one pole there is a large
oval nucleus with a very definite wall, and right up against
the wall, at the outer pole of the nucleus, there lies a small,
spherical nucleolus. The nucleus does not contain nearly so
many granules in proportion to its size as the body of the
ovum, but there is a quantity of coarsely granular material
chiefly aggregated towards the pole, remote from the nucleolus.
The nucleolus stains deeply, and is almost, if not quite, homo-
geneous. The longer diameter of the ovum measures 0-076 mm.,
of the nucleus 0 024 mm., and of the nucleolus 0 0048 mm.

350

ARTHUR DENDY.

(b) The Embryos.

As already stated, the embryos are all in pretty much the
same stage of development, although varying in diameter from
about 3 mm. to nearly 5 mm. The immense size of these
embryos as compared with those of other Sponges has already
been noticed by Mr. Carter1 (7), but he says scarcely a word
about their structure.

While out dredging with Mr. Wilson I picked a number of
these embryos out of the living Sponge with forceps, and tried
the effect of at once placing in fresh sea-water, but I could
detect no signs of motion of any kind.

All the embryos were solid, with the exception of one or
two of the smaller ones. These when examined in spirit
appeared to be hollow, but they were damaged, and I believe
the hollow character was a post-mortem condition2 due to the
escape or shrivelling up of the very delicate gelatinous, or
probably in the youngest stages more or less liquid, tissue
from the interior. I shall therefore not consider them apart
from the remainder.

When the surface of the embryo is examined with a pocket
lens it exhibits a minutely punctate appearance, due to the
presence of an immense number of shallow pits, somewhat
polygonal in outline, and separated from each other by low
ridges (fig. 18). Sections show that each one of these pits is
the imprint of one of the large epithelial cells of the embryo
capsule. The pittings were present, in parts at any rate, on
the smallest embryo examined, but they were not nearly so
well marked as on the older embryos. All my further observa-
tions were conducted by means of sections.

The embryo consists of an outer layer of rather large, closely
packed cells enclosing a mass of clear, transparent, jelly-like
substance, in which immense numbers of amoeboid wandering

1 Mr. Carter says : “ The largest embryo I have seen in the marine Sponges
is that of Stelospongus flabelliformis, Cart., .... where it is spherical
and one sixth of an inch in diameter.”

3 Cf. Quasillina.

STUDIES ON THE COMPARATIVE ANATOMY OE SPONGES. 351

cells are embedded. The outer layer, or ectoderm, consists of
a single layer of large, sac-shaped or somewhat flask-shaped
cells (figs. 19, 20, 21, 22) measuring about 0’024 mm. in
length. The narrower portion, or neck, of the cell is on the
outside of the embryo, and the swollen portion projects in-
wards into the gelatinous intercellular substance, and from its
inner extremity frequently sends out a few very short, slender
pseudopodial processes (figs. 19, 20, 21, 22). These processes
may possibly serve to transmit nutriment to the cells in the
interior, or they may simply indicate a tendency in the ecto-
dermal cells to become amoeboid. The body of the cell is
coarsely granular, but less so in the neck than in the swollen
portion. The greater part of the neck is, however, occupied
by a large, spherical nucleus, which appears to consist of a
hollow vesicle containing a few deeply-staining granules. I
have observed no nucleolus. The nuclei are all arranged at
just about the same level, so that the nuclei of adjacent cells
form a continuous row, which is a very conspicuous feature in
sections of the embryo.

Frequently the outer end of the neck of the ectodermal cells
may be seen to be drawn out into a short, slender, protoplasmic
process, which extends to the outer surface of one of the large
investing epithelial cells (fig. 21), and attaches itself to the
latter. Thus the ectodermal cells of the embryo often appear
to be suspended from the outer surfaces of the investing epi-
thelial cells by short protoplasmic processes, as shown some-
what diagrammatically in fig. 21. Judging from the number
seen in a single transverse section, it would appear that each
of the large epithelial cells may have a hundred or more sac-
shaped ectodermal embryonic cells hanging from its outer
surface.

The unusual length of time during which the embryo
remains within the mother Sponge, and the great size to which
it attains, necessitate some special arrangement whereby it
can be nourished. The peculiar relation of the ectodermal
cells of the embryo to the investing epithelium, and the very
unusual character of the latter, cause me to believe that the

352

ARTHUR DENDY.

investing epithelium lias for its function the nutrition of the
embryo, and that this is effected by the absorption of nutri-
ment through the elongated necks of the ectodermal cells.

Some of the ectodermal cells, however, exhibit no prolonga-
tions of the neck, but are smoothly rounded off at the free
end, and such cells may form a continuous layer over a con-
siderable area. In most sections, however, owing to the
forcible displacement of the nutrient epithelial cells and the
rupture of the delicate connections between them and the
ectodermal cells, the latter appear as if broken off at their
outer ends, just outside the nucleus (figs. 19, 20, 22).

The entire mass of the embryo within the ectodermal layer
is made up of a clear, jelly-like matrix, in which immense
numbers of large, amoeboid wandering cells are embedded
(figs. 19, 22). These cells appear somewhat larger than the
ectodermal cells, but I shall show presently that there is very
strong reason to believe that they are simply ectodermal cells
which have left their places in the outer layer, and, becoming
amoeboid, wandered into the central jelly. Between the large
amoeboid cells very delicate branching stellate cells may some-
times be seen (fig. 22, st. c.).

The amoeboid cells may put out pseudopodia in all directions,
but often they appear to be radially elongated and more or less
bipolar. I think my sections, and especially such as that
represented in fig. 22, show conclusively that the amoeboid
cells are derived from the ectodermal layer. They agree
firstly in all essentials with the cells of the latter, and in those
parts where the ectodermal cells, having the clearer, outer
end of the neck evenly rounded off, present a characteristic
feature, a precisely similar clear, rounded-off neck may often
be seen in the amoeboid cells immediately beneath the ecto-
derm. In fig. 22 two cells appear just leaving the ecto-
dermal layer and becoming amoeboid by the emission of
pseudopodia. The amoeboid cells are from the first highly
granular and, at what I believe to be an early stage of the
proceedings, each one has a spherical nucleus resembling that
which occurs in the ectodermal cells. Sometimes the amoeboid

STUDIES ON THE COMPAEATIVE ANATOMY OE SPONGES. 353

cells lying near the outside of the embryo have two or three
nuclei (fig. 22), and very rarely also even the ectodermal cells
appear to have two nuclei (fig. 20). At a later stage (fig. 23)
the entire amoeboid cell is seen to have become indistinct in
outline, and in place of one large cell we have an aggregation
of very minute spherical bodies about 00025 mm. in diameter,
each with a dark spot in its centre ; but each aggregation still
retains the form of the original amoeboid cell. In the same
sections which exhibit this condition many of the amoeboid
cells appear to have become rounded, their contents having
arranged themselves around a central cavity (fig. 23,/. c.), so
that we have a hollow chamber lined by small, spherical cells.
These chambers I believe to be the young flagellated chambers.
They are certainly very different in structure from the flagel-
lated chambers of the adult Sponge, and only about half the
size, measuring about 0024 mm. in diameter, but the dif-
ferences are easily accounted for by their embryonic condition.
I have not been able to trace the development of the chambers
any further, nor is it to be expected that the collars and
flagella would be developed before the young Sponge was set
free and required them.

Coincidently with the formation of the flagellated chambers
in the manner thus described, a slit-like invagination appears
on the surface of the young Sponge, and it is chiefly, if not
solely, around this invagination that chamber formation takes
place. This invagination is probably the commencement of a
communication between the chambers and the exterior. Un-
fortunately, I have only obtained a single embryo which is
sufficiently advanced to show the formation of the flagellated
chambers and the slit-like invagination from the exterior, but
I see no good reason for doubting the normality of the pheno-
mena above described.

The mode in which the embryos of Stelospongus escape
from the parent is still an enigma. It may be that by further
increase in size they rupture the walls of the oscular tubes in
whose immediate proximity they lie, and are then forcibly
ejected with the outgoing stream of water; or it may be that

354

ARTHUR DBNDY.

the Sponge dies down in the winter, and that the embryos are
then released by the decay of the maternal tissues.

I do not think it advisable at present to enter into any
speculations with regard to the general significance of the
development of Stelospongus, but prefer to wait for more
light on the subject.

In conclusion, I have to express my cordial thanks to
Professor Howes, of the Royal School of Mines, for kindly
undertaking to correct the proof-sheets of this paper in my
absence from England.

List of Memoirs referred to.

(1) Carter. — “Notes on the Species, Structure, and Animality of the

Freshwater Sponges in the Tanks of Bombay,” ‘ Ann. and Mag. Nat.
Hist.,’ ser. 2, vol. i, p. 303.

(2) Carter. — “ A Descriptive Account of the Freshwater Sponges (genus

Spongilla) in the Island of Bombay, with Observations on their
Structure and Development,” ‘ Ann. and Mag. Nat. Hist.,’ ser. 2,
vol. iv, p. 82.

(3) Carter. — “Notes on the Infusoria of the Island of Bombay,” No. 1,

“Organisation” (plate vi, fig. 41), ‘Ann. and Mag. Nat. Hist.,’
ser. 2, vol. xviii, pp. 115, 221.

(4) Carter. — “ On the Ultimate Structure of Spongilla, and Additional

Notes on Freshwater Infusoria,” ‘ Ann. and Mag. Nat. Hist.,’ ser. 2,
vol. xx, p. 21.

(5) Carter. — “ On the Nature of the Seed-like Body of Spongilla, &c.,”

‘Ann. and Mag. Nat. Hist.,’ ser. 4, vol. xiv, p. 97.

(f>) Carter.— “Descriptions of Sponges from the Neighbourhood of Port
Philip Heads, South Australia,” continued, ‘Ann. and Mag. Nat.
Hist.,’ ser. 5, vol. xv, p. 301.

(7) Carter. — “On the Reproductive Elements of the Spongida,” ‘Ann.
and Mag. Nat. Hist.,’ ser. 5, vol. xix, p. 350.

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 355

(8) Dendy. — “ Observations on the West Indian Chalininae, with Descrip-

tions of New Species” (Abstract), ‘Proc. Zool. Soc. Lond.,’ 1887,
p. 508.

(9) Dendy. — “ Studies on the Comparative Anatomy of Sponges, I, On

the Genera ftidleia, n. gen., and Quasillina, Norman,” ‘ Quart.
Journ. Micr. Sci.,’ N. S., vol. xxviii, p. 513.

(10) Goette. — “Untersuchungen zur Entwickelungsgeschichte von Spongilla

fluviatilis,” ‘ Abhandlungen zur Entwickelungsgeschichte derThiere,

III.’

(11) Lendenfelb. — “A Monograph of the Australian Sponges, Part VI,

The Genus Euspongia,” ‘Proc. Linn. Soc. New South Wales,’ vol.
x, part 2, p. 48.

(12) Metschnikoff. — “ Spongiologische Studien,” ‘ Zeitschr. fiir wiss. Zool.,’

xxxii, p. 349.

(13) Polejaeff. — ‘ Report on the Keratosa Dredged by H.M.S. “ Chal-

lenger.” ’

(14) Ridley and Dendy. — ‘ Report on the Monaxonida Dredged by H.M.S.

“ Challenger.” ’

(15) Schulze. — “ Untersuchungen iiber den Bau und die Entwicklung der

Spongien,” ‘ Zeitschr. fiir wiss. Zool.,’ xxx, et seq.

(10) Sollas. — “The Sponge-fauna of Norway,” ‘ Ann. and Mag. Nat. Hist.,’
ser. 5, vol. ix, p. 141.

(17) Sollas. — Article, “Sponges,” ‘Encyclopaedia Britannica,’ edition ix.

(18) Vosmaer. — “ Spongien (Porifera),” ‘ Bronn’s Klassen und Ordnungen

des Thierreichs,’ vol. ii.

356

ARTHUR DENDT.

EXPLANATION OF PLATES XXX, XXXI, XXXII, &

XXXIII,

Illustrating Mr. Arthur Dendy’s paper, “ Studies on the Com-
parative Anatomy of Sponges/’ II, “ On the Anatomy
and Histology of Stelospongus flabelliformis. Carter;
with Notes on the Development.”

(The following Explanation of llie Lettering applies to all the Figures.)

a. c. Amoeboid cell in the embryo, c. Collar of collared cell. c. c. Collared
cell. c. e. 1. Cell of external layer (ectoderm) in the embryo, cli. Choano-
some. col. Collum or neck of collared cell. eg. Cystenckyme. e. Embryo.
e. c. Embryo capsule, e. c. c. Exbalant cameral canaliculus, eel. Ectosome.

e. 1. Exbalant lacuna, e. 1. c. External layer of cells (ectoderm) of the em-
bryo. e. o. exbalant opening of flagellated chamber, f. Skeleton fibre.

f . Skeleton fibre projecting at the surface of the Sponge, f. c. Flagellated
chamber, fl. Flagellum of collared cell. g. Granule on the margin of the
collar of the collared cell. i. c. c. Inhalant cameral canaliculus, i. 1. Inha-
lant lacuna or channel, i. o. Inhalant opening of flagellated chamber, m. c.
Mesodermal cell. n. Nucleus, n. c. c. Nucleus of collared cell. n. e. c.
Nucleus of epithelial cell. nut. c. Nutrient epithelial cell from the lining of
the embryo capsule, nut. e. Nutrient epithelium from the lining of the
embryo capsule, o. Osculum. o. t. Oscular tube. p. Pore. p. a. Pore-
area. p.f. Primary fibre of skeleton, pt. Pit or depression on the surface
of the embryo, caused by a nutrient epithelial cell. r. b. Deeply staining,
radiately fibrous, globular bodies of unknown function, probably parasitic
organisms, r. m. Reticulate membrane in flagellated chamber, Sollas’s mem-
brane. s. f. Secondary fibre of skeleton, s. g. Sand grain, spb. Spongo-
blast. st. c. Stellate cell in the embryo.

PLATE XXX.

Stelospongus flabelliformis.

Fig. 1. — Specimen of Stelospongus flabelliformis, from which a
portion of one side has been cut off. Natural size.

Fig. 2. — Outer portion of a section at right angles to the surface of the
Sponge, showing the relations of the ectosome (eel.) to the ckoanosome (eh.),
and the proximal portions of the inhalant canal system commencing with the
pores (p.).

Fig. 3. — A small portion of the skeleton, showing the primary fibres (p.f.)
and the secondary fibres ( s.f ).

STUDIES ON THE COMPARATIVE ANATOMY OF SPONGES. 357

PLATE XXXI.

Stelospongus flabelliformis.

Pig. 4. — A small portion of the ectosome, removed from the surface and
examined from above as a transparent object after staining with borax carmine,
showing the pores (/>.) arranged in pore-areas (p. a.); the cystenchymatous
tissue (cy.) lying beneath is seen through the transparent epidermis. Drawn
under Zeiss’s C, ocular 2.

Fig. 5. — A solid section, showing the oscular tube ( o . t.) partially dissected
out, with the openings of its various branches ; an osculum (o.) and two
embryos ( e .). On the right of the oscular tube are two of the embryo cap-
sules ( e . c .) from which the embryos have been removed, x 2.

Fig. 6. — Portion of a section through the choanosome, showing the rela-
tions of the ultimate inhalant lacunae (i. e.), the flagellated chambers (/. c.)
and the ultimate exhalant lacunae ( e . L). Drawn under Zeiss’s F, ocular 2.

Fig. 7. — Section through a single flagellated chamber, showing Sollas’s
membrane (r. m .) as it very often appears in ordinary preparations. Drawn
under Zeiss’s F, ocular 2.

Fig. 8. — Small portion of an actual vertical section through the wall of a
flagellated chamber, showing three collared cells connected together at the
margins of the collars by a delicate membrane, seen iu section ( r . m.), Sollas’s
membrane. The granular bodies (y.) on the collars ( c .) may possibly represent
the last remnants of shrivelled-up flagella. Drawn under Zeiss’s F,
ocular 2.

PLATE XXXII.

Stelospongus flabelliformis.

Fig. 9.— Diagram of a portion of the wall of a flagellated chamber, showing
the various parts of the collared cells and their relations to the reticulate
membraue (Sollas’s membrane, r. m.) which connects together the margins of
the collars.

Fig. 10. — Diagram of a section through a flagellated chamber, passing
through the inhalant (i. o .) and exhalant (<?. o.) openings, showing the arrange-
ment of the collared cells ( c . c.) and the relations of the inhalant (*. 1.) and
exhalant ( e . 1.) lacunae, and inhalant (i. c. c.) and exhalant {e. c. c.) cameral
canaliculi.

Fig. 11. — A group of spongoblasts ( spb .), showing their form and relations
to the skeleton fibre (/.). Drawn under Zeiss’s F, ocular 2.

Fig. 12. — Portion of a plexus of stellate mesodermal cells from around a
grain of sand in the ectosome. Drawn under Zeiss’s F, ocular 2.

Fig. 13. — Cystenchyme from the ectosome, from a section stained with

VOL. XXIX, PART 3. NEW SER.

A A

358

ARTHUR DENDY.

borax carmine and mounted in Canada balsam. Drawn under Zeiss F,
ocular 2.

Fig. 14. — A single cystenchyme cell from the ectosome, showing the
nucleus ( n .) suspended in a network of protoplasmic threads. From a pre-
paration mounted in glycerine without embedding. Drawn under Zeiss’s F,
ocular 2.

Fig. 15. — Portion of the fibrous layer of an embryo capsule. The right
hand side of the figure corresponds to the inner portion of the layer, next to
the lining epithelium. Drawn under Zeiss’s F, ocular 2.

PLATE XXXIII.

Stelospongus flabelliformis.

Fig. 16. — Group of large nutrient epithelial cells, from the lining epithelium
of an embryo capsule. Drawn under Zeiss’s D, ocular 2.

Fig. 17. — Vertical section of one of the large nutrient epithelial cells from
the lining epithelium of an embryo capsule. Drawn under Zeiss’s F; ocular 2.

Fig. 18. — Portion of the surface of a large embryo, as seen with a hand
lens, showing the shallow pits on the surface (pi.), aud on the left of the
figure the nutrient epithelium from the embryo capsule [nut. e.) still adherent.

Fig. 19. — Portion of a radial section of an embryo, showing the ectodermal
layer of cells ( e . 1. c.), and the large amoeboid cells [a. c.) embedded in the
inner jelly-like mass. (The transparent gelatinous matrix is not represented.)
Drawn under Zeiss’s C, ocular 2.

Fig. 20. — Outer portion of a section similar to that represented in Fig. 19,
more highly magnified, with the remains of a nutrient epithelial cell [nut. c.)
still visible. The section corresponds to a transverse section through a single
one of the shallow pits represented in Fig. 18.

Fig. 21. — Vertical section through one of the large nutrient epithelial cells
[nut. c.), with the ectodermal cells of the embryo [c. e. 1.) attached to its inner
surface by means of slender prolongations of their necks. Drawn under
Zeiss’3 F, ocular 2.

Fig. 22. — Small portion of a section through an embryo, showing how the
ectodermal cells [c. e. 1.) become amoeboid [a. c.), and migrate inwards. Drawn
under Zeiss’s F, ocular 2.

Fig. 23. — Portion of a section through the interior only of the most
advanced embryo, showing how the large amoeboid cells [a. c.) break up into a
great many minute spherical cells, which arrange themselves around a central
cavity and form the young flagellated chambers [/. c.). Drawn under Zeiss’s
F, ocular 2.

'&M:*

<:r

.!'* t •’ * . '< «; f

* *•»§ u • • f •

Fi$- z-

nat. size.

Fig. I.

Arthur Bendy del

F Huth, Lith* Edmr

MwJtwrris%mXM&6. III/.

nee

7v.ec

Arthur

ne.c

ife

FLct 6

Tb C C.

T Hulh. Lilh' Ed>r?

Arlhur Dendy d,i

f Hufh.LifliT Edin'

Muyr. Jotum/%6 XXIX, M$i. XXXIII.

fe'AOS

^ej&'

F Huth. L:thT Edmr

Arthur Dendy del

SOME POINTS IN NATURAL HISTORY OP PUNGIA. 359

On Some Points in the Natural History of

Fungia.

By

J. J. Lister, M.A.

Duiung a visit to the Seychelles Islands at the latter end of
last and the beginning of the present year 1 was so fortunate
as to find species of the Madreporian Coral Fungia, both in
the fixed and free conditions. Owing to the shortness of the
time now at my disposal I have not been able to examine the
material I have brought home as completely as I hope to on
some future occasion, but as it throws light on some of the
stages of the life-history of this group of Corals which have
hitherto been obscure, a preliminary account may not be
without interest.

The Fungias are abundant in water from one to six feet
deep towards the outer edge of the broad fringing reef of
Malie, the principal island of the group. The reef is divided
by deep winding channels, whose sides are formed by living
Corals of many kinds. Between the channels the Coral comes
very near the surface, so near that the summits of the branches
of a blue tipped Madrepore, which is very abundant in these
shallow areas, are out of water at low tides, though apparently
without injury to the Coral.

Over some areas this Madrepore has died, and it was on the
dead branches that the fixed stocks of the Fungia were most
abundant. Among the free forms Fungia discus and F.
dentata were abundant.

The young fixed stocks of Fungia are attached by a broad

360

J. J. LISTER.

base, and have vertical thecal walls. The youngest that I
found have six septa conspicuously larger than the rest (Fig.
1), one at either end of the long axis, passing through the
mouth, and two symmetrically placed on either side. In the
intervals smaller septa have made their appearance, but they

Fig. 1. — A young fixed stock of Fungia(sp.F) X about 25. From a speci-
men preserved in spirit.

are lower and do not approach so near the centre as the six
large ones. In each interval there is one in the centre and
two small ones on either side of it. The appearance of fresh
septa does not, however, take place quite regularly, for while
in one interval between the primary septa there may be three
smaller ones well developed, in another the central one alone
may be only just discernible, at least in spirit specimens, in
which the skeleton is invested with the soft tissues.

The young stock has, as has been stated, vertical thecal
walls. After a certain, apparently very variable, height is
attained the upper part begins to widen out, forming at first a
very shallow cup with thecal walls facing outwards and down-
wards, and finally a disc, depressed in the centre, with the
thecal walls facing directly downwards. The cup or disc is
attached by the narrow stalk, the first formed part of the
stock.

SOME POINTS IN NATURAL HISTORY OP FUNGIA. 361

After the disc has become distinctly formed, though the
breadth it may have attained is very variable, a remarkable
process sets in, which results in its separation from the stalk
which has hitherto supported it. In a plane at right angles to
the axis of the stalk, at a point where the upper part is
beginning to widen out, absorption of the calcareous skeleton
takes place, which goes on till the disc is connected with the
stalk so weakly that a very small force is needed to set it free.
It often happens that the disc falls off when the object on
wThich the Fungia is growing is lifted from the water.

When the disc is set free it has a round scar in the middle
of the under surface which corresponds with a similar scar at
the summit of the stalk. In the scars the following parts of
the skeleton are exposed, with the soft tissues investing them.
On the outside there is a section of the thecal wall. Passing
from this towards the centre are sections of the septa, and
these unite with the trabeculae which fill in the middle. In
the disc there is no direct communication with the gastric
region, except through the interspaces among the trabeculae.
The surfaces of the calcareous structures where absorption has
taken place are white and opaque as compared with the general
appearance of the hard parts of the Coral.

The disc thus liberated is carried into some depression in
the reef, where it lies unattached, leading an independent
existence. The scar on the aboral surface becomes covered in,
and though it remains distinguishable for some time, ultimately
all trace of it is lost. On the separation of the disc the stalk
is left with a truncated top, slightly depressed in the middle.

The first change which takes place that is visible in dry
specimens is in the state of the septa (Fig. 2). These, instead
of terminating in broken edges, throw up delicate fluted laminae
with serrated edges, which project above the level of the other
structures of the scar.

A mouth is formed in the centre, and the lips appear, in
spirit specimens at least, to be almost in contact with the
trabeculae below. As the septal laminae rise higher a thecal
wall becomes formed round them, in some places continuous

362

J. J. LISTER.

with the thecal wall of the stalk, but generally springing a
little within its margin, so that the edge of the old thecal wall
remains as a prominent ridge round the stalk. A new cup is
thus formed, not as a bud but as a product of the growth of
the structures already existing in the base of its predecessor.
As its walls grow higher they become more and more expanded

Fig. 2. — A nurse stock of Fungia (sp. ?), from which a disc has been recently
separated X 3. The formation of a new disc has begun, as is indicated
by the laminae which have been formed on the edges of the septa in the
stalk of the old one. The ridge half way down the Coral indicates the
line of separation of an older disc. From a dried specimen.

outwards, until a new disc is formed supported on a short
stalk which springs from the point where the first disc was
separated.

This second disc in its turn is set free by absorption of the
calcareous skeleton at a point where the stalk begins to
widen into the disc, and in due course a third disc is formed.
As the process is repeated the stalk grows in height step by
step, each new disc that is formed being detached at a point
above that from which its stalk springs. These successive ad-
ditions are indicated on the common stalk by ridges which
mark the planes where discs have been separated.

The specimens obtained show many instances of the forma-

SOME POINTS IN NATURAL HISTORY OP PUNGIA. 363

tion of buds at the expanded bases of the fixed stalks, but in
none was there any evidence that the disc which grows on the
scar at the summit of the stalk is produced by budding. The
structures of the new disc are, as we have seen, the product of
the growth of the corresponding structures in the stalk of the
disc which went before it.

AMPHIOXUS LANCEOLATUS.

365

Contributions to the Knowledge of Amphioxus
lanceolatus, Yarrell.

By

£. Ray l>anli.ester, M.A., LL.l),, F.R.S.,

Professor iu University College, London,

With Plates XXXIY, XXXV, XXXVI, XXXVI A, & XXXVI B.

It is now fourteen years since I published in the f Quarterly
Journal of Microsc. Science ’ (vol. xv) some notes on the
structure of Amphioxus. I have delayed publishing until the
present date fuller illustrations of the facts then recorded, but
have made use of my material in annually recurring lectures
at University College.

One statement made in the notes above referred to I am not
able to confirm, and must withdraw. I refer to the supposed
confirmation of Johann Muller’s statement (1) that there is a
pair of apertures on either side of the oral sphincter (velum of
Huxley). I stated that these apertures lead from the pharynx
to the prseoral space. Muller had described them as leading
into the metapleural canals. In reality there are no such
apertures at all.

The “ brown funnels,” which were described in my original
note, are the most important structures which I have now to
illustrate. Curiously enough, they have escaped all subsequent
observers with the exception of Mr. William Bateson (2), of
St. John’s College, Cambridge, who confirmed my observation
as to their position and character, and has compared them
very significantly with the “ collar-pores” of Balanoglossus.

My purpose on the present occasion, in addition to that of

VOL. XXIX, PART 4. NEW SER. B B

366

E. RAY LANKESTER.

definitely exhibiting the position and form of the brown funnels,
is (1) to furnish a few numerical data of importance for the
anatomical discussion of Amphioxus ; (2) to correct some
errors which appear to be current as to the existence or non-
existence of spaces of one kind and another in the body and
gill-bars of Amphioxus; and (3) to submit some drawings
which represent, in a semi-diagrammatic form, the structure
of Amphioxus, not merely as seen in sections or dissections,
with all the imperfections necessarily arising from the action
of preservative media, but as reconstructed and corrected from
numerous specimens, so as to give as nearly as may be a true
conception of the undistorted organism.

External Marks and Numerical Characteristics.

The general outline and form of a living specimen of
Amphioxus lanceolatus is given in PI. XXXIV, fig. 4.
The drawing is constructed from sketches made by me at
Naples from the living animal, and has been corrected by sub-
sequent study of preserved material. When Amphioxus is alive
and at rest the atrial chamber is dilated in such a way that its
median ventral surface projects below the two lateral ridges,
for which I have proposed the name “ metapleura. ” I doubt
whether in life this surface is ever contracted to the extent
which it is in even the most carefully preserved specimens,
such, for instance, as that shown in PI. XXXV, fig. 3. That
specimen was treated with Kleinenberg's picro-sulphuric solu-
tion, followed by increasing strengths of alcohol ; and I have
not yet found any treatment which gives a less general distor-
tion of the body than this. Specimens placed when living into
alcohol assume the most extreme distortion, owing to the vio-
lent contraction of the transverse ventral muscle of the atrial
wall, and the shrinking of tissues and spaces. Such an extreme
contraction is exhibited in the figures of Rolph’s important
treatise on Amphioxus (3), and to a less extent in figs. 2 and 3
of PI. XXXIV accompanying this memoir. In PI. XXXVI
I have given a diagram of a transverse section with such form
and proportion of all regions and spaces as I have been led to

AMPHIOXUS LANCEOLATUS.

