FOR THE PEOPLE
FOR EDVCATION
FOR SCIENCE

LIBRARY

OF

THE AMERICAN MUSEUM
OF

NATURAL HISTORY

.*• «

I S- h' .

aUARTERLY

OF

JOURNAL

MICROSCOPICAL SCIENCE

EDITED BY

E. RAY LANKESTER, M.A., F.R.S., F.L.S.,

Fellow of Exeter College, Oxfords, and Professor of Zoology and Comparative
Anatomy in University College, London ;

WITH THE CO-OPEEATION OF

WILLIAM ARCHER^ F.R.S., M.R.I.A.,

Dublin.

F. M. BALFOUR, M.A., F.R.S., F.L.S.,

^ Fellow and Lecturer of Trinity College, Cambridge.

AND

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

Lecturer on Histology in St. Sartholometv's Hospital Medical School^ London,

VOLUME XX.— New Seeies.
lllwstralions on anb

L O B O N :

J. & A. OHURCIllLL, NEW BUETJNGTON STREET.

,C . '

^ '■

'' '* \

! A ^

CONTENTS

CONTENTS OF No. LXXVII, N.S., JANUARY, 1880.

MEMOIRS:

PAGE

On the Embryo-sac and Development of Gymnadenia compsea.

By H. Marshall Ward, Scholar of Christ’s College, Cam-
bridge. (With Plates I, II, and III) . . . I

Studies on the Pollen-Bodies of the Angiosperms. By Fred.

Elfving, of Helsingfors. (With Plate IV) . . .19

On the Development of the Conceptacle in the Fucaceee. By
F. 0. Bower, B.A., Trinity College, Cambridge. (With

Plate V) 36

On Certain Effects of Starvation on Vegetable and Animal
Tissues. By D. D. Cunningham, M.B., Surgeon, Indian
Medical Service, Fellow of the Calcutta University . . 50

On the Development of the Spermatozoa. Part I. Lumlricus.

By J. E. Bloomfield, B.A. Oxon. (With Plates VI and VII) 79
On the Spinal Nerves of Araphioxus. By F. M. Balfour, M.A.,
F.R.S., Fellow of Trinity College, Cambridge . . .90

The Bacillus of Leprosy. By G. Armaner Hansen. (With
Plate VIII) 92

NOTES AND MEMORANDA :

Development of Planorbis ..... 103

Origin of Sperm and Ova from the Cell-layers of Coelentera . 104

Muscular Fibre (Mesoderm) derived from Endoderm . . 106

Recent Researches on Bacteria ..... 106

Publications of the Zoological Institute of the University of
Vienna ....... 110

New Biological Journal ...... 110

PROCEEDINGS OP SOCIETIES :

Dublin Microscopical Club , . , , .111

b

IV

CONTENTS.

CONTENTS OF No. LXXYIII, N.S., APEIL, 1880.

I^IEMOIRS :

PAGE

The Coffee-leaf Disease of Ceylon. By W. T. Thiselton
Dyer, F.L.S., Assistant Director, Royal Gardens,

Kew. (With Plates IX, X, XI, XII, XIII, and XIV) . 119

On Shepheardella, an XJndescribed Type of Marine Rhizopoda ;
with a Pew Observations on Lieberkiihnia. By J. D. Siddall.

(With Plates XY and XVI) 130

Development of the Kidney in its relation to the Wolffian Body
in the Chick. By Adam Sedgwick, B.A., Scholar of Trinity
College, Cambridge ; Demonstrator in the Morphological La-
boratory. (With Plates XVII and XVIII) . . . 146

Notes on the Development of the Araneina. By P. M. Balfour,

M.A., P.R.S., Fellow of Trinity College, Cambridge. (With
Plates XIX, XX, and XXI) 167

A Contribution to the Biology of Bacteria. By Dr. L. Wald-
STEIN, of the Pathological Institute, Heidelberg . .190

Some Teachings' of Development. By E. Schafer, F.R.S.,
Pullerian Professor of Physiology ^ . . . . 202

On the Histology of Hydra fusca. By T. Jeffery Parker,

B.Sc., Professor of Natural History in the University of Otago 219

The Orthonectida, a New Class of the Phylum of the Worms.

By Alfred Giard, Professor in the Faculty of Sciences of
Lille. (With Plate XXII) 225

NOTES AND MEMORANDA:

The Origin of the Red Corpuscles of Mammalian Blood . 241

I. On the Mode in which Hydra Swallows its Prey. By M.

M. Hartog, M.A., B.Sc., ’F.L.S., of the Owens College,
Manchester ...... 243

II. Additional Note on Hydra. By the Same . . 243

III. On the Anal Respiration of the Copepoda. By the Same . 244
Dr. G. von Koch’s Method of Preparing Sections of Corals . 245

CONTENTS.

V

CONTENTS OP No. LXXIX, N.S., JULY, 1880.

MEMOIRS :

PAGE

On the Structure and Homologies of the Germinal Layers of
the Embryo. By F. M. Balfour, M.A., F.R.S., Fellow
of Trinity College, Cambridge .... 247
Hubrecht’s Researches on the Nervous System of Nemertines.

(With Plate XXIII) . . . . ^ . .274

On the Structure of the Nephridia of the Medicinal Leech.

By A. G. Bourne, Assistant in the Zoological Laboratory
of University College, London. (With Plates XXIV and

XXV) 283

On Intra-Epithelial Capillaries in the Integument of the Me-
dicinal Leech. By E. Ray Lankester, M.A., F.R.S., Pro-
fessor of Zoology in University College, London. (With

Plate XXVI) 303

On the Connective and Vasifactive Tissues of the Medicinal
Leech, By E. Ray Lankester, M.A., F.R.S., Professor
of Zoology and Comparative Anatomy in University College,
London. (With Plates XXVII & XXVIII) . . 307

On the Use of the Wenham Binocular with High Powers. By
Heneage Gibbes, M.B. ..... 318

On the Structure of the Human Spermatozoon. By Heneage
Gibbes, M.B. ...... 320

Some Disputed Points in Echinoderm Morphology. By P.
Herbert Carpenter, M.A., Assistant Master at Eton
College ....... 322

The Origin of the Red Blood-Corpuscles. By Professor Pouchet,
of the Jardin des Plantes, Paris . . ... 331

On Limnocodium (Craspedacustes) Sower biiy a New Tracho-
medusa inhabiting Fresh Water. By E. Ray Lankester,

M.A., F.R.S. (With Plates XXX and XXXI) . . 351

NOTES AND MEMORANDA :

On the Development of the Structure known as the ‘ Glome-
rulus of the Head-Kidney’ in the Chick. By Adam Sedg-
wick, B.A. . . , . . . .372

Bacterium anthracis ...... 374

Dr. Carl Rabl on the Pedicle of Invagination in Pulmonate
Gastropoda ....... 376

Development of Muscular Tissue from Epiblast in the Mammalia 377

PROCEEDINGS OF SOCIETIES :

Dublin Microscopical Club ..... 378

VI

CONTENTS.

CONTENTS OF No. LXXX, N.S., OCTOBER, 1880.

MEMOIRS:

PAGE

Larval Forms : their Nature, Origin, and Affinities. By F. M.

Balfour, M.A., F.R.S., Fellow of Trinity College, Cambridge 381 •
On the Classification of Cryptogams. By Alfred W. Bennett,

M.A., B.Sc., F.L.S., Lecturer on Botany at St. Thomas’s
Hospital ....... 408

A Reformed System of Terminology of the Reproductive Organs
of the Cryptogamia. By Alfred W. Bennett, M.A., B.Sc.,

F.L.S., Lecturer on Botany at St. Thomas’s Hospital ; and
George Murray, F.L.S., Assistant Botanical Department,

^ British Museum . . . . . .413

\J On the Laminar Tissue of Amphioxus. By Professor Pouchet.

(With Plate XXIX) . . . . . .421

The Peripheral Nervous System in Paloeo- and Schizonemertini,
one of the Layers of the Body -wall. By Dr. A. A. W.
Hubrecht, of Leyden. (With Plates XXXII and XXXIII) 431
The Eye of Pecten. By Sydney J. Hickson, B.Sc., Scholar
of Downing College, Cambridge. (With Plates XXXIV and

XXXV) 443

On the Terminations of Nerves in the Epidermis. By L. Ranvier,
Professor at the College de France. (With Plate XXXVI) . 456
On the Termination of the Nerves of the Mammalian Cornea. By
E. Klein, M.D., F.R.S., Lecturer on Histology at St. Bar-
tholomew’s Medical School. (With Plate XXXVI) . .459

Histological Notes. By E. Klein, M.D., F.R.S., Lecturer on
Histology at St. Bartholomew’s Medical School . .476

REVIEWS :

Atlas of Histology. By E. Klein, M.D., F.R.S., and E. Noble
Smith, L.R.C.P., M.R.C.S. . . . . .480

Freshwater Rhizopods of North America. By Joseph Leidy.

M.D., Professor of Anatomy in the University of Pennsyl-
vania, and of Natural History in Swarthmore College, Penn-
sylvania ....... 480

A History of the British Marine Polyzoa. By Thomas Hincks,

B.A., F.R.S. . . . • . . . .481

NOTES AND MEMORANDA :

Medusa and Hy droid Polyps living in Fresh Water . .488

On the Respiration of the Crustacea .... 485

Index

. 487

MEMOIRS.

On the Embryo-sac and Development of Gymnadenta
CONOPSEA. By H. Marshall Ward, Scholar of Christ’s
College, Cambridge. With Plates I, II, and III.

The hitherto received account of the origin, mode of for-
mation, and homologies of the embryo-sac in Angiosperms
was based on the researches of Hofmeister,^ for the most
part, and may be summed up to the following effect: — A
single cell situated in the apical part of the nucleus, and
usually the foremost of an axial row, enlarges, its nucleus
disappears, and a variable small number of free nuclei
appear in the protoplasm by free cell formation. Of these,
two are close to the apex of the enlarged cell, and are
known as germinal ” or embryonic ” vesicles, while a
very inconstant number of antipodal ” cells often, but not
always, forms in the lower end of the protoplasm. The
enlarged cell is the embryo-sac, and receives the pollen tube
at its apex, where one, or at times two, of the germinal vesi-
cles become fertilised by the contents of the tube. Schacht^
showed that one of the germinal vesicles acted as a
conductor between the tube and the fertile vesicle, or that
two acted as such towards a third germinal vesicle, and
were often marked by peculiar longitudinal striae.

This ‘‘ filiform apparatus ” and fertile germinal vesicle
were regarded as the homologues of the corpuscula^ of the
Gymnosperms by Strasburger and Pringsheim, and therefore
were regarded as representing rudimentary archegonia, in
accordance with Hofmeister’s views as to the rosette ” or
iieck-cellsin Conifers.'* The antipodal cells came to be regarded
as the last and fleeting representatives of the endosperm,
which arises in the embryo-sac of Coniferm, &c., and which

* “Neuc Beitra^e,” &c., in ‘ Abliandluiig d. Kouigl. siiclis Gescllsch. d.
AViss.,’ 1859 and 1801.

‘Jahrb. f. Wiss, Bot./ 1857.

® R. Brown (1834;), ‘ Misc. Bot. Works.’

* ‘ The Higher Cryptogams,’ Ray Soc., 1802.

VOL. XX. NEW SEK.

A

2

H. MARSHALL WARD.

had been recognised by Hofmeister as the homologue of a true
prothallium.i It was also generally admitted that the endo-
sperm of Angiosperms had nothing to do with that of Coni-
fers, but rather corresponds to what is found appearing late
in the developing macrospore of Selaginella, and independent
of the prothallium.2

Quite lately the necessity for a reconsideration of these
points has been forced upon us by the researches of Stras-
burger,^ of Warming/ and the papers of Vesque/ which
clearly showed that in a great number of types the embryo-
sac does not arise so simply as Hofmeister thought, and
further, that the processes going on between its formation
and the origin of an embryo are still more complex.

Before touching upon this or examining the different
views of the authors named, I propose to state what is found
to occur in Gymnadenia conopsea^ one of the commoner
Orchids of Europe. This has been partly done by Stras-
burger,® who has, however, more especially described the
processes in Orchis pollens and Monotropa hypopitys. The
later stages have not, to my knowledge, been before de-
scribed.

Except where stated otherwise, the method employed has
been to cut sections of the ovaries which have lain in abso-
lute alcohol at least twenty-four hours, and in a slowly
evaporating mixture of alcohol and glycerin at least twenty-
four hours longer. In a few cases staining with fuchsine
and acetic acid has been resorted to successfully. As is
well known, the ovules arise from three parietal placentae,
as outgrow^ths of a few cells projecting into the cavity
of the inferior ovary. Hofmeister traced the Orchid ovule to
one cell of the surface, whence arose the idea that it is
morphologically a trichome."^

At a period shortly before the flower bud is complete, trans-
verse sections of the ovary reveal the young ovules as straight
or very slightly curved structures, arising in considerable
numbers from each placenta and directed at right angles to
its surface. Each consists of an axial row of large polygonal
cells, surrounded by one layer of somewhat smaller cells. In
transverse section we see one large cell with five or six

^ Loc. cit.

2 Cf. also Sachs’ ‘ Text Book,’ Eng. Trans.

^ *Ueber Befrucbtuug und Zelltlieiluug,’ 1877.

“ ‘Ann. d. Sc. Mat. Bot.,’ 1878.

^ Ibid.

® Op. cit., and ‘ Die Gymnosp. und Angiosp,’ 1879.

^ ‘ Cf. also Sachs’ * Lehrbuch.*

EMBRYO-SAC OF GYMNADENIA CONOPSEA.

3

around it. The cells are thin-walled, and full of slightly
granular protoplasm, without any vacuoles, and each has a
large spherical nucleus with a bright nucleolus in the
centre.

At an early stage the terminal cell of the central series
increases considerably in size, its protoplasm becomes very
granular, and its nucleus remarkably bright. About the
same time certain cells in a zone above the middle of the
whole ovule divide by walls parallel to the periphery of the
ovule and form the rudiment of the inner integument. At this
period the ovule is represented in fig. 1, where, however, the
protoplasm has been mostly removed by amnionic hydrate.

It will be useful to name the central series of cells the

axial row,” and the large terminal cell of this row has
been called the embryo’sac mother-cell,” since the em-
byro-sac apparently arises as a daughter-cell by division of
this, as pointed out by Strasburger.^ Reasons will be given
later for considering the embryo-sac as possibly less simple
than here stated.

As the ovule lengthens its curvature increases, the integu-
ment grows upwards, and the embryo-sac mother-cell becomes
longer (fig. 2) ; this latter cell now divides by a transverse
wall appearing at about one third its length from the top,
and the smaller cell thus cut off appears like a cap sitting on
the larger one below. The lower and larger cell is mean-
while growing, and tends to compress this cap-cell against
the epidermal cells above. When it has reached a size about
equal to that with which it started, the larger cell repeats
the above division exactly as before, and so a second cap-cell
is thrown off, soon to be forced up against the other by
growth of its sister-cell below.

The specimen represented in fig. S shows this second
division in progress, the characteristic “ barrel figure” having
been fixed by absolute alcohol. A similar stage is figured
by Strasburger (‘ Angiosp. u. Gyrnnosp.,’ Taf. vi, fig. 90),
where, however, the division appears to follow more rapidly
upon the first one. The division walls thus established are
remarkably thick and bright, as if mucilaginous and swollen,
and the appearance they present is not easily indicated in a
drawing, especially at stages a little later, when the whole
becomes more homogeneous as the lower cell compresses the
uj)])er ones.

Meanwhile the ovule has become decidedly anatropous,
and the inner integument is already slowly closing in above ;
the outer integument has also appeared, by similar divisions

^ Loc. cit.

4

H. MARSHALL WARD.

of the outer cells of a zone, at the base of the inner integu-
ment (cf. figs. 2, 3, 4). The lower of the three cells pro-
duced by division of the “embryo-sac mother-cell” becomes
longer, and in its growth compresses, not only the two cap-
cells, but also the epidermal layer on all sides. It appears
to absorb the contents of these cells, and the exhausted
remains easily suffer compression to mere refractive homo-
geneous masses,' as their greedy neighbour grows. Since
this cell apparently becomes the embryo-sac we shall here-
after speak of it as such ; it has now again become about
as long as at first, and proportionately wider, and its granu-
lar protoplasm contains a large nucleus.

During the progress of the changes about to be described
the two integuments must be pictured as growing up, and
the inner one closely investing the nucleus and forming a
micropyle above ; the outer integument never closes in so
much, but leaves a large loose opening above in later stages.
Not many cells are formed, but as the embryo-sac becomes
completed the integument cells elongate very much, large
sap cavities form, and the nuclei are driven to their walls.
This is especially the case with the outer integument, and
at last results in the formation of a large air space between
the peripheral cells and the base of the nucleus and inner
integument (cf. figs. 10, 14, 30, &c.)

About now, or a little later, the pollen-tubes appear in the
upper part of the ovary as silver' threads, creeping slowly
down the tissue in the groove formed between the swollen
placentas and the carpellary walls. They may be few or
many, and in the latter case can be detected by the naked
eye on the walls of the ovary if torn open. Owing to the
prevalence of very cold and wet weather this summer I
cannot consider the question settled, but believe that the
further vigorous development of the embryo-sac and contents
depends on the presence of these tubes. Since insects were
rare during the bad weather I pollinated many flowers by
hand, and certainly got more vigorous ovules from the spikes
so treated, and there were many tubes in these ovaries. In
cases where the flowers had been pollinated naturally, how-
ever, the contents of the sac and the cells of the integument
were brown, and weakly developed. Selecting what appear
to he normal cases, the following changes occur in the
embryO'Sac, which by its growth has considerably compressed
the layer of cells around it, and is capped by the remains of
the two upper cells (figs. 5, 6, 7, &c.).

The protoplasm collects into two masses, one at each end
of the embryo-sac, leaving the centre of the sac filled by

EMBRYO-SAC OF GYMNADENIA CONOPSEA.

O

more fluid material for some time these masses are ill
defined, and they never form a distinct cell-wall, but, finally,
from the presence of a nucleus and well-marked contour, w^e
recognise in them naked cells in a protoplasmic matrix

Each of these soon divides into four smaller but otherwise
similar nucleated masses, by two divisions in planes crossing
at right angles. In some cases, at any rate, this occurs
simultaneously in both, but in others one or both divide first
into two, and then again each daughter mass into two others
(cf. figs. 8—12).

That this is the final result there can be no doubt, but
whether one is to suspect variations in the divisions as due
to manipulation, conditions of vigour, or peculiarities of the
species, must remain at present undecided. The important
fact is that eight nucleated masses of protoplasm result,
more or less isolated and complete, in groups of four at each
end of the sac (figs. 12, 13) ; this I consider demonstrated
by the specimen figured at 12, though the division of the
lower mass does not always appear to be completed.^
The relations of the products of division in the fore part
of the sac are, however, remarkably constant and surprising.
Two of them become elongated and packed close into
the top of the embryo-sac, as the Gehiilfinnen or “ Sy-
nergidae ” of Strasburger ; while one enlarges and rounds
off* as the egg-cell or oosphere,^^ and becomes suspended
laterally at the base of the “ Synergidae ” in the cavity of
the sac. The fourth mass also rounds off, and falls freely
into the lining protoplasm of the sac. All these masses have
acquired nuclei, that of the egg-cell being especially large
and bright, but no trace of a cell-wall appears around any of
them.

As to the fate of the products of division in the posterior
end, I can say very little positively. That four masses com-
mence to form appears certain, but only in a few cases have
they become completed ; as a rule, I find a mass of protoplasm
in this place with a variable number of nuclei in it, but in
some cases (fig. 13) four masses occur.

It is to be regretted that this point has not been more
successfully dealt with, and also that the exact origin of the

* Many facts suggest that this division is of the same order as the two
preceding divisions of the “en»bryo-sac motlier-cell.” If so, we must con-
sider that cell as suffering division into four, the last division wall being a
very weak and diffluent one.

’ The lower group may apparently suffer less complete division in other
plants also.

6

H. MARSHALL WARD.

embryo-sac nucleus in Gymnadeniareinains uncertain from my
drawings. According to Slrasburger, the typical process is as
follows — The fourth nucleus from the group in the anterior
end of embryo-sac travels down, and meets one of the four
from the posterior end, and, fusing together, these two form
one large nucleus — the nucleus of the embryo-sac. The
three nuclei left behind are ‘‘ antipodal cells or Gegen-
fusslerinnen.” This process I have failed to observe
directly in Gymnadenia, but may remark in this connection
upon figs. 14, 15, 16, and 17, in each of which are indica-
tions possibly of some such process.

In the first case (fig. 14), we have an ill-defined mass of
protoplasm in the posterior end of the sac, and two large
well-rounded nuclei close by the egg-cell above though
the bare possibility exists that one of these nuclei is on the
adjacent wall of an integument cell, since the case is not
isolated, yet it is offered here. In fig. 15 we have a large
bright nucleus below, and a faintly marked (badly preserved?)
body on the side wall of the sac. In fig. 16, both above and
below, appear good round nuclei. In fig. 17 the upper
nucleus has evidently travelled down, and now abuts upon
the mass of protoplasm at the base of the sac ; this mass is
well rounded, and presents four nuclei, or, more correctly,
two large nuclei, in a comm.encing stage of division.

Were it not for the fact that in Strasburger’s drawings
and description this process is so definitely put, and also that
in other cases (esp. Ranunculus, Lobelia, Anthericum,
Butomus, and Alisma),® I have often seen two nuclei free
in the sac besides the antipodal cells and a normal egg appa-
ratus, the above evidence would not deserve to be so insisted
upon ; it is incomplete, and is little bettered by the sugges-
tion that in these highly specialised plants a process of
reduction has become still more reduced, and that even the
rudiments are unusually imperfect and uncertain. However
this may he, in Gymnadenia conopsea there are formed a
normal “ egg- apparatus,” a large nucleus of the embryo-sac,
and a group of “ antipodal cells.”

The pollen- tubes should now be somewhere in the neigh-
bourhood, and in fig. 17 we have an example showing the

^ ‘ Ueber Befruchlung u. Zelltheilung/ p. .32, Orchia pa liens ^ Com-
pare Vesque’sdiflerent account of Orchis galatea, ‘Ann. des Sc. Nat.,’ 1878.

^ Is it possible tliat this is a case of two egg-cells? Strasburger figures
two embryos in the same sac, in a paper published in Jen. Zeitschr. f. Wiss.,
1878 — “ Ueber Polyernbryonie.”

^ In Butomus there can now be no doubt of the fusion of two nuclei.
I have every stage in the process. Sec a paper read before Linnsean Society,
Nov. 20th, 1879.

EMBRYO-SAC OF GYMNADENTA CONOPSEA.

7

act of fertilisation ; the pollen-tube, after a sinuous course
from the placenta, has made a sharp bend ere plunging into
the micropyle, and has then spread its broad apex over the

Gehiilfinnen,” apparently penetrating between the sac and
integument, but the difficulty of tracing so delicate an out-
line as it here presents is no ordinary one.* It does not
break through the top of the sac, but the well-marked
contour of the latter is to be seen through the transparent
tube, the walls of which are here marked by delicate longi-
tudinal striations, which appear to be folds, either caused by
contraction from reagents, or the closing in of the inner
integument ; I cannot identify it with the filiform apparatus
of Schacht, since it is on the pollen-tube, and has nothing
whatever to do with the embryo-sac contents. In the course
of the tube are developed the peculiar cellulose blocks pro-
jecting inwards from its walls, and serving apparently to
shut off the contents of the tube as it grows ; these “ Propfen
or stoppers were described by Strasburger^ and Elfving,^ and
occur in such quantities that a cross section of the pollen-
tube bundles appears marked here and there by waxy-
looking drops interspersed among silvery-like cellular walls
of the tubes. In the figure is one of these stoppers repre-
sented as it occurred in the tube just ere the final bend.

The outline of the embryo-sac is still marked by the
remains of the nucleus-cells which it compressed and des-
troyed in its growth, but the cap-cells appear to have quite
disappeared ; even the latter, however, persist for a long time
(cf. figs. 6 — 13) as a refractive cap on the apex of the sac, in
some cases (fig. 12) presenting a conical, or even beaked
appearance. If no pollen-tube enters the micropyle, the
whole ovule turns brown, shrivels, and the contents of the
sac become ill defined and decay ; the egg- cell persists
apparently longer than the Gehulfinnen (fig. 17, a). In all
these cases, and up to a much later period, we find the
remaining cells of the central or axial row persist beneath
the sac ; in fact, the air space already referred to is formed
beneath the lower one, and between it and the outer integu-
ment when the sharp bend is established.

The fact that the oosphere is fertilised is marked by the^
appearance of a thin cellulose coat around it; it elongates,
and its nucleus prepares to divide. The contents of the Gehul-
finnen and other parts of the sac, on the contrary, become

* Ilofmeisler represents the tube as entering the sac in Orc/a's morio.

‘ Vergl. Unters.,’ t. iv.

^ Loc. cit.

* ‘ Jtnaischer Zcitschr. fiir Naiurwisscnsch,’ 1878.

8

H. marshall ward.

cloudy and prepare for degeneration, and final disappear.
Very shortly the remains of the Gehiilfinnen form a lens-
shaped cap between the embryo and the apex of the sac,
while the decaying antipodal mass presents a similar appear-
ance below (cf. figs. 18 — 22)} In fig. 21 the appearance of
four nuclei in the antipodal mass — and a similar condition of
things is vaguely indicated in other cases, as fig. 19 — suggests
that the union of a fourth nucleus with one from above to
form the embyro-sac nucleus has not occurred, or if so, a
further division of the antipodal cells has taken place.

The nucleus of the elongated embryo-cell now divides, and
a wall appears between the new nuclei cutting it into an
upper and a lower cell. This horizontal lamella is thin and
sharply marked, and is nearly or quite perpendicular to the
long axis of the embryo. Each cell is full of fine grained
dark protoplasm, and its nucleus is very large, spherical,
and bright, and contains one or two brilliant nucleoli
(fig. 18.) On account of the different fates of these two cells
we must distinguish them from the first ; the upper one
becomes the pro-embryo or suspeiisor, and has but a tran-
sient existence ; the lower produces the true embryo.

Each nucleus repeats the process of division exactly as
before, and we have the embryonic body divided by two
more walls parallel to that first formed (6g. 20), and already
is established a physical diflference between the embryo and
pro-embryo, the latter being narrow and tapering somewhat,
and having difiluent thick walls, while the embryo rapidly
becomes stouter and more globular in accordance with the
distribution of its thin sharply marked cell-walls (cf. figures).
In fig. 19 the wall in the embryo is completed, and two
large nuclei again rounded off, but that in the pro-embryo is
only just appearing, its nucleus being fixed in the last stage
of division prior to the separation of the new nuclei, which
have already commenced to aggregate at the poles. In
fig. 20 the process is completed.

A further difference between pro-embryo and embryo is
now established, in that the next divisions in the embryo
are perpendicular to that already formed, whereas in the pro-
^^embryo all the divisions are horizontal, and parallel to the
first. In fig 24 the terminal cell of the pro-embryo is com-
mencing to divide as indicated by the condition of its nucleus,
while in fig. 27 the second cell has just divided, and the
last threads of protoplasm are still in contact with the new
wall. We thus demonstrate that each cell of the pro-embryo
divides, and its elongation is effected by intercalary growth.

^ Iji some endosperinous ovules the antipodal cells divide vigorously.

EMBRYO*SAC OF GYMNADENIA GONOPSEA,

9

In fig. 28 each of the four cells has doubled itself, and the
power of division in the pro-embryo seems now to become
exhausted, vacuoles begin to form, and the cells to elongate
as their nuclei go to the walls, and the vacuoles collect into
a sap-cavity (cf. figs. 28 — 30). Only in a few cases have I
seen more than eight cells formed by the pro-embryo; in the
specimen figured at fig. 30 are ten. As the cells elongate,
since the solid embryo soon completely fills the embryo-sac,
the apex of the pro-embryo becomes gradually pushed
through the top of the sac, and the loose tissue of the micro-
pyle allows it to escape into the cavity of the ovary (figs.
29, 30); its period of growth is now about completed, and,
as the last divisions are made in the embryo, the pro-embryo
turns brown and shrivels up, persisting as a mere ragged
appendage in the ripe seed.

To return to the embryo proper consisting of two cells (fig.
20). It becomes broader, and the nucleus of one cell divides
and a new wall cuts it into two equal cells arranged laterally.
The plane of this new division is always perpendicular to that
of the first (horizontal) wall, and passes through the longer
axis of the whole embryo ; it may appear first in the upper or
in the lower cell, but usually the latter (fig. 21, 22). The other
cell divides at the same time or very soon after by a wall,
also passing through the long axis, and also perpendicular
to the horizontal wall, but it is, in the majority of cases, if
not always, also at right angles to the other perpendicular
wall. Thus, in fig. 23 the first wall (horizontal) is cut at
right angles by the second (perpendicular), which lies in the
l)lane of the paper and in the upper cell ; the third (perpen-
dicular) will cut both of these at right angles or nearly so,
as shown by the dividing nucleus in the lower cell. Similar
relations are shown in figs. 24, 25. Each of the four cells
thus formed rapidly becomes again divided by a wall perpen-
dicular to all those which it cuts, and passing through the
long axis of the embryo, and in this manner the embryo
comes to consist of a nearly globular body cut into eight
octants, in each of which is a large round nucleus (fig. 26).

The next division walls are again horizontal, and may
appear first in the upper (fig. 29) or in the lower cells
(figs. 27, 28); thus, the embryo becomes cut into twelve, and
then sixteen cells, by walls in planes symmetrically related.
A series of walls very soon mark out a central from an epi-
dermal system; these (fig. 30) appear at about the same
lime in all the cells except the tour which abut upon the pro-
embryo, and lie in planes parallel to the outer wall of each.
They may well be called tangential, and mark the first in-

10

H. MARSHALL WAKD .

clication of tissue differentiation ; the cells of this outer layer
only divide further by -avails perpendicular to the outer sur-
face. The ovoid embryo now begins to have starchy and
other grandular matter deposited in its cells, and becomes
thereby too opaque for observation of the cell-walls until sub-
mitted to the action of warm potash and glycerin or other
clearing reagents. From the first no trace of vacuoles occurs,
hut the cells are tightly packed and full of fine-grained pro-
toplasm, with large bright nuclei, and thin firm cell-walls
surround them. By proper treatment, however, one recog-
nises in the last stages of the embryo (fig. 31) that yet
another series of tangential walls has appeared concentric to
the first, and thus the ovoid mass presents a central column
of cells, surrounded by a layer one cell thick, while over all
is another layer, also one cell deep.

About this time the cells are crowded with nutritive
matters, the pro-embryo and integument cells are empty and
shrivelled (fig. 82), the remains of nuclei appearing on their
dark brown walls, and the seed may be considered ripe.

In reviewing the processes above described, we may shortly
point out several views held of late as to their meaning, and
to render this more clear and complete, it may be advisable
to recall to mind some points more fully dealt with in the
larger text books.^

In the Ferns generally we haVe a spore developing a free
chlorophyll-bearing, and often large prothallus, on which
are produced antheridia and archegonia. When, as in Os-
munda, the prothallus only bears antheridia at times, we
may consider this the carrying to a step further a process
common to this and many other genera, where the arche-
gonia appear later than the very numerous antheridia. If
we suppose the appearance of the archegonia indefinitely
postponed the prothallus becomes unisexual — male ; if pro-
longed after all the antheridia have decayed the prothallus
is practically female in function.

Such a unisexal prothallus may be supposed diagramma-
tically represented in Fig. U as a section passing through
the germinating spore, prothallus, and archegonium. In the
long free neck of the latter are seen several masses of pro-

* Tlie English student has an excellent account in Sachs’ ‘ Text Book.’
Cf. also lloimeister, ‘ Higher Cryptogamia,’ Ray Soc., 1862, and Luerssen,
‘ Med. Pharm. Bot.,’ B. i. Also literature quoted.

2 Ti>e relative positions and sizes have not been insisted upon in the
diagrams. In all the figs. Ex. = exospore, En. = endospore, Arch.
points to neck of archegonium, and Can. to the canal-cells, Oos. = oosphere
or egg-cell.

EMBRYO-SAC OF GYMNADENIA CONOPSEA.

11

toplasmic substance [can.') which result from the partial
division of a small piece of the oosphere, which becomes
early cut off from that body ere it rounds off in the body of
the archegonium ; these masses are the ‘‘ canal cells of the
Germans, and by degradation become mucilaginous, and so
serve to fix the antherozoids.

Fig. 1. — Diagram of free protb allium of fern, with its archegonium pro-
jecting some distance exteriorly, and possessing a many-celled neck
in the canal of which are the “ canal cells.”

Although the spores of Ferns present no external charac-
teristics from which we can infer whether the prothallia to
be produced on germination will be predominantly male or
female, u e. in its whole course of existence will bear arche-
gonia or antheridia in excess, still we may see here indica-
tions of a differentiation of function which, attended hy
reduction and abbreviation of tbe prothallus and processes
peculiar to it, attains a limit in the highest plants. If in
this specialisation and distribution we see an economy of
material and energy, we can at the same time explain many
of the phenomena.

In the Rhizocarps the separation of the sexes has been
carried so far that from the spore itself we can predict
whether archegonia or antheridia will be formed on its ger-
mination, in some cases even the sporangia participating in
the separation, and being distributed on different parts
according as their products will yield male (microspore) or
female (macrospore) prothallia. The male or antheridium-
bearing prothallia are very small, becoming reduced to a
mere tube of two or three cells in Salvinia, and in others
being only represented by the trace of protoplasm left over
from the antherozoid mother-cells. We must, in fact, look
upon certain cell divisions in the microspore, preceding its
germination, as representing the formation of a rudimentary
prothallus and antheridia which form the few' antherozoids
then liberated.

In the niacrospore, though the prothallus is also not set
free, it is more obviously a cellular structure, producing one

12

H. MARSHALL WARD.

or more archegoiiia. We may take that of Marsilea as a
type. The diagram (Fig. 2) represents a section through
the germinating macrospore, with its only partially exposed
prothallus, bearing an archegonium which differs from that
of the Fern in several points.

Fig. 2. — Diagram of partially free $ protballium of Marsilea, it is re-
duced to little more than the archegonium, which hardly projects from
the surface, and has very few neck cells. The “ canal-cell,” however,
appears. The space ( X ) becomes filled with fluid.

In the first place, its neck, instead of being a long, freely
projecting structure of several tiers of cells, hardly pro-
trudes at all, and is formed of two tiers of four cells each ;
in plan these cells are arranged crosswise, and are almost
flush with the general level of the prothallium. Between
them, however, the young oosphere allows part of its sub-
stance, cut off as before, to penetrate as the canal cell.”

Besides a separation of the sexes, then, we have in Rhi-
zocarps a much smaller prothallus which never becomes
entirely free ; and as the prothallus tends to be withdrawn
(as it were) into the spore, so, too, the archegonia, &c.,
appear to be held back in the prothallus, and the neck to be
a less protruded structure. For a third stage in this re-
markable process we may select Selaginella. Here the
macrospores and microspores are not only produced in dif-
ferent sporangia, but the macro- and micro-sporangia are
borne on different leaves ; the sexes are further separated.

As before, the microspore undergoes the less extensive
development ; its contents becomes divided up into a few
cells, the majority of which produce antherozoids. The
spore then bursts and sets them free ; the process may be
considered as the formation of a rudimentary internal pro-
thallus reduced to little more than the antherozoid mother-
cells (antheridium). In the macrospore, a small prothallus
forms internally, and is just allowed to peep forth and

EMBRYO-SAC OF GYMNADENIA CONOPSEA.

13

expose its one or two rudimentary archegonia. These con-
sist of an oosphere, surmounted by four or eight neck-cells,
which open flush with the surface, and have a canal-cell ”
as before forced between them.

Fig. 3. — Diagram of endogenous protballium of Selaginella, with few ar-
chegonia, reduced each to an oosphere with four or eight neck-cells as
a “ rosette ” above, and with canal cell between. Before the rupture of
the apex to the expose prothallus, a large-celled, delicate “ endosperm ”
(end) forms below the prothallus, apparently independent of it.

These neck-cells form a kind of rosette,’’ as it has been
termed, capping the oosphere which is sunk in the pro-
thallus. But before the prothallus and its archegonia are
exposed at the ruptured apex of the spore, a process occurs
w^hich results in the formation of an apparently new
structure.

In Salvinia, Marsilea, &c., the space between the pro-
thallus and the endospore, filling up the major part of the
macrospore, becomes occupied by imbibed fluid, which accu-
mulates and serves to push the prothallus upwards to the
exterior, as it presses upon the diaphragm ” or membrane
separating them. In Selaginella, however, a formation of
large, thin-walled cells occurs in the fluid filling this space,
and thus the so-called endosperm ” (Fig 3, end.) is pro-
duced. At a later period its cells become crowded with
food material for the nourishment of the embryo, as this is
pushed down by its growth and that of its suspensor.”

Our next step is to the Gymnosperms — the Conifers and
their allies. Neglecting minor variations in detail, a typical
Conifer presents the following features :

Its microspores^ or pollen grains are produced not only in
special and separate sporangia or pollen- sacs, but also on en-

' For further information as to these homologies, &c., the literature
quoted may be consulted.

H. MARSHALL WARD.

]4

lirely different shoots, or even on other plants than those which
bear the female sexual organs. Each pollen grain on ger-
mination (a process which by special appliances is brought
about in close proximity to the female apparatus) emits a
tubular body, in which indications of division occur, while a
very few divisions are also established in the interior of the
pollen grain itself. It is now generally accepted that the
internal divisions represent a prothallium even more reduced
than that of Selaginella, while the pollen-tube contents must
be regarded as the representative of what becomes anthero-
zoids in vascular Cryptogams — structures which are here
rendered unnecessary as such, by arrangements already
referred to. Instead of shedding motile antherozoids, there-
fore, the pollen grain carries its sexual products right into
the region of the oosphere by means of the pollen-tube.

In the female apparatus we find the same principles of -
suppression carried still further ; the macrospore^ or '^primary
embryo-sac '' is never shed at all, but germinates, so to
speak, inside its sporangium — the so-called nucleus of the
ovule.

During the gradual discovery of the phenomena which we
are discussing, a number of synonyms have been introduced
into the nomenclature, and some confusion is apt to arise in
comparing these processes with what occur in Cryptogams ;
hence no apology is offered for the following summary :

The archegonia (the oosphere'S of which are the corpus-
cula of R. Brown, the secondary embryo-sacs^’ of Hen-
frey) are formed by division of peripheral cells of a delicate
prothallium formed of large, thin-walled, polygonal cells,
and termed endosperm this prothallus arises in the
protoplasm of a cavity which appears like an enlarged
cell ^ of the nucleus of the ovule, and is the primary
emhryo-sac.” From its bearing archegonia and other rela-
tions, we must regard this endosperm ’’ as an internal
prothallium similar to that which arises in Selaginella before
the endosperm ” of that plant is formed ; the term endo-
sperm” has been applied, therefore, to two structures which,
whatever relations they may have morphologically, are dif-
ferently distributed in time. The endosperm ” of Selaginella
arises after the prothallus of that plant is formed, and coexists
with it at the period of fertilisation ; the endosperm ” of
Conifers is the prothallus, and bears the archegonia.

The archegonium of Conifers consists of an oosphere, with

^ Or possibly several.

- And was so described by llofmeister. But cf. Strasburger, ‘Die
Angiosp. u. Gjfinuosp.’

EMBKYO-SAC OF GYMNADENIA CONOPSEA.

15

one or two tiers of four cells placed crosswise surmounting
it ; from the method of formation and the fact that in some
genera a canal-cell has been observed cut off from the oosphere
and forced up between these cells, this rosette ” may be

Fig. 4. — Diagram of wholly internal protliallium (so-called “ endosperm ”)
of Conifer, with an archegonium, consisting of an oosphere (“ cor-
pusculum,” “ secondary embryo sac ”), surmounted by a ‘‘ rosette ”
of four cells, placed cross-wise ; between these a “ canal-cell ” is forced.
A tendency to still further withdrawal within the macrospore (primary
embryo sac) is indicated by the funnel-like depression. Pr. Primary
embryo-sac at apex. Nuc. Nucleus of ovule.

regarded as the neck of the archegonium. As seen in the dia-
gram (Fig. 4) the archegonium is even more deeply withdrawn
into the prothallus than was the case in Selaginella, and in
some genera it becomes quite sunk into the prothallus.

We may now inquire what processes and structures in such
an ovule as that of Gymnadenia are related to those just
reviewed.

The first step appears to be to settle whether the ‘‘ embryo-
cell of the Angiosperm is the equivalent of the oosphere
of the Gymnosperm and Cryptogam.

Warming and Vesque argue somewhat as follows : — The
embryo-sac mother-cell becomes divided by various trans-
verse walls, just as the subepidermal cells of an anther
become divided by walls parallel to the epidermis to form
pollen mother-cells.” Hence, the cells into which the
embryo-sac mother-cell is cut up are so many spore mother-
cells,” i.e. each is equivalent to a pollen mother-cell. Vesque
says further, that two of these opposed cells form each a
group of four nuclei in its interior, just as the pollen mother-cell
forms four pollen grains in its interior, and that the cell-wall
between the two tetrahedral groups becomes absorbed, and
so eight nuclei are formed, in two groups of four each, at
the respective ends of the embryo-sac so formed by fusion of
the two mother-cells.

16

H. MAllSHALL WARD.

They further point out that just as the cell- walls between
the mother-cells of pollen are very deliquescent, so in the
formation of the embryo-sac the dividing lamellee are swollen
and soon absorbed in similar manner. Also the four nuclei
arising inside each cell pollen mother-cell or constituent
of embryo-sac) are arranged in tetrahedra in either case.
If this account be accepted, the eight nuclei are homologous
with pollen grains, i.e. spores, and the egg-cell is not
to be regarded as the hornologue of the “ oosphere ” of vas-
cular Cryptogams, but as a macrospore which never germi-
nates ; and may be regarded as containing in itself the
representative of the whole prothallus and archegonium.

But a careful survey of the facts in many different types has
convinced Strasburger that no such fusion of two mother-
cells occurs, and we see no such process in Gymnadenia as
Vesque describes in Orchis galatea / especially does he appear
to overlook the gradual compression and absorption of the two
upper (cap) cells. The analogy between the mode of divi-
sion in the embryo-sac mother-cell and the primitive cells of
pollen-forming layers in an anther can probably be ex-
plained in a totally different manner,^ and in any case we
can lay no stress on it so long as it is unsupported by other
evidence.

The following argument against this view appears to me
important, especially if we take into consideration the many
analogies tending to the conclusion that the Angiosperms
are a series in which the reduction of the prothallus
generation ” (oophore) is reaching its limit.

The pollen grain has been found to contain two nuclei,^ ap-
parently representing incipient germination changes — the
first division in the microspore to form rudimentary pro-
thallial structures. This being the case, from analogy with
Cryptogams and Conifers (where we find it is always the
microspore and male prothallus, &c., which suffer reduction
first) we may argue the probability that whatever represents
the macrospore will not have undergone more suppression
than the microspore, and probably much less. But I think
there is another reason for not imagining the female pro-
thallus to be entirely atrophied, and for holding that the
stages of reduction past a rudimentary archegonium art

* Loc. cit. Vesque’s account of the facts also differs from that oii
here.

2 See an attempt in paper to Linnsean Society, read Nov. 20th, 1879,
“ Contributions to Knowledge of Embryo-sac,” &c.

* Strasburger, loc. cit., and Elfving in ‘ Jenaische Zeitschrift,’ B. xiii.
See the account of Elfving’s researches in the present number of this
journal.

EMBRYO-SAC Of GVMNADENIA CONOPSEA.

17

almost inconceivable. For if the Synergidte represent two
cells of a protliallus, we may regard them possibly as
remnants of the neck of an archegonium, and the egg-
cell as its central cell ” (oosphere), and suppose that they
have persisted in virtue of their use in reproduction.

Strasburger says that one cell resulting from the division of
the embryo-sac mother-cell becomes the embryo-sac, that
its contents divide as described into eight nuclei by three
divisions in alternating planes ; this embryo sac must be
regarded as the macrospore — the equivalent of the embryo-
sac of Gymnosperms.

Looked at thus, the divisions in the embryo-sac represent
a rudimentary prothallus, and the egg-cell^’ may be re-
garded as the oosphere. The Synergidse were formerly
thought to be the representatives of the canal-cell ; but
Strasburger, who at first took this view, afterwards showed
that since they are sister cells of the egg-cell, and 7iot a
jyroduct of its division, such cannot be the case. He there-
fore considers them and the antipodal cells,^^ as well as
the nuclei which fuse to form the embryo-sac nucleus,’^ as
merely seven cells of the prothallus.

Strasburger, in criticising Vesque’s theory, points out
that no stress can be laid on the fact that the division walls
in the embryo-sac mother-cell are deliquescent ; for this,
and the similar appearance in pollen-cells simply result
from the imperfect nature of the process, since they will all
be very soon absorbed, on the one hand by the* enlarging
embryo- sac, on the other by the pollen grains. The same
remark applies to the divisions in the temporary pro-embryo
of Gymnadenia.

Further, we cannot lay much stress on the tetrahedral
arrangement of the nuclei, for the rule of rectangular
division applies very widely,^ and in the history of Gymna-
denia we have seen how the first divisions even in the
embryo follow a similar order ; but in trichomes, prothalli,
and other structures, the same obtains.

But another view appears possible. It will be remem-
bered that the first divisions across the embryo-sac mother-
cell follow one another in such a way that the two cap-
cells were spoken of as being cut off from the mother-cell
by diffluent swollen Avails ; and that, the lower cell having
enlarged, destroying the cap-cells, its protoplasm passes to
each end and a vacuole-like clear space forms betAveen.

This last division may probably be looked upon as merely

’ See also Sachs, in ‘Wurzburg Arbeiteii,’ “Auordnung d. Zelleu in
juugstcn Pflanzentheilen.”

VOL. XX. NEW SER.

18

H. MARSHALL WARD.

a third division across the embryo-sac mother-cell, and not
as the first division of the contents of the macrospore
(embryo-sac). In other words, we have here a division wall
still weaker than the two preceding, and the vacuole
is its expression. If this be so, it is possible that the
embryo-sac mother-cell is really the mother-cell of four
spores, two of which (the cap-cells) yield up their contents
to their more vigorous neighbours — to the other two, which
never completely separate, but form an embryo-sac,” and
its contained apparatus.

This suggestion does not exclude the view that the eight
nuclei derived from that of ihe embryo-sac mother-cell are
cells of rudimentary prothalli, but explains them as belong-
ing to prothallial structures instead of one ; the one pro-
duces a rudimentary archegonium (the egg-cell with its
synergidse may perhaps be an oosphere and two neck-cells),
and one vegetative cell ; the lower spore produces four vege-
tative cells. How to explain the subsequent fusion of one of
these from each group is not easy, unless some advantage
accrues to the large embryo-sac, by having a nucleus ren-
dered vigorous by material from two slightly different
sources. This is, however, too hypothetical to enlarge upon
here, and perhaps we are yet far from possessing the facts
necessary for an explanation of this remarkable process.

POLLEN-BODIES OF THE ANGIOSPERMS.

19

Studies on the Pollen-Bodies q/ Angiosperms. By
Fred. Elfving, of Helsingfors. ^ Y/ith Plate IV.

Until very recently botanists believed it to be a well-
established fact that the Pollen-bodies of the Angiosperms
were one-celled, that when once formed as tetrades in the
pollen mother-cell they underwent no further divisions.
This was thought, too, to form a direct contrast with the
Pollen-bodies of the Gymnosperms, in which, as is well
known, shortly before the period of pollination, one or more
so called vegetative cells are formed, which are regarded as
constituting a rudimentary male prothallus.

Strasburger^ has, however, quite recently shown that the
Pollen-bodies of several Angiosperms, both Mono- as well as
Di-cotyledons, possess two nuclei ; he has further shown
that one of these originally pertained to a small peripherally-
formed cell, and that it only became free by the subsequent
resolution of the partition wall ; so that here, as in the case
of the Gym.nosperms, a vegetative cell is formed in the Pollen-
body.

Strasburger further discovered in the case of the Orchideee,
which he examined, that the nucleus of the large cell is
always in front in the Pollen-tube.

Strasburger points out that Reichenbach had already
figured and described both nuclei in the Pollen-bodies of
some Orchids ; and that Hartig had also rendered the two
nuclei visible in the Pollen-bodies of several plants by the
use of a carmine solution. Yet these notices, because they
lay somewhat out of the beaten track, remained almost un-
observed.

I have attempted, at the request of Professor Strasburger,
to continue the researches into the stages of development of
the Pollen-bodies of the Angiosperms. My researches, with
this view, were made in the summer time of 1878 in the
Botanical Institute at Jena, under the direction of Professor
Strasburger; and this paper contains the result. In these
researches I had the advantage of Professor StrasburgePs
very kind and able help; and I now gladly seize the oppor-
tunity of expressing my warmest thanks to him therefore.

My task consisted on the one hand in following up the

^ Translated and condensed from the ‘ Jenaische Zeitschr.,’ 1879,
part 1.

^ “ Ueber Befruchlung und Zelltlieilung,” ‘Jenaische Zeitsclirift fiir
Naturwissenschaft,’ Bd. xi, Neue Edge Bd. iv, 1877, Heft. 4, p. 450.

20

FRED. ELFVING.

newly-discovered cell division in the Pollen-bodies, and on
the other in studying the relations of the nuclei in the
Pollen-tube.

The reason why the fact of these Pollen-bodies having
several cells has been overlooked by investigators, is pro-
bably altogether owing to the want of a suitable method
of investigation. For if the phenomena which take
place show' themselves with surprising clearness in some
plants, even without using particular reagents, yet, in
general, little can be gained by the examination of fresh
material ; and the reagents formerly used by botanists
rendered little or no assistance.

Osmic acid, wdiicli has lately come into use, proved to be
an invaluable aid in these investigations. This acid has
been used by Strasburger with great success in his investi-
gations concerning the Division and Fertilisation of Cells.
After I had tried several other means for clarifying my
specimens I used the osmic acid alone, and that only in a
solution one per cent, strong. It is always of great advan-
tage to add some kind of colouring material to the specimens
preserved in osmic acid ; sometimes it is quite indispensable
to do so. A solution of carmine, to w'hich a little glycerine
was added, usefully served as such. By these means pre-
parations can be obtained after a space of twenty-four hours
w'hich, in the way of clearness, leave nothing to be wished
for. It is very useful to break up the Pollen-bodies, if they
are large and richly filled with granular or oily contents,
immediately after the osmic acid has been added ; this can
be done by pressing on the glass cover ; for if left w'hole they
may colour slowly or not at all. This is especially useful
w'hen it is desired to see everything quickly. The nuclei
are in this way pressed out and immediately fixed, together
with the rest of the contents of the cell. It is often even
possible to get in this manner a view of nuclei in the act of
dividing ; colouring with carmine is naturally also here of
great use. Plants with many-flow'ered inflorescences are
especially suited for these investigations. If the first
blossoms of such a plant have opened, it is easy to find all
tlie younger stages of development in the buds above one
another. The vegetative cell is, however, always (the case of
the Cyperaceae alone excepted) formed in those Pollen-bodies,
which are separated from one another; that is to say, if they
ever do separate.

In order to study the relations of the nuclei in the Pollen-
bodies, I cultivated these in many and different solutions.
This I did in the usual way — viz. in suspended drops in a

POLLEX-BODIES OF THE ANGIOSPERMS.

21

moist chamber. At first I tried solutions of cane-sugar of
different degrees of concentration, but, as in many cases, no
satisfactory results could be obtained in this manner, I tried
other liquids. Van Tieghem,^ who also cultivated Pollen-
bodies, recommends the addition to the liquid of une petite
quantite” of acid tartrate of ammonia. T first of all decided
that this petite quantile’^ could not be above 1 per cent., as
stronger solutions would be simply deadly. Solutions of
*1, *25, *5, and 1* were, on the other band, of no special use.
The directions again of Van Tiegbem that inorganic salts,
sugar, gum, and etheric oils should be added, according to
the needs of the plants, are too general and uncertain to be
worth wasting time over in further trials. After I had made
trial also of a solution of glycerine one per cent, strong, and
after trying solutions of nitrate of potassium and of carbonate
of soda, I turned back to my solutions of cane-sugar (called
in the following for the sake of brevity sugar solutions), as the
most suitable. I made use of such solutions of different de-
grees of concentration (as 1, 8, 5, 10, 20, 30, and 40), for, as
might have been anticipated, it was soon apparent that the
maximum degree of concentration required is different for
the Pollen-bodies of different kinds of flowers ; while som.e
Pollen-bodies developed tubes in almost any solution, others
required a certain definite degree of concentration ; the indi-
vidual variations in this respect were very remarkable. I
is of great importance to use, for the purpose of cultivation
quite ripe, though not over ripe. Pollen-bodies.

Omitting minor variations I have given in the following
pages the degree of concentration which proved itself the
most useful in the culture of the Pollen of the several
species of plants studied. I have also given the time in
which the tube reached a certain length, under circumstances
which were otherwise normal. Where no other solution is
mentioned the cane-sugar solution was the one used.

In all culture experiments the Pollen tubes swell after a
time into club-like bodies, and finally finish off by bursting.

In older tubes the still growing point is found separated
by peculiar cellulose-plugs from the emptied posterior por-
tions (see Strasburger, 1. c., p. 456).

The culture experiments were undertaken with a tempera-
ture as warm as the usual one of a room and in the dark.
In many kinds of flowers the tube formation also took place,
and in a normal manner, in the daylight. I did not turn my
attention further to this point.

At the same time I was, in the case of many plants, unable
‘ Annales des Sc. Nat., 5 seric, t. xii, 1809, p. 318.

32

FRED. ELFVING.

to obtain tubes artificially. I then, sometimes successfully,
tried to make preparations thereof out of the already fer-
tilised pistils.

The Pollen-tubes which I obtained in either way were
immediately fixed with osmic acid or absolute alcohol. It
is well to separate the moisture which surrounds the tube
as much as possible before adding either of these reagents.
This can easily be done with a capillary glass tube ; and
thus powerful diffusion streams are avoided, which often
occasion the bursting of the Pollen-tubes. The fixing then
takes place almost instantaneously.

The Orchids, especially those with richly-flowered spikes,
offer splendid subjects for such investigations. I have ex-
amined Orchis latifolia, 0. masculaj 0. maculata, Ophrys
myodes, Platanthera hifolia, Gymnadenia conopsea, and
Serapias francogallica, and am able to confirm the state-
ments of Strasburger in every particular ; on which account
I may for these simply refer to his drawings (1. c., p. 450, 452,
plate t. xxvii, figs. 41 — 47). The formation and develop-
ment of the Pollen-bodies corresponded exactly with these in
every particular. The Pollen-body, originally provided with
a round nucleus, is completely separated into two cells, of
which the smaller is almost always to be found in one corner
of the Pollen-body. The nuclei ^f these two sister-cells are
round and almost of the same size ; the body of the nucleus
of the smaller cell is, however, always smaller than that of
the larger. No trace of a cellulose membrane between the
two sister-cells could be detected, either by the use of
reagents or by the crushing of the Pollen-body. Even the
separating plasma-layer is removed at a later stage, and in
the ripe Pollen-body the two nuclei lie free beside one
another.

The formation of the Pollen-tube takes a long time
in the case of the Orchids, compared to that required by
most other plants. Different species differ a little in this
respect ; the best results were obtained, as a rule, by a culture
of from twenty to forty hours, and in a solution of from 5 to
10 per cent, of sugar. It proved to be, as Strasburger first
found in these plants, that the nuclei wander together into
the tube, and that in doing so the nucleus of the larger cell
grows quite disproportionately. Both take a somewhat long
elliptic shape. In tubes the results of older cultures, the
anterior nucleus had elongated itself in a striking manner;
the nucleolus was, however, to be seen clearly. I am
inclined to regard this as only a change caused by cul-
tivation ; because in tubes extracted from styles one always

POLLEN-BODIES OF THE ANGIOSPERMS.

23

finds the nuclei of the same form^ and this also when they
are less developed.

I have tried to follow the nuclei until the moment of fer-
tilisation. In Gymnadenia conopsea, used as a favorable
object of research, I was able to make out definitely that the
nuclei were to be found som.e distance from the point in those
tubes which had already pressed their way into the micro-
pyle, and whose point already touched the inner tegument of
the ovule. In many cases one of the nuclei again divided
itself ; and so there were three nuclei. It ahvays seemed to
me to be the posterior nucleus which had divided, and in the
case of Orchis maculata I can state positively that it was the
posterior one. As soon as the fertilisation is over, which
can be seen by the changed condition of the envelopes in the
embryo sac, no further traces of the nuclei are to be found.
The whole end of the tube, whose point lies on the embryo
sac, often separated by a cellulose plug on the outside, is al-
together homogeneous and highly refractive.

I may as well state here that I have brought ovules,
which were capable of fertilisation, and tubes, when they
(the tubes) were growing vigorously, together into one drop
of saccharine solution; but that in no case could any entry
of the tube into the micropyle be observed, not to speak of
an act of fertilisation (Van Tieghem, 1. c., p. and

Strasburger, 1. c., p. 486). I will begin the description of
the stages of development in the Monocotyledons which I
examined, with the following, which I found to be one of the
best examples.

Antliericum ramosum. — The Pollen-bodies of this species
are almost semi-spherical. In a dry condition their convex
side is deeply folded in. The extine is here very thin, and
mostly broken through by the strong development of the
intine, which later will evolve the tube.

In examining flower buds of about 5 mm. in height Pollen-
bodies are generally found in which the vegetative cell is
already formed (Plate IV, fig. 1). It is very difficult in the
case of this plant to observe in detail the various stages of
division on account of the dense contents of the cells. There'
is, however, no difficulty in ascertaining that the large circular
nucleus places itself in the equatorial plain of the Pollen-
body, and there divides itself. The result of the division is
two cells. One, taking up by far the larger part of the
Pollen-body, has a far larger nucleus, which is very like the
original one, and like it has a very large nuclear body. The
other, and far the smaller one, is attached to one side of tht'
Pollen-body, and is separated from the sister-cell by a w'atch-

FRED. ELFVING.

. 24

glass shaped wall, to which the inline is attached ; it is
distinguished by its transparent almost homogeneous proto-
plasm, by its roundish oval nucleus, the nuclear body of
which, though large, is still smaller than that of the larger
cell. A partition by means of a cellulose membrane, as in
the Orchids, seldom takes place, as the result of this division.
The two cells, are oftener only separated by a layer of cortical
plasma, which originates from the cell })late.”