367

conclude are actually maintained in the living state. I do not
know of any one reagent which gives equally good results for
all parts of the Amphioxus body. I have found it necessary to
study specimens preserved in several different ways.

Grooves of the Ventral Wall. — Even when somewhat dis-
tended; as shown in PI. XXXIV, fig. 3, the ventral wall of the
atrium of Amphioxus exhibits longitudinal plaiting. These
folds have been observed and counted on living specimens at
my request, by my friend Balfour, and by others who have
had the opportunity of studying living Amphioxus at Naples,
since I was there in 1875. And there is no doubt that they
are not “ artifacts/’ but exist iu the living state, though
their depth is increased by the unnatural contraction caused
by preserving fluids. They are best seen in the ventral view
of a living specimen given in PI. XXXIV, fig. 4. As far
as I can ascertain they do not vary in number in the same
individual, except as the result of the general increase of the
animal’s size. All the folds do not extend the whole length of
the ventral surface : some stop short anteriorly. I have
counted from six to eight on each side of the middle line.
They entirely disappear when the ventral wall of the atrium is
fully stretched, as it is when the generative products are full
grown and ready for extrusion (see PI. XXXV, fig. 4). Their
production is accounted for by the insertion of some of the
fibres of the ventral transverse muscle into the somewhat
thick tegumentary connective tissue, in a series of lines cor-
responding to the grooves. This insertion can be readily
observed in thin transverse sections (see PI. XXXVI.4, fig. 2).

Absence of Canals below the Ventral Plaited Integument. —
The epidermis supported by a fine basement membrane fre-
quently becomes separated by the action of reagents from the
thick subjacent connective tissue of this region, and has given
rise to an erroneous conclusion, to the effect that thei’e is a
series of ventral canals underlying the plaited epithelium.
Or, on the other hand, the muscle becomes separated from
the deeper layer of the cutis, and a similar mistake has arisen.
This error is made by Stieda (4), Rolph (3), Langerhans (5),

368

B. BAY LANKESTER.

and Schneider (6). The true relations are shown in the
drawing, PI. XXXVI A, fig. 2.

Number of Myotomes. — The number of myotoines is an
important and fundamental numerical character of the species
of Amphioxus. In the Amphioxus lanceolatus of Naples
there are sixty-one of these myotomes. The last myotome is
extremely delicate, and it is by no means easy to count the
whole series with certainty. In some large specimens I have
counted sixty-two myotomes.1 Dr. Gunther (7) in his im-
portant account of the genus Branchiostoma (= Amphioxus)
in the ‘ Report on the Zoological Collections made in the Indo-
Pacific Ocean during the voyage of H.M.S. “ Alert/’ ’ published
by the trustees of the British Museum, 1884, gives sixty as
the number of myotomes in specimens of A. lanceolatus
of Naples, fifty-nine in one from Polperro, and sixty-one in a
specimen from the Scandinavian coast. The number in other
species varies as follows : — In A. elongatus from the coast
of Peru, 79; in A. bass an us from Bass’s Straits, 75 or 76; in A.
Belch eri from the coast of Borneo, 64 or 65 ; in A. caribaeus
from Rio de Janeiro, 60 or 59; in A. cultellus of Peters, 52.

It appears that the full number of myotomes is acquired by
Amphioxus at a very early period in its growth, even before
(?) the complete formation of the epipleural chamber. It is
not quite certain that the number of myotomes varies in A.
lanceolatus from fifty- nine to sixty-two, as would appear
from the numbers above given. The discrepancy may be due
to the difficulty of accurate counting, and to the recognition or
omission of the terminal myotome. The question, therefore,
needs some further study.

Position of Mouth, Atriopore and Anus. — The mouth of
Amphioxus is that small median aperture surrounded by a
well-developed sphincter muscle which is concealed by the oral
hood. It is not correct to call the margin of the wide space

1 I Lave a note of the number of myotomes counted in four large specimens
besides those referred to in the text, viz. respectively fifty-nine, sixty-two,
sixty-one, and sixty-one. In the figure in Plate XXXIV I have given sixty-
two myotomes, but sixty-one is the usual number.

AMPHIOXUS LANCEOLATUS.

369

bounded by the oral hood “ mouth,” since the true mouth
above indicated exists before the oral hood is formed. The
oral hood is the preeoral portion of the epipleural folds, which
post-orally give rise to the “atrial/’ “branchial/’ or “epi-
pleural” chamber. The true mouth is that which has been
compared by Huxley (9), whose nomenclature is followed by
Langerhans, to the velum palati of Cyclostome fishes. It
has twelve delicate tentacles projecting freely from its margin
backwards into the pharynx. The grouping of these has
not hitherto been satisfactorily figured in any account of
Amphioxus; they are represented in PI. XXXVI B, fig. 12.
They were seen by Rathke (8) and by Joh. Muller (1), who
figure them as seen when the sphincter is slit open ; their
minute structure is figured and described by Langerhans, who
calls them papillae.

It is difficult to assign a position in relation to the myo-
tomes to those organs which lie more ventrally than the
segmented musculature of the body wall. The myotomes are
separated from one another by connective-tissue septa, each of
which, instead of being vertical, is directed obliquely upwards
and backwards in the dorsal half, and obliquely downwards
and backwards in the ventral half, of its extent ; and as the
myotome becomes very narrow and almost horizontal before it
disappears ventrally, it is not possible to assert with any
assurance that structures lying below the region into which
the myotomes extend are behind or in front of any given one
of these obliquely directed structures.

I am inclined to the view that the oral sphincter is morpho-
logically in front of the first myotome, though its position
coincides approximately with a vertical line drawn through the
anterior angle of the seventh. This back-pushed position of
the ventrally placed organs in relation to the myotomes of the
body wall is characteristic of Amphioxus, and is connected
with the establishment of an independent metamerism of the
alimentary canal, which, after the early larval condition, seems
to be in no definite relation to the metamerism of the body wall.

The atriopore or ventral median aperture of the peri-

370

E. EAT LANKESTEE.

pharyngeal chamber formed by the down-growth and fusion of
the epipleural or opercular folds, is so placed that it is possible,
by carefully tracing back the obliquely directed myotomes, to
arrive at a definite conclusion as to its position. It appears to
me to coincide with the 36th myotome, whilst a vertical line
drawn from the anterior angle of the 41st myotome passes
through its posterior margin.

The anus coincides with the septum separating the 51st
from the 52nd myotome, and I count ten post-aual myotomes.
The series of numbers thus arrived at may be written thus :
36 + 15 + 10 = 61. 1 Dr. Gunther gives four different enumera-
tions of the myotomes of specimens of Amphioxus lan-
ce ol at us from different localities, none of which are pre-
cisely the same as that which I am inclined to regard as
characteristic for Neapolitan specimens, viz. 35 + 12 + 12 = 59
(Polperro) ; 36+14+11 = 61 (Scandinavia) ; 34+13 + 13 = 60
(Naples); 35 + 12 + 13 = 60 (Naples).

The Number of the Fin-Rays. — Amphioxus is provided with
a continuous dorsal fin which reaches anteriorly below the
extreme terminal portion of the notochord and becomes con-
tinuous with one side of the prseoral hood, viz. the
the right. Posteriorly the fin-like expanse is continued round
the notochord and runs forward on the ventral surface along
the median line, lying, however, to the right of the anus.
Coincidently with the last twelve myotomes the fin is expanded
both in its dorsal and ventral regions so as to form a lozenge-
shaped caudal fin. It runs forward on the ventral median line
as far as the atriopore. The base of the fin is supported by a
series of fin rays which are short cylindrical pieces of a kind
of connective tissue. The dorsal fin-rays are in a single
series ;2 those on the ventral surface between anus and atrio-

1 I have also notes of countings of Naples specimens which give

35 + 14 + 13 = 02; 35 + 14 + 12 = 61 ; 35 + 16 + 10 = 61 ;

36 + 15 + 11 = 62.

2 The very first fin-ray of the dorsal series is bifid at its base, as shown in
PI. XXXYI B, fig. 11, tending to show that the ventral series are not so
peculiar in their double character as is suggested in the text,

AMPHIOXUS LANCEOLATUS.

371

pore form a double or paired series. A very peculiar fact with
regard to these fin-rays is that whilst each is connected at its
base with a strong ridge of connective tissue which forms a
continuous median plate, springing from the roof of the
skeletal neural sheath, yet on all its other faces each fin-ray
is free, lying in a lymph space. The lymph space surrounding
the fin rays is not a continuous tube but is divided into com-
partments one to each fin-ray ; and each compartment is lined
with a pavement of endothelial cells which is extended over
both the wall of the compartment and the free surface of the
fin-ray. The liquid in the compartment separating the fin-ray
from the wall of the compartment is coagulable. The nuclei
of the cells on the free surfaces of compartment and fin-ray
may be readily observed in well-stained sections. It appears
that the compartments filled with lymph are antecedent
structures to the fin-ray which eventually comes to occupy
a large part of the space, since in both the anterior and the
posterior regions of the dorsal fin the fin-rays are relatively
small and occupy but little of the lvmph-space, whilst at the
extremities of the series the fin-rays actually disappear entirely,
leaving only the lymph-holding compartment to represent the
whole structure. Anteriorly, the fin-ray lymph-space extends
as far forward in the form of a fine canal as the notochord
itself, and is divided into five or six compartments devoid of
solid rays. Posteriorly I have not ascertained its precise
termination, but there are several compartments overlying the
last six myotomes which in adult specimens are devoid of fin
rays. I think that the number of compartments both ante-
riorly and posteriorly not occupied by fin-rays is larger in
half-grown than in fully-grown specimens, and that the volume
and solidity of all the fin-ravs is greater in the more fully-
grown individuals. Anteriorly, the fin-rays do not commence
until, in proceeding from before backwards, we have passed
that region of the nerve-cord which is in relation with the olfac-
tory pit. The figure given by de Quatrefages (10) of this region,
being a careful drawing from a living specimen, shows excellently
the condition of the first few fin-ray spaces and the first rays. I

372

E. RAY LANKESTER.

have not thought it desirable toreproduce that figure in the present
memoir nor to produce one like it, but the reader is referred
to the French naturalist’s drawing as one giving valuable data.
An extremely important fact with regard to the fin-rays of
the dorsal series is that they are between four and five times
as numerous as the myotomes, and the question arises whether
they have any definite numerical relation to the myotomes.
I have counted from 250 to 260 fin-rays in an adult
Amphioxus lanceolatus with sixty-one myotomes, no
rays being developed over the last six. Supposing we exclude
the imperfectly developed anterior and posterior regions, we find
that there are very nearly 220 fin-rays for forty-five myotomes,
approximately a relation of five to one. But I am unable to
accept the view that there is any real relation between the
metamerism of the fin-rays and the metamerism of the myo-
tomes. The fact that anteriorly there are less than five fin-
rays to a myotome, viz. four, and posteriorly more than five, is
opposed to such a relationship, whilst further, the numerical
features of the paired ventral fin-rays are entirely destructive
of any theory of the kind, for we find in the ventral series (on
an average) thirty-four pairs of fin-rays to twelve myotomes.

The paired fin-rays of the ventral post-atrioporal mid-line are,
like those of the dorsal series, contained in a series of compart-
ments, which are divisions of a lymph-space. The space is
not divided into a right and left half, but is simple. This
lymph-space is continued as a contracted canal with
coagulable contents along the mid-line posterior
to the anus for the space of several myotomes. I am not
able to say precisely where it terminates. A reinvestigation
of the tail by transverse sections would at once settle this point.1
The number of the paired fin-rays varies a little. There are

1 Anteriorly the dorsal fin-ray lymph-space ends with the notochord as a
very contracted canal overlying it. It is of some importance to note that in
this extreme anterior region there is a ventral lymph-space below the noto-
chord of the same nature as that above it, but devoid of fin-rays, though
divided into compartments, six in number. (See PI. XXXYI/f, fig. 3, and de
Quatrefages.)

AMPHIOXTJS LANOEOLA.TUS.

373

fifteen myotomes between the atriopore and the anus, and the
double fin-rays become exceedingly small and terminate before
the anus is reached. In three specimens counted they were
developed in relation to the first twelve of the fifteen myo-
tomes between atriopore and anus ; in one there were thirty-
four paired rays, in the second thirty-nine, and in a fourth
forty-one.

It is not improbable that the double series of ventral fin-rays
represent a posterior continuation of the same primitive lateral
fold on each side of the body, which in the anterior two thirds
of its extent becomes sufficiently large to wrap round the ven-
tral surface of the body, and by fusion with its fellow along the
mid-line to form the atrial chamber, whilst posteriorly its line
of offgrowth has descended on either side, so as to lead to an
approximation and ultimate fusion in the mid-ventral line,
forming the double ventral fin. It is of significance in this
connection that the paired structure of the ventral fin does not
extend beyond the anus, and that the azygos fin is continued
ventrally to the right of the anus, whilst anteriorly there is
a continuity of one lateral fold (the right half of the prseoral
epipleur or hood) with the dorsal azygos fin. This continuity
would be similar to that of the azygos fin (passing to the right
of the anus) with the series of paired ventral fins, if the view
should be established that the paired fins are the conjoined
post-atrioporal extensions of the epipleural folds.

It is important in this matter to distinguish the metapleural
canals and cartilages from the epipleura upon which they
develop (see PI. XXXIV, figs. 4 and 5). The metapleura, as
shown in the figures just cited, are continued posteriorly beyond
the atriopore and beyond the first two pairs of ventral fin-rays.
This is evidence in favour of the view that the paired ventral
fin-rays are continuations of the paired epipleura, for in front
of the atriopore the area between the two metapleura is formed
by the fusion of the two epipleura, and it is a legitimate
inference that behiud the atriopore what lies between the two
metapleura is also formed by fused epipleura.

The Number of the Gonad Pouches. — The coelomic sacs in

374

E. RAT LANKESTER.

which the reproductive cells develop are twenty-six in number
on each side of the body, and correspond to twenty-six of the
myotomes.1 On account of the oblique shape of the myotomes
it is not easy to decide precisely which of the thirty-six
myotomes between the anterior snout and the atriopore are
those to which the gonad pouches correspond. I am of
opinion that the last gonad pouch corresponds to the last prse-
atrioporal myotome, and this would make the first coin-
cident with myotome No. 10, and the last with myotome
No. 35.

The important fact is that the gonads are affected by the
same metamerism as that which affects the musculature.

Occasionally specimens of Amphioxus occur in which the
anterior one or two or the posterior gonad pouches are not
developed, whilst the others are in full ripeness ; and in speci-
mens taken in the autumn the entire series of gonad pouches
are usually in an extremely rudimentary condition.

Number of the Praeoral Tentacles. — The circular group of
pinnate tentacles to which Amphioxus owes its earlier name
of Branchiostoma, presents great numerical variations. The
tentacles increase in number as the Amphioxus increases in
size. Their first appearance is not known, but I have records
of small individuals with twenty, of middle-sized with thirty,
and of large individuals with forty tentacles. I am indebted
to Dr. Hugo Eisig for kindly counting specimens of various
sizes for me at Naples.

The addition of new tentacles appears to take place at the
middle point of the ventral side of the ring-like margin of the
prseoral hood, and they are formed in pairs, right and left, the
last formed being exceedingly small. There is no median
tentacle, either dorsal or ventral.

The Number of the Pharyngeal Gill-slits. — It seems from the
descriptions given by Kowalewsky (11) that the gill-slits which
first appear in the larva are in definite relation to the

1 In some specimens I have counted twenty-seven, and in some twenty-
nine ; in others again only twenty gonad pouches on one side, whilst twenty-
six are present on the other side.

AMPHIOXUS LANOEOLATUS.

375

myotomes, but that this relation is not subsequently main-
tained.

The accounts of the late larval condition of Amphioxus are
not sufficiently satisfactory to enable us to formulate a very
definite conclusion as to this early relation of the gill-slits to
the myotomes. It is, however, quite certain that after the
larval phase all relation between the number of the myotomes
and the number of the gill-slits is lost. The gill-slits go on
increasing in number by addition at the posterior end of the
series throughout the period of growth — probably as long as
the animal lives — whilst the full number of myotomes is
acquired at a very early period, and is not subsequently
increased. Owing to this fact it is possible in any Am-
phioxus to observe the mode of formation of the gill-slits, and
it is found that they originate as oval or nearly circular per-
forations of the proper body wall, which become divided each
into two by the growth from the dorsal margin of the oval slit
of a longitudinal bar or tongue, comparable to the tongue of a
Jew’s-harp, which thus divides each primary slit or gill
aperture into two.

The tongue bars can be distinguished throughout the series
by the fact that they are supported by a hollow chitinous rod,
whilst the adjacent bars separating primary slits from one
another are solid (PI. XXXYIjB, figs. 1 and 2). Also the
primary bars are provided with a plate-like projection on their
external border which becomes deeper dorsalwards and shal-
lower ventralwards. This plate-like projection is soft-walled
and hollow, containing a space which communicates with the
“ dorsal ” or supra-pharyngeal coelom. In my earlier paper
(12) I called these soft plate-like projections the pharyngo-
pleural folds. In the more dorsal or upper part of the bars
the pharyn go-pleural folds are so deep as to rest for some
distance against the inner face of the down-grown epipleura.
In consequence of the oblique and almost horizontal position
of the bars and slits throughout the middle third of the per-
forate region of the body, and, in consequence of the adhesion
of the pharyngo-pleural folds to the epipleura, the atrial

376

E. RAY LANKESTER.

chamber is divided, for a part of its extent, into a number of
nearly horizontal passages which may be compared to the
series of parallel adherent tube-like passages connecting the
gill pouches of Myxine with the branchial pore of that animal.
When the development of Myxine can be studied, I should be
anxious to inquire whether the tube-like passages in question
are formed by the septation of a primitively simple subopercular
cavity through the outgrowth of interbranchial septa as in
Amphioxus.

The number of gill-slits, counting each of the primary slits as
two, observed by Johann Muller in a small transparent speci-
men, was 50, in individuals of an inch long 80 to 100. In indi-
viduals a little over an inch in length I have counted 96 slits,
and in larger specimens (nearly two inches long), 124. To arrive
at the number of primary slits we have to halve these figures,
since each pair of slits is formed in the way above noted.

The independence of the gill-slits in relation to the meta-
merism of the body wall is related to the following facts.
(1) The myotomes increase in volume during growth but not
in number. (2) The whole pharyngeal region of the body
increases in volume during growth, and the point at which the
perforations cease, though it remains throughout life (after a
size of three quarters of an inch has been reached) in
approximately the same relative position to the superjacent
myotomes, viz. coincident with a vertical line drawn through
the anterior angle of the myotomes 27 to 29, yet advances
gradually backwards from the former to the latter as growth
goes on. (3) The pharyngeal slits do not increase in width,
and the increase of the pharynx is made by new local growth at
its posterior end. Accordingly new slits are formed in the new-
growing region of the pharynx. It thus results that organs which
are fixed in position in relation to a particular myotome — for
instance, the “ atrio-coelomic funnels,” to be described below, of
which I have spoken in my earlier paper as “ the brown canals,”
are found to vary in their relative position to the perforated
region of the pharynx. In small specimens the atrio-coelomic
funnels are in the same plane with the non-perforated termina-

AMPHTOXUS LANCEOLATUS.

377

tion of the pharynx ; in larger specimens they are seen in
sections coincidently with a full series of bars and slits as in
Plate XXXVI.

The independent metamerism of the body wall on the one
hand, and the gill-slits on the other in Amphioxus, is a matter
of some interest in relation to the metamerism of musculo-
skeletal axis and branchial bars in craniate Vertebrata.

The Spaces Enclosed in the Body of Amphioxus. — There are
three distinct kinds of spaces containing liquid in the living
state, which are to be met with in the study of transverse sec-
tions of Amphioxus. These are : (1) the atrial cavity; (2) the
enteric cavity ; (3) hsemo-lymph cavities. The last group is
divided into several sections which are more or less distinct
from one another ; they are (a) the vascular system, which, as
shown by Schneider, is in open continuity with ( b ) the supra-
pharyngeal and perienteric portions of the coelom ; (c) the
perivascular spaces of the dorsal aortse ; ( d ) the perigonadial
coelom; (e) the right and left metapleural lymph-spaces ; (/)
the lymph-spaces of the dorsal and ventral fin-rays ; (g) the
superior and inferior intra-notochordal lymph canals; (A)
the neuraxial canal ; (t) the myoccelomic pouches or intra-
muscular lymph-spaces of the head ; (A) the series of intra-
skeletal lymph-spaces of the myotomes.

As has been mentioned above, it is extremely difficult to
arrive at a correct conclusion as to the existence of spaces
within the body of Amphioxus, owing to the distorting action
of the reagents used for hardening specimens before cutting
sections. The chief errors which have been made by previous
writers — some falling into one mistake and some into another
— are the ascription to Amphioxus of a single wide ventral sub-
epidermic lymph canal, or of a series of such canals beneath
the plaited ventral region of the branchial chamber, the denial
of the existence of natural canals in the metapleura, the over-
looking of the intra-notochordal lymph channels, and the
assertion of a canalicular communication (“ godets ” of
Moreau) between the contents of the notochordal sheath and
the space enclosed by the superjacent, neuro skeletal tube.

378

B. EAY LANKESTER.

The structure of the pharyngeal bars, and the number and
character of the spaces contained in them, as shown in
transverse section, also have been the subject of divergent
and erroneous statements.

The atrial cavity can be readily traced in sections of well-
grown specimens, owing to the fact that the epiblastic epithe-
lium by which it is lined, is loaded very often with brown
pigment granules. For the purpose of tracing the atrial
cavity, a specimen should be chosen which has the brown
pigment well developed; it is more abundant in some indi-
viduals than in others. The general limitation of the atrial
cavity as seen in a transverse section about the twenty-seventh
myotome, is shown in the diagrammatic figure given in
PI. XXXYI. Other facts with regard to the atrial cavity are
shown in the “ reconstructed ” dissection of PL XXXIV,

fig- !•

A curious fact with regard to the atrium (first described by
Rolph) is the existence of a csecal prolongation of its cavity
beyond the atriopore posteriorly. This atrial caecum pushes
its way as a tapering blind sac into the perienteric coelomic
space behind the atriopore, and occupies a position between
the intestine and the musculature of the body wall. It
reaches as far back as the anus, where it terminates blindly.
It is represented in PI. XXXIV, fig. 1, for the first time as
exposed in a simple dissection, Rolph’s and Langerhaus*
figures showing it in transverse section.

The enteric cavity of Amphioxus presents three main
regions, viz. the pharynx, the intestine, and the caecum.
Owing to the enclosure of the true original surface of a large
part of the body by the atrial or epipleural folds, a misleading
nomenclature is apt to be applied to the regions of the body
thus enclosed ; we are led to overlook the fact that the wall of
the perforated pharyngeal region, the wall of the caecal region,
and the wall of the intestinal region as far as the atriopore, are
not the proper walls of pharynx, caecum, and intestine, but in
reality epidermis-clothed somatopleur or body wall, enclosing
within it more or less complete coelomic space, and the portion of

AMPHIOXUS LANOEOLATUS.

379

alimentary tract to which the body wall so closely moulds
itself. The relation of these parts is shown in PI. XXXIV,
fig. 1, and it is clear enough that we cannot in the prse-atrio-
poral region separate the various parts of the enteric canal
from the closely adherent body wall. The csecum appears not
to be enclosed in a portion of body wall common to it and the
pharynx ; but there is actually a complete diverticulum of the
body wall covering, and fitted to, the csecum, leaving a small
coelomic space between the somatic and splanchnic elements,
as shown in PI. XXXVI.

I do not propose to enter on the present occasion into detail
with regard to the structure of the successive regions of the
pharynx, but I may point out that whilst in the anterior region
it is broad and heart shaped, in section it becomes posteriorly
greatly compressed, as shown in PI. XXXVI. This shape
appears to me to be the natural shape during life; when the
gouads are enlarged the pharynx is necessarily compressed
throughout that region where it is accompanied by the
caecum.

The numerical relation of the caecum appears to be as
follows : — It is given off as a diverticulum on the right side of
the body about the 28th or 29th myotome, and reaches as far
forward as the 15th or 14th myotome. These figures apply to
adult specimens.

The vascular system of Amphioxus appears to be in a con-
dition of degeneration, since it presents a certain limited
development of vascular trunks, which do not appear to have
a physiological significance in their present relations.

I am not in a position to give a critical account of
the vascular system, but it is necessary to draw attention
very emphatically to the continuity of the vascular trunks
and lymphatic spaces of Amphioxus and their contents,
which make it impossible to decide with certainty in all cases
whether a given space with coagulable liquid contents is to be
considered as blood-vessel or lymph-vessel. Such a communi-
cation is described by Schneider and figured by him, showing
the free connection of the veins of the csecum with the dorso-

380

E. EAY LANKESTEB.

pharyngeal coelom. Such a communication is suggested by
Langerhans in his description of the capillary network on the
caecum. I am inclined to think that there are not distinct
capillaries and coelomic space around the caecum, but that the
space is capillariform.

Some main trunks of the vascular system of Amphioxus
are obvious enough in sections. The difficulty is to make out
definitely their connections. We have (1) the cardiac or
endostylar aorta, a highly contractile vessel lying in the wide
coelomic space below the hypobranchial ridge or endostyle of
the pharynx. Anteriorly this vessel dilates into the “ heart”
of Langerhans, placed just below the sphincter oris. From
this “ heart” are given off anteriorly a right and left vessel to
the oral tentacles, and laterally a single right so-called
“ aortic arch,” a large sinuous vessel, which runs forward and
upwards in the right prseoral epipleur (right side of oral hood),
until it reaches the level of the notochord, where it joins
(according to Langerhans) the right “ dorsal aorta.” This
sinuous aortic arch has been described by Rolph as a gland,
and in fact it appears to occupy the space in which a glandular
structure is developed in the larva.

The dorsal aortse are two vessels, right and left, underlying
the notochordal sheath, and placed on either side of the
hyperbranchial groove (see PI. XXXVI). They extend through-
out the length of the perforated pharynx, but unite to form a
single “posterior aorta” at the point where the alimentary
canal narrows and becomes intestine. This single median
vessel can be traced on the dorsal surface of the intestine
as far as the anus, beyond which point it appears to be
continued as a canal in the ventral part of the sheath of
the notochord, finally ending blindly near the extremity of
that organ. Similarly in the anterior region of the body
the left aorta is continued forward in front of the mouth
as a narrow canal in the left side of the notochordal sheath,
and finally, I am inclined to think, opens into the cavity of
one of the cephalic myotomes, the cavitary structure of the
mesoblastic somites surviving from the embryonic condition

AMPHIOXUS LANOEOLATUS.

381

in this region. The right dorsal aorta is said by Langerhans
to communicate with the right aortic arch, but I doubt this.
I am not very certain on the point, but I think that it ends
blindly. Its place is taken in the praeoral region by a branch
given off from the “ aortic arch,” which runs forward in the
substance of the notochordal sheath on the right side, parallel
with the forward continuation of the left aorta, and these two
vessels undoubtedly communicate beneath the notochord by a
transverse channel. Finally, the right-side vessel, like that of
the left, appears to communicate with the coelomic cavities of
the anterior myotomes.

Schneider has described a series of lateral vessels given off
from the dorsal aortae and running into the primary and
secondary (or tongue-like) bars of the pharynx, through
which they are supposed to communicate with the cardiac
endostylar vessel. I have not been able to trace these lateral
branchial vessels in transverse sections, though I have traced
a branch from the endostylar vessels into each primary bar
(see PI. XXXVI B, figs. 4 to 9).

Upon the inner face of the epipleura below the atrial tunic
a blood-vessel has been described by W. Muller (13), running
longitudinally. The vessel is seen especially in specimens
where the gonad pouches are rudimentary and is related to
their development. It furnishes capillaries to the testes, but
the connection between it and other blood-vessels has not
been observed.

Upon the wall of the intestine and upon the wall of the
caecum there are blood-vessels. Those on the intestine are
large and more numerous in its posterior region. They gather
together anteriorly and are continued into the endostylar or
cardiac subpharyngeal trunk, where the alimentary tract en-
larges to form the pharynx. The vessels on the caecum form
a network which has been described by Langerhans. They
give rise to a network of capillaries, and together with the
capillaries described by the same author in the testes, are
the only capillaries present (so far as my own conclusions go)
in Araphioxus. The vessels of the caecum communicate,
VOL. XXIX, PART 4. NEW SER.

c c

382

E. RAY LANKESTER.

according to Schneider, with the dorso-pharyngeal coelom at
the anterior extremity of the caecum through the coelomic
spaces within the pharyngo-pleural pouches of the primary
bars of the pharynx, which rest against and open into the
blood-holding cavity which surrounds the coelom. I can
confirm this observation from the study of transverse sections
made by my pupil, Mr. Willey.

The question as to how the blood which is brought by
veins into the cardiac aorta or great contractile blood-channel
underlying the hypopharyngeal ridge “ circulates,” or whether,
indeed, it circulates at all, has not been, in my judgment,
satisfactorily answered, and renewed investigations are needed.
This is in part due to the difficulty of investigating the structure
of the pharyngeal bars and of arriving at a certain conclusion
as to what are real natural spaces and channels and what are
artifacts.

The structure of the pharyngeal bars is shown in PI.
XXXVI B, figs. 1 and 2, which represent sections at right
angles to the length of the bars. As is well known, from the
observations of Muller and others, the bars are not all similar,
but of two kinds, viz. (1) those which correspond to the
division between primary gill-slits, the “ primary bars,” and
(2) those which form by a growth downwards from the dorsal
margin of a primary slit, dividing it into two secondary slits.
These are in relation like the tongue of a Jew’s-harp, and may
be called “ tongue-bars.” The development of these bars may
be seen in any Amphioxus continually in progress in the
posterior region of the pharynx. The chitin-like material
which forms the skeleton of the pharynx is deposited in the
form of rod-like tracts beneath the epithelium (in the cutis-
layer) boundiug the margin of the gill-slits. Accordingly
there is a double rod in each primary bar, one half corre-
sponding to each of the adjacent gill-slits. At each end of
the gill-slit this double rod bifurcates, and each half of the
fork runs parallel with the arch-like boundary of the gill-slit,
tending to meet the furcal half of the next double rod at the
summit of the arch. On the other hand, the rod of the tongue-

AMPHIOXUS LANCEOLATUS.

383

bar is not of a bifid or double character, but is a single hollow
rod, which is continued directly from the mid-point of the
upper chitinous arch. It does not at the lower end of the
primary gill-slit come into contact with the chitinous lower
arch, but simply joins the endostyle or median inferior area of
the pharynx, the chitinous material ceasing at the point of
junction. The general arrangement of the bars is shown
in Muller’s original plates, and is so well known that I have
not thought it necessary to figure it here.

The primary bars and the tongue-bars differ in other respects
besides the fact that the rod of the primary bar is essentially
bifid and that of the tongue-bar a hollow single rod. A.
Schneider and Langerhans have described the structure of
these bars as seen in transverse sections ; but I think that the
former has erred in assigning too many vascular passages to
the bars, whilst the latter has assigned too few. My own
conclusions are exhibited in the drawings given in PI.
XXXYI B, figs. 1 and 2.

Both bars are flattened like a lath, and are set with the nar-
rowest diameter parallel to the long axis of the Ampliioxus.
The atrial surface of the bars is clothed with the atrial epi-
thelium ( atr . ep.), the cells of which are especially deep and
large, whilst the brown pigment is limited to a strongly-
marked group of cells on each side ( pig .). The inner face of
the bars — that turned towards the lumen of the pharynx — is
provided with a peculiar epithelium arranged in three rows
( al ., am., ar.), the cells of which are very narrow and long, with
elongate deeply-staining nuclei. These cells resemble those
found in group al. and ar. of the endostylar epithelium (see
fig. 9, PI. XXXYI B), and like them carry short cilia.
The adjacent sides of the bars bounding the passage between
neighbouring bars are lined with columnar cells, which carry
very long cilia (col.). Below the outlines of these columnar
cells an immense number of closely aggregated nuclei, which
stain strongly with either hsematoxylin or carmine, are ob-
served. The superficial series of these (n.) probably belong to
the columnar cells. Whether the deeper nuclei (n'.) are all to

384

E. KAY LANKESTEK.

be reckoned to epithelium seems doubtful. A clear median
space or line exists (sept.) which must consist of a connective
tissue, and the deepest nuclei would in all probability be
referable to that tissue (and therefore to mesoblast).

The chitin-like rod lies near the atrial border of the primary
bar (PI. XXXVI .6, fig. 1 Rod), and similarly in the tongue-
bar (fig. 2 Rod). In both the rod is grooved on its inner
(pharyngeal) face, so as to form a small channel, which is
probably occupied by a blood-vessel marked Bl. vess. in the
figures. This is the only space which I can find in the
transverse section of the bars, excepting the larger space
marked coelom in the figures, and the fissure more or less
complete of the double rod of the primary bars ( fiss . in the
figures), and an occasional (by no means constant) minute
defect in the rod of the tongue-bar (sc. in the figures). The
blood-vessels which are given off right and left from the great
artery of the endostyle (see figs. 4 to 9, PI. XXXVI i?) pass
into the bases of the primary bars, where their rods bifurcate,
and are possibly and probably continued up the primary bars
in the channel marked Bl. vess. in fig. 1. It is, however, to
be noted that this channel is very narrow relatively to the
vessels given off from the median ventral artery of the
endostyle, and that the tongue-bars certainly receive no such
branches from the endostylar artery, although the channel
exists in them also. Schneider figures a vessel passing from
the dorsal end of each bar — both tongue-bars and primary
bars — into the dorsal aorta ; and possibly a communication
exists between the vessel of each primary bar and that of the
adjacent tongue-bars by means of the transverse junctions
which occur at intervals along the length of the bars. I have
not been able to satisfy myself as to the existence of the
communications with the dorsal aorta described by Schneider,
nor as to the existence of vessels in the transverse junctions.
At the same time it seems very probable that both exist, and
a little further investigation may enable us to recognise them
in sections.

Between the chitinous rod and the atrial epithelium of the

AMPHIOXUS LANCEOLATUS.

385

primary bar is a large space lined by an epithelium ( coel . ep.).
This is the coelom, and is in free communication dorsally with
the pharyngo-dorsal coelom, and ventrally with the coelom of
the endostyle. The space becomes deeper and its walls longer
as we ascend the primary bar, until it opens as a narrow but
greatly extended space into the pharyngo-dorsal coelom. It is
the raised-up walls of this space which form the pharyngo-
pleural folds or coelomic pouches of the primary bars (see PI.
XXXVI).

In the tongue-bars there seems at first sight to be nothing
which corresponds to the great coelomic channel of the primary
bars. But when we trace the connection of the tongue-bars
with the endostylar tract by means of transverse sections, we
find that the canal within the rod which distinguishes the rod
of the tongue-bar from the rod of the primary bar is in free
communication with the endostylar coelom (figs. 6, 7, 8, PI.
XXXVI B ). The canal within the hollow rod of the tongue-
bar probably opens dorsally into the pharyngo-dorsal coelom,
although the proof of this by means of transverse sections
remains to be obtained by future inquiry.

Variations in the amount and position of the chitinous
deposit forming the rods of the pharyngeal bars are frequently
found ; some of these are drawn in fig. 3, PI. XXXVI B. The
most noticeable is the tendency to form a complete chitinous
deposit embracing tbe supposed blood-vessels ( Bl. vess. of figs.
1, 2, 3), and this may be either fused with the chitinous rod or
detached from it as a separate piece. The bifid character of
the rod of the primary bars is more obvious towards its extre-
mities where it bifurcates (fig. 3, h ).

A comparison of the structure of the gill bars of Amphioxus
with the gill-filaments of the Lamellibranchs is instructive,
and the latter may throw some light on the former. It can
scarcely be maintained that the disposition of the blood-vessels
in Amphioxus lends itself to the conclusion that we have here
a highly efficient branchial respiratory apparatus. When the
existence of extensive communications between the large
coelomic spaces of Amphioxus and its blood-vessels are borne

386

E. RAY LANKESTER.

in mind, it becomes probable that the branchial apparatus as
we see it is modified as compared with an earlier condition in
which the blood-vessels played a more prominent part, and
were more largely and distinctly developed throughout the
organism. The probably degenerate condition of the vascular
system in Amphioxus has led me to doubt whether the spaces
marked Bl. vess. in figs. 1 and 2, PI. XXXVI B, are really
continued as distinct vessels to the dorsal aortse; it is not
unlikely that such a continuation exists, but it is also not
unlikely that the original branchial vessels have effected a
communication with the coelom. The generalisation that a
fragmentary vascular system is not in a primitive condition
but is in a state of degeneration appears to be warranted by a
survey of vascular systems in the animal series, and by the
a priori argument that a vascular system must be efficient as
a circulating and distributing apparatus in order to afford the
advantage necessary for the operation of natural selection.
The probable steps of the primary or ascending evolution of
a vascular system do not include a condition in which large
vessels are present without capillaries or are in free com-
munication with the coelom.

From the examination of the pharyngeal bars we may now
proceed to that of the median ventral tract of the pharynx,
which it is convenient to call in toto the “ endostyle,” the
name being justified by the undeniable identity of the peculiar
median ridge of epithelium with that which is recognised by
this name in the Ascidians.

Below the endostylar epithelium, as shown in the series of
sections, figs. 4 — 9, PI. XXXVIiJ, there is a chitinous plate
which has hitherto remained undescribed. It consists of
right and left moieties, and is segmented; that is to say, it
thins out and disappears for a brief space at intervals. This
endostylar skeleton, in fact, consists of a number of pieces
following one another, corresponding in number to the primary
gill-slits, each piece being composed of a loosely-joined over-
lapping right and left half. The endostylar skeletal plates
yest on the ends of the chitinous arches formed by the union

AMPHIOXUS LANCEOLATUS.

387

of the adjacent anterior and posterior halves of the furcal
extremities of the rods of the primary gill-bars.

Whilst the furcal ends of the rods of the primary bars of
the pharynx penetrate thus deeply below the endostyle, the
rods of the tongue-bars are shown, by the drawings referred to,
only to reach the margin of the endostylar tract. A large
coelomic space exists beneath the endostylar chitinous plates,
and around the furcal ends of the primary rods. This space
is seen by following the sections to communicate freely with
two structures of the pharyngeal bars, viz. (a) with the soft-
walled pharyngo-pleural fold of the primary bars, and (6)
with the cavity of the hollow chitinous rod of the tongue-bars.

The contractile endostylar artery or cardiac aorta is seen in
the sections either in the middle line or a little to the right or
to the left. A large branch is given off from it to each
primary bar, but the sections have not enabled me to trace
the vessel actually into the bar or along its length. No
vessel is given off to the tongue-bars.

I have not observed in sections of the endostylar region the
muscular tissue which Schneider has described as existing
there, and I doubt the correctness of his observation.

The structure of the deep part of the rods of the primary
bars, where their bifurcate extremities lie below the chitinous
plates of the endostyle, is remarkable. The substance of the
rods consists of a reticular tissue with scattered nuclei, and the
chitinous matter appears to be superficially deposited around
this axis (see figs. 4 — 9, Plate XXXVI jB). It is necessary to
bear in mind that in speaking of the rods of the pharyngeal
skeleton as “ chitinous, ” one is using that term without strict
justification, in order to indicate not the specific chemical sub-
stance “ chitin,” but a certain density and horn-like character in
a structureless skeletal deposit. The “ chitinoid” substance of
the pharyngeal bars and of the endostylar plates of Amphioxus
appears to be a special form of the subepidermic lamina of the
connective tissue, which is seen everywhere affording firm
support to the columnar cells of the body-surface. It is to be
regarded as a product of the connective tissue, and it is there-.

388

E. RAY LANKESTER.

fore intelligible that the furcal ends of the chitinoid rods
should gradually pass over into a gelatinous reticular form of
connective tissue.

The Supra-pharyngeal Coelom and its Perienteric Extension. —
I have but little to say in regard to this space. It is suffi-
ciently obvious in sections, and contains a coagulable fluid.
It is continued down the plaits or folds of the primary pharyn-
geal bars, and communicates through them with the coelomic
space surrounding the branchial aorta, viz. the endostylar
coelom. Anteriorly it ends blindly, acquiring a considerable
lateral and at last a ventral extension along the inner walls of
the epipleura, in the praegenital region of the body. Pos
teriorly, in the region where the perforations of the pharynx
cease, it forms a narrow space surrounding the intestine, and
in the post-atrioporal region expands to a much increased
proportionate volume. It ceases at the anus, and similarly it
is not traceable anteriorly beyond the sphincter oris. Its
relations are seen in the drawing (PI. XXXIV, fig. 1).

The perivascular space of the dorsal aortae has been alluded
to above in connection with those vessels. I will merely say
again' that they appear to me to be real spaces, and not arti-
facts, and that I have not traced any opening into them.
They unite when the aortae unite and form a single space.

The perigonadial coelom is, according to the observations and
speculations of Kowalewsky, Rolph, and Hatschek, a detached
downward continuation of the pharyn go-dorsal coelom, carried
downwards with the down- growing epipleura, and subse-
quently shut off from the pharyngo-dorsal coelom above. In
such a section as that given in PI. XXXVI we can see that a
slight horizontal splitting of the connective tissue would place
the two spaces in communication.

The Metapleural Lymph-Spaces. — These and the lymph-boxes
or spaces of the fin-rays appear to be traceable to the original
myocoelomic pouches.

The metapleural lymph-canals are large, well-developed
spaces, containing coagulable lymph. Their existence has been
denied by some observers in consequence of the action upon

AMPHIOXUS LANCEOLATUS.

389

them of absolute alcohol, which shrinks up the metapleura and
obliterates the space. Johann Muller thought that they opened
anteriorly, each by a pore, but it is admitted now that no such
pores exist.

The most important fact about the metapleura which has
been hitherto overlooked is that their space is abolished, and
their very existence as upstanding longitudinal ridges ceases
when the gonads attain their full size at the breeding season.
The stretching of the epipleural wall leads to a complete flat-
tening of the metapleura, as shown in PI. XXXV, fig. 4. It
seems not improbable that the albuminous fluid contained in
the metapleural canals may serve as a final supply of nutriment
for the enlarging gonads.

Were there any “ ventral canals ” such as have been sup-
posed to exist by nearly all writers on Amphioxus, this would
be the place, viz. in association with the metapleural canals, in
which to discuss them. Stieda’s specimens and figures showed
the whole of the epithelium of the plaited ventral wall of the
atrium “ blistered ” or raised from the subjacent connective
tissue. Accordingly he described the existence of a pair of
large ventral canals lying right and left between the two meta-
pleura. Eolph, Langerhans, and Schneider, described not
a single pair of canals but a number running parallel to
and corresponding with the longitudinal ridges of the surface.
The spaces which are frequently seen in this position are
really between the connective tissue and the epidermis and
are due to differential shrinking. Eolph indicates canals
below the layer of connective tissue in this position, that
is, between the ventral transverse muscle and the connective
tissue. It appears to me that no such canals exist. The
insertion of the fibres of the transverse muscle into the con-
nective tissue, and the excessive contraction of the muscle
under the influence of reagents, causes a deep plaiting of the
connective tissue and a tearing and separation of natural
adhesions in most specimens. But in such a preparation as
that drawn in (PI. XXXVI A, fig. 2,) we see that there is no
splitting of the connective tissue in the median ventral area

390

E. RAY LANKESTER.

corresponding to the splitting which forms the lymph-space of
the metapleura. The artifact canals which have been mistaken
for natural ventral canals (Bauchcanale) lie in one of two
situations either of which is impossible for a natural lvmph-
space, viz. between the connective tissue and the epidermis or
between the connective tissue and the muscular fibres.

The Lymph Spaces of the Dorsal and Ventral Fin-Rays. —
I have already spoken of these above in treating of the fin-
rays. Hatschek (f Anatom. Anzeiger,’ August 15th, 1888),
has shown that they are originally in continuity with the myo-
coelomic pouches (see Plate XXXVI A, figs. 6 and 7).
Rolph, Langerhans, and Schneider have recognised and
described the character of the “ fin-ray boxes” or lymph-
space compartments and their epithelial lining. Fine canals
passing from these spaces have been described and are
noted by Schneider. Such fine spaces and irregular canals
are to be seen in the thick connective tissue which forms the
substance of the fin-membrane (as distinguished from the
fin-rays) of the caudal fin and praeoral lobe. Schneider states
that he has not seen a coagulum in the “ fin-ray boxes,” but
such a coagulum occurs not unfrequently.

The Intra-notochordal Lymph Canals. — In sections of Am-
phioxus may be observed dorsally and ventrally within the
notochordal sheath a clear space, the natural shape and extent
of which appears to be that given in PI. XXXVI. The dense
laminar structure of the notochord is here deficient and replaced
by short intercrossing fibres. Adequate staining with hsema-
toxvlin reveals a number of small nuclei in the neighbourhood
of these spaces in connection with the fibres. Other nuclei
are seen in a series on either side in the lower half of the
notochord, bnt nuclei do not occur deeply nor generally within
the notochord (see PI. XXXVI A, fig. 1). The dorsal and
ventral space thus seen in sections are due to the existence
of a dorsal and ventral lymph-holding space which have not
sharply defined walls, but are bounded by loose fibres. The
more dorsal of these canals was first observed by Kossmann
(14), and the somewhat smaller ventral one by Camille Moreaq

AMPHIOXUS LANCEOLATUS.

391

(15). It is not my immediate purpose to discuss the histology
of the notochord of Amphioxus, but to determine the existence
of natural spaces within that animal which have to be dis-
tinguished from artifact spaces. The two intrachordal canals
are connected with the most violent distortions of the shape
of the notochord under the influence of reagents. The greater
or less rapidity with which osmotic currents are established
and the alternative distension or shrinking of the canals leads
to such alterations in the shape of the notochord as those
shown in outline in PI. XXXVI j3, fig. 10, a, b , c, d.

The sudden and powerful contraction of the muscles
attached to the connective-tissue sheath of the notochord, and
to the connective-tissue septa passing from it, also helps in the
distortion of the notochord. The extent of the distortion
caused by the contraction of the muscular fibres of the myo-
tomes, may be judged of by the large spaces which are fre-
quently left where they have torn themselves away from the
connective tissue. The undoubtedly artifact spaces thus
produced must be distinguished from the remarkable spaces
between myotomes and notochordal sheath, and again between
myotomes and neural skeleton, which have been described by
Schneider, and are related to the roots of the anterior and
posterior spinal nerves.

In view of the undeniable distortions of the notochord
which the muscular strains and the distension or shrinking of
the intrachordal canals must produce, I feel great hesitation
in admitting as natural structures the remarkable apparent
perforations of the sheath of the notochord, found dorsally on
either side of the dorsal intrachordal canal at regular inter-
vals ; and according to Moreau, who first described them under
the name “ godets,” placing the dorsal intrachordal canal in
communication with the neural canal, within which the nerve-
cord is contained.

I am inclined to consider the “ godets ” of Moreau as
naturally existing tubercles of the notochordal tissue, as shown
in PI. XXXVI A, fig. If. But it appears to me that they do
not completely perforate the sheath of the notochord, nor

392

E. EAY LANKESTER.

penetrate into the neural canal. They appear to be segmentally
arranged in pairs at regular intervals, as described by Rolph,
Langerhans, and Schneider ; and for the present their morpho-
logical and physiological significance is altogether unknown.

The Neuraxial Canal. — The central canal of the myelon
of Amphioxus must necessarily be cited in an enumeration
of the spaces within the body of that animal. Expanding
to the form of an oval cavity in the anterior region of the
myelon, which may justly be called the brain, the canal is
extremely small and narrow throughout the rest of the cord.
It does not become the seat of any distorting action in pre-
served specimens, and therefore no more need be said of it here.

It is worthy of remark that a perineural lymph-space,1 which
in some Vertebrates is largely developed between the myelon
and the neural skeletal sheath or spinal canal, seems to have
no existence in Amphioxus. The connective-tissue tube or
canal which forms the skeletal protection of the myelon in
this animal seems to adhere closely to the nervous tissue, and
it is rare to find, even under the influence of the most violent
action affecting other parts of its structure, a separation of the
skeletal sheath and the contained nervous tissue. In rare cases I
have observed such a dislocation, as also a case in which the
true notochordal tissue was also displaced or shrunk from its
investing connective-tissue sheath (in fig. 1, PI. XXXVI A).

The Intra- skeletal Lymph-spaces of the Myotomes and
Myoccelomic Pouches of the Head. — Professor Hatscliek, in
the ‘ Anatom. Anzeiger,’ August 15th, 1888, has published an
extremely valuable though brief account of his observations on
the development of the myotomes and skeletal tissue of Am-
phioxus. Two of Professor Hatschek’s figures are reproduced
in PI. XXXVI A, figs. 6 and 7. The division of the primary
segmental coelomic pouches each into a dorsal portion (proto-
vertebra or Urwirbel) enclosing the “ myocoel,” and a
ventral portion (lateral plate or Seitenplatte) enclosing the
“ splanchnocoel,” is described.

1 In front of the termination of the nerve-cord there is a small space within
the neural sheath filled with coagulable liquid (PI. XXXVI A, fig. 3 b).

AMPHIOXUS LANCEOLATUS.

393

It is pointed out that the dorsal pouches enclosing each a
“ myoccel55 undergo the following modification : — The parietal
Avail (subjacent to the epidermis) becomes cutis, and is called,
therefore, the “ cutis-layer,5’ whilst epithelial cells of the
mediad wall (adjacent to the notochord) become individually
elongated and converted into muscle-cells. The cavity of the
myoccel remains for a time, and the muscular tissue is a truly
epithelial tissue.

Professor Hatschek does not deal with the adult animal. It
may therefore be stated that the myoccel cavities totally dis-
appear in the full-grown Amphioxus, excepting at the extremi-
ties of the body. Some of the spaces, which are invariably to
be found in transverse sections of Amphioxus between the
connective tissue and the muscular masses of the myotoms, are
artifact, and due to the contraction of the muscular fibre.
These spaces can be distinguished from the natural intra-
skeletal lymph-spaces of the myotomes (first described and
figured by Schneider) by the fact that they are not limited by
connective-tissue epithelium.

Professor Hatschek states that after the myoccel is formed a
secondary pouch forms by a folding inwards and upwards of
the ventral wall of the myocoel, which, as development ad-
vances, makes its way as a double fold between the notochord
and the muscle-layer, Avhilst at the same time the pouch grows
downwards between the lateral plate (the cavity of which is
the splanchnocoel) and the epidermis. The cells of this offset
of the myoccel pouch give rise to the skeletal tissue, Avhich
invests the notochord and the nerve-cord, as well as the fin-
rays, the cavities of which are part of the myocoel, whilst it
also furnishes the fascia to the muscle-fibres. The arrange-
ment is explained by the two diagrams in PI. XXXVI A,
figs. 6 and 7, copied from Hatschek5s paper. In the adult
Amphioxus the space between muscle fascia and notochord
sheath persists as a series of large lymph-holding spaces in
connection with the roots of the spinal nerves (see PI. XXXVI
and PI. XXXVI A, fig. 1). The walls of the space become
adherent in parts, but leave considerable regions as cavities

394

E. RAY LANKESTER.

filled with blood-lymph (Blut-haltende Raum of Schneider).
Their exact shape and extent in each myotome requires further
careful study by means of a series of sections, since they are
liable to distortion by osmotic action.

In the first three or four myotomes, which are traversed in
transverse sections in the neighbourhood of the eyespot and
olfactory pit, it appears to me that the myocoel cavities are
permanently preserved, and that the spaces as seen in
PI. XXXVI A, figs. 3, 4, 5, are not artifact but natural. The
myocoel has, in fact, never been obliterated by the adhesion of
its opposite walls.

The Atrio-ccelomic Funnels or Brown Canals. — These struc-
tures, which I discovered and described in 1874, are a pair of
short wide funnels placed in the 27th myotome, right and left,
at that region where the pharynx narrows to form intestine.
The wider end of the funnel is open to the atrium, the narrower
end is within the dorso-pharyngeal coelom, and the axis of the
funnel is parallel with the long axis of the body (see PI. XXXV,
fig. 1). It is difficult to decide whether the narrow end is
actually perforate, but I am inclined to think that it is.
The funnels are lined internally by the pigmented epithelium
which characterises the atrial wall. Each funnel adheres by
one side to the roof of the dorso-pharyngeal coelom, as shown
in the transverse section, PI. XXXV, fig. 2. The wall of the
funnel is formed by a firm connective tissue with nuclei in
addition to the lining layer of pigmented epithelium. The
funnels always exhibit longitudinal folds as though they were
capable of dilatation.

It depends on the size of the Amphioxus whether these
funnels are met with in sections with many bars to the pharynx,
or in sections where the bars and slits are few in number and
extent, and the pharynx reduced in volume. Probably in quite
young Amphioxus the atrio-ccelomic funnels do not occur in the
same vertical plane as any of the gill-slits, but as growth goes
on the pharynx extends farther back, carrying with it the wide
mouths of the funnels, the pointed extremities of which remain
in the 27th myotome. Thus in a full-grown specimen a

AHPHIOXTJS LANCEOLATUS.

395

vertical plane passing through the narrower part of the atrio-
ccelomic funnels will also cut the deepest and most fully deve-
loped region of the perforations of the pharynx as shown in
PI. XXXVI.

I am not able to offer any suggestion as to the function
of the atrio-coelomic funnels, based on positive characters.
Their inner cell-lining appears not to be glandular, and their
connective-tissue tunicis equally devoid of any special characters.
It is possible that they may serve either to admit water to the
coelom or to remove the coelomic liquid under conditions of
tension. The structure and position of one of these funnels
render it probable that were there greater tension of liquid in
the coelom than in the atrium its walls would be pressed together
and the funnel closed. On the other hand, were there greater
tension of the sea-water contained in the atrium than of the
coelomic fluid, the funnel would be dilated and sea-water would
flow into the coelom until the tension was equalised.

Morphologically, the atrio-coelomic funnels are paired short
tubes placing the coelom in continuity with the exterior, for
the atrial cavity is morphologically external surface. In so
far they correspond with the abdominal pores of craniate Verte-
brata. Mr. Bateson (2) has shown that they have a remarkable
correspondence in other respects to the collar-pores of Balano-
glossus. The collar of Balanoglossus is, like the epipleura of
Amphioxus, an outgrowth of the body wall. It may be com-
pared to a dice-box open at each end and fused with the body
wall of the Balanoglossus (over the head of which it has been
pushed) all round the inner surface of its constricted middle
third. It is thus in fact not one collar but two, one having its
free circular margin directed forward and the other having its
free circular margin directed backwards. The anterior collar
surrounds and conceals the mouth and the base of the pro-
boscis ; the posterior collar surrounds the commencement of
the pharyngeal perforated region and overhangs two or three
gill- slits. It is on the under surface of this posterior collar
that the pair of collar-pores are placed.

The epipleura of Amphioxus, like the double collar of

396

E. RAY LANKESTER.

Balanoglossus, are fused to the body wall immediately behind
the mouth. In front of this region they project as an almost
complete collai’, the so-called prseoral hood ; behind it they
are not short and annular in direction, as in the posterior
collar of Balanoglossus, but are extended horizontally so as to
enclose the whole perforate region, and their free margins fuse
together below the ventral wall of the pharynx. Still we can
easily imagine a reduction of the epipleural folds of Amphi-
oxus which would give us them in the form of an incompletely
annular fold, overhanging only the three or four anterior gill-
slits. Now, if we consider the position of the atrio-ccelomic
funnels, we find that they are in the base of the epipleural
folds, and therefore, with the reduction and shrinking of the
epipleura, would come to lie very much in the position occupied
by the collar-pores of Balanoglossus.

Whether the atrio-coelomic funnels of Amphioxus, the
collar-pores of Balanoglossus, and the abdominal pores of
Craniata are to be considered as modified nephridia, is a
question upon which I am not prepared to enter. Our con-
ception of the nephridium as a unit of structure common to all
Coelomata, is at the present moment undergoing development
and extension. But whilst we now refer to this category the
genital ducts of Arthropoda and Mollusca, as well as glandular
tubes with excretory functions, and whilst our notions as to
the limitation of the number of nephridia in one individual or
one segment are greatly modified, we must be careful not to
assume too hastily that every opening in the body wall of a
coelomate animal communicating with its coelom, is neces-
sarily the opening of a nephridium. It is not impossible
that so wide a generalisation as this may be established, but
in the meanwhile it seems possible to distinguish such aper-
tures as the dorsal pores of Lumbricus from nephridial
openings, and so long as the former are not shown to be
related by origin to nephridia, it will be necessary to admit
the existence of a category of pores which have not, and never
had, any relation to “ a specialised tubular portion of coelom,
the lining cells of which have an excretory function.” The

AMPHIOXUS LANOEOLATUS.

397

words in inverted commas constitute the definition of a primi-
tive or typical nephridium if we add to them these additional
words, “ the tube opening at one end on the surface of the
body, at the other into the general coelom.” We recognise a
variety of modifications of this primitive structure, more
especially the loss of one or both of the openings of the tube,
and the cessation of excretory glandular activity on the part
of its lining cells. We can even admit the dwindling of the
tube and its total disappearance, with survival of the external
aperture only. But in all these modifications we start with
the conception of a tubular modification of part of the coelom,
open to the exterior, and renal in the function of its lining
cells. A pore which has had no such antecedent history is
not nephridial, nor to be classed with nephridial structures.