The vegetative cell soon separates itself from the inline
altogether, and appears as a spherical formation in the inside
of the Pollen-body (figs. 2 and 3). This cell elongates
itself considerably in length and becomes spindle shaped
v,dth pointed ends, which are bent in (figs. 4 — 6) ; its nucleus
remains almost unchanged. If the Pollen-bodies are care-
fully crushed in a saccharine solution (5 per cent, strong),
this nucleus is easily found intact among the contents, which
are pressed out. When pressed out of rather young Pollen-
bodies, it generally becomes rounded and assumes a spherical
shape, although it may have had already in the Pollen-
body the spindle shape. In ripe Pollen-bodies, on the other
hand, it is tolerably resistant, and keeps its form. Its
nucleus appears clearer than the surrounding protoplasm if
osmic acid is added ; it, however, stands out darker. When
a strong solution of sugar — the best is one of 20 per cent. —
is used, or, when a weak agent is a,dded capable of absorbing
water, the inner plasm structures will be seen to contract and
to leave behind as a membrane the outer cortical plasm.
The whole is coloured brown by a chloride of zinc solution.

While the vegetative cell changes in the manner just
described, the nucleus of the big cell remains at first unal-
tered. It afterwards becomes longish, and in doing so often
bends upon itself. Upon this its nucleolus also vanishes.
It is very difficult to point out the nucleus in the ripe Pollen-
* body without using staining materials. The nucleus appears
then as an irregularly-shaped, often crumpled body, or as
membranous and shrunk together (fig. 6).

Antherictm liliago corresponds with the above species,
except that the large nucleus keeps its round form. In a
Pollen-body, which lay already on the stigma, but had pro-
duced no tubes, the nucleus was apparent, when fixed with
osmic acid and coloured with carmine, as a strangely star-
shaped body (fig. 8), and gave one the idea that it might
have performed amoeboid movements while it was being
fixed. Direct observation on the living Pollen-bodies was
impossible on account of the thick contents of their cells.

Pollen-tubes could not be obtained from these two species

POLLEN-BODIES OF THE ANGIOSPERMS.

25

by culture^ I therefore prepared them independently from
pollen-besprinkled pistils. In Anthericum liliacjo the whole
vegetative cell is found in the tubes, together with the
nucleus of the big cell, which is very much elongated
and often looks like a very fine thread. It is usually
this nucleus which goes first, though not without excep-
tion. Fig. 7 shows such a tube, which also exhibits
the rather rare instance of branching. On the other hand,
in Anthericum ramosum I cowld observe no nuclei as soon
as the tubes w'ere formed.

It is very much the same with Glohha bracteata^ only that
in this case one can indicate no special place in which the
vegetative cell will be formed, owing to the Pollen-body
being spherical and provided Avith a membrane equally thick
all over. In its ripe condition the Pollen-body is shaped
like that of Anthericum (fig. 4). The Pollen-bodies, in a
solution of 5 per cent, of sugar, produce short tubes; these,
however, stop too short to allow of conclusions about the
behaviour of the nuclei.

Tulipa Gesneriana (figs. 9 — 14) corresponds in most
particulars with Anthericum. The vegetative cell is enor-
mously developed (fig. 12 shows two pressed out), and its
nucleus is often provided Avith several nucleoli. In
the ripe Pollen-body it immediately strikes one from its
size and half-moon shaped form. Occasionally I have found
a division of the vegetative cell in the younger Pollen-bodies
(fig. 18). In the tubes (1 — 8 per cent, sugar solution;
eighteen hours) the nucleus of the big cell Avent before, and
was folloAved by the elongated vegetative cell (fig. 14).

In Ornithogalum pyramidale the vegetative cell is also
formed in a corner of the Pollen-body, opposite to the
opening in the extine. In one instance, Avith osraic acid
and carmine preparations, the division Avail betAveen
the tAvo cells clearly appeared double-contoured and was
highly refractive like the intine, into Avhich it undoubtedly
passed over (fig. 15). I have no doubt that in th-s case a
cellulose membrane Avas really formed.

Its further development corresponds with that Avhich has
already been described. The vegetative cell becomes finally
much elongated by stretching, and has pointed, often bent
in, ends. The greatest part of it is occupied by the almost
cylindrical nucleus Avhich possesses no nucleolus (figs. 16,
17). On crushing out the contents of a ripe Pollen-body the
vegetative cell appears as a hyaline body in the midst.
Both ends are filled Avith yellowish coloured little bodies ;
in general it has a strong resemblance to a cell nucleus, for

26

FRED. ELFVING.

which it was taken in the like case of Narcissus poeticus
by Strashurger. Osmic acid dissolves the yellow bodies,
but brings out the nucleus clearly in the midst (figs. 18,
18o). The nucleus of the larger cell undergoes consider-
able changes, whereby the greatest part of its substance
becomes -dissolved, so that finally only a small, irregularly
shaped, often whipcord* like remainder of it can be made
out (figs. 16, 17). ^

Ornithogalum Ecklonii corresponds with the preceding.
The Pollen-bodies in both species produced no tubes in the
various solutions I used. There were no traces of nuclei to
be seen in the tubes of 0. Ecklonii, which I prepared from
the styles, and which could be followed throughout their
whole length ; the ends of the tubes were thickly filled with
fine granular protoplasm, and were cut oif from the empty
upper part and from the nucleus by the usual cellulose
plugs.

Leucojum (Bstivum is a very favorable subject for exami-
nation, where the whole course of development, almost with-
out using re-agents, can be pursued. The formation of the
vegetative cell takes place in tolerably old buds. It
is marked off by a thick arched-in wall of cortical plasm.
The contents appears almost homogeneous (fig. 19) ;
the addition of osmic acid causes, however, the nucleus,
which possesses no nuclear body, to appear very distinctly
(fig. 20). The question whether the vegetative cell is formed
in a particular part of the Pollen-body I must leave un-
answered, as I did not direct my attention to that point in
examining it. Neither does my drawing allow me to come
to any definite conclusion. The nucleus of the large cell is
provided with distinct nucleoli.

The vegetative cell now loosens itself from the intine ; at
first spherical (fig. 21), it soon assumes the shape depicted
in figs. 22 and 23. The largest part of the cell is occu-
pied by the now elliptic nucleus, which appears pellucid
in the fresh Pollen-bodies, while the protoplasm, permeated
by dark-coloured granules, is pushed away almost entirely
into the ends, which are bent in like horns. Fig. 23 shows
such a vegetative cell after treatment with osmic acid.
Fig. 24 shows the two nuclei crushed out and treated with
osmic acid and carmine.

Shortly before the period of fertilisation, the wall of the
vegetative cell is absorbed, and at the same time the nucleo-
lus of the larger cell disappears, so that the two nuclei are
hardly to be distinguished.

Pollen-tubes are to be obtained easily by cultivation (3 to

POLLEN-BODIES OF THE ANGIOSPERMS.

27

5 per cent, strong, six hours). The nuclei travelling into
them are thereby elongated and become altogether similar
to one another (fig. 25).

Narcissus poeticus. — Here the development is precisely
similar, only that the vegetative cell is more spindle shaped,
almost like that in Ornithogalum.

r obtained Pollen-tubes of N. poeticus after four hours’
cultivation in a solution of 3 — 5 per cent, of sugar. In
general, the nuclei, as soon as they have entered the tubes,
cannot be distinguished, as they have become elongated early,
often even in the Pollen-body (fig. 26). In many cases, how-
ever, viz. where the nucleolus of the nucleus belonging to the
larger cell was still preserved, I was able to determine with
certainty that sometimes the nucleus of the vegetative cell and
sometimes that of the large cell went in front. I once saw,
what is very exceptional, the doubling of the posterior nucleus

(fig-

Iris sihirica is very difficult of examination owing toe
the richness of the cell contents and the thickness of th
extine. The extine can be removed and the conditions of
the investigation made easier if the Pollen-bodies are put
into a drop of water or of sugar solution, and the covering
glass is repeatedly raised up and down by means of a pair
of small pincers. By such manipulation the extine will
separate from several Pollen-bodies. If the Pollen-bodies
of older buds are subjected to this treatment it is sometimes
possible to press the whole contents of the large cell out,
so as to leave the small vegetative cell attached to the intine
(fig. 29). It must, therefore, be surrounded by a tolerably
resisting (cellulose?) membrane, and I actually succeeded
once in demonstrating such a membrane by a cautious
crushing (fig. 28). The vegetative cell is always formed on
the flattened side of the Pollen-body. Its nucleus has a small
nucleolus, the nucleus of the larger cell is a little larger and
has also a larger nucleolus. In the ripe Pollen-body the
two nuclei are almost unchanged, the vegetative one is some-
times naked and sometimes covered with a hyaline proto-
plasmic mass of a spindle shape, which represents the
vegetative cell set free.

Pollen-tubes were obtained after a cultivation of six hours
in a solution, 30 — 40 per cent, strong, of sugar. As soon
as the tube formation begins all signs by which one can
distinguish the two nuclei vanish. They travel rather
late, and often beside one another into the wide tubes, and
can then only be pointed out as bits of protoplasm, coloured
deeper than the rest by carmine, which have no clear outline.

28

FRED. ELFVING.

Iris xiphium ; the development is similar to that of I. sih~
irica. I once observed one of its vegetative cells^ whose
nucleus had divided. The nucleus of the big cell sometimes
goes first into the tubes and sometimes that of the small
one. They also often go beside one another. The substance
of the large nucleus becomes very much elongated on enter-
ing, and approaches the point of the tube by bending and
twisting itself. The vegetative nucleus retains its shape
even after the dissolution of its surrounding protoplasm
(fig. 30).

Tradescantia virginica ; here the nuclei are very strangely
formed (fig. 37), as Hartig has already remarked. One of
them is very long and narrow, curved, with bent-in, almost
rolled-in, ends ; it is, indeed, often of a trichina form. The
other is round and finally of the same star-shaped form
which is met with in Anthericum liliago. Both are devoid
of nucleoli. The developmental history tells us that the
former nucleus originates from the larger cell; while the
other sausage-shaped one is the considerably modified vegeta-
tive nucleus. I may here mention in passing that the manner
of division corresponded with that described by Strasburger
for the integument-cells of Nothoscordum fragrans Ueber
Befruchtung und Zelltheilung,’ p. 517, Taf. xxxiii, figs.
47—54).

I examined the Pollen-tubes, w'hich were prepared out of
the style. These tubes grow out of one end of the Pollen-
body. Already, whilst the two nuclei lie in the Pollen-body,
the round one becomes much elongated, and after the two have
entered the tube they are in every respect similar, becoming
very long and drawn out. I discovered, however, some
instances in which the vegetative nucleus lay in the Pollen-
body with its characteristically rolled- up ends, while the
other had already left it.

Sparganium ramosum shows a similar development to
that of Typha angustifolia. The formation of starch
(which hinders observation in Typha) begins, how-
ever, later, so that I can state positively that the vegeta-.
tive cell generally separates itself and assumes a spindle
shape (figs. 41, 42). The vegetative nucleus only becomes
free after further development (fig. 43), after which its
nucleolus vanishes (fig. 44). The nucleus becomes deeply
coloured by carmine. The nucleolus of the large nucleus,
deeper coloured, can generally be found in the ripe Pollen-
body.

The nucleus of the large cell sometimes goes first into the
Pollen-tubes (5 — 10 per cent., eighteen hours) and sometimes

POLLEN-BODIES OF THE ANGIOSPERMS.

29

the nucleus of the vegetative cell (figs. 45, 46). It, however,
also happens that the latter may remain behind in the
Pollen-body (fig. 48). The vegetative nucleus divides later
on in the tube, and this occurs botli when it is in front
as well as when it is behind (figs. 45, 47). I never found it
thus dividing while in the Pollen-body.

In P'othos Olfersii the two nuclei are similar also, being
both elliptic and lying beside one another.

Not to mention the division of the vegetative nucleus in the
Pollen-tubes of the Orchids and in Sparganium, I found that
division of the vegetative nucleus may occur in Iris xiphiumy
and a duplication of the vegetative cell itself in the Pollen-
body of Tiilipa gesneriana. In these species this is excep-
tional. In some Monocotyledons such a formation of several
vegetative cells is normal. This is well seen i\\ Andropogon
campanus. The Pollen-bodies of this species are spherical,
with a thin extine which, as-in all the Gramineae, is pierced
by a small orifice, through which later the inline grows out
into a tube. In the stage of development in which the
vegetative cell is formed, the Pollen-body contains only a
thin parietal layer of finely granular protoplasm, surround-
ing a large vacuole. These Pollen-bodies are very favor-
able objects for examination, as starch and other bodies
sometimes enclosed in protoplasm are altogether absent.
Indeed, I have found no other plant which showed the
formation and development of the vegetative cell so
plainly, without making use of reagents. Pollen-bodies can
be found in almost every anther ; even in blossoms which
are already ripe for pollination many Pollen-bodies appear
in a retarded state.

Originally the Pollen-body carries only one single nucleus
with a nucleolus. A small vegetative cell is then formed
diametrically opposite the orifice in the extine. It is of the
usual shape with clear protoplasm, and with a spherical or
oval nucleus, which is provided with a small but distinct
nucleolus. The nucleus of the large cell is generally
disk-shaped, and has a large, strongly refractive nucleolus
(figs. 54, 55). Typically, the vegetative cell next divides
itself into two equal sister-cells, one of which often divides
a<zain, so that we have finally three vegetative cells (figs. 58,
59). The nucleus of the large cell remains unchanged.
AVhen this is over the protoplasm increases in volume and the
Pollen-body fills itself with starch grains. Soon the vege-
tative cells are detached and float about freely (figs. 60,
61). Before the Pollen-bodies reach their ripe condition
several nucleoli are absorbed, the nuclei themselves undergo

30

FRED. ELFVING.

a lengthening out whereby the large nucleus visibly dimin-
ishes in volume.

Bromus erectus corresponds in many particulars with
Andropogon. The original cell-nucleus has often as many as
four nucleoli. The two daughter-cells have often more.
Division takes place in this species tolerably late, as
seems to be the case in most of the Graminese, in this case
when the ’anther is about a centimeter in length. The
vegetative cell, which is also here formed diametrically
opposite to the orifice in the extine (figs. 61 — 63), divides itself,
not immediately, as in the case of the last-named plant, but
after the nucleolus is resorbed, first loosens itself away
from the extine and now appears free in the surrounding
plasm in which it sooner or later assumes a somewhat elon-
gated form (figs. 64, 65). Then it becomes hidden by a
copious mass of starch grains I have only once seen the
division into two of the vegetative nucleus in the uninjured
Pollen-body (fig. 66), it is, oii the other hand, very easy
when the Pollen-bodies are placed in a 5 per cent, sugar
solution, to find conditions like those in figure 67, where the
vegetative cell has become divided into two still attached
cells, forming a sickle-shaped body. The two vegetative nu-
clei are alike oval, and without any nucleolus. Later on they
become considerably elongated, together with their surround-
ing cells, which are often found in the crushed-out contents
hanging on by their ends, the nuclei being then often crum-
pled up ; the nucleus of the larger cell with its larger nu-
cleolus being still unaltered (fig. 68). At the last even
this nucleolus becomes dissolved, whereupon the nucleus
stretches and curves itself, so that finally it is not to he dis-
tinguished from the two vegetative nuclei, the surrounding
protoplasm of which has meanwhile disappeared (fig. 69).

In Butomus umbellatm , as in all the Liliacese, the vegeta-
tive cell is formed directly opposite the opening in the extine,
and somewhat late, namely, just before the buds are un-
folded (fig. 75). The Pollen-body may remain in this state
until the formation of the Pollen-tubes ; a division of the
vegetative nucleus into two generally takes place, however,
either when it is still enclosed in the vegetative cell (fig. 76),
or after the partition wall is absorbed. Lastly, these two
vegetative nuclei are small and roundish, with an indistinct
nucleolus ; they are more deeply coloured by carmine than
the nucleus of the large cell, which preserves its nucleolus
a long time.

Tubes were formed, but rather sparsely, in a solution of
sugar, 5 per cent, strong, after a space of twenty to thirty

POLLEN-BODIES OF THE ANGIOSPERMS.

31

hours, and I very often found tubes which had reached a con-
siderable length, and appeared otherwise quite normal, to
have within them the nucleus of the large cell, while the
vegetative cell, or vegetative cells, remained on the intine
quite intact (figs. 78, 79). In those cases in which the
nuclei of the vegetative cells were free they generally,
though not always, entered the tube, and that later than the
large nuclei (fig. 77).

Alisma plantago corresponds wholly with Bromus, except
that the place where the vegetative cell is formed cannot be
precisely indicated on the spherical Pollen-body. Attempts
at cultivation were fruitless.

In Arum ternatum the vegetative nucleus divides also into
two. This can easily be seen in Pollen-bodies, which are
crushed, and then immediately fixed with osmic acid (figs.
70 — 72). The two small vegetative nuclei are then often still
surrounded vritli their vegetative protoplasm (fig. 73). The
nucleus of the large cells appears wrinkled up in the fresh
Pollen-bodies ; when it is subjected to pressure it assumes
the most strange forms, which are more or less star-shaped,
often even the glove-shaped” form of Hartig (fig. 74).
Cultivation proved barren.

In the Cyperacece I found the most complicated phenomena
of development. I specially examined Eleocharis palustris.
Almost all stages of development can be found in the large
spikelets. At the time when the lowermost blossoms have
opened, the Pollen-bodies in the upper ones are still united
and polygonal ; they have then a large nucleus in which
one almost always sees several nucleoli (fig. 82). Later
on the nucleus -divides, and, indeed, it would seem is
complete usually before the separation of the individual
Pollen-bodies. As soon, however, as the Pollen-bodies have
become free and have taken their definite, almost spherical
shape, a division takes place of that nucleus which lies
towards the apex of the cell (fig. 84). One of the so
formed sister-cells divides again, so that we thus have
four nuclei in the Pollen-body, three small, oval ones
lying close to one another in the point of the Pollen-body,
and one large, more central (figs. 85, 86). In most
nucleoli, often several, are to be seen ; that of the cen-
tral nucleus surpasses the rest in size. Only exceptionally
are four nuclei found in the point, giving a total of live.
The protoplasm, which fills the point and surrounds the
three little nuclei, often appears brighter than ^hat in the
other part of the Pollen-body, and occasionally indications
of plasmatic partition walls can be seen, which mark the

32

FRED. ELFVING,

boundaries of the nuclei ; these, however, disappear soon,
and the protoplasm then appears altogether homogeneous in
the whole Pollen-body.

The central nucleus now divides (figs. 87 — 89). One of
the sister nuclei is larger, and has a distinct nucleolus ;
the other is like those in the point, and, like these, has a
small nucleolus. The three small nuclei are gradually ab-
sorbed after this division ; they take colour in carmine solu-
tion, gradually less and less, and finally disappear alto-
gether (figs. 90 — 92). As the two remaining nuclei often
show several nucleoli, one might perhaps conjecture that
here it is not a resorption of the smaller but a conjugation
of the larger with the smaller nuclei that takes place, such
as is the case in the embryo-sac. Observation, however,
shows that several nucleoli may be already present before
the small nuclei are absorbed.

Of the two nuclei which are now present, the one pro-
vided with the smaller nucleolus finally divides again
(fig. 93) ; so that we have definitely three nuclei. One
larger, with distinct nucleolus, and two small, which
generally have either none or a very small nucleolus
(fig. 94).

When these divisions have been accomplished the Pollen-
body has become ripe. The extine has become clearly dif-
ferentiated, and the contents inblude starch granules, and
sometimes one finds ripe Pollen-bodies which appear to have
two cells (fig. 98) ; closer examination, however, shows
us that we have not here to do with a cell division in the
usual sense of the word. The partition wall has rather
been formed by an accidental fusion of peculiar inner
thickenings of the intine, which is normally somewhat
strongly developed in the apex of the Pollen-body (figs. 95
— 97). In one instance I observed a nucleus in this so-
formed chamber, which nucleus was clearly one of the three
originally lying in the apex ; its resorption had been pre-
vented by the early formation of the partition w’all.

The Pollen tubes (5 — 10 per cent., 5 hours) always grow
either from the base or from the side of the Pollen-body,
never from the apex. The nucleolus of the large nucleus
vanishes at this time, so that it can only be distinguished
by its size, and, in most instances, by its clearer colouring
from the two others. All three suffer a diminution in size.
The nuclei do not go into the tube in any particular order.
In most cases, perhaps, it is the large nucleus which goes in
front, and it is often elongated (figs. 99, 100). But other
combinations may take place (figs. 101, 102).

POLLEN-BODIES OF THE ANGIOSPERMS.

33

Of the other Cyperaceee examined Carex vulpina and
Cyperus hadius are, perhaps, less suited for examination
than Heleocharis. I can, however, state with certainty that
in all the development is precisely similar.

I have intentionally fully discussed the stages of develop-
ment in the Monocotyledons. These can generally be more
clearly seen in them than in the Dicotyledons, in which the
richer contents of the cells and the small size of the nuclei
make the examination more difficult. Besides, the Dicoty-
ledons show nothing extraordinary ; nothing w'hich does not
also occur in the Monocotyledons. I can, therefore, record
my observations of them in a small space.

In the Pollen-bodies of the Dicotyledons we know that in
general many orifices in the extine are formed. This, to-
gether with the more or less spherical shape of the Pollen-
bodies, makes it impossible to indicate the place where the
vegetative cell is formed with the exactness that one could
do in the Monocotyledons. In the ellipsoid Pollen-bodies of
the Papilionaccse and Umbelliferse, where the orifices for the
tube are formed in an equatorial circle, the vegetative cell
occupies a polar position. Even in the case of other Dico-
tyledons the position of the vegetative cell was never ob-
served just under the extine orifices. Nevertheless, it seems
to be the large cell that grows out to form the tube.

The vegetative cell is separated by a more or less convex
partition wall of cortical plasm from the large cell. Its
nucleus and nucleolus are always smaller than the same
parts in the large cell.

Th® vegetative cell soon separates from the intine and
appears as a spherical formation free in the Pollen-body.

As in many Monocotyledons, a division of the vegetative
nucleus can still take place ; so that three nuclei are pre-
sent in the ripe Pollen-body. This occurs in Samhuctis
race77iosus, Fedia cornucopicB^ Dahlia Merckii^ Nymphaea
alba, Biscutella eidgerifolia, Gei'aniu^n Hookerianum, A7'e?i~
aria laricifolia, Foeniculum officinale.

The two vegetative cells resemble one another, and after
a time have no nucleoli.

In Nymphaea alba the nucleus of the large cell gene-
rally preserves its nucleolus in the ripe Pollen-body, often
even in the tubes (1 — 5 per cent., twenty hours). I can
certify that this nucleus goes into the tube earlier than the
two other vegetative ones which lie beside one another. It
was only on very rare occasions that the opposite order
obtained.

In the other plants mentioned above, the three nuclei can-

VOL. XX. NEW SER. C

34

FRED. ELFVING.

not be distinguished with certainty during the process of
tube formation.

In most Dicotyledons no further divisions take place in
the Pollen-body. In the following only two nuclei are
found in the ripe Pollen-body, namely, Gilia tricolor^
Nicotiana tabacum, Salvia verticillata, Digitalis lanata^
Gloxinia lixjbrida, Torenia asiatica, Plantago media, Campa-
mda rapuncidoides, Bryonia alba, Lysimachia punctata,
Erica tetralix, Monotropa hypopitys, Peperomia claytonioides.
Cannabis sativa, Rhus glabra, Rida angustifolia, Ricinus
communis, Hippuris vulgaris. Ranunculus muricatus. Del-
phinium decorum. Clematis viticella, Papaver dubium, Viola
tricolor, Helianthemum polifolium, Ampelopsis hederacea,
Oxalis lasyandra, Malva caroliniana. Polygonum rubrum,
Begoniae sp., Sedum hybridum, Clarkia pulchella, Spirea
villosa. Mimosa brachybotrya, Lathyrus silvestris.

The development of all of these can be described in a very
few words. The vegetative cell, which has become free, either
remains spherical or becomes, as is more usual, spindle shaped.
The parietal layer of cortical plasm sooner or later vanishes, j

and the two nuclei become mutually freed. These nuclei are, |

in their younger conditions, very easily distinguished. The
nucleolus of that nucleus, w^hich originates from the large I

cell, remains much longer thap the vegetative one. This ■

distinguishing sign vanishes, however; and in, or even earlier |

than, the tube-formation period, by successive metamorphoses '

both in form and size, the two nuclei become so like, that it
is impossible to distinguish them. Only in Cannabis sativa
is the nucleolus of the nucleus of the large cell preserved
in the tubes (10 per cent., twelve hours). In this case some-
times one, sometimes another nucleus goes in front. In
Monotropa the two nuclei preserve their nucleoli very long
(5 — 30 per cent., twenty hours). An absolute distinction
between them w’as nevertheless impossible.

Almost all possible differences occur in respect to the form
of the nuclei, from the round’ one of Rhus to the greatly
elongated, almost thread-shaped nucleus of Sedum. Gene-
rally they are almost elliptic in shape.

Putting together the chief results of my examinations it
would appear that in a particular stage of development before
fertilization the Pollen-body of the Angiosperms is divided
into two cells — a larger and a smaller, the ^G^egetative ”
one — which latter, by further divisions, can form a two- to
thrce-celled tissue (tliallus).

This vegetative cell, or these vegetative cells are separated

j?OLLEN-BODIES OF THE ANGIOSPERxMS.

35

as a group from the larger cell only by a wall of cortical
plasm. In some isolated cases this may become a resisting
(cellulose ?) membrane.

The Pollen-tube is formed from the large cell. It may
happen that the vegetative cell or cells have nothing to do
with this occurrence, so that the nucleus and contents of the
large cell alone immigrate into the tube. The separating wall
is generally, however, absorbed. It may vanish altogether
after the division; in most cases it however remains for a
certain time ; the whole vegetative cell or cells loosen them-
selves from the inner wall of the Pollen-body, and are then
surrounded by the large cell, which appears strangely
spindle- or half-moon shaped. The vegetative cell may
remain in this state a longer or shorter time, or its nu-
cleus may divide, and thus free-swimming vegetative cells
be formed. In either case the "wall of plasm is finally
dissolved ; this may take place either in the Pollen-body or
after the vegetative cell has gone into the tube. A division
of the naked vegetative nucleus may take place after the
disappearance of the wall, and this may also occur either
in the Pollen-body itself or in the tube.

The nuclei have often strange forms. I have noticed no
division of the nuclei of the large cell, except in the
Cyperaceae.

A particular order is not generally kept in the migration
into the tube. The nuclei are dissolved sooner or later ; but
in any case before fertilisation takes place. The large cell
of the Pollembody and its nucleus appear to be of more im-
portance for fertilisation than the vegetative one. I arrive
at this conclusion from the following facts, viz. that it is
the large cell which grows into a Pollen-tube ; from the
circumstance that in some cases the nucleus of the large cell
always goes in front, while I never saw the opposite case as
a constant characteristic; that in those plants where this is
not the case, the nucleus of the large cell still goes oftener
ill front than the other; finally, that in some instances the
vegetative cell remains in its original position without
entering the tube at all.

36

F. O. BOWER.

On the Developmej^t of the Concept acle in the FucacEvE.

By F. O. Bower, B.A.., Trinity College, Cambridge.

With Plate V.

The first careful notice of the conceptacle which I have
met with is that in the ^ Phycologia Generalis ’ of Kiitzing
(1843).

He describes (p. 98) it as a roofed-in sorus (einges*
tiilpter sorus) ; while the tissue which lines the cavity is, he
says, nothing else than a slightly modified continuation of
the limiting tissue^ (cortical schicht).

Speaking (p. 92) of the Fasergrubchen ” (a word which
has as yet no English equivalent), he says they seem to stand
in a certain relation to the conceptacle (Hiillenfrucht),
although they are found on such of the brown seaweeds
as have no conceptacles (cf. Alaria esculenta).

Agardh (^Species et ordines Algarum,’ 1848, vol. i,
p. 101) suggests that the Fasergrubchen ” may be the
equivalent of the conceptacles (scaphidia) in the fertile part
of the plant.

Sachs Lelirbuch,’ 1874, p. 283) remarks that the layer
of cells lining the cavity is a continuation of the outer limit-
ing layer of the Thallus.

The first attempt at an accurate description of the de-
velopment of the Fasergrubchen or conceptacle was made
by Reinke (^Bot. Zeit./ 1875; and ^ Nachrichten der K.
Ges. d. Wiss. zu Gottingen,^ 1875, p. 230). His results
were republished in PringsheinTs ‘Jahrbuch,’ x, 1876,
p. 317.

Speaking of Fucus vesiculosus, he says (p. 337) that the

Fasergriibchen,^’ which he seems to take as the type of
these structures, originates, as seen from above, by a sepa-
rating of four or five neighbouring cells of the limiting tissue
from one another. He compares this process with the forma-
tion of the resin passages in the Coniferse. The ‘‘ intercel-
lular space” thus formed is filled with mucilage. Longitu-
dinal sections showed him that not only cells of the limiting
tissue, but also cells of the subadjacent cortical tissue, separate

‘ RostaGnski (‘ Beitrii^e zur Kenntniss der Tange,’ Heft i, 1876, p. 5)
has already objected to the use of the term “epidermis,” in the case of
Fucus, on the gropd of the outermost layer of cells being capable of con-
stant tangential division. Pie uses the term “ aussenrinde,” which I pro-
pose to render loosely by the term “ limiting tissue,” reserving the term
“ cortical tissue,” for his “ innenrinde.”

DEVELOPMENT OF CONCEPTACLE IN THE FUCACE^. 37

from one another. An intercellular canal” is thus formed,
stretching through the limiting tissue into the cortical tissue.
He further tells how the cavity thus formed becomes flask
shaped, and the cells which line it put out papillae, which
further develop into hairs. He assigns to the conceptacle
a similar history, and hence concludes that it is ‘‘ homolo-
gous” with the Fasergrubchen.” Though he records
observations on a considerable range of allied plants, he finds
no noteworthy deviation from the type of Fucus vesciculostis.
He gives no figures illustrating these observations.

Luerssen Handbuch der Syst. Bot.,’ 1879, Bd. i, p. 105)
gives a different history. He says small spots of the surface
of the thallus are so overgrown by the surrounding tissue
that only a narrow opening remains.

Thuret does not mention the subject of development of the
conceptacle in any of his publications to which I have had
access.

Preliminary Pemarks.

Many of my preparations of Fucus serratus were made
from materials collected in August, 1878 ; these were treated,
while fresh, with a dilute solution of chromic acid in water,
and afterwards preserved in alcohol. The results obtained
from these materials have been verified in other specimens,
collected in August, 1879. These were preserved in a satu-
rated solution of common salt, and then hardened in alcohol.
The latter method of preparation has been used for all the
other plants of the group with which I have worked. The
sections were in all cases mounted in glycerine and acetic
acid.

The youngest stages of development of the conceptacle are
naturally to be found close to the apex of the branch, and
only on those branches which still retain an active apical
growth. Such branches may be recognised by their possess-
ing a well-marked depression at the apex.

Regarding the sexual conceptacle as the type of such
structures, I have studied the development of it first, and
tlien compared with it the development of the “ Faser-
griihehen,” which I regard as an incomplete sexual
conceptacle.

In describing the planes of section relative to the thallus,
I use the terms vertical longitudinal section ” and trans-
verse section ” in the same sense as Rostafinski Beitriige
z. Kennt. d. Tange.,’ Heft, i, 187G, p. 17), the former being
a longitudinal section in a plane perpendicular to the
fiattened sides of the thallus; the latter cutting the organic
axis at right angles.

38

F. 0. BOWER.

0

Fucus serratus.

I chose F. serratus as the species best fitted for following
the stages of development of the conceptacle for these reasons.
(1) The fertile branches (Bliithen of Reinke) are less swollen
than in other species^ and it is therefore easier to obtain
sections through the apex ; while the tissues are more com-
pact. (2) The hairs, which in later stages fill the cavity
of the conceptacle, are not developed so early here as in
other species. (3) F, serratus presents conceptacles of all
ages in the month of August, at which time only I have had
the opportunity of making collections.

The first traces of the conceptacle are best observed in
vertical longitudinal sections.

In £he fertile branches of F. serratus I have noticed no
deviation from the form of the apical cells, and the succession
of their segments, described by Rostafinski (loc. cit.) for the
sterile branches of F, vesciculosus. The division of the lateral
segments cut off from the apical cells also appears to follow
the type of his fig. 14, each segment dividing first by a wall
parallel to its free surface Basalwand ; the outer of the
two cells thus formed again divides in two planes at right
angles to one another and to the Basalwand.” Four cells
are thus formed, each of which may again divide according
to the same law as the original segment ; this is the constant
law of division of the cells of the normal limiting tissue of
the younger branches, be they sterile or fertile (cf. fig. 1,
groups marked 1).

It is, however, in a modification of this succession of
divisions that the first traces of the conceptacle make their
appearance. Just as noticed by Rostafinsky (loc. cit., p. 9)
for older branches, the division of the outer cell by vertical
walls into four ceases in certain cases. The division by
walls parallel to the surface (Basalwande) , however, continues.
A linear series of cells is thus formed which may be traced
some distance into the tissues, but which is terminated by a
single cell only (fig. 1). Later, the activity of division in
the horizontal direction also ceases, and as the terminal cell
of the series does not increase in size, the result is that it is
surpassed by the tissues surrounding it.

The terminal cell of this series we may call the initial
cell of the conceptacle, the cell immediately beneath it may
be termed the basal ” cell.

As the cells which abut laterally on the initial cell retain
^ Thereby conveying no functional connection with an a))ical cell.

DEVELOPMENT OF CONCEPTACLE IN THE FUCACEJE. 39

the usual law of division, the result is that it is bounded
laterally (as in fig. 1) not only by cells of the limiting tissue
(aussenriiide), but partly also by cells of the subjacent cortical
tissue (innenrinde).

Up to the stage represented in fig. 1 no important change
is to be observed in the initial cell. It contains a nucleus
and plentiful protoplasm, and its walls are normal.
There may, however, be noticed at the apex the beginning
of a separation of its cell-wall from those of the cells around
it. In the latter the only modification of the usual divisions
is an inclination of the vertical wails towards the initial
cell.

Conceptacles of this stage of development are to be found
on the inner side of the lips bordering the terminal cavity of
the fertile branch.

Fig. 2 represents part of a vertical longitudinal section
passing through a slightly older conceptacle. Here the
initial cell has lost its internal tension. The cell-walls have
collapsed, while the substance of the cell-walls themselves
has been converted into the swollen mass which fills the
cavity thus formed. This mass is continuous with the layer
which covers the surface of the thallus ; it already shows
signs of the change which it undergoes later.

The upper part of the initial cell has shrivelled, and
appears to have no cell contents ; the lower part, which
adjoins the basal cell, still contains^ a small quantity of
vacuolated protoplasm. The basal cell has meanwhile
increased in size, but has not yet divided. If figs. 1 and 2
are compared, it will not be difficult to assign their origin to
the cells which border on the cavity in fig. 2.^ It will be
seen that, as Heinke rightly observed (loc. cit., p. S87), they
are of different origin.

As foreshadowed in fig. 1, the cell- walls of the limiting
tissue are strongly inclined in fig. 2 tOAvards the centre of
the conceptacle. The group marked I being most so, it is
between the group I and the group c that the boundary of
origin of the tissues lies, the group I being derived from
the limiting tissue, the group c by the division of a cell .
of the cortical tissue. Those cells, then, which border the
lower part of the cavity of the conceptacle (including the
basal cell) are derived from the cortical tissue ; those which
line the upper part from the limiting tissue.

A surface-view of a conceptacle, intermediate in develop-

‘ 1 do not mean to imply by such a comparison that the order of the cell
divisions is constant. A mere glance at the figures will be sullicient to
preclude such an idea,

40

F. O BOWER.

ment between those of figs. 1 and 2, is represented in fig. 3. 1

Among the normal cells of the limiting tissue, which appear
thin walled and full of granular protoplasm, one is seen with
contracted contents and swollen cell-wall. This is the
initial cell.

The further development of the conceptacle as shown by
section is marked by continued shrinking of the initial cell,
and a gradual decomposition of the swollen substance which
fills the cavity. This is accompanied by divisions in the
basal and surrounding cells. In fig. 4 it will be noticed
that the basal cell has divided in a plane parallel to the
axis of the conceptacle. In the swollen mass which fills the
cavity may be recognised irregular patches where the sub-
stance has undergone alteration, accompanied by change in
optical properties. This is still more evident in fig. 5, which
is part of a transverse section of the thallus.

Here the change of the substance filling the cavity has
advanced so far that the unaltered portion immediately
surrounding the remnants of the initial cell forms a
central column of irregular outline. This stretches from
the basal cell to the neck of the conceptacle, and is con-
nected with the walls of the conceptacle by thin strings
which, like itself, have remained as yet unchanged.^

' We must draw a distinction betweei\ these two stuffs : — that which
forms the central column and the layer which overlies the whole surface of
the plant ; and h, that which is formed at first in irregular patches, but
which finally fills the greater part of the cavity of the conceptacle. The
reactions, which are described later, bring to light a third body (c), which
constitutes the outermost layers covering h at the exterior of the thallus
(cf. reaction 6). The following observations will serve to point out their
characteristics :

1. The substance a has different optical properties from h, so that tliey
are easily distinguished under the microscope ; a usually has a yellowish
appearance.

2. The substance a does not change volume to any marked extent on
dehydration ; h, when dehydrated, contracts strongly, so that the concep-
tacle, when mounted in alcohol, appears only partially filled. When water
is added the contracted mass swells.

3. Solution of iodine does not colour a, b, or c.

4. When treated with a solution of chromic acid in water, the substance

h becomes more transparent, and swells strongly, a is only slightly swollen ;
and since its appearance is not much altered, it becomes more prominent,
owing to the change in b. j

5. With Schultz’s solution neither a, b, nor c give a blue colour, b swells
strongly ; a is more resistant. [N.B. With this reagent the contents of
the initial cell sometimes give the brown reaction of protoplasm; but this '*
colouration is not constant, and probably depends upon the state of decom-
position of the contents.]

0. Treated with concentrated sulphuric acid, b is immediately dissolved ;
a resists for a time, and is later dissolved ; but a central portion remains

1

DEVELOPMENT OF CONCEPTACLE IN THE 1’UCACE.E. 41

Besides the division shown in fig. 4, the basal cell at this
time divides also in two planes at right angles to this and to
one another, giving rise to a group of eight cells, four of
which only can be seen at once in a section (fig. 5). These
cells, during the further development of the conceptacle,
appear to differ functionally in no respect from the other
tissues bordering on the cavity. These tissues, having
increased principally by divisions at right angles to the
inner surface of the conceptacle, form a layer of thin w'alled,
closely compressed, protoplasmic cells, which completely line
the cavity; by these peculiarities this tissue is pretty sharply
marked off from the adjacent cortical tissue.

Coincident with this increase in number of the constituents
of the lining tissue is a change in form of the whole con-
ceptacle. The low-er part of the cavity becomes wider, so
that the whole conceptacle assumes a flask form, which may
be recognised in fig. 5, but becomes more pronounced at the
stage represented in fig. 6.^

In the stage represented in fig. 6 the central column may
often be found, still continuous, from the point, where the
initial cell is fixed, to the neck of the conceptacle, which is,
however, still closed. The column shows a jagged irregular
outline, owing to the rupture of the strings which originally
connected it (as in fig. 5) with the walls of the conceptacle.
It may be noticed that the margin of the cavity is still
unbroken by any outgrowth of hairs.

which is then seen to be continuous witli the outer portion of the thickened
covering of the limiting tissue, this we have designated c.

The conclusions to be drawn from these observations are, that 'a is a
swollen form of cell-w'all, which is not true cellulose, but is similar to the
central portion of older cell-w'alls of the tissue of the larger tangles
(“Innenlamelle,” Luerssen, p. 101 ; ‘"Gelin” (?), Kiitzing, ‘Phyc. Gen.,’
p. 31). That 6 is a substance coincident with mucilage (“ Schleim,”
Kiitzing, ‘ Phyc. Gen.,’ p. 30) ; that is a substance akin to cuticle.

^ It seems to me probable that the power of swelling with water possessed
by tlie mucilaginous contents of the conceptacle, and the consequent internal
pressure exercised on the internal surface of the conceptacle, has some con-
nection with this form. Suppose such an internal pressure. The concep-
tacle is completely closed. If a yielding occurs it will naturally be at the
point where the resistance is least, and that will be where the tissues are
least crowded. This is, however, the case in the interior of the lhallus.
The cavity will then enlarge most at its base. Hence the flask shape.

This explanation is further supported by the fact that where the curva-
ture is greatest (and the form of the conceptacle is usually less regular than
in our fig. G) there the cells of the lining tissue are largest. Hence it may
be concluded, other things being equal, that there the external resistance
was least. This supports our view. Again, it is universally the case that
around the mouth of a conceptacle the surface of tlie thallus is raised so
as to form a slight liillock (cf. figs. C, 7). This would naturally be the
result if an internal tension existed

42

F. O. BOWER.

Not having the opportunity to make preparations from
fresh materials, I was forced, in verifying these results,
obtained with chromic acid materials, to use specimens pre-
served in saturated solution of common salt, and afterwards
hardened in alcohol. Sections from these, mounted in
glycerine and acetic acid, gave the same results in all impor-
tant points ; the distinction between the central column
and the mucilaginous mass surrounding it being as distinct
as in the chromic acid materials. When first mounted,
however, in glycerine, the mucilage was marked by a very
definite and beautiful system of striation and stratification ;
this was more striking towards the periphery of the mass,
towards the centre the marking gradually lost its regularity.
After the sections had lain in glycerine for a short time the
appearance faded and disappeared, probably owing to a
gradual swelling.

The direction of the systems is represented in fig. 7. In
this figure it will be noticed that the outline of the cavity is
irregular. This is due to the outgrowth of lumps of tissue,
which, especially in male conceptacles, precede the formation
of hairs.

Here we come upon the first evidence of a sexual difference.
The conceptacles have thus far, in the dioecious plant,
offered no variety in development ,according as they are male
or female ; but in the development and character of the
hairs there is an appreciable difference. If the conceptacle
be male the formation of hairs is preceded, as in fig. 7, by
the outgrowth of irregular masses of tissue. The surface
cells of these grow out into papillae, which divide and form
hairs, these hairs branch according to a monopodial racemose
system ; owing to their mode of origin these primary hairs
are usually associated in bunches.

If the conceptacle be female the hairs arise more uniformly
over the inner surface of the conceptacle, and are, as a rule,
not branched. It is only after these hairs have begun to be
formed that the initial cell and central column are thrown
off. Up to this stage it may, in thick sections, be almost
always recognised. Though the central column is detached,
still the conceptacle remains closed at the neck for some
time longer.

The antherida are not formed on the end of the upper
branches of the primary hairs ; the lower lateral branches of
them are, however, at an early stage terminated by antheri-
dial cells. Antheridia are formed also on the apex of small
hairs developed for that special purpose by the outgrowth of
cells of the tissue lining the cavity, and these show, in its

DEVELOPMENT OF CONCEPTACLE IN THE FUCACE^. 43

simplest form, the mode of branching which accompanies
the formation of antheridia (fig. 8, i, ii, hi).

A single cell grows out to form a papilla. This divides
(8, i) by a transverse wall. An apical antheridial cell is
thus formed, which is supported by a single pedicel cell.
The latter may now continue growth and put out a lateral
papilla (8, ii), which is functionally equivalent to the
original papilla.

In fig. 8, hi, a represents a system of antheridial branching
of a higher order. i is an antheridium from which the
antherozoids have already escaped, the primary axis was
then p — a i, a ii is an antheridium which was formed on
the apex of a lateral outgrowth below a i. a Hi is the lateral
papilla which precedes the formation of a third antheridium.
The whole system is then, in this case, a sympodial system, of
the type represented by Sachs in his ^Lehrbuch,’ 1874,
fig. 136, D (cf. 8, iii, b). It is only rarely that the branch-
ing is as simple as in the cases represented, since it is
often not confined to one plane, and the antheridia are
usually more crowded together (compare Thuret’s figures,
especially in his ^ Etudes Phycologiques,'’ plates xv, 6 ; xxii,
8 — 10, xxiv, 6). Often the system approaches irregularly
the type figured by Sachs (loc. cit., fig. c), but, however
irregular, the antheridial branching seems always to be
sympodial.

On the mode of formation of the oogonia I have no remarks
to offer.

The difference between the Fasergriibchen ” and the true
conceptacle is clearly marked in the mature state both by
the form and the contents. But as we pass back to the
earlier stages of development the distinction becomes less and
less, till w^hen we reach the condition in which the formation
of hairs has begun in neither of them, it is impossible to say
definitely, from mere inspection of the structure, whether
the cavities will produce sexual organs or only hairs.

Still, if we take young plants and study the development
of the cavities formed at their apex we may, with the greatest
probability possible in such a case, regard such cavities as
early stages of the Fasergriibchen.”

This I have done, and found their development to differ in no
point of importance from that of the sexual conceptacle. The
initial and basal cells are present ; the form of the cavity is
the same. The central column appears just as in the sexual
conceptacle.

The hairs themselves are multicellular but unbranched.
The neck is usually wider than in the sexual conceptacle.

44

F. O. BOWER.

•It does not open till the apices of the hairs reach the swollen
mass, which closes it and push their way through it. The
mass itself has been previously undergoing a process of
disorganisation and swelling similar to that which takes
place at an earlier period around the central column.

Seeingthen that, i, in their development these two structures
are so similar; ii, that Fasergrlibchen do not occur
among the sexual conceptacles on a fertile branch, t. e, the

metamorphosis ” is complete ; iii, that no function has
been assigned to the Fasergriibchen, except that they act
physiologically as root hairs (Reinke, loc. cit., p. 321) ; iv,
that they are completely absent from some plants, e. g.
Halerica (Reinke, loc. cit., p. 360) and Himanthalia lorea ;
I think the homology of the two structures is so clearly
proved that I shall be justified in proposing instead of a
translation of the German word Fasergriibchen,” we
should use the name neutral conceptacle ” to denote
these neutral structures. This may be objected to as a con-
tradictory term, but I think it is important to convey at once
the relation which appears to exist between them and [the
true conceptacle. Since, however, the term conceptacle has
now obtained a morphological rather than a physiological
significance, the absence of the main function can hardly be
taken as ground for refusing the term.

Fucus platycarpus.

The sexual conceptacles (^. e. those developed on the
swollen branches) show in their earliest stages a structure
corresponding closely with that of the sexual conceptacle of
F. serratus, having an initial cell, terminating a basal series,
which may, as in .F. serratus^ be traced into the substance of
the thallus. In later stages, however, the initial cell appears
in some cases to retain vitality and to divide, forming a
central hair, of which the apical cell is partially decomposed
and shrivelled ; in other cases it remains inactive, and
decomposes as in F. serratus. The central column is not
so clearly marked as in F. serratus^ since it is very soon
thrown off, owing to the early formation of hairs lining the
cavity. For the same reason the identity of the initial cell
is often lost at an earlier stage than in F. serratus. The divi-
sion of the basal cell appears to be less definite and regular
than in the former species.

In no point of importance were the neutral conceptacles
found to differ from the sexual ; the hairs are formed earlier
from the lining tissue than in the sexual conceptacles.

DEVELOPMENT OF CONCEPTACLE IN THE FUCACE.E. 45
Fucus vesciculosiis.

The conceptacles of F. vesciculosus resemble those of
F. platijcarpus rather than those of F. serratus, but the
division of the initial cell is less frequent than in the former
species. The development of the neutral is similar to that of
the sexual conceptacles.

Ozothallia nodosa.

My materials, collected in August, 1879, supplied only the
youngest stages of conceptacles. In these the initial cell and
the divisions of the surrounding tissue correspond to those in
the genus Fucus. The basal cell is, hoAvever, not so promi-
nent as in F. serratiis. In this point, and also in the general
form of the conceptacle, Ozothallia appears to hold an inter-
mediate position between Fucus and Himanthalia.

Halidrys siliquosa, ^

Viewed from above the young conceptacle of Halidrys
appears in all respects similar to that of Fucus. In longi-
tudinal section the initial cell appears in some cases as in
Fucus ; but, in the large majority of cases, divisions occur in
it, resulting at an early period in two or, in some cases, more
than two cells. The basal cell divides at first by avails
strongly inclined to one another. The cells thus produced
do not at an early stage divide by walls parallel to the
surface of the cavity. The result is that the conceptacle
usually appears, as in fig. 9, as though lined by a layer of
cells continuous with the limiting layer ; but as part, at
least, of this tissue is derived from the basal cell, this con-
clusion is inadmissible. Meanwhile, the initial cell (or grou])
of cells) has been completely thrown off by the s^velling of
the wall dividing it from the basal cell.

Later, as in the other plants of the group, the cells of the
lining tissue put forth papillae which develop further into
hairs.

Himanthalia lorea.

Here we have a departure from the type of Fucus. Close
to the apical cell of the fertile branch the cells of the limiting
tissue, by means of repeated divisions parallel to the surface
of the thallus, grow^ out into multicellular hairs, almost
every cell bearing one. As the cells of this tissue con-
tinue to divide, some of the hairs arc thrown off, and the rest
remain scattered over the surface. These isolated hairs
consist of an apical part, in which the cells are in various
stages of decomposition ; a central part where the cells form

46

F. O. BOWER.

a beaded series ; a basal part consisting of cells with only
small quantity of protoplasm and slightly swollen cell-walls ;
this lower part of the hair gives the impression of having
shrunk similarly to the initial cell of the conceptacle of
Fucus. Tracing this hair back it may be seen, when viewed
from the surface, to pass into a cavity between the adjoining
cells of the limiting tissue (fig. 10). This cavity is formed,
as in Fucus, by the shrinking of the lower cells of the hair
and swelling of the contiguous cell-walls, so as to fill the
space thus formed. Fig. 11 shows a conceptacle of Himan-
thalia at a slightly older stage in longitudinal section. The
upper part of the hair has broken away, leaving only a single
cell of the series behind. This cell is shrivelled and the
cell-walls swollen, giving it the same appearance as the
initial cell of Fucus, of which it is undoubtedly the equiva-
lent. The form and size of this cell is very inconstant ;
sometimes two such cells of the initial series are left behind.
There is no cell in Himanthalia which may be easily
recognised as corresponding to the basal cell of Fucus. This
is probably owing to the distortion of tissues which often
occurs at a little distance from the apical cell.

During this change in the original hair, the cells of the
limiting tissue, which immediately surround it, do not grow
as rapidly as the adjacent tissue ^they only divide rarely in
a direction parallel to the surface ; but more frequent and
active division takes place in a direction perpendicular to it.
The result is a depression of the surface around the initial
cell. The tissue lining the cavity thus formed is therefore
derived only from the limiting tissue ; no part of it is derived
from the cortical tissue. The further development of the
conceptacle is carried on chiefly by divisions according to
the same rule as at first. The region of greatest activity
is at the base of the cavity. Thus the conceptacle attains
the form of a long tube, which widens later at the base,
so as to assume the usual flask form. As this widening pro-
ceeds, hairs are formed, as in Fucus, by outgrowth of single
cells of the lining tissue. In a female conceptacle these
hairs are multicellular but unbranched. In the male con-
ceptacle the hairs are branched, at first on a monopodial
system developed in a racemose manner ; i. e. the apical cell
divides transversely, and the upper cell grows on as before ;
the lower cell puts out a protuberance immediately below
the dividing cell-wall. This does not, however, overtop the
apical cell (cf. fig. 12). The apical cells of such hairs do not
usually form antheridia; only the lower branches undergo
the change. Coincident with the formation of the an the-

DEVELOPMENT OF COXCEPTACLE IN THE FUCACEyE, 47

ridia appears an alteration in the system of branching from
inonopodial to sympodial_, as will be seen from fig. 12.
Each of the lower lateral axes is terminated by an anthe-
ridium (1) which is cut off* by a transverse cell-wall. This
stops further growth in that direction. As before shown in
the case of Fucus, below this cell-wall appears a lateral
branch, which is in its turn again terminated by an anthe-
ridium (2), and so on. The system of branching is thus
rendered sympodial by the formation of antheridia ; the
apparent axis being a pseudo-axis, composed of a series of
pedicels of successive antheridia. As to the arrangement of
the successive antheridial cells, the same remarks will apply
here as were made upon the antheridia of Fucus serrahis.
As in Fucus, the antheridia are not only formed on the
lower branches of primary hairs ; unicellular papillae are
here also formed by outgrow’th of single cells of the lining
tissue ; these develop as described in the case of Fucus
serratus (cf. fig. 8, i, ii, iii).

General remarks.

In drawfing our conclusions from the facts before us, the
most striking point is that, in all cases described, the forma-
tion of the conceptacle is preceded hy the decay of one or more
cells lohich occupy a central position with regard to the
changes which folloio. The number of the ceils thus re-
moved is various, and the manner of their destruction is not
constant ; but the fact remains in all cases. I must own
inability to suggest a parallel to this.^

A point which is not so obvious, but 'which appears of
similar constancy, is that the cell or cells tchich decay are in
all cases members of a linear series. It depends upon the
activity of division, in a direction tangential to the surface
of the thallus, how this series is characterised ; -whether, as
in Fucus, w'here the division is slow, and even ceases, the
apical cell of the series hangs behind the surrounding tissue;
or whether, as in Himanthalia, where the division is often
repeated, the series is elongated, and, protruding beyond the
surface of the thallus, is called a hair. The small import-
ance of this difference is shown by the variety -which exists
in this respect in specimens of the same species, under dif-

‘ Reinke (‘ Eut\A\ Unters ; iiber die Dictyotaceeii,’ 1878, p. 47) speakiug
of the hairs, whicli he termed “ Sprossfadeii,” remarks that they arc always
unbranched ; they are grouped in slieaves or rows, and grow, at least later,
exclusively by division of their basal cells; they form the precursors of the
reproductive cells, and fall off when these appear. How far these may be
compared with the initial cell or hair of the Fucaceai it remains for closer
observation to decide.

48

P. O. BOWER.

ferent circumstances ; e. g. the bunch of hairs at the apex of
young Fucus plants is due to an unusual activity of tangen-
tial division in the cells of the limiting tissue. This pheno-
menon again appears at the apex of older branches of
Fucus, whose activity of apical growth is in any way
hindered (Reinke, loc. cit., p. 338). And Rostafinski (loc.
cit. p. 9) notes how the cells of the limiting tissue of older
branches divides by tangential rather than longitudinal
walls. The activity of tangential division being then so
variable, before asserting that the condition described in the
case of Himanthalia is the normal one for that species, it
wdll be necessary to compare the results here obtained with
those drawn from materials collected at a different time of
year. I think it quite possible that the plants I have used,
being collected in August, were in a semi-dormant state
previous to their period of active grow’th and reproduction
in the winter. If so, the hair, as represented in fig. 10, may
be compared with those at the apex of the young plant of
Fucus, or of the older branch of Fucus living under diffi-
culties. On this ground I shall not attempt to suggest
which is the simplest form, or the typical mode of forma-
tion of the conceptacle.

This variation in the activity of tangential division ac-
counts for the want of uniformity in number of the cells
thrown off in different species, and even in the same species
(this was especially noticed in Halidrys and F. platycarpus).
In viewing the formation of the conceptacle of the whole
group (as far as studied), we may say generally, that the
differences in mode of development (in the early stages at
least) depend upon the difference in activity of tangential
division of the cells of the central series ; for it will be seen
that upon the behaviour of the central series in relation to
the surrounding tissue depends also the origin of the lining
tissue of the conceptacle.

In Fucus the terminal cell of the central series (initial
cell), ceasing to divide tangentially and being left behind by
the surrounding tissue (fig. 1), when that cell decays a
cavity is formed, which extends further into the tissues than
the base of the cells of the surrounding limiting tissue. The
tissue lining the cavity is therefore in this case derived in
its basal part from the cortical, in its upper part from the
limiting tissue. In Himanthalia, however, the cavity thus
formed only extends at most to the base of the cells of the
limiting tissue ; the layer lining the cavity is thus derived
only from the limiting tissue. This distinction of origin
must not, however, be put on a level with that from Der-

k

DEVELOPMENT OF CONCEPTACLE IN THE FUCACEvE. 49

matogen or Periblein of the higher plants. The mere fact
that the cortical tissue of Fucus is derived by constant
direct division from the limiting tissue is sufficient to show
that the two are in the closest genetic connection, and that
therefore the difference of origin of the lining tissue of the
conceptacle is of very slight importance.

Upon the homology of the sterile conceptacle Faser-
griibchen”), and of the sexual conceptacle, I need here
offer no further remarks.

In comparing the neutral hairs in the male and female
conceptacjes of dioecious species, I find it the rule that no
hranching occurs in the neutral hairs of the female concep-
tacle. It may be noticed in connection with this that no
secondary lateral formation from the pedicel of the oogo-
nium has been observed. In the male conceptacle even
before the formation of antheridia the hairs hra7ich, they do
so according to a monopodial racemose system. It has been
seen (cf. fig. 12) how the formation of antheridia modifies
the branching to a sympodial system. The antheridia were,
however, observed to arise also on the apex of outgrowths
from single cells bordering on the cavity of the conceptacle
(fig. 8). Comparing antheridia in the state represented by
fig. 8, i, with the oogonia as represented by Thuret (Etudes
Phycologiques, plate xii), we cannot fail to see a complete
morphological identity. Both are produced by the out-
growth of a cell bordering on the cavity of the conceptacle.
In both cases a cell-wall tangential to the inner surface of
the conceptacle divides an apical sexual cell from a basal
neutral cell. It is a matter of but minor importance that
the basal cell of the antheridium may by lateral growth
produce a second antheridial cell. This may be regarded as
an outcome of the general rule, that where the male organ
is endowed with free motion, it is produced in greater pro-
fusion than the female organ.

We may conclude then that the antheridium of the Fucus
group is, at least in some cases, morphologically identical with
the oogonium.

Lastly, I must record my thanks to Professor de Bary, of
Strasburg, for his kindness in allowing me the use of his
laboratory, and also for his interest and suggestions during
part of my work.

VOL. XX. NEW SER.

D

50

D. D. CUNNINGHAM.

0)1 Certain Effects of Starvation on Vegetable and Animal
Tissues.^ By D. D. Cunningham, M.B., Surgeon, Indian
Medical Service; Bellow of the Calcutta University.

That insufficient supply of nutritive material is closely con-
nected with fatty degeneration of tissue is a well-ascertained
pathological phenomenon. Rindlleisch, in his work on ^Patho-
logical Histology,^ expresses himself as follows in regard to this
subject: — ^^This (the exclusively pathological category of fatty
metamorphosis) includes all cases of disproportion between the
means of nutrition and the parenchyma to be nourished. Such
a disproportion may be caused either by a diminution of the
nutriens, or by an increase of the nutriendum. If a minute
vessel in the brain is plugged, the circulation in the area which
it supplies is not wholly suspended, owing to the manifold
anastomoses with neighbouring vessels ; nevertheless, a very
considerable retardation of the current takes place, which may
even give rise to temporary stasis and to hsemorrhage, and this
suffices to disturb nutrition, and so to cause fatty metamor-
phosis.^^^

It is easy to find other concrete instances of fatty degeneration
in association with impaired supply of nutritive material. The
fatty degenerations of various tissues occurring in old age, in
disused organs, and in tissues and ofgans where increase in bulk
has been disproportionate to vascular supply, are currently cited
as examples of it.