It appears to me that we have not at present any grounds
for assigning the atrio-coelomic funnels of Amphioxus to either
category. It is not improbable that the developmental
history of the later stages of the Amphioxus-larva will furnish
the necesary data.

The Connective Tissue. — I am desirous of saying a few words
about the connective tissue of Amphioxus before concluding
these observations.

In Plate XXXVI I have, for diagrammatic purposes, coloured
the connective tissue and the gonad-cells of a uniform purple
tint. Nevertheless we can readily distinguish in Amphioxus
varieties of the connective tissue differing from one another in
density and massiveness. The varieties pass over into one
another at several points. One of the most important state-
ments that can be made about all of them is that, like the other
tissues of Amphioxus, they differ very greatly from the cor-
respondingly placed tissues in other Vertebrates, and do not
closely resemble those of any other animal. I am not
acquainted with any chemical examination of either the con-
nective tissues or the notochord of Amphioxus.

The structural varieties presented by the connective tissue of
Amphioxus may be enumerated as (1) the lamellar; (2) the
gelatinous ; and (3) the cartilaginoid. When we examine

VOL. XXIX, PART 4. NEW SER.

n n

398

E. RAY LANKESTER.

well-stained specimens of Amphioxus we find that the nuclei of
the cells which give rise to these varieties of skeletal substance
are always arranged in simple layers, in fact are epithelial in
character, and, as Hatschek (16) has recently shown, are in fact
the epithelia bounding the primitive myocoel pouches or their
outgrowths, as explained above (see PI. XXXYI A, figs. 6, 7,
copied from Hatschek).

Beneath the epidermis we find a dense lamina supporting the
epithelial cells ; beneath this, again, a softer, less dense gela-
tinous substance, and more deeply a second very delicate lamina,
on which we find the connective-tissue cells. These four
laminae constitute the cutis. The cells of the deepest layer
are the only cells of the connective tissue (see PI. XXXYI A,
fig. 1), and must be regarded as the matrix-cells of the various
layers of skeletal substance superficial to them.

In the expanded regions of the median fin at the two
extremities of the body the substance of the fin is formed by
the gelatinous tissue, which is excavated by small irregular
canals and spaces clothed with the epithelial connective-tissue
cells. Thus a cartilaginoid tissue is produced, no longer a
plane lamelliform deposit, but a tissue which increases in three
dimensions (see Pouchet, this Journal, vol. xx, p. 421).

The thickened mass of cutis bounding the outer wall of the
metapleural canal is formed by an increase in the gelatinous
layer, which not only is thickened but contains numerous
fibrillse.

The fin-rays consist of a fibro-gelatinous substance, which is
invested by an epithelial layer. The fin-ray boxes or compart-
ments are, according to the important observations of Hatschek,
survivals of the myoccelomic pouches, and are, at one time (as
they are permanently at the extremities of the series) simple
cavities lined with the myoccelomic epithelium. In the floor
of the cavity beneath the epithelium the fibro-gelatinous fin-
ray is formed, and gradually grows up into the compartment
clothed with the epithelium. There are no canalicular spaces
in the fin-ray, and no cells sunk in its substance.

The skeletal tissue which surrounds the notochord and

AMPHIOXUS LANCEOLATUS.

399

forms the intermuscular septa and neural sheath, is, like all the
rest, of epithelial origin, according to Hatschek. A special
diverticulum of the myocoelomic pouch pushes its way, at an
early stage of development, between the muscle and the noto-
chord, and also descends between the epidermis and the lateral
plate (see PI. XXXVI ^4, figs. 6, 7). This outgrowth is called
by Hatschek the sklerotome.

It is indicated in the adult by a layer of connective-tissue
cells of epithelial character, which surround the notochord, and
a second layer, which rests on the surface of the muscular
mass facing the notochord. Between the two was originally
a space continuous with the myocoelomic pouch. This space
is obliterated in part, but in part persists as the intra-
skeletal or myoskeletal lymph-spaces of the myotomes. The
disposition of these layers of cells will be best understood
by a comparison of Hatscbek’s figures and tbe drawing of a
well-stained section of an adult Ampbioxus (PI. XXXVI A ,
fig. 1).

The sheath of the notochord deposited between the epithelial
cells and the proper notochordal tissue is similar to the cutis in
structure. Next to the notochordal tissue forming the inner-
most layer of the sheath is a dense, highly-refringent substance,
which gradually passes over into a more distinctly laminate
layer (corresponding to the gelatinous layer of the cutis), and
then follow the cells. In some preparations the staining of
these two layers is very different ; in others they are not dis-
tinguishable from one another.

The dense innermost layer has been described by some
observers as the cuticle of the notochord, just as the dense
lamina below the epidermis is regarded as being formed by the
epidermis cells. I am, on the contrary, inclined to reckon both
to the connective tissue, and do not find in the adult Am-
phioxus any recognisable and distinct notochordal cuticle,
though such may exist in the embryo.

The lamination of the connective tissue, both of cutis and
notochord, will be best understood by the examination of
figures 1 and 2 in PI. XXXVI A , as interpreted by the

400

E. BAY LANKESTER.

important diagrams of Hatschek, which are reproduced side by
side with them.

A curiously modified tract of connective substance is to be
observed forming that part of the notochordal sheath which is
attached to the dorsal wall of the pharynx (PI. XXXVI .<4,
fig. 1 x.). It has an irregular granular appearance, quite dis-
tinct from that of the connective substance in any other
region, excepting a similar tract on the upper median surface
of the notochordal sheath ( y . in same figure).

The reticular tissue with nuclei which forms the axis of the
furcal portions of the primary bars in the endostylar region of
the pharynx seems to be distinct in character from all the other
skeletal tissues of Amphioxus (PI. XXXVI B, figs. 4 to 9).

The Notochordal Tissue.— The series of vertical laminae
which build up the notochord of Amphioxus have often been
described and figured. I desire here merely to draw attention
to the disposition of nuclei within the notochord as shown in
well stained preparations. There are no nuclei in the position
described by Moreau (15) towards the axis of the notochord ;
the nuclei are confined to two perfectly definite regions. In a
transverse section a series is seen lying in a single row near
the circumference of the notochord, and extending along the
inferior third of its area. A second smaller group of nuclei is
seen dorsally on either side of the dorsal intrachordal canal.
I have already referred to the superior and inferior intra-
chordal canals. I may again state here that the notochordal
tissue does not appear to me to form itself any cuticle or
investing sheath. Such a cuticle may exist in the embryonic
condition before the connective-tissue sheath is developed, but
it would be difficult to attribute any part in the formation of
the adult notochordal sheath to a delicate envelope of the kind,
owing to the enormous increase in the bulk of the notochord.

Summary. — The present memoir by no means professes to be
a monographic treatment of Amphioxus nor even to deal
exhaustively with parts of the structure of that animal. It
must be regarded as a contribution to the knowledge of
Amphioxus, detailing a few new facts, offering evidence towards

AMPHIOXUS LANCEOLATUS.

401

the decision of some doubtful questions, and above all pointing
out a number of matters in which further observation is
needed in order to clear up uncertainty. The chief points
brought to notice are —

1. The number of the myotomes, of the dorsal fin rays, of
the ventral fin-rays, of the prseoral cirrhi.

2. The size and importance of the post-oral tentacles, or
tentacles of the sphincter oris.

3. The non-existence of the so-called “ ventral canals ”
beneath the plaited ventral wall of the atrium.

4. The actual existence of the metapleura and metapleural
lymph-canals, but their obliteration during complete distension
of the atrium.

5. The actual existence as natural spaces of (a) the fin-ray
compartments, ( b ) dorsal and ventral intrachordal canals, (c)
the intraskeletal lymph-spaces of the myotomes.

6. The structure of the gill-bars and endostyle.

7. The position and form of the atrio-ccelomic funnels or
brown-canals, now figured for the first time.

8. The general form of the body as corrected from distor-
tion by reagents, and the disposition of parts as shown in a
reconstructed dissection.

List of Memoirs referred to.

1. Muller, Joiiann. — “ Ueber den Bau und die Lebenserscheinungen des

Brancliiostoma lubricum, Costa, Amphioxus lanceolatus,

Yarrell,” ‘Konigl. Akad. der Wissenschaften,’ Berlin, 1844.

2. Bateson, W. — “ The Later Stages in the Development of Balauoglos-

sus Kowalewskii,” ‘Quart. Journ. Micr. Sci.,’ vol. xxv (Supple-
ment p. 110), 1885.

3. Bolph. — “ Untersuchungpn ueber den Bau des Amphioxus lauceo-

latus,” ‘ Morphol. Jahrbuch,’ Bd. ii, 1876.

4. Stieda. — “ Studien ueber den Amphioxus lanceolatus,” ‘Memoires

de l’Acad. Imp. des Sciences de St. Petersbourg,’ tome xix, No. 7,

1873.

5. Langeruans. — “ Zur Anatomie des Amphioxus lanceolatus,”

‘ Archiv fur Mikroskopische Anatomie,’ vol. xii, 1875.

6. Schneider, Anton. — ‘Beitrage zur Anatomie und Entwickel. der Wir-

beltbiere,’ Berlin, 1879.

402

E. RAY LANKESTER.

7 Gunther, A. C. L. G. — In the ‘ Report on the Zoological Collections,
made in the Indo-Pacific Ocean during the Voyage of H.M.S. Alert,’
1884.

8. Rathke, H. — ‘Bemerkungen ueber den Bau des Amphioxus lanceo-

latus,’ Konigsberg, 1841.

9. Huxley, T. H. — ‘Proceedings of the Royal Society,’ vol. xxiii, 1874,

10. Quatkefages, A. de. — “Memoire sur le systeme nerveux et sur l’histo-

logie du Branchiostome ou Amphioxus,” ‘Annalesdes Sciences
naturelles,’ 3me ser., “ Zoologie,” tome iv, Paris, 1845.

11. Kowalewsky. — ‘ Memoires de l’Acad. Imp. des Sciences de St. Peters-

bourg,’ vii scrie, tome xi, No. 4, 1867.

12. Lankester, E. Ray. — ‘Quart. Journ. Micr. Sci.,’ vol. xv, 1875, p. 257.

13. Muller, W. — ‘Beitriige zur Anatomie und Physiologie als Festgabe an

Carl Ludwig,’ 1875.

14. Kossmann. — ‘ Arbeiten aus den zoologisch-zootomischen Institut in

Wurzburg,’ 1874.

15. Moreau, Camille. — “ Recherches sur la structure de la corde dorsale

de l’Amphioxus,” ‘Bulletins de l’Acad. Royale de Belgique,’ tome
xxxix, No. 3, 1875.

16. Hatscbek, B. — ‘ Anatom. Auzeiger,’ Aug. 15th, 1888.

17. Rohde. — ‘ Schneider’s Zoolog. Beitrage,’ vol. ii, 1888.

EXPLANATION OF PLATES XXXIV, XXXV, XXXVI,
XXXVI a*, & XXXVI B,

Illustrating Professor Lankester5 s “ Contributions to the
Knowledge of Amphioxus.5’

PLATE XXXIV.

Eig. 1. — Semi-diagrammatic drawing of a dissection of Amphioxus lan-
ceolatus. The animal is resting on the dorsal surface: the ventral half of the
body has been separated by a horizontal cut, reaching on each side from near
the mouth to the anus, and has been thrown over to the left side of the
animal. The perforated region of the primitive body wall exposed by thus
removing the epipleura has been severed from its median dorsal attachment
and pulled over to the animal’s right side, the dislocation being aided by
cutting through the body transversely. The pharyngo-dorsal coelom right and
left of the median attachment of the pharynx to the notochordal sheath is thus
exposed, the deeply folded wall formed by the pharyngo-pleural pouches being
cut through. The lining membrane of ccelomic spaces is everywhere coloured
blue ; other surfaces are left either colourless or of brownish tint. The

AMPHIOXUS LANCEOLATUS.

403

drawing is intended to show especially the relation of the pharyngo-pleural
pouches to the pharyngo-dorsal coelom, the position and relations of the atrio-
ccelomic funnels (“ brown-pigmented canals ”), and the post-atrioporal exten-
sion of the atrium as a caecal tube running side by side with the intestine
as far as the anus. A rod, E, is introduced through the atriopore into the
atrial chamber, and a second rod, F, is passed from the post-atrioporal enlarged
coelom through the natural passage into the peri-enteric coelom of the prae-
atrioporal region. The letters are a , d, d' indicate the parts similarly marked
in the transverse sections figs. 3 and 4 ; v. marks the folded cut edge of the
body wall corresponding to the pharyngo-pleural pouches of the primary bars
of the pharynx.

Fig. 2. — Diagrams of sections through the lines A B of Fig. 1.

Fig. 3. — Diagrams of sections through the lines C B of Fig. 1.

Fig. 4. — Amphioxus lanceolatus viewed from the right side, and
magnified about five times linear. The animal is represented as nearly as
possible in its living proportions and shape ; the oral hood and tentacles are
expanded, the atrial cavity is dilated, and the atriopore open. The drawing
shows the number of the myotonies, sixty-two (this is probably an exceptional
number, sixty-one or even sixty being more frequent) ; the number of the
oral tentacles, the number of the dorsal and ventral fin-rays, the number of
the gonads, the position of the atriopore and the anus.

Fig. 5. — Amphioxus lanceolatus viewed from the ventral surface.
The specimen is the same as that drawn in Fig. 4, and the drawing is intended
to show especially the plaits of the ventral epipleural surface, the position of
the metapleura, the double series of ventral fin-rays, and the form of the
praeoral hood.

PLATE XXXV.

Fig. 1. — Untouched sketch of a horizontal section through the region of the
atrio-coelomic canals. The section is in a plane passing below the notochord
and just cutting the top of the pharynx and the caecum. It shows the two atrio-
coelomic funnels, with their widely open posterior mouths and their narrow
anterior terminations, the left a little in front of the right.

Fig. 2. — Untouched sketch of a vertical section passing through the two
atrio-coelomic funnels. The division of the muscular fibres of the myotome
into two groups, a deeper and a more superficial, is seen on the left side. The
two sets of fibres are separated by connective tissue, but the distinction does
not depend, as the lettering would imply, on the direction of the fibres
themselves but on the arrangement of the groups of fibres; in both groups
the direction of the actual fibres is essentially longitudinal.

Fig. 3. — Untouched sketch of the section in which the atrio-coelomic funnels
were first observed. The Amphioxus had been hardened by Kleiuenberg’s

404

E. KAY LANKESTEK.

picro-sulphuric solution, followed by alcohol, and was not treated with any
staining reagent ; consequently tbe brown pigment of tbe atrial tunic, which
was richly developed in this specimen, was very obvious, as shown in the
figure. Artifact spaces, together with the natural lymph-spaces in the
neighbourhood of the notochord and nerve-cord, have produced a distortion of
this region. The space marked “ artifact space,” on the left of the figure,
is incorrectly labelled, being probably a natural lymph-space (myoskeletal
lymph-space) slightly distended by the action of reagents.

Fig. 4. — Untouched sketch of a section through the branchial region of a
female Amphioxus at the breeding season. The specimen was treated with
Muller’s fluid, followed by alcohol, and is remarkable for the great distension
of the epipleura by the enlarged gonads, leading to obliteration of the meta-
pleural ridges and canals and of the ventral plaits. The pharynx was badly
preserved, and the sub-notochordal region much distorted. The artist has
represented these parts in a rough and undetailed way.

Fig. 5. — Diagram of a portion of the body of Amphioxus, from which the
notochord, roof of the pharyngo-dorsal coelom, and upper part of the myo-
tonies have been removed by horizontal section, so as to show the atrio-
coelomic funnels in position. A rod is passed from the peri-enteric coelom into
the pharyngo-dorsal coelom.

Fig. 6. — Enlarged transverse section of an atrio-ccelomic canal, to show the
cell structure, after staining with hsematoxylin.

Fig. 6'. A piece of the same before staining.

PLATE XXXYI.

Diagram of a transverse vertical section through an adult female Amphioxus
at the region of the atrio-ccelomic funnels. The muscular tissue is coloured
yellow, the connective tissue and gonad cells purple, the hypoblastic cell-layer
and notochord green, and the atrial epithelium light brown ; the epidermis
and nerve-cord are uncoloured, as also are the clots present in the meta-
pleural canals, the fin-ray space, the pharyngo-dorsal ccelom, the myoskeletal
lymph-spaces, and the dorsal aortse.

The drawing is intended especially to show the form and position of parts
when corrected from drawings of actual sections, so as to allow for local
shrinking and distortion. The correct estimation of the spaces found in
sections between the skeletal sheath of the notochord and the muscular tissue,
and again between the latter and the neural skeletal tube and its skeletal crest,
is exceedingly difficult, ar.d has perhaps not been altogether rightly carried
out. It has been necessary to select a particular section for illustration, the
myo-skeletal lymph-spaces and the position of the inter-muscular septa being
necessarily different in each one of a series of sections traversing a given
segment of the organism. The seel ion selected shows part of a posterior

AMPHIOXUS LANCEOLATUS.

405

nerve-root and a portion of a group of anterior nerve-roots. The small
lymph-space to the left of the neural crest appears to be part of one of a
series (alternately right and left) not in continuity with those situate below
the nerve-roots. The lymph-space below the posterior nerve-root on the
right, and that below the anterior nerve-root on the left, are parts of corre-
sponding spaces lying asymmetrically on each side of the skeletal tissue of
the notochord-sheath and nerve-tube. The relation of the anterior nerve-
fibres to the deep layer of the myotome is seen on the left side, but probably
the correct and undistorted relation of these parts is not quite exactly ascer-
tained (see Rohde, 17). The division of the more ventral portion of the
myotomes into two groups of fibres, separated by delicate connective fascia, is
prominently shown.

The continuity of the gelatinous layer of the cutis between the ventral
epithelium and the transverse muscle, and the consequent absence of “ventral
cauals,” is prominently shown.

The median artery of the endostyle is not drawn. This is an omission
which is rectified by the series of sections in PI. XXXVI B, Pigs. 4 to 9.

PLATE XXXVI A.

Pig. 1. — Portion of a section through the notochord and adjacent region of
Amphioxus. The specimen had been stained with logwood solution. Some
of the structures shown are named in full on the drawing.

The letters have the following signification : — a. Artifact space, produced
by the shrinking and tearing of muscular fibres from their attachment to the
muscle fascia, 6. Left myoskeletal lymph-space, between the fascia of the
muscle of the myotome and the skeletal sheath of the notochord. Whether
this space is opened out by the action of reagents, or is in its natural condi-
tion, is uncertain. In either case it represents the similarly situated space of
Pig. 7, which represents an early larval condition, c. Similar space of the
right side. d. Upper or para-neural portion of the myoskeletal lvmph-space,
which is shown by other sections to be in continuity with 6. It contains a
clot. e. Similar space of the right side, elsewhere in continuity with c.
f. One of the “godets” of Moreau, or dorsal tubercles of the notochordal
tissue, g. Artifact space caused by the shrinking of the notochordal tissue
from its connective-tissue sheath. h. Dorsal canal of the notochord,
with nuclei and trabeculae, i. Veutral canal of the notochord, k. Thin
connective-tissue septum (fascia), separating the inner from the outer group
of muscle-fibres of the myotome. 1. Ventrally placed nuclei of the proper
notochordal tissue, x. Peculiar granular tract of the connective-tissue sheath
of the notochord, lying above the plane of adhesion of the pharynx, y. Simi-
lar but smaller tract of granular-looking tissue at the opposite face of the
notochord.

406

E. RAY LANKESTER.

Fig. 2. — Portion of a transverse section of Amphioxus, to show the
metapleur and the region of the supposed “ ventral canals ” (non-existent).
The various structures shown are named on the drawing. A small artifact
space is seen between the transverse muscle and the thick gelatinous layer of
the cutis, which is similar in position and origin to larger rents which have
led to the erroneous inference of the existence of “ ventral canals.” The
nuclei of the transverse muscle and the vertically fibrillated non-nucleated
structure of the gelatinous layer of the wall of the metapleur are noteworthy.
The same fibrillated structure is seen in the gelatinous layer of the cutis in
other sections through other parts of the body wall, especially where the
transverse intermuscular septa join the sub-epidermic cutis.

Pigs. 3, 4, 5. — Three vertical sections through the anterior extremity of
Amphioxus, the first in front of the termination of the nerve-cord, the
second through the olfactory pit, and the third through the eye-spot. The
drawings were made by Mr. Herbert Thompson, M.A., in the zoological
laboratory of University College.

a. Pin-ray lymph-space. b. Cavity of the neural sheath or skeletal tube
in front of the termination of the nerve-cord, occupied by a clot. c. Anterior
nerves in transverse section, d. Notochord, e. Sub-notochordal lymph-
space, probably similar in nature to the dorsal fin-ray lymph-space, f. Irre-
gular lymph-spaces lined with connective-tissue epithelium, excavated in the
gelatinous tissue of the dorsal and ventral fin-plates, g. The ventral prae-
oral fin-plate, becoming here the right half of the prseoral hood. h. The
unequally developed left-side outgrowth, which becomes the left half of the
praeoral hood. i. The ciliated olfactory pit, an inversion of the epidermis of
the left side, connected with a short olfactory bulb or nerve given off from
the left side of the nerve-cord. k. The wall of the nerve-cord enclosing the
enlarged neur-axial canal. /. The eye-spot, consisting of distinct spherical
granules of brown-black pigment, m. The unobliterated myoccelomic pouch
of the first myotome, identical with the cavity marked m. in Pig. 7- n. The
muscular epithelial cells of the mediad wall of the pouch, o. The myoccelomic
pouch of the second myotome.

Pig. C. — Diagram of a transverse section through an early larval condition
of Amphioxus, showing the division of the primitive coelomic pouches into a
dorsal myoccel and a ventral splanchnoccel. The origin of the cutis
from the peripheral and of the muscular tissue from the mediad wall of the
myoccel is indicated. Prom Hatschek.

Pig. 7. — Diagram of a transverse section through a later larval condition
of Amphioxus, showing the origin of the skeletal tissue from a secondary
pouch, “ the sklerotome,” which grows out from the myoccel and pushes its
way between the notochord and mediad muscular wall of the primitive myoccel,
and downwards between the wall of the splanchnoccel and the epidermis.
Prom Hatschek.

AMPHIOXUS LANCEOLATUS.

407

PLATE XXXVI jB.

Eig. 1. — Transverse section of a “primary bar” of the pharynx of
Amphioxus.

Pig. 2. — Transverse section of a “tongue bar” of the pharynx of
Amphioxus.

Letters in Pigs. 1 and 2. — al. Left inner epithelial band. ar. Right inner
epithelial band. nm. Median inner epithelial band. col. Columnar lateral
cells, with long cilia, n. Superficial nuclei. »' . Deeper nuclei (? all epithelial
or some mesoblastic). sept. Clear septal tissue. Bl. vess. Supposed blood-
vessel, connected in the primary bar with the lateral branches of the median
endostylar artery, and ending blindly at the ventral extremity of the tongue
bars ; probably connected dorsally iu both to the dorsal aortas. Rod. The
chitinoid skeletal rod, bifid in the primary bar, hollow iu the tongue bar.
fiss. Fissure, due to the bilateral origin of the rod of the primary bar. x.
Sporadic cavity of the rod of the tongue bar. Coelom. The coelomic cavity
occupying the pharyngo-pleural extension of the primary bar, but enclosed in
the rod of the tongue bar. Ccel. ep. Coelomic epithelium of the pharyngo-
pleural pouch of the primary bar. air. epith. Atrial epithelium (epidermic),
clothing the external face of both bars. pig. Lateral groups of pigment in
the atrial epithelium of both bars.

Fig. 3. — a. to e. Various forms of the chitinous substance of the tongue
bar, as seen in transverse sections, f. to h. Various forms of the chitinous
substance of the primary bars.

Pigs. 4 to 9. — Six consecutive vertical sections across the endostyle of the
anterior third of the pharynx of Amphioxus. The sections are numbered from
behind forwards. The first section (Pig. 4) passes on the side marked “ Right,”
through the junction of the anterior half of the fork of the rod of the
primary bar A with the posterior half of the fork of the rod of the primary
bar B. Opposite to this, on the side marked “ left,” the section traverses
the actual fork of the rod of the primary bar C, at the point where a lateral
vessel is given off from the median endostylar artery, and runs up into the
fork (see Schneider’s figure of the bars and blood-vessels, loc. cit.). The
grouping of the epithelial cells is shown in all the figures; they are lettered
in Pig. 9 as follows : — al. and ar. Right and left lateral epithelium (similar to
that marked al., am., ar., in Pigs. 1 and 2). bl. and hr. Right and left groups
of less-staining cells, cl. and cr. Right and left interposed groups of darker-
stained cells, dl. and dr. Mediad right and left groups of less-staining cells.
m. Median group of darker-stained cells, with a special very long tuft of
cilia. In all nine groups of cells.

In all the sections the chitin-like endostylar plates, immediately underlying
the nine-zoned endostylar epithelium, are seen ; like the chitinoid substance
of the pharyngeal rods they are coloured yellow, having a slightly yellow

408

E. KAY LANKESTER.

tinge and highly refringent character in the actual sections. The bilateral
character of these plates is clearly seen, also their connection with the rods,
and their “ thinning out ” at intervals ( e . g. Eig. 5, where the left endostylar
plate is very thin, and Fig. 4, where it is practically absent).

Following the sections according to their numbering, we see in the second
the median endostylar artery, with its clot coloured black for diagrammatic
purposes. The section traverses a free portion of the primary bar belonging
to rod A, and within the endostylar area we see the posterior half of rod B
on the one side, and the anterior half of rod C on the other side.

In the third section (Fig. 6) we come across the junction of a tongue bar
with the endostylar tract on the right side ; whilst the primary bar belonging
to rod C on the left side is seen much in the same relation as is the primary
bar belonging to rod A on the right side of Fig. 4.

In the fourth section (Fig. 7) the tongue bar on the right side, and the
primary bar on the left side, are cut more largely ; whilst a tongue bar extremity
is seen on the left side also.

In the fifth section (Fig. 8) the junction of the forks of two rods, C and D,
is traversed on the left side (actual left, but right of the drawing) ; and the
median artery is giving off a lateral branch to the incipient fork of the primary-
bar rod B.

In Fig. 9 the lateral vessel is more largely involved, the tongue bar of the
right side is cut through its free region, and the tongue bar of the left side
is seen in apposition to the endostylar tract.

The whole series of figures serves to demonstrate the continuity of the
coelomic space of the endostyle, with the pouch of the primary bars and with
the cavity of the hollow rods of the tongue bars.

Fig. 10. — a, b, c, d. Outlines showing various forms of distortion of the
notochord, in transverse section, brought about by the action of reagents.

Fig. 11. — Section through the fin-ray lymph-space of the first dorsal
fin-ray, to show the bilateral base of the fin-ray and its consequent simi-
larity to the double fin-rays of the ventral series between atriopore and anus.

a. Epidermis. 1. Outer lamella of cutis, c. Gelatinous layer of cutis.
d. Spaces in gelatinous layer of cutis, e. Inner lamella, with nucleated cells
of cutis, f. Right and left basal pieces of the fin-ray. g. Fin-ray lymph
compartment, h. Neural sheath.

Fig. 12. — The twelve post-oral tentacles of Amphioxus or tentacles of the
sphincter, seen from the post- oral surface. Drawn by Mr. Arthur Willey,
student of University College, London, from a preparation made by him.

These tentacles depend from the oral sphincter (velum of Huxley and
Langerhans) into the pharynx, and are provided with numerous special sense-
organs described by Langerhans (5).

muscles perigonaaiai

®loin

post atrio-poral coelom

mctspleural lymph canal

. — — — ~ atnopore

median lower face of the sheath of the notochord
: to which the dorsal aorta is attached

floor of atrial chamber

itaploural lymph canal

//all of tho po*
extension ol

alnoporal

atnopore

floor ofatnal chamber

:®lom

top of pharynx

pharyngeal Woouchtv

wall of intestine

c®lomic

funnel

muscle

•trio copIoti
funnel

muscle

top of pharynx

right dorsal artery

coelom

muscle

muscle

primary pharyngeal bar with ccelomic pouch
ondary pharyngeal bar or tongue

coccun.

ventral ridg
of pharynx

metaplci

mttapli

metapleur

position of i

myotome

notochord

myotome

notochord

plaited ventral wall of atrium formed by the

junction of down-grown epipleura left metaplei

iction

rntaeles

posterior «i
of notochor

mvotorne

double ventral fin

portenor end of sotocn

?OubU doaH
,lr,°pore 4 ,IC ra

Orattentaclei

left metaplei

ventral wall of atrium

openings of ctelomic pouches

XXXV.

sheath of notochord

notochord

muscle

left atrio-ccelomic
funnel.

circum-aite
lymph sp

atrium

right atrio -caelotuic
funnel

left atrio -caelomic
funnel.

atrium.

artifact spac

nerve chord

notochord

coecum

cbnn.tisr.

nucleus

deep ver
muse. fib

gonad pouch.

artifact

space

coelom

l. top of pharynx
i detached from its
\ adhesion to the
1 sheath of notochord

ccelom

left atrio-ccelomic
funnel.

. . Tight atrio-ccelomic
\ ’ funnel .

muscle of
"body -wall

ventral muscle

cut edge of
orig body-wall
detached fromx

dorsal artery

metapleural
lympn space

ova

V ) .//Si J j? :■ ‘

plaited ventral wall

atrium

E Ray Lankesler del

atrium

atrium

Jfoyr. fcwm^oi. WZISM. IlfVf.

connect, tissue ( cutis).

fin-ray..

.. Epidemics

. . . lymph space of fin-ray with clot.

space with

anterior nerve roots ... /

superior canal of. .
notochord.

space with clot

supra,-neur ad crest.

(d

muscular fibre,
posterior nerve root.

. nerve cord,

spare with clot

Tio to chord.

. conn. tiss. sheath
^ of notochord,

ext. muse, fibre

.. mi . muse, fibre
A. transv. skelet. ridge
right dors, artery
circum,. arterial.

canal,
-ccelom
.alrio coel. funnel
atrium

left metaphor

chit/nous hollow rod

chitinous solid, rod

venlral ridge of

In ■xalom.hcldUig
' pouch, of pksryn
1 : -peal bar.

. coelom, around
coecum,

Ip. . blood-vessels
[p of caecum,

.perigonadium,

(coelom)

. transv muscle
V * of epipleur.

iggonadcells

P lymph space
! ■ af clot of
metapleur

metapleuraj.

seberite..

..airiim.

oUrial epi-
thelium.

V right meiaplettr

blood- spare (coelom,)

- sub -epidermic conn,, tissue of
ventral, wall of epipleur.

pharynx with long edict, ■ epidermis thrown, into ridges by
vartiad contraction, of the transverse

A ay LanV.ester del

endostylar sheielafplade, .

muscle.

F. Huth.LSK EdmT

conn, tiss
Septum,

\ epidermis

outer lametta- of cutis
\V' .gelatinous layer (f d?
A\ ... inner lamella, <f d°

gonad, Mood-vessel

Neural canal

per ip on. coelom

atrial, epithelium.

transverse muscle.

cutis

mediad

raphe

inner lasridLa ofcutis
gelatinous layer of d °
outer lamella, (f d,°
epidermis

epidermis

outer' lametta of cuds ;

MYOCGL gelatinous layer of cutis

( dorsal fin-J J J J

spare

inner lamella ofcutis
with, nuclei.

rtslztog. layer
(nerve sheath, J

MYOCGL

J.... . muscle fasc

epidermis
... cutis -layer
CL. muscle- layer

Space h

orry i "
notochord.

...MYOCGL
. cuds

muscle notochord. .

MYOCGL

skeletog layer
. MYOCGL

fj skeletoaenous lay<
L nplaruhncplsur
■ SPLA NCJi NO C CP L
sorrMtopUur

SPLANCH NOCG L

intestine

MYOCGL
vtnlralf wj
scare

E Ray Lankester del.

F Huth.Lith* Edinr

XX1X,N.AxXXVI.b

Bl.vess.

B l . vess.

Blvess.

Blvess.

Blvess.

Blvess.

*•**«»*

Cadcm.'

Coelom.

Coelom.

Codom.

Coelom.

Coelom

Coelom.

Coelom,

RIGHT

H*Vh;

$5#

RIGHT

primary bar

7-

Tongue, bar

Fig. 10.

*V! »U«
•& »!**
«V*

RocLC

9*1 yrf~\

?■ hollow rod

yjfjp CceJcrn,
f end of tongue

bar rod

i tndostylar plate
/ ( left suU)

re/ecular Conner/.,
tissue

endostylar

plate.

Rod. A ■'
ant half

Bl.vess.

.// Rod C post.half /

Rod. B

artery to fork. post.half '

of primary Ictr

endostylar plate

hi. vess.

Coelom,

Coelom

Red.# post, half

Coelcm,

median axle ry

Rod ■ C an t, half

Reticula/' connect
tissue

Fig. 8.

tongue bar i \yt

ad/: eyoitlv

Coelom,

hollow rod

uidoslylargla/e,
(left side J

RodC ant half

air epUh,.

hod A

j i tongue bar

//■■ Rod, D
/ post, half

RodC
ant. half

Rod B
post half

Coelom,

Codom-

rnedian artery

Ccdom,

Rod B post, half

endostylar plale (right, side)

tongue bctr

tongue bar

hollow rot

\f / hollow rod

RodB 'Vs
post, half \

Rod. B
ant half

Cotlom, 0

tongue bar Rod

Rod, Bpost half

Rod D

/ post, half
endostylar

plate

Rod C ant. half

artery lo fork, of '

pruruuy bar mjfim„tarUy

.ankester del.

F Hut'n, Lithr Sdinr

STUDIES IN THE EMBRYOLOGY OF ECHINODERMS. 409

Studies in the Embryology of the Echinoderms.

By

II. Bury, B. A., F.I*.S.

Fellow of Trinity College, Cambridge.

With Plates XXXVII, XXXVIII, and XXXIX.

In the following pages I propose to give a comparative
account of the structure and development of certain organs in
young Echinoderm larvae. I am not here concerned with the
earliest stages (gastrula, &c.), nor with those later ones in
which the pentamerous symmetry of the adult is already
assumed, but only with that bilaterally symmetrical stage
which is more or less clearly represented in all Echinoderm
larvae, and to which Semon (28) has given the convenient name
“ Dipleurula.”

At this stage but few organs are developed, and one of them
— the alimentary canal — is too similar in structure throughout
the group to need any comparative description. I have there-
fore confined myself to the following points :

I. The primary divisions of the coelom, starting
from a stage in which at least two enterocoel pouches
are already present.

II. The Hydrocoel: its development and connections.

III. The Skeleton, so far as it is developed in the dip-
leurula stage.

I. Primary Divisions of the Ccelom.

Up till quite lately only three main divisions of the coelom
were usually recognised in Echinoderm larvae — the right and

410

H. BUEY.

left enterocoels and the hydrocoel; for though some observers
asserted the presence in Ophiurid and Echinid Plutei of a
second hydrocoel, yet the existence of this was not generally
accepted.

In my paper on Antedon (7) I showed that there exists in
the larva of this animal a separate cavity (anterior body-cavity),
median, or nearly so, with which the water-pore is related, and
into which the hydrocoel subsequently opens by means of the
water-tube (= stone-canal) ; and I pointed out the existence in
Asterina gibbosa of an apparently homologous cavity. I
shall now try to show that this cavity is always represented in
Echinoderm larvae, but it is not always unpaired, as in Ante-
don, a very distinct fellow to it on the right side being fre-
quently visible. This condition, with two anterior enterocoels,
I consider to be probably the most primitive, and I shall
therefore begin my account with a description of those forms
in which it obtains.

Ophiurids. — The youngest Ophiurid Pluteus I was
able to obtain had already a pair of cavities lying beside the
oesophagus, but none as yet beside the stomach ; and was, in
fact, in the stage described and figured by Metschnikoff (18,
p. 21, pi. v, fig. 2). The only account we possess of the origin
of these cavities is that of Apostolides (3), who believes that
both in Ophiothrix versicolor and in Amphiura squa-
mata they are formed by delamination in the mesoblast.
Since, however, he describes the gastral cavity of these forms
as also formed by delamination, while Kowalevsky (13, p. 5) in
Ophiura (sp. ?), Selenka (27) in Ophioglypha lacertosa,
Balfour (4) in Ophiothrix fragilis, and Fewkes (9) in
Ophiopholis aculeata, find it to be formed by invagina-
tion, we may be permitted to think either that Apostolides is
mistaken, or that he has studied exceptional forms ; and that
in some Ophiurids, at least, the two cavities beside the oeso-
phagus arise as pouches of the archenteron, as in other Echino-
derms. In any case the homology of these cavities with the
similarly placed pair in Echinids and Asterids is hardly likely
to be disputed.

STUDIES IN THE EMBRYOLOGY OE ECHINODERMS, 411

In the next stage observed there was present, in addition to
the pair of cavities already mentioned, another pair beside the
stomach. I was unable to trace the origin of this new pair,
but I have no reason to doubt the correctness of Metschnikoff ’s
statement (18, p. 21 ; 19, p. 62) that they are directly derived
from the first-formed pair.

At this stage (and possibly earlier) the left anterior entero-
coel, which is not, as has been stated, a hydrocoel, opens to the
exterior by a pore (“ water-pore ”) at its posterior end on the
dorsal surface of the larva. Examination of the living animal,
under a high power, shows that this pore is formed by a single
elongated cell, perforated throughout its length, and lined with
cilia. It is important to notice that at this stage we have a
larva with almost complete bilateral symmetry (broken only
by the water-pore), and with two pairs of enterocoel pouches
assuming a metameric arrangement.

In fig. 1 I have represented one of the earliest stages in the
development of the hydrocoel, and one which has escaped the
notice of previous observers. On the right side the anterior
and posterior enterocoels remain as in the previous stage,
but on the left side a third vesicle has now made its
appearance between the two previously formed : this third
vesicle is the hydrocoel. For a long time I remained in
doubt whether it was derived from the anterior or posterior
enterocoel ; in the specimen figured, however, it was quite
separated from the former though still connected with the
latter; and in fig. 18 is shown, on a larger scale, part of a
larva in which its walls are closely fused with those of the
posterior enterocoel, while the anterior enterocoel (represented
by a dotted outline) is apparently entirely distinct. These
and other specimens have convinced me that the hydrocoel is
derived from the left posterior euterocoel, though I confess
this conclusion has surprised me, since, as we shall see, this
condition is not found in any other group of Echinoderms.
The stage just described is a very transitory one: almost im-
mediately the hydrocoel separates itself completely from the
posterior enterocoel, and, assuming an elongated form, pushes

412

H. BURY.

its way towards the anterior end of the larva, immediately
ventral to the left anterior enterocoel. Prom the latter it is
distinguishable by its more regular outline, the greater dis-
tinctness of its lumen, and the more refringent character of
its walls ; indeed, it is altogether so much more conspicuous
an object than the overlying enterocoel, that it is easy to over-
look the latter altogether, or to mistake it for a solid mass of
mesoderm cells — an error into which previous observers appear
to have fallen.

A section, through a larva in this stage, just behind the
water- pore, is given in fig. 19, and confirms the observations
made upon the living animal. It is true that in this section
the lumen of the right anterior enterocoel is not seen ; but
this, when the small size of the cavity is considered, is not
surprising. I have abundant evidence that the cavity still
exists.

Echinids. — The first stage which concerns us here is that
of the young Pluteus with two completely separated enterocoels
lying beside the oesophagus ; a satisfactory figure of this stage
is given by Prouho (24, pi. xxiv, fig. 3). Whether these two
cavities are separate from the first (24, p. 234), or whether they
are at first united (26, p. 49), is a question which need not
detain us now. Even at an early period the bilateral sym-
metry is rendered incomplete by the development of a ciliated
pore (water-pore) at the posterior end of the left enterocoel.

The next stage in development is marked by the division of
each of the primary enterocoel pouches into two lobes, one of
which remains beside the oesophagus, while the other extends
back to the side of the stomach. This was the latest stage ob-
served by Prouho (24), but in all my larvae (belonging
to Echinus microtuberculatus, Strongylocentrotus
lividus, Spliaerechinus granularis, Echinocardium
cordatum, and other unknown forms) each cavity soon
divided completely into two, so that a stage was reached such
as MetschnikofF (18 and 19) has already described, in which
there exist two anterior and two posterior enterocoels, the only
difference between my account and MetschnikofF’ s being that

STUDIES IN THE EMBEYOLOGY OE ECHINODEEMS. 413

he describes the anterior enterocoels as “ Wassergefassan-
lagen” (= hydrocoels), which I shall presently show to be
incorrect.

But whatever name we may give to these cavities I am
entirely in agreement with Metschnikoff as to the reality of
their existence, which has been frequently questioned, or even
denied. Thus Gotte (11, p. 609) denies their presence in
Spatangids, and asserts that the two primary enterocoel cavi-
ties simply shift their position down to the stomach, while part
of the left one at the same time forms the hydroccel. This
view is also taken by Ludwig (17, p. 141), and apparently by
Selenka (26), who does not otherwise account for the fact that
the only two cavities noticed by him are at first situated beside
the oesophagus, and afterwards beside the stomach.

Though Metschnikoff’s account of this stage is therefore the
most correct one yet given, yet in the next stage he has, as in
his account of Ophiurids, fallen into error by confounding the
hydrocoel with the anterior enterocoel ; it must, however, be
admitted that the phenomena are here even more difficult to
make out, and that a stage is soon reached in which, but for
the evidence afforded by the Ophiurids, the fundamental dis-
tinction between the anterior enterocoel and the hydrocoel could
hardly have been recognised.

In fig. 8 is represented the earliest stage in which I have been
able to determine the relations of the hydrocoel. It consists,
as in Ophiurids, of a vesicle with thick and well-defined walls,
lying between the anterior and posterior enterocoels on the left
side ; but it differs from its homologue in Ophiurids in that its
cavity is already connected with that of the left anterior en-
terocoel. Whether it is derived from this anterior enterocoel
or from the posterior one, as is stated by Gotte (11, p. 609),
Ludwig (17, p. 141), and Selenka (26, p. 49), is a question
which, in spite of much time spent upon it, I am unable to
answer satisfactorily. The latter is suggested by the analogy
of the Ophiurids, but on the whole my observations seem to
support the former, and to indicate that the communication
above mentioned between the anterior enterocoel and hydrocoel

VOL. XXIX, PART 4. NEW SER.

E E

414

H. BURY.

means that they have never become entirely separated.
However this may be, in considering the difference between
Metschnikoff’s account and nomenclature and mine, it is im-
portant to notice that on the right side of the oesophagus there
still exists a cavity exactly comparable to that which I have
called the left anterior enterocoel (except that it has no pore),
whereas the structure to which I have here confined the term
hydrocoel is entirely unrepresented on the right side.

Asterids. — In the young Bipinnaria we find beside the
oesophagus a pair of peritoneal vesicles, one of which early
opens to the exterior by a pore situated at its posterior end.
Both vesicles soon extend back to the stomach, over which
these posterior lobes spread dorsallv and ventrally, while the
anterior lobes, lying beside the oesophagus, have no such dorso-
ventral extension; it is at the junction of the anterior and
posterior lobes on the left side that the water-pore lies. It is
clear that we have here the representatives of the anterior and
posterior enterocoels of Ophiurids and Echinids, though they
are not as a rule separated from one another.

In many forms of Bipinnaria and in Brachiolaria the two
anterior lobes of the enterocoels grow forwards and unite in
front of the mouth, and are then continued as a common cavity
into the large prseoral lobe ; but they are always separate in
young larvse, and in certain forms of Bipinnaria described by
J. Muller (21) and Metschnikoff (18, pp. 32 — 40) they never
unite at all. I obtained a few examples of such a form at
Naples, and they are so instructive that it will be well to
describe their anatomy in some detail.

In fig. 14 one of these larvae is represented as seen from the
dorsal side. The hydrocoel is already present as a pouch
opening into the left anterior enterocoel, though its exact mode
of origin was not traced. But the great peculiarity of this
larva lies in the fact that on the left side the anterior and
posterior enterocoels are entirely separated from one
another. In Asterina (17) there is a dorsal communication
of these two cavities just above the hydrocoel and just behind
the pore, and a ventral communication just below the hydrocoel;

STUDIES IN THE EMBRYOLOGY OF ECHINODERMS. 415

the latter communication is absent in all the Bipinnarise
which I have examined, but the dorsal one is usually present.
In the small form here figured, however, I convinced myself,
both by observation of the living animal and by sections, that
no communication existed, between these two cavities at this
stage.

It is interesting to notice that this larva gives us two cha-
racters, not possessed by other Asterid larvae, in which it
resembles the Plutei of Opliiurids and Echinids : (1) firstly,
the anterior continuations of the enterocoels are never united
in the prseoral lobe ; (2) secondly, on the left side there is a
complete separation of the anterior and posterior enterocoels
just behind the water-pore. No separation of the anterior and
posterior enterocoels on the right side was ever observed. It
should further be noted that in its external form this larva
gives evidence of being primitive, in that it retains up to the
time of metamorphosis the simple outline characteristic of all
young Bipinnarise, without developing the gigantic prseoral
lobe and elongated arm-like processes which so entirely alter
the appearance of other Bipinnarise in their later stages.

In fig. 13 is given a lateral view of the same larva, in which
it will be seen that the hydrocoel occupies nearly the same
position as in young Ophiurid and Echinid Plutei namely,
between the anterior and posterior enterocoels ; but in remain-
ing open to the anterior enterocoel it approaches the Echinid
rather than the Ophiurid condition.

The Bipinnarise with a large prseoral lobe and terminal fin
(Bipinnaria asterigera), and Brachiolaria, do not differ
much internally from the form just described, except in the
already-mentioned communications (1) between the anterior
and posterior enterocoels of the left side, dorsal to the hydro-
coel, and (2) between the right and left anterior enterocoels ;
the latter is clearly secondary. But in Asterina, Ludwig (17)
describes a very different arrangement ; though, as I shall sub-
sequently show, my observations do not entirely agree with his.
According to him, the right and left enterocoels are from the
first connected with a single large anterior cavity occupying

416

H. BUEY.

the whole of the praeoral lobe ; and the hydrocoel, which arises
as a pouch on the left side at the junction of the left enteroccel
with the cavity of the praeoral lobe, has at first a portion of the
coelom intervening between it and the wall of the stomach — a
condition which I have not observed in any other Echinoderm.
Subsequently the hydrocoel comes to lie close to the stomach,
having, as already mentioned, a ventral as well as a dorsal
communication between the anterior and posterior divisions of
the coelom (17, p. 147, pi. ii, fig. 37).

In a previous paper (7, p. 38) I put forward the view that
the single large anterior enterocoel of Asterina might be primi-
tive ; but I ought to have remembered its formation in Bipin-
naria and Brachiolaria by the fusion of two primarily distinct
cavities. Now, the larva of Asterina is probably a modified
Brachiolaria (17, p. 154), so that we shall be more correct in
assuming that in this, as probably in many other points,
Asterina exhibits a secondary and abbreviated form of develop-
ment.

Crinoids. — The earliest formation of the enterocoel pouches
and hydrocoel in Antedon rosacea (the only Crinoid yet
studied), exhibits some peculiarities with which we are not at
present concerned (see 5 and 7) ; we will pass at once to the
stage in which the following divisions of the coelom are pre-
sent : (1) a single median anterior enterocoel (called “ caual de
sable ” by Barrois). At the close of the free-swimming stage,
this opens to the exterior at its posterior end by the water-
pore, situated on the left side of the larva ; (2) right and
(3) left posterior enterocoels, one on each side of the stomach;
these do not long retain their lateral position ; and (4) the
hydrocoel, on the left side, between the anterior and the left
posterior enterocoels ; at first it opens into the anterior en-
terocoel, but this communication is soon closed. Although
this anterior enterocoel usually occupies a median position in
the praeoral lobe, I have little doubt that it is the homologue
of the left anterior enterocoel of other Echinoderms, and,
indeed, in many cases it lies distinctly on the left side.

Holothurians. — For the early stages of development of

STUDIES IN THE EMBRYOLOGY OF ECHINODERMS. 417

the coelom in this group I must refer to the well-known
accounts of Metschnikoff (18) and Selenka (25). In a well-
developed Auricularia, we find a pair of enterocoels beside the
stomach, and a single vesicle on the left side at the level of
the oesophagus ; this vesicle, which opens to the exterior by
the water-pore, is usually looked upon as the hydrocoel, but I
shall endeavour to show that it contains also the rudiment of
an anterior enterocoel.

Fig. 22 gives a lateral view of this vesicle in an abnormal
specimen which first attracted my attention to the subject : it
will be seen that between the straight tube leading from the
pore and the thick-walled inner portion (which subsequently
becomes lobed, and is undoubtedly the hydrocoel), there inter-
venes a thin-walled section, which extends but slightly behind
the pore, but is considerably elongated towards the anterior
end of the larva.

In fig. 21 we have a dorsal view of the same part of another
larva, showing that the thin-walled cavity has no great lateral
extension. It is evident that this cavity has precisely the
position and relations of the anterior enterocoel of an Echinid
Pluteus, or of the small Bipinnaria above described; it is,
therefore, important for us to see how far it is represented in
normal larvae. My attention was not called to this point till
rather late in the season, when Auriculariae were becoming
scarce ; but from time to time I managed to obtain a fair
number of larvae, and there was not one in which I was not
able to recognise some vestige of this anterior enterocoel,
though before noticing the abnormal one (fig. 22) I had never
detected a trace of such a structure. The cavity is extremely
variable in size, but fig. 23 represents what appears to he a
fairly typical development, and is useful in illustrating the
extreme difficulty of observation. Owing to the form of the
larva, the only positions in which it is possible to get a steady
and prolonged observation, give us a directly dorsal or directly
ventral view ; but in a dorsal view (fig. 23) the thin-walled
anterior enterocoel lies directly over the thick- and refringent-
walled hydrocoel, and is consequently extremely difficult to see,

418

H. BURY.

its component cells being hardly distinguishable from the
surrounding mesoderm cells, which lie scattered over the
surface of the hydrocoel, while other mesoderm cells (purposely
omitted in. this figure) collected round the water-pore and the
tube leading from it (“ pore canal ”) further obscure it. A
lateral view (fig. 24) is far more satisfactory, but is by no
means easy to obtain ; the only method known to me is to
place the larva in a watch-glass and roll it over until it assumes
the required position ; but of course it cannot then be kept
perfectly steady, and drawing with the camera is impossible.

To the later stages of development, in which this cavity is
still easily traceable, I shall return presently. Enough has, I
think, been said to show that in Auricularia a cavity is
present between the hydrocoel and the water-pore, which,
though usually rudimentary, we have reason to regard as the
representative of the left anterior enterocoel. We shall sub-
sequently see that it is also present in Cucumaria, though
perhaps not at such an early stage.

Summary and Conclusions.

Ophiurids. — Here we find two pairs of enterocoels, meta-
merically arranged. The anterior enterocoels retain the
position of the primary pair of peritoneal vesicles, and one of
them (the left) opens to the exterior at its posterior end by
means of the water-pore. Besides these there is formed some-
what later a hydrocoel, lying on the left side between the
anterior and posterior enterocoels, and apparently derived
from the latter ; at this stage it has no communication with
the anterior enterocoel.

E chin ids. — These have two pairs of enterocoels and a
water-pore, as in Ophiurids ; the hydrocoel occupies the same
position as in that group, but appears to arise from the
anterior enterocoel, and to retain its communication with it.

As ter ids. — The anterior and posterior enterocoels are dis-
tinguishable on both sides, but are not usually separated,
though they are so on the left side in one form. The water-
pore and hydrocoel occupy their usual positions, but the latter

STUDIES IN THE EMBRYOLOGY OF EOHINODERMS. 419

remains open to the anterior enterocoel, from which it probably
arises.

Crinoids. — Only one anterior enterocoel is present. The
hydrocoel is at first connected with this anterior enterocoel, but
subsequently becomes independent. The water-pore opens to
the anterior enterocoel, and a pair of posterior enterocoels lie
beside the stomach.

Holothurians. — The left anterior enterocoel appears to be
present, but rudimentary, and connected from the first with
the hydrocoel. Two posterior enterocoels exist as in other
groups.

From these facts we may arrive at the following conclu-
sions :

(1) A pair of anterior enterocoels was probably originally
present in all Echinoderms. So long as the left anterior
enterocoel of Ophiurids and Echinids was confused with the
hydrocoel, doubts were frequently expressed as to whether it
ever had a fellow on the right side (4, p. 458; 11, p. 609;
17, p. 141 ; 26, p. 49) ; and even when this was admitted to
exist, it was hinted that this bilaterally symmetrical arrange-
ment might be pathological (17, p. 142). However plausible
this supposition may formerly have seemed, it appears to me
absolutely untenable in the face of the new evidence here
advanced. Pathology may account for such obviously mon-
strous forms as that observed by Metschnikoff (19, p. 64), but
the term is clearly inapplicable to a condition which obtains
with the utmost regularity in every individual member of two
groups (Ophiurids and Echinids). Nor is it any more satis
factory to assume that a portion of the enterocoel is cut off on
the right side merely for the sake of symmetry (26, p. 50) and
is then allowed to atrophy without further development. The
only tenable view, as it seems to me, involving a secondary
origin for these anterior enterocoels, is that they are derived by
a species of segmentation from such a condition as is found in
many Asterids, in which the oesophageal and gastral sections
of the coelom are continuous on each side of the body ; but the
apparently primitive character of the Bipinnaria, in which the

420

H. BURT.

anterior and posterior enterocoels of the left side are distinct,
is, as already remarked, opposed to the idea that the other
Asterids are primitive in this respect. If union of the cavities
originally obtained (as is not improbable), there is strong
reason for supposing that before the separation of the existing
groups of Echinoderms, the segmentation of each lateral
cavity into an anterior and a posterior part had already
occurred.

(2) The hydrocoel is generally formed distinctly later than
the other cavities ; indeed, the only apparent exception to this
is afforded by the Holothurians, and in these it is quite as
reasonable to consider the anterior vesicle to be an anterior
enterocoel as to follow previous writers in regarding it as the
hydrocoel. This being so, it seems probable that the hydrocoel
is of later phylogenetic origin than the enterocoels. The entire
absence of any trace of a right hydrocoel makes it improbable
that this organ was ever paired ; but we must not lay too
much stress on this evidence, seeing that in Holothurians and
Crinoids the right anterior enterocoel has entirely disappeared.
In its mode of origin the hydrocoel varies, but its normal posi-
tion, when formed, seems to be between the anterior and
posterior enterocoels, not separated by either of them from the
wall of the stomach.

(3) The water-pore always (with the possible exception of
Holothurians) arises in connection with the posterior end of
the left anterior enterocoel, and only communicates indirectly,
if at all, with the hydrocoel. In all pelagic larvae it ap-
pears exceedingly early — probably always before the hydro-
coel (except, perhaps, in Holothurians). In Asterina, on the
other hand, it is formed later, simultaneously with the hydrocoel ;
while in Antedon it does not make its appearance till after the
hydrocoel has become five-lobed, and the larva has escaped from
the vitelline membrane. It might be thought from this that
the water-pore was really a late development, which had become
precociously formed in pelagic larvae on account of its physio-
logical importance, and that its real time of formation was
about the same as, or later than, that of the hydrocoel. This

STUDIES IN THE EMBRYOLOGY OF ECHINODERMS. 421

may be so, but the evidence seems to me to point rather to
the water-pore having existed in a very early stage in the
history of Echinoderms, probably before the hydrocoel had
arisen.

A word must be said here as to the probable function of the
water-pore. Bearing in mind Hartog’s experiments and
remarks (12) upon the madreporite of the adult, I carefully
observed not only the apparent motion of the cilia but also the
action of the currents produced by them, as indicated by the
motion of particles suspended in the water. The apparent
motion of the cilia was inwards, which, as we know, indicates
that the real current produced is exhalent. If we need an
illustration of this we have only to turn from the water-pore to
the oesophagus, and observe the motion there ; for in this case,
while the apparent motion is outwards, particles suspended in
the water show clearly that the current passes inwards. With
regard to the water-pore also particles suspended in the water
will guide us as to the direction of the current, though not
so readily as in the case of the oesophagus. No particles were
ever observed to pass in through the pore, though there was
nothing in their size to prevent them ; on the other hand, it
was difficult to observe an exhalent current owing to the rapid
motion imparted to the particles by the external cilia of the
larvae; nevertheless, in a few cases in Echinid Piute i, in
Bipinnaria, in Auricularia, and in Tornaria, a definite repulsion
of particles from the pore was noticed. Taking this in connec-
tion with the apparent motion of the cilia (also observed in
Ophiurid Plutei) it seems safe to assert that the current
passing through the water-pore is an exhalent one, though,
from the very slight disturbances produced by it in the sur-
rounding water, I conclude that it is not usually very strong.
It is, of course, not proved that the current is never inhalent;
but until such a reversal of its direction has been definitely
observed we have no particular reason for supposing that it
ever occurs.