While this is the case, however, it is by no means generally
recognised that fatty degeneration of tissue is an accompani-
ment of the general deprivation of nutritive supply incident on
starvation. On the contrary, we find such a distinguished
authority as Bauer writing as follows : — During a period of
thirty days of starvation the body loses the greater part of its
organic albumen (Organeiweiss) without the cells becoming
incapable of performing their functions; they continue to act
and are capable of full recovery on the addition of material, and
death results when the actions produced by the decompositions
are no longer sufficient to permit of vital phenomena. There is

* This article forms a portion of a Report regarding the pathological
changes observed by the author in persons who had died during the famine
in the Madras Presidency in 1877. The complete paper appears as an
Appendix to the recently published ‘ Fourteenth Annual Report of the
Sanitary Commissioner with the Government of India.’ — Editor ‘ Quart.
Journ. Micro. Sc.’]

2 ‘ Manual of Pathological Histology,’ by Dr. Eduard Riudfleisch
(English translation), vol. i, p. 29.

STARVATION OF VEGETABLE AND ANIMAL TISSUES.

51

lure a gradual emaciation or atrophy of the organic albumen,
but no breaking down of the cells into detritus.’’^^

Although practical experience in regard to the effects of
insufficient nourishment on the population of famine- stricken
areas, as well as many phenomena in the organs of those dying
of diseases incident on starvation, rendered the accuracy of such
conclusions very questionable, it yet appeared necessary to in-
vestigate the subject experimentally, more especially as the inter-
pretation of the phenomena presented by the tissues of subjects
of famine dying of special diseases, such as diarrhoea and dysen-
tery, is beset with certain fallacies dependent partly on the
presence of the diseased conditions themselves as distinct from
non-complicated starvation, and partly on the remedial treatment
which may have been employed. The results of experimental
inquiry into the subject occupy the earlier sections of this report
as an introduction to the details regarding the phenomena
observed in the human subject.

In regard to all the higher forms of life, whether animal or
vegetable, it is now generally recognised that they are composed
of living material, and of various structures developed from and
by this material, but that these structures, whilst retaining
more or less distinct structural evidence of their genetic relation
to living material, cannot be regarded as themselves alive. Such
are the walls of cell* cavities or of tubular or filamentous struc-
tures, &c., in animal and vegetable tissues. These are, no
doubt, in many cases dependent for their integrity on the presence
of the living material connected wdth them, but in others they
may persist for various periods practically unchanged after the
disappearance of the parent material, and in any case they have
entirely ceased to exhibit any of the essential characteristics of
living matter. They are never developed except under the
influence of living material, but the latter can be developed
independent of them, and can even persist and thrive for inde-
finite periods entirely apart from any of them. Changes occurring
in such structures as the result of insufficient supply of nutritive
material can only be induced mediately and through the direct
influence of such deficiency on the living matter. The primary
object of the experiments in the present instance was, therefore,
the determination of the effects of insufficient nutritive supply on
the living material, or, in other w'ords, on the protoplasmic
constituents of the tissues.

Professor Huxley long ago defined protoplasm as ^^the
physical basis of life,^^ and insisted on its general uniformify in
character in w'hatever group of living beings it may be studied,

^ “ Dcr Stoffumsatz bei der Pliospborvergiftung,'*’ Von Dr. Jos. Bauer,

‘ Zeitschrift fiir Biologic,’ Bd. vii, S. 63 — 85.

52

D. D. CUNNINGHAM.

at the same time affirming that it thus becomes evident that
all living powers are cognate, and that all living forms are funda-
mentally of one character.^^^ It cannot be supposed that any
difference of opinion should exist on this point, and, such being
the case, it must be evident to all that the effects produced on
vegetable protoplasm by insufficient nutrition are of a nature
analogous to those occurring under similar conditions in the pro-
toplasm of animal tissues, and that the study of the former may
afford much information in regard to the latter. Taking this
into consideration, the reason that a series of experiments should
have first been carried out on the influence of starvation on vege-
table protoplasm becomes manifest.

In selecting particular vegetable organisms as the subject of
experiment, the primary desiderata appeared to be — 1st, to
employ such as were capable of easy and accurate observation
under the various conditions to which they were to be subjected ;
and, 2nd, to employ such as were familiar to the observer, so as
to secure the best opportunities of fairly estimating the actual
influence of various alterations in these conditions. Two species
of mucorine fungi were accordingly selected. Both of these had
been special objects of study for some years, and both are readily
susceptible of artificial culture under conditions extremely
favorable to accurate observation. An additional recommenda-
tion of these organisms as subjects of experiment lay in the fact
that, due to their fungal nature, the processes of nutrition in
them s-how a more accurate correspondence to those in most
animals than they do in vegetable tissues containing chlorophyll.

In regard to animal tissues the great point appeared to be to
secure such as could be submitted to observation during the life
of the organism to which they belonged, or in any cases under
circumstances involving the smallest possible amount of manipu-
lative interference. In order to meet these requirements, the
animals selected as subjects of experiment were tadpoles. Brom
the extreme transparency of the tissues in certain parts of these
organisms, and the ease with which large numbers of them may
be procured and kept under observation, they afford exceptional
facilities for the conduct of experiments ; and although it may
not be permissible to deduce exact conclusions on all points from
the phenomena observed in cold-blooded animals regarding those
which will present themselves in higher forms under similar
conditions, yet in so far as the general effects of defective supply
of nutritive material on the living elements of the tissues are
concerned, there is good reason to suppose that the phenomena
in both cases must be much alike.

> ‘ On the Physical Basis of Life,’ Lay Sermons, Addresses, and Reviews,
by T. H. Huxley, LL.D., P.R.S., pp. 120 — 146.

starvation of vegetable and animal tissues. 53

In regard to both animals and vegetables the procedure em-
ployed in carrying out the experiments was similar, consisting
in retaining them for various periods in fluids in which nutritive
material was present in various amounts, or almost entirely
absent, the latter condition being secured by employing distilled
water as a medium. With these preliminary observations a
detailed account of the results of experiment may be entered
upon.

I. — Experiments on the Effects of Deficient Sn,pply of Nutritive
Material on Vegetahle Tissues.

The plants subjected to experiment were, as before mentioned,
both members of the mucorine order of fungi ; one of them
being the sole known representative of the genus Choanephora,
the other the well-known Pilobolus crystallinus. Both of these
fungi are of constant occurrence in the neighbourhood of Cal-
cutta, and both are readily susceptible of cultivation in media so
liquid as to allow of the easy removal of germinating conidia
and spores, or portions of mycelium and fructification, with a
minimum of disturbance or injury. This is naturally a matter
of moment when it becomes necessary to submit such structures
to minute examination, or to transfer them to new media. In
so far as Choanephora, however, is concerned, the nutritive fluid
itself, a strong decoction of the corollse of Hibiscus, is such as
to allow of continuous observation of all the stages of growth
in the fungal structures while still in it, and of ready dilution
with, or entire substitution by, distilled water without any dis-
turbance of the mycelium.

The conidia of Choanephora when introduced into the nutri-
tive fluid germinate almost immediately, and under ordinary
circumstances produce an abundant crop of conidial fructification
within twenty-four hours. Immature conidia are full of a
coarsely granular oily protoplasm, but they clear up greatly as
they ripen, and ultimately the oily matter is only represented by
aggregations of yellowish granules, one of which is generally
situated toward either extremity of each oval conidium. When
the conidia are subjected to conditions permitting of their ger-
mination, the first change which occurs in them is a gradual
diffusion of the fat granules, and the establishment of streaming
movement in the protoplasm. Ultimately the granules dissolve
and almost entirely disappear, and the conidium, now full of
shining, active protoplasm, emits a germinal tube. The latter
grows and ramifies rapidly if the surrounding fluid contains
sufficient nutritive material, and within the course of a few hours
gives rise to an abundant mycelium. Where the fluid is deficient

54

D. D. CUNNINGHAM.

in nutritive material, the mycelial development undergoes various
modifications, and where nutritive material is absent, as in dis-
tilled water, many conidia refuse to germinate ; and, in those
cases where germination occurs, the germinal tubes are abortive,
their growth being arrested so soon as the store material of the
parent conidia is exhausted. The mere addition of moisture is
sufficient to induce germination in many fungal spores and
conidia, but where nutritive material is wanting, or where the
means providing for its assimilation are absent, the amount of
growth which takes place is determined by the amount of ma-
terial stored up in the reproductive body. The phenomena
observed in regard to the conidia of Choanephora afford an
example of the result of absence of nutritive material, whilst
those recorded regarding the germination of lichen-spores apart
from algal elements appear to be referable to absence of assimi-
lative capacity.

The germinal tubes and actively-growing young mycelial fila-
ments of Choanephora are full of a shining protoplasm, in which
sap vacuoles are present in varying numbers. "When examined
under high powers, the protoplasm presents a dimly clouded
aspect, and a small number of brilliant, sharply-defined granules
are seen scattered through it at wide intervals {vide fig. 1, a).

Fig. 1.— Germinal tubes of Choanephora. x 1000. A. Healthy
tube. B. Starved tube.

In such tubes it is almost impossible to detect any movement
in the protoplasm, as, owing to the fulness of the tube and the
general absence of distinct granules, there is nothing to serve as
an index of movement save the slow changes in the form and
distribution of the vacuoles. In older portions of mycelium,
however, an abundance of granular matter accumulates, and
here active streaming of the protoplasm is very distinct in many
cases. Granular accumulation increases rapidly preparatory to
the development of the fertile filaments, and these and the young

STARVATION OF VEGETABLE AND ANIMAL TISSUES.

55

conidia arising from them are densely packed with coarsely
granular matter.

This granular matter may be determined to be of an oily
nature by the use of appropriate reagents^ and its appearance
in the fructifying mycelia and filaments of fungi is a phenomenon
of frequent occurrence. It is usually ascribed to accumulation
of oily matter derived from the nutritive medium, and to this,
no doubt, it may in many cases be in great part due, but that it
is entirely and invariably to be accounted for in this way is ren-
dered very questionable by the phenomena which will presently
be recorded as occurring in the protoplasm of starved mycelia.
An excessive supply of nutritive material is certainly not favor-
able to the development of fructification, for where a portion of
mycelium is retained in a concentrated and frequently renewed
nutritive medium, the appearance of fertile filaments is in-
definitely delayed, and the mycelium continues to grow luxuriantly,
while a crop of reproductive bodies may generally be readily
secured from it by diluting the medium or ceasing to renew it,
and thereby allowing its nutritive properties to become ex-
hausted.

Various other remarkable effects are produced by variations in
the amount of nutritive supply, the special form assumed by the
fructification in any cultivation being to a great extent directly
determined by the quantity and quality of the nutritive medium.
These will be referred to again in the descriptions of the effects
of starvation of the mycelium, and a detailed account of them
would occupy more space than can be devoted to it here. The
general points to be borne in mind are — 1st, that during the
evolution of the young plants from the reproductive cells there
is an involution and disappearance of oily matter ; 2nd, that in
young mycelia, in which vegetative growth is at a maximum
and nutrition most abundant, the protoplasm contains a minimum
of differentiated oily matter ; 3rd, that in older mycelia, and
especially in the fructifying filaments and young reproductive
bodies, a great accumulation of oily matter occurs ; 4th, that
excessive nutrition appears to be antagonistic to the development
of reproductive bodies.

Keeping these facts in view, the phenomena attending diminu-
tion, or deprivation of nutritive material, may now be described.
The experiments regarding these may be referred to the follow-
ing subdivisions : — 1st, cultivation of conidia in distilled water;
2nd, cultivation of germinal filaments in distilled water; 3rd,
cultivation of more or less developed mycelia in distilled water.

1st. Cultivations of conidia in distilled water. — In many
cases, as previously mentioned, the conidia refused to germinate
at all, while their contents became gradually resolved into a

56

D. D. CUNNINGHAM.

mere loose aggregation of granular matter. In other instances,
however, germination occurred. Evolution in such cases was
limited to the development of short germinal tubes, and no
formation of mycelium ever ensued. The germinal tubes having
been formed, two final results occurred. These consisted either
in a granular disintegration of the protoplasm, or in the forma-
tion of chlamydosporic reproductive bodies. In the one case the
contents of the tubes ultimately consisted of a collection of
swarming oil granules ; in the other they became condensed into
shining, isolated masses of oval or fusiform outline, consisting of
dense aggregations of protoplasmic material, invested by a
limiting layer, and in other respects resembling similar repro-
ductive bodies occurring in other mucorine fungi.

2nd. Cvltivations of germinal tubes. — In this case the conidia
were introduced into nutritive fluid, and, at the close of some
hours, when germination had freely taken place, the fluid was
washed out and replaced by distilled water. The healthy germinal
tubes (vide fig. I, a) contain an abundance of dimly clouded
protoplasm, with numerous vacuoles and a few distinct granules.
The primary effect of the substitution of water for the nutritive
fluid is a great increase in vacuolation, the protoplasm becoming
contracted as absorption progresses, and ultimately in the
majority of tubes being reduced to an irregular peripheral layer,
lining the walls of the tubes, and to a series of connecting
processes extending across the cavity and forming irregular
anastomoses with the lining layer. Whilst these changes occur
in the distribution of the protoplasm, ^another series of alterations
affect its composition. The number of granules present in it,
which originally is very small, rapidly increases, and this increase
combines with the altered arrangement in distribution to render
the streaming motion of the protoplasm very evident as the
granules are hurried along the peripheral layer and reticular
processes. The granules continue to increase in size and number,
and assume a more or less distinctly oily aspect, and after the
lapse of a few hours are very abundant and conspicuous. After
twelve hours of starvation all protoplasmic activity has ceased in
most of the tubes, and the contents are now represented by a
mere network of filaments studded with bright granules (fig. 1, b),
but in some instances movement continues with great vigour,
and is rendered particularly clear by the abundance and size of
the granules. Einally, however, movement ceases in these tubes
also, and they come to present the same appearances as the
others. The peripheral layer of protoplasm appears to become
gradually detached from the walls of the tube, so that the con-
tents become reduced to a mere tangled aggregation of filaments
and granules, Eventually the connecting material disappears.

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 57

and the granules swarm free in the tube cavity, or unite into
large oil globules, which may be dissolved out in ether, and aro
redeposited as globules and fat crystals on the evaporation of
the reagent.

3rd. Cultivations of mycelium in distilled water, — The pro-
cedure followed in these experiments was similar to that in the
case of those with the germinal tubes; the only diflPerence lying
in the fact that here the fungal elements were allowed to remain
in the nutritive medium for a longer period, and until a consider-
able development of mycelium had occurred. As a rule, whilst
the nutritive fluid was replaced by water in three or four hours
from the period at which the conidia were sown in the experi-
ments on the germinal tubes, the substitution was not carried out
here until after the lapse of eight or ten hours. The primary
effects of the treatment in the present series were identical with
those described as occurring in the former one, but while in
that series the process ran rapidly through various stages of fatty
transformation and disintegration without the occurrence of any
further development, one result of depriving a formed mycelium
of nutritive material is almost invariably the development of
fructification within the course of the next twelve hours.

The form of fructification characterising such cultivations is
sporangial. As a result of previous experience, this form had
been determined in the case of Choanephora to be developed
under circumstances of defective nutrition dependent on various
causes — on disproportion between the numbers of conidia sown
in a nutritive fluid and its amount, on employing a fluid wdiich
was originally weak or had become exhausted by previous culti-
vations, &c. ; but no other procedure appears to afford such
certain production of a crop of sporangia. When the mycelial
filaments have not been entirely emptied of their contents during
this development, the condition of these is identical with that
occurring in the case of germinal tubes which have been exposed
to starvation for a similar period, and the development is pro-
bably to be ascribed to a utilisation by one portion of the proto-
plasm of materials derived from the disintegration of another
portion.

In all three series of experiments the results were similar in
showing that fatty change and ultimate disintegration of the
protoplasm are direct results of insufficient nutrition. The
development of reproductive bodies coincidently with these
processes is very interesting, but by no means so anomalous as
it might at first sight appear. There are many facts showing
that the antagonism between individual growflh and reproduction
is not to be ascribed solely to expenditure in one direction
counterbalancing that in another, lloot-pruning and poor soil

58

D. D. CUNNINGHAM.

will often force plants into flower and fruit, which, with a liberal
supply of nutritive material, have persisted in mere continuous
vegetable growth. This, however, is a subject hardly calling for
detailed discussion here, but it is of importance to note that the
coincidence of the formation of reproductive bodies with an oily
condition of the protoplasm is of very frequent occurrence in
fungi, and that the present experiments demonstrate that the
presence of oil in such cases cannot be taken as necessarily indi-
cative of excessive accumulation of nutritive material.

The other plants in which the effects of insufficient nutrition
were studied — Filobolus crystallinm — is in various respects not
such a good subject for experiment. It cannot be satisfactorily
cultivated in a medium allowing of continuous study without
disturbance of the growing tissues ; it is impossible without great
interference with and injury to the growing tissues to free them
completely from extraneous matters from which they may derive
nutrition; and the healthy growing protoplasm in almost any
case contains a considerable amount of differentiated oily matter.
Allowing for these drawbacks, the experiments yielded results
essentially similar to those in regard to Choanephora.

The materials for cultivation were furnished by the spores of
the plant developed under natural conditions. The spores are
at any time readily attainable in Calcutta, as portions of fresh
cowdung kept for a day or two in a moist chamber hardly ever
fail to produce an abundant crop of the fungus. Tresh spore
masses, secured shortly after their discharge from the summits of
the parent filaments, were introduced into cowdung, whichhad been
diluted with water and subsequently boiled to secure the destruc-
tion of extraneous fungal elements. By this means a clean crop
of Pilobolus was secured, developed in a basis sufficiently fluid
to render the removal of germinating spores or mycelial filaments
an easy matter. In removing mycelia it was found impossible
to free the filaments entirely from adherent portions of the basis
without injuring them so greatly as to complicate the results of
experiment, but the amount of nutritive material was reduced to
a minimum by the adoption of the following method. A mass
of the cowdung containing mycelium was carefully removed, and
introduced into a watch-glass containing distilled water. Gentle
agitation served to wash out much of the cowdung ; the water
was then poured off and replaced by a fresh supply, renewed
agitation applied, and this process being repeated several times,
the mycelium was at last obtained relatively free from adherent
particles. It was then allowed to remain in distilled water, and
the phenomena presented by it, at various intervals from the com-
mencement of the experiment, carefully studied. As in the case
of the other species, the main effect following the deprivation of

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 59

nutritive material seemed to lie in an oily transformation in the
protoplasm, and the ultimate reduction of the contents of the
filaments to mere accumulations of swarming fat granules. The
nature of these granules can generally be readily determined by
the action of appropriate reagents, and the quantities of yellow
oil which can be extracted is very remarkable.

The spores of Pilobolus, as a rule, entirely refuse to germinate
in distilled water, and the contents merely pass on into granular
degeneration.

In these experiments on vegetable tissues the primary effect of
starvation seems to consist in a gradual analysis of the complex
amalgam constituting the protoplasm, and of a precipitation of
oily matter from it. This is followed by transformation affect-
ing the albuminoid constituents, and ultimately leading to their
more or less complete conversion into fat. Vegetable proto-
plasm, according to most recent observations, consists of a
combination of albuminous substances with water and small
quantities of incombustible material. In most cases it also con-
tains, as may be concluded on physiological grounds, consider-
able quantities of other organic compounds, belonging probably
to the series of carbo-hydrates and fats. These admixtures are
distributed through its mass in an invisible form.^^^

The rapidity with which oil granules appear in many instances
renders it probable that those first produced are the result of a
separation of pre-existent oil. That the oily material subse-
quently produced is a new formation is demonstrated by the
entire disappearance of all other elements from the interior of
the cells or filaments.

The contrast between the phenomena attending evolution and
involution is strikingly exhibited in these experiments. In ger-
minating spores we find a solution and disappearance of differ-
entiated oily matter ; in starved filaments, a precipitation of oily
matter from a previously homogeneous protoplasm.

II. — Uxperime7its on the Effects of Deficient Supply of Nutritive
Material on Animal Tissues,

The animals employed in the experiments on this point were
the larvae of the common toad of this part of India {Biifo ?nelano-
sticlus) and of Rana tigrina^ the so-called Bullfrog of Anglo-
Indians. The procedure followed consisted in retaining the
larvae in water containing varying amounts of nutritive material,
or, as far as possible, entirely devoid of it; freshly distilled water

* ‘Text Book of Botany, Morphological and Physiological,’ by Julius
Sachs. Translated and annotated by A. W. Bennett aud \V. T. Thistleton
Dyer. Oxford, 1875, p. 37.

D. D. CUNNINGHAM.

60

being employed in the latter case^ as in the experiments on
starvation of fungal tissues. A very large number of careful
observations were‘made on the toad-larvae ; those on the larvae of
the bullfrog were few in number^ and only undertaken with a
view to compare the general results with those of the former
series. A detailed account will therefore be given regarding the
phenomena observed in the toad-larvae, with brief notices of
those occurring in the others.

Before describing the effects produced by abnormal conditions
of nutrition it is necessary that some account of the healthy
tissues and organs should be given. A complete description of
the entire anatomy of the larvae would occupy much space, and
seems unnecessary for a correct apprehension of the most impor-
tant effects of starvation ; anatomical details are therefore given
in regard only to the transparent portions of the tails and to
the intestinal canal. The state of the tissues in these was made
the object of special study in all the experiments, the tissues of
the fin affording exceptionally favorable opportunities for obser-
vations conducted during the life of the animals, and those of
the intestinal canal being readily studied without the employ-
ment of means calculated to introduce manipulative fallacies.

The lateral portions of the tail in the larvae of Bnfo melano-
stictus may be described in general terms as consisting of a sheath
of epidermis containing a network of ramified connective-tissue
elements, together with vascular and nervous structures, and a
number of free cells or bioplasts characterised by their irregular
ramified outline and amoeboid changes of form. A few cells
containing pigment are also present in some cases, but their
number and distribution are very uncertain.

The epidermis consists of two layers of epithelial cells. In
the outer of these layers the constituent cells are flattened, so as
to form polygonal plates. The edges of these plates are straight,
or at utmost only show slight shallow sinuosities (vide fig. 2, a).

The cells vary greatly in size and form in various portions of
the fin, but are apparently everywhere closely adapted to one
another, forming a continuous, uninterrupted sheath over the
deeper structures. Each cell contains a large oval nucleus, and
this, together with the portion of the cell immediately surround-
ing it, appears to project on the under surface of the plate.
Due to this circumstance, whilst the external surface of the
entire layer of epithelium is flat and smooth, the under surface is
uneven and covered with prominences. These prominences come
into close contact with the upper surface of the inner layer of
epidermis, while a series of shallow spaces is left elsewhere cor-
responding with the thin portions of the plates.

The inner layer of epidermis differs in many respects fj-om that

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 6l

just described. The cells of which it is mainly composed
resemble those of the outer layer in their polygonal outlines,

Fig. 2. — Epidermal cells, x 1000. A. Cells of superficial layer.

B. Cells of deep layer.

and in the size and form of their nuclei. They are not, how-
ever, nearly so flat and thin, but present a tabular in place of a
scale-like form, each cell being of some depth, and more or less
convex interiorly. The margins of the cells are deeply and
closely sinuate, the prominences and depressions of neighbour-
ing cells being closely adapted to one another [vide fig. b) .
There is not the same uniformity throughout the whole extent of
this layer of the epidermis as in the case of the outer layer.
Intercalated among the cells which have just been described are
cell spaces, which stand out conspicuously, due to their dark

62

D. D. CUNNINGHAM.

■granular contents, and, as a rule, considerably exceed the sur-

JFig. 3. — Deep layer of epidermis, x 180. a, a. Granule cells,
b. Empty cell-space.

The granules with which they are filled vary somewhat in size
in different cases, in some being very coarse, and in others fine
and molecular ; when in mass they appear of a dark yellow or
brown colour. In addition to these cells or cell spaces, for in
many instances there seems to be no evidence of any distinct
cell membrane, blank spaces or hiatuses in the continuity of the
layer of cells present themselves here and there, which in size
and form so exactly resemble the cells containing the granular
matter as at once to suggest the idea that they owe their origin
to the discharge of the contents from such structures.

The nature of the contents and the phenomena of development
of these granule cells was carefully investigated, and I believe
that there can be no doubt that they are merely epithelial cells
which have undergone fatty change and which are ultimately
removed to make room for new structures. At first sight they
recall the Schleimzellen originally observed by Leydig in the
integument of fish and amphibia, and subsequently described by^
various other authors, and specially by Eranz Schultze and
Langerhans.^ Such cells are generally regarded as unicellular
glands, which, according to some observers, discharge their
secretion on the surface ; according to others they never reach
the surface, but by means of their thick mucoid secretion assist
in effecting the desquamation of the older layers of epidermis.
The granule cells in the present case resemble the Schleim-
zellen ” in being modified epithelial cells, but here the resem-
blance ceases. They differ entirely from the Schleimzellen

^ “ Ueber die Haut einiger Siisswasserfische.” Von Dr. Franz Leydig.

‘ Zeitsclirift fiir wissenschaftlicbe Zoologie/ Bd. iii, S. 2, 1851. “Ueber
Organe eincs seclisten sinnes.” Von Dr. Franz Leydig. ‘ Verhandlungen
der Kaiserliclien Leopoldino-Carolinischen deutschen Akademie der Natur-
forschcr,’ Bd. xxxiv, Dresden, 1868. “ Epitliel iind Driisen-Zellen.” Von

Franz Eilliard Schultze. ‘Archiv fiir Mikroskopisclie Anatomic,’ Bd. iii,
S. 137, 1807. “Ueber die Haut der Larve von Salamandra maculosa.”
Von J)r. Paul Langcrhans. ‘ Archiv fiir Mikroskopisclie Anatomic,’ Bd, ix

rounding cells in size {vide fig. 3).

a

S. 745.

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 63

in the nature of their contents ; they cannot be regarded as
unicellular glands unless all cells normally undergoing degenera-
tion and destruction are to be regarded as such^ and they
certainly do not serve to facilitate desquamation,, as they may be
developed and ultimately disappear in excessive numbers without
any tendency to desquamation.

The nature of the granular matter is easily determined by
means of alcohol and ether, the granules being completely and
rapidy dissolved under the influence of these reagents. The
first effect of the solvents is to cause active swarming of the
granules, which either melt together into larger granules and
globules previous to dissolving, or merely vanish, of a sudden,
without the occurrence of such a process. The swarming
granules occasionally become dispersed ere dissolving and spread
for some distance in the interspaces between the two layers of
epidermis ; but I have never seen them discharged externally,
even where the parent cells have been so much distended as to
cause elevations of the surface. After the dissolution of the
granules open spaces are left in the deeper layer of epidermis
precisely similar to those occurring normally in it, save in cases
where the cells have only undergone partial transfcrmation ere
the application of the reagents, when traces of an atrophied
cell-wail and nucleus are more or less distinctly recognisable in
the spaces.

It is easy to trace the various stages of transition by which
the normal epidermal cells are transformed into mature granule
cells. The earliest symptom of impending transformation con-
sists in an increased refractiveness and density of the cell
contents. They acquire a yellowish tint, granules begin to
appear, and, continuing to accumulate, from a dense yellow or
brown mass, which ultimately occupies the entire cell cavity
{;pide fig. I, a and b).

Fig. 4. — Stages in the development of granule eells. x 1000, A Young
eell, B. Fully developed cell.

64

D. D. CUNNINGHAM.

During this accumulation the cell generally increases con-
siderably in size^ so as to fill a space twice or three times as great as
that which it originally occupied, and the nucleus is at the same
time pressed to one or other side of the cell. After the highest
stage of development has been attained, a certain amount of
condensation usually occurs, and the granular mass shrinks
and lies more or less free in the space which it formerly occupied.
The nucleus and cell-wall finally disappear, and a mere mass of
fat granules remains behind, lying in a large space in the epidermal
tissue. Such appears to be the normal course of development
of these structures, but in some cases more or less pigment is
deposited along with the fat, and persists after the solution of
the latter.

The only other bodies which these cells in some degree resemble
are the granule cells discovered by Kolliker in the epidermis of
the lamprey. They differ from them, however, in their simple
structure, having no processes connected with them which can be
regarded either as gland ducts (Kolliker) or sensory filaments
(T. Schultze).

In the normal state of the healthy larvae, the granule cells are
developed in varying, but never in excessive numbers. In so
far as a careful study of their development and subsequent
history affords any information on their functions they appear to
subserve important ends in the processes of growth of the tissue
in which they occur, as well as in the involution of those
portions of the latter which are peculiar to the larval condition,
and disappear on the assumption of the adult form. As pre-
viously mentioned, the granule cells during the course of their
development increase considerably in size, forcing apart the
surrounding cells of the epidermis, and, on their dissolution,
leaving spaces much larger than those occupied by the unaltered
elements of the tissue. Under normal conditions these spaces
are now filled up from below by the intercalation of new struc-
tures. These may in some cases be derived from the system of
sub-epidermal free amoeboid cells; but, as a rule, they are, I
believe, derived from the connective-tissue system. I have
frequently been able to trace a direct connection between fusi-
form nuclear bodies lying in the blank spaces and the common
connective-tissue system of the interior of the fin; and in
cases where this cannot be done, the appearances presented
by the new elements in spaces which are about to be filled
up cannot fail to render such an explanation of their origin
probable. The peripheral ramifications of the connective
tissue appear to be everywhere closely connected with the inner
surface of the epidermis, a system of connective-tissue corpuscles
lying scattered over the latter, while many processes appear to

STARVATION OF VEGETABLE AND ANIMAL TISSUES,

65

be directly affixed to the epidermal cells themselves. On the
formation of a blank space or hiatus in the epidermis^ one or
more connective-tissue corpuscles, according to the size of the
area to be filled up, are drawn into it. Tiie connective-tissue
elements at first form an open reticular tissue, which, as the
nuclei enlarge and the processes diminish, is gradually converted
into a coherent patch of new cells intercalated among the older
epidermal elements, and occupying a space originally formed by
the degeneration and removal of one cell. If this interpretation
of the phenomena be correct, one main function of the granule
cells under ordinary circumstances is mechanically to enlarge the
area of the inner layer of the epidermis.

Observations did not decide how the outer epidermal layer
accommodates itself to the increased area of the inner one. The
occurrence of any general desquamation was never observed, and
there were no appearances of the addition of new elements from
beneath. Trom the extreme variations in the form and size of
the cells of the outer layer it seems not improbable that a good
deal of the accommodation is accomplished by means of mere
extension and flattening of the constituent cells.

It is evident that the granule cells may also play an impor
tant part during the resorption and disappearance of the fin, for
if developed in large numbers without equivalent addition of new
elements from the deeper tissues, they must tend to the ultimate
removal of the deep layer of epidermis. This, however, is a
subject not calling for discussion on the present occasion. The
important points to recollect here are, that when developed in
due proportion, the granule cells provide for the growth of the
tissue in which they occur, and that in such proportion their
development is strictly physiological for the organism, although
pathological for the individual ceils.

The remaining structures in the tail of the larvae call for little
detailed notice, as their characters are familiar to all physio-
logical observers. The branched amoeboid cells beneath the
epidermis are very inconspicuous during the life of the animal,
and may readily escape notice unless carefully looked for {vide
fig- 5, a).

The connective-tissue elements present no special peculiarities.
The vascular loops are full, and the current of blood is strong
and rapid. Many of the blood-corpuscles are characterised by
containing two or three minute shining oil granules.

This brief account of the normal anatomy of the tissues may
serve to render an account of the ])henomena associated with
dq)rivation of nutritive material intelligible. The following
description is taken from the notes recorded during the examina-
tion of a tadpole which had been for a fortnight in distilled

VOL. XX. NEW SER. E

66

D. D. CUNNINGHAM.

water : — '' The larva is very feeble and tends to turn over on its
back when at rest. The thin portions of the tail to the naked

Tig. 5.— AmOBboid cells* A & B x 1000, C X 180. A. Healthy cell.

B. Portion of a starved cell. C. Starved cells*

eye appear of a whitish colour in place of presenting the trans-
parent aspect normal to them> and when examined with a simple
lens are found to be covered with innumerable white points.
Under higher powers these points are resolved into granule cells in
various stages of development. Large numbers of fully-matured
granule cells are everywhere present, and the number of partially
developed ones is very great. Some of the latter are merely begin-
ning to be distinguished from the surrounding epidermal cells by
their shining yellowish appearance, while others present various
degrees of granular accumulation. The tissue of the deep
layer of epidermis appears, as it were, opened out, the consti-
tuent cells not being in such close contact with one another as
they normally are. Empty granule-cell spaces abound, and here
and there large areas occur in which the deeper epidermal tissue'
is either entirely absent, or replaced by loose reticular tissue con-
sisting of nuclear elements connected together by slender pro-
cesses {^ide fig. 6).

'' Granule cells sometimes occur, as it were, suspended in the
midst of such reticular areas. The outer layer of epidermis
appears little affected. There is no opening out of its texture,
no hiatuses occur in it, the cells are everywhere closely in contact
with one another, and the only abnormal appearances are pre-
sented by the nuclei, which in many cases are broken up into a
collection of large shining granules. The branched sub-epider-
mal cells are greatly altered. They have lost their normal soft
dim refraction and have become very conspicuous, appearing to

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 67

be distended with material of a yellow colour and containing

Fig. 6. — Reticular tissue replacing the deep layer of epidermis in a case of
starvation, x 180.

numerous refractive granules (fig. 5, b & c). The connective-
tissue Corpuscles appear empty and shrunken, and together with
their connecting filaments are in some cases studded with bright
granules. The amount of blood in the vessels is very small, the
current is very feeble, and the corpuscles appear to escape from
the vessels with abnormal readiness. Many vascular loops
remain entirely empty for considerable periods at a time, and in
others isolated corpuscles slowly follow one another at wide
intervals. Almost all the corpuscles are studded with large
granules and globules of oil, and many of them are quite bleached
and colourless. Some in process of distintegration can only be
recognised by means of the characteristic arrangement of the oil
globules which they contain, and free oil globules afford evidence
of the complete destruction of others. Alcohol and ether pro-
duced their ordinary effects on the granule cells, and prolonged
application of the same reagents showed that the yellow matter
and granules with which the sub-epidermal cells were filled was
also of an oily nature. The granules first vanished, and the
contents of the cells then gradually escaped as large yellow
globules, which gradually dissolved and disappeared. The coarse
granules in the nuclei of the outer layer of epidermis seemed to
be little, if at all, alOfected by the reagents.

Numerous careful examinations of specimens in all stages of
starvation afforded results substantially in accordance with those
just recorded, variations in degree of change being naturally
present in accordance with the varying periods of starvation.
The phenomenon which presented itself most conspicuously
and frequently at first was an increase in the number of the
granule cells. This increase cannot be ascribed to mere accumu-
lation of the cells formed in normal proportion, but failing to
undergo complete transformation, for — 1st, the increase occurs

68

D, D. CUNNINGHAM.

‘without any evidence of any equivalent increase in persistence
of the cells, blank spaces being present in increased rather than
diminished quantity ; 2nd, the increase in number of granule
cells is often extremely marked wdien there has been a very con-
siderable destruction of the epidermal tissue over wide areas. In
proportion to the period of starvation there is a steady increase
in the number of blanks in the deeper layer of epidermis and of
areas occupied by reticular tissue. This tissue, which in areas
formed early in the course of starvation, is comparatively close
and approaches new epithelial tissue more or less nearly in its
characters, alters in nature in the later stages. The nuclear
bodies are then small and far apart, and the connecting processes
proportionately lengthened. Ultimately, in the last stages of
inanition, the tissue in many places fails to be developed at all,
and large blank spaces are left, in which the internal tissues are
only covered by the outer epidermal layer.

The increased formation of granule cells is soon and constantly
accompanied by changes in the blood supply. The amount of
blood becomes visibly diminished, and the blood current flows
with diminished rapidity. The oil granules in the corpuscles
increase in size and numbers, so as to present themselves very
conspicuously even to low-power observations. The destruction
of the corpuscles, however, goes on very gradually until a late
stage of starvation, at w^hich it appears to assume a more active
course. The colouring matter escapes from the corpuscles, and
the latter ultimately break up, setting their contained oil globules
free in the serum {vide^^. 7).

Fig. 7. State of the blood-corpuscles in advanced starvation, x lOOO,

Somewhat less constant than either of the above- described
phenomena, but yet of very frequent occurrence, is a fatty
change in the ramified sub-epidermal, amoeboid cells. They
increase greatly in size, and appear to become distended with
yellow matter and numerous granules. These changes are so
conspicuous in many cases as to render the cells one of the
inost striking features in the starved tissues. The contents of
the cells are dissolved out by ether, leaving mere shrunken
skeletons behind.

No such conspicuous changes affect the connective-tissue
elements even in very advanced stages of starvation. The nuclei

STARVATION OF .VEGETABLE AND ANIMAL TISSUES. 69

lose their plump, shining aspect, and numerous fat granules
frequently present themselves, both in them and in the con-
necting processes, but no great accumulation of fatty matter or
destructive change seems to occur.

The principal effect of defective nutritive supply on the tissues
in these experiments, as in those in regard to vegetable tissues,
seemed to lie in a fatty change. This change or degeneration
specially affected epithelial elements containing a large proportion
of living material, and the blood cells. The destructive changes
in the blood cells must be regarded as a direct effect of defective
nutrition, but it is not easy to determine how far the changes in
the epithelial tissues are to be ascribed to this and how far to
defective oxidation and consequent accumulation dependent on
the changes in the blood. The persistence of the outer epidermal
layer is very remarkable, and is probably to be ascribed to the
extent to which it is composed of formed, more or less cuticu-
larised material.

The normal anatomy of the intestinal canal and the changes
occurring in it during starvation remain now to be described.
The intestinal canal of the larvae is of such simple structure and
such transparent texture as greatly to facilitate the determination
of these. All the structures can be readily subjected to obser-
vation without disturbing their normal relations, or interfering
with the intestine in any way beyond removing it from the body
of the animal. In normal healthy specimens the spiral coil of
the intestine appears conspicuous externally, shining through the
transparent abdominal walls. It consists of three coats, the
external of which is muscular, the internal epithelial, and the
intermediate one composed of reticulate adenoid tissue

fig. 8).

Fig. 8. — Normal structure of the coats of the intestine, x 180.

The muscular coat consists of the two layers of fibres, those in
the outer layer running parallel to the length of the canal, those in
the inner transversely arranged. The epithelial coat is composed
of large cylindrical cells. These call for no special note, as they
closely resemble those occurring in the intestines of other animals.
Their contents consist of soft, clouded protoplasm with a few
scattered granules, and their only peculiarity lies in the fact that

70

D. D. CUMNINGHAM.

their inner free extremities do not present such a distinctly
defined border as that apparent in many intestinal epithelia. The
reticular layer between the other coats is formed of a network of
fine filaments^ the meshes of which are occupied by free nuclear
bodies. It is very thin, and under normal circumstances is
very inconspicuous, lying, as it were, compressed between the
other two coats. When the latter, however, are separated from
one another by pressure, or by means of reagents w^hich cause
the inner one to contract, the meshes of the reticular tissue are
stretched and open out, showing that in many places several
strata of reticula are present. In some places, on the other
hand, there appears to be only a single layer, the bands of tissue
seeming to pass direct from the inner surface of the muscular
coat to the outer surface of the epithelial one. Fig. 8 repre-
sents a specimen in which several strata of reticuli and nuclei
are present. The nuclei are plump and shining, and after treat-
ment with alcohol appear softly molecular.

The appearances presented by the intestinal canal of a larva
in the last stage of starvation were the following : — The intes-
tine cannot be detected from the exterior of the abdomen, and
on its removal from the body it is seen to be very slender and
short. The canal is quite empty throughout almost its entire
length, but here and there an isolated brownish-yellow granular
mass is present. The epithelial coat is everywhere entirely
wanting. The reticula of the adenoid coat appear not to have
been in any way affected, but the nuclear elements present most
abnormal appearances. In place of being quite colourless and
smooth, they are rendered extremely conspicuous by being of
various shades of yellow and brown, due to the presence of
accumulations of granular matter within them. In some places
comparatively few nuclei are present, and the reticula are empty
or contain collections of free granules. The changes in the
adenoid tissue are most marked in the upper portion of the
intestinal canal, and in the large intestine a large number of
unaltered nuclei are still present. On treating the specimen
w’ith alcohol and ether, active swarming of the granules in the
reticular meshes and in the nuclei occurred ; and on continuing
the application of ether, many of the nuclei and nuclear masses
were entirely broken up, all the yellow matter being dissolved
and disappearing, while an abundance of deep brown pigment
granules remained behind. Many of the pigment granules lay
loose in the reticular spaces, but others were contained within
nuclei, or formed dense aggregations corresponding apparently
with nuclei which they had replaced. The yellow granular
matter was also dissolved out from the masses of intestinal con-

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 71

tents, and the oil thus extracted was subsequently precipitated
in the form of abundant oil globules and crystals."’^

The phenomena presented in this case were not peculiar to it,
but may be taken as fairly typical of the intestinal canal in
advanced stages of starvation. They show the ultimate effects
of starvation to consist in general atrophy of the intestine, entire
removal of the epithelial coat, and extensive degeneration and
destruction of the nuclei of the adenoid tissue. The degree of
general atrophy may be estimated from the fact that, while in
the healthy tadpoles which were examined the average length of
the intestine from the liver to the commencement of the large
intestine was T36 inch, and its average breadth 0*018, in the
starved specimens the corresponding averages were only 0*57
and 0*011 inch.

The absence of epithelium appears to be invariable in cases of
advanced starvation, and in them the destructive change has
generally been so complete as to leave no evidences capable of
explaining by what processes it has been effected. Careful
observations on specimens in earlier stages of starvation, how-
ever, clearly determined that they essentially consist in a fatty
transformation of the contents of the cells, followed by gradual
atrophy and disappearance of the tissue which they compose.
Even after very brief periods of deficient nutrition changes
begin to manifest themselves in the cells. The processes of
change do not, however, affect the entire extent of the epithelial
surface simultaneously. They first declare themselves over an
area extending from a point at some distance from the upper
extremity of the small intestine to another towards its lower
extremity. Erom this area the changes spread gradually both
upwards and downwards ; but, until the epithelium has entirely
disappeared, differences in the degree of change can be distinctly
traced in various regions, and the area primarily affected is often
entirely denuded at a time when the terminal portions of the
small intestine and the whole of the large one show a consider-
able amount of recognisable epithelium.

The first sign of change is shown by the cells acquiring a
yellow tint towards their free extremities. This tint soon
becomes very marked, and forms a distinct coloured band along
the margin of the epithelial coat when viewed sectionally. The
colouring is due to the accumulation of a thick granular matter
within the cells, which is first deposited immediately around the
nucleus, and gradually spreads thence throughout the cell. It
consists of fat, which can be dissolved out by treatment with
alcohol and ether. With regard to the precise nature of the
])rocess by means of which the altered cells are finally removed,
1 am unable to give any definite opinion. During the advance

72

D. D. CUNNINGHAM.

of the transformation over an area of epithelium the thickness of
the layer of cells seems gradually to diminish until, just before
the complete disappearance of the coat, it is represented by a
mere narrow granular, yellow band on the surface of the reticular
tissue ; but whether the individual cells are gradually atrophied
and consumed away, or whether the phenomena are due to the
process first affecting the largest, fully developed cells, and, sub-
sequent to their destruction, invading younger less developed
structures, I was unable to determine. Whatever the precise
nature of the process may be its ultimate effect is undoubtedly
an entire destruction of the epithelium {vide fig. 9).

Fig. 9. — Coats of the intestine in advanced starvation, showing complete
absence of epithelium, x 180.

The debris of the destroyed tissue appears in great part to
enter the cavity of the gut, and passes on to form an important
constituent in the evacuations which continue to be passed in
perceptible amount up to the later stages of starvation. Careful
examinations were, on several occasions, made of these excreta.
They consisted^ in great part, of amorphous particles, probably
derived from dust which had entered the fiuid, but they always
contained considerable quantities of soluble oily matter. The
infusoria, too, which are almost invariably present in the intes-
tinal canals of the larvse, were generally full of oil globules and
granules in cases where their host had been starved.

The changes in the nuclear elements of the reticular tissue
are always extremely marked. As in the degeneration of the
epithelium, the process of change does not appear to occur
simultaneously over the entire extent of the intestine, and as it
does not go on to complete destruction and removal of the tissue,
differences in the degree of affection in different portions of the
intestinal surface are evident, even in the last stages of starvation.
In this coat the changes seem to occur first, and to attain their
highest development in the upper portion of the canal, while in
other parts, in proportion to their distance from this portion, they
apj)ear later and are less complete. The earliest symptom of
change here also is a change in the colour of the affected struc-
tures. The nuclei, in place of presenting their normal soft,
colourless appearance {vide fig. 10), assume a yellow colour,
and become distended with a thick yellow material. This
material in some cases is distinctly granular, but in others it
appears rather as though it were a thick fluid. As the change

STARVATION OF VEGETABLE AND ANIMAL TISSUES. 73

coDtinues, the nucleus appears to become converted into a mere
aggregation of this substance; and_, judging from the presence of

Fig. 10. — Healthy nuclei of the adenoid tissue, x 1000.

free granules and masses of it in the meshes of the tissue, the
process seems in many cases to end by the breaking up of the
mass. The material, like that in the altered epithelium, can be
readily determined to be of an oily nature.

In the earlier stages of starvation such material alone is
deposited in the nuclei ; but in the later stages of the process it
is apparently invariably associated with more or less pigmentary
deposit. The amount of pigment varies greatly in different
cases, in some being comparatively scanty, in others so exces-
sive as to form the most conspicuous feature in the tissue, and
greatly to obscure the fatty deposit with which it is associated
{vide fig. 11).

Fig. 11. — Appearances of the nuclei of the adenoid tissue in a case of
advanced starvation, x 1000.

Its distribution over the intestinal surface is similar to that of
the fatty deposit, the upper portion of the small intestine being
the most constant site of its extensive occurrence. In some
cases the nuclei appear to be reduced to mere masses of pig-

74

D. D. CUNNINGHAM.

inent granules, and numerous free granules lie loose in the
meshes of the tissue.

The extensive deposit of pigment is probably due to the great
destruction of blood-corpuscles accompanying starvation. The
time of its appearance coincides with that period of starvation in
which distinct evidences of considerable destruction in the blood
present themselves ; and observation seemed even to indicate a
certain amount of correspondence between the degree of blood-
destruction and of pigmentation in individual cases.

Starvation also produced most marked efPects on the liver.
The size of the organ is greatly diminished ; it assumes a dark
brown colour in place of the pinkish-yellow tint normal to it ;
and its substance appears granular. When examined micro-
scopically it is found to be bloodless, and its substance reduced
to a mere amorphous mass of dark yellow and brown oil
granules, with a sprinkling of bright red concretions scattered
through it in some cases. The gall-bladder is invariably full,
and very frequently is greatly distended. The contained fluid, in
place of being almost colourless, as it ought to be, is green. The
depth of the colour varies considerably, but in some cases is so
intense as to cause the gall-bladder to appear as a deep emerald-
green body, w'hich shines prominently through the abdominal
walls. The gall-bladder, in any case, usually contains some
fatty concretionary matter ; but in starved cases this is present
in excessive quantity, and is frequently of a deep green colour.

In the case of the intestine, as in that of the tail of the larvae,
the ultimate effect of starvation consisted in a destruction of
tissue associated with fatty change and subsequent disintegration
of the component elements. The principal difference between
the two cases was one of degree rather than of kind, the amount
of destruction in the intestinal tissue being much greater than
that in the tail. In both cases the destructive changes w^ere
specially manifest in active epithelial elements, that is, in
structures in which the living protoplasm bore a high proportion
to the amount of formed material. The occurrence of fatty
change in the free sub-epidermal cells flnds a parallel in that
presenting itself in the nuclear elements of the adenoid tissue of
the intestine, and in both cases the connective -tissue structures
w'ere little affected and remained persistent after the occurrence of
extensive destruction in other parts. The phenomena in both
cases, and specially those occurring in the intestinal canal, show
conclusively that during the course of starvation the changes are
not limited to mere atrophy, but that in some tissues extensive
destruction of the component elements occurs, rendering the
latter incapable of recovery on the subsequent addition of nutri-
tive supply. So long as any living material remains within the

starvation of vegetable and animal tissues. 75

affected elements the possibility of recovery remains ; but when
they have absolutely broken up, or where they are converted
into mere aggregations of dead matter, it is clear that restoration
of the tissue of which they form a constituent can only be
effected by the formation of new elements.

It is also clear that the amount of transformation of tissue
taking place in the organism is not of necessity directly indicated
by the mere chemical determination of the amount and nature of
the materials leaviug the body in the various excretions, for the
products of transformation need not necessarily pass off, but
may accumulate in large quantities, as in the case of the pig-
mentary and fatty deposits occurring in the intestinal tissue as
recorded above.

The experiments which were conducted with larvm of Rana
tigrina showed that in these also the destructive effects of
starvation were specially centred on the intestinal canal, and that
the changes in the tissues of the mucous membrane were essen-
tially similar to those just described.

In so far as the present experiments afford any information on
the subject the changes effected by starvation in animal and
vegetable tissues are fundamentally the same, and the variations
presented by the phenomena attending the process in individual
cases are mainly dependent on variations in the amount and
character of the formed material present. Where this is small
the effects of destruction and removal of the protoplasmic con-
stituents are rendered rapidly • conspicuous, but where it is
present in large amount and is of a resistant nature the occur-
rence of change is to a great extent masked, and large portions
of tissue which, in so far as active function is concerned, have
in great measure ceased to serve any purpose in the organism,
may, to outward appearance, remain almost unaffected.

III. — The Conclusions drawn from a comparison of the data
acquired from the Autopsies of persons who have died through
want with those obtained by Experiment,

After a detailed description of the post-mortem lesions ob-
served by the author in the relief camps of the famine districts
of Madras, which went to show that the general result of the
entire series of observations was that the disease conditions
under investigation were specially characterised by extreme
general anaemia and destructive processes affecting the mucous
membrane of the intestinal canal, the report is brought to a
conclusion as follows: — We have now to consider what the
phenomena observed in the post-mortem examinations of cases

76

D. D. CUNNINGHAM.

of famine- diarrhoea and dysentery indicate in regard to the
es'sential nature and origin of the disease processes proximately
causing death. The prominent symptoms observed during life
are those of diarrhoea and dysentery ; we must attempt to ascer-
tain whether and how far the diseases merit the titles of famine-
diarrhoea^^ and famine-dysentery.^^ Any forms of disease
occurring during periods of famine may, in a sense, be termed
famine-diseases ; conditions of scarcity and distress must, no
doubt, more or less favour the prevalence of all diseases.
Cholera and smallpox, for example, probably find in the famine-
stricken a most favorable field, but there is no evidence to prove
that famine alone can produce either the one or the other. In
other words, these and other diseases are not directly due to
famine, but they may and probably do become much more pre-
valent in consequence of famine than they otherwise would have
been. But is there evidence to show that the so-called ^^famine-
diarrhoea and ‘‘famine-dysentery” possess characters, either
in their symptoms or in their pathological appearances, sufficient
to distinguish them as the results of insufficient nutrition ?
Does the evidence regarding them justify us in believing that
special forms of intestinal disease prevail during periods of
scarcity, essentially dependent on the effects of insufficient
nutritive supply ? It appears to me that it does. Subjects of
famine are of course exposed to the ordinary exciting causes of
diarrhoea and dysentery, and may probably be specially suscep-
tible to their influence ; but the pathological phenomena charac-
terising many cases of such disease in them appear to indicate
very distinctly that the predisposing cause is starvation, and that
the symptoms are fundamentally due to destructive changes in
the mucous membrane of the digestive canal induced by imper-
fect nutrition.

On comparing the results of the autopsies in cases of these
diseases with those obtained by experiment, it appears clear that
the changes effected by defective nutritive supply in the human
subject are closely analogous to those occurring in the amphibian
larvae. In both cases a fatty change and subsequent disappear-
ance of tissue elements occur ; and in both this change is
specially pronounced in the tissues of the intestinal apparatus.
The phenomena observed in the human subject show as distinctly
as those in the amphibian larvae that the effects of famine on the
tissues are to produce an actual destruction of tissue, and not
a mere atrophic diminution of bulk, as affirmed by Bauer. The
great diminution in the mass of the blood, which forms such a
conspicuous phenomenon in starvation, would of itself seem to
indicate that destruction of tissue elements occurs, unless it be
demonstrated that the anaemia is merely relative and solely due

STARVATION OF VEGETABLE AND ANIMAL TISSUES.

77

to diminution of the fluid constituents of the blood ; but the
phenomena occurring in the intestinal canal appear to be quite
decisive. ^Yithout the experimental data it might have remained
a more open question whether the conditions in the human
subject were not rather secondary phenomena^ dependent on the
fatal diarrhcea and dysentery^ or due to some occult cause
inducing the latter ; but when taken along with the phenomena
of uncomplicated starvation in the larvee they appear unequivo-
cally traceable to the influence of defective nutrition. The
clinical experience obtained in the conduct of relief camps also
affords most conclusive evidence on the question. Had all the
tissues retained their integrity and capability of performing their
proper functions, the excessive mortality and the general futility
of the most careful dietetic treatment among cases of advanced
starvation would remain utterly inexplicable.

When, however, we recognise the existence of destructive
change in the tissues of the intestinal canal, these problems
admit of ready solution, and all occasions to call in arbitrary
explanations of the prevalence of fatal diarrhoea and dysentery
among the subjects of starvation ceases. It has been usual to
ascribe these symptoms to the influence of noxious materials in
the blood derived from abnormal transformations of tissue, and
such materials may, no doubt, be present, but the local effects
produced on the intestinal tissues appear to be quite sufficient
to account for. the symptoms. It has also been indicated, as a
curious effect of starvation, that it should produce an incapacity
for the assimilation of food when supplied.^ With the present
anatomical data it would rather be a matter of wonder if it did
not. With the degeneration and destruction of the epithelial
and glandular elements of the mucous membrane of the digestive
system, the digestion and assimilation of nutritive materials
supplied in the food must necessarily be impaired or destroyed,
according to the degree of morbid change. The food-elements,
not being submitted to their normal transformations, become
mere foreign bodies liable to undergo decomposition, and well
adapted to cause irritation, especially on surfaces which have
been more or less denuded of their normal protective coverings,
as is the case with the intestinal mucous membrane after the
destruction of its epithelium.

That symptoms of intestinal irritation should set in under such
circumstances is only what might be expected, aud that these
symptoms should have been especially liable to occur in people
shortly after admission into relief camps is readily ex})licable.
AVliile they were outside and actually suffering from extreme
privation, the primary destruction of tissue was no doubt
* Carpenter; op. cit., p. 112.

78

D. D. CUNNINGHAM.

advancing, but the amount of nutritive material ingested was
correspondingly reduced, and was so small as to be within the
control of the remaining digestive tissues. On admission into
camp a larger amount of food was supplied ; the digestive and
absorptive apparatus which had formerly sufficed was now rela-
tively greatly reduced, and the surplus food-elements became
mere sources of irritation. It has long been recognised that
great caution is necessary in regard to food-supply after even
comparatively slight starvation. Where changes in the tissues
have only been slight, careful dietetic treatment, so as to avoid
irritation, may suffice to tide over the danger, and ultimately
effect a cure. Where, however, extensive destruction of the
tissue-elements of the mucous membrane has taken place, it is
clear that no dietetic treatment, however carefully carried out,
can be expected to effect recovery ; dietetic treatment may save
a weakened mucous membrane ; it cannot make a new one.

All the phenomena observed in the present series of investiga-
tions point to the absolute necessity of great caution in regard
to dietetic experiments, dietetic systems of punishment, &c.
They show that it is not safe to push such procedures in the
belief that, so long as no evident active evil results present them-
selves, we can at any time pull up and restore things to their
normal state. The fact that in so many cases the fatal diarrhoea
and dysentery first manifested itself in people after their
admission into the relief camps is very significant. The people
in such cases were admitted into camp, showing, no doubt,
symptoms of extreme general mal-nutrition, but suffering from
no active symptoms of disease. The mischief had, however,
been irrevocably accomplished, and it only require a change — a
favorable change too — in conditions to cause it to manifest
itself.

The insidious character of the mischief has a most important
bearing on the practical question of the management of famines.
Due to it relief camps may, to a great extent, be rendered useless
by the people failing to have recourse to them until it is too late.
They, too, are likely to be deluded by the idea that, where no
active symptoms have appeared, no permanent damage has been
done, and that they may safely delay until their distress has
counter-balanced their natural inertness and dislike of disturbing
their ordinary habits.

Calcutta ; October^ 1878.

ON THE DEVELOPMENT OP THE SPERMATOZOA.

79

On the Development of the Spermatozoa. Part I. Lumbncus,
By J. E. Bloompield, B*A. Oxon. With Plate YI.

The investigation into the structure of the testis and the de-
velopment of the spermatozoa, of which the present chapter on
Lumbricus is an instalment^ was undertaken at the suggestion
and with the kind supervision of Prof. Lankester, and has been
carried out partly in the laboratory of Exeter College, Oxford,
partly in the new zootomical laboratories at University College,
London.

The true testes of the earth-worm were first described by
Hering (' Zeitschr. fiir wiss. Zool.,^ vol. viii), who did not, how-
ever, figure them in position. They occupy much the same
position in the tenth and eleventh rings of the body as do the
ovaries in the thirteenth ; that is to say, they are placed in
pairs in those rings well forward, so as to be in relation to the
anterior septal wall and near the neural median line. Each
testis is a pure white, translucent body, of irregularly quadran-
gular form, rarely more than -pVth of an inch in diameter,
much flattened, and attached by one side to the coelomic epithe-

80

J. E. BLOOMFIELD.

liura, of whicli it appears to be a local modification. The
position of the testes is seen in the diagram woodcut. Each
testis consists of a mass of spherical cells (Plate YI, fig. 1),
those at its free edges tending readily to separate and fall into
the coelomic space or body cavity. These readily separated cells
have, moreover, often advanced in development beyond those
which lie more deeply in the mass, and have attained the con-
dition of spermospheres or sperm-polyplasts described below.
I have not been able to determine the line of demarcation
between coelomic epithelial cells of the ordinary kind and those
which build up the substance of the testis. There does not
appear to be a complete investment of the testis by normal
coelomic epithelium as there undoubtedly is of the ovary.

The large lobular, yellowish- white masses which are frequently
taken for the true testes of Lumbricus are, as Bering pointed
out, really seminal reservoirs or vesicles^ into which the incom-
pletely developed cells thrown off by the true testes are received,
and where they undergo further development. I am indebted
to Mr. A. G. Bourne, assistant in the Zootomical Laboratory of
University College, London, for the demonstration of the truth
of Hering^s views on this subject. Mr. Bourne showed that
the true testes are usually overlooked, owing to the fact that one
naturally selects a well-grown specimen of an earth-worm with
fully developed cingulum for dissection. In such specimens the
periodic development of the seminal Vesicles or reservoirs has
proceeded so far that they form the two well-known anterior
bilobed and posterior unilobed pairs of so-called testicular sacs,
the middle portions of which stretch across the middle line
coincidently with the septum between rings 10 and 11, and
with that between rings 11 and 12. In this state the true
testes are completely hidden from view, and being at this time
completely enveloped by the enlarged seminal reservoirs cannot
be demonstrated. Mr. Bourne, however, found in a series of
earth-worms dissected for the purpose of tracing the development
of the seminal reservoirs, when these bodies are in an incom-
plete or in the periodic undistended condition, that it is quite
easy to exhibit the four testes, as represented diagrammatically
in the woodcut here given.

The SEMINAL VESICLES or RESERVOIRS are seen in immature
specimens of Lumbricus as six small light-coloured vascular
growths on the three septa 9-10, 10-11, 11-12, arranged in
three pairs. The anterior pair grow forwards so as to project
into the ninth ring, the second grow backward into the
eleventh ring, and the third pair grow backward into the
twelfth ring (see woodcut). In the tenth and the eleventh ring
are the four ciliated rosettes or expanded mouths of the seminal

ON THE DEVELOPMENT OF THE SPERMATOZOA.

81

ducts, a pair in each (see woodcut). It is suggested to me by
Professor Lankester that the seminal vesicles possibly originate
as pocket-like outgrowths of the side-w^alls of these rosettes,
the anterior pair carrying each an anterior and a posterior out-
growth, whilst the posterior pair have only each a posterior out-
growth. The developing sperm-cells shed by the testis are
collected in the ciliated folds of the corresponding rosettes, and
in all probability conveyed by their agency to the increasing
seminal reservoirs. As sexual maturity approaches the three
primitive pairs of seminal vesicles become larger and larger, and
finally the four anterior pairs meet in one mass in such a way
as to form a central body, covering in the rosettes and testes of
the tenth segment, and also encroaching upon the eleventh seg-
ment ; to the four corners of this central oblong body are
attached four lobes corresponding to the anterior and middle
pairs of the primitive seminal vesicles. A similar coalescence
of the proximal portion of the posterior pair has taken place in
the eleventh segment, with invasion of the area of the twelfth
segment ; but there are only two lobes— the backward-growing
pair of vesicles which appeared on the septum between segments
11 and 12.

The minute structure of the seminal vesicles is remarkable and
has not hitherto been described. Trom their earliest appearance
they are essentially highly vascular pouches, but they do not
exhibit at any time a simple cavity or lumen. The whole of the
interior of the organ is traversed by vascular trabeculae, which
consist of excessively delicate connective tissue and exceedingly
fine blood-vessels. The larger vessels are seen in Plate VI, fig.
4. The vascular trabeculae form a sort of sustentaculum, in
which the developing sperm-cells are packed.

To gain an idea as to the structure of this sustentaculum it is
a good plan to peel off, wash, and stain a portion of the delicate
tunic with which the seminal reservoir is covered. On comparing
this with the sustentaculum seen in sections it is found that they
agree in essential structure, and, moreover, the inner supporting
portion is continuous with and formed by the indipping of the
outer tunic. This outer coat of the reservoir consists of a mem-
brane supported by fibres irregularly disposed and varying in
shape. Many of them are spindle-like with a well-stained nucleus
at the equator, and sometimes branching at the ends (fig 2),
others are longer sinuous fibres. Among the fibres are seen a
number of nuclei which, by treatment with nitrate of silver, are
found to belong to the large flat cells whicli cover the exterior
of the reservoir, and are similar to the cells elsewhere forming
coclomic lining of the perivisceral cavity.

This tunic dips down into the reservoir and forms a Ira-

VOL. XX, NEW SER, F

82

J. E. BLOOMFIELD.

becular sustentaculum, dividing it into a number of small,
irregular compartments, in which the true sperm cells undergo the
changes requisite for the development of the spermatozoa, and
accompanying these trabeculse are seen blood-vessels and capil-
laries in section.

If a seminal reservoir be placed directly in alcohol, so that the
red fluid is coagulated, and sections made without staining, the
blood-vessels will be brought out as if they had been injected ;
and such sections will show that the blood is not confined to the
surface, but enters into the substance of the reservoir with the
trabeculse (fig. 4) ; in fact, the trabeculse are little more than
networks of blood-vessels. One large vessel runs along the
internal aspect of the reservoir, and from this main trunk are
given off the secondary vessels which ramify over and into the
organ.

Development of the Spermatozoa.

An account of the development of the spermatozoa of the
lower animals has been for some time a want. While the ovary
and ova have been the subject of many investigations, the testis
and its contents seem not to have met with the attention they
deserve; and as the earth-worm is such an easily attainable
specimen, and its testicular products so easy of examination, it is a
wonder that it has been so long neglected ; for if a portion of the
contents of a seminal reservoir are examined in salt solution, a
great many of the stages of the developing spermatozoon are
exhibited in one field.

The method of examination and preservation which I have
found to answer best for the delicate ceils is to expose them in
salt solution to the vapour of osmic acid, stain with picrocarmine
and mount in glycerine ; and for the testes themselves and the
seminal reservoirs I cut sections of the tenth, eleventh and
twelfth segments, when, if all the sections are kept, the testes and
also the reservoirs cannot fail to be found. Examination after
treatment with osmic acid and staining reagents I have found
more satisfactory than viewing them in the fresh state, as the
cells are of such a delicate nature that even salt solution has a
slight effect on them. A good method, when it was particularly
needful to be sure that the corpuscles were perfectly unaltered,
was to use the hsemolymph from the perivisceral cavity only, with
the precaution that there was no confusion of the corpuscles
peculiar to that liquid with the testicular cells.

As each cell of the testis itself is the source of many sperma-
tozoa it is needful to employ terms which shall distinguish the
diflerent stages of the sperm-cells. Professor Lankester has

ON THE DEVELOPMENT OF THE SPERMATOZOA.

83

suggested the following terminology. The term spermatospore
(parallel to oospore, denoting the ovarian egg-cell) is applied to
the constituent cells of a testis, derived from the primitive germ-
epithelium. These, by division of their nuclei, form spermato-
spheres or sperm-polyplasts. Each constituent of a sperm-
polyplast is a spermatoblast, and when the process of division is
over each spermatoblast becomes a spermatozoon.

The spermatoblasts, as a rule, stand out like buds from the cell
which generates them, hence the name sperm-buds (spermato-
blasts) applied to them. The whole of the spermatospore does not
appear to be used up in the process of division to form spermato-
blasts ; a central or eccentrically placed portion, which may or may
not be nucleated, remains passive, and serves to carry the spermato-
blasts. This body, which, as will be seen below, has a central
position in the earth-worm, is to be called the ' sperm blasto-
phor ^ or ^ blastophoral cell.'’ The terms thus defined find
their application in a variety of different animals, and it appears
probable, from my investigations on Lumbricus, Tubifex,
Hirudo, Helix, Arion, Paludina, Eana, Salamandra, and Mus,
that the primitive spermatospore always give rise to a passive
blastophor and peripheral spermatoblasts, which latter only are
directly converted into spermatozoa.

The earliest condition of the spermatospores (excluding their
embryonic phases) is seen in a teazed testis. They are spherical
corpuscles, averagiug *01 mm. in diameter. The nucleus is com-
paratively very large and possesses a well-marked nucleolus ; the
thin coat of enveloping protoplasm is granular, and often of a
prickly appearance (Plate YI, fig«. 6 — 10).

In a preparation of a young seminal reservoir the next stages
of the formation of the spermatozoa will be seen. Many of
the primitive corpuscles are seen with two nuclei, and an
intermediate constriction of the enveloping protoplasm, which
is often very scanty, but distinct when it is marked by
granules. The thorny processes of the protoplasm may still
continue (figs. 16 — 20). Next to this in a normal series

comes the corpuscle with four nuclei (figs. 21, 22), but abnor-
mal forms with three do occur (figs. 23). Growth of the
whole cell continues, so that each of the segments in this stage
may be as big or bigger than the primitive spermatospore ; the
nucleoli are large and distinct, often situated near the periphery of
the nuclei. The amount to which the protoplasm is constricted in
agreement with the nuclei varies slightly ; sometimes the segments
look as if they were cpiite separate, at other times as if they
were bound and held together in an investing matrix.

The spermatosphere or polyplast, which exhibits eight seg-
ments, is the normal follower of the (quadripartite spheroid (figs.

J. jS, BLOOMFIELD.

Si

24, 25) ; but a not uncommon form is that with six, which is
to be regarded as abnormal.

It is in this stage that there is first any indication that, as the
spermatoblasts are being formed, a slight quantity of protoplasm
is being left in the centre of the generating polyplast, which, as
development proceeds, will form a cushion on which the sperm
rods may rest. It is best seen in polyplasts which have been
subjected to pressure, when the filament- cells or spermatoblasts
will be squeezed asunder, but remain connected with the central
substance by fine strands of protopalsm (fig. 26). This central
mass is the ^ blastophor.'

The general outline of the eight-divided cell is circular, but
often oval.

The following stages up to the appearance of the complete sper-
matozoa may be made out in a preparation of a well-developed
seminal reservoir from a large worm. The youngest stage in
such a preparation is a flattened, oval, plate-like corpuscle, which,
in the fresh state, appears to be composed of finely granular
protoplasm, with very slight indications of the several spermato-
blasts of which it consists, except a curious vacuolation, which,
having a radiate direction, seems to mark the separate cells
(fig. 29) . These corpuscles are highly refractive, and have sharp,
w'ell-defined outlines. On treatment with acetic acid, or the
mode recommended above, these corpyiscles become broken up
into their component spermatoblasts. Besides these, in the fresh
state occur others, in which the component cells are more distinct,
owing to projecting pieces of protoplasm, which give them an
angular appearance.

It is not possible to determine accurately the number of sperma-
toblasts which form one of these corpuscles, as only one side is pre-
sented to view, but it probably varies from sixteen to sixty-four,
or more. Viewed in optical section laterally, the spermatoblasts
are seen to be placed on a central mass of protoplasm — the sperm-
blastophor, w^hich is flat (fig. 29), in accordance ^ with the
flattened shape of the whole corpuscle. These flattened plate-
like forms have arisen by division of the nucleus from the
primitive phase of the spermatospore, w^hich was a single nucleated
cell. Though the primitive cell was spherical, yet, after its first
division, when there are two nuclei and two cells, it is oblong
and flattened ; the quadripartite form is flattened, and so is the
one with eight divisions, or nearly so, as it is possible in these
stages to count the component cells.

The most interesting point in this stage is the division of the
whole s})ermatosphere. In the fresh state the indication of this
division is very conspicuous, and as the component spermatoblasts
are not distinguishable, it looks as if one large cell were under-

ON THE DEVELOPMENT OF THE SPERMATOZOA.

85

going primary division. The sign of this division is a cleft on
both sides of the spermatosphere, not running transversely, but
generally diagonally. The cleft on one side is, moreover, deeper
than on the other. Sperm-polyplasts of this kind are found in
varying states of division till two new polyplasts result, each
consisting of eight or more spermatoblasts, which continue to
go on growing (fig. 29).

I am doubtful as to whether all the sperm-polyplasts divide
when they reach this state, or whether some do not continue their
development without a break. I am inclined to think that this
is the case, as the final form, from whence the sperm rods arise,
has two shapes — either a spherical or an oval. The first would
result where the corpuscle had undergone division at this
stage ; the second, where there had been no multiplication of
the kind.

After this division the peripheral spermatoblasts continue to
multiply by fission in planes radial to the whole polyplast. The
whole corpuscle increases slightly in size, and the central mass
of protoplasm or blastophor begins to attain some magnitude
(figs. 30 — 36). Henceforward it can be distinguished as a

large central corpuscle.

When the spermatoblasts have reached a suitable size,
the coat of protoplasm which has been enveloping the nucleus
begins, in each, to collect in a small cap or knob-like mass at
the distal end (fig. 37), and, from its high refractive power, con-
stitutes a very conspicuous part of the spermatoblast, even in the
fresh state. From this knob proceeds a small whip-like fila-
ment of protoplasm, at first very fine and short. It soon grows
out into a lash, and constitutes the vibratile tail of the mature
spermatozoon (figs. 38 — 42).

It does not stain readily, and requires, at all times, a good
light to be seen distinctly in its early stages.

The spermatoblast at this stage may be described as pear-
shaped, with the stalk of the pear turned outwards, and consti-
tuted by the whip-like filament, the body of the pear by the
nucleus which joins the blastophoral corpuscle by another thin
process of unstained matter.

The sperm-polyplast now is spherical or oval in shape, con-
sisting of a number of elongated spermatoblasts, varying in
amount according to the size of the sphere, supported like pins
on a pincushion on the central blastophor, which has swollen up
considerably, and can be seen in optical section, or even better
where rough usage has rubbed some of the spermatoblasts off
or flattened out the whole polyplast.

When the polyplast has reached this stage all further multi-
plication of the spermatoblasts ceases, and tlieir further dc-

86

J. E. BLOOMFIELD.

velopment consists in a growth in length, both of the rod-like
and lash-like portion.

The pear-like shape of the spermatoblast gives place to an
elongated oval with a conspicuous knob-like head, from which
proceeds the lash-like filament. The elongated oval grows into a
short rod with a prominent cap. The short rod then grows into
a comparatively long one, till the length of a mature spermato-
zoon is attained, though the distinction of nucleus “cap” and lash
may be traced as long as the sperm rod is attached to the central
blastophor (figs. 43 — 48). In the mature spermatozoa the
nuclear portion is only distinguishable from the rest by its rather
greater thickness. When the sperm cells reach the stage of
mature spermatozoa they fall off from the blastophor and find
their w^ay into the ciliated vasa deferentia. The further fate of
the blastophor it is hard to decide. It has no nucleus, but
sometimes a vacuole is visible (fig. 49). Often it is pigmented,
and it is probable that it atrophies, having served its purpose of
a support to the spermatoblasts.

The blastophor or central mass of the earth-worm's sperm-
polyplast possesses a very high interest, and will be found to be
represented, I believe, in the development of the spermatozoa of
most members of the animal kingdom. I have seen it in
Hirudo, in Helix (PI. VI, fig. 73), and in Rana (fig. 74) and Sala-
mandra. In the two latter it is nuclecdedy a difference from the
earth-worm, which I endeavour to explain in the next paragraph.

General conclusions. — I have found no consistent account of
the development of the spermatozoa of the earthworm, though
isolated phases of the ^ sperm-polyplasts ’ have been from
time to time noticed and figured in text-books of histology.
Prom observations, which I am still pursuing, upon other animals,
I am inclined to think that what we find in the earthworm is
fairly typical of a large number of other organisms, including
Mollusca and Yertebrata. The general notions on the
subject appear to have made little or no advance since
Kolliker's paper, published in 1856 (^Zeitsch. fur wiss.
Zool.,' Bd. vii), which is, as might be expected from its date,
w'anting in accuracy. Kdlliker gives very unsatisfactory
drawings of the development of the spermatozoa of the bul-
lock, the pigeon, the frog, and the carp. His ^ mother-cell '
corresponds to my ^ polyplast,' but he complicates the whole
history by erroneous notions as to endogenous cell formation
and the conversion of nuclei into spirally rolled spermatozoa.
The result of my observations is that, to begin with, the nucleus
of the prmUive cell or spermatosporc in the young testis is of
unusually large size, and that the secondary nuclei to which it
givL's rise stand out around the central mass or blastophor of the

ON THE DEVELOPMENT OF THE SPERMATOZOA.

87

generating spheroid with very little protoplasm clothing them.
The nucleus undoubtedly becomes the rod-like head of the earth-
worm^ s spermatozoon^ and the filament is as undeniably formed
from non-nuclear protoplasm. The sperm-blastophor of the
earth-worm is not nucleated, and it atrophies and disappears after
it has shed its crop of spermatozoa. This must be brought into
relation with the fact that the development of the spermatoblasts
proceeds not in the testis itself, but after the spermatospores
have been shed from the testis and taken into the seminal re-
servoirs.

On the other hand, we find in the frog and salamander that what
corresponds to the blastophor is a mtcleated body (see fig. 7 4). The
blastophoral nucleus was indeed seen and figured by Kolliker in
fig. 5 V of his memoir. In the frog and salamander the nucleated
portion of the blastophor (which in these animals is not central
but lateral) remains adherent to the loall of the seminal tube or
crypt, and only a portion of the corpuscle breaks off, carrying
with it the elongated nuclei, which become heads of spermatozoa.
Thus, in the frog and salamander, a portion of the sperm-poly-
plast, the nucleated blastophor, remains every year in the period
succeeding the breeding season, and is very probably ready to
resume its activity and produce a new crop of spermatozoa after
one crop has been cleared away. These nucleated blastophors
are seen forming the lining of the testicular tubes in the frog.
On the other hand, the primitive testis cells or spermatospores of
the earth-worm pass away from the testis into another organ (the
reservoir) in order to undergo their development ; the whole mass
of the sperm-cell is. detached from its site before the blastophor
and spermatoblasts have been differentiated. Hence the central
position of the blastophor and its temporary, evanescent cha-
racter.

There is another kind of corpuscle vrhich occurs in prepara-
tions of a well-developed seminal vesicle (reservoir) of an
earth-worm, and for a long time it was a puzzle to connect
it with any part of the previously described history of a sperm-
cell, but the final conclusion to which I have arrived is
that it has no connection with it at all. What the function
of these cells in the reservoirs is I do not know ; it is
possible that they form a kind of packing material during the
growth and development of the spermatozoa, and may be con-
nected with the nutrition of the developing cells and the periodic
atrophy of the vascular framework of the reservoir.

In size these cells vary, but attain in the larger specimens a
diameter of -s-TToth inch, and may be called, even wlien pale in
tint ‘Hhe brown corpuscles.^^ They have very generally a brown

88

J. E. BLOOMFIELD.

colour due to granules. In shape they would be flat plates of
an oblong outline, were they not generally bent on themselves
and curved in various places. Often they are very much elon-
gated, the longitudinal dimensions exceeding the transverse four
or five times.

An interesting fact in regard to them is that they are com-
posed of two distinct substances, one a hyaline transparent
groundwork of a refractive index, very little differing from that of
water, in which is held an irregular network of a denser substance
with large granules, many of which, as shown by Klein for intra-
nuclear granules, are the ends of rods of the network in optical
section ; but this is not the case with all. In one part of the
cell, varying in position, but in the corpuscles of an elongated
shape, generally situated at one of the ends, is a nucleus : some-
times two. Of these nuclei there is little indication in the fresh
state except that at a certain spot the protoplasm is slightly more
dense and opaque in appearance (figs. 51 — 59). Under
treatment with osmic acid and picrocarmine the rods of the net-
work split up into granules, often not quite completely, so that a
moniliform outline is the result (figs. 60 — 65).

Salt solution causes a slight shrinking of the network.

The effect of water is instructive. When it has been added to
a preparation in salt solution, the first effect is that the network
loses its clearness, but continues to be visible for some little
time. Then the corpuscle begins to swell until it becomes a
hyaline sphere, containing granules which dance in rapid
Brownian motion, and the nucleus is exposed to view. Dilute
acetic acid has the same effect (fig. 66).

Osmic acid, when applied in solution and picrocarmine, swell up
the corpuscle in a similar way owing to their water, but when
glycerine is applied (as is usual for preserving the preparation) it
is reduced again in size.

The vapour of osmic acid causes the network to disappear,
leaving granules in its place.

The proportionate amount of tliis network varies a good deal.
In some corpuscles the meshes are large and distinct and the
rods of large dimension, but in others the meshes are small and
irregular, and often the rods very fine.

At all times the arrangement of the network is hard to de-
termine accurately, and it is only possible to give an idea of it
in a drawing by diagrammatic treatment.

Of the function of these corpuscles I can only, as I have said,
make a guess. The stages in the development of the seminal
reservoirs in which they are found are the following : in large
well-developed reservoirs they are large and numerous, but
in reservoirs taken from those large dark worms, which seem

ON THE DEVELOPMENT OF THE SPERMATOZOA.

89

to be in a state of decline after the performance of the generative
function they are particidoHy numerous in proportion to the other
corpuscles^ The actual number of them in this case may not be
increased, but owing to the absence of the other cells which have
developed into spermatozoa it seems that there are more of them.
This fact seemed to indicate that they have a connection with the
blastophoral corpuscle, but a fatal objection to this view is that the
latter possesses no nucleus, and would have to go through a long
series of changes before it would resemble one of these ^ brown
corpuscles,^ of which intermediate phases there are no indica-
tions whatever. I have repeatedly searched for such phases and
have never found them. In the very youngest stages of the re-
servoirs there is no sign of them, but in preparations where the
sperm-cell segmentation has reached the binary stage it is seen
that they are represented in a young condition, developing to-
gether with the true spermatoblasts and the fusiform cells of the
reticulum. They are in this stage much smaller than the mature
brown corpuscles,^ and the network is so dense that it is
impossible to make out its arrangement in detail. I have re-
presented them at this stage in figs. 67 — 71.

90

F. M. BALFOUR.

On the Spinal Nerves <?/■ Amphioxus. By F. M. Balfour,
M.A., E.E.S. Fellow of Trinity College, Cambridge.

In an interesting memoir devoted to the elucidation of a
series of points in the anatomy and development of the Verte-
brata, SchneideF has described what he believes to be motor
nerves in Amphioxus, which spring from the anterior side of the
spinal cord. According to Schneider these nerves have been
overlooked by all previous observers except Stieda. *

P myself attempted to show some time ago that anterior
roots were absent in Amphioxus ; and in some speculations on the
cranial nerves, I employed this peculiarity of the nervous system
of Amphioxus to support a view that Yertebrata were primitively
provided only with nerves of mixed function springing from the
posterior side of the spinal cord. Under these circumstances,
Schneider’s statement naturally attracted my attention, and I
have made some efforts to satisfy myself as to its accuracy. The
nerves, as he describes them, are very peculiar. They arise
from a number of distinct roots in the hinder third of each
segment. They form a flat bundle, of which part passes up-
wards and part downwards. When they meet the muscles they
bend backwards, and fuse with the free borders of the muscle-
plates. The fibres, which at first sight appear to form the nerve,
are, however, transversely striated, and are regarded by Schneider
as muscles ; and he holds that each muscle-plate sends a process
to the edge of the spinal cord, which there receives its innerva-
tion. A considerable body of evidence is requisite to justify a
belief in the existence of such very extraordinary and un-
paralleled motor nerves; and for my part I cannot say that
Schneider’s observations are convincing to me. I have attempted
to repeat his observations, employing the methods he describes.

In the first place, he states that by isolating the spinal cord
by boiling in acetic acid, the anterior roots may be brought into
view as numerous conical processes of the spinal cord in each
segment. I find by treating the spinal cord in this way, that
processes more or less similar, but more irregular than those
which he figures, are occasionally present ; but I cannot per-
suade myself that they are anything but parts of the sheath
of the spinal cord which is not completely dissolved by treatment
with acetic acid. By treatment with nitric acid no stick processes

1 ‘Beitrage z. Anat. u. Entwick. d. Wirbelthiere/ Berlin, 1879.

2 “ On the Spinal Nerves of Amphioxus,” ' Journ. of Anat. and Pliys. ’
vol. X, 1870.

ON THE SPINAL NERVES OP AMPHIOXUS.

91

are to he seeuy though the whole length and very finest branches
of the posterior nerves are preserved.

By treating with nitric acid and clarifying by oil of cloves, and
subsequently removing one half of the body so as to expose the
spinal cord in situ, the origin and distribution of the posterior
nerves is very clearly exhibited. But I have failed to detect
any trace of the anterior nerve-roots. Horizontal section, which
ought also to bring them clearly into view, failed to show me
anything wdiich I could interpret as such. I agree with Schneider
that a process of each muscle-plate is prolonged up to the anterior
border of the spinal cord, but I can find no trace of a connection
between it and the cord.

Schneider has represented a transverse section in which the
anterior nerves are figured. I am very familiar with an appear-
ance in section such as that represented in his figure, but I
satisfied myself when I previously studied the nerves in Amphi-
oxus, that the body supposed to be a nerve by Schneider was
nothing else than part of the intermuscular septum, and after
re-examining my sections I see no reason to alter my view.

A very satisfactory proof that the ventral nerves do not exist
would be found, if it could be established that the dorsal nerves
contained both motor and sensory fibres. So far I have not
succeeded in proving this ; I have not, however, had fresh
specimens to assist me in the investigation. Langerhans,^ whose
careful observations appear to me to have been undervalued by
Schneider, figures a branch distributed to the muscles, which
passes off from the dorsal roots. Till the inaccuracy of this
observation is demonstrated, the balance of evidence appears to
me to be opposed to Schneider’s view.

^ ‘ Archiv f. Mikros. Anatomie/ vol. xii.

92

ARMANER HANSEN.

The Bacillus of Leprosy. By G. Armaner Hansen.
With Plate YIII.

It was not ray intention to make any of ray investigations
on this subject public at present, but as not only Dr. Edlund,
to whom in the preceding year I showed preparations, and
mentioned that I considered leprosy a parasitic disease, in
his little work on ^ Leprosy ’ speaks of its precise origin as
something that he has discovered in the form of micrococci,^’
but also Dr. Neisser, of Breslau, who passed some portion of
this summer in Bergen has just published the result of his
investigations of those preparations that he made while here,
and as these results also point out that in general, the pre-
parations are filled with bacilli ” which he supposes to be
peculiar to leprosy, and as its contagium ” — I feel myself

called upon to announce what I have attained to, up to the
present time, in my researches after the same contagium,”
and, this, partly to assert my priority with reference to this
discovery, and partly in order to advance those details in
research which 1 omitted to announce on account of the
still uncertain result in my report to the Medical Society in
Christiania, 1874, concerning my investigations into the
etiology of leprosy. In this report I Iiave briefly stated that
I often, indeed generally, found, when seeking for them in the
leprous tubercles, small rod-shaped bodies in the cells of
the swelling,! whilst, on the contrary, I never found
such bodies ” in blood newly taken from leprous patients,
in which, however. Dr. Edlund declares he has seen the
microccoci ” described by him. This observation of Dr.
Edlund I must, however, after having examined several
• times, quite lately, blood taken from a leper, consider as
unreliable. I found very often, on the contrary, after pre-
serving the blood-preparations in a damp room, that in the
course of a few days there appeared articulated threads,
which, I believe, must be considered as a fungous growth,
and which never appeared in blood-preparations taken from
either healthy or syphilitic people. After having employed
myself for a lengthened period in these investigations of the
blood, I proceeded to those of the tubercle, and shall com-
municate, as follows, a few of the memoranda I made during
that time. •

Case No. 755. — Johannes Giil, vigorous nodules;
February 28th, 1873. A nodule taken from each side of
the nose with scissors, and laid in a carefully cleaned watch-
• Vide ‘ British and Foreign Medico-Chirurgical Be view,’ April, 1875.

THE BACILLUS OF LEPROSY.

93

glass. Cut through them ; no softening ; scraped th^ surface
with the edge of the knife^ and placed the parts removed on
an object-glass, and without any addition of fluid covered over
with a glass cover. There were seen almost only round-cells,
very few with granules of fat, some finely granulated, others
containing rod-shaped bodies, which are sometimes bordered
by parallel lines, and sometimes pointed at both ends. In
this latter instance they are about twice as thick in the
middle as the others. Such bodies are to be found
where fluid-spaces are formed by the pressure of the glass
cover surrounded by the dense mass of cells. In these
spaces the bodies move after the manner of bacteria.”
Other preparations prepared in the same way were examined
in a drop of distilled water with Hartnack No. 9, but with no
result. The round cells are the regressive elements of
brown colour, which I have both described and drawn in
my previous contribution to Leprosy’s Characteristics,”

^ Nord. Med. Art.,’ vol. i. No. 13; these drawings are
reproduced in “Leprous Diseases of the Eye,” by O. B. Bull
and G. A. Hansen, Christiania, 1873. In the next place
I scraped the surface of the tubercle as above, and put
the scrapings into a drop of water. In such preparations
an unusually large number of the small bodies show them-
selves, which have besides more or less lively movements.
The cells mostly swell up considerably in the water, and in
the swollen cells the rod-shaped bodies are much more easily
found; many cells show themselves completely loaded with
them; at the first glance it seems as if the cells were filled with
coarse granules, but on closer examination, on the contrary,
these apparent granules are found to be small, oblong, rod-
shaped bodies. Several preparations of each kind are then
placed on the bottom of a glass basin which is covered by
a larger one ; the bottom is covered with damp sand, and
over all is placed a glass plate (two preparations with water,
three without).

March 1st. — No examination of the preparations.

2nd. — Preparations appeared as on February 29th.

3rd. — In one of these preparations with water a mass 'of
articulated threads was to be found in one place exactly
corresponding to those which are found in the blood of
certain lepers after cultivation.

4th. — In such preparations the articulated threads are to
be found ; in one of them without water they were found at
the edge of the preparation, as well as in the spaces filled
with serum amongst tlie densely packed cells ; these were not
long threads, only two or three joints fixed to each other.

94

ARMANER HANSE.V.

and swimming partly free in the fluid. In one part of one
of the preparations there was a countless number of oscillat-
ing rods, and one penicillium.” Two of the preparations
without water did not fill the whole space under the glass
cover ; in the water which had fallen in drops in the vacant
space no trace was to be found of either bacteria or any-
thing else.

7th. — Kristian Lotuft. Vigorous tubercles which have
steadily increased in the last year. One tubercle was
removed and divided ; one half was immediately placed in
1 per cent osmic acid, and preparations were made of the
other, and the blood oozing out from the cut surface was
collected in a small glass vessel.

Preparation, No. 1. — Serum containing blood-corpuscles ;
from this preparation a particle of the tubercle was teased
out with needles, and examined under the microscope ;
numerous oscillating rods and rod-like bodies were to be
observed in some of the cells.

Preparation No. 2. — Serum containing blood-corpuscles ;
here and there a white body ; extremely few oscillating
granulated masses.

Preparation No. 3. — Serum with numerous blood-cor-
puscles, and admixed epithelium from the edge of the
cutting ; numerous bacteria. All these were kept in a moist
granulated chamber.

Preparation No. 4. — In a drop of distilled water pieces of
the picked-out tubercle ; numerous bacteria ; immovable
bacteria in the cells. The preparation was enclosed with
oil along the edges of the glass cover.

Preparation No. 5. — Serum containing blood-corpuscles,
enclosed with oil as No. 4.

Preparation No. 6, — Serum containing blood-corpuscles,
•with pieces of the picked-out tubercle also enclosed with oil,
and containing numerous bacteria.

8th. — -Nothing to remark.

9th. — No. 1. Leptothrix ; articulated threads abundant.

No. 2. Nothing.

No. 3. Nothing.

No. 4. An occasional articulated thread here and there.

No. 5. Nothing.

No. 6. Like No. 4.

10th. — No. 1. Still more chains in large groups.

Nos. 2 and 3. Nothing.

Nos. 4, 5, and 6. Like those of yesterday.

18th. — Case No, 705. A tubercle "taken from the
under lip ; softening commencing.

THE BACILLUS OF LEPROSY.

95

Preparation No, 1. — Procured after incision by pressure on
the tubercle which is tolerably juicy, contains a consider-
able number of large brown elements and numerous bacteria ;
these could also be indistinctly seen within the cells w'hich
are not of a very deep brown, and I think now and then
that I could distinctly see in these a long stripe in the
apparently granulated masses.

No. 2. Like No. 1

No. 3. Like No. 1 (with water). The cells for the most
part were swollen, and in these rod-shaped bodies in large
numbers were distinctly to be seen. The large brown
elements w'ere not much affected by water, occasionally a
little swollen, and thus it can be tolerably clearly seen that
in general a large portion of the apparent granulations are
oblong and rod-shaped.

All these preparations were kept in a moist chamber. The
addition of acetic acid does not help at all. The preparations
become more opaque by coagulation.

With acetate of potash all the oscillating rods are killed ;
they become instantly more highly refractive after it has
penetrated, shrivel, and lie exactly like corpses all over
the preparation ; the brown ones shrivel considerably, and
become exceedingly refractive, shining like >vax. Their
whole contents become as if kneaded together ; no rods to be
distinctly seen unless action be just about to commence. The
rod-like bodies in the cells are to be seen best by teasing
out a piece of the tubercle in 1 per cent. “ osmic acid.”

20th. — In many places in all the three preparations arti-
culated threads of greater or less extent are to be found ; in
some places a single thread wound a few times round itself,
in others such a confusion of threads that to follow them w^as
impossible.

Of the tubercle which, after obtaining the above-men-
tioned preparations on March 18th, was laid in 1 per
cent, osmic acid, a preparation was taken to-day, which
showed rod-shaped bodies in most of the cells. I took
a drop of the reduced acid in which the tubercle was
lying ; no oscillating bodies were to be found ; a smtill
particle of the tubercle w’as placed into this drop and teased
out, and on investigation many oscillating rods, of a length of
O'OOlo — 0 006 mm., were to be found. On repeated knocking
on the cover glass until almost the whole of the above particle
had fallen to pieces, it was found swarming with oscillating
rods, and in the broken edge of a large cell which had
been fractured by the knocking a few rods were seen pro-
truding into the fluid.

96

ARMANER HANSEN.

— Numbers of articulated threads in all the prepara-
tions like those of yesterday.

April 1st. — Olive Bjorhaug. Spots six weeks old ; strong
retrogression. A piece cut from a spot on the forearm,
divided into two pieces, one teased out in fresh water, one
in salt water ; the sudoriparous glands appear large and
easily teased out; round cells amongst the windings of the
canals, also here and there, between the bundles of connective
tissue, many cells interwoven with strongly granulated pro-
toplasma.” In both preparations, principally along the
edges of the fragments of the teased-out sweat-ducts, a
quantity of pale, round, angular and oblong molecules,
with oscillating movements, were found. It cannot be posi-
tively decided whether some of the more lengthy ones were
really bacteria ; highly refractive, round, and oblong small
bodies could be found in some few of the sweat-ducts ; kept
in a moist chamber. Besides this a preparation in osmic
acid was made of the blood oozing out from the cut surface
of the spot.

3rd. — In the blood preparation, which is almost dried up,
numerous articulated threads were found. In the other two,
along the edge of them, numerous large bacteria with lively
movements, but no articulated threads.

4th. — The blood preparation quite dried up. The two
others like those of yesterday, although the bacteria were
not in such lively movement.

7th. — The preparation with water dried up ; the prepara-
tion with salt water, connective tissue and cells unchanged.
No bacteria to be discovered ; on the contrary, numerous
highly refractive, small articulated threads were found, with
from two to five articulations, immovable.

I7th. — Christian Loluft. Obtained a preparation from a
tubercle on the cheek by puncture, principally blood, with
some greater or smaller tubercle cells ; unable to discover
any free bacteria.

1 8th. — The red blood-corpuscles a little shrivelled ; the
cells, however, about normal, some of them a little dis-
coloured.

20th. — The cells extremely pale and somewhat shrivelled,
and occasionally hydropic ; no bacteria to be seen in them ;
the large pale brown cell-contents unchanged ; nothing more
to be discovered than on the first day ; granules and doubt-
ful rods.

22nd. — Chains in many places, but the preparation not
to be relied on, on account of water having penetrated in one
place.

THE BACILLUS OP LEPROSY.

97

Anne Sahingetar, died March 10th ; post-mortem examina-
tion 11th March, 1873. Some rather shrivelled tubercles
from the face were examined ; large, dry brown bodies were
found in parts, formidably large, and easy to be seen with
the naked eye. In the microscopical preparations oscillating
rods were to be found everywhere, and after the addition of
potash-lye I was fortunate enough to see in several of the
large brown elements a sort of striation amongst the apparent
granulations ; the brown ones adhered strongly to the glass,
and when the glass cover was lifted up and removed from
one place to another, many pieces were found remaining
on it in many places ; these pieces showed themselves to be
composed of small rod-shaped bodies, which crossed each other
in every direction,

March 17th. — No articulated threads have formed them-
selves in any of those preparations which were kept in a damp
room.

21st. — Christian Loluft. Eruption over the whole
body rather severe ; two newly-formed tubercles were
punctured with a needle ; a drop of something resembling

pus ” was squeezed out at the same time ; the drop was
viscid, tough, and did not float out by pressure on the
glass cover ; only a few cells at the edges became free by the
addition of distilled water, also when the glass cover was
repeatedly lifted up and set down on the one side. It was
also evident from microscopical examination that the prepara-
tion contained a few blood-corpuscles. After the addition of
water and tlie glass cover still remaining removed, nothing
remarkable is to be found in the preparation ; if, however,
on the contrary, the glass be moved from side to side so
that a portion of the preparation is floated, a not incon-
siderable number of oscillating rods of different sizes make
their appearance in the fluid. Two similar preparations
were made, one with and one without water (in a moist
chamber) ; besides these a preparation of blood was made,
in which, on examination with Hartnack No. 11, no bacteria
could be discovered ; finally, two other preparations of the
contents of the tubercle without water.

23rd. — No articulated threads to be found in any of the
preparations ; those without the addition of water still
appear as quite fresh.

24th. — The same as above.

25th. — Also the same.

28th. — Still no fungus. The preparation without water
kept remarkably well. The cells had a perfectly fresh
appearance. The preparation containing serum was also

VOL. XX. NEW SER. O

98

ARMANER HANSEN.

perfectly dried up along the edge, so no water could pene-
trate.

S9th, 30th, 31st. — On account of other engagements the
preparations were not examined.

April 1st. — Nothing is seen in the blood preparation,
which is mostly dried up ; fungus in both the others.
The preparation without water, especially deserves atten-
tion ; here right in the middle of it, with so especially well-
preserved cells that they appear perfectly fresh, were found
in four places large masses of fungus, of finely granulated
appearance, from the edge of which there shoot out fine
articulated threads, whose several joints measured 0*0006 —
0*0007 mm.

4th. — This preparation exactly like that of the 1st. The
masses of fungus have not become larger ; a rather large
brown one, which is broken on one side, remained un-
changed in these three days. Its place was carefully noted
in order to watch if any articulated threads should grow out
of it. Along the edge of the glass cover, upon the dried up
edge of the preparation, a brown fungus has shot out a rich
network of threads with fructification, by division of the
points.

7th. — The fungus mass along the edge of and in the pre-
paration is unchanged, the edges ^et free, and alcohol and
ether added, which only, by repeatedly lifting the glass
cover, penetrate the preparation ; by this the fungus mass
was divided into lumps ; little by little, by the action of the
ether, they took the same appearance as the brown ones,
viz. the granules and rods were kneaded together into a
shining wax-like mass.

Hospital, Johannes Giil. — Two tubercles from the nose,
taken out with entire epithelium.

Preparation No. 1. — Scraped with a knife the mass of cells
in the blood-serum which lies on the surface; obtained
several doubtful bacteria with slow movements.

Preparation No. 2. — With a glass tube, freshly blown,
from an incision in the tubercle teased and sucked out small
pieces of the cell mass ; considerable number of blood cor-
puscles, also bacteria.

Preparation No. 3. — In like manner; fewer blood-cor-
puscles.

Preparation No. 4. — Ditto, with as much water ; much
more numerous bacteria, and here and there in the swollen
cells rod-shaped bodies ; many brown ones in a lump, none
floating.

Preparation No. 5. — Preparation of blood collected in a

THE BACILLUS OF LEPROSY.

99

watch-glass ; serum containing blood-corpuscles, here and
there a white blood-corpuscle ; discovered a few bacteria.
On teazing a preparation in osmic acid I find many cells,
which enclose longish rods.

10th, Nos. 1, 2, and 3. — Present no alterations, except
that where the longish rods can be seen they are motion-
less.

Preparation No. 4. — In many of the cells that have been
swollen by water are found more or less numerous granules
of various sizes, showing molecular movement ; in some of
the cells between the oscillating granules a rod with slower
movements can be observed, which appears independent of
the movements of the granules ; in other cells no granules
are to be found ; on the contrary, many rods and one or two
of these with slow serpentine movement.

Preparation No 5. — No alteration, except that the blood-
corpuscles are more contracted.

11th. — The preparations are unchanged.

12th. — The same as 11th, only that the cells have no
longer anywhere the same uninjured appearance as before ;
in many the small rods are not to be seen.

14th. — In No. 1 — 3 the cells are still well preserved.
The nuclei stand more boldly out than before, more homo-
geneous. In the dense crowd of cells is found, in many
places, what appears to be a film over the preparation ; this
film seems to consist of a finely dotted mass, and also of
small rods, which cross one another in all directions. In
No. 2 is found in the serum two small chains of
monads. In one place in No. 3 a well-defined granulated
mass, which at the edge shows itself to be a number of
chains of monads. In No. 4 only a few of the cells are
preserved in one side of the preparation ; an immense
number of bacteria right in the middle of the preparation ;
in several places a mass of immovable rods crowded to-
gether. It is difficult here to ascertain the nature of these,
for in many places are found small lumps of stearine crys-
tals, but the former are found to be much more highly
refractive and of a more irregular arrangement than the
stearine crystals.

18th, Nos. 1 — 3. — The cells have disappeared, best pre-
served in No. 3 ; the finely granulated fungus mass not par-
ticularly increased, coloured a strong brown black with
osmic acid, like the brown bodies. No more chains of
monads in Nos. 1, 2.

20th, Nos. 1, 2. — The cells always more fallen away
and kneaded together in a mass, in which the single cells

100

ARMANER HANSEN.

could not be distinguished. No 2 best preserved, still no
chains. In No. 4 are masses of bacteria ; at the edge a
large mass of zoogloea, has made its appearance, amongst
which are found a few penicillium threads, so at all events a
part of the bacteria may be supposed to have come in from
the outside, but this bacteria* containing zone has not en-
croached farther into the preparation than before.

21st. — In Nos. 1 and 2 in many places, in the middle of
the preparation, masses of zoogloea.

March 24th, 1873. — Uatel Espeland. Eruption. Punc-
tured an umbilicated tubercle that had become tender by
the eruption ; blood and the whitish contents collected.
Two preparations were made. In both gigantically large
brown elements to be found ; no bacteria. Water was added
to the one preparation. As long as this flows quietly in,
and the glass cover be not moved, no bacteria appear, whilst,
however, a large portion of the cells swell up; by moving
the glass cover up and down, and pushing it a little to one
side, a large portion of the cells break, and now bacteria
appear in the fluid.

25th. — Preparation unchanged.

27th. — In the preparation to which water has been added
numerous articulated threads are especially remarkable,
issuing from a brown body ; the connection confirmed by
the movement of the glass cover. By these means it was
ascertained that the whole mass in the lower element, to-
gether with the apparent articulated threads, remained con-
nected in all positions of the former (fig. 5).

April 18th. — Ratel Espeland. From a puncture of a
tubercle on the forehead ; pressed out contents of it ; this, so
firm that it must be teased out with needles. One prepara-
tion was made in salt solution, and one in distilled water ; in
both an immense quantity of bacteria ; the cells large. In
the preparation with distilled water bacteria are to be seen
in almost all the swollen-up cells, some wdth rather lively
movements. The situation of the bacteria within the cells
ascertained in the most evident manner, by getting the
cells to roll in the fluid ; no granules for the most part can
be found in the cells, and where such is the case, they
oscillate with far greater speed than the slow-moving
bacteria.

20th. — Many bacteria can be found in the water prepara-
tion, which are those with 3 — 4 — 8 articulations. In the
salt-water preparation the cells are not a little shrivelled ; by
the addition of water a large portion of them are very
quickly set free ; a crowd of granules stream out in

THE BACILLUS OF LEPROSY.

101

greater or smaller clumps, amongst these a number of small
rods.

22. — In both the preparations abundant masses of zoogloea.

April 10th. — From Iver Sorlide were taken three samples
of blood from the cheek, in freshly blown capillary tubes,
which were secured at one end and hung up on a string in
the room.

20th. — The contents were examined, coagulated in all
three tubes, and were difficult to get out. The blood-cor-
puscles had a healthy appearance ; the red ones rather
spherical, the white, for the most part, kneaded together in
clumps. In one place in one tube in the coagulation was
found a mass of zoogloea ; in the other two nothing could be
discovered.

From the various notes of my investigations in 1873
every one will be easily able to see that I had good
reason for supposing that bacteria appear in leprous pro-
ducts, but also that I, supported alone by these investi-
gations, could not propound a theory on this subject,
and still more decide whether these bacteria really were
the virus which, introduced into the system, produced the
disease. In order, if possible, to arrive at a decision in this
matter, I tried to inoculate rabbits with leprosy by intro-
ducing portions of the leprous growths, especially of the
tubercles, under the skin of the animals. I was not lucky
in any of these attempts, which, however, as a matter of
course, is a proof against the supposition that the above-
named bacteria are the real virus. I have not repeated this
attempt later, and only now and then, by examining the
tubercles, have become convinced that my observations with
regard to the occurrence of the oscillating rod-shaped bodies
were correct, until I, on reading this spring Dr. Koch’s
w'ork, ‘ Untersuchungen uber die -dEtiologi der Wund-
infections krankheiten,’ and by being enabled to see some of
Dr. Koch’s preparations of Anthrax bacilli^ learned to
know this author’s excellent system for the demonstration of
bacteria. It occurred to me this summer to try ether in the
investigations, in order, if possible, to prove the bacterium-
nature of the rod-formed bodies, and their presence every-
where where leprous productions are to be found. But I
have hitherto endeavoured in vain to obtain good and con-
vincing preparations, except in one case, when I was able to
obtain perfectly convincing results. Guided by the above-
recorded observations, viz. that the small rods become more
distinct by the treatment of the tubercles with osniic acid,
I placed an extirpated tubercle in this acid, and have

102

ARMANER HANSEN.

obtained from it, by colouring the sections with methyl
violet, a few preparations in which the brown cells, easily
visible, also without colouring, appeared very sharply dis-
tinguished from the surroundings by their violet colour ;
upon closer inspection they show a finely granulated and
partly striated appearance, as if they consisted of small rods.
As a reason for my small success in this work I must (after a
letter from Dr. Koch, with whom I communicated respecting
it) presume that the reason is either that the methyl violet
1 use is not as it ought to be, or that I have allowed the
colouring matter to act too short a time. These mistakes.
Dr. Koch informs me, were from the first committed also by
Dr. Neisser. Dr. Neisser’s later and more successful results
probably arise from his having been so fortunate as to have
Dr. Koch’s valuable guidance. I have, by this preparation,
obtained conformation of my earlier supposition that the
large brown bodies after all are nothing else than either
masses of zoogloea or collections of bacilli which are enclosed
in cells. By looking at fig. 4, which represent tumour-
cells treated with osmic acid, drawn from preparations made
in 1873, one is easily able to form an idea how these same
cells, by a constantly increasing number of small rods, at
last become quite overloaded, and thus obtain the appearane
of being filled with fine granules', since the single rods
cannot then be distinguished. I have already in my first
communication to “Leprosy’s Characteristics” (‘North.
Med. Archiv,’ vol. i. No. 13) stated that I was inclined to
regard those brown elements as peculiar to leprous growth,
both on account of their most striking appearance, and
because they were always to be met with in all the parts
affected with leprosy. Should the above-mentioned supposi-
tion concerning the true nature of the brown elements prove
in time to be correct, their peculiarities will, at the same
time, be demonstrated, and it will be important to ascertain
the conditions of existence of these bacilli, in order, finally,
with full reliability to remove all doubt as to the real cause
of leprosy ; and this, shall as before, be the goal of my work.

Since writing the above I have also been so fortunate as to
obtain bacilli finely coloured in a section of a tubercle hardened
in absolute alcohol, and, acting upon Dr. Koch’s advice,
stained with a stronger staining fluid. Baccilli are found in
all parts of the section, either singly, or more frequently
in groups, fully corresponding to those occurring in the cells.
I furnish a drawing of two groups taken with Zeiss’s
immersion system ^^^d eye- piece No. 4.

NOTES AND MEMORANDA.

Development of Planorbis. — Dr. Carl Rabl, of Vienna, has
published in Gegenbaur’s ^Morphol. Jahrbuch,’ vol. v, a
very carefully illustrated account of the development of the
fresh-water snail Planorbis. In essentials the history does
not differ from that of Limnseus, on which subject Dr. Rabl
some time since published a less complete and less accurate
memoir (‘ Jenaische Zeitschr.,* 1875). In the present
memoir Dr. Rabl abandons his erroneous interpretation of
the shell-gland and recognises the correctness as well as the
priority of ray descriptions of that organ in the Mollusca
generally and of the velum of Pulmonata. Dr. Rabl has,
however, failed to trace the derivation of the hinder part of
the embryonic intestine from a group of cells adhering to
the margin of the blastopore (constituting a “ pedicle of in-
vagination ” as in the Lamellibranchs Pisidium and Ano-
don) ; on the contrary, he considers this portion of the enteric
tract to be a caecal outgrowth of the invaginated endodermal
sac, which outgrowth pushes forward so as to touch the
body-wall. The endodermal sac (archenteron) is only in
contact with the blastoporal margin (according to Dr. Rabl)
at the anterior or oral end of that orifice. Satisfactory evi-
dence of this is not adduced. Drawings of a very beautiful
series of sections are given in two plates as well as numer-
ous transparent views of young embryos. The drawings of
sections are perfectly accurate, as I am able to state from
the examination of a number kindly sent to me by Dr.
Rabl. The sections are masterpieces of embryological mani-
pulative skill, but they do not tend to substantiate the view
with regard to the origin of the hinder part of the embry-
onic intestine which Dr. Rabl maintains. They are equally
favorable to the view to which I was led by observations
recorded in this Journal in 1874. Dr. Rabl takes so
strongly adverse a position in reference to the connection of
the hind-intestine with the blastopore, that he emphatically
denies the accuracy of my observations on the connec-
tion of the blastopore and anus in another Gasteropod, viz.
Paludiiia.

104

NOTES AND MEMORANDA.

He is fully aware of the fact that Professor Butschli
examined the development of Paludina (‘Zeitsch. wiss. Zool./
vol. xxix) for the express purpose of testing the accuracy of
my statement, and came to the conclusion that the matter
was as stated by me. Dr. Rabl is scarcely well advised to
impugn the correctness of an observation, thus supported,
without offering illustrative drawings^ showing what it is that
he has seen in the history of Paludina.

The formation of the primitive nephridia of the embryo
Planorbis, by perforated cells, similar to those forming
the nephridia in adult Lumbricus and Hirudo, is admir-
ably shown by RabPs sections and drawings. — E. Ray
Lankester,

Origin of Sperm and Ova from the Cell-layers of Coelentera. — •
The researches of the brothers Hertwig (‘ Organismus der
Medusen/ Jena, 1878) upon the histology of a number of
Medusae belonging to the group known as Hydromedusae led
them to the conclusion that, uniformly in Medusae of that
group, the genital products are derived from the Ectoderm.
This conclusion, corresponding with that of Kleinenberg
relative to Hydra, led the Hertwigs to consider it pro-
bable that the observations of Ed^ van Beneden, Ciamician,
and others, which attributed the ova, or, in some cases, the
sperm to Endoderm in the sporosacs of certain hydroid
polyps, were erroneous, and that the derivation of the ova
and sperm from Ectoderm might be considered as general for
all the Hydrozoa.

Claus, however, found the spermatozoa of the medusa
Chrysaora developing directly from cells of the Endoderm
over a large surface of the gastric pouches (Polypen und
Quallen der Adria, ‘ Wiener Denkschr.,^ 1878), and now it
appears, from further researches of the Hertwigs, upon
Lucernaria, Pelagia, and Charybdoea (‘ Jen. Zeitschr.,’
vol. xiii, 1879), that in the Medusae of the Lucernarian type
(Acraspedae of Gegenbaur, or better called Scyphomedusae,
as opposed to Hydromedusae), the ova as well as the sper-
matozoa uniformly take their origin from cells of the Endo-
derm. This being the case it seems quite possible that,
after all, both the observations of Ed. van Beneden and those
of Ciamician may be accurate, and that in the Hydrozoa
there is no absolute uniformity as to the origin of the gene-
rative products. This, however, is not the view taken by
the Hertwigs. They have extended their observations to
certain — a very limited selection — of the Anthozoa, and find
that in them (various Actiniae, Cerianthus, and Alcyonium)

NOTES AND MEMORANDA.

105

the ova and the spermatozoa both take their origin from cells
of the Entoderm. Accordingly, the Hertvvigs propose to
generalise their observation — a proceeding which appears to
us to be as yet without justification. They still maintain —
in spite of the observations of Ed. van Beneden and others —
that all the Hydromedusae develop their genital products
from Ectoderm. This really rests only on the observation of
a certain number of medusa forms and of Hydra, and is op-
posed to equally careful observations on Hydractinia, &c.
Further, they expand their observations on Scyphomedusae
(Acraspeda) and Anthozoa to the dimensions of a statement
that all members of both these groups uniformly develop
their genitalia from Endoderm. Very possibly both these
large statements are true, but such is by no means demon-
strated at present. However, the Hertwigs proceed further,
and on this basis propose to remodel the groups of the nema-
tophorous Ccelentera. They propose to abandon the division
into Hydrozoa and Anthozoa, and to constitute two groups,
one to contain the Hydromedusae which have ectodermal
genitalia, the other to contain the Scyphomedusae and
Anthozoa, which have endodermal genitalia, thus denying
the supposed close relationship between the two kinds of
Medusae, and making them mere homoplasts, one of the
other.

For the two new groups they propose the names Ectocarpa
and Endocarpa respectively. Very possibly such a grouping
may have to be ultimately adopted. There is no doubt that
a sharp line cannot be drawn which shall separate Hydrozoa
and Anthozoa, and the new division proposed would not be
more wanting in definition than is the present. At the
same time, should the grounds put forward by the Hertwigs
prove really solid, the old terms proposed by Wilhelm Rapp,
of Tubingen, for an exactly similar classification, will have
to be adopted, namely, ‘^Exoarii and Endoarii,”

It is strange that the Hertwigs should have overlooked
their countryman’s contribution to this subject, published as
long ago as 1829, and not so far from Jena Ueber die
Folypen in Allgemeinen und die Actinien insbesondere,^
WeimaT). Yet further do the Hertwigs err in their reference
to the history of the investigation of Coelenterate mor-
phology, for, whilst discussing the appropriateness of the
old tenns, Phanerocar])a and Cryptocarpa, for the designa-
tion of the groups which they propose to establish, they refer
those well-known terms to Edward Forbes (loc. cit., p. 624)
insteail of to the illustrious Eschscholtz.

Thp difficulty which presents itself most obviously in the

106

NOTES AND MEMORANDA.

way of the proposed new grouping of the Ccelentera nemato^
pJiora is found in the Ctenophora. These have been re-
cently shown by Haeckel to be most readily explained as a
peculiar modification of the Hydromedusa3 ; at the same time
there seems no reason to doubt that their genitalia develop
from the endodermal cells of the ctenophoral canals. If
this be so, the origin of the genitalia from this or that layer,
would prove to be — in the Ccelentera at any rate — a matter
of no genealogical import, instead of one of primary signi-
ficance.

An important character, serving to unite the Medusce acra^
spedcB with the Anthozoa, is found by the Hertwigs in the
homology of the gastral filaments of the former with the
mesenterial filaments of the latter.

Muscular fibre (mesoderm) derived from Endoderm. — For
some years the record of observed fact has pointed to the
exclusive derivation of muscular and fibrous tissues in the
Hydrozoa and Actinozoa from the ectoderm, the endoderm
being employed solely as an epithelium lining the gastro-
vascular system. Recently, however. Professor Claus, who
has in recent years by varied and most valuable researches
done so much towards placing our knowledge of the Hy-
drozoa on a secure histological basi«, showed that Halis-
temma possesses in its axis a delicate layer of ring muscula-
ture derived from Endoderm (‘ Arbeiten aus d. Zool. Institut
der Universitat,’ Wien, vol. i, 1878). This observation
has been followed by the definite observation on the part of
the Hertwigs (loc. cit.), to the efifect that in the Anthozoa
(Actinia, Cerianthus, &c.) the musculature is chiefly derived
from Endoderm, and only a small layer from Ectoderm.

If this observation should prove to be thoroughly well
founded, we have a new basis for the interpretation of the
facts now known as to the origin of muscular and connec-
tive tissues in the higher animals, the Coelomata or Tri-
ploblastica.

Recent Researches on Bacteria. — In a resume which
we recently gave (this Journal, vol. xviii, p. 455) of the
recent contributions to our knowledge of this most important
poup of organisms we omitted to make any mention of the
important researches of Dr. Alb. Fitz, of the Chemical
Institute in the University of Strasburg. These have been
published from time to time in the ^ Berichte der Deutschen
Cbemische Gesellschaft zu Berlin ’ from 1876 up to the
present time. Dr. Fitz has investigated from the chemical

NOTES AND MEMORANDA.

107

point of view the fermentations caused by various Schizo-
myceta, and also that due to Mucor racemosus. One of his
most important results (published in 1876) is that glycerin
in the presence of calcic carbonate enters into fermentation
by the agency of a Bacterium. The chief products of the
fermentation, in addition to carbonic acid and hydrogen, are
normal butyl-alcohol and normal butyric acid. As by-pro-
ducts are produced also in very small quantity ethyl-alcohol
and one of the higher fatty acids, probably capronic acid.
The Bacterium of this, particular fermentation has the form
of a Bacillus. It is on the average two micromillimeters
broad and five or six long. A smaller Bacillus occurs with it
in glycerin fermentation, to which is probably due the pro-
duction of the ethylic alcohol. Dr. Fitz found, moreover, that
the remarkable Bacterium which gives the blue green colour
to old pus, caii also excite fermentation in glycerin. He culti-
vated this organism from pus by transference to a solution
containing cBcium lactate and food-salts (ammonium chlo-
ride as a nitrogen food). In a successful cultivation he
obtained nof^SL coloured solution, but a colourless reduction-
product of the colouring matter. Only on the surface was
there a blue' tint. This was due to the access of oxygen to

the surface, for when the solution was agitated with atmo-
spheric air, it immediately assumed a deep blue colour,
like that of the solution of copper sulphate. This pus-blue
colouring niatter is soluble in water, is turned red by acids,
and by alkalies again blue. It is, therefore, similar to litmus.
The Bacteria which produce it, and multiply luxuriantly in
the solution, are small elliptic bodies, about one to one and a
half micromillimeters in length, and generally occur in
couples.

Dr. Fitz’s further researches on the fermentation of
erythrite, mannite, gelatin, and albumen, as well as those
on butyric and propionic fermentation, are of the greatest
importance and interest.

The separate tract of Dr. Robert Koch, entitled, ^ Unter-
suchungen fiber die ^tiologie der Wundinfectionskrank •
heiten ’ (Lei[)zig, 1878), is worthy of the most careful study
from all persons interested in the progress of knowledge
relative to the all-important Bacteria. In a first series of cx-
])erimei]ts we have an investigation of septicaemia produced
in mice by injection of putrid blood or infusion of meat
beneath the skin. Death in some cases resulted in from four to
eight hours, being due 7iot to the mxdiiplication of Bacteria
but to the poisonous matter (sepsin) produced by them and
injected with them. In other cases, death resulted after

]08

NOTES AND MEMOR.iNDA.

fifty hours, the Bacteria (a bacillus form, one micro-
millimeter in length) having multiplied in the blood and
crowded the vessels of all organs of the body. One-tenth
of a drop of the blood of an animal thus affected sufficed to
convey the infection, and it was handed on thus for seven-
teen generations.