As many of the larvae observed by me had as yet no hydrocoel,
we arrive at the conclusion that in Echinoderms as well as in

422

H. BURY.

Balanoglossus the water-pore and the short tube by which it
communicates with the enterocoel represent a primitive nephri-
dium. It would seem from Hartog’s observations that in the
adult the nephridial function is transferred to the water-
vascular system.

The existence of a pair of anterior enterocoels in Echinoderms
is to some extent opposed to the homology which I formerly
attempted to establish between the anterior enterocoel of An-
tedon and that of the larva of Balanoglossus. It is true that
the existence of two pores belonging to the anterior cavity in
the larva of B. Kupfferi (30) is evidence of the paired nature
of this cavity ; but, on the other hand, two distinct anterior
cavities are never, so far as I know, present in Balanoglossus,
while no instance is yet known in which a pore normally occurs
in connection with the right anterior enterocoel of Echinoderms.
The present paper, therefore, adduces no new evidence in favour
of the phylogenetic connection of the Echinodermata and
Enteropneusta, though it does not seriously weaken the proba-
bility of such a connection.

II. Further Development of Hydroccel : Water-Tube.

Ophiurids. — We left the hydroccel of this group as a
closed elongated vesicle stretching forwards for some distance
under the anterior enterocoel, and backwards as far as the
posterior enterocoel. The next change consists in the forma-
tion of five lobes on its outer (left) border. This has been
already described by previous observers, and need not be dwelt
upon here ; it is only necessary to add that the water-pore
usually lies at first nearly at the level of the third lobe, but by
a further shifting forwards of the hydroccel it afterwards
comes to lie over the interval between the fourth and fifth, or
even over the fifth (posterior) lobe itself. When this stage
has been reached we notice the first formation of the water-
tube (= stone-canal), which arises as an outgrowth of the
posterior end of the hydrocoel, between the fourth and fifth
pouches, and has, like the rest of the hydrocoel, a columnar

STUDIES IN THE EMBRYOLOGY OF EOHINODERMS.

423

ciliated epithelium. It runs directly dorsalwards, and after an
extremely short course* opens into the comparatively thin-
walled anterior enterocoel almost immediately below the pore.
From the fact that it is almost impossible to get anything but
a directly dorsal or directly ventral view of a living Ophiurid
Pluteus, the relations of the parts just described are not
always easy to make out ; and owing to the extreme minute-
ness of the cavities concerned* sections are even less satis-
factory than the living objects. Nevertheless, with favourable
living specimens (Pluteus paradox us is one of the most con-
sistently satisfactory forms* but there is much individual
variation) placed in an extremely small quantity of water, and
examined with a high power (Zeiss* Obj. E or F), it is not
usually a very difficult matter to see the cilia working in the
water-pore at the surface, and in the up-turned mouth of the
water-tube at a deeper level.

In fig. 3 I have attempted to give some idea of a dorsal
view of this region of the body, but it is impossible in such a
figure to convey a correct notion of the differences of level,
and the arrangements of the parts will be better understood
from the diagram (fig. 2).

Shortly before the final metamorphosis of the whole larva
into the pentagonal form, the hydroccel grows round the
oesophagus into the form of a ring. This has been already
described by Metschnikoff, but I have not found it easy from
his description to tell in which direction this growth takes
place, and as others may have shared my difficulty, I may be
excused for adding some details to his account.

Fig. 4 gives nearly all that is necessary ; the fourth and
fifth lobes retain their places, while the three anterior ones
grow across to the right side on the dorsal side of the
oesophagus. On reaching the right side, the first lobe passes
underneath (ventral to) the anterior enterocoel and the oeso-
phagus, and so nearly joins the fifth, which at the same time
bends slightly in under the oesophagus as if to meet it.

The further history of these parts cannot be described with-
out entering into a detailed account of the metamorphosis, but

424

H. BURY.

it is important to notice that fig. 4 represents a stage in
which the hydrocoel has already formed a ring round the oeso-
phagus with five tentacular outgrowths, while the rest of the
body still retains its bilateral form.

Echinids. — After the stage represented in fig. 8 the hydro-
coel pushes its way farther back (not forward, as in Ophiurids)
until it comes to lie in the centre of the left side. In doing
this it does not pass on the outside of the posterior enterocoel,
but lies close to the wall of the stomach, while the posterior
enterocoel forms a kind of horse-shoe round it, as represented
in fig. 7.

In this figure the hydrocoel already possesses five lobes, aud
is itself curved into the form of a ring, incomplete towards the
posterior end. The tube connecting it with the anterior en-
teroccel is now much longer than before, and has acquired a
columnar epithelium ; it is, as already mentioned, the water-
tube (= stone-canal), and enters the hydrocoel ring anteriorly
and slightly dorsally. Shortly after this the hydrocoel ring
closes completely, leaving a central perforation (19) through
which, at a much later period, the oesophagus grows.

We must now return to the anterior enterocoel. It has
usually been stated (1, p. 714 ; 18, p. 42 ; 14, p. 40 ; 9, p. 137)
that the water-tube (= stone-canal) opens directly to the
exterior by the water-pore, the hydrocoel being supposed to be
formed either by a direct metamorphosis of the left oesophageal
enterocoel (1 and 18), or as an outgrowth from the left posterior
enterocoel (11, p. 609; 17, p. 141 ; 26, p. 49), but in his last
note on the subject (19) MetschnikofF tells us that, though
most of the cavity by the oesophagus goes to form the hydro-
coel, a considerable portion forms a pulsating vesicle into which
the pore at first opens. It might be supposed that this pul-
sating vesicle was the anterior enterocoel of my account, but
this does not appear to me to be the case.

In fig. 9 I have given a view of the pore aud its surround-
ings so far as I have been able to make them out. On the left
is seen the water-tube (= stone-canal) coming up from the
hydrocoel, and opening into a swollen portion of the anterior

STUDIES IN THE EMBRYOLOGY OF ECHINODERMS. 425

enterocoel, which I have marked “ ampulla,” and this in its turn
opens to the exterior by a conspicuous median pore. Also in
the median line, and partly hidden by the pore, is a large pul-
sating vesicle, which is doubtless the same as that seen by
Metschnikoff ; it is overlaid by a reticulated calcareous plate
(not represented), which surrounds the pore, and makes it
extremely difficult to determine the relations of the subjacent
parts. That the pore does not open directly into the pul-
sating vesicle I am almost certain, while I have seen its open-
ing into the non-contractile ampulla in a large number of
specimens. It is certainly, however, possible that there may
exist a communication between the ampulla and the pulsating
vesicle which has escaped my notice ; the wall separating them
is undoubtedly very thin, and a small valvular aperture in it
would be exceedingly hard to see in the living animal, while
it would be practically impossible to detect in sections. I
have not as yet been successful in tracing the origin of this
pulsating vesicle, but, as far as I can make out, it is at first
more widely separated from the anterior enterocoel than in the
stage figured, and I have no reason to think that it is derived
from this enterocoel ; I am more disposed to believe that it is of
schizocoel origin, and that it is at no time connected either
with the water-pore or the ampulla.

Most of the parts represented in fig. 9 are subject to con-
siderable variation in size, not only in different forms of
Plutei, but even in Plutei belonging to the same species ; it
may, however, be stated generally that the right anterior
enterocoel and the anterior continuation of the left anterior
enterocoel (i. e. all except the ampulla) are difficult to see in the
living animal, since they lie rather under the oesophagus,
though they are easily visible in sections. In the specimen
figured the water-tube (stone-canal) and the thick-walled
tube from the water-pore (pore-canal) are continuous on one
side ; but in some other Plutei, and especially in Spatangids,
they are more or less widely separated ; in fact Pewkes (9) has
evidently seen only the water-tube, and has described its
opening into the anterior enterocoel as the water-pore; the

426

H. BUBY.

real water-pore is always median or even somewhat to the
right, except in the very earliest stages, but it is certainly not
easy to see in the opaque Spatangid Plutei. What Fewkes
describes as a movement of the pore is a growth of the water-
tube (stone-canal) to join the pore-canal.

Asterids. — In this group the primary opening of the
hydroccel into the anterior enterocoel continues to exist after
the former has acquired its five primary tentacular lobes; but
it does not, as in Echinids, give rise directly to the water-tube.
The latter arises after the appearance of the tentacular pouches,
and in the small Bipinnaria above described, in which the
anterior and posterior enterocoel are distinct, runs close to the
surface of the stomach in the mesentery separating these two
cavities, and opens into the anterior enterocoel beside the water-
pore : it occupies a correspondiug position in other Bipinnariae
and in Asterina, in which the mesentery in question is in-
complete.

Not only is the primary connection of the hydroccel with
the anterior enterocoel distinct from the water-tube, but they
are in different interradii ; this has been proved by Ludwig
for Asterina, and is apparently true also of Bipinnaria, though
I cannot assert this with any confidence.

Crinoids. — The hydrocoel forms a ring (long, incomplete)
through which the oesophagus grows. The water-tube (stone-
canal) starts from one end of this incomplete ring, and opens
into the anterior enterocoel (7, p. 21). It is, of course, im-
possible to say whether this new opening is in the same position
as the primary one, since the latter is closed before the forma-
tion of the tentacular lobes.

Holothurians. — In Auricularia the primary opening be-
tween the anterior enterocoel and the hydrocoel persists as the
water-tube (stone-canal), but instead of remaining short, as in
figs. 22 and 24, it elongates rapidly just before metamorphosis
into the “Pupa,” and forms a tube with columnar epithelium.
At the same time the cells forming the wall of the anterior
enterocoel become rounded and increase in number, at the
expense of the cavity, so as to form a bunch of cells, which

STUDIES IN THE EMBRYOLOGY OP ECH1NODERMS. 427

MetschnikofF (18, pi. iii, fig. 20), at a later stage, described as
mesodermic; it is really present in the oldest Auricularim,
though in these it escaped the notice of MetschnikofF and of
Semon (28). In fig. 25 I have represented part of a section
through a larva just entering into the pupa stage, in which the
opening of the water- tube into the anterior enterocoel is clearly
seen. This section is also useful as illustrating the various
parts into which the tube, usually spoken of as the “ stone-
canal/’ stretching from the water-vascular ring (hydrocoel) to
the water-pore, is divisible; (1) water-tube (= stone-canal of
Asterid), (2) anterior enterocoel, and (3) pore-canal; these
three parts are distinguishable in all Echinoderm larvse,
though the second has been frequently overlooked, and the
first and third consequently spoken of as one.

The remnant of the anterior entei’ocoel is also traceable in
Cucumaria (fig. 26), though I have not yet followed its develop-
ment. In the series of sections from which this figure is
taken the water-tube and pore-canal are cut transversely ; the
columnar epithelium on one side of the cavity in the section
figured, is continuous with those of the pore- canal and water-
tube, which appear respectively in sections above and below
this. The same continuity of epithelia is seen in Synapta
(fig. 25) in some Echinid Plutei (fig. 9), and in Asterina
(17, fig. 72).

Shortly before the metamorphosis of Auricularia into the
pupa the hydrocoel sends out five pouches (primary tentacles),
and almost immediately afterwards six smaller ones (alternat-
ing with the former), five of which become the five longitu-
dinal water-vessels, while the sixth becomes the primary
Polian vesicle. The groAvth of the hydrocoel into a ring
accompanies metamorphosis.

There is some difficulty in determining the position of the
water-tube in relation to the closing point of the water-vas-
cular ring. Semon (28, p. 196) states that it lies opposite one
of the smaller pouches (longitudinal water- vessels), and on this
he bases his determination that these longitudinal vessels are
interradial, and that the five primary tentacles are radial, and

423

H. BURY.

homologous with the five primary tentacles of other Echino-
derms. His evidence seems to me insufficient ; on p. 197 he
refers to pi. viii, fig. 3, as proving the interradial position of
the longitudinal vessels; yet in this figure the water-tube
(stone-canal) is distinctly adradial, i. e. between a tentacle
and a longitudinal vessel, and no figure whatever is given of a
stage (supposed to precede this) in which the water-tube is
midway between two tentacles. For my own part I have
never seen such a stage as this, but have always found the
water-tube to be adradial from the first, though I do not agree
with Semon as to which adradius it occupies.

In fig. 27 I have given the result of my observations in a
diagrammatic form comparable to Semon’s fig. 3 (pi. viii) :
it will be seen that in my figure we have to cross two primary
tentacles and one longitudinal vessel in passing from the water-
tube to the Polian vesicle ; now, according to one of Semon’s
figures (fig. 3), we have to pass two primary tentacles and two
longitudinal vessels ; but another of his figures (pi. viii, fig.
2), when carefully examined, gives the same results as mine;
and the same position is assigned to the water-tube in Baur's
figures (6). Whatever doubts may exist on this point after
the examination of Auricularia are easily set at rest by
sections through young Synaptee, for in these it is no difficult
matter to ascertain the positions of the water-tube and Polian
vesicle, while the primary tentacles and longitudinal vessels
are exceedingly conspicuous. It is surprising that Semon did
not adopt this method of inquiry, which would also have set at
rest his doubts (28, p. 305) as to which end of the hydrocoel
formed the Polian vesicle.

Summary and Conclusions.

We will now summarize the facts above related, and see
what conclusions can be drawn from them as to the nature
and origin of the liydroccel ; in doing so we shall repeat for
the sake of clearness some of the arguments used on p. 420.

(1) Origin. — The hydrocoel always arises on the left side
as a derivative of one or other division of the coelom.

STUDIES IN THE EMBRYOLOGY OE ECHINODERMS. 429

In 0 phi u rids it is formed from the posterior enterocoel.

In Echinids probably from the anterior enterocoel.

In Asterids (Bipinnaria) from the anterior enterocoel; the
case of Asterina is easily reducible to this.

In Crinoids it and the anterior enterocoel come off together
from the gut, and then separate.

In Holothurians it and the anterior enterocoel are not at
first distinguishable from one another, and are always
connected.

From the fact that in the first three groups there is a period
in which the hydrocoel does not yet exist, though the anterior
enterocoel is already formed, we may assume that the former
is of later phylogenetic origin ; so that in Crinoids and Holo-
thurians, where the two cavities arise together, the hydrocoel
may be regarded as a derivative of the anterior enterocoel.
The frequency with which this origin of the hydrocoel occurs
in ontogeny might seem to indicate that it was of phylogenetic
significance, but it is not easy to see why, if so, there should
be any departure from this condition in Ophiurids, which, in
possessing two well-developed anterior enterocoels, seem some-
what primitive. On the other hand, the variation in the
ontogeny of the hydrocoel may mean that when it originally
appeared the anterior and posterior enterocoels were connected,
as they are in Asterina and some Bipinnarise, and that in
separating them the other groups have adopted different
methods of producing the hydrocoel ; but the fact that the
Bipinnaria in which these two divisions of the coelom are
separated is in other respects primitive, is opposed to this view.
A slightly more satisfactory result is arrived at by regarding
the condition found in Ophiurids as primitive. Then we must
assume that, on account of its physiological importance to the
free-swimming larva, an opening was formed at an early stage
into the anterior enterocoel (before the water-tube could arise),
and that this in time gave rise to the ontogenetic derivation
of the hydrocoel from the anterior enterocoel, such as we find
in all forms except Ophiurids. The fact that this communica-
tion is not kept open in Crinoids till the formation of the
VOL. XXIX, PART 4. NEW SER. F F

430

H. BURY.

water-tube, might be explained by their having compara-
tively recently acquired a large amount of food-yolk, and
passing a very short free-swimming existence. But this is
merely a suggestion, and it is difficult to understand why the
pelagic larvae of Ophiurids should have been able to do with-
out the early connection of the hydrocoel with the anterior
enterocoel, supposed to be of such importance to other larvae ;
or why this connection should not have been made to coincide
with the water-tube (stone-canal) in Asterids as it does in
Echinids and Holothurians. Possibly the true explanation is,
that the hydrocoel originally arose in some more complicated
way, which has since been simplified, independently, by each
of the different groups, and is no longer repeated in the
ontogeny of any one of them.

(2) Connection with the Anterior Enterocoel. — The
hydrocoel never has an external pore of its own, but always at
some time opens into the anterior enterocoel, and so forms an
indirect communication with the exterior ; there are two ways
in which this communication may be established :

(a) In those cases in which the hydrocoel is derived from the
anterior enterocoel, there is, of course, a communication from
the first.

(b) At some period or other, but usually late, a water-tube
(stone-canal) is formed as an outgrowth from the hydrocoel ; it
has a columnar ciliated epithelium.

These communications (a and b) coincide in Holothurians,
and probably in Echinids ; but even in these groups the forma-
tion of a definite columnar epithelium in the water-tube occurs
somewhat late — after the appearance of the primary tentacles
in Holothurians (compare figs. 7 and 27).

In Asterids the two communications coexist but do not
coincide, being in different interradii. The water-tube is formed
after the pouching of the hydrocoel.

In Crinoids (Antedon) the primary communication (a) closes
too early for its exact position to be determined. The water-
tube appears after the primary tentacles.

In Ophiurids the connection (a) never exists at all

STUDIES IN THE EMBRYOLOGY OP ECHINODERMS. 431

The water-tube is formed after tbe hydrocoel has become
lobed.

These facts seem to show that even if the derivation of the
hydrocoel from the anterior enterocoel is of phylogenetic sig-
nificance, there must have been a subsequent time when the
two cavities were entirely separated, otherwise it is difficult to
understand why the primary communication does not always
coincide with the water-tube, as it does in Holothurians and
Echinids, but is sometimes so far distinct from it as to be in a
different interradius (Asterids). It seems more rational to
regard the condition found in the two former groups as secon-
dary (perhaps as a physiological hastening of the connection
between the hydrocoel and the exterior), and to suppose that
the water-tube is a secondary structure belonging to a compara-
tively late stage in the phylogeny of Echinoderms.

(3) Closure of Water-vascular Ring. — Ludwig (14,
p. 45) has already alluded to the variation in the point of
closure of the water-vascular ring with regard to the position
of the water-tube (stone-canal) ; but as I have been able to
collect some data which he did not possess, it will be well to
review the whole matter with some care, and for this purpose
it seems to me that the diagram (fig. 28) will be of more use
than the most detailed description of the facts. For reasons
which will be given later the interradius of the water-tube is
placed anteriorly ; the anterior part of the water-vascular ring
lies on the dorsal side of the oesophagus, while the posterior
part lies beneath it. The positions marked for the closure of
the water- vascular ring rest principally on my own observations,
but that of Asterina is given on Ludwig’s authority, while that
of Ophiurids has been already described by Metschnikoff, whose
account I can confirm. Barrois’s account for Autedon differs
from mine in that he puts the point of closure on the other
side of the water-tube, though in the same interradius (5,
p. 608). The case of Holothurians cannot be settled till we
know whether the radii are marked by the primary tentacles
(Holoth., I), or by the longitudinal vessels (Holoth., II).

Bipinnaria has not been very satisfactorily studied, and it is

432

H. BURT.

not yet certain whether in it the hydrocoel ever forms a horse-
shoe curve shut off from the anterior enterocoel, as it does in
Asterina, or whether the new oesophagus simply grows through
it and perforates it as Metschnikoff asserts (18 and 19).

All this variation is certainly very puzzling ; but if, as seems
to me necessary, we regard the interradius of the water-tube
as a fixed point, we are almost bound to conclude that the
present position of closure of the water-vascular ring is secon-
dary, at any rate in most groups. We have already seen
reason to doubt whether the derivation of the hydrocoel from
the anterior enterocoel, which obtains in the ontogeny of most
groups, is really phylogenetic, and we are now tempted to
ask whether the whole development of the hydrocoel, up to the
time when it forms a complete ring round the oesophagus (the
earliest stage in which all Echinoderms agree), has not under-
gone secondary changes which completely mask its true phylo-
genetic history. It is to be hoped that further investigations
may throw some more light on this point, which at present
forms one of the most insoluble, as well as the most important,
questions in Echinoderm morphology.

III. The Skeleton.

The greater part of the development of the skeleton belongs
to the pentagonal stage, and with this we are not at present
concerned ; nor need we mention the purely larval skeletons of
Ophiurid andEchinid Plutei — only such parts of the per-
manent skeleton as are developed in the Dipleurula will be
considered here.

Ophiurids. — No satisfactory observations have hitherto
been made on the first appearance of the skeleton in this
group. Ludwig (16), in his valuable studies on the skeleton
of Amphiura, found that the radials and terminals were present
before any other plates of the aboral surface, but he was unable
to determine which of these sets was the first to appear.
Fewkes (10), working on the same animal, states positively (p.
139) that the radials appear before the terminals, though he

STUDIES IN THE EMBRYOLOGY OE ECHINODERMS. 433

admits that this statement rests only on the relatively small
size of the terminals in the young pentagonal embryo. On p.
132 he further states that the first and second adambulacral
plates appear before the terminals, but probably after the
radials. I have myself worked out the first appearance of
these plates in Amphiura squamata, and am convinced
that the plates in the bilateral stage, which Fewkes took for
the adambulacrals (10, p. 131, figs. 7, 8, and 10), are really
terminals ; but as the whole subject is far more easily studied
in the various forms of Plutei, I shall begin with a description
of these.

Soon after the formation of the water-tube (stone-canal),
and shortly before metamorphosis, ten skeletal plates make
their appearance simultaneously in the mesoderm surrounding
the posterior enteroccels; five plates accompany each cavity,
and are arranged along it in a straight line antero-posteriorly,
three being dorsal and two ventral, as shown for the left side
in the diagram (fig. 2) ; it will be convenient to state at once
that those on the right side are the radials, and those on the
left the terminals. In some few cases the terminals appeared
before the radials, and several times the dorsal plates of both
series appeared before the ventral. It is possible that these
peculiarities may be constant for certain forms of Plutei, but
of this I have not sufficient evidence. Sometimes simul-
taneously with, but usually some hours later than, the radials
and terminals a plate appears in the middle of the right side,
which is destined to form the dorso-central. Later again than
this there arises another plate on the left side, just in front of
the water-pore; it is the madreporic plate, or first oral (fig. 3).
These twelve plates are all that were ever observed in the
bilateral larva. It is not until after metamorphosis has com-
menced that the adambulacral plates make their appearance.

The development of the plates in Amphiura squamata
so closely resembles that already described, that the details of
it will be postponed to a future paper. The opacity of this
larva, the excessive development of the larval skeleton, and
certain irregularities to which its skeleton seems peculiai’ly

434

H. BUEY.

liable, render it a far more troublesome object for study than
the transparent Plutei.

It is not my intention in the present paper to enter into a
detailed account of the metamorphosis of the Pluteus into a
pentagonal Ophiurid, but it will be well to give a few facts in
support of the statement that the terminals are developed
round the left enterocoel. In fig. 4 we see that the rapid
growth of the right and left enterocoels to meet one another
in the middle dorsal line, has caused the formerly longitudinal
series of plates to become more or less bowed, and the dorso-
central to appear distinctly on the dorsal surface of the larva.
At the same time or rather later, the terminals assume a
peculiar form (see 18, pi. vi, fig. 11, p1 — p3), and over each of
them is developed a marked thickening of the ectoderm.
Without pausing to describe the next stages we will at once
pass on to a much later one, represented in fig. 5. The whole
of the right enterocoel has shifted so far onto the previously
dorsal face of the larva that the fourth and fifth radials (count-
ing from before backwards, as in fig. 2) are visible from the
dorsal side. In correspondence with this the left enterocoel is
passing round towards the ventral surface, so that the three
dorsal terminals lie close to the left margin. Besides this
there has occurred a great shortening of the anterior region
of the larva, and the most anterior terminal has shifted for-
wards and across towards the right, so that it now forms the
anterior median point of the body of the larva (compare 23,
pi. iii, figs. 1, 3, and 4 f1). The hydrocoel, too, has undergone
great changes, which cannot be fully described here, but a com-
parison of figs. 4 and 5 will give some indication of them :
it will be noticed that in the latter the madreporic plate has
shifted its position, and lies anteriorly and to the right, hence
it will be easily understood that that tentacular pouch, which
formerly (fig. 2) lay just behind the water-tube, now lies at
the anterior end of the body, immediately under the anterior
terminal ; while that pouch, which was the most anterior in
fig. 2, and is on the right in fig. 4, now (fig. 2), lies under
the second terminal plate, by which its extremity (“ unpaired

STUDIES IN THE EMBRYOLOGY OF EOHINODERMS. 435

tentacle ”) is about to be embraced. Between the first and
second terminals, the second adambulacral plates are now
visible ; the first pair belonging to this interradius, and both
pairs of other interradii lie at too deep a level to be shown in
this drawing.

If my description has been followed up to this point, it will
be an easy matter to follow the plates, which I have identified
as the radials and terminals, into the stage represented in
fig. 6, and the correctness of the identification will then have
been sufficiently proved.

Asterids. — In the small Bipinnaria seen in fig. 14 the five
terminal plates are already present, though the lobes of the
hydrocoel are not yet developed. In Asterina, however, the
primary tentacles are formed before any skeletal plates appear,
and certain plates in connection with the water-vascular ring
appear as early as, if not earlier than, the terminals. These,
however, need not detain us now.

Fig. 14 shows clearly that the terminals are developed round
the left enterocoel, as in Ophiurids, and I have supplemented
this view of the whole larva by sections through this and
another form of Bipinnaria (B. asterigera), in both of which
this relation of the plates to the enterocoel was quite evident.
The case of Asterina, which offers some difficulties, will be
considered presently.

Up to the present time it has, I believe, been invariably
assumed that the terminals of Asterids and Ophiurids belonged
to the right enterocoel. No attempt, so far as I know, has
been made to prove this for the latter group, but for Asterids
we have the authority of Agassiz (2, p. 32), and it is worth
while to spend a few moments in explaining how his mistake
arose. At the time when he wrote, the ultimate fate of the
two primary enterocoel pouches was still very imperfectly
understood, and he believed that the left pouch (“ left water-
tube,” as he calls it)1 gave rise solely to the water-vascular

1 It will be noticed that Agassiz’ use of the term “ water-tube ” is very
different from that adopted in this paper. Here it is substituted for the

436

H. BURY.

system, while the right pouch formed the whole of the adult
body-cavity. Now, the terminal plates (“ brachial plates ” ol
his description) are undoubtedly formed to the right of the
water-vessel (hydroccel), and hence Agassiz was led to speak of
them as formed round the right {f water-tube ” (enterocoel) ;
and although subsequent writers must long have been aware
that the left enterocoel pouch enters largely into the forma-
tion of the adult body-cavity, yet they have made no fresh
investigation into the relations of the terminals. It is
not a little curious that the dorsal mesentery, separating
the right and left enterocoels, is actually represented in
one of Agassiz’ figures (2, pi. v, fig. 6) to the right of the
terminals.

The next plate to appear is the madreporite. It arises close
to, but nearer the median line than, the water-pore, which
before long it embraces. In one form of Bipinnaria, which I
more than once obtained, it lay in the same straight line with
the terminals ; but usually it is more to the right, as shown in
fig. 14. In the larva from which this figure was taken the
terminals only were present, but I have added the madreporite
from another larva, in which it was precociously developed.
It really belongs to a somewhat later stage.

Seeing that the terminals, both in Asterids and Ophiurids,
belong to the left enterocoel, and not, as hitherto supposed, to
the right, it obviously becomes important to reconsider the
position of the madreporic plate in the two groups. This
plate and the other orals of Ophiurids have always been
assumed to belong to the left enterocoel, and it is, I think,
practically certain that this view is correct, though the only
fresh evidence I can offer of it is derived from the position of
the madreporite in the bilateral larva, in which it lies over the
left anterior enterocoel. The question now arises, is it not
possible that the madreporite of Asterids may also belong to
the left side, and not, as hitherto asserted, to the right ? If
this were so, the position taken up by Ludwig on other

inappropriate expression “stone-canal” (= Steinkanal = canal de sable).
Agassiz applies it to the coelom, and its derivative the hydroccel.

STUDIES IN THE EMBEYOLOGY OP ECHINODEEMS. 437

grounds (15, p. 79), that the madreporic plates of Asterids and
Ophiurids are homologous, would receive new and striking
support.