A distinct disease — gangrene — was induced in some of the
mice experimented on, and this was found to be accompanied
by the appearance of micrococci (minute spherical Bacteria
occurring in chains and masses). These are, in all proba-
bility, a form-phase of the septicaemic Bacilli, which are
their necessary predecessors, but the Micrococci differ not
only in form but in pathological activity from the Bacilli
which generated them.

In rabbits by injection of putrid blood a subcutaneous
abscess could be produced without infection of the blood.
The disease could be propagated by injections of the fluid of
the abscess which contained micrococci. From putrid fluid
obtained by macerating a mouse-skin in W’ater, pyaemia was
produced in rabbits which could be propagated by injection of
y^th of a drop of the blood. Micrococci (not forming chains nor
zooglaea masses) characterised this particular kind of pyaemia.
Another septicaemic condition was obtained in rabbits by
subcutaneous injection of putrid tneat, and found to be
characterised by large micrococci (one micromillimeter in
diameter) . This was communicable ; but another local
erysipelatous condition of the tissues, obtained by injecting
mouse-dung infusion beneath the ear of the rabbit, was not
inoculable, although the tissues of the ear contained large
numbers of Bacilli, three micromillimeters in length.

These investigations are of the highest importance, as
showing firstly that a variety of inflammatory diseases may
take their origin by the access of certain Bacteria to wounds ;
secondly, as showing that the particular form of Bacterium
in each case can be distinguished. The derivation of the
micrococcus of necrosis of the mouse from the Bacillus of
septicaemia of the same animal is also a result which has
great significance.

Dr. Koch is not able to admit in the case of septicaemia
that the virulence of the poison increases with each succes-
sive inoculation. This he admits only up to the third inocu-
lation as a rule, and attributes it in accordance with the most
probable analogy to the history of various beer-ferments
studied and separated by Pasteur, to this fact, namely, that
just as with beer-yeasts so with septicaemic ferments, culti-
vation by iterative minimal inoculation in their specially

NOTES AND MEMORANDA.

109

appropriate media, serves to purify the specific ferment from
injurious interfering ferments of other species.

M. Paul Bert (‘ Comptes Rendus Soc. Biologie,’ 1879,
p. 355) has been led by some experiments on the blood of
animals infected with Anthrax {Bacillus anthracis) to con-
clude that such blood may contain two poisonous elements, the
one similar in nature to that of cow-pox and glanders, non-
organised, and precipitable by alcohol and of a highly viru-
lent character ; the other consisting of the micro-parasites
known as Bacillus anthracis. The first poison which can
be obtained free from the second by subjecting Anthrax-
blood to compressed oxygen for two days (which destroys
the Bacillus) causes death when introduced into a guinea-
pig in about twelve hours ; the second requires about thirty
hours. These results are quite in accordance with the germ-
theory of such diseases, and with what is known in general
as to the properties of organised ferments. The Bacteria or
organised ferments undoubtedly produce poisonous matter,
such as the sepsin, now recognised for some years, and above
noted by Dr. Koch. Such matter may act directly as a
poison or the poisonous effects may be retarded until the
Bacterium has multiplied and produced larger quantities of
the toxic material.

It is exceedingly probable that soluble ferments are pro-
duced in a similar way by many Bacteria associated with
fermentative changes such for instance as that of urine.

The Bacteria which normally infest the human body must
be exceedingly numerous and varied in form and properties.
At present they have not been fully recognised. Mr. Butlin
(' Proc. Royal Society,’ 1879) has made a beginning in de-
scribing those which infest the mouth, and which are the
chief constituents of the fur of the tongue. Mr. Butlin finds
Sarcina, Bacterium, and Spirochaeta in the mouth, but chiefly
micrococcus and Bacillus subtilis.

We have yet to obtain from some careful worker with
high power objectives an account of the forms of Bacteria
present in the contents of the large intestine and other
passages of the body. Further, the different protected
surfaces of the skin offer an interesting field for study in
relation to this subject. The axillae, pubes, and even the
scalp, have their special saprogenous Bacteria, whilst the
epidermis of the prepuce and of the plantar surface of the
foot is occu])ied by Bacteria-ferments, specifically connected
with the higher fatty acids.

110

NOTES AND MEMORANDA.

Publications of the Zoological Institute of the University of
Vienna. — We desire to draw the attention of our readers to
an important series of memoirs published under the direc-
tion of Professor Claus, of Vienna. A very large amount
of first-rate zoological research is now being carried on
under the direction of the Vienna professor of Zoology. He
not only holds his important chair in that city but has the
zoological station of Trieste under his direction, and both
he and his pupils have made ample use of these fine oppor-
tunities. Among the memoirs which have appeared in the
‘ Arbeiten aus d. Zool. Instit. der Univ.,’ Wien, are those of
Claus on Halistemma, on Charybodaea, and on Phronima.
These memoirs are not published in any other periodical or
society’s proceedings.

New Biological Journal.— On the first of January will
appear a new biological journal, published in Brussels.
Professor Edouard Van Beneden is one of the editors; the
journal will be chiefly devoted to original memoirs and
similar in scope to our own. We have great pleasure in
commending the new Belgian journal to the attention of
histologists and embryologists.

PROCEEDINGS OP SOCIETIES.

Dublin Miceoscopical Club.

24dTi April y 1879.

Blodgettia. — Dr. E. Perceval Wright exhibited some freshly-
mounted specimens of Blodgettia confervoides of Harvey, which
he had prepared from specimens recently collected at Key West
by Mr. Farlow,and sent to him preserved in alcohol. These very
clearly showed that strings of spores referred to by Harvey were
parts of a parasitic plant living in the cells of GladopJiora
ccespitosa. This parasite would seem to consist of a series of
delicate threadlike hyphse, with here and there enlargements.
For this Fuugoid form Dr. Wright would retain the generic
term Blodgettia, of course with a new diagnosis, and he would
venture to call the species after Dr. Bornet. In addition Dr.
Wright read a letter received that very morning from Dr.
Bornet. Alluding to the memorandum about this plant in the
Club Minutes for March 23rd, 1876, Dr. Bornet wished to
correct the statement referred to there as having been made by
him, on the authority of Mr. Farlow, to the effect that Blod-
gettia was an unicellular form of parasite. It was most obviously
nothing of the sort, but a fungus with branching hypba3.

Goscinodiscus Soly WaMich, from Sea of Javay exhibited. — Bev.
E. 0‘Meara exhibited specimens of the rare diatomaceous form
Goscinodiscus Soly Wallich. These were found on slides from
material gathered in the 8ea of .lava and mounted by Professor
T. P. Cleve, of Upsala. The form is not enumerated by Pro-
fessor Cleve among the many new and interesting species
discovered by him in gatherings from the Sea of Java.

Anthelm Turneriy Dumortier, from Kenty exhibited. — Dr. Moore
showed an example in fruit of the very rare British species
Anthelia Turneriy Dumortier — Jungermannia Turneriy Hook.,
Br. Jung This was first found near Bantry by the late Miss
Hutchins in 1811. Dr, Taylor remarks in Flor. Hib. (1836),
“ This plant has never been seen growing but by its discoverer.”
Dr. Mo(>re stated he had often searched for it when botauising
in the South of Ireland, but never succeeded in finding it. It
has, hovever, been found in France, according to Dumortier in
‘Hepat Europ.’ (1874). The specimen now shown came from

112

PROCEEDINGS OF SOCIETIES.

North Kent, and was found by H. Davies, Esq., of Brighton,
who sent it to Dr. Moore. Like its congeners, this forms a
very neat and pretty little object, though the present example
had become too dried and brittle to enable one fully to realise
its elegauce.

Cross Section of the Spine of Acanthaster echinites, Ellis et
Solander, was exhibited by Mr. Mackintosh. The genus is re-
markable for the possession of numerous madreporic bodies and
spines with double articulation, which in section show a pith
with rather larger interspaces than are seen in the peripheral
part, which is covered externally by a tolerably thick skin. There
were no indications of the solid wedges which form so charac-
teristic a feature in the spines of Sea-Urchins.

Problematic cotton-like branched fibrous structure^ exhibited.
— Mr. Archer drew attention to a problematic piece of organi-
zation, met with in an aquatic gathering, resertbling a little
“ tree ” of colourless cotton fibre, having enclosed in the stem and
branches and running longitudinally therein one or more opaque
and shining wire-like fibres, these apparently very brittle, being
often here and there seemingly broken into short lengths ; the
apices of the branches, which proceeded in a tufted manner from
the central stem, were apparently granuliferous, tie parts lower
down hyaline, and more or less flattened and twisted — the
question was, what could this thing be ?

22nd May, 1879.

Spermatic state of Volvox, showing the groups of soermatozoids
and the yellow oospores, were exhibited by Mr. Archer; no
meeting of the two elements had been noticed.

Berkleyia.—J)T. E. P. Wright read a letter which he had re-
ceived from Dr. Griinow in reference to the algal form exhibited
by Dr. Wright to the Club in May, 1876. This form, from Dr.
Harvey’s Friendly Island Collection of Algae, had been described
by Dr. Griinow in the Botany of the Novara Reise as Berkleyia
Barveyana, which fact had been overlooked by Dr. Wright in
1876. At Dr. Griinow’s request he had again carefully and in
many ways examined the type-specimens, and he now exhibited
mounted preparations thereof, some in glycerine, sone simply
burnt, and some cleared with weak acid. In none of these
was there a trace of silex to be seen, unless here and there an
odd stray frustule of a Navicula, a Diatoma, or a Suriiella. In
addition Dr. Griinow had very kindly forwarded some mounted
specimens of his species, of the nature of which there could be no
doubt, they being decidedly siliceous fruatules, in all probability
quite correctly assigned to the genus Berkleyia by Dr. Grriinow,
so that now the question stood thus : Berkleyia Rar^wyanus
Griinow, was a diatom from the Friendly Islands. lut the

DUBLIN MICROSCOPICAL CLUB.

113

Alga? quam paradoxa! of Harvey was not a synonym thereof,
and this latter, a mass of semi-confused congeries of green (phyco-
chromaceous r) cells, was still to be described.

Section of Larynx of Human Foetus, exhibited. — Mr. B. Wills
Bichardson exhibited with a Quekett’s dissecting microscope and
one of Charles Baker’s 2-inch objectives only, so that the whole of
the object might be seen, a cross section of the larynx of a nine
months humaa foetus. In the specimen the most striking
structures were (1) sections of the thyroid and the arytenoid
cartilages ; (2} the laryngeal pouches ; (8) the pharynx ; (4)
portions of some of the intrinsic muscles of the larynx ; (5) the
laryngeal columnar epithelium ; and (6) mucous glands.
The specimen was a triple staining made with carmine, picric
acid, and macder, the latter being an excellent medium for
staining cartilage cells, their nuclei and nucleoli.

The section was a striking illustration of the great assistance
that may be derived from the use of the freezing microtome in
minute anatomical research, for it is difficult to conceive that so
even a section could be made without the greatest difficulty by
any other methods with which we were familiar. Moreover,
Mr. Richardscn observed that scarcely a bit of the larynx was
wasted, as he succeeded without any trouble in cutting the whole
of it, the tenuity of all the sections being about the same as the
one he exhibited. It was mounted in Klein’s dammar solution.

Cross secticn of the spine of Toxopneustes variegatus (Larnk.)
was shown ly Mr. Mackintosh, wffio called attention to the
differences mticeable between it and the spine of T. maculatus,
A. Agass. : tlese are chiefly the lesser extent of epithelium in the
former, the nore gradual increase in size of the solid wedges and
the marked igure-of-8 shape of the peripheral wedges, those of
T. maculatush^mg oval.

Mr. Archer announced that the Boyal Microscopical Society of
London had done the Club the honour to “ nominate it under
the bye-law relating to ex officio Fellows, and the same having
been approved of by a Gieueral Meeting, its President for the
time being is entitled to the privileges of an ex off do Fellow of
the Society;” and it was resolved that the marked thanks of this
Club be mmmunicated through the Secretary of the Royal
Microscopcal Society to that body for the distinguished honour
thus confured.

19M June, 1879.

The shted meeting appointed for above evening was not held
owing tc the very sudden and much-lamented death of Dr. David
Moore, F.L.S., M.R.I.A., Director of the Botanic Gardens,
Glasnevn, for many years a highly esteemed member of the
Club.

VOL XX. Ni:w SER.

U

114

PROCEEDINGS OF SOCIETIES.

July 24nth, 1879.

Xanthidium Rohinsoniamim, Archer, shown from a new midland
Irish locality. — Mr. Crowe showed examples of Xanthidium
Hohinsonianum^ being the fourth occasion this form has as yet
been met with ; first found by Mr. Archer in co. Armagh,
then in co. Donegal, afterwards by Mr. Crowe at Glengarifie, and
now in co. Kildare, this well-marked little species has not yet
been taken out of Ireland. Although its distribution is thus
wide in this island, it must still be accounted as very rare. The
form is small, semicells trapezoid, the rounded angles crowned by
a few minute spines, central elevation of the semicells bearing a
few irregularly disposed, rather conspicuous grinules ; lower .
margins straight, isthmus slightly elongated, constriction thus
forming a parallel and straight division between the semicells.
A more detailed description will be given on another occasion.

Trichia clavata, Persoon, was exhibited by Mr. G. Pirn ; this was
found growing on moss in an orchid basket in his sto’^e. This form
comes under the genus Uemiarcyria of Rostofinski which differs
from Arcyria in having the capillitium in spirals, as in Trichia,
and from Trichia, defined by him in having the spral combined
into a dense network similar to Arcyria. This proved to be the
fully-developed form of the immature myxomycete shown at a
previous meeting of the Club. It w^as abundant on the wood and
moss for many weeks. It formed a very pretty cbject viewed
either with a low power, so as to emblace the whole, or with a
higher one, so as to exhibit the structure of the capilitium.

Section of Limestone from Ogradina, Banks of the Danube,
exhibited. — Professor Hull exhibited a thin slice of limestone
from the banks of the Danube, near Ogradina, considered to
be of Cretaceous age. The rock from which this was taken rises
in massive cliffs above the left bank of that river, where it
traverses the Eastern Carpathians. As Dr. Hull could discover
no traces of fossils, he had hoped a thin section would probably
(as in the case of our chalk) exhibit a foraminifera. structure,
but this expectation had not been fulfilled, as no organic
structures were visible under the microscope in tie sections
exhibited. The entire field consists of small particles of calcite,
which polarise with the usual opalescent colours, and fi. into each
other along irregular, but generally straight margins A high
power shows the particles of calcite to be cellular in sime cases,
and to be traversed by cleavage planes in others. Veinsof calcite
also occur. We must assume the absence of organic stucture to
be due to the entire replacement of the original rock b/ foreign
matter, but of similar composition, assuming, what is generally
admitted, that all limestone formations are of organii origin.
The replacement of the original organised structure by crystalline
calcite is evidently a subsequent process of “ metamorphsm,” of
which the specimen exhibited affords an illustration.

DUBLIN MICROSCOPICAL CLUB.

115

23rc? OctoheTi 1879.

Peculiarity in Histological Structure of Spine of JEchinothrioe
Turcarum, Peters. — Prof. H. W. Mackintosh exhibited a cross sec-
tion of the spine of HcTiinothrix Turcarum, Peters, and called atten-
tion to a peculiarity in its structure, of which this was, so far, the
only example he had met. In the normal spine the solid wedges
are coextensive with the network between them, but in this form
they were produced beyond it, reminding one somewhat of the
arrangement in the spine of Echinus lividus. The specimen from
which the spire had been taken formed part of a collection made
by Professor E. P. Wright in the Seychelles, which had been
presented by kim to the Museum of Trinity College, Dublin.

A new TInitellular Epiphytic Green Marine Alga, exhibited. —
Dr. E. P. Wfight showed specimens of a new unicellular epi-
phytic green marine alga found attached to larger forms in rock-
pools at Howth. This was a sub-globose or balloon-shaped, or
sub-quadrilateral thick-walled, form often showing empty cells still
attached by tleir lower extremity, their contents evacuated as
zoospores. Tius this would be to a great extent like Characium,
but the forms there included are stipitate. It called to mind a
form found bf Mr. Archer some few years ago in moor-pools in
Connemara aid attached to confervoids, but the cells were cubical,
thus presentiig a quadrilaterul outline and the examples attached
(sometimes ii rows) by the whole of the base in it, forming the
longest diamder.

Roothairs inside the rootsheaths in Azolla pinnata. —
Dr. McNab exhibited roots of Azolla pinnata, showing the
sheath with the numerous root-hairs developed inside the inner
sheath, and leld down by it with their apices pointing to the
apex of the loot until they are liberated by the elongation of the
root. Thes« hairs are produced in regular whorls close to the
growing-poirt, but by growth of the axis their regularity is
disturbed. Strasburger mentions that in Azolla pinnata the
sheath breiks up and gives the root a plumose appearance.
Nothing life this could be observed either in young roots or in
those whidi have lost the root cup. As Strasburger does not
describe ary root hairs on these plants, it is probable that they
have beenoverlooked.

Micrasirias furcata, exhibited. — Mr. Archer showed an example
of Micraierias furcata, one of the most rare as well as the most
beautiful of the Desmidiem. He had himself seen probably in
all not Hlf a dozen examples of this fine species, always from
Connema’a, and which appears quite distinct from M. Crux-
Melitenis, itself very rare, but cropjiing up sometimes, mainly in
CO. Wesimeath gatherings.

Tricim fallax, exhibited. — Mr. Greenwood Pirn showed Trichia
fillax, /ar. cerina, from his stove at Monkstowii. This form
(lifiers Torn T. chrysosderma in having much finer and more
delicaU spiral elaiers, which are cylindrical and taper abruptly

116

PROCEEDINGS OP SOCIETIES.

to a point, while those of T. cJirysosperma tapered gradually
from the middle to each end — the spores being quite smooth in
the former and minutely echinulate in the latter.

Zygospore of Xanthidium Hobinsonianum^ exhibited. — Mr.
Crowe, showed the zygospore of this species in continuation of
his exhibition of the parent form at last meeting ; these he had
since discovered in another gathering made on the same occasion.
They are orbicular, beset with elongate very slender spines, trifid
at apex, slightly dilated at base ; the contents of the spore
present a somew^hat olive-green or bronzy hue, rendering the
appearance somewhat marked. It is therefore interesting to
have found the zygospore of this species, which is really rare
in Ireland, though, as seen, widely diffused, but jet it does not
seem to have been met with out of this island.

Xodose Hairs. — Mr. B. Wills Bichardson exhibited two mounted
slides of human nodose hairs which occurred in tie practice of
Dr. Walter Smith. The specimens on one of tie slides were
from the head of a young lady of nineteen, and on tie other were
from the head of a boy of seven years.

Similar specimens from the girl’s head were exhibited by Dr.
Smith at the recent meeting of the Medical Associition in Cork,
and excited much attention. These hairs do not correspond
with any of the descriptions of nodose hairs pub! shed, at least
as far as a very prolonged research could ascertaii. Indeed it
may be assumed that they are unique, from the fict that Mr.
Erasmus Wilson requested Dr. Smith -to give hire some speci-
mens for the Museum of the Boyal College of Surgeons in
England ; and Dr. Lionel Beale, to whom he (Mr. Eichardson)
sent a slide of the hairs, wrote to him to say that hehad not seen
anything of the kind before, and asked for some siecimens for
the Eoyal Microscopical Society of London.

The boy’s hairs are not so uniformly nodose as th? girl’s ; still
his head afforded many perfect specimens, and prettier micro-
scopic objects than hers, being almost altogether free from
colouring matter. Many of the imperfect specimeis had only
a couple of nodes, the remainder of the shaft beng normal.
There is not the slightest trace of a fungus on oi in any of
the specimens, neither is there any appearance to had to the
inference that the nodes could have been caused by such. At
the constrictions the imbrication of the scales can be asily made
out, but only at these points.

The girl’s hair, to the naked eye are regularly chec'ied, black
and white; the boy’s have no such appearances, owiig to the
absence of pigment. The specimens were mounted h Klein’s
dammar solution. He might mention before they were ixamined
that those from the boy’s had not been previously exhibted.

Tongue of Cyclothurus. — Mr. P. S. Abraham, exhibited!, section
of the tongue of Cyclothurus^ the Central American ‘ Tee Ant-
eater,’ a microscopical examination of which had not hitheto been
made. The section showed the general arrangement of the nuscular

DUBLIN xMICROSCOPICAL CLUB.

117

bundles, remarkable for the great distinctness with which ‘ Cohn-
heim’s areas ’ came out on the cross section of the fibres. The
stratified mucous epithelium was well seen and the submucosa
was apparently destitute of mucous glands ; the shiny secretion
which lubricates the tongue being probably the product chiefly
of the enormous sub maxillary glands.

Aorta in Bdminants. — Dr. Purser showed specimens of the
aorta of ruminants, cows and sheep. In these animals the
muscular tissue of the middle coat does not consist of single
film or small groups of films equally distributed between the
elastic plates or membranes, but in parts the elastic plates are
interrupted £nd the tissue is composed almost altogether of
muscular fibre, with only a few elastic threads interposed. In
these muscular patches the fibres run, for the most part, trans-
versely, but ii some cases obliquely or longitudinally.

Section of Tillous Tumour. — Dr. Charles Ball exhibited some
sections of a villous tumour of the rectum, which showed one
reason why these growths have been so frequently confounded
with true cacinomata, namely, that in consequence of the nu-
merous involitions of the superficial epithelium sections would
present the ippearance of endogenous growths of epithelium,
while in realty these appearances were due to the section having
traversed the bottom of superficial sulci.

Encysted Sate of Vacuolaria virescens^ Cienkowski, exhibited,
— Mr. Archer showed a condition of this conspicuous flagellate,
forming in i'self a pretty object, and one which, met with in an
isolated way had long puzzled him as to what it could be.
Here, however, were examples of this flagellate partially encysted,
and others vholly so, the latter still encircled by the original
cuticular ctA’ering or skin, standmg ofi? as a separate pellicular
investment. The inner body appeared as a “ spore-like ” struc-
ture, conteits dense, green, and quite destitute of any vacuolar
appearance, its bounding wall very thick, and covered by
a number of large sub hemispherical hyaline prominences.
This, withoit the cast-ofl’ parent, simply membranous coat, is, as
has been -*aid, not unlike a possible zygospore of say some
'desmidian &c., but something always told one it could scarcely
80 be. E'idently Cienkovvski’s original examples had not gone
on to th( formation of this thickened and unduly margined
inner wal, or he would have figured this condition : he repre-
sents onk a still, smooth, thin-walled condition, not uncommon.
Indeed i would only be by obtaining examples in several con-
ditions fiat the connection between the two extremes could be
Been, liese bodies, apart from their being still surrounded
simply ly the parent “ cellwvall,” could not be regarded as truly
“spores’ or, as in any way resulting from conjugation; if they
were fomed by the union of two individuals, they would have
been lager ; on the contrary, they appeared rather smaller, but
more dtise, than the ordinary freely swimming examples. They

118

PROCEEDINGS OF SOCIETIES,

probably merely represent a “ resting ” or “ encysted ” condition
of this organism.

It was then

Hesolved, “ That we record with unfeigned sorrow the death of
our colleague, Dr, David Moore. The records of the Club testify
to the number and value of his scientific contributions ; but our
members have in addition to mourn the loss of one whom we
each and all esteemed as a personal friend, and whose vacant
place will leaye a serious blank in our limited ranks.”

MEMOIES.

The Coffee-leaf Disease of Ceylon. By W. T. Thiselton
Dyer, M.A., F.L.S., Assistant Director, Royal Gardens,
Kew. With Plates IX, X, XI, XII, XIII, and XIV.

During the last ten years a parasitic fungus has devas-
tated the coffee plantations of Ceylon and Southern India,
and is now slowly spreading eastward through the coffee-
growing countries of the old world. Its history commands
our attention as the enemy of one of the most important
industries of the tropical possessions of the empire. In
itself, moreover, it is an interesting biological study, supply-
ing, as it does, a striking analogy in its first appearance,
progress, and effects, to the epidemic diseases of animals, in
regard to which, however, so far, there is only a presump-
tive probability in favour of the view that the physio-
logical disturbances or organic lesions which are charac-
teristic of them are due to the introduction into their
tissues and subsequent multiplication thereof some extraneous
organism.

The early history of the disease is well ascertained in
Ceylon. Before 1869 it was altogether unknown, and the
fungus which causes it had not been detected, notwithstand-
ing that the plants of this group had been diligently collected,
no less than 1190 species (of which two thirds are endemic)
having been described.* In May of that year, in a com-
paratively new estate in the Madulsima district, which
occupies the south-western portion of the hilly country,' a
few plants were noticed which had light-coloured patches
on the under side of their leaves ; and these patches, on
closer examination, were noticed to be dotted over with
orange-coloured dust. In July following two or three acres
were found to be affected in this way. From this time it
rapidly spread, till in 1873 there was probably no estate in
the island entirely free from it.

* Berkeley and Broome, ‘ Journ. Linn. Soc. Bot ,’ vol. xiv.

VOL. XX. NEW SER. I

120

W. T. THISELTON DYER.

From the first appearance of th<^ disease its serious charac-
ter was recognised. Specimens of the diseased leaves were
sent to this country, and were examined by the Rev. M.
J. Berkeley, who named it Hemileia vastatrix, and published
a short notice describing it in the ^ Gardener^s Chronicle ’ for
1869 (p. 1157) with a figure.^ He not merely determined
it to be quite new to science, but with difficulty referable
to any recognised section of fungi.” He considered it to be
intermediate between Mucedines and TJredinei, a view
which is not sustainable since the Rev. R. Abbay ^ and
Mr. Morris, after working upon the plant in a fresh
state, have both agreed in affirming that the reproductive
bodies regarded by Mr. Berkeley as spores are in reality
sporangia. This observation widely alters any possible
view of the affinity of the fungus, but it has not, so far,
led to any very definite systematic position being assigned
to it.

The coffee tree is an evergreen, and the speedy effect of
the attack of the fungus upon its young leaves is to cause
them to fall off, leaving the tree almost bare. After a time
it throws out fresh healthy-looking foliage. Sooner or later
this in its turn exhibits the fatal spotting, and is shed pre-
maturely. It is said that there is no known instance of
coffee trees being actually killed by the attacks of the
disease, but they become much enfeebled by the repeated
loss of foliage and eventually become worthless as crop
producers.

The following pages give in a condensed form the net
result of the investigations to which the life-history of the
Hemileia has been subjected. The plates have all been
drawn from drawings by Mr. Morris, and have been borrowed
from a book which he is preparing for the use of the Ceylon
planters.

To the naked eye the first appearance of the Hemileia
is indicated by a slight transparency or palish discolour-
ation,” easily noticed when the leaf is held up to the light.
These transparent spots indicate the points where infection
of the leaf has begun. As the ‘‘ spot ” becomes larger and
older it assumes a faint yellow colour ; ultimately on the
under side of the leaf it becomes covered with a bright
yellow dust, and this later on changes to a bright orange.
(Plate IX, fig. 1, represents the under side of a coffee-leaf
in a diseased state.) The discoloured patches are irregular

1 Reproduced in ‘ Quart. Journ. Mic. Sc.,’ 1873, p. 80.

2 * Journ. Linn. Soc. Bot.,’ vol. xvii, p. 176.

THE COFFEE-LEAF DISEASE OF CEYLON, 121

in size, but always preserve a circular outline, except where,
as occasionally happens, the accidental coalescence of two
patches produces one of irregular form (PL IX, hg. 2, a).
They are most numerous towards the apex of the leaf, the
tissue there being somewhat younger and more succulent,
and therefore more easily attacked. The centrifugal deve-
lopment of the patches is sometimes arrested by the barrier
presented by the midrib or a vein, but this is not always the
case (PL IX, fig. 1, a).

When a diseased patch is examined with a low power it is
seen to be covered with the somewhat symmetrically disposed
clusters and sporangia, the most mature being in the centre
(PL IX, fig. 4). Further examination shows that these
clusters occupy stomata, PL IX, fig. 5 ; and fig. 6 shows a
cluster in which some of the uppermost sporangia have
fallen off, allowing the oblique attachment of the rest to the
mycelial threads which fill up the openings of the stomata
to be seen. Mr. Morris describes the fully ripe and detached
sporangium as an ovate, oblong, or reniform body, covered
externally on all sides but one with globose or wart-like
papillae. The smooth surface of the sporangium is under-
neath on the side of attachment, and the fact that on all
other sides the sporangium is covered by wart-like papillae
gives it a peculiar character, which, according to Mr. Berke-
ley, is found in only one other fungus, viz. in Badhamia.^*
The immature sporangia are destitute of papillae.

The Hemileiai^ itself not without its enemies. In PL IX,
fig. 1 A, a dark-coloured patch is shown where the sporangia
have been attacked by some other fungus, possibly an Asper-
gillus. In fig. 3 the sporangia are colourless towards the
centre of the spot, their contents having been destroyed by
some other agency of the same kind; in this patch also is
represented the larva of a small undescribed dipterous insect
(fig. 7), which likewise feeds on the ripe sporangia. This
larva was first noticed in 1874 ; Dr. Thwaites then remarked :
‘‘ Sometimes these little maggots are very numerous, and it
could be wished they were still more so, and that thus the
fungus spores might be altogether consumed and the pro-
pagation of this terrible pest arrested.^’

If a vertical section be made through one of the diseased
patches of a coffee-leaf the connection of the mycelium with
the masses of sporangia can be traced and the destructive
eflect of the former upon the leaf tissue readily studied. The
bundle of mycelial filaments, to each of which a sporangium
is attached externally, forms a felted mass, like a plug, which
fills up the cavity, into which a stoma opens above. This

1^2 W. T. THISELTON DYER*

mass apparently becomes so consolidated that the fact of its
composition from compacted mycelium eluded the observation
of Mr. Abbay, who thought it was a sac-shaped dilatation
of a single mycelial filament.^ To Mr. Morris is due the
credit of clearing up this pointy though Mr. Berkeley, in the
figure which accompanied his first description, clearly indi-
cated that the sporangia (or, as he thought them, spores)
sprang from individual threads.

Tracing the mycelial filaments further into the leaf, which
is, of course, reversing the course of their development out-
wards, they are seen to assume a branching coral-like
habit (PI. XII, fig. 2, c). They feed on and destroy the
spongy parenchymatous tissue of the leaf where they come in
contact with the chlorophyll-containing cells. These are
broken down, cavities are produced (PL XII, figs. 1 and 2,
h), and the leaf at these points appears consequently of a
paler colour.

When mature the sporangia appear to be readily detached,
and are liable to be dispersed by the wind or by animals, or
even the clothes of persons working in or visiting the plan-
tations. Probably the papillae with which they are covered
on all but one side give them a mechanical adhesiveness.
At first the sporangia are filled with dark-red or orange
granular matter ; when fully ripe they are seen to contain
a varying number of minute nucleated spherical sporidia,
ranging, apparently, from one or two to fifteen or twenty.
According to Mr. Abbay, the sporangia have two coats, an
outer one, bearing the papillae which disintegrates,” while
the inner membrane opens out and slowly dissolves away.”
Mr. Morris, however, finds that in very ripe sporangia the
smooth surface is ruptured, and the sporidia by this means
escape. When the sporangia are placed on a moist slip of
glass resting on damp blotting paper the sporidia begin to
germinate within twenty-four hours, and for the most part
inside the sporangium, the mycelial filaments being pro-
truded through its wall, which softens and gives way before
it (PI. X, fig. 2, d) ; some germinating sporidia, however,
are generally found also, which have escaped from the spo-
rangium (fig. 2, c). The result is that everywhere, after an
attack of leaf disease, in the neighbourhood of diseased coffee
trees, whether on fallen leaves, moist surfaces, or on the
stem or branches, a web of mycelium is produced, by the
continuous development of which the whole of the exterior
of the coffee tree is completely invested with the toils of its

* Loc. cit., p. 176.

THE COFFEE-LEAF DISEASE OF CEYLON. ' 123

enemy, although there is nothing to the naked eye which
reveals its presence. The extent to which this takes place
maybe readily judged from the two figures in Plate XI;
fig. 1 is a portion of the epidermis of a young branch, and
fig. 2 is part of the under surface of a coffee leaf. Accord-
ing to Mr. Abbay there is a mode of propagation in this
stage by conidia which are produced during the wet weather
which is so favorable to the progress of the mycelial filaments.
These conidia are developed at the ends of the branches
of the mycelium in the form of radiating necklace-like
strings of minute spherical bodies, the whole arrangement
closely resembling the fructification of an Aspergillus,
Whether these conidia are really part of the life-cycle of
Hemileia must still be regarded with some doubt, inasmuch
as fungus-cultivations are peculiarly liable to error from the
introduction and development of the spores of other species
than that which is the object of study.

Mr. Morris finds that during the four or five months pre-
ceding what the planters recognise as an attack of leaf
disease,” and which is, in reality, nothing more than the
period in the life-history of the Hemileia when the sporangia
are developed, nearly every part of the coffee tree is invaded
by mycelial threads. The absence, at the end of this period,
of any empty sporangia leads him to think that the mycelium
has travelled from some distance to finally invade the young
foliage. If these were infected by sporangia blown on to
them by the wind there would be some indication of their
presence ; none, however, are found, and it may be concluded,
therefore, that the mycelium has travelled to them from the
stem and branches, or even from the ground. As this
amount of growth could not be sustained by the small quan-
tity of nutrient material contained in an individual sporadium,
it must be concluded that the mycelium feeds in its course
upon the organic matter dissolved in the moisture which,
during the wet season, copiously bedeWs the sufaces over
which the mycelium passes.

As long as the mycelium is external to the coffee plant it
apparently does it no harm, however copiously it ramifies
upon its surface. But having reached the young foliage
during the moist season the filaments begin to invade
their internal tissues, and, in fact, to commence the
ravages which are so injurious to the health of the coffee
tree.

The stomata of the leaves are closely set, and the mycelium
is no less closely woven into an interlacing web. Sooner or
later; therefore; a filament finds its way into a stoma, or,

124

W. T. THISELTON DYER.

passing over it, sends down into it» opening a lateral branch.
The mycelial filaments, however derived, now enter the
intercellular air passages of the leaf and immediately alter
their habit of growth. From being slender, filamentous, and
attenuated, they become thicker, more copiously furnished
with short branches, and exhibiting, in fact, what has been
called the coral- like habit” (PL XII, fig. 1, g). They
excavate the parenchymatous tissue bounding the inter-
cellular air passages, which is, of course, to attack the plant
in a most vital and important point. The branches may be
seen to impinge on one of the chlorophyll-containing and
starch-forming cells, the wall undergoes absorption, and the
contents are speedily evacuated. These injuries being in-
flicted when the leaf is still young and its new foliage
scarcely matured, the lesion appears to cause a disturbance
of the health of the leaf, at first sight scarcely explicable
by the comparatively small proportion which the tissues
actually destroyed bear to those which are unaffected. But
the internal cavities produced intercept the paths taken
towards the plant by the carbohydrates fabricated by the
leaf, and towards the leaves by the water charged with saline
matter brought up to them from the roots. The functions
of the leaves are completely disorganised, and, as always
happens when this is the case, when the struggle ceases any
longer to be possible, they are shed.

The young shoots which bear the new foliage commonly
perish, and, at any rate, the coffee berries which they bear
are not matured ; those on the older part of the branches
are badly developed, and the crop is both reduced and de-
teriorated by the admixture of ill-developed berries. The
effect upon the coffee industry has been very serious. Mr.
Abbay estimates that up to 1871 the average yield for five
years over the whole of Ceylon had been 4*5 cwt. per acre,
whilst for the five succeeding years the average has only
been 2*9 cwt. For the ten years during which the disease
had existed from 1869 to 1878 he estimated the total loss
in crops alone due to the disease as from £12,000,000 to
£15,000,000.

In order to conclude the preceding sketch of the life-
history of the Hemileiuj it is only necessary to add that, after
feeding for some time on the leaf tissues, the mycelial
filaments send off lateral branches into the air chambers
beneath the stomata, and the aggregation of these grow up
side by side in a closely compacted bundle. On emerging
from the mouth of the stomata each filament develops from
its apex a sporangium. The object of this process of fructi-

THE COFFEE-LEAF DISEASE OF CEYLON. 125

fication is apparently to enable the Hemileia to bridge over
the dry season, the sporidia immediately germinating on the
approach of wet weather.

Mr, Morris, from whose forthcoming handbook most of the
facts put together in the preceding sketch are taken, was
during the greater part of last year placed on special duty
for the purpose of coffee-leaf disease inquiry. He was of
course extremely anxious to discover some vulnerable point in
the life-cycle of the Hemileia at which it would be open to
attack. It is to his persevering investigations that we owe
the clear account of its habits which we now possess. Mr.
Morris recommends that during the period of external
development, when the mycelium is merely mechanically
attached, as it were, to the surface of the coffee tree, it
should be treated with some remedy which will destroy its
vitality without injuring the coffee tree itself. It is not
necessary in the present paper to discuss the attempts and
experiments which have been made in this direction, more
especially as Mr. Morris will discuss them in his forth-
coming publication. But as far as can be judged from
experiments on a small scale the destruction of the filaments
can be completely effected by dusting the whole coffee plant
with a mixture of one part of sublimed sulphur and two parts
of lime ; the effects are very well shown in Plates XIII and
XIV, the former of which shows the epidermis of a branch,
and the latter that of the under side of a leaf traversed by
Hemileia mycelium, both before and after treatment with
the mixture. Mr. Morris having been transferred by the
Colonial Office from his post as Assistant Director of the
Royal Botanical Gardens, Peradeniya, Ceylon, to the Direc-
torship of the Botanical Department of the Island of
Jamaica, Mr. H. Marshall Ward has proceeded to Ceylon
to take up the inquiry where Mr. Morris left it. There are,
of course, wide practical differences between the conditions
of a laboratory experiment and the treatment of a planta-
tion. But whatever is the ultimate result, to Mr. Morris
belongs the credit of having applied a strictly scientific
method of investigation to the problem of coffee-leaf disease,
and of having devised the only method of treatment which
so far affords any rational hope of success.

It is only fair to observe that Dr. Thwaites, the distin-
guished late Director of the Peradeniya Gardens, whose
opinions must carry great authority, hesitates in accepting
Mr. Morris’s results. He does not, indeed, gainsay them, but
he apparently thinks that the account of the life-history of
the Hemileia given by Mr. Morris is incomplete, and that the

126

W. T. THISELTON DYER.

remedial measures which he has based upon his view of the
matter are inadequate.

Dr. Thwaites, in a letter to the Ceylon Government, written
September 2Tth, 1879, points to the fact, which is undoubtedly
striking, that the disease, after its first appearance at Madul-
sima in 1869, showed itself at widely separated points and
over considerable areas in the two following years. He thinks
this must, no doubt, have been due to the presence of
inconceivably minute germs floating in the atmosphere and
carried by the wind/’ These minute germs are, however,
as far as I have been at present able to understand, com-
pletely hypothetical, and, as will be seen, it is possible to
explain the transference of the disease to remote points from
its original centre of dispersion in accordance with known
means. Before the investigations of Mr. Abbay had first
brought to light the important fact of the extensive develop-
ment of the Hemileia in the mycelial stage, Dr. Thwaites,
at a loss to divine how the coffee plants could be infected
when the spores had all been shed, hazarded the remark.

In what way the coffee tree receives the infection remains
to be ascertained, and from the subtlety of the operation this
will have to be inferred, rather than discovered, by direct
observation of the process. It would seem most probable
that the infecting matter contained in the spore is absorbed
by the tender rootlets of the coffee tree, although it is
possible to conceive it might also be introduced into the
tree through the very young foliage.” This was in 1874;
in the following year Mr. Abbay, having demonstrated the
universal prevalence of the mycelium. Dr. Thwaites felt
himself constrained ” to give up the idea of an “ absorp-
tion of the fungus matter by the roots.” He is, however,
still unable to account for the fact that after a coffee plant
has had all its leaves removed and has been treated with
sulphur and lime, its young foliage speedily exhibits once
more the evidence of the disease. Mr. Morris attributes
this to direct infection by wind from neighbouring planta-
tions or wild coffee trees on which the Hemileia is in a state
of ripe fructification, and he points out reasonably enough
that, till all plantations unite in preventive methods, and
wfild and neglected coffee is destroyed and rooted out, no
method of remedial procedure has a fair trial. Dr. Thwaites,
however, in 1874, expressed the opinion that the fungus
was present in the growing tissues generally of the coffee
pl^nt in diffused form,” and in his latest communication on
the subject, dated January 8th of the present year, he states
that the phenomena of the earliest appearance of the leaf

THE COFFEE-LEAF DISEASE OF CEYLON.

127

disease seem to be very analogous to what are

observed in vaccinic or other such contaminations.”

Anything falling from Dr. Thwaites on such a subject
naturally carries great weight. As has been already re-
marked, it is possible that the remedial methods suggested
by Mr. Morris may prove practically useless. But it may
be claimed for them that they are based on intelligible facts,
and if they fail the practical proceeding is to try and ascertain
why, and improve the curative treatment accordingly. But
Dr. Thwaites’s criticisms are based, not so much on facts, but
on theoretical conceptions. How, it may be asked, can
‘‘ infecting matter ” be absorbed by the roots of a plant to
subsequently give origin to a definite organism in its tissues ?
How can this organism exist in those tissues in diffused
form ?” And what is the precise meaning to be gathered from
these expressions and from the supposed analogy of the
Heinileia to “ vaccinic contamination ?” These quotations are
taken from Government documents officially published for
the purpose of communicating to the planting community
the voice of science on a subject of the gravest importance
to them. Dr. Thwmites certainly owes it to himself to
express, in a more precise way, and to support by scientific
evidence, the views which he has at present only somewhat
vaguely indicated, but which seem to lean towards an ex-
planation of the disease as a persistent constitutional in-
fection.

The leaf disease seems to have speedily shown itself in
Southern India after its first appearance in Ceylon. In
1876 it appeared in Sumatra, in 1879 in Java and Bencoolen,
and this year news has reached this country of its appearance
in Fiji, where, however, it was first noticed in May of last
year. The outbreak in the latter colony is a misfortune
much to be regretted, as coffee promised to be one of its most
successful enterprises. Very full details have been sent by its
Government to the Colonial Office, and the whole narrative
of the outbreak affords a most striking analogy to the introduc-
tion of an epidemic, such as that of measles, into Fiji itself.^

The disease was in all probability introduced directly from
Ceylon to Fiji with a box of seeds packed in earth early in
1879. A few months later it made its appearance in the
Great Amalgam Plantation, and subsequently spread to other
plantations on the Ilewa river, in the apparently capricious
manner which Dr. Thwaites found so difficult of compre-
hension in Ceylon. Fortunately, in Fiji the Government
placed the matter in the hands of Dr. W. McGregor, the
^ Moseley ‘ Notes by a Naturalist,’ p. 341,

128

W. T. THTSELTON DYER.

Chief Medical Officer, and under his directions a very bold
and skilfully conducted attempt has been made to stamp out
the disease. His whole report w'ould be well worth printing
as a model of procedure based upon strictly scientific prin-
ciples. The following passages may be quoted :

On each plantation, the Great Amalgam included, it is
easy to see at what point the disease first took hold, and that
it is radiating from that spot. The great probability is,
therefore, that the disease is carried about from one place to
another by physical agencies, a view of its distribution that
accords with the glutinous properties [papillose surface] of
the spores [sporangia], that cause them to adhere readily to
any object with which they are brought into contact. Many
people have visited the great Amalgam Estate, curious to see
the coffee-leaf disease, and several of them have afterwards
visited the other patches of coffee in the same district, and
the presumption is that measures were not in each case
adopted to prevent the spores of the disease being thus
carried from one place to another.’^

*****

On the 22nd I saw the estate of Viti. It was easy to
see the disease had there begun on the trees nearest the
landing-place from the river, as the trees growing at that
spot were covered by the disease for a short distance around
them, but the disease became gradually less prevalent as the
distance from the landing-place increased, and at the further
end of the plantation was not apparent.”

The stamping out” having been sanctioned by the
Government, Dr. McGregor gives the following account of
the method adopted :

1 proceeded next day to near Viti, the place to be dealt
with first, as being the largest and most liable to be tra-
versed by natives and others. As a description of the pro-
ceedings undertaken at Viti will give a correct idea of how
the same work was carried out at each of the other places, I
shall give the former in detail. Our camp was pitched on
the other side of the river, nearly opposite Viti. Food was
supplied from the native town of Na-Toa-Ika, on the same
side of the river as the encampment. In the morning
each of the natives employed put aside his sulu, substituting
one of banana leaves. I myself and the other two Eu-
ropeans with me — Mr. Vaughan and Mr. Gibb — took a spare
suit of clothes, which we left in the boat by which we
all crossed to Viti, and which was anchored in mid-stream
during the day when we were at work. At night the men

THE COFFEE-LEAF DISEASE OF CEYLON.

129

left their banana sulus on the ground, and were thoroughly
washed in the river before they could get near the boat.
The Europeans walked into the river, where they and
their clothes were thoroughly washed before entering the
boat to put on clean clothes. By these means the camp
was kept as clean as possible under the circumstances, as
this method was rigidly carried out.”

The whole of the infected coffee was burned with all the
herbage on the ground which could retain the sporangia, and
additional fuel was carried on to the ground where necessary,
so as to enable a continuous wave of flame to pass over it.

The extraordinary prominence which the Hemileia has
suddenly assumed is at first sight not a little astonishing ;
but the ravages of pai*asitic upon cultivated plants, it is not
difficult to see is one of the consequences with which we
have to reckon in disturbing the established order of nature.
Hemileia vastatrix is a parasitic plant which has probably
for ages lived in a state of constrained commensalism with
some other members of the family of Hubiacecey to which
coffee belongs. The constituents of tropical forests are not
gregarious ; the native host of the Hemileia is, perhaps, some
plant of little but botanical interest, of which individuals
- are scattered at intervals through the forest. Parasitic
plants in nature obviously cannot extirpate their hosts,
because, if they did so, they would determine their own ex-
tinction. They only exist at all, therefore, by doing the
minimum of injury to their hosts which is compatible with
their own continued existence. But when man places
within their reach vast aggregations of plants ready for in-
vasion the latent capacity for mischief of the parasite is
brought out to the uttermost, and it runs riot at man’s ex-
pense. If Ceylon were abandoned to nature the coffee planta-
tions would lapse again into mixed forest ; coffee would
either succumb to the Hemileiay or would only be repre-
sented by those individuals which proved able to carry on a
commensal existence with it.

It is remarkable that though the coffee districts of the
New World are free so far from the Hemileiay they are
also ravaged by a leaf disease. In this also the parenchyma
of the leaf is devoured and destroyed with parallel consti-
tutional effects. The parasite is, however, in this case not
a fungus but the larvae of a minute moth {Cemiostoma
coffeellum'). This seems to have first taken to its mis-
chievous vocation in the West Indian Islands, and thence
to have been introduced on the South American continent,
in all the coffee- growing districts of which it is a serious pest.

130

J, D, SIDDALL,

On Shepheardella, an Undescribed Type of Marine

Rhizopoda; with a Few Observations on Lieberkuhnia.

By J. D. SiDDALL. With Plates XV and XVI.

A FRESH interest has attached to the Rhizopoda of recent
years, as extended research has furnished data for the more
systematic study of the group. But the labours of Hertwig,
Cienkowski, Archer, F. E. Schulze, and other eminent histo-
logists, so admirably summarised by Professor Allman in his
two Presidential addresses to the Linnean Society, seem
rather to suggest how much there must be still to learn re-
specting the life-history of sarcode organisms, than to render
us satisfied with what is already known concerning them.
Under these circumstances it has appeared to me that the
following notes on certain little understood forms, especially
on one that does not appear to have fallen under the notice
of previous observers, may be worth placing on record.

On the 22nd of August, 1879, I received from my friend
Mr. Shepheard two bottles of sea-water containing Polyzoa,
Hydrozoa, and Sponges, which had been collected at low tide
on the previous day from the shore netir the Lydstep Caverns,
Tenby. For their better examination I placed a numl^m* of
the specimens in cells, and had the pleasure of finding that
most of them were Kving, notwithstanding their long railway
journey in small bottles. On the following day my atten-
tion was arrested by an organism, apparently a Rhizopod,
which had separated itself from the other things in the cell,
and was adhering to the glass, and which, but for the pseudo-
podia extended from its two extremities, would probably
have been passed over as a worm. With a view to the closer
study of this creature, a further supply of material from the
same locality was subsequently obtained for me by Mr.
F. Walker, of Tenby, who, during the months of September
and October, made various gatherings, from each of which I
obtained specimens not only of the species originally found,
but also of one or two other Rhizopods having features of
interest.

To Mr. Shepheard and Mr. Walker, therefore, I am pri-
marily indebted for the material from which'the present paper
was written; and, as the organism I purpose in the first place
to describe appears to be new, I have much pleasure in
associating it with the name of my friend and fellow-worker
Mr. Shepheard,

ON SHEPHEARDELLA, A NEW TYPE OF MARINE RHIZOPODA, 131

Shepheardella t^niformis. — Nov. Gen. et sp.

General description. — The body of the Rhizopod is uni-
cellular, elongated, and abruptly pointed at both ends. Flat-
tened and ribbon-like when in a state of activity, rounded
and worm-like when at rest. It is furnished with a flexible,
transparent, colourless integument of considerable flrmness ;
and the whole tubular cavity is densely filled with yellowish,
coarsely granular protoplasm, having a very distinct oval nu-
cleus, and occasionally also a few scattered non-contractile
hyaline vesicles. Thesarcode rotates in a regular stream around
the interior of the integument, and carries the nucleus along
with it ; the current performing a complete circuit within the
cell. The nucleus is seldom carried entirely round the cell,
being as a rule intercepted in its course along one stream
and passed over into the opposing one before it has travelled
far from the centre in either directions. The opposing
streams of sarcode thus formed on the two sides of the cell,
are not separated from each other by any clear line as in
Characece, though otherwise much resembling the pheno-
menon observable in the living vegetable cell. The prin-
cipal difference between the two is in the direction of the
current, that of Chara advancing in a spiral direction, whilst
that of Shepheardella completes its circuit in one plane, the
two currents slightly overlapping each other. The in-
tegument is perforated at each end by a minute aperture,
through which’ some of the sarcode passes and collects into a
small mass. From this mass a very delicate coating spreads
over the whole exterior surface of the integument, and this
thin layer occasionally throws out a pseudopodium. But
the great network of inosculating and branching pseudopodia
are at the two extremities of the organism, extending them-
selves from the terminal masses of sarcode to a distance
considerably exceeding the whole length of the body, as in
PI. XV, fig. 2. The circulation of the finer granular sar-
code in the pseudopodia is easily traced. It is very rapid,
advancing and returning in single, double, or even triple
streams, according to the breadth of the pseudopodium
traversed. So that from the rotation of the chief mass of
sarcode within the test and the external circulation in the
pscudopodia, it is evident that every part of the body-
contents, the nucleus excepted, is in turn brought in contact
with the surrounding water. The combined movements
present an evidence of vigorous life rarely exhibited among
the Rhizopoda, and the transparency of the “test^* permits
the internal functions of the organism to be as easily followed
as the external.

i32

j. D. SIDDALt.

Size, — The first specimen observed measured in length of
body, without the pseudopodia, *07 inch (1*75 millim.), and
in breadth *0016 inch (0*042 millim.). Since then I have
measured a number of others, the largest of which is *31 inch
(7*5 millim.) in length, and *2 inch (0*5 millim.) in breadth.
The rest vary between these two extremes. (PI. XV, fig. 1.)

Two days after finding the first I observed a second much
larger specimen, and examined it carefully whilst living with
powers varying from 280 to 800 diameters (^ in. objective
with A to D eye-pieces). Nothing fresh was noticed in the
sarcode, except that it contained, in addition to the dark
granules of irregular size and shape, a considerable number
of spherical and oval masses, relatively larger in size, of clear
colourless protoplasm. These were carried along by the
rotating current, but did not travel so quickly as the finer
portion of the sarcode, probably owing to their greater size
and weight. No traces of a contractile vesicle could be seen
although carefully sought for. Neither structure nor surface
markings of any sort could be discerned in or upon the in-
tegument, which presented a perfectly homogeneous texture,
and may therefore be regarded as a simple layer of hardened
or otherwise slightly modified protoplasm.

The nucleus proved to be a far more difficult subject for
study, but careful observation in upwards of twenty examples
has within the last few weeks revealed its structure. Ex-
amined in the living state very little can be made out satis-
factorily. In appearance it is a simple oval sac or cell
possessing a wall of exceeding tenuity, and filled probably
with fluid of greater transparency than the sarcode which
surrounds it. The chief difficulty arose from certain dark
lines, continually travelling over its surface ; not ceasing
either when the nucleus was carried along with the rotating
sarcode, or momentarily arrested. Some of the varying ap-
pearances thus presented are shown in PL XV, fig 4. After
watching these lines for many hours I was led to attribute
them to alternate wrinkling and distension of a nucleus-
membrane, taking place without any regularity as to time.
They appeared generally to open out from the centre to the
edges of its longer diameter, or vice versa. This seemed to
prove definitely the existence of a distinct wall of greater or
less resistance coating the nucleus, and it appeared to me
that whilst the lines were chiefly the result of the unequal
external pressure of the impinging currents of sarcode, they
might also be in some degree due to inherent causes ; and
continued observation has led me to believe that the latter
rather than the former is the correct explanation.

ON SH£PHEARDELLA_, A NEW TYPE OF MARINE RHIZOPODA, 133

On the 12th of March last I noticed a fine Shepheardella
entangled among the tentacles of a Hydratuba,” and
placed it between two thin cover-glasses for examination
with higher powers^ care being taken not to press it. Urti-
cating threads of the lasso-cells had pierced the integument
of the rhizopod in many parts_, but did not seem to cause
any inconvenience. The nucleus was finely developed, and
was watched most closely, in order to determine, if possible,
whether its movements were the same, or an effect of the
rotation of the sarcode. These movements proved, however,
more puzzling and erratic than ever, the difficulty being
increased by the density of the sarcode. Finding that no
satisfactory results could be obtained otherwise, I subjected
the creature to sufficient gentle pressure to force out most of
the sarcode, when I had the satisfaction of observing that
the nucleus continued its meandering course, somewhat after
the fashion of a large Paramaecium, for about half an hour
after all motion had ceased in the sarcode which surrounded
it. The rotation of the sarcode was at once interrupted and
speedily stopped altogether by the pressure applied ; the
nucleus gradually ceased to move to and fro in about thirty
minutes afterwards, a few signs of life being exhibited by
its protoplasmic contents for a few minutes longer. The
nucleus membrane was deeply contricted and wrinkled into
a constantly varying series of elevations and depressions,
forming avenues, through which the granular sarcode passed,
and at times appeared to receive increased impetus in doing
so. In consequence of these ever-changing appearances the
various parts of the nucleus could only be distinguished
from each other at rare intervals. Sufficient, however, had
been seen to prove that the nucleus possesses the power of
independent movement, but probably not in so high a degree
as the sarcode, in the more powerful stream of which it is,
as already stated, carried along with irresistible force.

With two exceptions the specimens of Shepheardella exam-
ined have each possessed one nucleus. One of the ex-
ceptional specimens was rather small and pale coloured, and
contained no visible nucleus ; the other was of full size and
colour, and possessed three nuclei, all of which were of full
size and exhibited the same phenomena of moving lines.
They were carried along independently in the sarcode stream,
sometimes coming into collision and becoming blocked for a
short time, and then forced on again with a jerk by the col-
lected force of the choked current.

This particular specimen was first observed on the side of
a bottle, to which it adhered so firmly by its extended pseu-

134

J. D. SIDDALL.

dopodia, that it was necessary to use a camel-hair pencil to
transfer it to a cell. In so doing I twisted it twice in the
course of its length, thus causing stoppages of the sarcode-
current. Before long, by a series of spasmodic movements
of its body, the creature succeeded in stretching itself out
quite straight upon the glass, when the jerking movement
at once ceased. As the internal rotation resumed its course,
slowly at first, and then more quickly, pseudopodia were
emitted plentifully. Since then I have observed the same
irritability in several other specimens, and believe that the
source of such movements is the nucleus, which furnishes
the impetus required to produce them. I have never
before seen any Bhizopod exhibit the same sort of irrita-
bility, or make such manifest attempts to rectify itself when
placed in an abnormal position. The power of curving
laterally, combined with that of crawling by the aid of its
pseudopodia, enables Sliephear della to move from place to
place with considerable rapidity ; but its habit is, when once
fixed, to remain for many days in the same spot with its
pseudopodia fully extended.

Abnormal forms. — Several specimens observed during the
months of October to December differed considerably from
the normal form and also from each other. Two of these
are represented in PI. XV, figs. 5 anfi 6. For some time I
looked upon them as distinct species, but having since
traced other specimens of Shephear della through similar con-
ditions during their breaking up,^’ I am disposed to con-
sider them as representing an early stage in that process.
One of them (fig. 6) was found on Nov. 29th. Its body
was continually changing in form, pseudopodia being ex-
tended from no less than thirteen distinct orifices in its
integument ; some of the main branches of pseudopodia
reaching to the extraordinary length of three-fourths of an
inch (20 millim.), and capturing infusoria in all directions.
Most of the food so captured was digested outside the body,
in temporary masses formed by the accumulation of sarcode
from the pseudopodia. The bodies of the abnormal forms
were also full of dark- coloured food particles, diatoms being
occasionally noticeable among them, but the sarcode of the nor-
mal Shephear della seldom seemed to contain any alimentary
or other foreign matter. On one occasion I detected a typical
specimen in the act of conveying towards an aperture a
relatively large mass of apparently faecal matter, so large
that it choked the orifice for a considerable time and was
eventually ejected suddenly and passed away by an excurrent
pseudopodium, and finally thrown off at some distance from

ON SHEPHEARDELLA, A NEW TYPE OF MARINE RHIZOPODA. 135

the body. The chief food of the animal appears to consist of
Infusoria, which are not as a rule conveyed into the body,
but digested among the pseudopodia outside the test.

Life history. — The phenomena connected with the life
history and reproduction of this remarkable Rhizopod have
not yet been fully made out, so that the remarks now
offered are in some measure provisional, and subject to cor-
rection at some future time. They may be given chron-
ologically, beginning with a note made August STth,
when two Shepheardellce under observation in a cell dis-
appeared entirely during that day and the following night,
leaving as their only traces a very great number of Amoebae
of various forms, some of which are reproduced in PI.
XV, figs. 16, 17, 18. What had become of the integu-
ments and coloured granular portion of the sarcode I failed
to find out. All the Amoebae were quite colourless ; most of
them possessed a nucleus, or more than one, and a contrac-
tile vesicle. They were very active in their movements for
from four to six weeks, increasing considerably in size
during the earlier portion of the period ; but towards the end
of that time their movements became slower and more
feeble, and finally ceased altogether, when they collapsed
into motionless specks of very finely granulated sarcode (PI.
XV, fig. 19), ultimately becoming so indistinct as to be quite
unrecognisable amongst the organic debris with which the
cell had got somewhat obscured.

Another point noticed was that if some of the sarcode
from the extended pseudopodia of a Shephear della was taken
up on the point of a camel-hair pencil, and broken up into
a number of particles in a drop of sea water on a glass slide,
each separate particle presently put out long filamentous
quivering pseudopodia ; and if pseudopodia from two such
particles happened to touch each other they at once coa-
lesced. After a very short time these characteristic pseudo-
podia were all retracted, and the particles became quiescent,
presently resuming active life, not as before, but as Amoebae,
crawling by lohose prolongations.

With a view to ascertaining if possible the mode of repro-
duction, I put two fine Shephear dellce into a cell on December
9th, and on the 10th, found they were alive and in all respects
similar to PI. XV, fig. 2. By December 15th, at 7 p.m., one of
these had assumed the nearly spherical shape of PI. XV, figs.
9, a, i, but was still encased by its integument. Two and a
half hours later it had stretched out again, and become in
form nearly like PI. XV, fig. 6, pseudopodia being put out
from ten different orifices. The changes of form during the

VOL. XX. NEW SER. K

136

J. D. SIDJ>ALL.

ensuing two hours were very considerable. By the follow-
ing morning it had again become nearly spherical, and
showed a tendency to divide into four parts by constriction.
On the 17th, at 9 a.m., all the contents were found to have
been expelled from the test during the preceding night, and
the now naked sarcode had divided into four unequal por-
tions (PI. XV, fig. 10), the empty, cast-off integument lying
near to them in a wrinkled mass. The four separated por-
tions changed shape very slightly and slowly, delicate blunt
extensions of transparent protoplasm occasionally protrud-
ing from the edges of each. In the evening of the same day
rapid changes were continually going on ; the sarcode passing
in sheets from one portion to the other. The drawing at
PI. XV, fig. 11, gives a good general idea of the phenomena at
this stage, though the actual condition varied every instant.
The principal masses literally poured their sarcode one into
the other. New centres towards which all would flow were
formed every few minutes. Some of these for a short time
would form separate and distinct masses, and then reunite
by the coalescing of the sarcode extensions ; so that there
was sometimes but a single mass, which within a few
minutes might be broken up into half a dozen. These
alterations continued for several hours, during which time
no trace of a nucleus, which had become invisible on the
first constriction of the Rhizopod, could be found. On the
following morning, December 18th, 9 a.m., the w'hole had con-
tracted into one principal mass (from which both lobose and
filamentous pseudopodia w^ere freely extended) except some
minute detached patches of the sarcode which were lying at
some distance from it. These exhibited feeble amoeboid
movements so long as they remained free, but were at once
reabsorbed into the principal mass on coming in contact
with any of its pseudopodia. Examined again in the even-
ing of the same day it \Vas found in the same condition, but
a still larger number of the outlying amoeboid particles of
sarcode had during the interval become detached. One
very small spheroidal particle (PI. XV, figs. 15, a, 5), contain-
ing rather larger granules than those usually present in the
sarcode, was removed to a considerable distance, and care-
fully watched until January 25th, by which time it had
gradually given off all its contents, and apparently melted
away. By December 18th, a very large number of Amoebae,
differing in no recognisable way from the ordinary forms, and
evidently separated from the sarcode which had been con-
stantly in motion for the past nine days, w'ere travelling
about all over the cell cover. These, when once fairly

ON SHEPHEARDELLA^ A NEW TYPE OP MARINE RHIZOPODA. 137

established and difierentiated into eiidosarc and ectosarc,
crawled with impunity into actual contact with the pseudo-
podia of the other Shephear della, which had up to this date
retained its normal form, without in any one instance
becoming absorbed, as was the case in their earlier stages.

During the night preceding December 18th the second
specimen began to change its form. Its appearance at 9 a.m.
on that day is represented in PL XV, fig. 7. The sarcode had
entirely left one portion of the integument, and collected at
the other extremity, which was much swollen in conse-
quence. The nucleus could still be seen clearly, and the
internal rotation went on in the sarcode, but, instead of two
apertures, it now possessed at least four, from each of which
pseudopodia w'ere emitted. The contraction of the sarcode
continued slowly and steadily throughout the whole of this
day and the next, the nucleus became more obscure as the
protoplasm became denser, and finally was lost to view alto-
gether, the pseudopodia being at the same time gradually
withdrawn from the wrinkled end. At 7 p.rn. on the 19th
the now much altered Shepheardella adhered to the cell hy a
few pseudopodia from one end only, and had become broadly
oval in contour, as in PI. XV, fig. 8, a, h. Examined twenty-
four hours later, the integument had become completely
emptied and lay in a wrinkled mass on the free end of the ex-
uded sarcode (PI. XV, fig. 9, h, The naked body was

adherent to the glass by its opposite extremity, and was deeply
constricted in a quadripartite manner, and evidently about
to subdivide. Further observations on this specimen were
unfortunately prevented by the partial leaking of the cell
and the consequent loss of the object.

To return now to the first specimen, whose changes up to
the evening of December 18th have been already noted. I
found on the 19th, at 9 a.m., that it had once more divided
into two subglobular masses, which continued to give out
pseudopodia freely, especially under the influence of the
warmth and light of a lamp, until the 27th, when the only
remaining evidence of life observable in them ^vas a feeble
swarming among the granules of the sarcode, and this also
ceased after a day or two. After remaining motionless
from December 29th until February 15th one of the masses
suddenly gave evidence of life by emitting, whilst under
observation,’ a lohose extension of protoplasm containing three
enlarged granules. In a short time this travelled away as
an Amoeba, within whose body the granules appeared as
nuclei, no vesicle or vacuole being then present, hut, ex-
amined again next day, the granules were found to be re-

138

J. D. SIDDALL.

placed by one very large hyaline vesicle containing denser
matters within it. Watched for some time this vesicle was
seen to travel towards and ultimately protrude beyond the
protoplasm of the Amoeba, then rupture and discharge its
contents into the surrounding water, wherein they appeared
as bright spots held together by a viscid substance, the
vesicle afterwards collapsing and disappearing entirely. From
one of the two masses a thin layer of protoplasm is still
(March, 1880) extended, immersed in which are a number
of very transparent and much enlarged nucleated granules,
of tolerably regular form. In the larger of these a general
slow movement of the contents is discernible, and some
which became detached afterwards put out Actinophrys-X^^
pseudopodia, as represented in PI. XV, fig. 13, ajh,c. These
new forms generally develop a nucleus and small internal
contractile vesicle, and in one instance a specimen divided
into three distinct individuals. Like the Amoebse, most of
these have now become quiescent, and the tv/o masses of
sarcode have become smaller and paler in colour, in con-
sequence of the detachment of particles from them.

From the foregoing account it will be gathered that little
beyond the dissolution oiShephear della into amoeboid particles
has yet been quite satisfactorily traced. No attempt at
fission, encystation, or anything approaching to either, and
no development of special reproductive bodies, unless we
accept as such the minute spheroid drawn on PI. XV, fig. 15,
having yet been observed. The loss of the specimen pos-
sessing three nuclei was a matter of much regret, as I had
hoped, judging from what has been noticed in other simple
organisms having more than one nucleus, that it might ul-
timately divide into three distinct individuals, and by so
doing give conclusive evidence of at least one process of
reproduction. As it at present stands, the life history of
Shepheardella may be looked upon as a chain, a few links of
which are here presented, the major portion being still
missing.

The similarity of Shepheardella, in point of external form,
with the Gregarina gigantea of M. Van Beneden, has
attracted my attention ; but as it is not known that the Gre~
garinidoe, except in the very earliest stages of their existence,
resemble the Rhizopoda in possessing the power of extend-
ing pseudopodia, the idea which presented itself that the
animal I have described might possibly be only one stage of
some large parasitic creature of this class, has not yet been
seriously followed up. Should there be anything in its life
history to verify or controvert this superficial resemblance

ON SHEPHEARDELLA, A NEW TYPE OF MARINE RHIZOPODA. 139

it will doubtless become apparent in continued observa-
tions.

Histology » — Beyond what has been stated nothing has been
elicited from living specimens concerning the structure of
either integument, sarcode, or nucleus, but the examination
of a series of mounted examples, some of which had been
previously treated with reagents, has revealed some further
interesting details in the two portions last named.

As to the integument, further investigation only corrobo-
rates the description already given, that it is a perfectly
transparent, structureless, homogeneous membrane, flexible
and elastic. Viewed in optical section with high powers,
that is to say magnified from 600 to 1000 diameters, it
appears to be double under some conditions of illumination ;
but this is probably an optical illusion, caused by the unequal
density of the surfaces, for when torn across it does not ex-
hibit a double structure, nor is it separated into distinct
layers by the action of reagents. Dilute acetic acid softens
and swells it, and iodine stains it brown, but it is unaffected
by carmine.

Lying immediately within the integument, and closely
adhering to it, is an exceedingly thin stratum of very finely
granulated colourless protoplasm. Acetic acid causes this sub-
cutaneous layer to rise into clear spherical masses destitute
of granules but furnished with a large number of vacuoles.
A camera tracing of it, mounted in glycerine jelly, magnified
600 diameters, is given on PI. XYI, fig. 1. The sketch is
taken from a specimen in which the denser sarcode occupy-
ing the interior had retreated naturally from one end,
leaving the integument with this delicate lining layer ex-
posed. The nucleated granules and masses of firmer proto-
plasm seen in the sketch are detached portions of the internal
sarcode remaining upon it.

The densely granular yellowish sarcode which occupies
the whole interior of the body is affected in a very marked
manner by osmic acid, dilute alcohol, and picro-carmine,
the use of which was suggested to me by Professor Lankes-
ter. By the application of these reagents each granule
is rendered distinct and separate, and has the appearance of
a nucleated cell, the central speck taking the carmine colour,
whilst the peripheral portions remain yellow. The drawing
(PI. XVI, fig. 2) is a camera tracing of a portion so treated,
mounted in glycerine, and magnified 600 diameters. Some
of the granules exhibit different stages of multiplication
into two or more by fissiparous division, and their minute
size will be realised by the figure referred to,

140

J. O. SIDDALL.

The number of these nucleated granules contained in each
perfect Shepheardella is, of course, very great. They are
packed so closely that the sarcode seems to owe its density and
yellow colour to their presence, and yet it is not difficult to be-
lieve that each one of them may, by the continued exercise of
the process of free-cell formation, and possibly, also, the exu-
dation at the same time from each cell of a coating of living
protoplasm, possess the power of becoming, under favorable
circumstances, a new individual like the parent Shepheardella.
The probability of this is increased by the great enlargement
in size which takes place before the granules are liberated as
separate living organisms, as shown in PL XV, fig. 13 h.

The nucleus, which in living specimens appears to be so
simple, is found by the more deliberate examination of
mounted specimens to be of exceedingly complex structure.
It consists of three portions, viz., a lenticular oval body,
which may be termed the nucleus proper, an embracing but
not continuous inner coat, and an outer membranous sac
enclosing the whole. The nucleus proper is an oval biconvex
cell, bounded by a very delicate wall and filled with proto-
plasm, which in PI. XVI, fig. 5 h and a will be seen to be
contracted and coagulated by the action of the reagents round
the edges of the cell ; and at fig. 4 to be dispersed over its
whole interior surface. It possesses within it a distinct
secondary nucleus of somewhat denser protoplasm, and in
this again a central spot or nucleolus of its own. Of the
two surrounding envelopes, the inner one is a simple trans-
parent membrane very slightly roughened or granulated in
texture, apparently composed of firmly coherent specialised
protoplasm, and on this the nucleus lies ; the free thickened
edges of the envelope enfolding but not completely covering
it (see PI. XVI, figs. 3 4 S, 5 c, 6 c). The outer envelope

is a simple membranous continuous sac, completely enclosing
both nucleus proper and inner coat (PI. XVI, figs. 3 c, 5 c?,
and 6 d). It is perfectly structureless, and so transparent
that when tightly stretched by the widely opened inner coat
it cannot readily be distinguished, but is easily seen as a
somewhat wrinkled thin skin when the inner coat is enfolded.

Figures 4 and 5 are taken from specimens treated with
dilute acetic acid and subsequently stained with carmine
solution. This method of treatment appears to answer better
than any other for rendering the structure clear, and at the
same time staining the nucleus proper and its contents, whilst
the coverings retain pretty much their original colour.
Picro-carmine, after osmic acid and dilute alcohol, renders
the nucleus generally very distinct, staining it red, while

ON LIEBERKUHNIA WAGENERI.

141

the sarcode remains yellow ; but it has not so good an
an effect in other respects. Figures S and 6 are from selected
specimens mounted like the last two in glycerine jelly, but
not previously treated with reagents. In the former the
nucleus proper is plainly discernible, but its coverings are
somewhat confused, whilst fig. 6 shows the two coats very
clearly but not the nucleus itself.

From the appearances that have been described, it is evi-
dent that the moving lines noticed in the nuclei of the living
specimens are caused by the alternate opening and closing
of the folds of the inner envelope, and that the lines indicate
the approximation and separation of its free thickened edges.
This movement is probably initiated by the contraction and
dilation of the nucleus proper, thence communicated to the
embracing envelope, and by it rendered apparent.

The object of this movement, which takes place without any
regularity as to time, it is not easy to conjecture. It does
not appear to have any very close connection with, or effect
upon, the rotation of the sarcode within the body of the
animal ; but that it serves some definite purpose, probably
an important one in the economy of the creature, can scarcely
be doubted.

All the sketches of the nuclei were drawn originally twice
the size of the engraved figures, as it was found that the
details of structure were better displayed under higher mag-
nifying power.

At present it would be premature to assign to Shepheardella
t(£niformis any systematic position, nor could this be attempted
satisfactorily until the particulars of its life history are more
fully known. What has been already made out appears to
indicate that it belongs to the Rhizopoda, but there seems
no type hitherto described with which it has any close
affinity. Apart from its remarkable nucleus, there would be
little reason why it should not be placed near Lieherkiihnia
Wageneri^ in company with which it occurs at Tenby ; but
the presence of the complex body, the structure and move-
inents of which have been described, suggests a somewhat
higher organization than we are accustomed to associate with
Rhizopod life. Upon this point, as well as upon others
which have been mentioned, I trust to be able before long to
add to the notes now offered, which must be regarded as
tentative only.

The Other Rhizopod to which I have referred, Lieherkiih-
nia Wagenerij was first found in the neighbourhood of Berlin
by MM. Claparede and Lachmann, and was described and

142

J. O. SIDDALL.

figured in their ^ Etudes sur les Infusoires et les Rhizopodes.’^
The original figure (reproduced by Dr. Carpenter in his
'Introduction to the Study of Foraminifera/ PI. II) is a
lateral view only. I am now able to give some details in
respect to the mouth, as well as concerning other points in
its structure, from specimens obtained from the Welsh
coast.

It is singular that so remarkable an organism does not
appear to have been observed, except by its original dis-
coverers, until found by me in 1878. My specimens were
obtained from the sides of a bottle of sea-water, in which
I was keeping some Hydrozoa and Polyzoa collected
on the shore, at low water-mark, near the Little Orme’s
Head, Colwyn Bay. It may be also of some interest
to observe that the same gathering furnished fine specimens
of Haliphysema Tumanowiczii. Its non-appearance for so
long an interval may be accounted for by the fact that in
English books it is invariably stated to have been originally
found in " unknown fresh water from Berlin but having
found it in 1878 in sea-water from Colwyn, and subsequently
in sea-water from Tenby, either the Berlin habitat must be
erroneous, or the creature lives indilferently in marine and
fresh water, the former being the likeliest hypothesis. If
vessels containing marine Algae and other organisms, which
have remained undisturbed for a few weeks in the autumn
of each year, were carefully examined, both Lieherkuhnia
and Shepheardella would probably be found to be not un-
common British species.

In general contour Lieherkuhnia is ovate or rounded, but
the outline is constantly slowly changing by the movement
of the enclosed sarcode. The actual change in shape is not
very great, and consists principally in the gradual dilation of
one part or other of its circumference, causing it to vary
from an elongated oval or pyriform to a subglobular form.
The milky- white, semitransparent, coarsely granular proto-
plasm, of which the body consists, is invested by a thin pel-
lucid membranous integument, the interior surface of which is
lined by a transparent " subcutaneous layer ” of finely granu-
lated protoplasm. Its exterior surface is set with a number of
highly refractive, short, rod-like spicules, disposed at various
angles upon it, like those described by Greeff in Heliophrys
variahilis. These spicules are indistinguishable in the living or-
ganism,being covered by the coat of motile protoplasm, but are
brought into view when this is absent, as in dead specimens.
The investment is sufficiently flexible to follow closely all
* Geneva, 1850—1861.

ON LIEBERKUHNIA WAGENERI. 143

the changes in contour of the sarcode mass, and therefore
not immediately apparent in the living animal, but is occa-
sionally exhibited very plainly by the contraction of the
sarcode within it (PI. XVI, fig. 8). The sarcode rotates con-
stantly within this integument, and emits at the same time
from a terminal orifice a main stem of protoplasm, from
which the freely inosculating and branching pseudopodia
principally arise. Pseudopodia are also numerously put
forth by the coat of living sarcode which, spreading from the
main aperture, completely invests the integument. No dis-
tinct nucleus, such as that of Shephear della, can be detected
in the body, but there are several clear spaces, as shown in
fig. 8. These are very distinct, as they are carried along by
the rotating protoplasm, and are persistent in form and size,
partaking more of the nature of vacuoles than contractile
vesicles. They do not appear to possess any wall, and are
not seen in a mounted specimen. In addition to these
vacuoles, a large number of vesicular nuclei” are dispersed
throughout the sarcode, consisting apparently of a simple
cell enclosing a clear fluid and one or more nucleoli. These
are especially distinct in a specimen mounted in glycerine,
a camera tracing of a portion of which is given on PI. XVI,
fig. "Sy showing in optical section the integument with its
spicula {a), the vesicular nuclei [b), the granular protoplasm
(c), and the subcutaneous layer” {d).

The vesicular nuclei” are apparently identical with the
bodies so designated by Prof. Ray Lankester in his descrip-
tion of the sarcode of Haliphysema Tumanowiczii} They are
probably not quite so large in Lieherkuhnia, in which their
average diameter is about of an inch (0*013 mm.),

the size of those present in Haliphysema being about
of an inch. They seem to be equally abundant in both
cases, but there appears to be some difference in the manner
of their distribution throughout the sarcode. Prof. Lan-
kester having observed that those present in Haliphysema
were “ scattered in the protoplasm, .... being most abun-
dant in the basal portion of the core,” whereas in Lieher-
kuhnia they are principally found resting on the outer surface
of the internal protoplasm, just within the subcutaneous
layer. I have not been able to detect as such any of the
‘‘ egg-like bodies” of the same author, but as the only speci-
men I at present have is mounted entire in glycerine jelly
without previous treatment by reagents, they may possibly
be there, although indistinguishable from the nuclei.

This apparent identity of the structure of the sarcode of
1 ‘ Quarterly Journal of Microscopical Science,’ October, 1879,

144

J. O. SIDDALL.

Lieherkiihnia with that of Haliphysema appears to me inter-
estiug, as corroborative of the conclusions suggested by Prof.
Lankester’s paper, viz. that Haliphysema is correctly placed
among arenaceous” Foraminifera. The affinity existing
between these two forms is further demonstrated by the
scattered spiculiform bodies with wffiich the integument of
Lieherhuhnia is invested.

The orifice or mouth of Lieberkuhnia, as seen when the
creature was crawling mouth upwards on the thin glass
cover of a cell, is situated at its broader extremity in a
depression formed by the infolding of the four lobes of
the integument, taking the contour of a square whose
angles are carried outwards into projecting points (PL XVI,
fig. 9). The sarcode stolon is shown in the sketch as a
small oval spot in the centre of the square — the pseudopodia
which arise from it having been omitted to avoid confusion.

To the life history of LieherkhUnia I can add little to
what has been already recorded, but the following may not
be devoid of interest as illustrating its voracity and its
general appearance whilst feeding.

On the 5th of October last I found a very fine specimen
anchored to the side of a bottle, and transferred it to a cell by
means of a camel-hair pencil. When examined under the
microscope it was found to be in the act of trying to swallow
a living worm quite twice as long as itself, which it had
evidently seized whilst in the bottle, retaining its hold
upon it even when taken up by the brush for transfer to
the cell. The means adopted to accomplish deglutition
were somewhat singular. Having got one end into its
mouth, the Lieberkuhnia passed a thick stream of sarcode
along the worm to its furthest extremity, the creature mean-
while making feeble efforts to escape from the glairy mass.
Then, having got well hold of its prey, and anchored itself to
the glass by an outspread network of pseudopodia, it pro-
ceeded slowly and deliberately to envelop it, partly by re-
tracting the thick band of sarcode, and partly by advancing
its own body. This process continued until the pointed end of
the worm was pressed so tightly against the lower extremity of
the Lieberkuhnia as to almost burst the integument, when,
finding it could engulf it no further, more sarcode flowed from
the interior of the Rhizopod over the worm, until it was
completely enclosed (PI. XVI, figs. 10 and 11). Matters re-
mained thus for some time, the rotation of the sarcode going
on with unusual rapidity meanwhile, and the pseudopodia
being spasmodically emitted and retracted in all directions.
After a couple of hours, by a process the reverse of swallow-

ON LIEBERKUHNIA WAGENERl. 145

iiig, the worm was slowly ejected, quite dead, and much
changed in appearance, a large proportion of its contents
having apparently been sucked out. For some hours sub-
sequently the Rhizopod displayed remarkable activity, the
pseudopodial filaments stretching out on all sides until they
eventually entirely filled the field of a 3 in. objective, the
centre being occupied by the creature itself. The movement
of the sarcode to and fro both in its body and in every rami-
fication of the pseudopodia was much increased in rapidity
by the unwonted feast, and it was not until six or seven hours
had elapsed that the rapidity and force of the streams were
sensibly diminished, after which the animal gradually as-
sumed an almost dormant condition.

In conclusion, I may just relate that in transferring this
specimen from the bottle to the cell, I had some difficulty in
detaching it from the pencil point, and in doing so accident-
ally separated two small portions of the naked external pro-
toplasm. These detached pieces began as soon as they were
settled in the cell to put out pseudopodia in all directions,
until at last no central mass was left in either. Each had
become an interlacing network of pseudopodia. The parent
Lieherkuhnia at the same time began to put out its pseudo-
podia, which ultimately reached those extended from the
detached portions, and the moment actual contact took place
they coalesced, the fragments again becoming part of the
original body — thus repeating what had been previously
observed and followed a stage ffirther in Shephear della.

Chester, March, 1880.

146

ADAM SEDGWICK.

Development of the Kidney in its relation to Wolffian
Body in the Chick. By Adam Sedgwick, B.A., Scholar
of Trinity College, Cambridge ; Demonstrator in the
Morphological Laboratory. (With Plates XVII,
XVIII.)

This paper contains an account of observations on the
development of the excretory system of the chick, made with
a view of elucidating the relation 'which the kidney bears to
the Wolffian body.

The Wolffian body in the embryo chick attains to a very
great development, but almost completely atrophies in the
adult, a small part only persisting in the male as part of the
testicular apparatus.

In the embryos of lower Vertebrates, viz. most of the
Icthyopsida, there is present, similarly, an organ called the
Wolffian body, which, however, much more completely
persists in the adult, functioning in part as kidney and in
part as semen carrier.

The separation into an urinary part and into a sexual part
is much more complete in some forms than in others. In
the Elasmobranchii, for instance, the posterior part of the
embryonic Wolffian body gives rise in the adult to a well-
developed gland, the kidney, while the anterior part attains
a far less development ; in fact, more or less retrogrades in
the adult ; but in the male a part of it enters into connection
with the testis.

In the Amniota the case is different. In them an embryonic
organ, called the Wolffian body, does not function at all in
the adult as an excretory organ ; it almost completely atro-
phies from its embryonic perfection, only a small part per-
sisting in the adult male as the epidydimis. The organ
which functions as kidney in the adult arises at a relatively
late stage, and is not apparently, as in Elasmobranchs, a
modified part of the hind end of the embryonic Wolffian
body. What, then, is the kidney of the Amniota ? Is it an
organ which has arisen de novo in the Amniota, or can it, by
a more accurate study of its development, be traced into
relation with the embryonic excretory system ? In other
words, can any evidence be obtained by the study of develop-
ment proving that the kidney of the chick phylogenetically
has been modified from part of the same primitive organ as
that from which the Wolffian body developed, as is the case
in the Icthyopsida ?

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK, 147

To obtain an answer to these questions I have been
obliged to make a close study of the earliest stages in the
development of the kidney and Wolffian body. The results
obtained with regard to the latter are so different from
those obtained by the latest observers, that I have recorded
them in full in the following account.

Peculiarities in the early development of the Avian
Woffian body necessitated an examination of the early de-
velopment of the Wolffian tubules in other Vertebrates.
This examination I was enabled to make in the case of
Elasmobranchii owing to the great kindness of Mr. Balfour,
who placed at my disposal the whole of his Elasmobranch sec-
tions. The result of this examination was to convince me that
the account given of the earliest stages in the development
of the Elasmobranch Wolffian body is in some respects
erroneous.

Before proceeding to an account of the observations
made upon these heads it will be well to give a short
historical account of the progress of our knowledge on this
subject, i.e. the development of the Wolffian body and
kidney.

The later views as to the homologies of the parts of the
excretory system found in the different members of the
Vertebrate group dates from the work of Balfour^ and
Semper^ on the embryology of Elasmobranchs.

The independent discoveries of these two investigators
gave an impulse to the study of the development of the
organs in question in other animals, and as a result it has
gradually become clearer as the embryology of more animals
became known that a great similarity in the development of
these organs characterised the Vertebrata.

The earlier observers, Remak^ and Rathke,^ maintained
that the tubules of the Wolffian body developed indepen-
dently of the Wolffian duct in a blastema of mesoblast cells
adjoining the inner side of the duct.

Waldeyer, in his well-known work,^ asserted, from his
observations, that the tubules of the Wolffian body developed
as outgrowths from the duct, and that the Malpighian bodies
arose independently in the adjoining mesoblast. The views
of other observers, before 1874, were identical with one or
the other of these.

* ‘ Monograph on the Development of Elasmobranch Fishes.’

’ ‘ Urogenitalsystem der Plagiostomen Arbeiten/ vol. ii.

’ * Entwickelung der Wirbelthiere,’ &c.

* ‘ Entwickelungsgeschichte der Wirbelthiere,’ Leipzig, 18G1.

‘ ‘ Eierstock und Ei,’ 1870.

148

ADAM SEDGWICK.

Since 1874 the work of Gotte^ and SpengeP on Am-
phibia, Kolliker^ on Aves, BrauiA on Lacertilia, and Fiiv-
bringer on Teleostei, Amphibia, and Aves^ has shewn that
the excretory system of all these animals is developed on a
type seen in its simplest form in Elasmobranchs.

Kolliker first discovered in Aves structures composed of
strings of cells connected with the Wolffian duct and peri-
toneal epithelium, and placed just ventral and internal to the
former. These he compared to the early segmental tubes
described in Elasmobranchs. From the similarity of these
structures to those seen in Elasmobranchs and from his
own observations he was led to assert for them a deve-
lopment similar to that described for Elasmobranchs,
viz. from segmental involutions of the body-cavity epi-
thelium.

In this he was followed by Fiirbringer,® except in a small
detail, the latter observer denying that these cell strings had
any lumen opening into the body-cavity.

So far as 1 know, no ideas as to the morphological
meaning of the Amniote kidney was held before 1874.

Conflicting statements were then put forward by different
observers with regard to the actual embryonic development.
Remak^ and Kolliker® maintained that the whole of the
epithelium of the kidney tubules, including that of the
collecting and secreting tubules ^ and the Malpighian
bodies, was derived from a simple outgrowth from the
ureter.

The condensed mesoblast tissue which lies near the ureter
and its offshoots, in their opinion, only gives rise to the
connective and vascular elements of the kidney.

Kolliker has expressed this view in the second edition of
his great work on the development of Vertebrates. Lbwe^
has also recently arrived at the same conclusion from his
observations on Mammals.

^ ‘ Entwickelungsgesch. d. Unke.’

2 “ Das Urogen. system d. Arnphibien,” ‘ Arb. a. d. Zool. Inst.’
Wurzburg, Bd. 3, 1876.

•'* ‘ Ent. gesch. d. Menschen u. d. hoheren Thiere.’

^ “ Das Urogen-system d. d. Einheimischen Reptilien,’ Semper’s ‘ Ar-
beiten,’ Bd. 4.

^ “ Zur vergleichenden Anat. u. Entwickelungsgeschichte d. Excre-
tionsorgane der Vertebraten,” ‘ Morphol. Jahrbuch,’ Bd. 4. The reader
is referred to this admirable essay for the literature, and a complete ac-
count of our knowledge of the excretory organs of Vertebrates.

Loc. cit.

‘ Entwickelung der Wirbelthiere.’

8 Loc. cit.

® ‘ Centralblatt fur die Med. Wissenschaften,’ Oct., 1879.

KIDNEY IN RELATION TO VYOLFIAN BODY' IN THE CHICK. 149

Kupfter,^ Bornhaupt/ and Braun/ on the other hand^
assert that the secretory tubules and Malpighian bodies are
formed independently of the ureter in the condensed meso-
blast tissue mentioned above^ the outgrowths from the ureter
merely giving rise to the collecting tubules.

I shall return, when I have described the kidney develop-
ment in the chick, to a consideration and discussion of
the various hypotheses which have been held concerning the
Amniote kidney.

Development of Wolffian body, — The ages of the younger
embryos from Avhich the sections figured in the accompany-
ing plates (XVII and XVIII) were taken are indicated by
the number of protovertebrse. In the older embryos this
was not possible. In most cases the place in the body,
from which a section figured was taken, is indicated by the
number of the segment^ in which it occurred, counting
the first segment behind the auditory involutions as the
first.

These determinations have been made with some care by
mounting all the sections in order, and then by observing the
protovertebrse, arranging them into groups corresponding to
each protovertebra, beginning the process always in front.

The observations here recorded do not extend to any part
of the Wolffian body in front of the fourteenth segment, nor
to the development of the Wolffian duct. I have made
some observations on both these parts, but they are not yet
sufficiently complete to enable me to understand certain
remarkable appearances in their development. The Wolf-
fian body, like most other organs, develops first of all in
front and then gradually backwards, so that supposing the
development behind were the same as in front, the process
might be shewn by a series of sections from a single chick
of the proper age. But this is not the case. In the chick
the development of the Wolffian tubules behind is very
different to that in front. This fact has apparently been
overlooked by the most recent observers.

The development of the Wolffian body in the duck is
much more completely similar throughout than in the chick,
and reference will be at first made to figs. ^ — 5, taken from
a duck embryo with thirty-one or thirty-two proto vertebra3,
in the following description.

^ ‘ Arch. f. Mic. Anat.,’ Bd. 1.

‘ Untersuchungen iiber d. Entwick. des Urogen. systems beiin
liuhncben,’ Diss. Inaug., liiga, 1867.

^ Loc. cit.

^ The term segment is used as equivalent to protovertebra, muscle
plate.

150

ADAM SEDGWICK.

The tubules of the Wolffian body do not develop from
serial involutions of the peritoneal epithelium, but from the
cells of the intermediate cell mass. The intermediate cell
mass is so familiar to all students of Avian embryology that

Fig. 1.’ — Section from a chick of the second day, showing the inter-
mediate cell-mass. M. c. Medullary canal. P.v. Protovertebra.
}F. d. Wolffian duct. p.p. Body-cavity, c. h. Notochord, a. o. aorta.

1 For the woodcuts, figs. 1 and 2, I have to thank Mr. Balfour. Fig. 1
is from the * Elements of Embryology.’

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK. 151

it is hardly necessary to say anything about it. It is a
mass of cells^ (woodcut, fig. 1), stretching between the proto-
vertebra (P. V.) and the dorsal inner angle of the body-
cavity (P’JPO- formation continuous with

the peritoneal epithelium. Its relation to the protovertebra
is obscure, and I have been unable, so far, to make it out
satisfactorily. There is one point, however, to be borne
in mind concerning this intermediate cell mass ; it is con-
tinuous, t. e. is not divided by the lines of segmentation
into areas corresponding to each protovertebra.

Very soon after the intermediate cell mass is established
it undergoes a change. It becomes at some points more dis-
tinctly continuous with the peritoneal epithelium, at others
less so. And finally this culminates in a clear continuity, as
seen in fig. 2, and a marked discontinuity, as seen in fig. 3.
In fig. 2 we have what practically amounts to a continuation
of the body-cavity into the intermediate cell mass (^ c m.) ;
in fig. 3, on the other hand, the intermediate cell mass is
distinctly disconnected with the peritoneal epithelium, and
lies as a mass of cells between it and the protovertebra. Al-
though these figures are not taken from contiguous sections,
fig. 2 being taken from the thirtieth segment, and fig. 3 from
the twenty-ninth, yet for all the important details fig. 3
represents exactly a section next or next but one to fig. 2.
The intermediate cell mass is then now present as a cord of
cells continuous from segment to segment, and continuous
at intervals with the peritoneal epithelium. Fig. 4 repre-
sents the two conditions of the intermediate cell mass, as
seen in a single section taken from the twenty-sixth seg-
ment.

At the next stage of development the intermediate cell
mass entirely breaks away from the peritoneal epithelium,
and lies as a cellular blastema just internal to the Wolffian
duct. It may be called the Wolffian blastema.

The Wolffian blastema almost directly breaks up into the .
structures constituting the first rudiments of the Wolffian
tubules (fig. 5).

The development of the Wolffian blastema in the chick
needs further description.

In the anterior region of the Wolffian body, as far back as
the nineteenth or twentieth segments, the above description
of the conversion of the intermediate cell mass into the

^ The term intermediate cell mass in this account is only used to in-
dicate the cell mass connecting a 'protovertebra with the peritoneal epi-
thelium, and never refers to the cell mass occupying the same position
before segmentation has given rise to protovertebrso.

VOL. XX. NEW SER.

L

152 ADAM SEDGWICK.

Wolffian blastema applies {vide fig. 1) ; but behind this
region the development is different.

Here^ even before the segments are formed, it is found
that the cell mass, which on segmentation gives rise to the
intermediate cell mass, is distinctly separate from the thick
epithelium of the body-cavity, but attached to the cell mass,
which will give rise to the protovertebra (fig. 6, taken
from a chick with twenty-six protovertebrse behind the last
protovertebra) .

And this separation is apparently retained through later
development. The cell mass {i c m' ^ fig. 6), in the next stage
(fig. 7, taken from twenty-ninth segment of a chick with
twenty-nine protovertebree) is obviously the Wolffian blas-
tema which, in a still later stage (fig. 10, from the twenty-
ninth segment of a chick with thirty-four), gives rise to the
commencing Wolffian tubule.

The same fact may be seen by comparing figs. 8 and 9,
fig. 8 being taken from the twenty-fourth segment of a chick
with twenty-six, and fig. 9 from the twenty-fourth segment
of a chick with twenty-nine.

It will be observed, by inspection of figs. 6, 7, 8, that the
peritoneal epithelium which adjoins the Wolffian blastema is
thick, as it is elsewhere ; while later (figs. 9, 10) it is in
the same spot thin, as it is, or will be, in most other parts
of the body-cavity. '

No doubt this thick epithelium does, in the process of
becoming thin, bud off cells, which travel inwards, and some
of which may help to form the definite Wolffian blastema
(figs. 9, 10). But this process takes place everywhere. In
fact, at an early stage of development almost all the meso-
blast cells are represented by the thick lining of the body-
cavity ; and it is by a process of growth inwards of cells from
this that most of the connective tissue, &c., of the wall of the
body and gut is derived ; and, therefore, from the analogy of
the fate of the cells growing out from the body-cavity wall
in other places, we might fairly assume that those growing
out from that particular part adjoining the intermediate cell
mass or Wolffian blastema give rise, not to the Wolffian
tubules, but to the connective tissue and blood-vessels of the
Wolffian body. This is rendered highly probable from a
consideration of some observations I have made on the de-
velopment of the segmental tubes in Elasmobranchs, which
point to the conclusion that the epithelial lining is derived
from the cells of the intermediate cell mass. However that
may be, of one fact there can be no doubt, viz. the cells from
which the Wolffian tubules develop are not derived from

KIDNEY IN RELATION TO WOLLFIAN BODY IN THE CHICK. 153

serial ingrowths of the body-cavity epithelium, for this thin-
ning of the peritoneal epithelium, adjoining the region where
they will appear, is continuous, i. e. cells must grow in along
a line extending the whole length of that part of the body-
cavity which the Wolffian body adjoins.

The development of the Wolffian blastema, so described, is
continued as far hack as the » opening of the Wolffian duct
into the cloaca, which occurs in the thirty-fourth segment.
In fig. 12 it may be seen in the thirty-second segment (kh).

The Wolffian blastema of the chick then develops in two
slightly different ways.

In the anterior part, about as far back as the twentieth
(fig. 1) segment, that process of development which has been
described at length in the case of the duck (in which animal
it is apparently the only method of development) is passed
through in the chick.

Posteriorly from the twentieth segment the intermediate
cell mass has never any connection with the peritoneal epi-
thelium, and gives rise to the Wolffian blastema quite inde-
pendently of the peritoneal epithelium. This latter process
is clearly an abbreviation of that which takes place through-
out in the duck and in the anterior part of the chick.

I have mentioned the twentieth segment as about the limit
between the two. I cannot fix the exact limit.

It has been stated above that in the case of the duck and
the anterior part of the chick (figs. 1, 2, 4) the intermediate
cell mass becomes, at certain points, very markedly con-
tinuous with the peritoneal epithelium, and appears to en-
close a prolongation of the body-cavity (fig. 1 and fig. 2).
Such connections are undoubtedly rudiments of the nephro-
stomata seen in other Vertebrates. They do not occur seg-
mentally, being situated as often as not between the
protovertebrse.

Rudiments of these rudimentary nephrostomata occur in
the posterior part of the chick’s Wolffian body ; that is,
although a fairly sharp line can always be drawn between
the Wolffian blastema and the peritoneal epithelium, yet
the cells of the latter, at certain points, arrange themselves
just as they do in front, where the line of the body-cavity is
continued into the intermediate cell mass. These latter
rudiments are very obscure, and I have been unable to make
any satisfactory determination of their number. They may
be due merely to an accidental arrangement of the cells,
which might occur in consequence of the bend in the peri-
toneal epithelium at this point. Whether the rudimentary
nephrostomata in the anterior part of the chick’s kidney

154

ADAM SEDGWICK.

(fig. 1^) are continued as small channels in the intermediate
cell mass or Wolffian blastema, and remain, enlarging when
the Wolffian tubules are formed later, I have been unable to
ascertain. Neither have I been able to satisfy myself as to
another interesting point, viz. Do those rudimentary nephro-
stomata correspond to the Wolffian tubules subsequently
developed? In the chick’s twentieth segment never more
than three, at the most, nephrostomata can be made out,
yet there are four or five primary Wolffian tubules later.

Has this increase been caused by the development of more
tubules than there were nephrosomata, e, by intercalation,
or has it been caused by a change in the relation of the parts
to one another, due either to an elongation of the proto-
vertebra or to the travelling forward of the tubules as they
are developed behind ?

I shall return to the consideration of this point in a future
paper.

The mode in which the Wolffian blastema behind the
twentieth^ segment breaks up into tubules, so far as I have
been able to ascertain it, is the following : — A number of
vesicles, either oval or circular, lined by columnar cells, and
lying just internal to the Wolffian duct, make their appear-
ance (fig. 5). In longitudinal sections it may be seen that
these vesicles closely adjoin one another antero-posteriorly.
They are developed from those cells of the Wolffian blastema
immediately adjoining the inner border of the Wolffian duct.
By a study of transverse sections it appears that each vesicle
is continuous ventrally with that part of the Wolffian blas-
tema which has not undergone conversion into the walls of
the vesicle, and which dies just internal to the vesicle. The
cells of this part of the Wolffian blastema very soon arrange
themselves round what appears as a continuation of the
original vesicle (fig. II). From the inner and dorsal wall
of the last-formed structure a glomerulus is ultimately de-
veloped. The whole structure grows enormously, and gives
rise to the Malpighian body and complicated coils of the
later Wolffian tubule. At about the stage of development
represented in fig. 11 the tubules acquire an opening into
the Wolffian duct.

The question as to whether or no there are outgrowths
from the Wolffian duct to meet the independently developed

^ Vide also fig. 125, of Kolliker’s ‘ Entwick. gesch. der Menschen u.
der. h. Thiere.’

^ I reserve an account of the development of the tubules in front of the
twentieth segment, as my observations on this point are not yet suffi-
ciently complete to enable me to speak with certainty.

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK. 155

Wolffian tubules is not easy to answer. I leave it open now,
but hope to be in a position to give a definite answer soon.
Now I will merely state that there are appearances in my
sections which incline me to the opinion that there are out-
growths from the Wolffian duct which, in the case of the
primary Wolffian tubule, are solid, but hollow in the case of
the secondary and tertiary tubules.

The above description of the development of the primary
Wolffian tubules difiers from the most recent account of
Kdlliker^ and Fiirbringer.^ I have stated above the views
which these distinguished observers hold as to the develop-
ment. They have described perfectly correctly one stage in
the development of the anterior part of the Wolffian body.
I have often seen the appearances given by Kolliker in fig.
l2o of his work, and have given myself a similar represen-
tation (fig. 1). But if I understand them correctly they
have given an erroneous account of. the earlier development
of these structures. Fiirbringer says of them (p. 67) : “ Sie
finden sich in reihenweiser Anordnung als solide Urnieren-
strange die von dem parietalen Peritoneum ventral und
medial vomWolff’schen Gange ausgehen. . . . Sehr bald losen
sich diese Urnierenstrange von dem parietalen Peritoneum
ab undliegennun als rundliche solide Zellenmassen retrope-
ritoneal neben dem Wolffischen Gange, ein Stadium das die
Beobachtungen der mersten Autoren deckt.” It is easy to
see how Fiirbringer has been misled. He has seen in trans-
verse sections in a fairly young chick the S-shaped strings
of cells (solide Urnierenstrange) in connection with the peri-
toneal epithelium. He has also seen in an older chick the
Wolffian blastema of the posterior region of the Wolffian
body (rundliche solide Zellenmassen). Both these observa-
tions I can entirely confirm. But apparently he has not
examined the condition of these structures at an earlier
stage, assuming that they originate as solid outgrowths of
the peritoneal epithelium. This assumption my observa-
tions prove to be unwarranted.

The older observers (see above) were quite correct in their
statements of the origin of the Wolffian tubules, as struc-
tures developed in the intermediate cell mass, independently
of the Wolffian duct, and later acquiring an opening into it,

AValdeyer’s^ statement that the Malpighian body thus
develops, the rest of the Wolffian tubule developing as out-
growths from the Wolffian duct, is in my opinion erroneous.
For if there be an outgrowth from the Wolffian duct it
does not give rise to the whole tubule, exclusive of the
* hoc. cit. 2 ‘ Morpli. Jabrbucb,’ B<1, 4, ’ Loc. cit.

156

ADAM SEDGWICK.

Malpighian body, the structure developed independently
in the intermediate cell mass certainly giving rise to more
than the Malpighian body.

I now pass to the development of the secondary tubules,
&c. Fiirbringer^ derives them also from peidtoneal in-
growths. He has not, however, given, so far as I know, any
figures showing this. I have examined this point with
some care, but have quite failed to discover any traces of
these secondary ingrowths.

The secondary tubules appear to me to arise from a small
mass of cells. They occupy, at a slightly earlier stage, the
position of in fig. 11. In this figure the vesicular
rudiment of a secondary tubule has appeared in this mass of
cells. The tertiary and quaternary, &c., tubules appear to
arise successively at a slightly later stage from similar
small masses of cells, which are always placed just dorsal to
the last-formed tubule. The later development of these
secondary, tertiary, &c., Wolffian tubules is very similar to
that of the primary.

The time of development of the primary tubules relatively
to that of the secondary, &c., tubules varies in different parts.
Anteriorly the primary tubule is much more developed (in
fact, has acquired an opening into the Wolffian duct, not
figured in fig. 11), before the first trace of the secondary
tubule arises (fig. 11) ; while posteriorly a secondary and ter-
tiary tubule have appeared almost before the primary tubule
has lost its vesicular structure (fig. 13).

The development of the secondary, &c., Wolffian tubules in
the chick appears to be very much abbreviated.

Whatever may have been their development in phylogeny,
no light is thrown upon it by their ontogeny. Nor even can
a comparison be made between their development in the
chick, and that in other forms in which it is possible to
suppose the development is less abbreviated. In Elasmo-
branchs the secondary tubules, as Balfour^ has shewn,
develop in connection with the Malpighian bodies of the
primary tubules, as outgrowths from them, which eventually
open into the collecting tubules of the segment in front.
Neither Balfour nor, as far as I know, any other observer,
have elucidated the development of the tertiary, &c., tubules
in Elasmobranchs.

In the Salamander Furbringer^ has shewn that they
develop as they do in the chick from cell masses closely

* Loc. cit.

’ Balfour, < Elasmobranch Fishes/

Loc, cit.

KIDNEY IN RELATION TO WOL?riAN BODY IN THE CHICK. 157

adjoining the primary tubules, and from an inspection of his
ficrures it is evident that these cell masses are situated close to
the Malpighian body of the primary tubule. In the chick I
have sought in vain for some clear sign in the development
of these cells which would enable a comparison to be insti-
tuted with Elasmobranch development.

In the chick the cells of the Wolffian blastema do not all
aeem to be used in the formation of the primary tubule.
Those, that are not, seem to collect at a special point, e.
just dorsal to that part of the primary tubule which will
become eventually the Malpighian body. The cells of the
primary tubule are especially thick at this point, and per-
haps they give rise to some of the cells for the secondary
tubule (fig. 11). Even if this is the case — and we may look
upon the secondary tubule as an outgrowth from the primary
tubule — it is impossible to say where the secondary tubule
so formed opens into the Wolffian duct.

This brings me to another difference between the dorsal
tubules of the chick and those of Elasmobranchs. In the latter
they all open into the collecting part of a primary Wolffian
tubule. In the former they all open independently into the
Wolffian duct, or it may be into an outgrowth from it, but
separately from the primary tubule. This latter point I am,
as above stated, obliged to leave open at present. The
number of primary tubules present in one segment seem to
be fairly constant, five or six to each segment, throughout
the Wolffian body, except quite in front, where there seem
to be fewer. All segments, from the twentieth to the
thirtieth inclusive, contain five or six primary tubules. In
front of the twentieth segment they seem gradually to de-
crease. In the first segment in -which a fully developed
tubule appears there seems only to be one, the number
increasing rapidly to the twentieth.

The dorsal tubules appear in greater number behind
than in front. In the twenty-eighth segment I have
counted as many as four, but more are possibly developed
later. They correspond in number to the primary tubules,
i. e. if there are five primary tubules in the twenty-eighth
segment, there are twenty secondary tubules (five sets of
four). The most anterior segment in wffiich a secondary
tubule appears is the twenty-first or twenty-second; I have
not been able, however, to localise it exactly.

Develojwient of kidney. — The development of the Wolffian
blastema from the intermediate cell mass has been described
as far back as the thirty-fourth segment ; i. e. to the opening
of the Wolffian duct into the cloaca; it is never seen

158

ADAM SEDGWICK.

behind this point. But it does not all undergo the above
development into Wolffian tubules. It breaks up into
Wolffian tubules as far back as the thirtieth segment. Behind
this pointy ^. e. from the thirty-first to the thirty-fourth
segments inclusively, the Wolffian blastema undergoes quite
a different fate. It remains for some time almost quite pas-
sive and ultimately gives rise to the epithelium of the per-
manent kidneys. In consequence of this I have called that
part of the Wolffian blastema between the thirty-first and
thirty- fourth segments the kidney blastema ; and in future
shall refer to it by that name (figs. 12, 15, 16, 17, hh). It
is important to notice that this kidney blastema develops
in an exactly similar manner to the Wolffian blastema.

It is not until well into the fourth day, when the ureter
has appeared, that it is possible to draw the line between
the two.

Fig. 1^ is taken from the thirty-second segment of a chick
with thirty-four protovertebrae ; it shows a blastema of cells
lying just internal to the Wolffian duct. Fig. 10 is taken
from the tw’enty-ninth segment of the same chick. It shows
the hindermost trace of a Wolffian tubule 1 could find at
this stage. In all the sections between figs. 10 and 12 there
is present, just as in fig. 12, a blastema of cells lying just
internal to the Wolffian duct.

In a slightly older embryo the hindermost trace of a
Wolffian tubule would be in the thirtieth segment. In still
older embryos secondary tubules would have appeared in the
thirtieth segment, but no trace of a primary tubule in the
thirty-first, and so on in later stages, Wolffian tubules never
appearing in the thirty-first segment. In the embryo, from
which figs. 13 to 17 were taken, the ureter had not appeared.
In examining a series of sections from the posterior part of
this embryo, some of which are figured (figs. 13 to 17), the
following points are noticeable, illustrating what has just
been stated.

A primary and secondary tubule are present in fig. 14, and
it is almost the last section in which any trace of a Wolffian
tubule can be seen (the two above tubules are cut in the
next two sections). The tubules adjacent anteriorly to these
are three in number (fig. 13), consisting of primary, secon-
dary, and tertiary. Supposing the Wolffian body were going
on developing in the region behind that from which fig. 14
was taken, we ought at the least to find at this stage primary
tubules in that region, for the formation of primary and
secondary tubules is always separated by an interval of time.
But no such primary tubules are seen. Behind fig. 14 (figs.

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK. 169

15 to IT) a blastema of cells is still present, precisely
similar in its appearance and position with regard to the
Wolffian duct to the Wolffian blastema seen anteriorly at
earlier stages, and to the blastema seen in the same region
at later stages (fig. 12). And this can be traced back to the
opening of the Wolffian duct into the cloaca (fig. 17).

To this blastema of cells I have given the name kidney
blastema ; it is at this stage perfectly continuous anteriorly
with the hinder Wolffian tubules, the junction between the
two lying in one of the two sections intervening between
figs. 14 and 15. I ought rather to say the line of future
separation, for so far they have been always continuous,
having developed so. The continuity between the kidney
blastema {kb) and the hinder part of the Wolffian body
may be seen in fig. 21, which is taken from a chick of nearly
the same age as that from which figs. 13 — IT were.

In this figure the section has passed through the hind end
of the Wolffian duct and through the kidney blastema, and
has just shaved the hinder end of the Wolffian body, in conse-
quence of which the hinder Wolffian tubules are only indis -
tinctly visible.

The next change to notice is caused by the appearance of
the ureter. It arises as a growth forward from the dorsal
border of the enlarged end of the Wolffian duct. This has
been generally recognised since Kupffer’s^ account. The dila-
tation of the hind end of the Wolffian duct occurs in a very
slightly later embryo than that from which fig. 21 was taken.

The kidney blastema is now found not ventrally close to
the body-cavity, but lies dorsal to its former position, just
internal to the dorsal extremity of the dilated Wolffian duct
(fig. 20). From this dilatation there grows forward a duct,
the ureter (fig. 19), on the inner side of which lies the kidney
blastema.

The ureter, at this stage, has not a very great extent, and
is only seen for a few more sections ; in fig. 18, still behind
the Wolffian body, the ureter is not visible; but the kidney
blastema occupies a dorsal position, as it did in the posterior
section in which the ureter was present (fig. 19).

In tracing it forward it gradually descends and becomes
continuous with the hind end of the Wolffian body.

In yet older embryos, in which the ureter is more developed
and overlaps the hind end of the Wolffian body, the kidney
blastema has quite broken off from the Wolffian body, and
invests the anterior end of the ureter, so that in a series of
transverse sections through a chick at this age we should see
» ‘ Arch. f. M. Anat.,’ Bd. 2.

160

ADAM SEDGWICK.

posteriorly, in the same section with the now complicated
Wolffian body, a dorsal mass of cells. Gradually travelling
backwards a duct would appear cut across lying in the
mass of cells ; further back still we should see no Wolffian
body, but merely a duct with a mass of cells lying just
internal to it, placed well dorsal to the Wolffian duct. This
mass of cells is the kidney blastema ; and the duct is the
ureter.

Such would be seen in a chick at the end of the fifth day.

On the sixth day the ureter grows in length, the kidney
blastema accompanying it, and enveloping its anterior
extremity.

The ureter now dilates at intervals, and the kidney blas-
tema especially collects round these dilatations. From the
latter, the number of which I have not determined, the
kidney tubules grow out. In a chick of the seventh day the
tubules are just beginning to grow out from these dilata-
tions. The two posterior tubules are, however, far more
advanced than the anterior.

The ureter is now a small duct lying just dorsal to the
Wolffian body ; except at its anterior extremity, where it
is rather more dorsal, and is completely surrounded by the
kidney blastema.

Almost immediately in front of the hind end of the ureter
a tubule is given off, which runs dorsalwards and outwards.
The kidney blastema no longer adjoins the ureter, but is
disposed round the branches of this tubule. The ureter is
continued forwards through a considerable number of sec-
tions, giving off no tubules, and unaccompanied by the kidney
blastema. It now becomes continuous with a tubule, which
has already been seen in many sections surrounded by kidney
blastema, and which, though not so much branched as the
most posterior tubule above mentioned, is more developed
than any tubule met with in front.

The ureter continues as a small duct lying just dorsal to
the Wolffian body.

In this embryo (seventh day), travelling forwards, several
dilations could be made out. The appearance presented
by such a dilatation in transverse section and its position
with regard to the Wolffian body, may be gathered from an
inspection of fig. 22.

The lateral walls of the dorsal part of the dilated ureter
are closely applied, the lumen being very indistinct.

Around the dorsal part of the dilatation the kidney
blastema is present.

In the next section, or in the next section but one, either

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK. 161

backwards or for^vards, the dorsal dilatation w^ould be no
longer visible, but occupying the position of the dorsal part
of the dilatation, would be seen a tubule surrounded by the
cells of the kidney blastema (fig. 23).

In the next section to this the walls of the tubule become
indistinctly marked off from the kidney blastema (fig. 24).
Some of the large columnar cells of the kidney tubule become
branched, the processes being continuous with the processes
of the branched cells of the kidney blastema. In fact, every
stage of cell shape between a columnar lining cell of the
tubule and a branched cell of the blastema is visible.

The lumen of the tubule is no longer distinct, it not being
possible to say what is an intercellular space and what the
lumen of the tuhule.

In the next section no trace of a tubule is visible, its
place being occupied by the cells of the blastema.

Fig. 23 is taken from a section next but one to fig. 22.
Nine such dorsal dilatations of the ureter, with commencing
tubules, growing from them, could be made out in the embryo
under consideration.

In front of the most anterior the tubules open directly into
the ureter which in this region has become more dorsal with
regard to the Wolffian body. Tubules in this region, more-
over, are given ofi* from the ventral side of the ureter, corre-
sponding almost exactly to those given off from the dorsal side.
Four pairs could be made out ; after which the ureter ended
closely surrounded on all sides hy dense kidney blastema.

The next stage, which I have closely examined, w-as in an
eight-day chick. The kidney had reached a great compli-
cation of structure. Malpighian bodies had, however^ not
distinctly appeared. The tubules were still surrounded by
kidney blastema which was especially conspicuous at their
growing ends. The appearance of the latter, which was ex-
actly similar in all essential details to the growing points of
the tubule last described, is represented in fig. 24.

Before considering the bearing of the above facts upon the
questions asked at the outset, I will recapitulate the more
important points in the development of the Avian kidney
and Wolffian body.

1. The cells which give rise to the Wolffian and kidney
tubules do not develop as involutions of the peritoneal epi-
thelium, hut from a blastema of cells derived from the inter-
mediate cell mass.

2. The blastema of the kidney is at first perfectly con-
tinuous with that of the Wolffian body, and cannot be dis-
tinguished from it.

.162

ADAM SEDGWICK.

3. Wolffian tubules do not appear in any part of the blas-
tema behind the thirtieth segment. Primary, secondary, and
tertiary, &c., tubules are developed in that part of it placed
in the thirtieth and anterior segments as far as the twenty-
first or twenty-second, and primary tubules in yet anterior
segments.

4. The blastema in the thirty-first to thirty-fourth segments,
on the appearance of the ureter, moves dorsalwards from the
Wolffian duct, breaking away from the hindermost Wolffian
tubules and enters into close relation with the ureter.

5. This part of the blastema — the kidney blastema — espe-
cially collects round swellings on the ureter, from which
kidney tubules grow out.

6. These kidney tubules burrow into the kidney blastema,
their growing points being continuous with the cells of the
blastema.

Five years ago Balfour ^ and Semper ^ independently put
forward the hypothesis that the kidney of the Amniota holds
the same relation to the embryonic Wolffian body as does the
adult kidney in Elasmobranchs.

Balfour wrote then The last feature in the anatomy of
the Selachians which requires notice is the division of the
kidney into two portions, an anterior and posterior. The
anatomical similarity between this arrangement and that of
higher Vertebrates (birds, &c.) is very striking. The anterior
one precisely corresponds, anatomically, to the Wolffian body,
and the posterior to the true permanent kidney of higher
Vertebrates ; and when we find that in the Selachians the
duct for the anterior serves also for the semen, as does the
duct of higher Vertebrates, this similarity seems almost to
amount to identity.’’

The development of the kidney of the bird has never
been fully w'orked out, so that this hypothesis, arrived at from
a consideration of the facts of comparative anatomy and Elas-
mobranch embryology, has hitherto not been tested by the
facts of Avian embryology. The observations described above
were undertaken with a view of testing this hypothesis, and,
in my opinion, it has fully stood the test.

The development of the kidney in the chick points most
decidedly to the conclusion that it is merely the posterior
part of the Wolffian body — or, perhaps, it would be better to
say, of a primitive organ, the anterior part of which is now

^ “Urogenital Organs of Vertebrates/’ ‘Journal of Anatomy and
Physiology/ vol. x,

2 Loc. cit.

3 P. 27.

Sidney in relation to wolffian body in the chick. 16S

seen as the Wolffian body — the whole of this primitive organ
in Elasmobranch embryos being termed Wolffian body.

The most important fact in favour of Balfour’s hypothesis
is the primitive continuity in early stages in the bird of the
cells from which both Wolffian body and kidney arise.

The differences in later development cannot be looked
upon as a serious difficulty when we remember the immense
differences which many undoubtedly homologous organs
show in their embryonic development in various animals.

It has been stated (see above, p. 148) by many students
of Avian embryology that the kidney tubules develop as
outgrowths from the ureter, and that the cells of the kidney
blastema merely give rise to the vascular elements of the
glomerulus. This view, whether considered a priori, or with
reference to the facts of development, cannot for a moment
be maintained.

If Balfour’s hypothesis as to the relation of the kid-
ney to the Wolffian body be accepted, and I do not see
how it can be rejected, assuming the truth of the facts
of development recorded in this account, it would re-
quire very strong proof indeed to establish the fact that
the cells of the kidney blastema give rise merely to the
vascular elements of the glomerulus, and take no part in
the formation of the secretory epithelium of the kidney
tubules, such as is taken in the formation of tubules of a
very similar organ by cells developed in a precisely similar
way. Such proof is not forthcoming, and would be very
hard to give.