Agassiz (2) and Gotte (11, p. 620) both describe the basals
of Asterids as formed round the right enterocoel, but as the
value of their testimony is in some degree weakened by their
mistake with regard to the terminals, it will be well to give
some further evidence ; without, therefore, referring to the
numerous arguments to which Ludwig’s suggestion has given
rise, I shall give the result of my own observations on Bipin-
naria and Asterina. In the former most of the basals are
formed late, and I have not obtained specimens which show
their position satisfactorily. At first sight the case above
mentioned, in which the madreporite and terminals are all in
the same straight line, seems to indicate that the former also
belongs to the left side ; but on cutting sections we find that
it does not lie over the enterocoel, but over another cavity
which has not been noticed by previous observers ; this cavity
is situated in the median line, and is utterly unconnected with
the enterocoel in any stage in which I have observed it ; and
though I have not with certainty traced its formation, I
believe it to be of schizocoel origin, like the similarly situated
“ pulsating vesicle ” in Echinids ; it is, however, not contrac-
tile, but contains a few corpuscles which are kept in move-
ment by cilia on the walls of the sac ; it is most probably the
rudiment of the blood-vascular system, but I cannot at present
assert this positively.

Since nothing could be determined as to the position of the
basals from such a larva as this, I next turned my attention to
Asterina gibbosa, in which, as Ludwig had already made
known, all five basals (including the madreporite) appear at an
early period. Here, however, I met with an unexpected diffi-
culty, for it was soon evident that my larvae (obtained at
Naples in May, 1888) did not at all agree in their internal
anatomy with Ludwig’s description. I must therefore, for the
present, ignore his account (accepting only such parts as relate
to the external form, and to the position and nomenclature of

438

H. BTTEY.

the plates1), and briefly describe what I find in larvae of the
seventh and eighth days of development. Fig. 17 is a partly
diagrammatic view of such a larva from the dorsal (and partly
left) side, the larval organ being considered anterior, and the
larval mouth ventral. An oblique mesentery, separating two
cavities, runs across the posterior part of the stomach, and on
the left side of it are seen the three dorsal terminals (compare
fig. 14), while on the right are three basals, the most anterior
being the madreporite. The cavity on the left is identified
from Ludwig’s description as the left enterocoel, and this is
supported by its relations to the terminals ; but the right cavity
is not mentioned by Ludwig. According to his description it
too should be part of the left enterocoel, and he recognises no
mesentery in the position here represented. To confirm this
view of the whole animal, and to determine the relations of this
right cavity, I have cut a number of transverse and longitudinal
sections, and in every case the same result is obtained. This
cavity is entirely shut off from that of the larval organ, and,
indeed, from every other cavity; it begins a little behind the
water-pore, and runs back to the extreme posterior end of the
larva, being separated from the rest of the enterocoel by a ven-
tral mesentery, as well as by the dorsal one here represented ;
it is correctly shown in transverse section by Ludwig (17,
pi. ii, fig. 37) as a comparatively small cavity exactly opposite
to the hydrocoel, but there is nothing in his description which
will enable us to understand how it gets there. I have repre-
sented it again diagrammatically in fig. 16, in which is shown
its relation to the dorso-central and basal plates, and the close
parallelism of the terminals to the mesentery enclosing it. I
have not yet fully traced its origin or subsequent fate, and
until I have done so I am unable to determine its true cha-
racter, or to point out the full extent of the difference between
Ludwig’s account and mine ; but, bearing in mind the posi-

1 This statement requires modification : the madreporic plate is called by
Ludwig the fifth basal, the one immediately behind it being the fourth, and so
on. I begin with the madreporite and count backwards, as in the case of
the terminals (fig. 13).

STUDIES IN THE EMBEYOLOGY OF ECHINODEEMS. 439

tion of the terminals in Bipinnaria (figs. 13 and 14), and that of
the dorso-central in Ophiurid Plutei (figs. 4 and 5), we
can, I think, have little hesitation in identifying the cavity on
the right of the mesentery in fig. 17 as the right posterior
enterocoel. In any case it is quite certain that the basals are
not related to the left enterocoel, as the terminals are ; and I
fully agree with Carpenter (8, p. 386) and Sladen (29, p. 37),
that this affords the strongest possible argument against the
homology of madreporic plate of Ophiurids with that of
Asterids, which Ludwig has attempted to establish.

Crinoids. — Here again, as originally pointed out by Gotte
(11, p. 395), whose account I can fully confirm, we find a
bilateral arrangement of the primary skeletal plates, corre-
sponding to the bilateral division of the enterocoel ; that is to
say, five orals are developed round the left enterocoel, and five
basals round the right enterocoel. The water-pore is at first
unconnected with any plate, but after a while becomes sur-
rounded by one of the orals. In his paper on Comatula,
Barrois (5, p. 634) gives a very different account ; he begins
with the statement that it is a recognised fact that in Eehiuids,
Ophiurids, and Asterids the primary plates are developed
asymmetrically, only on one side of the body. This is difficult to
reconcile with the writings of Carpenter and Sladen above
referred to, and is absolutely opposed to the facts related in
this paper. Then, after pointing out that the plates of
Antedon do not at first reach the ventral side, Barrois attempts
to prove that at the time of the formation of these plates the
right body-cavity is wholly dorsal, and the left wholly ventral,
so that the plates belong solely to the right side. I consider
this supposition to be entirely negatived by the observations of
Gotte and myself (11 and 7).

Eehiuids. — Although certain plates have long been known
to exist in Echinid Plutei, no one, so far as I know, has
traced them into connection with the plates of the young
pentamerous Echinus. On the right side of an advanced
Pluteus of Echinus microtuberculatus we find the
plates shown in fig. 10 ; two of them bear pedicellarke

440

H. BUEY.

(the dorsal plate may have two) as well as one or two
spines, the number of which varies slightly. Just anterior
to the ventral pedicellaria is a third plate bearing a spine
but no pedicellaria. I have ascertained, by means of sec-
tions, that these three plates lie immediately over the right
enteroccel, and the relation to this cavity of the two which bear
pedicellarise is shown in fig. 20.

In one form of larva, of which I obtained but few specimens,
there was also a terminal pedicellaria at the posterior end, and
in Echinus microtuber culatus a small calcareous nodule,
the remnant of the posterior end of one of the larval skeletal
rods, occupies a similar position, and appears to develop later
into a plate. On the dorsal side of a fairly young Pluteus
lies a tri-radiate skeletal rod, which can be seen in fig. 8,
where it is the only part of the larval skeleton represented.
Later on a reticulated plate is formed round the median pos-
terior arm of this rod, and before long envelops the water-pore,
and renders the observation of the subjacent parts extremely
difficult ; it also possesses a spine which is seen in figs. 7 and
10. This plate is clearly the madreporite, and here, as in
Asterids, we are called upon to decide whether this plate
belongs to the right or left enteroccel, since it lies in a median
position over what is, apparently, a schizocoel cavity. This
question seems to me settled, by the fact that the five plates
enumerated become the five basals (genitals) of the adult ; and
since three of these plates are unquestionably formed round
the right enteroccel, it seems reasonable to suppose that the
other two are related to the same cavity.

A comparison of figs. 10, 11, and 12 will clearly show that
the five plates of the Pluteus become the basals of the adult.
In fig. 11, which represents a young Echinus a few hours
after its metamorphosis from the Pluteus, the plates in ques-
tion have changed their positions, but are otherwise not
materially altered; but in fig. 12 a much later stage is shown
in which the pentamerous arrangement is more clearly
marked ; the number of spines and pedicellarise, however, still
remains unchanged, except that one plate has two of the

STUDIES IN THE EMBRYOLOGY OE ECHINODERMS. 441

latter ; but this, as already mentioned, sometimes occurs in the
Pluteus. A large dorso-central is also present.

Besides these five basals several other plates are formed on
the left side of the Pluteus ; at first I imagined that all these
were related to the hydrocoel, but closer examination of them
has led me to suspect that some of them should be regarded
as developed round the left enteroccel ; their numbers and
positions are, however, extremely difficult to determine, and at
present I cannot attempt to describe them. The spines
belonging to some of them are shown in fig. 11.

Hoi othurians. — The only plates in this group which are
known to be present in the bilateral larva are the “ Kalkrad-
chen ” and plates of the water-vascular ring in Auricularia,
but as none of these can be homologised with plates in the
other groups, it is useless to pursue the matter further at
present. Our knowledge of the skeleton of other Holothurian
larvae is sadly deficient.

Summary and Conclusions.

Not only has it been shown in the foregoing pages that
many skeletal plates are developed in the bilateral larva
(Dipleurula), and that they bear a definite relation to the
body- cavities, but the discovery that the terminals lie on the
left side enables us to establish a typical bilateral form from
which all the conditions found in existing larvae may have been
derived ; this typical form has five radial and five interradial
plates on each side, in definite relation to the body-cavities, as
shown in the following Table, in which are also given the
names by which the plates in question are usually known.

r

Position . .

l

Right Enteroccel.

Left Enteroccel.

Radial

Interradial

Radial

Interradial

Name . . . £

Primary

Radials

Basals

Terminals.

|

Orals.

442

H. BURY.

In Ophiurid Plutei the ten radial plates (primary radials
and terminals) and one of the orals (madreporite) are early
developed ; but the basals and most of the orals do not appear
till the pentagonal stage is reached. In Crinoids, on the
other hand, the ten interradials (basals and orals) are the first
to appear, while the primary radials arise late ; no terminals
have yet been recognised, but now that we know where to look
for them it is not impossible that they may be discovered. In
Asterids the terminals are usually the first plates to show
themselves, though in Asterina the basals arise simultaneously
with them ; in other forms the basals (except the madreporite)
and the primary radials are late in appearing, and it is not yet
certain whether the orals are ever developed. In Echinids the
basals appear early, but we know nothing at present of the
primary radials, terminals, and orals ; the ocular plates have
usually been identified as primary radials, but some regard
them as terminals ; embryology alone can decide this question,
and at present my material is not sufficient for it; I am, how-
ever, strongly disposed to believe that the oculars are ter-
minals, and that the primary radials are entirely absent.

The definite relation borne by the plates to the body-cavities
is a fact of great morphological importance, and while it is
absolutely opposed to Barrois’s statement that all the skeletal
plates of Ophiurid, Echinid, Asterid, and Crinoid larvae are
developed round the right enteroccel, it also throws consider-
able doubts on Semon’s sweeping assertion (28, p. 282) that
no homologies are to be found between the primary plates in
the different groups.

Another point of considerable morphological importance
receives great light from the study of the development of the
calcareous plates. If we look at the lateral views of Ophiurid
and Asterid larvae (figs. 2 and 18) we shall see that the plates
on each side may be regarded as forming a longitudinal series
dorsal to the alimentary canal ; since those plates which are
apparently ventral do not reach as far forward as the anus, and
may be conceived to have reached their present position in
connection with a general curvature of the body. The arrange-

STUDIES IN THE EMBRYOLOGY OF EOHINODERMS. 443

merit, in fact, strongly suggests segmentation, but I cannot
discuss in the present paper whether the pentamerism of the
adult Echinoderms arose in this way in the skeleton, or
whether it first made its appearance in the hydroccel; the
latter appears to me more probable. It is further evident
from figs. 2 and 13 that, taking the terminals as marking the
radii, the mouth, anus, and water-pore are at this stage all in
the same interradius. It may be asked, Why should the
terminals rather than the water-vascular pouches be taken to
mark the radii? The fact is that during metamorphosis the
hydrocoel undergoes such extraordinary changes of position
that it is doubtful whether any reliance can be placed on the
position of its pouches in the larva as indicating any per-
manent relations to’ the rest of the body ; at any rate, it is
certain that the water-pore is much more constant in its
relations to the mouth and anus than to the water-vascular
pouches. Thus we know that in Asterina (17) the most
anterior tentacular pouch (in front of the water-pore) is
eventually embraced by that terminal which seems to be
morphologically the most posterior, i. e. the one just behind
the anus in fig. 13. It is possible that this is also the case in
Bipinnaria, but if so, it is the tentacular pouch alone which
shifts its position; for in fig. 15 we see that, after the connec-
tion of the terminals with their respective water-vascular
pouches, the mouth, anus, praeoral lobe, water-pore, and water-
tube still lie in one and the same interradius (see also 2 and 1,
pi. vii, fig. 6). The same figure also indicates what I shall
prove more fully in a future paper, that the left anterior
enteroccel becomes the so-called “ Schlauchformiger Kanal”
of the adult. The Ophiurids are still more remarkable than
Asterina in the behaviour of their hydroccel ; as already
pointed out, the whole hydrocoel is pushed forward and round
the oesophagus in such a way that the tentacular pouch imme-
diately behind the water-tube (most posterior in fig. 2) unites
with the most anterior terminal, while the pouch just in front
of the water-tube unites with the terminal which is nearest
the anus. In this case it is clear that to distinguish the ante-

444

H. BURY.

rior and posterior radii by means of the tentacular pouches,
before they have selected, so to speak, their respective ter-
minals, would only involve us in confusion.

In a previous paper (7, p. 294) I imagined that the arrange-
ment found in Antedon, in which the mouth and anus are
actually in the same interradius in the adult, was arrived at by
a secondary shifting of the anus. This appears to be Ludwig’s
idea (14, p. 54), and is also advocated by Barrois for Antedon
(5, p. 638). Now, however, I am compelled to regard this
position of the anus as primitive, though of course it is still
possible that in this particular group it may have been
secondarily derived from such a condition as is found in
Asterids.

Of the view here advanced, that the mouth, anus, and water-
pore belong primarily to the same interradius, some further
support is afforded by an examination of the Echinids. Turn-
ing to fig. 10, already described, we can without difficulty trace
a series of five basal plates, beginning with the madreporite as
the most anterior, working backwards by way of the dorsal
pedicellaria, and ending up with the plate which bears a spine
but no pedicellaria. These plates are of course interradial,
but by taking tbe interspaces to represent the radii, we shall see
that there is some reason for thinking that here again the
mouth, anus, and water-pore occupy the same interradius.

Another point to be noticed is, that within this interradius the
water-pore and the mouth frequently occupy adradial positions.
In Ophiurids they clearly lie on opposite sides of the same
interradial plate (figs. 2 and 3). The same is true of Echinids,
though it is not evident from my figures ; for in them too the
madreporic plate is at first situated in front and to the right of
the pore. In Crinoids, again, this fact is very distinct (7,
fig. 45), and in these, as in Ophiurids, the adradial position of
the water-pore is long marked by its excentric position in the
madreporic plate.

Among Asterids I have not been able to obtain any clear
evidence of this, for the madreporic plate generally seems to
arise opposite the pore, and just to the right of it (fig. 14), or

STUDIES IN THE EMBRYOLOGY OP EOHINODERMS. 445

even slightly behind it (fig. 17). It is to be noticed that in
Asterids, as well as in Echinids, the madreporic plate loses its
relation to the body-cavity, which the other four basals possess,
as if it were dragged out of its natural position in order to
embrace the pore.

Owing to our imperfect knowledge of the development of the
skeleton in Holothurians we are at present unable to trace in
this group the relation of the mouth and anus to the radii.

Enough has, however, been said to make it probable that in
all groups (except perhaps Holothurians) the radii of the
abactinal part of the body (including the regions of the right
and left posterior enterocoels) bear a very definite relation to
the mouth, anus, and water-pore of the larva ; that, in fact,
these organs mark out an interradius which, since it contains
both mouth and anus, might be called ventral, or, as it is an-
terior to the system of radial plates and contains the prseoral
lobe (where this is present), may be called anterior. The latter
term seems to me preferable, since we can with less confusion
apply it to the adults, though, of course, in seeking for an ante-
rior interradius in them, we must be guided by the situation
of the water-pore rather than by the indefinite and variable
positions of the mouth and anus.

VOL. XXIX, PART 4. NEW SER.

G G

446

H. BURY.

LITERATURE.

1. A. Agassiz. — “ Revision of the Echini,” Cambridge, U. S., 1872-74.

2. A. Agassiz. — “North American Starfishes,” ‘Mem. Mus. Comp. Anat.

and Zool.,’ Harvard, vol. v, No. 1, 1887.

3. N. C. Apostolides. — “Anatomie et Developpement des Ophiures,”

‘Arch. Zool. Exp. et Gen.,’ vol. x, p. 121, 1882.

4. F. M. Balfour. — ‘ Treatise on Comparative Embryology,’ vol. i, London,

1880.

5. J. Barrois. — “ Becherches sur le Developpement de la Comatule (C.

mediterranea),” ‘Recueil Zool. Suisse,’ vol. iv, No. 4, p. 546,1888.

6. A. Baur. — ‘ Beitrage zur Naturgeschichte der Syn apt a digitata,’ 2te

Abhand., Dresden, 1864.

7. H. Bury. — “The Early Stages in the Development of Antedon

rosacea,” ‘ Phil. Trans.,’ 1888, vol. 179.

8. P. H. Carpenter. — “ Notes onEchinoderm Morphology,” No. V, ‘Quart.

Journ. Micr. Sci.,’ vol. xxii, p. 371, 1882.

9. J. W. Fewkes. — “Preliminary Observations on the Development of

Ophiopholis and Echinarachnius,” ‘ Bull. Mus. Comp. Zool.,’ Harvard,
vol. xiii, No. 4, p. 105, 1885-86.

10. J. W. Fewkes. — “ On the Development of the Calcareous Plates of

Amphiura,” ‘ Bull. Mus. Comp. Zool.,’ Harvard, vol. xiii, No. 4,
* 1887.

11. A. Gotte. — “ Yergleichende Entwicklungsgeschichte der Comatula

mediterranea,” ‘Arch. f. Micros. Anat.,’ vol. xii, p. 583, 1876.

12. M. M. Hartog. — “The True Nature of the Madreporic System of

Echinodermata, with remarks on Nephridia,” ‘Ann. and Mag. Nat.
Hist.,’ vol. xx, 5th ser., p. 321, 1887.

13. A. Kowalevsky. — “ Entwicklungsgeschichte des Ampliioxus lanceo-

latus,” ‘Mem. Acad. Imp. des Sciences de St. Petersbourg,’ series
vii, vol. xi, 1867.

14. H. Ludwig.— “Ueber den primaren Steinkaual der Crinoideen, nebst

vergleichend-anatomischen Bemerkungen fiber die Echinodermen fiber-
haupt,” ‘ Morph. Stud.,’ vol. ii, p. 34, Leipsic, 1880-82. (And ‘ Zeitsch.
f. wiss. Zool.,’ vol. xxxiv, p. 310, 1880.)

15. H. Ludwig. — “Neue Beitrage zur Anatomie der Ophiuren,” ‘Morph.

Stud.,’ vol. ii, p. 57. (And ‘ Zeitsch. f. wiss. Zool.,’ vol. xxxiv, p
333, 1880.)

STUDIES IN THE EMBRYOLOGY OP ECHINODERMS. 447

16. H. Ludwig. — “ Zur Entwicklungsgeschichte des Ophiurenskelettes,”

‘ Morph. Stud.,’ vol. ii, p. 91. (And ‘ Zeitsch. f. wiss. Zool.,’ vol.
xxxvi, p. 181.)

17. H. Ludwig. — “Entwicklungsgeschichte der Asterina gibbosa,”

‘Morph. Stud.,’ vol. ii, p. 111. (And ‘Zeitsch. f. wiss. Zool.,’ vol.
xxxvii, 1882.)

18. E. Metschnikoff. — “ Studien liber die Entwicklung der Echinodermen

und Nemertinen,” ‘ Mem. de l’Acad. Imper. de St. Petersb./ vol. xiv,
ser. 7, No. 8, 1869.

19. E. Metschnikoff. — “ Embryologische Mittheilungen liber Echinodermen,’

‘ Zool. Anz./ Jahrg. 7, p. 43, 1884.

20. Joh. Muller. — “ Ueber die Larven und die Metamorphose der Ophiuren

und Seeigel,” ‘ Abh. d. k. Akad. d. Wissensch. zu Berlin aus den
Jahre 1846/ Berlin, 1848.

21. Joh. Muller. — “ Ueber die Larven und die Metamorphose der Holo-

thurien und Asterien,” * Abh. d. k. Akad. d. Wissensch. zu Berlin aus
den Jahre 1849/ Berlin, 1850.

22. Joh. Muller. — “Ueber die Larven und die Metamorphose der Echino-

dermen,” ‘ IVte Abh. d. k. Akad. d. Wissensch. zu Berlin a. d. Jahre
1850/ Berlin, 1852.

23. Joh. Muller. — “ Ueber die Ophiurenlarven des adriatischen Meeres,”

‘Abh. d. k. Akad. d. Wissensch. zu Berlin a. d. Jahre 1851/ Berlin,
1852.

24. H. Prouho. — “Recherches sur le Dorocidaris papillata,” ‘Arch. d.

Zool. Exp. et Gen./ vol. v, 2nd ser., p. 213, 1888.

25. E. Selenka. — “ Zur Entwicklung der Holothurien,” ‘ Zeit. f. wiss. Zool./

vol. xxvii, p. 155, 1876.

26. E. Selenka. — “ Keimblatter und Organanlage der Echiniden,” ‘Zeit. f.

wiss. Zool./ vol. xxxiii, p. 39, 1879.

27. E. Selenka. — “ Keimblatter der Echinodermen,” ‘ Stud. iib. Entwick. d.

Thiereu/ Wiesbaden, 1883.

28. R. Semon. — “ Die Entwicklung der Synapta digitata, und die Stam-

mesgeschichte der Echinodermen,” ‘Jen. Zeitsch. f. Naturwiss./ vol.
xxii, new ser., xv, 1888.

29. W. P. Sladen. — “ On the Homologies of the Primary Larval Plates in

the Test of the Brachiate Echinoderms,” * Quart. Journ. Micr. Sci./
vol. xxiv, new. ser., 1884.

30. J. W. Sfengel. — “ Zur Anatomie des Balanoglossus,” ‘Mitt. a. d. Zool.

Sta. Neap./ vol v, p. 494, 1884.

448

H. BURY.

EXPLANATION OF PLATES XXXVII, XXXVIII,

& XXXIX,

Illustrating Mr. H. Bury’s paper on “ Studies in the Embry-
ology of the Echinoderms.”

Reference Letters.

Ant. Anterior. Post. Posterior. Dors. Dorsal. Vent. Ventral. R.
Plight. L. Left.

Fig. 1. — Dorsal view of a young Ophiurid Pluteus, showing the
arrangement of the enteroccels and the origin of the hydroccel. x 300.

Fig. 2. — Diagrammatic view of an Ophiurid Pluteus, from the left
side.

Fig. 3. — Part of an Ophiurid Pluteus, dorsal view, x 300.

Fig. 4. — Dorsal view of an Ophiurid Pluteus, just before metamor-
phosis. X 180.

Fig. 5. — Dorsal view of an Ophiurid Pluteus, undergoing metamor-
phosis. The terminal plates are really far more complicated in their structure
than is here represented. X 300.

Fig. 6. — Young Ophiurid still retaining some of the arms of the Pluteus.
X 300.

Fig. 7. — Diagrammatic view of the left side of an Echinid Pluteus.
X 180.

Fig. 8. — Dorsal view of an Echinid Pluteus, showing the arrangement
of the enterocoels and the origin of the hydroccel. X 180.

Fig. 9. — Part of a Pluteus of Echinus microtuberculatus, seen from
the dorsal side. X 300.

Fig. 10. — Pluteus of Echinus microtuberculatus, seen from the right
side, x 180.

Fig. 11. — Young Echinus microtuberculatus, a few hours after
metamorphosis from the Pluteus. The calcareous plates at the bases of the
marginal spines are omitted. X 180.

Fig. 12. — Plates at the aboral pole of a much older specimen of Echinus
microtuberculatus (diam. 75 mm.), x 75.

Fig. 13. — Diagrammatic view of the left side of the same larva. X 100.

Fig. 14. — Dorsal view of a Bipinnaria. x 100.

STUDIES IN THE EMBRYOLOGY OF EOHINODERMS. 449

Fig. 15. — Diagrammatic view of the left side of a large Bipinnaria, the
larval arms and most of the prseoral lobe being cut away. Outline drawn with
camera lucida. X 20.

Fig. 16. — The same, from the right side. X 100.

Eig. 17. — Larva of Asterina gibbosa on the seventh day of develop-
ment, seen from the dorsal side. X 100.

Fig. 18. — Part of the left side of an Ophiurid Pluteus, seen from the
dorsal side, showing the origin of the hydroccel. X 510.

Eig. 19. — Transverse section of an older Ophiurid Pluteus, passing
through the posterior part of the oesophagus, x 540.

Fig. 20. — Transverse section of a Pluteus of Echinus microtubercu-
latus, showing the relation of the pedicellarise to the right body-cavity.
X 180.

Fig. 21. — The same part of another abnormal Auricularia, dorsal view.
X 180.

Fig. 22. — Part of an abnormal Auricularia, lateral view. X 180.

Fig. 23. — Dorsal view of the same part of a normal Auricularia. Treatment
with chloral-hydrate has caused retraction of the pseudopodia of the meso-
derm cells. X 180.

Fig. 24. — Lateral view of the same. The mesoderm cells are omitted.
X 180.

Fig. 25. — Part of a transverse section through an Auricularia, just entering
into the “pupa” stage. X 300.

Fig. 26. — Part of a transverse section of a larva of Cucumaria Plauci.
X 300.

Fig. 27. — Diagram of the hydrocoel of an old Auricularia, seen from the
dorsal side.

Fig. 28. — Diagram of the closure of the water-vascular ring in different
groups of Echinoderms, dorsal (aboral) view.

.

Madrtpt

K&M c%fzw< 't?A9¥zum’

Ant

Orscphxgus

) Ant EnterocceJ.

T)-\l ’

Yatcnporc

v . Stomach

\ Poll.

Post cnt fro cool.

Fig 10.

Li'h&Imr '-amt Sci Inst Co

rnj.sM. xl.

RESPIRATORY ORGANS IN DEOAPODOUS CRUSTACEA. 451

On the Ancestral Development of the Respira-
tory Organs in the Decapodous Crustacea.

By

Florence Buchanan,

(A Paper read to the Biological Society of University College, London.)

With Plate XL.

In the Crayfish, so well known to all students of zoology,
our attention is attracted to the varying positions of the gills,
and, in reading Professor Huxley's book, especially devoted to
this Crustacean, we are led to compare the different positions
of these organs and their relations to each other in the dif-
ferent segments in many of the Decapods. In so doing the
question arises as to how they have come to be situated as
they are, and it is to this question that I propose, in the
following paper, to attempt an answer. It must, I am afraid,
be only a suggestion as to the real answer, since several details
in the development of the various forms would have to be care-
fully worked out in order to prove the whole theory true.

For a great many of the ideas and facts cited I am indebted
to a paper by Professor Claus in the ‘ Wiener Arbeiten * for
1886. Other facts I have derived from the account by Professor
Sars of the Schizopods brought up in the “ Challenger ” expe-
dition. (The figures also are mostly from both these sources.)

In order to explain the present positions of the branchiae
of such forms as Astacus, it is needful first to find out what
their past positions were, and for this we look not only to the
early history of the individual, but also to the early history of
the race.

The Chaetopod-like ancestor of the Crustacea probably had

452

FLORENCE BUCHANAN.

no special respiratory organs, but the vessels of the vascular
system were distributed equally on the whole surface of the
body, thus enabling the liquid they contained to absorb
oxygen through the thin wall separating them from the water
in which the creature lived. This vascular surface then be-
came concentrated, and it would naturally be concentrated to
those parts of the body which are most brought in contact
with fresh supplies of oxygen, and, consequently, of water.
Thus when certain limbs became especially modified for swim-
ming, it was the parts of the body-surface behind those limbs
that first became especially vascular and branchial. It was then
of advantage to the animal to have this vascular surface in-
creased ; the skin therefore, at or on the base of the swimming
appendage, became folded, and we find it thus as a simple plate
in the nearest living representatives of the Crustacean ancestor,
namely, the Phyllopods, as exemplified by Apus. The typical
Apus thoracic appendage (PI. XXXVII, fig. 1) consists of a
basal or axial portion (protopodite), sometimes divided into
two, three, or even four parts, of six “ endites,” and of two
<e exites,” viz. a flabellum and a bract. Of these the fifth
endite probably represents the endopodite of the Crayfish limb ;
the sixth represents the exopodite; the flabellum or large
swimming plate the epipodite ; and then we have immediately
behind the swimming plate a flattened fold of the skin or bract,
shown by Lankester to be devoid of muscles and by Claus to
be of different constituency to the rest of the limb by the
rapidity with which it stains with dilute osmic acid. It is this
that is, it seems probable, homologous with the branchiae of the
Decapod. Anyway this vascular respiratory outgrowth is quite
independent of what corresponds to the epipodite in the Cray-
fish or Lobster (see fig. 9). It is worthy of note that in
the oostigite, where the flabellum is especially modified for
carrying the eggs, and therefore not used for swimming, the
bract is very rudimentary. Whether it is vestigial (i. e. the
remains of a bract that was once well developed) or rudi-
mentary (i. e. that it never was more developed), depends, of
course, on whether the special modification of the oosti-

RESPIRATORY ORGANS IN DEOAPODOUS CRUSTACEA. 453

gite took place after or before this respiratory fold was
established.