Considering the very late development of the posterior
part of the Wolffian body (kidney) in the chick with refer-
ence to that of the anterior part, it surely cannot be a matter
of surprise if the development has been modified, the
walls of the tubule arising from the cells of the blastema ;
the lumen, however, not as in the anterior part, first appear-
ing as an independent cavity, which opens later into the
duct, but being from the first continuous with the lumen of
the ureter.

Fiirbringer’s suggestion, that the Amniote kidney is de-
rived from dorsal tubules of the Wolffian body, is based
mainly on the fact that it lies dorsal to the Wolffian body,
and an observation of Braun’s for Lizards. Braun has
stated for these animals that the kidney blastema develops
Irom irregular ingrowths of the peritoneal epithelium, at a
period when the secondary dorsal tubules of the Wolffian
body are developing.

With regard to the first point it is to be noted that the

164

ADAM SEDGWICK.

dorsal position of the kidney is clearly a secondary change,
appearing only late in development, and due obviously to
the great size the kidney attains. Moreover, according to
Fiirbringer’s view, one would expect to find some kind of
continuity between the developing kidney and dorsal part
of the Wolffian body ; but no trace of any such connection
can ever be seen.

Finally, in view of the facts of development here recorded
for the chick, and of those about to be mentioned for Elas-
mobranchs, Braun’s observations on the development of the
kidney blastema of Lizards from peritoneal ingrowths cannot
be accepted without further evidence. The irregularly
scattered cells lying between the Wolffian duct and the peri-
toneal epithelium, which Braun has figured, are by no means
proof of ingrowth of cells from the peritoneal epithelium.
Such an irregular arrangement of cells can be seen anywhere
adjoining the body-cavity epithelium.

Kblliker’s view that the kidney of the Amniota is an
organ sui generis, which was not present in any form in the
excretory system of the common ancestor of Icthyopsida and
Amniota, needs in my opinion no refutation ; for if true it
can only be established by proving all other hypotheses con-
cerning the kidney to be untenable.

Development of segmental tubes in Elasmohrancliii. — I
should hardly have been bold enough to publish these obser-
vations on the development of the chick’s Wolffian body,
opposed as they are to statements supported by great
authority, had I not had the opportunity of examining the
early development of the parts in question in Elasmabranchs.

I was thus enabled to confirm suspicions which I had
entertained since examining the development of the Wolffian
body of birds, as to the correctness of the description of the
earliest stages in these fishes. It is well known that the
Wolffian tubules of Elasmobranchii are derived from the seg-
mentally-arranged segmental tubes. These latter were said
to arise by an invagination, at first solid but subsequently
becoming hollow, of the peritoneal epithelium just internal
to the segmental duct into the cells of the intermediate cell
mass. Tlie intermediate cell mass was said to be produced
by the coming together of the splanchnic and somatic layers
of that part of the body cavity, which at an earlier period
existed connecting the general ventral body cavity with the
dorsal continuations of it in the muscle plates.

On examining specimens of young Elasinobranch (Scyt-
lium, Pristiurus, Torpedo) embryos, I found that the
passage connecting the general body-cavity with that in the

KIDNEY IN RELATION TO WOLFFIAN BODY IN THE CHICK. 165

muscle-plates persisted later than had been described.
Its connection with the ventral dilatation of the muscle-
plate cavity is carried ventralwards as far as the outer dorsal
corner of the segmental duct ; so that it appears as a canal
opening into the body-cavity just internal to the segmental
duct, and thence curling round its dorsal wall to open into
the muscle-plate cavity. The ventral outer wall of this
passage is formed of large columnar cells, the inner and
dorsal wall of much flatter cells (woodcut, fig. 2), as seen in
transverse sections.

Fig. 2.J — Transverse section through a young embryo of Scytlium. mjj.

Muscle-plate. s^. Segmental tube. sd. Segmental duct. c/i.

Notochord, sp. v. Alimentary canal.

At the next stage of development the passage becomes
quite separated from the muscle-plate cavity, and now lies
as a blind tube, opening into the body- cavity internal to the
segmental duct, its blind outer end being applied to the
ventral dilatation of the muscle-plate body-cavity (woodcut,
fig. 2). This blind tube is the commencement of a segmental
tube.

I have traced it through a series of older embryos to the
fully-formed segmental tube. The adjoining woodcut (fig. 2)

’ Fig. 2 is from Mr. Balfour’s forthcoming treatise on ‘ Comparative
Embryology.’ It is copied from a figure in his ‘ Monograph on the
Development of Elasmobranch Fishes.’

166

ADAM SEDGWICK.

represents a section from a Scytlium embryo, in which the
segmental tube has just broken aw^ay from the muscle plate ;
in a slightly younger embryo, or perhaps in posterior sections
of the same embryo, the cavity of the segmental tube (.9^)
would communicate wuth the ventral dilatation of the muscle
plate (mjo), at the point where they are in contact in the
figure.

This mode of origin of the segmental tubes of Elasmo-
branchii renders the origin of the same structures in the
chick less extraordinary than it at first sight seemed.

I refrain now' from the discussions and perhaps hypotheses
which this observation on the development of the Elas-
mbbranch segmental tubes suggests. On one point, however,
there can be little doubt, viz. that segmental involutions of
the peritoneal epithelium which have been described in the
Teleostei, Amphibia, Ganoids, Mammals, will have to be
given up. At any rate the current statement on this point
cannot be accepted without further proof.

I have made observations on the Teleostei and Sturgeon
(Mr.Balfour having kindly put his specimens at my disposal)
which tend to show', not, however, with certainty, that in the
embryos of these forms at any rate there are no serial invo-
lutions of the body-cavity epithelium to form the segmental
tubes. The method of development in these forms appears
to me to be very much modified, if continue to regard the
development in Elasmobranchs as primitive.

The Wolffian body in all those animals whose ova have a
relatively small food yolk seems to be retarded in develop-
ment ; while the head kidney, the relation of which to the
rest of the excretory system is not understood, attains an
early development and functions as the larval excretory
organ. Very possibly a clue to the explanation of the re-
tardation in the development of the Wolffian body and of the
morphological meaning of the head kidney, in Teleostei,
Amphibia, &c., may be obtained by a consideration of this
coincidence and others hitherto apparently overlooked.

I hope in another paper to discuss these questions and some
others which have been passed over here, and to describe the
development of the anterior part of the Wolffian body in the
chick.

In conclusion, I wish to acknowdedge the great obligations
I am under to Mr. Balfour. Not only have I to thank him
for his kindness in placing all his preparations and specimens
at my disposal, but, for what has been far more valuable,
the help and encouragement he has given me through the
whole course of this investigation.

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 167

Notes on the Development of the Araneina. By F. M.

Balfour, M.A., F.R.S., Fellow of Trinity College,

Cambridge. With Plates XIX, XX, and XXI.

The following observations do not profess to contain a
complete history of the development even of a single species
of spider. They are the result of investigations carried on
at intervals during rather more than two years, on the ova
of Angelena labyrinthica ; and I should not have published,
them now, if I had any hope of being able to complete them
before the appearance of the work I am in the course of
publishing on Comparative Embryology. It appeared to me,
however, desirable to publish in full such parts of my
observations as are completed before the appearance of my
treatise, since the account of the development of the Aran-
eina is mainly founded upon them.

My investigations on the germinal layers and organs have
been chiefly conducted by means of sections. To prepare
the embryos for sections, I employed the valuable method
first made known by Bobretsky. I hardened the embryos
in bichromate of potash, after placing them for a short time
in nearly boiling water. They were stained as a whole
with hematoxylin after the removal of the membranes, and
embedded for cutting in coagulated albumen.

The number of investigators who have studied the de-
velopment of spiders is inconsiderable. A list of them is
given at the end of the paper.

The earliest writer on the subject is Herold (No. 4) ; he
was followed after a very considerable interval of time by
Claparede (No. 3), whose memoir is illustrated by a series
of beautiful plates, and contains a very satisfactory account
of the external features of development.

Balbiani (No. 1) has gone with some detail into the
history of the early stages; and Ludwig (No. 5) has pub-
lished some very important observations on the development
of the blastoderm. Finally, Barrois (No. 2) has quite re-
cently taken up the study of the group, and has added some
valuable observations on the development of the germinal
layers.

In addition to these papers on the true spiders, important
investigations have been published by Metschnikoff on
other groups of the Arachnida, notably the scorpion.
Metschuikoff’s observations on the formation of the ger-

VOL. XX. NEW SER. M

.168

F. M. BALFOUR.

minal layers and organs accord in most points with my
own.

The development of the Araneina may be divided into
four periods : (1) the segmentation ; (2) the period from the
close of the segmentation up to the period when the segments
commence to be formed ; (3) the period from the com-
mencing formation of the segments to the development of the
full number of limbs ; (4) the subsequent stages up to the
attainment of the adult form.

In my earliest stage the segmentation was already com-
pleted, and the embryo was formed of a single layer of large
flattened cells enveloping a central mass of polygonal yolk-
segments.

Each yolk-segment is formed of a number of large clear
somewhat oval yolk-spherules. In hardened specimens the
yolk-spherules become polygonal, and in ova treated with
hot water prior to preservation are not unfrequently broken
up. Amongst the yolk-segments are placed a fair number,
of nucleated bodies of a very characteristic appearance.
Each of them is formed of (1) a large, often angular, nucleus,
filled with deeply staining bodies (nucleoli ?) . (2) Of a layer

of protoplasm surrounding the nucleus, prolonged into a
protoplasmic reticulum. The exact relation of these
nucleated bodies to the yolk-segments is not very easy to
make out, but the general tendency of my observations is to
show (1) that each nucleated body Delongs to a yolk-sphere,
and (2) that it is generally placed not at the centre, but to
one side of a yolk-sphere. If the above conclusions are
correct each complete yolk-segment is a cell, and each such
cell consists of a normal nucleus, protoplas-m, and yolk-
spherules. There is a special layer of protoplasm surrounding
the nucleus, while the remainder of the protoplasm consists
of a reticulum holding together the yolk -spherules. Yolk-
cells of this character are seen in PI. XX and XXI, fiers.
10—21.

The nuclei of the yolk-cells are probably derived by divi-
sion from the nuclei of the segmentation rosettes {vide Lud-
wig, No. 5), and it is probable that they take their origin at
the time when the superficial layer of protoplasm separates
from the yolk-columns below to form the blastoderm.

The protoplasm of the yolk-cells undergoes rapid division,
as is shewn by the fact that there are often two nucleated
bodies close together, and sometimes two nuclei in a sino^le
mass of protoplasm (fig. 10). It is probable that in some
cases the yolk-spheres divide at the same time as the proto-
plasm belonging to them; the division of the nucleated

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 169

bodies is, however, in the main destined to give rise to fresh
cells which enter the blastoderm.

I have not elucidated to my complete satisfaction the next
stage or two in the development of the embryo ; and have
not succeeded in completely reconciling the results of my
own observations with those of Claparede and Balbiani. In
order to show exactly where my difficulties lie it is necessary
briefly to state the results arrived at by the above authors.

According to Claparede the first difi“erentiation inPholcus
consists in the accumulation of the cells over a small area to
form a protuberance, which he calls the primitive cumulus.
Owing to its smaller specific gravity the part of the ovum
with the cumulus always turns upwards, like the blasto-
dermic pole of a fowPs egg.

After a short time the cumulus elongates itself on one
side, and becomes connected by a streak with a white patch,
which appears on the surface of the egg, about 90° from the
cumulus. This patch gradually enlarges, and soon covers
the whole surface of the ovum except the region where the
cumulus is placed. It becomes the ventral plate or germinal
streak of the embryo, its extremity adjoining the cumulus
is the anal extremity, and its opposite extremity the cephalic
one. The cumulus itself is placed in a depression on the
dorsal surface of the ovum. Claparede compares the cumu-
lus to the dorsal organ of many Crustacea.

Balbiani (No. 1) describes the primitive cumulus in
Tegenaria domestica, Epeira diadema, and Agelena labyrin-
thicuy as originating as a protuberance at the centre of the
ventral surface, surrounded by a specialised portion of the
blastoderm (p. 57), which I will call the ventral plate. In
Tegenaria domestica he finds that it encloses the so-called
yolk-nucleus, p. 62. By an unequal growth of the ventral
plate the primitive cumulus comes to be placed at the
cephalic pole of the ventral plate. The cumulus now
becomes less prominent, and in a few cases disappears. In
the next stage the central part of the ventral plate becomes
very prominent and forms the procephalic lobe, close to the
anterior border of which is usually placed the primitive
cumulus (p. 67). The space between the eumulus and the
procephalic lobe grows larger, so that the latter gradually
travels towards the dorsal surface and finally vanishes.
Behind the procephalic lobe the first traces of the segments
make their appearance, as three transverse bands, before a
distinct anal lobe becomes apparent.

The points ^^hich require to be cleared up are, (1) what
is the nature of the primitive cumulus? (2) where is it

170

F. M. BALFOUR.

situated in relation to the embryo? Before attempting to
answer these questions I will shortly describe the develop-
ment, so far as I have made it out, for the stages during
which the cumulus is visible.

The first change that I find in the embryo (when examined
after it has been hardened)^ is the appearance of a small,
whitish spot, which is at first very indistinct. A section
through such an ovum (PI. XX, fig. 10) shows that the cells
of about one half of the ovum have become more columnar
than those of the other half, and that there is a point {pr. c.)
near one end of the thickened half where the cells are more
columnar, and about two layers or so deep. It appears tome
probable that this point is the whitish spot visible in the hard-
ened ovum. In a somewhat later stage (PI. XIX, fig. 1) the
whitish spot becomes more conspicuous (j»c), and appears as
a distinct prominence, which is, without doubt, the primitive
cumulus, and from it there proceeds on one side a whitish
streak. The prominence,as noticed by Claparede andBalbiani,
is situated on the flatter side of the ovum. Sections at this
stage show the same features as the previous stage, except
that (1) the cells throughout are smaller, (2) those of the
thickened hemisphere of the ovum more columnar, and (3)
cumulus is formed of several rows of cells, though not
divided into distinct layers. In the next stage the appear-
ances from the surface are rather more obscure, and in some
of my best specimens a coagulum, derived from the fluid
surrounding the ovum, covers the most important part of the
blastoderm. In PI. XIX, fig. 2, I have attempted to repre-
sent, as truly as I could, the appearances presented by the
ovum. There is a well-marked whitish side of the ovum,
near one end of which is a prominence {pc), which must,
no doubt, be identified with the cumulus of the earlier stages.
Towards the opposite end, or perhaps rather nearer the
centre of the white side of the ovum, is an imperfectly
marked triangular white area. There can be no doubt that
the line connecting the cumulus with the triangular area is
the future long axis of the embryo, and the white area is,
without doubt, the procephalic lobe of Balbiani.

A section of the ovum at this stage is represented in
PI. XX, fig. 11. It is not quite certain in what direction
the section is taken, but I think it probable it is somewhat
oblique to the long axis. Ho\vever this may be, the section

^ I was unfortunately too much engaged, at the time when the eggs
were collected, to study them in the fresh condition ; a fact which has
added not a little to my difficulties in elucidating the obscure points in
the early stages.

NOTES ON THE DEVELOPMENT OP THE ARA.NEINA. 171

shows that the whitish hemisphere of the blastoderm is
formed of columnar cells, for the most part two or so
layers deep, but that there is, not very far from the middle
line, a wedge-shaped internal thickening of the blastoderm
where the cells are several rows deep. With what part
visible in surface view this thickened portion corresponds is
not clear. To my mind it most probably corresponds to the
larger white patch, in which case I have not got a section
through the terminal prominence. In the other sections of
the same embryo the wedge-shaped thickening was not so
marked, but it, nevertheless, extended through all the sections.
It appears to me probable that it constitutes a longitudinal
thickened ridge of the blastoderm. In any case, it is clear
that the white hemisphere of the blastoderm is a thickened
portion of the blastoderm, and that the thickening is in part
due to the cells being more columnar, and, in part, to their
being more than one row deep, though they have not become
divided into two distinct germinal layers. It is further clear
that the increase in the number of ceils in the thickened
part of the blastoderm is, in the main, a result of the mul-
tiplication of the original single row of cells, while a careful
examination of my sections proves that it is also partly due
to cells, derived from the yolk, having been added to the blas-
toderm.

In the following stage which I have obtained (which
cannot be very much older than the previous stage, because
my specimens of it come from the same batch of eggs), a
distinct and fairly circumscribed thickening forming the
ventral surface of the embryo has become established. Though
its component parts are somewhat indistinct, it appears to
consist of a procephalic lobe, a less prominent caudal lobe,
and an intermediate portion divided into about three seg-
ments; but its constituents cannot be clearly identified with
the structures visible in the previous stage. I am inclined,
however, to identify the anterior thickened area of the pre-
vious stage with the procephalic lobe, and a slight protu-
berance of the caudal portion (visible from the surface) with
the primitive cumulus. I have, however, failed to meet
with any trace of the cumulus in my sections.

To this stage, which forms the first of the second period
of the larval history, I shall return, but it is necessary
now to go back to the observations of Claparede and Bal-
biani.

There can, in the first place, be but little doubt that what
I have called the primitive cumulus in my description is the
structure so named by Claparede and Balbiani,

172

F. M. BALFOUR.

It is clear that Balbiani and Claparede have both failed to
appreciate the importance of the organ, which my observa-
tions show to be the part of the ventral thickening of the
blastoderm where two rows of cells are first established, and
therefore the point where the first traces of the future meso-
blast becomes visible.

Though Claparede and Balbiani differ somewhat as to the
position of the organ, they both make it last longer than I
do : I feel certainly inclined to doubt whether Claparede is
right in considering a body he figures after six segments are
present, to be the same as the dorsal organ of the embryo
before the formation of any segments, especially as all the
stages between the two appear to have escaped him. In
Agelena there is undoubtedly no organ in the position he
gives when six segments are found.

Balbiani^s observations accord fairly with my own up to
the stage represented in fig. 2. Beyond this stage my own
observations are not satisfactory, but I must state that I feel
doubtful whether Balbiani is correct in his description of the
gradual separation of the procephalic lobe and the cumulus,
and the passage of the latter to the dorsal surface, and
think it possible that he may have made a mistake as to
which side of the procephalic lobe, in relation to the parts
of the embryo, the cumulus is placed.

Although there appear to he grounds for doubting whether
either Balbiani and Claparede are correct in the position
they assign to the cumulus, my observations scarcely warrant
me in being very definite in my statements on this head, but,
as already mentioned, I am inclined to place the organ near
the posterior end (and therefore, as will be afterwards shown,
in a somewhat dorsal situation) of the ventral embryonic
thickening.

In my earliest stage of the third period there is present,
as has already been stated, a procephalic lobe, and an indis-
tinct and not very prominent caudal portion, and about
three segments between the two. The definition of the
parts of the blastoderm at this stage is still very imperfect,
but from subsequent stages it appears to me probable that the
first of the three segments is that of the first pair of ambu-
latory limbs, and that the segments of the chelicerae and pedi-
palpi are formed later than those of the first three ambula-
tory appendages.

Balbiani believes that the segment of chelicerse is formed
later than that of the six succeeding segments. He further
concludes, from the fact that this segment is cut off from the
procephalic portion in front, that it is really part of the

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 173

procephalic lobe. I cannot accept the validity of this argu-
ment; though I am glad to find myself in^ at any rate,
partial harmony with the distinguished French embryologist
as to the facts. Balbiani denies for this stage the existence
of a caudal lobe. There is certainly, as is very well shown
in my longitudinal sections, a thickening of the blastoderm in
the caudal region, though it is not so prominent in surface
views as the procephalic lobe.

A transverse section through an embryo at this stage (PI.
XX, fig. 1^) shows that there is a ventral plate of somewhat
columnar cells more than one row deep, and a dorsal portion
of the blastoderm formed of a single row of flattened cells.
Every section at this stage shows that the inner layer of
cells of the ventral plate is receiving accessions of cells from
the yolk, which has not to any appreciable extent altered
its constitution. A large cell, passing from the yolk to the
blastoderm, is shown in fig. 12 at y. c.

The cells of the 'central plate are now divided into two
distinct layers. The outer of these is the epihlast, the
inner the mesohlast. The cells of both layers are quite
continuous across the median line, and exhibit no trace of a
bilateral arrangement.

This stage is an interesting one on account of the striking
similarity which (apart from the amnion) exists between a
section through the blastoderm of a spider and that of an
insect immediately after the formation of the mesoblast.
The reader should compare Kowalevsky’s (^ Mem. Acad.
Petersbourg,’ vol. xvi, 1871) fig. 26, PI. IX with my fig.
12. The existence of a continuous ventral plate of meso-
blast has been noticed by Barrois (p. 532), who states that
the two mesoblastic bands originate from the longitudinal
division of a primitive single band.

In a slightly later stage (PI. XIX, fig. 3 a and 3 h) six
distinct segments are interpellated between the procephalic
and the caudal lobes. The two foremost, ch and pd (especially
the first), of these are far less distinct than the remainder,
and the first segment is very indistinctly separated from
the procephalic lobe. From the indistinctness of the
first two somites, I conclude that they are later formations .
than the four succeeding ones. The caudal and procephalic
lobes are very similar in appearance, but the procephalic lobe
is slightly the wider of the two. There is a slight protuber-
ance on the caudal lobe, which is possibly the remnant of
the cumulus. The superficial appearance of segmentation
is produced by a series of transverse valleys, separating
raised intermediate portions which form the segments. The

174

F. M. BALFOUR.

ventral thickening of the embryo now occupies rather more
than half the circumference of the ovum.

Transverse sections show that considerable changes have
been effected in the constitution of the blastoderm. In the
previous stage, the ventral plate was formed of an uniform
external layer of epiblast, and a continuous internal layer
of mesoblast. The mesoblast has now become divided along
the whole length of the embryo, except, perhaps, the proce-
phalic lobes, into two lateral bands which are not continuous
across the middle line (PI. XX, fig. 13 me). It has, more-
over, become a much more definite layer, closely attached
to the epiblast. Between each mesoblastic band and the
adjoining yolk there are placed a few scattered cells, which
in a somewhat later stage become the splanchnic mesoblast.
These cells are derived from the yolk-cells ; and almost
every section contains examples of such cells in the act of
joining the mesoblast..

The epiblast of the ventral plate has not, to any great
extent, altered in constitution. It is, perhaps, a shade
thinner in the median line than it is laterally. The division
of the mesoblast plate into two bands, together, perhaps,
with the slight reduction of the epiblast in the median
ventral line, gives rise at this stage to an imperfectly
marked median groove.

The dorsal epiblast is still formed of a single layer of flat
cells. In the neighbourhood of this layer the yolk nuclei
are especially concentrated. The yolk itself remains as
before.

The segments continue to increase regularly, each fresh
segment being added in the usual way between the last
formed segment and the unsegmented caudal lobe. At the
stage when about nine or ten segments have become estab-
lished, the first rudiments of appendages become visible.
At this period (PI. XIX, fig. 4) there is a distinct median
ventral groove, extending through the whole length of the
embryo, which becomes, however, considerably shallower
behind. The procephalic region is distinctly bilobed. The
first segment (that of the chelicerge) is better marked off
. from it than in the previous stage, but is without a trace
of an appendage, and exhibits therefore, in respect to the
development of its appendages, the same retardation that
characterised its first appearance. The next five segments,
viz, those of the pedipalpi and four ambulatory appendages,
present a very well-marked swelling at each extremity.
These swellings are the earliest traces of the appendages.
Of the three succeeding segments; only the first is well

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 175

differentiated. The caudal lobe, though less broad than the
procephalic lobe, is still a widish structure. The most
important internal changes concern the mesoblast, which is
now imperfectly though distinctly divided into somites,
corresponding with segments visible externally. Each meso-
blastic somite is formed of a distinct somatic layer closely
attached to the epiblast, and a thinner and less well-marked
splanchnic layer. In the appendage-bearing segments the
somatic layer is continued up into the appendages.

The epiblast is distinctly thinner in the median line than
at the two sides.

The next stage figured (PL XIX, figs. 5 and 6) is an im-
portant one, as it is characterised by the establishment of
the full number of appendages. The whole length of the
ventral plate has greatly increased, so that it embraces
nearly the circumference of the ovum, and there is left
uncovered but a very small arc between the two extremities
of the plate (PI. XIX, fig. 6 ; PI. XX, fig. 15). This arc
is the future dorsal portion of the embryo, which lags in its
development immensely behind the ventral portion.

There is a very distinctly bilobed procephalic region {pr, 1)
well separated from the segment with the cheliceree {ch). It
is marked by a shallow groove opening behind into a circular
depression (^L) — the earliest rudiment of the stomodseum.
The six segments behind the procephalic lobes are the six
largest, and each of them bears two prominent appendages.
They constitute the six appendage-bearing segments of the
adult. The four future ambulatory appendages are equal in
size : they are slightly larger than the pedipalpi, and these
again than the chelicerm. Behind the six somites with pro-
minent appendages there are four well-marked somites, each
with a small protuberance. These four protuberances are
provisional appendages. They have been found in many
other genera of Araneina (Claparede, Barrois). The segments
behind these are rudimentary and difficult to count, but
there are, at any rate, five, and at a slightly later stage
probably six, including the anal lobe. These fresh segments
have been formed by the continued segmentation of the anal
lobe, which has greatly altered its shape in the process. The
ventral groove of the earlier stage is still continued along
the whole length of the ventral plate.

By the close of this stage the full number of post-cephalic
segments has become established. They are best seen in the
longitudinal section (PI. XX, fig. 15). There are six anterior
appendage-bearing segments, followed by four with rudimen-
tary appendages (not seen in this figure), and six without

176

F. M. BALFOUR.

appendages behind. There are, therefore, sixteen in all.
This number accords with the result arrived at by Barrois,
but is higher by two than that given by Claparede.

The germinal layers {vide PL XX, fig. 14) have by this
stage undergone a further development. The mesoblastic
somites are more fully developed. The general relations of
these somites is shown in longitudinal section in PI. XX,
fig. 15, and in transverse section in PL XX, fig. 14. In the
tail, where they are simplest (shown on the upper side in fig,
14), each mesoblastic somite is formed of a somatic layer of
more or less cubical cells attached to the epiblast, and a
splanchnic layer of flattened cells. Between the two is placed a
completely circumscribed cavity, which constitutes part of the
embryonic body cavity. Between the yolk and the splanchnic
layer are placed a few scattered cells, which form the latest
derivatives of the yolk-cells, and are to be reckoned as part of
the splanclinic mesoblast. The mesoblastic somites do not
extend outwards beyond the edge of the ventral plate, and
the corresponding mesoblastic somites of the two sides do
not nearly meet in the middle line. In the limb-bearing
somites the mesoblast has the same general characters as in
the posterior somites, but the somatic layer is prolonged as
a hollow papilliform process into the limb, so that each limb
has an axial cavity continuous with the section of the body
cavity of its somite. The description given by Metschnikoff
of the formation of the mesoblastic somites in the scorpion,
and their continuation into the limhs, closely corresponds
with the history of these parts in spiders. In the region of
each procephalic lobe the mesoblast is present as a continuous
layer underneath the epiblast, but in the earlier part of the
stage, at any rate, is not formed of two distinct layers with
a cavity between them.

The epihlast at this stage has also undergone important
changes. Along the median ventral groove it has become
very thin. On each side of this groove it exhibits in each
appendage-bearing somite a well-marked thickening, which
gives in surface views the appearance of a slightly raised
area (PL XIX, fig. 5), between each appendage and the
median line. These thickenings are the first rudiments of
the ventral nerve ganglia. The ventral nerve cord at this
stage is formed of two ridge-like thickenings of the epiblast,
widely separated in the median line, each of which is con-
stituted of a series of raised divisions — the ganglia — united by
shorter, less prominent divisions (fig. 14, vg). The nerve
cords are formed from before backwards, and are not at this
stage found in the hinder segments. There is a distinct

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 177

ga7iglionic thickening for the chelicerm quite independent of
the procephalic lobes.

In the procephalic lobes the epiblast is much thickened
and is formed of several rows of cells. The greater part o^
it is destined to give rise to the supra-oesophageal ganglia.

During the various changes which have been described the
blastoderm cells have been continually dividing^ and, together
with their nuclei, have become considerably smaller than at
first. The yolk cells have in the meantime remained much
as before, and are, therefore, considerably larger than the
nuclei of the blastoderm cells. They are more numerous
than in the earlier stages, but are still surrounded by a
protoplasmic body, which is continued into a protoplasmic
reticulum. The yolk is still divided up into polygonal seg-
ments, but from sections it would appear that the nuclei are
more numerous than the segments, though I have failed to
arrive at quite definite conclusions on this point.

As development proceeds the appendages grow longer and
gradually bend inwards. They become very soon divided by
a series of ring-like constrictions which constitute the first
indications of the future joints (PI. XIX, fig. 6). The full
number of joints are not at once reached, but in the ambu-
latory appendages five only appear at first to be formed.
There are four joints in the pedipalpi, while the cheliceree do
not exhibit any signs of becoming jointed till somewhat later.
The primitive presence of only five joints in the ambulatory
appendages is interesting, as this number is permanent in
Insects and in Peripatus.

The next stage figured forms the last of the third
period (PI. XIX, fig. 7 .and 7 a). The ventral plate
is still rolled round the egg (fig. 7), and the end of the tail
and the procephalic lobes nearly meet dorsally, so that there
is but a very slight development of the dorsal region. There
are the same number of segments as before, and the chief
differences in appearance between the present and the
previous stage depend upon the fact (1) that the median
ventral integument between the nerve ganglia has become
wider, and at the same time thinner ; (2) that the limbs have
become much more developed ; (3) that the stomodaeum is
definitely established ; (4) that the procephalic lobes have
undergone considerable development.

Of these features, the three last require a fuller descrip-
tion. The limbs of the two sides are directed towards each
other, and nearly meet in the ventral line. The chelicerae
are two-jointed, and terminate in what appear like rudi-
mentary chelae, a fact which perhaps indicates that the

178

F. M. BALFOUR.

spiders are descended from ancestors with chelate chelicerae.
The four embryonic post-ambulatory appendages are now at
the height of their development.

The stomodseum (PI. XIX, fig. 7, and PI. XX, fig. 17, st)
is a deepish pit between the two procephalic lobes, and dis-
tinctly in front of segment of the chelicerse. It is bordered
in front by a large, well-marked, bilobed upper lip, and be-
hind by a smaller lower lip. The large upper lip is a tem-
porary structure, to be compared, perhaps, with the gigantic
upper lip of the embryo of Chelifer (cf. Metschinkoff) . On
each side of and behind the mouth two whitish masses are
visible, which are the epiblastic thickenings which constitute
the ganglia of the chelicerae (PI. XIX, fig. 7, ch. g).

The procephalic lobes {p7\ 1) now form two distinct
masses, and each of them is marked by a semicircular groove,
dividing them into a narrower anterior and a broader
posterior division.

In the region of the trunk the general arrangement of the
germinal layers has not altered to any great extent. The
ventral ganglionic thickenings are now developed in all the
segments in the abdominal as well as in the thoracic region.
The individual thickenings themselves, though much more
conspicuous than in the previous stage (PI. XX, fig. 16, v, c),
are still integral parts of the epiblast. They are more widely
separated than before in the middle line. The mesoblastic
somites retain their earlier constitution (PI. XX, fig. 16).
Beneath the procephalic lobes the raesoblast has, in most
respects, a constitution similar to that of a mesoblastic somite
in the trunk. It is formed of two bodies, one on each side,
each composed of a splanchnic and somatic layer (PI. XX,
fig. 17, sp. and so), enclosing between them a section of the
body-cavity. But the cephalic somites, unlike those of the
trunk, are united by a median bridge of mesoblast, in which
no division into two layers can be detected. This bridge
assists in forming a thick investment of mesoblast round the
stomodeeum [st)»

The existence of a section of the body cavity in the
prseoral region is a fact of some interest, especially when
taken in connection with the discovery, by Kleinenberg, of a
similar structure in the head of Lumbricus. The procephalic
lobe represents the praeoral lobe of Chaetopod larvae, but the
prolongation of the body cavity into it does not, in my opinion,
necessarily imply that it is equivalent to a post-oral segment.

The epiblast of the procephalic lobes is a thick layer
several cells deep, but without any trace of a separation of
the ganglionic portion from the epidermis,

J^OTES ON THE DEVELOPMENT OF THE ARANEINA. 179

The nuclei of the yolk have increased in number, but the
yolk, in other respects, retains its earlier characters.

The next period in the development is that in which the
body of the embryo gradually acquires the adult form. The
most important event which takes place during this period
is the development of the dorsal region of the embryo, which,
up to its commencement, is practically non-existent. As a
consequence of the development of the dorsal region, the
embryo, which has hitherto had what may be called a dorsal
flexure, gradually unrolls itself, and acquires a ventral
flexure. This change in the flexure of the embryo is in
appearance a rather complicated phenomenon, and has been
somewhat dififerently described by the two naturalists who
have studied it in recent times.

For Claparede the prime cause of the change of flexure
is the translation dorsalwards of the limbs. He compares
the dorsal region of the embryo to the arc of a circle, the
two ends of which are united by a cord formed by the line
of insertion of the limbs. He points out that if you bring
the middle of the cord, so stretched between the two ends of
the arc, nearer to the summit of the arc, you necessarily
cause the two ends of the arch to approach each other, or, in
other words, if the insertion of the limbs is drawn up dor-
sally, the head and tail must approach each other
ventrally.

Barrois takes quite a different view to that of Claparede,
which will perhaps be best understood if I quote a transla-
tion of his own words. He says: ^^At the period of the
last stage of the embryonic band (the stage represented in
PI. XX, fig. 7, in the present paper) this latter completely
encircles the egg, and its posterior extremity nearly ap-
proaches the cephalic region. Finally, the germinal bands,
where they unite at the anal lobe (placed above on the
dorsal surface), form between them a very acute angle.
During the following stages one observes the anal segment
separate further and further from the cephalic region, and
approach nearer and nearer to the ventral region. This dis-
placement of the anal segment determines, in its turn, a
modification in the divergence of the anal bands ; the angle
which they form at their junction tends to become more
obtuse. The same processes continue regularly till the anal
segment comes to occupy the opposite extremity to the
cephalic region, a period at which the two germinal bands
are placed in the same plane and the two sides of the obtuse
angle end by meeting in a straight line. If we suppose a
continuation of the same phenomenon it is clear that the

180

F. M, BALFOUR.

anal segment will come to occupy a position on the ventral
surface, and the germinal bands to approach, but in the
inverse way, so as to form an angle opposite to that which
they formed at first. This condition ends the process by
which the posterior extremity of the embryonic band, at first
directed towards the dorsal side, comes to bend in towards
the ventral region.”

Neither of the above explanations is to my mind perfectly
satisfactory. The whole phenomenon appears to me to be
very simple, and to be caused by the elongation of the
dorsal region, ^. e, the region on the dorsal surface between
the anal and procephalic lobes. Such an elongation neces-
sarily separates the anal and procephalic lobes ; but, since
the ventral plate does not become shortened in the process,
and the embryo cannot straighten itself on account of the
egg-shell, it necessarily becomes flexed, and such flexure can
only be what I have already called a ventral flexure. If there
were but little food yolk this flexure would cause the whole
embryo to be bent in, so as to have the ventral surface con-
cave, but instead of this the flexure is confined at first to the
two bands which form the ventral plate. These bands are
bent in the natural way (PL XIX, fig. 8, b), but the yolk
forms a projection, a kind of yolk sack as Barrois calls it, dis-
tending the thin integument between the two ventral bands.
This yolk sack is shown in surface view in P1. XIX, fig. 8,
and in section in PI. XXI, fig. 18. At a later period, when
the yolk has become largely absorbed in the formation of
various organs, the true nature of the ventral flexure
becomes apparent, and the abdomen of the young Spider is
found to be bent over so as to press against the ventral
surface of the thorax (PI. XIX, fig. 9). This flexure is
shown in section in PI. XXI, fig. 21.

At the earliest stage of this period of which I have ex-
amples, the dorsal region has somewhat increased, though
not very much. The limbs have grown very considerably and
now cross in the middle line.

The ventral ganglia, though not the supra-oesophageal,
have become separated from the epiblast.

The yolk nuclei, each surrounded by protoplasm as before,
are much more numerous.

In other respects there are no great changes in the internal
features.

In my next stage, represented in PI. XIX, figs. 8 a, and
8 a very considerable advance has become effected. In
the first place the dorsal surface'has increased in length to
rather more than one half the circumference of the ovum.

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 181

The dorsal region has, however, not only increased in length,
but also in definiteness, and a series of transverse markings
(fig. 8 a and which are very conspicuous in the case of
the four anterior abdominal segments (the segments with
rudimentary appendages), have appeared, indicating the
limits of segments dorsally. The terga of the somites may,
in fact, be said to have become formed. The posterior terga
(fig. 8 d) are very narrow compared to the anterior.

The caudal protuberance is more prominent than it was,
and somewhat bilobed ; it is continued on each side into
one of the bands, into which the ventral plate is divided.
These bands, as is best seen in side view (fig. 8 h),
have a ventral curvature, or, perhaps more correctly, 'are
formed of two parts, which meet at a large angle open to-
wards the ventral surface. The posterior of these parts bears
the four still very conspicuous provisional appendages, and
the anterior the six pairs of thoracic appendages. The four
ambulatory appendages are now seven-jointed, as in the
adult, but though longer than in the previous stage they
do not any longer cross or even meet in the middle line, but
are, on the contrary, separated by a very considerable in-
terval. This is due to the great distension by the yolk of
the ventral part of the body, in the interval between the
two parts of the original ventral plate. The amount of this
yolk may be gathered from the section (PI. XXI, fig. 18).
The pedipalpi carry a blade on their basal joint. The che-
licerse no longer appear to spring from an independent post-
oral segment.

There is a conspicuous lower lip, but the upper is less
prominent than before. Sections at this stage show that
the internal changes have been nearly as considerable as the
external.

The dorsal region is now formed of a (1) flattened layer of
epiblast cells, and a (2) fairly thick layer of large and rather
characteristic cells which any one who has studied sections
of spider’s embryos will recognise as derivatives of the yolk.
These cells are not, therefore, derived from prolongations
of the somatic and splanchnic layers of the already formed'
somites, but are new formations derived from the yolk. They
commenced to be formed at a much earlier period, and some
of them are shown in the long section (PI. XX, fig. 15).
In the next stage these cells become differentiated into the
somatic and splanchnic mesoblast layers of the dorsal region
of the embryo.

In the dorsal region of the abdomen the heart has already
become established. So far as I have been able to make

182

F. M. BALFOUR.

out it is formed from a solid cord of the cells of the dorsal
region. The peripheral layer of this cord gives rise to the
walls of the heart, while the central cells become converted
into the corpuscles of the blood.

The rudiment of the heart is in contact with the epiblast
above, and there is no greater evidence of its being derived
from the splanchnic than from the somatic mesoblast ; it is,
in fact, formed before the dorsal mesoblast has become dif-
ferentiated into two layers.

In the abdomen three or four transverse septa, derived
from the splanchnic mesoblast, grow a short way into the
yolk. They become more conspicuous during the succeed-
ing stage, and are spoken of in detail in the description of
that stage. In the anterior part of the thorax a longitu-
dinal and vertical septum is formed, which grows downwards
from the median dorsal line, and divides the yolk in this
region into two parts. In this septum there is formed at a
later stage a vertical muscle attached to the suctorial part
of the stomodseum.

The mesoblastic somites of the earlier stage are but little
modified ; and there are still prolongations of the body
cavity into the limbs (PI. XXI, fig. 18).

The lateral parts of the ventral nerve cords are now at
their maximum of separation (PI. XXI, fig. 18, v, g).
Considerable differentiation has already set in in the con-
stitution of the ganglia themselves, which are composed of
an outer mass of ganglion cells enclosing a kernel of nerve
fibres, which lie on the inner side and connect the successive
ganglia. There are still distinct thoracic and abdominal
ganglia for each segment, and there is also a pair of separate
ganglion for the chelicerse, which assists, however, in forming
the oesophageal commissures.

The thickenings of the praeoral lobe which form the supra-
cesophageal ganglia are nearly though not quite separated from
the epiblast. The semicircular grooves of the earlier stages
are now deeper than before, and are well shown in sections
nearly parallel to the outer anterior surface of the ganglion
(PI. XXI, fig. 19). The supra-oesophageal ganglia are still
entirely formed of undifferentiated cells, and are without
commissural tissue like that present in the ventral ganglia.

The stomodeeum has considerably increased in length, and
the proctodeeurn has become formed as a short, posteriorly
directed involution of the epiblast. I have seen traces of
what I believe to be two outgrowths from it, which form
the Malpighian bodies.

The next stage constitutes (PI. XIX, fig. 9) the last which

i

KOTES ON THE DEVELOPMENT OF THE ARANEINA. 183

requires to be dealt with so far as the external features are
concerned. The yolk has now mainly passed into the
abdomen, and the constriction separating the thorax and
abdomen has began to appear. The yolk sack has become ab-
sorbed, so that the two halves of the ventral plate in the thorax
are no longer widely divaricated. The limbs have to a large
extent acquired their permanent structure, and the rings of
which they are formed in the earlier stages are now replaced
by definite joints. A delicate cuticle has become formed,
which is not figured in my sections. The four rudimen-
tary appendages have disappeared, unless, which to seems me
in the highest degree improbable, they remain as the spin-
ning mammillae, two pairs of which are now present. Be-
hind is the anal lobe, which is much smaller and less con-
spicuous than in the previous stage. The spinnerets and
anal lobe are shown as five papillae in PI. XIX, fig. 9. Dor-
sally the heart is now very conspicuous, and in front of the
chelicerae may be seen the supra-oesophageal ganglia.

The indifferent mesoblast has now to a great extent
become converted into the permanent tissues. On the dorsal
surface there was present in the last stage a great mass of
unformed mesoblast cells. This mass of cells has now
become divided into a somatic and splanchnic layer (PI. XXI,
fig. 22). It has, moreover, in the abdominal region at any
rate, become divided up into somites. At the junction be-
tween the successive somites the splanchnic mesoblast on
each side of the abdomen dips down into the yolk and forms
a septum (PI. XXI, fig. 22 s). The septa so formed, which
were first described by Barrois, are not complete. The septa
of the t^vo sides do not, in the first place, quite meet along
the median dorsal or ventral lines, and in the second place
they only penetrate the yolk for a certain distance. In-
ternally they usually end in a thickened border.

Along the line of insertion of each of these septa there is
developed a considerable space between the somatic and
splanchnic layers of mesoblast. The parts of the body
cavity so established are transversely directed channels pass-
ing from the heart outwards. They probably constitute the'
venous spaces, and perhaps also contain the transverse aortic
branches.

In the intervals between these venous spaces the somatic
and splanchnic layers of mesoblast are in contact with each
other.

I have not been able to work out satisfactorily the later
stages of development of the septa, but I have found that
tbcy play an important part in tlie subsequent development

VOL. XX. NEW SEU.

N

184

F. M. BALFOUR.

of the abdomen. In the first place they send ofi* lateral off-
shoots, which unite the various septa together, and divide
up the cavity of the abdomen into a number of partially
separated compartments. There appears, however, to be
left a free axial space for the alimentary tract, the meso-
blastic walls of which are, I believe, formed from the septa.

At the present stage the splanchnic mesoblast, apart from
the septa, is a delicate membrane of flattened cells (fig. 22,
sp). The somatic mesoblast is thicker, and is formed of
scattered cells (50).

The somatic layer is in part converted, in the posterior
region of the abdomen, into a delicate layer of longitudinal
muscles, the fibres of which are not continuous for the whole
length of the body, but are interrupted at the lines of junc-
tion of the successive segments. They are not present in the
anterior part of the abdomen. The longitudinal direction of
these fibres, and their division with myotomes, is interesting,
since both these characters, which are preserved in Scorpions,
are lost in the abdomen of the adult Spider.

The original mesoblastic somites have undergone quite as
important changes as the dorsal mesoblast. In the abdominal
region the somatic layer constitutes two powerful bands of
longitudinal muscles, inserted anteriorly at the root of the
fourth ambulatory appendage, and posteriorly at the spinning
mammillse. Between these two bands are placed the nervous
bands. The relation of these parts are shown in the section in
PI. XXI, fig. 20 c?, which cuts the abdomen horizontally and
longitudinally. The mesoblastic bands are seen at m., and the
nervous bands within them at ah. g. In the thoracic region
the part of the somatic layer in each limb is converted into
muscles, which are continued into dorsal and ventral muscles
in the thorax {vide fig. 20 c). There are, in addition to these,
intrinsic transverse fibres on the ventral side of the thorax.
Besides these muscles there are in the thorax, attached to
the suctorial extremity of the stomodseum, three powerful
muscles, which I believe to be derived from the somatic me-
soblast. One of these passes vertically down from the dorsal
surface, in the septum the commencement of which was
described in the last stage. The two other muscles are
lateral, one on each side (PI. XX, fig. 20 c).

The heart has now, in most respects, reached its full
development. It is formed of an outer muscular layer, within
which is a doubly- contoured lining, containing nuclei at
intervals, which is probably of the nature of an epithelioid
lining (PI. XXI, fig. 22 ht). In its lumen are numerous
blood-corpuscles (not represented in my figure). The heart

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 185

lies in a space bound below by the splanchnic mesoblast,
and to the sides by the somatic mesoblast. This space
forms a kind of pericardium (fig. pc)^ but dorsally the
heart is in contact with the epiblast. The arterial trunks
connected with it are fully established.

The nervous system has undergone very importa^nt
changes.

In the abdominal region the ganglia of each side have
fused together into a continuous cord (fig. SI ah.g,). In
fig. SO, in which the abdomen is cut horizontally and longi-
tudinally, there are seen the two abdominal cords (ab. g.)
united by two transverse commissures ; and I believe that
there are at this stage three or four transverse commissures
at any rate, which remain as indications of the separate
ganglia, from the coalescence of which the abdominal cords
are formed. The two abdominal cords are parallel and in
close contact.

In the thoracic region changes of not less importance
have taken place. The ganglia are still distinct. The two
cords formed of these ganglia are no longer widely separated
in median line, but meet, in the usual way, in the ventral
line. Transverse commissures have become established
(fig. 20 c) between the ganglia of the two sides. There is
as little trace at this, as at the previous stages, of an ingrowth
of epiblast, to form a median portion of the central nervous
system. Such a median structure has been described by
Hatschek for Lepidoptera, and he states that it gives rise
to the transverse commissures between the ganglia. My ob-
servations show that for the spider, at any rate, nothing of
the kind is present.

As shown in the longitudinal section (PL XXI, fig. 21),
the ganglion of the chelicerse has now united with the supra-
msophageal ganglion. It forms, as is shown in fig. 20 h {ch, g.),
a part of the oesophageal commissure, and there is no sub-
oesophageal commissure uniting the ganglia of the chelicerae,
but the oesophageal ring is completed below by the ganglia
of the pedipalpi (fig. 20 c,pd, g.).

The supra-oesophageal ganglia have become completely^
separated from the epiblast.

I have unfortunately not studied their constitution in the
adult, so that I cannot satisfactorily identify the parts which
can be made out at this stage.

I distinguish, however, the following regions :

(1) A central region containing the commissural part,
and continuous below with the ganglia of the chelicerae.

(2) A dorsal region formed of two hemispherical lobes.

186

F. M. BALFOUR.

(3) A ventral anterior region.

The central region contains in its interior the cornnhs-
saral portion, forming a punctifonn, rounded mass in each
ganglion. A transverse commissure connects the two {vide
fig. 20 If).

The dorsal hemispherical lobes are derived from the part
which, at the earlier stage, contained the semicircular
grooves. When the supra-oesophageal ganglia become sepa-
rated from the epidermis the cells lining these grooves
become constricted off with them, and form part of these
ganglia. Two cavities are thus formed in this part of the
supra-oesophageal ganglia. These cavities become, for the
most part, obliterated, but persist at the outer side of the
hemispherical lobes (figs. 20 a and 21).

The ventral lobe of the brain is a large mass shown in long
section in fig. 21. It lies immediately in front of and almost
in contact with the ganglia of the chelicerse.

The two hemispherical lobes agree in position with the
fungiform body (pilzhutfdrmige Korpern), which has attracted
so much the attention of anatomists, in the supra-oesophageal
ganglia of Insects and Crustacea; but till the adult brain of
Spiders has been more fully studied it is not possible to state
whether the hemispherical lobes become fungiform bodies.

Hatschek^ has described a special epiblastic invagination
in the supra-oesophageal ganglion of Bombyx, which is pro-
bably identical with the semicircular groove of Spiders and
Scorpions, but in the figure he gives the groove does not
resemble that in the Arachnida. A similar groove is found
in Peripatus, and there forms, as I have found, a large part
of the supra-oesophageal ganglia. It is figured by Moseley,
^ Phil. Trans.,’ vol. 164, pi. Ixxv, fig. 9.

The stomodseum is considerably larger than in the last
stage, and is lined by a cuticle ; it is a blind tube, the blind
end of which is the suctorial pouch of the adult. To this
pouch are attached the vertical dorsal, and two lateral muscles
spoken of above.

The proctodseum (/?r.)has also grown in length, and the two
Malpighian vessels which grow out from its blind extremity
(fig. 20 e, mp. g.) have become quite distinct. The part now
formed is the rectum of the adult. The proctodaeum is sur-
rounded by a great mass of splanchnic mesoblast. The mesen-
teroii has as yet hardly commenced to be developed. There
is, however, a short tube close to the proctodaeum (fig. 20
7/IC5), which would seem to be the commencement of it. It

* “ lieitrjige z. Eiitvvick d. Lepidopteren,” ‘ Jenaische Zeit.,’ vol. xi,
p. 124.

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 187

ends blindly on the side adjoining the rectum, but is open
anteriorly towards the yolk, and there can be very little doubt
that it owes its origin to cells derived from the yolk. On
its outer surface is a layer of mesoblast.

From the condition of the mesenteron at this stage there
can be but little doubt that it wdll be formed, not on the sur-
face, hut in the interior of the yolk. I failed to find any
trace of an anterior part of the mesenteron adjoining the
stomodaeum. In the posterior part of the thorax (vide fig.
20 d), there is undoubtedly no trace of the alimentary tract.

The presence of this rudiment shows that Barrois is
mistaken in supposing that the alimentary canal is formed
entirely from the stomodaeum and proctodaeuin, which are
stated by him to grow towards each other, and to meet at
the junction of the thorax and abdomen. My own impres-
sion is that the stomodaeum and proctodaeum have reached
their full extension at the present stage, and that both the
stomach in the thorax and the intestine in the abdomen are
products of the mesenteron.

The yolk retains its earlier constitution, being divided
into polygonal segments, formed of large yolk vesicles.
The nuclei are more numerous than before. In the thorax
the yolk is anteriorly divided into two lobes by the vertical
septum, which contains the vertical muscle of the suctorial
pouch. In the posterior part of the thorax it is undivided.

I have not yet been able clearly to make out the eventual
fate of the yolk. At a subsequent stage, when the cavity of
the abdomen is cut up into a series of compartments by the
growth of the septa, described above, the yolk fills these
compartments, and there is undoubtedly a proliferation of
yolk cells round the walls of these compartments. It would
not be unreasonable to conclude from this that the compart-
ments were destined to form the hepatic cseca, each caecum
being enclosed in a layer of splanchnic mesoblast, and its
hypoblastic wall being derived from the yolk cells. I think
that this hypothesis is probably correct, but I have met with
some facts which made me think it possible that the thick- ^
enings at the ends of the septa, visible in PI. XXI, fig. 22,
w’ere the commencing hepatic caeca.

I must, in fact, admit that I have hitherto failed to work
out satisfactorily the history of the mesenteron and its appen-
dages. The firm cuticle of young spiders is an obstacle both
in the way of making sections and of staining, which I have
not yet overcome.

168

F. M. BALFOUfl.

General Conclusions,

Without attempting to compare at length the develop-
ment of the spiders with that of other Arthropoda, I propose
to point out a few features in the development of spiders,
which appear to show that the Arachnida are undoubtedly
more closely related to the other Tracheata than to the
Crustacea.

The whole history of the formation of the mesoblast is
very similar to that in insects. The mesoblast in both groups
is formed by a thickening of the median line of the ventral
plate (germinal streak).

In insects there is usually formed a median groove, the
walls of which become converted into a plate of mesoblast.
In spiders there is no such groove, but a median keel-like
thickening of the ventral plate (PI. XX, fig. 11), is very pro-
bably an homologous structure. The unpaired plate of
mesoblast formed in both insects and Arachnida is exactly
similar, and beomes divided, in both groups, into two bands,
one on each side of the middle line. Such differences as
there are between Insects and Arachnida sink into insigni-
ficance compared with the immense differences in the origin
of the mesoblast between either^ group^ and that in the
Isopoda, or, still more, the Malacostraca and most Crustacea.
In most Crustacea we find that the mesoblast is budded off
from the walls of an invagination, which gives rise to the
mesenteron.

In both spiders and Myriopoda, and probably insects, the
mesoblast is subsequently divided into somites, the lumen of
which is continued into the limbs. In Crustacea mesoblastic
somites have not usually been found, though they appear
occasionally to occur, e.g, Mysis, but they are in no case
similar to those in the Tracheata.

In the formation of the alimentary tract, again, the
differences between the Crustacea and Tracheata are equally
marked, and the Arachnida agree with the Tracheata.
There is generally in Crustacea an invagination, which gives
rise to the mesenteron. In Tracheata this never occurs.
The proctodseum is usually formed in Crustacea before or,
at any rate, not later than the stomodceum.^ The reverse is
true for the Tracheata. In Crustacea the proctodaeum and
stomodseum, especially the former, are very long, and usually
give rise to the greater part of the alimentary tract, while
the mesenteron is usually short.

* If Grobben’s account of the development of Moina is correct this
statement must be considered not to be universally true.

NOTES ON THE DEVELOPMENT OF THE ARANEINA. 189

In the Tracheata the mesenteron is always considerable,
and the proctodseura is always short. The derivation of the
Malpighian bodies from the proctodseum is common to
most Tracheata. Such organs are not found in the
Crustacea.

With reference to other points in my investigations, the
evidence which I have got that the chelicerae are true post-
oral appendages supplied in the embryo from a distinct
postoral ganglion, confirms the conclusions of most previous
investigators, and shows that these appendages are equivalent
to the mandibles, or possibly the first pair of maxilla of
other Tracheata. The invagination, which I have found, of
part of a groove of epiblast in the formation of the supra-
oesophageal ganglia is of interest, owing to the wide exten-
sion of a similar occurrence amongst the Tracheata.

The wide divarication of the ventral nerve cords in the
embryo renders it easy to prove that there is no median in-
vagination of epiblast between them, and supports Kleinen-
berg^s observations on Lumbricus as to the absence of this
invagination. I have further satisfied myself as to the
absence of such an invagination in Peripatus. It is probable
that Hatschek and other observers who have followed him
are mistaken in affirming the existence of such an invagina-
tion in either the Chaetopoda or the Arthropoda.

The observations recorded in this paper on the yolk cells
and their derivations are, on the whole, in close harmony
with the observations of Dohrn, Bobretsky, and Graber, on
Insects. They show, however, that the first formed meso-
blastic plate does not give rise to the whole of the mesoblast,
but that during the whole of embryonic life the mesoblast
continues to receive accessions of cells derived from the cells
of the yolk.

Araneina,

1. Balbiani^ “Memoire sur le Developpement des Araneides,” ‘Ann.
Sci. Nat.,’ series v, vol. xvii, 1873.

2. J. Barrois^ “ Recherches s. 1. Developpement des Araignees,”
‘ Journal de 1. Anat. e. de la Physiol.,’ 1878.

3. E. Claparede, ‘ Recherches s. I’Evolution des Araignees,’ Utrecht,

4. Ilerold, ‘De Generatione Araniorum in Ovo,’ Marburg, 1824.

5. II. Ludwig, “ Ueb. d. Bildung des Blastoderm,” bie d. Spinnea,

‘ Zeit. f. wiss. Zool.,’vol. xxvi, 1876,

■ 190

DR. L. WALDSTBIN.

A Contribution to the Biology of Bacteria. By Dr. L.

Waldstetn, of the Pathological Institute, Heidelberg.^

The experiments here recorded were not undertaken to meet
Dr. Bastian on his favourite field of research, nor does this
communication partake of the polemical nature of so many of
the recent publications on the same subject. The wide import-
ance, quite irrespective of abiogenesis, which attaches itself
to Dr. Bastian’s work,^ is here recognised in so far as the method
employed by him is concerned.

Bacteria have been assigned a prominent position in pathology,
more especially since the theories of Pettenkofer and of Lister
have introduced them as causally connected with infection. They
are to-day certainly made responsible for more than can be well
proven with our present knowledge of the conditions of their
existence. This can be gathered from the perusal of but a very
small portion of the voluminous literature, which has grouped
itself about the many theories and hypotheses as to their patho-
genic properties, based on experiments which, in many instances,
have given quite as frequently negative as positive results.

The study of the conditions under which Bacteria multiply,
as well as of those which prove to be unfavorable to their develop-
ment, is, clearly in view of these circumstances, of the highest
interest, and all researches in this direction ought to be examined
with great care, in an objective spirit. Dr. Bastian, in the
course of his researches, has not strictly confined himself to these
questions alone. He has also drawn conclusions which, as he
himself remarks, tend towards a complete revolution in the
current opinions held with respect to the role taken in pathology
by micro-organisms.

Dr. Bastian — citing the experiments of Gerhardt, which go to
prove that alcohol mixed with a little potash becomes converted
into vinegar and a brown resinous substance, whereas in its
natural state it can be exposed to the air indefinitely without
becoming acid — proposes to show that potash added to sterilised
and well guarded urine produces conditions favorable to the
spontaneous generation of micro-organisms. He adds the alkali
in a certain proportion, determined in each case by the acidity of
the urine.

* The author desires to refer the reader to his more extended memoir
on the same subject in Virchow’s ‘ Archiv,* vol. 77, 1879.

* “ On the Conditions favouring Fermentation, and the appearance of
Bacilli, Micrococci, and Torulac in previously Boiled Fluids,” by H. Charl-
ton Bastian. ‘Journal of the Linnean Society (Zoology),’ Oct. 24, 1877.
London.

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA. 191

I refer to the original for an account of the details of the
experiment.

The result was invariably that the urine in the control experi-
ments remains clear and apparently unaltered for an indefinite
time; whilst where the potash had been allowed to act upon the
sterilised fluid it became turbid, lighter in colour, and swarmed
with organisms in from eighteen to twenty-six hours, on the
average.

In repeating these experiments I have closely followed Dr. Bas-
tian’s directions, with the exception that the potash tubes were made
by drawing out test tubes in the middle, charging them with the
requisite amount of liquor potassae, and then hermetically sealing
them by melting, and thus separating the upper part of the
tube. This procedure is much more convenient than the original
method; and the sterilisation called for is ensured fully by
keeping such tubes in boiling water during twenty minutes, as,
says Dr. Bastian, in these experiments I soon found that the
longer or shorter duration of the period of boiling of the liquor
potassae tubes did not appreciably influence the results.