The primitive position, therefore, of the respiratory organ is
behind the swimming organ. We shall see that this is the
case in the Schizopods, Stomapods, and also in the higher
group of the Isopods. The Archimalacostraca, however, which
developed from the same source as the Phyllopods (i. e. from
an Archi-entomostracous form), probably had not a settled
respiratory organ ; this might or might not be present. For
we find in Nebalia, which of all living forms most nearly repre-
sents the Malacostracan ancestor, that there are no special
branchial organs. The epipodites (fig. 2) of the thoracic
limbs, as in the Entomostraca, and as in no other Mala-
costracan forms, are still used for swimming, and probably
move rapidly enough to keep the surface of the body sufficiently
well supplied with oxygen.

While the genus Nebalia has been modified but slightly
from its original form, another form, very closely allied to it,
and having probably behind its swimming epipoditic plates
plates or folds of the skin for branchial purposes, has been
subject to more severe competition, and has become changed
and modified in many different ways. This form probably
had the typical Malacostracan number of segments and appen-
dages (nineteen and a telson, the telson being possibly divided
externally, so as to give the appearance of either one or two
extra segments, as in Nebalia), and an elongated heart. From
this a more stable form has developed, which we may call the
“ Archischizopod,” not, however, until it had given rise to a
form which became the ancestor of the Stomapoda. The
Archischizopod, when formed, would have acquired a fixed
number of segments, and would differ chiefly from the Archi-
malacostraca in its different mode of swimming; for it is no
longer the epipoditic plate which is the swimming organ, but
the other primitive branches of the Crustacean limb, and espe-
cially the exopodite, which is developed and modified for this
purpose. Consequently the epipodite, having lost its primitive
signification, and at this stage being of no special advantage to

454

FLORENCE BUCHANAN.

RESPIRATORY ORGANS IN DECAPODOUS CRUSTACEA. 455

the animal, may or may not be present. The respiratory
organs would, as usual, be developed at the bases of the swim-
ming appendages, i. e. on all those of the thoracic segments.
Here, as in Apus, they do not represent the epipodites, and
consequently in all normal Schizopod forms we find either the
epipodite and branchia coexisting (see fig. 5) or the branchia
alone left, the functionless epipodite having disappeared, or
possibly, in exceptional cases (Bentheuphausia), having been
formed at the same time, and so nearly in the same position as
the gill as almost to have become part of it.

Most nearly related to the Archischizopod of living forms
are the Euphausidae, all of which have branchiae attached to
the bases of their thoracic appendages, and called “podo-
branchiae.” One of the lowest of these, Thysanopoda (fig. 3),
has thoracic limbs very closely resembling those of Nebalia
(fig. 2) only with gills, and with no epipodites. These gills
are developed from plate-like outgrowths at the base of the
appendages, which gradually become branched ( Resp . Org. 1),
while behind each there arises another branch (Resp. Org. 2),
which rapidily becomes more complex. Thysanopoda is typical
of most of the Euphausidae, but in one form, Nematoscelis
(fig. 4), we find that this second branch (Resp. Org. 2), which in
form exactly resembles that of other Euphausidae, is attached
not to the same joint of the limb as the first but to the next
joint (i. e. to the basopodite instead of to the coxopodite) . This, I
take it, is merely due to the fact that the gill developed while the
coxopodite and basopodite were undivided, and the two branches
were carried apart by the separation of these two joints, and
appear, therefore, as separate branchiae. I have not been able to
find any actual mention of this being the case, but, on compar-
ing the limb of such a creature as Thysanoessa or Thysanopoda
with that of Nematoscelis, no other explanation seems possible ;
and, considering the small size of the embryo at the stage
when the gills are developed, the point as to whether the
two basal joints are united or not at this period may have been
easily overlooked or not thought worthy of notice. If it is
once admitted that this change of position is due to the exact

456

FLORENCE BUCHANAN.

time at which the gill is developed, and to the condition of the
appendage at that time, I think we can account for all the
different positions of branchiae in the higher Crustaceans.

In some of the Euphausidae the branchiae of the hinder
thoracic appendages may be highly complex, the second branch
developing other branches on it, but only in Nematoscelis are
these branches apart from each other. In all they are unpro-
tected, and not covered by the carapace, the secondary
branch usually reaching over on to the ventral surface of the
animal.

All the Euphausidae undergo a free metamorphosis, hatching
in the Nauplius condition, and passing through a number of
stages, varying from four to eight, before reaching their adult
and permanent form. The first traces of gills appear in all
cases to arise soon after, and in some cases even before, the
development of the limb to which they are finally attached.
This is all that Sars (who gives a nearly full account of the
development of the Euphausidae found in the “ Challenger”
expedition) states about the development of the gills. Their
appearance before the leg to which they are attached seems
to point to the existence at an earlier period of an epipodite
which was used in swimming, and which has now disap-
peared. In all the Euphausidae the endopodite develops
before the exopodite, as is not the case with the higher
Crustaceans.

The structure of the gill of the Euphausidae (figs. 3, 4),
is very simple, consisting merely of branching lobes with no
secondary branches.

In the Lophogastridae, which are also Schizopods, but of
a higher order than the Euphausidae, the primary lobes of the
stem are themselves lobed, and thus the gill is more complex
in structure (fig. 5). The final ramifications may be either
foliaceous, as in Lophogaster itself, or vesicular as in Gnatho-
phausia (Sars). The arrangement of the gills is also different :
instead of being attached to the limb itself, each complex gill
is attached to the arthrodial membrane near the base of the
appendage, as first stated by Boas and afterwards confirmed by

RESPIRATORY ORGANS IN DEOAPODOUS CRUSTACEA. 457

Sars. Boas inferred from this that the gill was therefore quite
a different structure to that of the Euphausidse, but Professor
Claus (in the paper I have already referred to) has shown that
it would be absurd on account of mere difference of position
to call the two gills separate structures. The development of
the Lophogastridse has not as yet ever been thoroughly made
out, but it seems probable that the position of the gills on the
arthrodial membrane is due to the fact of their developing at
a later period than in the Euphausidse after the full formation
of the appendage, and to the subsequent sinking of the coxo-
podite into the body wall. This also seems likely from the
fact that the Lophogastridae have no longer a free metamor-
phosis, and therefore it is of no advantage to the embryo to
have its gills developed early. Each of these gills is compound,
consisting of either three (Lophogaster) or four (Gnathophausia)
distinct lobes springing from the same point. If the arthrodial
membrane to which they are attached were to be stretched we
can easily see how these four lobes would be pulled apart, and
would thus assume the appearance of separate gills. The
variability of the boundaries of the arthrodial membrane is
shown by Professor Claus in the development of one of the
Brachvurous larvae to which I shall afterwards refer (p. 461).
Of the lobes of the gill three are covered in by the carapace,
and the fourth projects freely beneath the trunk, meeting its
fellow in the middle line (fig. 10).

Let us now see what has become of the epipodite during this
development. In the Euphausidae, in all the forms mentioned
by Sars (of which there are six or seven genera), this is entirely
absent, although Sars, and also Boas, regards the gill as re-
presenting the epipodite. Only in one form mentioned by
Sars (Bentheuphausia), which has a very complicated gill, it is
difficult to say whether epipodite is or is not present as well as
gill. Since, in the higher Crustacea, we have in so many
instances the epipodite present as well as, and perfectly distinct
from, the gill, it seems probable that this vascular outgrowth,
although in outward form very closely resembling an epipodite,
is really to be regarded as an independent organ rather repre-

458

FLORENCE BUCHANAN.

senting the bract than the flabellum of Apus. The develop-
ment of the gill may have led to the suppression of the
epipodite when this first lost its primary significance, without,
however, representing it either in function or structure
but in position only. The absence of the epipodite in the
Euphausidse, therefore, seems to show that, on account of its
loss of function, it has tended almost entirely to disappear.
In the Lophogastridse the epipodite, on all the hinder thoracic
limbs (vii — xm), is either absent or rudimentary, being, when
rudimentary, a projection quite independent of the gill and
attached to the basal joint of the limb (fig. 5). On the maxil-
lipede, however (vi), in all the Lophogastridse, the epipodite is
well developed, and on this appendage there is no gill to cause
its suppression. It has been retained, probably, because it has
acquired a new function, namely, that of producing movement
of the water in the branchial cavity into which it projects, and
thus keeping the gills well supplied with oxygen. This lanceo-
late epipodite, with the same function, is also present in the
third group of the Schizopods, the Mysidse — where, how-
ever, there are no gills. According to Professor Claus, the
Mysidse probably once had gills and were much larger animals
than they now are. (They are now usually only one third or
sometimes two thirds of an inch in length.) On account of
this reduction in size the gills have been lost, and some forms
have acquired peculiar foldings of the integument round the
bases of the thoracic limbs, which probably have branchial
functions. These are covered by the carapace, underneath
which the water is kept in continual motion by the long
epipodite of the sixth appendage. The other thoracic append-
ages, besides having no gill, also have no epipodite. The
presence of the epipodite of the sixth appendage, in both
Mysidse and Lophogastridse, and of the rudimentary epipodite
in some of the Lophogastridse, seems to show that they
sprung from the Euphausia-stem before the epipodites were
entirely lost.

The next group that we come to, the De capo da, of which
the Crayfish and Crab are well-known examples, originated

RESPIRATORY ORGANS IN DECAPODOUS CRUSTACEA. 459

either still lower on the Euphausia stem, since epipodites (see
fig. 9) are often found on all the thoracic feet as well as gills ;
or, as seems more probable, from some earlier form of the
Lophogastridae which had epipodites, represented more fully
than they now are, on all the thoracic limbs. In reference to
this matter, the reader is referred to the genealogical tree on
p. 4, which is nearly the same as that given by Claus with
reference to other structures as well as the branchiae. He
does not, however, discuss the special relations of the separate
Schizopod groups either to each other or to the Decapods.

In the Decapods the thoracic feet have no longer a swim-
ming function. Consequently one branch of the biramose
limb (the exopodite) has either become vestigial or is alto-
gether wanting. The three anterior thoracic limbs have now
become maxillipedes, instead of only one as in the Schizo-
pods, while the hinder thoracic appendages (ix — xm) have an
ambulatory function. The swimming function is therefore
left to the swimmerets or appendages of the abdomen, which in
the Schizopods were in all cases very small although as a rule
present. In most of the Decapods these swimmerets also
attain no very great size, and in the long-tailed forms the
telson with the two appendages of the penultimate segment is
largely used in swimming, while the short-tailed forms
scarcely swim at all. This change in function of the appen-
dages does not, however, affect the gills, since these have
already become fixed to the thoracic region in the Schizopod
stage, which is gone through both phylogenetically and onto-
genetically by the Decapod. One change, however, though
apparently a slight one, does affect them, and this is the
increase in length of the epimeral walls, as may be seen by
comparing the two diagrammatic sections of Gnathophausia
(fig. 10) and Astacus (fig. 11).

In consequence of the raising of the pleura the epimeral
walls, and with them, it appears probable, the arthrodial
membrane at the base of the appendages, has become
stretched. Thus the gills from being situated close together
have become separated. But we find an indication of their

460

FLORENCE BUCHANAN.

being formed close together in the larval form of one of the
Decapods, Calliaxis (fig. 7), where we see two of the branchiae
(b and c) being formed almost from the same spot. How far
they are apart from each other depended in all probability
originally upon the exact time at which the special branch
developed in relation to the stretching of the membrane cover-
ing the joint of the limb. When their position had become
definitely established they finally, in most cases, but yet not
quite in all, as we see from the Calliaxis larva, developed
straightway in their respective places even when developing
at the same time. Calliaxis and the forms allied to it (i. e. all
the Thalassinidte) never have gills on the epimeral wall,
though they have very well-developed ones on the arthrodial
membrane ( b and c) as well as on the epipodites (a). This as
well as the structure of the gills seems to poiut to a more
ancestral condition than that of most other Decapods.

The advantage in the branchise being situated immediately
on the bases of the appendages has ceased to exist, as it had
also in the higher Schizopods, for the branchiae have become
more complicated (in comparison with the simple ones of the
Euphausidse), and require protection from surrounding ob-
jects, though, of course, still requiring continuously to be
bathed with fresh supplies of water. Already in the Lopho-
gastridse we find three branches of the compound gill covered
in by the carapace (fig. 10), while the fourth is bent over on
the ventral surface, and is thus also to some degree protected
from being hurt by anything with which the creature comes in
contact. In the Decapods we find all the gills, even those
attached to the basal joint of the limb, protected by the cara-
pace (fig. 11), and so closely closed in by it in most cases as to
lie in a special branchial chamber through which water is
driven, as in the Crayfish, by the continuous movement of
the scapliognathite or exopodite of the second maxilla. This
covering in of the gills for protection may have been accom-
panied by a tendency to vary in the position assumed, and
when once varied natural selection may have favoured the
variation.

RESPIRATORY ORGANS IN DECAPODOUS CRUSTACEA. 461

. The position of the different branchiae with regard to each
other in the same segment tends to vary very greatly.
Huxley has classed the different kinds of branchiae, in com-
paring them with those of the typical Astacus, as podobranchs,
anterior and posterior arthrobranchs and pleurobranchs.
Claus, however, has shown that these names do not, in all
cases, apply, but that, owing to the undefined limits of the
arthrodial membrane, they should vary in the different
families if they are to be strictly correct. Thus what Huxley
calls “ posterior arthrobranchs ” he calls “ anterior pleuro-
branchs •” since in Penseus, which is probably a more ancestral
form than Astacus, the third branchiae (fig. 6, c) are attached
to the body wall and not to the arthrodial membrane in the
adult. Claus holds that in the ancestral Decapod the distal
branchia was, as it still is, a true podobranch. The middle
one was probably also a podobranch, which in the shortening
of the coxopodite has moved to the arthrodial membrane ;
while the two proximal ones were attached near the base of
the appendage either to the membrane or to the body wall.
Claus also explains how an arthrobranch may become a pleuro-
branch by the moving of the arthrodial membrane away from
the body wall along the limb, so that the proximal portion of
the membrane may become part of the pleural wall. That a
podobranch may become an arthrobranch is shown by the
condition in the larva of one of the Brachyurous Decapods —
Acanthocaris (fig. 8), as well as in that of Pengeus (fig. 6, a
and (3 ), where the second gill ( b ) is developed on the basal
portion of the limb, and only afterwards becomes moved
backwards to the arthrodial membrane. All such variations
as these seem to point to an earlier approximation in position
of the gills, so that from a compound four-fold gill, not unlike
that of one of the Lophogastridge (fig. 5), the various gills of
the Decapod may be derived.

The structure of the gill itself may also, as shown by Claus,
be derived from that of a Schizopod, and both the typical
forms of gill observed in Decapods may be so derived. The
Crayfish and the greater number of long-tailed Decapods

VOL. XXIX, PART 4. NEW SER. H H

462

FLORENCE BUCHANAN.

have feathery filamentous gills known as t richobranchs.
The development of such a gill from a Schizopod gill is seen
in the individual development of Stenopus, a form closely
allied to Penaeus. Here a gill consisting of a shaft with
two opposite rows of rays is first formed, resembling the
gill of one of the Euphausidae (fig. 3). These rays, instead of
becoming lobed, become longer and narrower, and other new
rows of lobes appear on the shaft, which in their turn increase
in length and decrease in width. These secondary rays do not
spring so regularly from the shaft as the primary ones. Thus
a typical trichobranch is formed.

The other kind of gill, that of the Crab and most short-tailed
Decapods, as well as of some of the long-tailed forms (e.g.
Palaemon), consists of a stem on which are lamella-like plates
lying upon each other like the leaves of a book. Such a form
is known as a phvllobranch, and its derivation is seen in
the individual development of Penaeus (Claus). Here the pri-
mary rays lengthen and grow round on the side away from the
body, so as to enclose a sort of canal running parallel to the
shaft and open at the ends. Secondary rays ariseonthe outer side
of the primary ones, i. e. projecting into the canal, in a single
row turned towards the base of the gill. These may split so
as to appear as though they sprang separately. The flattening
of these secondary rays into leaf-like plates and the enlarge-
ment of the primary ones would give rise to a phyllobranch.

According to phylogenetic development the podobranch is
the most ancestral of the gills, and it is therefore, at first
sight, surprising that in the adult Penaeus (a form which is so
very typical of the whole Decapod group) no trace of podo-
branchs is to be found, while the pleurobranchs are well de-
veloped. Looking, however, to the larva of Penaeus, we see
that podobranchial rudiments are developed (fig: 6), and,
indeed, they are the first to develop, while the rudiments of
pleurobranchs ( d ) develop in a later larval stage than all the
others (fig. 6'). Whilst, however, the arthrobranch and pleuro-
branchs (6, c, d) go on developing the podobranchial rudi-
ments (c) disappear, the epipodites, which develop later, being

RESPIRATORY ORGANS IN DEOAPODOUS CRUSTACEA. 463

left alone on the basal joint of the leg. In which, although
very peculiar, appears to be another form, very closely allied
to Penseus (Cerataspis), the podobranchs go on developing
with the rest, and are present in the adult on nearly all the
thoracic segments (vii-xii) as well as, and very close to, the
epipodites.1 It is therefore probable that Penseus, in the course
of phylogenetic development, has lost its podobranchs. This
loss, I think, can be explained by the fact that Penseus hatches
at an earlier period than Cerataspis and other Decapods, for in
comparing the different gill formulae of the Decapods, espe-
cially those mentioned in Professor Huxley’s book on the
Crayfish, we notice that it is in forms in which, like Astacus,
the young is hatched only when fully developed, that the podo-
branchs are the most fully and the pleurobranchs the least
fully developed.

Thus we see that in Astacus and in all the Astacidae very
nearly the full number of podobranchs is present, while the
number of pleurobranchs varies, there sometimes being none
at all, sometimes one or two rudimentary ones and one well-
developed one as in Astacus.

In Homarus, and those Decapoda macrura whose young
are hatched rather earlier than in Astacus, more pleurobranchs
are developed. Penseus, which has lost its podobranchs, and is
well supplied with pleurobranchs, is, you will remember, the
only Decapod which hatches in the nauplius or earliest larval
form. This at first seems not to be in accordance with what we
find in the only other Malacostracan forms which hatch in the
nauplius condition, namely, the Euphausiidse, where, as we have
already seen, the gill is always attached to the base of the limb,
and is truly podobranchial, though in the other Schizopods (Lo-
phogastridse), which hatch at a later stage, these have moved to
the arthrodial membrane. But when we take into account the
more delicate structure of the gill of these higher Malacostra-
cans, and the consequent need of protection, this want of
agreement can, I think, be fully explained. From the ancestor

1 See Dolirn “Untersuchungen iiber Bau und Entwickelung der Arthro-
poden,” ‘ Zeitsclir. f. wiss. Zool.,’ vol. xxi, 1871 (fig. 32).

464

FLORENCE BUCHANAN.

of Penaeus, which probably had its podobranchs well developed,
forms like Astacus, Homarus, &c., are to be derived. These
carry on the ancestral development within the egg, and the
different gills develop in the ancestral order, the pleurobranchs
being formed last, and therefore often being not needed by the
time the creature hatches. In a form like Penaeus, however,
which has continued to hatch in an early ancestral form, the
podobranchs which are formed originally at an early stage, and
before the carapace has grown down to cover them, are wholly
unprotected, and therefore apt to get more harmed than those
branchiae which develop later, and are more shielded by the
carapace. Thus it has come to have been of more advantage
to the embryo not to develop these outer gills, and natural
selection has favoured those forms which do not develop them,
though we still find indications of their having once been
present in that the rudiments are found in the embryo.

It would take too long to go into the different branchial
formulae of all the different groups of Decapods, but I think
that, taking into account the stretching of the arthro-
dial mem brane and the time at which it took place,
the need of protection to the branchiae, the condi-
tion of the larva when hatched, and probably also
the condition of the tissues of the creature (some
tissues requiring more oxygen for the maintenance of the
individual than others), we can explain all the various
positions of the branchiae found. One group that I
might mention particularly is that of the short-tailed Decapods
or Brachyura. These, as you will remember from the instance
of the Crab, all have a very much reduced number of branchiae.
If we look to the development of these forms we find that it is
very much hurried, and that at the stage in which the gills are
developed the embryo is so cramped that its thoracic legs
appear to spring one above the other on the sides of the body
wall. This would easily account for the suppression and the
irregularity of the suppression of some of the gills. Turning
our attention to the epipodites of the Decapods we find that
these are as a rule present on all the thoracic segments, and it

RESPIRATORY ORGANS IN DECAPODOUS CRUSTACEA. 465

is near their base that the podobranchs spring. The epipodite
has not disappeared as it has in the Euphausidse, nor become
rudimentary as in the Lophogastridse : it has acquired a new
function, and is of sufficient importance to be preserved. In
Pengeus it probably has the function of keeping the gills clean,
but in higher forms (Homarus, Astacus) this office is performed
by special setse attached to the coxopodite of the appendage,
and the epipodite has been transformed into a broad lamella
which serves to separate the gills to some extent and prevent
their entanglement. In most cases, as in Homarus (fig. 9), the
podobranch remains quite distinct from this lamella ; but in
the case of Astacus and some allied forms, all of which inhabit
fresh water, and are hatched only when fully developed, the
podobranch exists as a tuft on the epipodite, which at its
extremity is known as the “ lamina,” and there are branchial
filaments, exactly similar in structure to those of the tuft, on
the epipodite itself. This condition is probably due to the
fusion of the two organs on account of the small compass in the
egg for the development of each separately. Such fusion is to
be found in the early life of other Decapods besides the Asta-
cidae, e.g. Calliaxis (fig. 7) and Calocaris. In the larvae of
both these forms the podobranch has the appearance of being
merely a differentiated portion of the epipodite, while in the
adults the two organs are easily distinguishable from one
another, although the gill still remains attached to the epipodite
and does not spring independently from the protopodite, as I
have been able to verify from a specimen of Calocaris which,
owing to the kindness of Mr. Pocock, of the British Museum,
I have been allowed to examine. It appears probable that the
simultaneous development of the two organs almost on the
same spot has caused their fusion : in Calliaxis and Calocaris,
where the larva is free-swimming, separation has soon taken
place, though the indication of a common origin is maintained;
in Astacus and its near allies, on the other hand, where develop-
ment continues in the egg coverings and, consequently, in a
much limited space, separation takes place to a very small
extent only. This separation is at the fore end, and the greater

466

FLORENCE BUCHANAN.

parts of the two organs remain fused throughout life, giving the
appearance of an epipodite forming gill filaments. In some cases
(Astacoides) separation never takes place at all. In Homarus,
although there is the same tendency of the two organs to be
formed together, as shown by the attachment of the podo-
branch to the base of the epipodite, yet, as the larva has
become free-swimming before differentiation takes place,
separation takes place at the same time, and the podobranch
never is fused, except just at its base, to the epipodite.

To refer now briefly to the other groups of the Malacostraca,
which I have until now put aside in considering the histori-
cally most interesting group of the Decapods, we come first to
the Stomapods, of which Squilla is an example. These prob-
ably are to be derived (see classif.) from a Malacostracan form,
whose swimming and respiratory organs were not yet fixed to
the thoracic region as they are in the Archischizopod. We
know that the Stomapod does not undergo the same changes
in development as the forms with which we have hitherto been
dealing, but that, instead of the midbody being developed last
as it is in the Schizopod and Decapod, this becomes developed
before the hind body. The thoracic appendages therefore
develop early, and probably before any special respiratory
apparatus begins to be needed, while the abdomen is only
afterwards developed, and its appendages become the chief and
most active swimming oi’gans. The swimming function, there-
fore, which in the Archischizopod is the part of the thoracic
limbs, is here undertaken by the abdominal appendages ; and,
as in the Schizopod, the gills have developed behind the
thoracic swimming appendages, so in the Stomapod they
have developed behind the abdominal swimming appendages
and are present as branchial tufts attached to the exopodite,
not in any way representing an epipodite. It is worthy of
note that it is not only the respiratory organs but also the
heart, generative organs, &c., in the Stomapods that develop
in the abdominal instead of in the thoracic region. This
probably has also to do with the reversion in the development
of the two regions.

RESPIRATORY ORGANS IN DECAPODOUS CRUSTACEA. 467

The Cumacea are probably degenerated from forms not far
removed from the Archischizopod, and have only one gill
remaining.

The Arthostraca, comprising the two groups of Amphipods
and Isopods, are also probably to be derived from the Archi-
schizopod, but having from the beginning taken a different
line of descent from the true Schizopods. The Amphipods
(of which Talitrus is a well-known example) have a plate-like
outgrowth serving for respiration at the base of each thoracic
limb. This resembles the bract of Apus, and very probably
represents the branchia of the Archischizopod, and is therefore
the homologue of the Schizopod and Decapod gill. The
Isopods have lost all traces of gills in their thoracic append-
ages, this being probably owing in some degree to their general
modifications to suit a terrestrial life. The branchial function,
as an after development, has been undertaken by one branch
(the endopodite) of the appendages that are used in swim-
ming, namely, the abdominal appendages. This endopodite
has therefore developed branchial filaments, which, however,
bear no relation whatever to the branchial tufts on the abdo-
minal appendages of the Stomapods.

Thus, the positions of the respiratory organs in the different
groups of the Crustacea are, to some extent, explained,
although in some instances very imperfectly. I have not
gone into the relations of the numerous groups of Decapods
nor quoted their branchial formulae. Their relation to each
other is, however, very fully given by Claus, and this paper
merely offers a suggestion as to how the different formulae may
be explained, whilst its chief purpose is to draw the attention
of my fellow-students to an interesting field of morphological
theory and observation.

Before concluding, I must thank Professor Lankester for the
help he has given me in showing me how to treat the subject,
and in referring me to the memoirs which I have cited.

■

INDEX TO YOL. XXIX,

NEW SERIES.

Actiniaria, two new types of, by
Fowler, 143

Amphibians, development of, by Orr,
295

Amphioxus, contributions to a know-
ledge of, by E. Ray Lankester, 365

Bahamas, pelagic organism from, 1

Beard on the development of the
peripheral nervous system of Ver-
tebrates, 153

„ on the parietal eye of Cyclo-
stome fishes, 55

Beddard on three new species of
Earthworms and on morphology of
Oligochseta, 101

„ on Urochmta and Dichogas-
ter, and on nephridia of Earth-
worms, 235

Bury on the development of Echino-
derms, 409

Blastopore, fate of, in Rana tem-
poraria, by Sidebotham, 49

Buchanan on the gills of Decapod
Crustacea, 451

Crustacea decapoda, the gills of,
by F. Buchanan, 451

Cuttle-fishes, by Weiss, 75

Cyclostome fishes, parietal eye of, 55

Dendy, studies on Sponges (Stelo-
spongus), 325

Dichogaster, by Beddard, 235

Earthworms, nephridia of, by Bed-
dard, 235

„ three new species of,

101

Echinoderms, development of, by
Bury, 409

Eye, parietal, of Cyclostome fishes,
55

Fat-bodies of Rana, by Giles, 133

Fowler on two new types of Actiniaria,
143

Fungia, natural history of, by Lister,
359

Giles on fat-bodies and pronephros of
Rana, 133

Haplodiscus, by Weldon, 1

Lankester, E. Ray, on Amphioxus,
365

Laurie, on the organ of Verrill in
Loligo, 97

Lister on the natural history of Fungia,
359

Loligo, organ of Verrill in, 97

470

INDEX

Minckin on a new organ in Peripla-
neta, 229

Nephridia of Earthworms, by Bed-
dard, 235

Nervous system, development of, by
Beard, 153

Oigopsid cuttle-fishes, by Weiss, 75

Oligochseta, morphology of, by Bed-
dard, 101

Ornithorhynchus, Poulton on the true
teeth and horny plates of, 9

Orr on the development of Amphibians,
295

Parietal eye of Cyclostome fishes, 55

Peripatus novse-zealandise, de-
velopment of, by Lilian Sheldon,
283

Periplaneta, a new orgau in, and hy-
podermis of, by Minchin, 229

Poulton on the teeth of Ornithorhyn-
chus, 9

Pronephros and fat-bodies of Rana,
by Giles, 133

Rana, fate of blastopore in, 19
„ temporaria, fat-bodies of, by
Giles, 133

Sheldon on the development of Peri-
patus novse-zealandise, 283

Sidebotham, fate of blastopore in
Rana, 19

Sponges, studies on, by Arthur Dendy,
325

Stelospongus, by Dendy, 325

Teeth of Ornithorhyncus, by Poulton,
9

Urochseta, by Beddard, 235

Weiss on some Oigopsid cuttle-fishes,
75

Weldon, on Haplodiscus, 1

PEINTED BY ADLAED AND SON, BAETHOLOAIEW CLOSE.

9