Dr. AV. Koberts and M. Pasteur could not obtain the same
results, and it is furthermore not absolutely certain that all germs
were kept out of the retorts when Dr. Bastian boiled the tubes
in oil during a period of twenty hours. He himself mentions the
difficulty of properl}" cleaning them then and, even if it be con-
ceded that those contained within the tubes were destroyed,
fresh germs might possibly adhere to the exterior, whereas the
heat and the period of immersion of the urine retort in boiling
water would be the only safeguard against a contamination with
such germs.

If the views of all experimenters in regard to the degree of
heat and the duration of its action are compared, it may be
allowed to doubt that the germs in the vessels are all destroyed.
It appears that micrococci and the so-called spores of Bacillut
suhtilis and Bacillus anthracis retain their vitality in consider-
ably higher temperatures than do the fully developed Bacillus
and shorter rod-forms.

It is true that the concentration of the fluid is changed if
heated longer than to ebullition before sealing, and therefore one
may entertain doubts of complete sterilisation so long as Dr.
Bastian cannot either eliminate this possibility or disregard the
change. One is strongly reminded of Needham^s objections to
Spallanzani^s experiments in the last century, when Dr. Bastian
says that if Pasteups modification to continue the boiling were
introduced, the urine might undergo a chemical change. But
not alone is the uncertainty of a thorough sterilisation of great
weight in the a priori estimation of the final success of such

192

DR. L. WALDSTEIN.

experiments. The amount of oxygen remaining in the fluid is
also considered by different authors a factor of greater or less
moment with regard to the different forms of organisms or their
earlier stages. The more recent investigations on this subject
made known since the first publication of the present communi-
cation, as well as those of the best authorities previously, have
shown that as the different forms retain their vitality with varied
tenacity under the influence of higher and lower degrees of heat,
so many are more or less independent of the amount of oxygen
present. It may be freely admitted that the greater part of the
gas is well driven out of the retorts, but some may still remain,
and if so-called germs and lower forms are yet within the fluid
or adherent to the inner surface of the retorts or to the exterior
of the liquor potassse tubes, they will find oxygen enough for
their development. The unfavorable conditions into which
such germs are brought by the treatment of the vessels and
their contents, may well limit the degree of fertility ; on the other
hand, as Dr. Bastian himself has found, the incubator tempera-
ture of 122° B. (40 — 45° C.) is a very favorable agent.

Notwithstanding the justice of these objections, which could
be multiplied by an extended and a detailed reference to the
Bacterium literature, w^hat would appear as most remarkable in
Dr. Bastian^s numerous repetitions of the experiment, is the uni-
formity of the difference between the contents of the retorts with
urine and unbroken tubes and of those with the mixed fluids
produced by breaking the potash tubes within the retorts after
closure and ebullition. The critics of Dr. Bastian have singularly
enough not laid the proper stress upon this, the real point at
issue, for even if all their objections were just, the fact of such
a difference would be all the more interesting. In their trials of
the method they have not scrupulously followed his directions,
and he is w^ll justified in refusing to accept their adverse state-
ments for this if for no other reason.

Before entering upon my own experimentation I would with a
word call attention to the apparent unchanged condition of the
fluid in the one set of retorts. I have found, and the same
observation has been made by others, that the degree of turbidity
is not in direct proportion to the multiplication of Bacteria.
Certain specimens of perfectly clear and apparently unchanged
fluids may swarm with organisms, whilst in turbid urine with
much sediment few if any organisms can with absolute certainty
be discovered with the microscope. This, it is true, is not the
rule, but as it may occur the fact of an apparently unchanged
apjiearance alone is not a sufficient test. It appears that the
formation of so-called zoogloea masses is in most cases the cause
of turbidity.

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA. 193

The fresh filtered urine wliich I used never showed precipita-
tion of phosphates or albumen on boiling, and in order to
ascertain the degree of acidity a previously boiled specimen of it
was tested. This latter precaution, and the addition of the two-
thirds proportion of liq. pot., was only adopted, however, after
the result was found to be other than that recorded by Mr.
Bastian, with urine whose degree of acidity had been determined
with portions previously unboiled and mixed in the prepared
retorts with liquor potassse of the proportion of three-fourths of
its acidity. To facilitate the determination of the degree of
acidity a burette was constructed, with a capillary termination
by means of which drops of known weight (0*0075 grms.) of
the liquor potassse could be mixed gradually with 30 c. c. of
urine. The equality of the drops was insured by suitable pre-
cautions, the liq. pot. changed and the instrument well cleaned
from time to time. In the first series of experiments litmus
paper was used, but ever after rosolic acid, a very delicate re-
agent. In determining whether or not the contents of the
retorts contained organisms, three to five specimens from every
vessel were examined, and only then was their presence noted
when they could either be distinguished by their characteristic
movements or other signs of life, such as, for instance, an
increase in the size of the zoogloea masses, and when caustic,
liq. potassge, and strong acetic acid failed to dissolve the sus-
pected bodies. In all more than fifty retorts were charged and
examined.

The fluids, which were mixed with the proportionate amount
of liq. pot., and which were examined at intervals between
twenty-four to seventy -two hours and fourteen days, were all
found turbid and showed sedimentation, but not one of them
contained organisms; the sediment was, moreover, found to
consist of phosphates, crystalline, and amorphous masses on
being examined by means of micro-chemical tests. Their odour
was not appreciably changed, not even in those specimens which
contained triple phosphate crystals.

This result remained the same after using urine from other
individuals, and after adding peptone to some specimens. In
order to make quite sure that organisms were not contained in
the fluid, at least wdthin the limits of all means of examination,
portions of the same were allowed to form sediment in narrow glass
tubes with a capillary termination, and the first drops carefully
searched through under the microscope. In all these trials the
result remained the same.

After remaining in the incubator during one month the
specimens containing unbroken liquor potasssc tubes were found
clear with very few exceptions, and, when examined, contained

194

DR. L. WALDSTEIN.

no Bacteria, tlie sediment dissolving after addition of acetic
acid ; the degree of acidity had been in all such cases remark-
ably low.

I could not so far corroborate the statements of Mr. Bastian,
and was unable to explain this remarkable divergence from the
expected result. The disappointment was the more dishearten-
ing as the repetition of his experiments were made from the
beginning more with a view to an explanation than as test
experiments designed to determine the truth of his statement of
fact.

There still remained seventeen retorts in the incubator, of
which some contained mixed and others unmixed fluids. They
were opened and their contents examined after they had been
exposed to a temperature of 120° B. (40° — 45° C.) between
65 to 126 days. They all contained more or less flocculent
suspensions and a whitish-yellow sediment, which adhered to
the walls of the retort ; their reaction with rosolic acid proved
to be alkaline and several smelled of ammonia. One only
had a putrid odour, and was registered as not entirely without
blemish. On addition of acetic acid the greater part of the
sediment, which proved to consist chiefly of amorphous masses
and of few crystals, was dissolved ; here and there were found
peculiar zoogloea-like bodies of yellowish tinge and a remarkable
power of resistance to reagents, the exact nature of which I
was unable to determine, and which frequently appeared in a
later series of experiments ; and lastly, every specimen of urine,
with and vnthout admixture of liq. pot., and such also with
perfectly clear contents, together with one whose fluid showed acid
reaction, contained innumerable organisms of various forms.

Micrococci and small spheroid bodies were most numerous,
and appeared either as individuals or in torula- and zoogloea-
form ; together with these, and next in number, smaller rods
(vibrio), single and in biscuit form, and lastly, variable quantity
of bacilli of all lengths, free or articulated. These organisms
showed the characteristic movements, or were grouped together
in more or less dense masses! In no case was the degree of
turbidity in direct proportion to the number of organisms, but
it seemed to stand in relation to the zoogloea after the action of
acetic acid had dissolved the other suspensions. It follows
from this that the long-continued action of the incubator tem-
perature is favorable to the multiplication of Bacteria, even
when the remaining conditions introduced tend to retard their
development. In what manner, if in any, the presence of the
alkali acts as a factor was not thus far apparent, especially as
organisms appeared where it was not allowed to mix with the
pabulum just as freely as where it was added.

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA. 195

The next step was, therefore, to ascertain if there existed
a causal connection or simply a concomitance of the pheno-
mena. For this object the following elementary experiments
were made. Fresh-filtered urine was put into test tubes,
their opening closed with cotton and dipped into the water
bath of 122° F. (40 — 45° C.). After thirty hours the
reaction was alkaline, ammonia was generated, and countless
Bacteria found in it. After some time the Bacteria sank to
the bottom, and the urine was again clear and limpid. In
spite of the lifeless condition of the organisms ammonia was
constantly generated, and the degree of alkalescence enhanced.
In order to determine the latter I made use of a burette, con-
structed as that for liquor potassse, and added sulphuric-acid
solution (5 per cent.), drop by drop, to a certain quantity of the
urine, to which a small quantity of the above-mentioned alco-
holic solution of rosolic acid was added. The change from a
beautiful rose-red tint to a bright straw-like yellow, indicated
that the first drop of acid in excess had been added. Perhaps
this simple experiment goes far to show that the decomposition
of urea is not so directly dependent upon living organisms as
the vitalistic theory implies ; but, on the contrary, that the con-
tinued action of the temperature of 122° F. (40 — 45° C.) is a
favorable circumstance, if not an etiological factor of the
same. Nor can the urea ferment of Musculus be its cause, for
its action is said by the discoverer to be neutralised in 80° C.
and by alkalies, both of which are made to act upon the con-
tents of the retorts in Dr. Bastian’s experiment. It is well
known that urea is decomposed when its aqueous solution is
heated in sealed glass tubes, or when in the dry state it is melted
with an alkali hydrate. Mindful of the latter fact, I charged
four test tubes with an aqueous solution of urea, and added to
two of these a small quantity of dilute liquor potassae. Two
tubes, one with urea and the other with urea and the alkali,
were sealed in the flame and placed in the incubator at 122° F.
(40 — 45° C.). The remaining pair was dipped into the water
bath of the same temperature, and closed with a cotton plug.
The result was :

I. Test tube with urea and liquor potassm, cotton plug.
Generation of ammonia after seventeen hours in water bath.

II. As I, but without alkali. Feeble generation of ammonia,
alkaline reaction after 139 hours in water bath.

HI. Test tube with urea and liq. pot., sealed. Feeble genera-
tion of ammonia after twelve days in incubator.

IV. As III, but without alkali, was opened after fifty-seven
days. Generation of ammonia not distinctly determinable, but

196

DR. L. WALDSTEIN.

alkaline reaction was shewn with rosolic acid, whereas litmus
paper was not changed.

It is clear, therefore, that at the temperature of the incubator
and of the water bath urea is decomposed in a comparatively
short space of time, and that the presence of the alkali serves to
facilitate it. The slowness of the generation of ammonia in
tube IV is of especial interest when brought in relation with
the contents of the retorts of the first series. Here, also, when
other tests for free ammonia are not available, a little of the
solution of rosolic acid on a cotton plug, held over the mouth
of the vessel, is a delicate reagent; care must naturally be
taken that the plug does not touch the tube, nor that it be
exposed to the air too long, for the smallest traces of atmospheric
ammonia will show the colour reaction.

With reference to the change of the reaction, attention ought to
be called to one retort, mentioned above, which contained or-
ganisms, and, notwithstanding, showed acid reaction. This
retort was charged, July 11th, with urine of the acidity of eighty
drops, together with a liq. pot. tube, which remained intact. On
October 17th it was opened and examined, when the acidity was
found to be reduced to thirty-five drops of 5*85 percent, liq. pot.
Dr. Bastian has already called attention to the reduction of the
degree of acidity under similar conditions.

Notwithstanding the minute description which Dr. Bastian has
given of his method it is, as in all cases of experimentation with
fluids of unknown or inconstant composition, never possible to
repeat the tests and draw the same conclusions. The simpler the
pabulum, and the more we have it in our power to modify its
composition, the more uniform will be the results of cultivations
of organisms. I have, therefore, sought to study the question,
making use of Mayer^s solution,^ modifying it in such a manner
that, instead of adding tartrate of ammonia, I introduced urea
(1*06 grm. to 30 c. c.), after the solution of the other salts
had been boiled and filtered. After dissolving the urea the solu-
tion was again filtered. Although Cohn has found that this is
not a very good pabulum, I would, in reference to my results,
call attention to the fact that he himself has seen Bacteria
develop and multiply in a simple solution of the milk sugar
of commerce, and that he believes that the nitrogen was absorbed
from the surrounding atmosphere. With this solution a number
of test tubes were charged, sealed, and placed in the incubator
122° 1\ (40 — 45° C.). After sixty-five days all specimens were
turbid and swarmed with organisms ; the degree of acidity of
30 c. c. was reduced from 110 drops to 66 drops. It seemed

1 •! grm. of potassium phosphate, ’1 grm. crystallised magnesium sul-
phate, *01 grm. tribasic calcium phosphate to 100 c. c. distilled water.

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA. 197

that the solution was too strongly acid^ and the generation of
ammonia too slow in order to change the reaction, In the next
series a much weaker solution was employed. Eetorts charged
with the same, and prepared in the manner of Mr. Bastian’s urine
retort, i. e, with and without liq. pot., in proportion to the degree
of acidity. Here, also, the result was similar to that in the case
of the urine experiment. After ten to fourteen days the solu-
tions mixed with liq. pot., and after fifty- seven days those with
unbroken tubes all contained innumerable organisms.

As regards the forms of Bacteria which I found in all the
above-described series of experiments, the greatest number were
micrococci (Billroth), both in their characteristic movements as
separate individuals, or linked in varying numbers and at rest,
singly or in zoogloea colonies. The rods and bacilli were not so
numerous ; their contents appeared to be granular. In several
specimens I also found forms much resembling ascococcus (Bill-
roth). Here also I met with the zoogloea-like masses already men-
tioned, but am unable to determine whether or not they were
organisms, as they showed no kind of movement, nor could I
observe any multiplication of their elements.

It still remains to be proven that the multiplication of
organisms in these fluids, treated by high temperature, and in a
minimum of free and absorbed oxygen, is causally connected with
the decomposition of urea and the generation of ammonia, which
here seems to be the source for nitrogen.

The change in the following series of retorts, kept in the in-
cubator during forty -one days, will conclusively demonstrate that
ammonia takes such a role :

I. In dilute MayeHs solution and urea (3’0 grms., 200 c. c.).
The fluid remained clear, reaction alkaline, contained micrococci,
spheroids, rods, and bacilli.

II. In dilute Mayer's solution and urea, with forty drops
liq. pot. (5*85 per cent.) ; little sedimentation, fluid clear,
reaction alkaline, few moving micrococci and spheroids, together
with crystals and amorphous, granular masses. It appears
here as if the process of multiplication had already passed over.

III. In dilute Mayer's solution with dilute solution of caustic
ammonia, which was introduced in the manner as the liq. pot.
tubes in the first series, and then broken. Strong and diffuse
turbidity, reaction alkaline, immovable organisms (micrococci,
spheroidal forms, rods, and bacilli).

IV. In dilute Mayer's solution and urea, and a tube containing
fifty drops of sulphuric acid (5 per cent.), which was introduced
and broken as above. Fluid clear, low degree of acidity, here
and there a micrococcus.

Finally, a portion of the contents of retort IV was divided and

198

Dll. L. WALDSTEIN.

put into two test tubes, into one of which a stream of air charged
with ammonia was led, until the alkaline litmus reaction \vas
obtained, the other remaining intact. Both were then closed by
means of a cotton plug and examined after forty clays. The
contents of the second tube w^ere in appearance and reaction un-
changed, whereas the first was turbid and contained much
zoogloea, and literally swarmed with different forms of organisms.
It is worth noting that litmus paper indicated neutral reaction,
but rosolic acid solution gave a very strong alkaline reaction.

A review of the above^-described experiments shows, as regards
Dr. Bastian’s work, that in urine and the modified Mayer^s solution
containing urea, micro-organisms are developed or multiply after
exposure to 212° B. (100° C.) for a certain length of time, after
evacuating the greater part of oxygen and placing the retorts in
an incubating temperature 122° F. (40 — 45° C.). But in the
series here recorded this effect was produced only after a com-
paratively long time. It has also been found that it is not safe
to draw conclusions regarding the presence of organisms in fluids
from their outward appearance, i. e. the clearness or turbidty of the
same, as the suspensions and sediments are frequently composed of
other bodies, and as perfectly clear fluids often contain numberless
organisms. Dr. Bastian found that micro-organisms appeared
alone in such of the retorts as held urine mixed with liquor potassae,
and that after eighteen to thirty- six hours, w hilst the urine un-
mixed invariably remained apparently unaltered. My repetitions
of his experiments have led to a different result ; in all specimens,
irrespective of the admixture of liq. pot.. Bacteria were developed,
but only after a much longer sojourn in the incubator. The
fact that they appeared sooner w'here the alkali had been added
is explained without difficulty by the accelerated decomposition
of urea and generation of ammonia, which provided the necessary
nitrogen ; in so far, it is true, liquor potassse acts as a favorable
factor.

As long as a doubt can be entertained regarding a per-
fect sterilisation, so long is it inadmissible to make use of these
results as arguments in favour of abiogenesis.” I am inclined
to consider these researches of some interest, inasmuch as the re-
tardation of a development of low^r forms or so-called germs
w'hich may have remained in the vessels can be naturally ex-
plained by the high temperatures and exclusion of free oxygen,
wdiilst the multiplication of organisms can be understood w^hen the
incubator temperature and the generation of a simple nitro-
genous body are considered as favorable factors. Urea may not,
according to Cohn, be such a body, but when ammonia is gene-
rated in presence of the necessary inorganic substances Bacteria
can multiply even without an unlimited supply of free oxygen.

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA., 199

II.

With a view of arriving at a more definite knowledge of the
effects of an addition of ammonia to a solution of salts free from
nitrogen, I proposed next to study its influence under the
microscope, introducing this method by the following more ele-
mentary series of experiments.

A number of retorts were prepared like those of Dr. Bastian,
being charged with Mayer^s salts solution, minus the vehicle for
nitrogen, which latter was represented by twenty drops in each of
a dilute solution of ammonia (ten drops caustic ammonia, 100
c. c. distilled water) in sealed tubes, which were afterwards
broken and their contents mixed with those of the sealed retorts.
The result proved to be, in numerous repetitions of the same
procedure, that after remaining about seven days in the tem-
perature of the laboratory the fluid contained numberless organ-
isms of various forms.

Continued observation under the microscope was facilitated
by a simple chamber consisting of a glass slide, upon which
was fastened by means of Canada balsam a square of hard rubber
with a central boring measuring 4 mm. in thickness. On each
of two opposite sides a fine metal tube was let into the square in
such a manner that a stream of air could pass through the
chamber. A drop of the salts solution, minus nitrogen, having
been brought on a thin cover glass, which was held down by a
thin film of pure olive oil on the upper surface of the rubber
square, the field of observation was now in position, the drop
hanging in the chamber, on the bottom of which a drop of dis-
tilled water presented evaporations. A rubber tube of small
calibre is drawn over one of the metal tubes and connected with
a AVoulff^s flask in such a manner that a stream of ammoniated
atmospheric air can be gently forced into the chamber. In
order to prevent aspiration when the afferent rubber tube is
closed, a short piece of rubber tubing is attached to the opposite
metal tube, whose distal end, provided with a glass mouth piece,
is kept plunged under water.

With this simple apparatus, which can easily be kept scru-
pulously clean and can be taken apart when necessary, the
development of micro-organisms and its modification when
ammonia is provided or withheld can be observed for many days
in succession. Here, also, a small cotton plug soaked with a
little rosolic acid aiid applied where the current leaves the
chamber is of value in determining whether or not ammonia
passes through the chamber.

The results of my observations have led me to believe that the
first stage in the development of micrococci, rods, and bacilli, as

VOL. XX. NEW SER. O

200

DR. L. WALDSTEIN.

well as of gloeo-coccus or gloeo-bacteria (Billroth) is represented
either hy what could only he distinguished as an extremely small
speck in the field or of spheroidal bodies much resembling small
fat-globules in appearance, but not in their chemical nature, so
far as could be determined by means of chemical reagents. In
one case the smallest speck assumes lively motion, in another it
remains where first seen, and little by little a conglomeration is
formed of many like bodies ; I was unable to determine whether
this was a process of division. In another part of the field a
spheroidal body is gradually elongated, and finally a bacillus is
found in its stead, and presently a thin septum is found to part
it into two individuals, and so on until the chain of bacilli
extends in some cases from one end of the field to the other.
When the zooglcea colonies have reached a certain size they sink
to the bottom of the drop. I have found that in those parts
nearest the under surface of the cover glass, micrococci and
zooglcea generally develop most abundantly, whereas bacilli are
most numerous in those parts of the drop nearer to the atmo-
sphere of the chamber. Other globular bodies send out pro-
cesses of comparatively considerable thickness, which are often-
times drawn in repeatedly or after assuming a knob-like shape,
at their extremity are gradually separated from the mother cell ;
this mode of prolification closely resembles that of yeast cells,
but careful measurements have shown that the organisms in
question are smaller than Torulee.

When the rubber tube was compressed during seventeen
hours, and in this manner ammonia and a fresh supply of atmo-
spheric air was withheld, no bacilli were to be seen in the fluid,
only micrococci globules and short rods; but after allowing the
ammoniated air free access, and gently forcing it through the
chamber, bacilli appeared after twenty-four hours; the move-
ments of the micrococci and of the rods wholly ceased, however,
when the metallic clamp was allowed to compress the afferent
hose for a longer period of time. The movements of the bacilli
and of the rods were such that the body proper remains un-
bent ; it is, therefore, very natural to suppose that they are sup-
plied with a ciliary appendix; I was, however, less fortunate
than several other investigators, and could not discover them.
When a second Woulff^s flask, containing distilled water with
a small known quantity of carbolic, acetic, hydrochloric acid or
camphor was connected with the apparatus, so that the current
of air passed through these liquids before reaching the chamber,
the development of Bacteria was seen to begin only after at
least five days, and only a few bacilli appeared within a week.
During this time the current was made to pass gently through
the chamber every two hours in the daytime, just as in the

A CONTRIBUTION TO THE BIOLOGY OF BACTERIA. 201

operation with ammoniated air above mentioned. Here the
rosolic-acid cotton plug was not reddened, no free ammonia
reaching the chambers. As a control experiment, the same fluids
were added to solutions contained in test tubes, and which
swarmed with organisms. No diminution in the violent move-
ments of the Bacteria ensued, nor was their number apparently
diminished. It is clear, therefore, that the vapours of these
so-called antiseptic agents were not the cause of the retardation
in the development or multiplication of organisms. That the
organisms appear at all, but only comparatively late, is easily
understood when it is borne in mind that Cohn observed in a
simple solution of milk sugar a multiplication of Bacteria which
derived their nitrogen, according to his hypothesis, from the
atmosphere, to which fact I have already called attention, whilst
Schonbein found that when water evaporates in atmospheric air
nitrogen is made free.

The success in practice of the treatment of wounds according
to Mr. Lister^s method is now very generally admitted, but the
hypothesis on which this method is based is not so universally
accepted. My experiments are, in this respect, of some interest,
but I am not willing to claim for their results any very general
application. Observers of high rank still assert, after careful
researches, the ubiquity of organic germs, and it is clear that
such a view cannot be logically made to harmonise with the views
held by Mr. Lister and his more ardent disciples.

The antiseptic properties of the substances which I have used
being conceded, I hold that they are harmful to the development
and multiplication of organisms from the readiness with which
they unite with substances indispensable to the existence of the
organisms rather than on account of their direct antizymotic pro-
perties. I freely grant that these researches do not conclusively
])rove this supposition, but they seem to indicate such a relation.

202

E. A. SCHAFER.

So?ne Teachings ^Development. By E. A. Schafer, E.B.S.,
Eullerian Professor of Physiology.^

I.

Development is the term applied to the process of evolution of
an organism from its simplest to its most complex phase of
existence.

In all but the very lowest animals the development of the
individual begins with the ovum or egg, this being the elemental
portion or cell of the parent animal which is set aside for the
special purpose of reproduction. In most cases the separated
cell is incapable by itself of reproducing the parent ; it must first
blend with a separated part of another parent. This blending
with the ovum (female reproductive element) of the anthero-
zoid (male reproductive element) constitutes the process of
fertilisation.

The commencement of development follows immediately upon
the completion of fertilisation. Once commenced the pr cess is
usually continuous, and being accompanied by a general increase
in size, the result is the perfect or adult individual. Although
these two processes (development, ^. increase of structural and
functional complexity, and growth, i. e. increase of size) generally
go hand in hand, we must be careful to distinguish between
them. Eor mere increase of size may often give the appearance
of greater complexity, when in reality the organism has not
advanced either in function or in. intimate structure. This is
the case sometimes when the increase of size takes the form of
budded outgrowths which are merely repetitions of the structure
of the parent organism. The common freshwater polyp con-
stantly presents us with an illustration of this, for its numerous
buds produce an appearance of complexity which disappears
when the buds detach themselves as independent organisms. In
the compound hydroid polyps, in which the buds remain attached,
the appearance of complexity remains throughout the life of the
organism, and is accompanied in some by an actual progress in
functional and structural complexity.

In the case just alluded to the increase of size accompanying
development takes the form of an arborescent repetition of all the
chief parts of which the developing individual is composed. But
it may instead take the form of a serial repetition such as we see

* This article contains the substance of the last two of a series of
twelve lectures on Animal Development, delivered at the Royal Insti-
tution in Jan., Feb,, March, 1879 ; and has not been modified since that
date.

SOME TEACHINGS OF DEVELOPMENT.

203

in the Strobila stage of development of the larger Medusse^ and
which presents itself in striking contrast to the arborescent repe-
tition of the corresponding stage of development of the Craspe-
dote Medusge. In the so-called Strobila the developing indi-
vidual has become partially separated by transverse constrictions
into a series of perfectly similar, saucer- shaped segments, each
of which comprises within itself in a rudimentary form, a repre-
sentative of all the parts of the originally simple parent or-
ganism. So long as these remain united the Strobila is a com-
pound individual constituted by a serial repetition of similar
parts ; but they soon prove their individuality by breaking away
from the compound body and from one another, and maintaining
existence as independent organisms.

This serial repetition is seen in still more characteristic form
amongst the Annelids, where almost every one of the many seg-
ments which compose the body is complete within itself, and
any one might be taken as a representation of the primarily
simple embryo which produced them all. And viewing them by
the light of the illustrations furnished us by the Coelenterata, it
seems not unreasonable to regard all animals in which such serial
repetition is found as compound rather than as simple indi-
viduals— to a certain extent comparable to the condition of the
united polyps just referred to, but far more closely joined together
and interdependent ; rendered thereby incapable of maintaining
themselves as separate organisms, but, on the other hand, able
the better, by their united powers, to carry on the struggle for
existence. And the importance of this serial repetition is indi-
cated by the fact that it is found as the main characteristic of
the two most highly organised branches of the animal tree — the
Arthropoda and the Yertebrata, masked though it often is in
them by that tendency to recoalescence which at first sight seems
a recurrence to more simple types of existence, but which, when
attentively considered, proves to be merely the expression of the
closer union and combination of the separate segments, the
better to work together for special objects or for their mutual
benefit.

I have stated that, once commenced, the process of develop-
ment of an animal is a continuous one. Nevertheless, it is
customary to regard it as advancing by separate steps or stages
— to look upon it as a staircase or ladder, the stairs or rungs
of which are successively attained, and the number of which
depends upon the complexity of organisation of the adult indi-
vidual. The familiar phrases “scale of development,” “scale
of organisation” (Latin, scalm, a staircase), are a suflicient in-
dication of the mode in which the subject is generally viewed.
And although I think that the continuous nature of the process

204

E. A. SCHAFER.

should on no account be lost sight of^ yet it is so manifestly
convenient for purposes of description and of comparison to
maintain the idea of grades, that I shall here adopt a like simile,
choosing a succession of telescopic tubes, instead of a staircase,
to represent the succession of developmental phases, because
those can be imagined to be more or less drawn out, and shifted
upon one another, so as to illustrate the varying relations which
the typical successive phases bear to one another in the develop-
ment of different animals. It will further be necessary to imagine
the tubes constituted of some very elastic material, rendering it
possible for them to be extended to two, three, or many times
their proper length, or, on the other hand, to become shortened
so much as to be scarcely perceptible. Each tube of which the
telescope is composed may be taken to represent a particular
stage of development, and the greater the complexity of the
organism the more tubes must be added to the series, and the
greater the length of the developmental telescope.

If we now proceed to compare the development of various
animals with one another, we are struck with the resemblance —
with the identity I might rather say — of the first stages in all.
To begin with all take origin in an ovum ; and although at the
first glance this appears so different in many (compare the ovum
of a mammal with the egg-yolk or pvum of the bird), exami-
nation shows it to be in all cases similar in essential composition.
For the difference depends solely upon the amount of nutritive
material which has been stored up within and in connection with
the reproductive protoplasm, and has relation to the greater or
less facility with which a supply of food can be obtained from
without during the active developmental process. To use a fa-
miliar comparison, the difference is precisely the same as that
between an army operating in a fertile country where supplies are
readily obtainable, and a force which is to operate in a region
barren or devastated, and far removed from the base of supplies.
In the one case pabulum sufficient merely for immediate use is
required, in the other a great store of nutriment must be accu-
mulated and must accompany and proportionately hamper every
movement of the army. Moreover, since to a part of the force
must be entrusted the duty of guarding the store and distributing
it as required, a greater number of men is required to bring the
army up to the same proportionate strength as that of the unen-
cumbered force. So in those ova which are provided with a
large store of nutritive material, do we find that when the active
movements which characterse development commence, those
movements are hampered by the presence of the inert food-ma-
terial, and it is precisely as the store becomes lessened that the
apparent difference between such ova as these, and those others

SOME TEACHINGS OF DEVELOPMENT.

205

which from the first have scarcely any store of nutriment, dis-
appears. The difference is one of degree only and not a difference
in the essential nature of the ova ; this view is confirmed by the
existence of every shade of transition between them, as well as by
observing that the difference obtains in the ova of animals which
are otherwise closely allied.

Further, in the ovum of all animals there follows immediately
on fertilisation the successive division or cleavage of the single
cell into first two, then four, and so on until an indefinite number
of cells is formed which arrange themselves around a central space
— the cleavage-cavity. This, in the majority of cases, is distinct
enough, being occupied merely by clear fluid and enclosed by a
single layer of cells; but in some instances it is partially or
wholly filled up by the excess of food-material and its immediately
attendant protoplasm, which may or may not, according to its
relative bulk, have participated in the active changes which have
been occurring in the rest of the protoplasm, and of which the
cleavage is the result. To revert to our military comparison it
is as if the two armies had broken up into detachments and
arranged themselves so as to enclose circular areas of ground ;
in the one case the detachments unencumbered and active on all
sides, with the enclosed area left clear ; in the other case the
materiel^ with the camp followers, packed away in the centre
and partly occupying one side of the space, whilst the least
encumbered detachments are chiefly accumulated on the side
where most activity is required.

Bearing in mind this difference, and regarding it as accidental,
we may take the completion of the cleavage process, as far as
the arrangement of the resulting cells around a central area in
the manner described, as the first stage in the development of
all animals. It may be termed the stage of the unilaminar
blastoderm^ or the hlasto sphere. For the name blastoderm has
long been applied to the cellular membranes of wdiich the devel-
oping ovum is formed at early periods ; and the term blasto-
sphere may be used, because at this stage a typical ovum, i.e.
one wuth a minimal amount of nutritive material, consists as we
have seen of but a single uniform layer of cells forming the wall
of a hollow sphere.

We meet with this typical condition of the first stage in Gas-
trophy seraa (according to Haeckel), in Phallusia, and in Am-
phioxus and other animals. But even in those ova w^hich con-
tain only a slight excess of pabulum we find that there are
differences observable in the cells which form the wall of the
blastosphere, for the food-material becomes chiefly collected at
one part of the layer, rendering the cells at this part more
granular and larger than the others. Tliis difference may be so

206

E, A. SCHAFER,.

slight as to be almost imperceptible, as in the Holothuriaii
blastosphere, or it may be more obvious as in that of the
Sponge and Paludina. And every transition is found between
these cases and those in which the amount of pabulum is so
great in proportion that it can no longer be contained in the
typical single layer of cells, but becomes accumulated in the
cavity of the sphere as well, either still enclosed in definite cells
as in the Amphibian, or for the most part unenclosed in cells as
in the Pish, the Bird, and the Crustacean.

The next most important change that occurs in the typical
unencumbered ovum is the invagination of one part of the simple
wall of the blastosphere, in such a manner as to convert the
hollow single-layered vesicle into a cup with double walls. The
cleavage- cavity or cavity of the blastosphere is concomitantly
reduced in size or even obliterated as it is encroached upon by the
invaginated part of the wall, and its place is taken by the cavity
of the cup. The orifice of this, at first widely open, becomes
gradually closed up by the continued growth of the two layers
at the mouth of the cup, until but a narrow aperture remains,
which persists for a time, but at length in many cases becomes
closed. This aperture is termed by Haeckel the primitive mouth
or protostoma, by Lankester the blastopore.

The result of the completion of this process of invagination is
again a hollow vesicle, but its walls are now composed of two
layers of cells instead of one. The inner or invaginated layer is
termed entoderm and the outer or enclosing layer ecto-
derm.’^ The cavity which they enclose always becomes the
future alimentary cavity, and hence the name gastrula ” has
been applied by Haeckel to this stage of development. We may
also speak of it as the cup^ stage or as the stage of the hilaminar
blastoderm.

In typical ova the original blastosphere is uniform through-
out, and it would be impossible to point out the part of the wall
which is to be invaginated. But in those ova which have even a
small excess of nutritive material and in which therefore, as
already mentioned, some of the cells of the blastosphere are
characterised by containing this excess, we always find that it is
this particular part of the wall of the vesicle which is involuted
to form the inner layer of the cup. So that even in the stage of
the blastosphere we can predict which cells are to become ento-
derm and which ectoderm.

Finally, in those ova in which the nutritive material largely
}>repon derates we find that even when the cavity of the blasto-
sphere is not entirely filled up by that material (and of course
also when it is so filled up), the mechanical hindrance to any
invagination is so great that the process cannot be e|Fected»

SOME TEACHINGS OF DEVELOPMENT.

207

Still, even in such cases as these the cup-stage is not absent.
The same result is obtained, but in a different way. Tor the
enclosure of the pabulum-containing entoderm by the ectoderm
is effected solely by the growth of the latter, without any accom-
panying invagination of entoderm. In these cases the cavity of
the gastrula also is necessarily occupied by the pabulum, and
this if greatly in excess may even project beyond the cup-orifice.

We may then take the gastrula or cup-phase as a second stage
and note that, like the first, it is met with (modified only by the
accidental presence of food material) in all animals above the
Protozoa.

There are two apparent exceptions to the general rule of for-
mation of the gastrula stage by invagination. These are met
with in the development of the freshwater polyp (Hydra) as
described by Kleinenberg, and that of one of the Medusse (Ger-
yonia) as described by Metschnikoff and by Pol. In these
animals the bilaminar blastoderm is said to be formed by the
splitting into two portions, inner and outer, of some or all of the
cells of the single layer which forms the wall of the blastosphere,
the inner parts becoming collectively the entoderm, the outer
remaining as ectoderm. Eegarding these exceptions the ques-
tion naturally suggests itself. Are they due to defects of obser-
vation ? and further, supposing that there is’no flaw in the facts,
are they nevertheless explicable as modifications only in the
ordinary mode of formation of the gastrula ? Certainly as far as
concerns Hydra, the recorded observations of Kleinenberg are
too incomplete on this part of the development for his conclu-
sions to be accepted without demur. And it is very generally
admitted that the development of Geryonia needs careful rein-
vestigation with the aid of microscopic sections. It is not
improbable that the result of such reinvestigation might prove
the existence of some sort of invagination at a much earlier
period than that at which the so-called delamination or splitting
occurs.

The cavity of the cup at first communicates, as we have seen,
with the exterior, by means of the aperture of invagination or
protostome, and when this becomes closed, or even before its
closure, another aperture appears, and in most cases yet another.
These secondary openings into the cavity of the gastrula become
respectively the anterior and posterior orifices of the alimentary
canal, and their formation is always accompanied by an ingrowth
of ectoderm towards the endoderm. One or other of these aper-
tures may be formed in the place which was before occupied by
the now obliterated protostome, or in some cases the latter may
itself remain persistent as one of the secondary orifices.

\\ e may now pass to the consideration of the third essential

208

E. A. SCHAFER.

forward step in development. This is the separation or segre-
gation of some cells from either one or both primary layers to form
an intermediate set, which acquires greater proportional import-
ance the further upwards we trace it in the scale of organisation.
Since the cells of this intermediate set in most cases, and certainly
in typical ova, are not of independent origin, but are derived in-
directly from one or other of the two primary layers, it is clearly
a mistake to regard the layer which they form as of equal im-
portance with the other two — an idea which is distinctly implied
by the name mesoderm’^ usually given to it. The two primary
layers — the foundation membranes” of Huxley; ‘^ectoderm

and endoderm” of Allmann — are distinct and well defined. The
so-called mesoderm is often very different in these respects, so
much so that in some animals parts which are looked upon as
mesodermic by some morphologists are regarded as still belonging
to the two primary layers by others.

This mistake of regarding the mesoderm as of equal morpho-
logical importance with ectoderm and entoderm has originated
in the study of the development of the higher animals, in which,
in many cases, the relative time of its origin has become shifted,
so that it may begin to appear almost simultaneously with the
primary layers. And even in some animals, not very high in the
scale, the same shifting may be found. Thus, for example, in
the Holothurian, the separation of some of the ectoderm cells to
form part of the intermediate or mesodermic set has already
begun, even before the commencement of invagination, and, there-
fore, whilst the general development has not advanced beyond
the blastophere stage. And the same shifting is carried to a
still greater extent in Clepsine, where we find that even when the
cleavage of the ovum has advanced but a step or two, certain
portions are separated to produce, as the cleavage process con-
tinues in them, the whole of the mesoderm. In this case,
then, the third stage has begun almost as soon as the first stage
itself.

This premature separation or precocious segregation (Lan-
kester) of parts which analogy would lead us to expect later is
a very common feature in animal development. It is illustrated
in Clepsine in a still more remarkable way, in the premature
separation of the portions of the dividing ovum from which the
cells of the nervous system are derived. And it is on account of
this tendency which the phases of development exhibit to become
shifted to periods earlier than usual for their appearance that I
have compared them to the tubes of a telescope, for we can so
much the better illustrate them in their relations to one another.
Thus, in the case of the Holothurian, we should shift the third
tube representing the formation of the mesoderm downwards, so

SOME TEACHINGS OF DEVELOPMENT.

209

as to overlap the second one representing the gastrula, and in
the case of Ciepsine so as almost to reach the base upon which
the telescope rests.

II.

At the end of the previous lecture we had arrived at the con-
sideration of what we, for convenience, regarded as the third
stage in the developmental staircase, the third tube in the de-
velopmental telescope — the formation, namely, of the mesoderm ;
and I insisted that the term is a misnomer, because it obviously
places itself side by side in the mind with the names which have
long been applied to the two primary layers, ectoderm’^ and
entoderm,” whereas it is secondary to these, being derived
from them. The mesoderm consists, in fact, of cells which
have been separated or segregated from amongst the cells of the
primary layers for the performance of special functions. In the
lowest animals in which such a separation occurs it may take
place at any part of the primary layers, or even over the whole
of their extent. We see this in the sponges and jelly-fish. In
these, when intermediate cells are separated, they lie in a jelly-
like substance, and the main purpose which the segregation sub-
serves is that of support and connection. Now, it is obvious
that this sustentacular function might be almost as well per-
formed by the inert jelly alone — in fact, the soft protoplasm of
the cells can be of but little assistance, so that we should natu-
rally look upon the cells in this primitive mesoderm rather as
ministering to the nutrition of the jelly than as agents in the
performance of its function. This view is strengthened by ob-
serving that it is the intermediate jelly-like matrix of this primi-
tive connective tissue which is the first to appear, the cells (when
they do occur, for in many cases they are absent throughout life)
wanderiog into it subsequently. In the larger Medusae (Aurelia)
they come, according to the testimony of most observers, from
the entoderm ; in the Sponge they are derived from the outer of
the two layers which are found in the Olynthus stage — the one
which is generally regarded as ectoderm.^

^ There seems to be some uncertainty about the interpretation of the
two layers of cells which are found in the Olynthus stage of development
of the Sponge. According to the descriptions of various observers, and
especially that given by F. E. Schulze with regard to Sycandra raphanns,
the segmentation of the ovum takes place, and a blastosphere becomes
formed much in the usual manner. Of the cells of this blastosphere
those of one hemisphere are smaller clearer, and provided with long cilia,
those of the other and lesser hemisphere being larger, more granular,
and without cilia. Presently the latter become invaginated, but the
cupping thus produced proves a temporary condition merely ; fluid again
accumulates in the cavity of the blastosphere and obliterates the cup-

210

E. A. SCHAFER.

Another proof, if one were needed, that the jelly is the pri-
mary and the cells which wander into it the secondary part of
the purely sustentacular mesoderm of these lowly organised
animals is to be met with in the fact that in some — e. g. Hydra
and the smaller Jelly-fish — the jelly is the only part of the layer
present. The nutrition of the jelly is administered directly by
the entoderm cells.

We see then that in the lower forms the only function which
is delegated to the intermediate layer is the mechanical one of
support. But in all the higher animals the segregated mesoderm
cells are deputed to perform other and more important functions
as well, for, besides the connective and supporting structures of
the body, the actively contractile tissues which are concerned in
effecting the movements of the body are derived from them. In
the Medusae or jelly-fishes this function is still performed by a
tissue which is undoubtedly part of the ectoderm, although at
first sight it seems to constitute a distinct layer of cross- striped
fibres beneath the ectodermal epithelium. For if these fibres
are carefully investigated, it will be found that they are many of
them placed in the interior of large cells which project to the

form. After a certain interval a more decided cupping again takes
place, but this time it is the clear ciliated cells that are invaginated. This
condition remains permanent. The cup-like sponge settles down and
the orifice of the cup becomes closed; a jelly-like substance accumulates
between the two layers of which the cup is formed, and cells pass into it
from the outer layer ; one or more canals become bored through the
jelly so as to effect a communication between the cavity of the sponge
and the external medium ; and the sponge may be regarded as formed.
Now it is generally assumed that the first cupping being temporary is a
mere accident, and that the second represents the gastrula stage of other
animals. If this is really the case there is of course no fault to find with
the generally received interpretation which sets down the large cells of
the ciliated chambers of a sponge as entoderm and all the rest of the
sponge (viz. the flattened cells which cover the external surface and line
the water-canals, and the general thickness of its jelly with its included
branched cells), as ectoderm (and mesoderm). But it is difficult to avoid
the thought that by ignoring the first cupping and setting it down as
accidental or at least as unimportant on account of its transciency, an
error may have been made and a divergent if not a retrograde process
of development regarded as a normal and progressive "one. It is
admitted that the microscopic appearance of the ciliated hemisphere of
the sponge-blastosphere is in favour of its being looked upon as ectoderm
rather than entoderm, and conversely with regard to the non-ciliated
hemisphere. The origin of the branched cells in the jelly from the non-
ciliated cells is a fact pointing in the same dii’ection. And although our
knowledge of the physiology of the sponge is too imperfect to be of much
service in deciding the question, nevertheless the vegetative mode of life
which sponges exhibit would lead one to expect a proportionately in-
creased entodermal and diminished ectodermal development.

If the conjecture thus sketched out is a sound one it follows that
sponges, as compared with other animals, are turned inside out,

SOME TEACHINGS OF DEVELOPMENT. 211

surface between the cells of the general layer of ectoderm, which
covers the under surface of the umbrella, and they evidently
belong to that layer. In other words, the deeper parts of some
of the ectoderm cells are modified to form the muscular struc-
tures. In Hydra, according to the descriptions of Kleinenberg
and Korotneff, this is still more clearly the case — in fact, almost
all the large superficial cells of the ectoderm are thus modified
in their deeper part, although the muscular development is less
characteristic. Now, supposing these muscular cells of the
jelly-fish, in place of remaining amongst the rest of the ectodermal
cells, to sink deeper into the surface of the jelly, or even to
become imbedded in its substance, they would then appear
altogether distinct ; and although originally derived from the
ectoderm, would be reckoned as part of the mesoderm. This is,
indeed, the condition which is actually found in many other
Coelenterates.

We have seen that in the lower forms of the Metazoa the
cells of the intermediate layer are derivable from almost any
part of the primary layers. But in all the higher forms the
mesoderm is developed from one part only of those layers, and
this part is very frequently close to the contracted orifice of the
gastrula, at the place where ectoderm and entoderm pass round
into one another.^ And from this place of origin the mesodermic
cells — consisting of those which are to minister to support, and
those which are to minister to movement intermixed— spread out
between the two primary layers. We see a typical instance of
this in Paludina and also in Unio. Now, if we suppose that the
segregation appeared first at one particular point at the margin
of the protostome, and afterwards spread in all directions from
that point, we can comprehend how bilateralism might have
appeared as the result of the separation of mesoderm cells (acci-
dentally ?) at one part only of the margin of the protostome.
But whatever might have been the original conditions, the meso-
derm as we now actually see its development, appears from the
first in two halves, and bilateralism makes its appearance simul-
taneously with these.

Our third stage, then, such as we see it in all animals above
the Coclenterates, — the segregation, namely, of an intermediate

If we agree with Haeckel in surmising that in the earliest stage of
animal evolution in which the cup-form or gastrula appeared, the oriGce
functioned as a mouth, we can readily understand how it is that a
general mesodermic cell-segregation should have tended to localise itself
in the neighbourhood of that aperture. For this would probably, earlier
than other parts, become the seat of special active functions; e.g. the
opening and closing of the orifice or even the sucking in of aliment, such
as we see occurring in the early embryo of the earth-worm.

213

E. A. SCHAFER.

set of cells from the entoderm and ectoderm, destined to sub-
serve the functions of support and motion, and termed meso-
derm — has in reality in all probability been produced by the
coincident occurrence of at least two distinct segregations. Per-
haps one of these, the muscular, is precocious, and has blended
with the other, the sustentacular, which should have preceded it.
If this is the case each of the two segregations should be re-
garded as constituting a distinct stage in development. But they
are so completely blended in their origin in the higher animals
that it is impossible to differentiate them.

There is no reason to suppose that the two layers into which
the mesoderm subsequently splits, in animals in which a body-
cavity, or coelom becomes formed, have anything to do with this
supposed primary double segregation of the layer. Por the
segregated cells are entirely intermixed as the mesoderm spreads ;
although they may afterwards become again partially separated in
groups (for the constitution of muscles, cartilages, &c.), according
to the function for which they are destined. The formation of
the coelom is a distinct forward stage in development, and is the
first step in the direction of the formation of a circulatory
system. The cells which bound both the coelom and its offsets
are in most animals segregated from the general mesoderm, and
belong to the set of sustentacular cells, but in the Holothurian,
according to the description given by Selenka (and also in
Sagitta and Amphioxus as shown by Kowalewsky), they are
derived directly from the entoderm of the alimentary cavity, and
already before their severance from this, enclose a commencing
coelom. There is not sufficient ground for regarding these lining
cells of the coelom and vascular system as constituting a special
segregation. In vertebrates they are certainly of the same
nature as those of the sustentacular part of the mesoderm (the
connective tissue), and the same is probably the case in the other
animals where they are found.

Another well-defined forward stage in development is the
occurrence of a special segregation of ectodermic cells for the
performance of the nervous function. These differ from those
which are destined for the muscular function in that they are
never blended in their origin with the mesoderm, and indeed do
not in any animal lose their primitive connection with the
ectoderm until development is comparatively advanced, if even
they do so then.

As in the case of the muscular segregation the first indication of
a separation of some of the ectodermic cells for the performance
of nervous and special sensory functions is met with in the
Coelenterata. This takes the form of a prolongation of the
attached ends of some of the ectoderm cells into branched

SOME TEACHINGS OE DEVELOPMENT.

213

fibres, which interlace with those of neighbouring cells, and
probably serve to convey any impressions received by the cells
of which they form a part, to deeper lying muscular cells which
have lost their place in the superficial ectodermal layer and their
connection with the external medium.

The nerve-epithelium cells thus formed may themselves
tend to sink below the general surface of ectoderm. If this
happens they lose the characteristic shape of epithelium cells,
and become rounded, with extensions in the direction of the
nerve-prolongations. In fact, like the muscular tissue, they
assume a minute structure which is similar to that of the nerve-
cells and nerves of the higher animals. Nevertheless these
nerve-cells in the jelly-fishes in no case separate themselves
from the ectodermal layer. They lie, in fact, between it and the
muscular layer in those parts in which the latter occurs.

And the sense-organs in this branch of the animal kingdom
also show that their essential place of origin is from the ecto-
derm. For visual organs first make their appearance as patches
of ectoderm-cells filled with coloured pigment, and some of
them with nerve-fibre processes connecting, them with adjacent
nerve-epithelium cells ; auditory organs as ectoderm-cells, con-
taining crystals in their interior, and similarly connected; and
olfactory organs as little pits lined by ciliated ectoderm-cells,
and connected likewise to a nerve-epithelium.^

Amongst other animals we observe that in Sagitta also the
segregation of ectoderm to form a nervous layer is in the first
instance general and not localised. Eventually this segregation
becomes accumulated mainly in two situations to form the cephalic
and abdominal ganglia. In Amphibia, too, the separation of
ectodermic-cells to form a nervo-sensory layer is at first general.
But in most animals, e, g. the Earthworm, Euaxes, Ascidia,
Amphioxus, the segregation in question begins at one part only
of the edge of the protostome, as is the case with the meso-
dermic segregation — this situation having been possibly deter-
mined by the high functional importance of this orifice (assum-
ing as before with Haeckel that it originally served as a mouth).
Extending from this situation, the nervous separation would no
doubt be chiefly guided by the arrangement of the pre-formed
muscular segregation, and would hence tend to assume a bilateral
condition such as we see it to possess.

^ It is the opinion of Professor Claus that the depression which is
seated above the base of each lithocyst in Aurelia and allied forms of
Medusm represents an olfactory organ. And the comparative researches
of l)rs. O. and R. Hertwig have rendered it more than probable that
the otoliths of the jelly fish even when they appear, as in Aurelia, to be
connected with entoderm alone, are originally derived from the ecto-
derm.

214

E. A. SCUAFER.

We also find that cells which are set aside for the sole per-
formance of special functions invariably become modified in the
arrangement of their living substance, and in some cases also
in its chemical nature. The structural changes which thus
accompany and indicate the assumption of any special function
by cells, constitutes the science of histology. It would be
carrying us too far from our immediate subject to enter into a
detailed account of the nature of these structural changes here,
so I wiU only point out two facts wdth regard to them. One is
that the special modifications of^structure which the cells assume,
that have been set aside for the sole performance of special
functions, are similar throughout the animal kingdom in all
essential features. The second is that in the lowest types
where they occur the structural characters appear before the
segregation of the cells is complete, whilst on the other hand
in the development of an individual belonging to the higher
types such segregation may have been long effected before either
the functions or the accompanying structural changes in the cells
begin to be manifest ; another instance of premature segregation.

We have thus far been speaking of the separation of special
sets of the cells of which a simple organism is composed for
the performance of special duties. But in the united or com-
pound organisms, before considered, ;t very frequently happens
that one or more of the units of which the compound body is
composed becomes altogether specially modified for the perform-
ance of one duty to the exclusion of others. Thus we see that
some of the individual buds, of which the compound organism
of a hydroid polyp consists are adapted for purposes of prehen-
sion and alimentation, whilst others are adapted solely for
reproduction. This localisation of function in the different buds
of a compound organism is carried to an extreme degree in the
Siphonophora — animals which on the whole resemble the hydroid
polyps, but instead of being fixed to a rock at the bottom of
the sea, float near the surface of the water, looking like strings
of beautifully iridescent hyaline beads. In them we find the
specialisation of individuals to have proceeded so far that whilst
one of the innumerable individuals which have been produced by
budding from the original single one, is set aside to perform the
purely mechanical purpose of suspending the united colony near
the surface of the sea, and is changed into a minute balloon
enclosing a bubble of air ; others are deputed to propel the
organism through the water and become wholly transformed into
so many pulsating bells ; others again have merely to receive
tactile impressioiis from objects in the external medium, and are
chiefly metamorphosed into long feelers ; others are occupied
solely with the seizure and ingestion of victims for the food of

SOME TEACHINGS ON DEVELOPMENT. 215

the whole compound organism, and mainly consist of a stomach
with a long stinging trailer appended ; others have confided to
them the reproduction of the organism ; whilst all the more
delicate of the thus modified individuals can shrink for shelter
under certain other of the members of the colony which have
degenerated into firm protecting scales. And in that higher type
of compound organism which is characterised by serial instead
of budding repetition, whole individual segments are often as
distinctly set apart for the performance of a special function,
although from their close union one with another it is generally
difficult to trace such complete specialisation. The extensively
found adaptation for prehensile and masticatory purposes ofi-ofie
or more of the anterior segments in so many of these organisms
readily comes to mind as an illustration of this.

But let us revert to the consideration of the sets of cells which
become segregated for special purposes from the primary blasto-
dermic layers. Once formed they are found to proceed through
stages of their own, varying in number with the extent of devel-
opment of the organs which they compose in different animals.
In one group of animals it will often be found that one or more
such segregations have progressed in their functional, and corre-
spondingly in their structural development much further than
others. In illustration of this we may notice that in the Yerte-
brata it is the nervous segregation, in the Arthropoda the mus-
cular segregation, which above all others has attained the
greatest development. What mammal, for instance, is capable
of the tenth part of the activity in proportion to its size, which
is evinced by the sustained flight of the dragon-fly or the pro-
digious leap of the flea !

As an illustration of the progression in development of a
special segregation, it will for our present purpose be most
instructive if we trace out the early stages of formation of the
nervous system in one of the highest types, and if we compare
those stages with what we find in animals lower in the scale.

In the Toad, which we may select as a typical Vertebrate,
there is at first a uniform layer of cells which is separated at an
early period from the ectoderm ; then a thickening of this layer
occurs on either side of the axis of the embryo, extending for-
wards from the protostome, the two thickenings forming the
boundaries of a groove which lies between them ; we next find
the groove becoming roofed over by upward extensions of ecto-
derm, both from behind the protostome and on either side, and
thus converted into a canal which is at first open in front and
communicates behind through the protostome with the alimen-
tary cavity ; next (or even before the roofing in is completed) we
observe that the anterior part of the nerve-tube becomes en-

VOL. XX. NEW^ SEIl. P

216

E. A. SCHAFER.

larged and subdivided to form the primary parts of the brain,
and finally, the wall of the tube becomes differentiated to form
nerve-cells and nerve-fibres.

If now we compare these stages of development of the nervous
svstem of the Toad with temporary or permanent conditions of
the same system in certain animads lower in the scale, we are
struck at once by the fact that the various stages described are
more or less represented by those conditions.

Without again referring in detail to the mode of formation of
the nervous system as a general segregation from the ectoderm,
which is met with in the Medusae and in Sagitta — for it might
perhaps be accounted too bold to attempt a comparison between
these and the earliest stage, that of general segregation, in the
Toad — we find that in the Earth-worm the nervous system com-
mences exactly as in the Toad, in the form of two thickenings of
the ectoderm which extend forwards from the cup-orifice and form
the boundaries of a shallow groove. But the development does
not proceed further in the same way as in the Y ertebrate, for the
groove is never converted into a canal.

In the Ascidian we observe as the first stage in the develop-
ment of the nervous system the formation of a neural groove,
with its boundaries of thickened ectoderm — so far, as in the
Earth-worm ; but the development proceeds a stage further. The
groove becomes gradually roofed in frdm behind forwards, forming
a tube which long remains open in front and is traceable behind,
through the protostome, into continuity with the alimentary
cavity. But ^though the anterior extremity becomes enlarged,
nevertheless the development of the nervous system in Phallusia
turns aside from the Vertebrate road, and passes through a series
of transformations which are special to the Tunicate type. Indeed
these are of such a nature that the tubular character of the nervous
system soon becomes no longer recognisable.

In Amphioxus exactly the same early stages are passed through ;
there is first the groove, then the enclosure of this to form the
neural canal, open at first in front, and communicating with the
alimentary cavity behind. But these apertures before long
become closed, and the neural canal shut off as a distinct simple
tube. In this condition it remains permanently, only that the
walls of the tube become much thickened, so that the ca\dty is
almost obliterated. Although there is no very apparent de-
parture from the path which we have seen that the nervous system
of the Toad takes in its development, nevertheless there is no
further progression ; the anterior enlargement to form the brain
never appearing.

This brief sketch is sufficient to show that the various stages
in the development of the nervous system of a Vertebrate are

SOME TEACHINGS OF DEVELOPMENT.

217

represented by transitory or even by permanent conditions of the
nervous system of a series of animals lower in the scale of organi-
sation. If we were to analyse in the same way the development
of any other specialised segregation of one of the higher animals,
we should find that exactly similar results would be arrived at.
Did time and space allow, it would be easy to trace the corre-
spondences of development in the case of the heart, of the central
axis or notochord, of the branchial slits, of the renal organs, and
so on. And if this correspondence of permanent conditions in
lower animals, with stages of development in higher animals,
is thus found to obtain to minute detail in the separate
and specialised parts, it stands to reason that it must be
found also in the aggregate which those parts compose. And
indeed, in many cases, even in the mere matter of external form,
the correspondence is often such as to strike the most ordinary
observer. Compare, for example, the polyp stage of development
of the jelly-fish with the permanent conditions of some of the
Hydroid polyps ; compare the Scyphistoma stage of Aurelia with
the permanent condition of Lucernaria ; compare the tailed larva
of Phallusia with the permanent condition of Appendicularia ;
compare the several stages of transformation of one of the higher
Crustacea with permanent conditions of the lower Crustacea ;
compare the various stages of the developing Amphibian with
either transitory or permanent conditions of AYorms, of Tunicates,
of Amphioxus, and of lower Yertebrates !

In conclusion I will attempt to formulate as briefly as pos-
sible, some of the general results at which we are able to arrive
from a consideration of the facts we have had before us :

I. If we compare the processes of development of any two
animals, from Sponges upwards, we find complete correspondence
up to a certain point ; from which point they may diverge from
one another. This point is sometimes placed near the bottom of
the development-scale, sometimes near the top ; or, it may be, in
any intermediate position.

II. Development is essentially localisation of function and
concomitant or consequent modification of structure ; such modi-
fication being accompanied by segregation of the cells concerned
with the function localised.

III. The path of development of all the more important of
these segregated parts is the same up to a certain point in the
development of each segregation. Trom this point it may, in any
animals or group of animals, diverge from the rest ; or may remain
stationary, whilst in the others, specialisation and modification
progress further.

IV. The various stages or phases of development of an animal,
as well as of its specialised parts, are often found to correspond

ns

E. A. SCHaFER.

with either permanent or transient conditions of animals lower
in the scale.

V. Since the phases of development of individual animals are
often seen to be representations of the permanent conditions
which are met with in a series of animals belonging to lower
grades of organisation^ it is impossible not to infer that these
successive phases in the development of the individual represent
similar phases in the process of formation or development of the
race to which the individual belongs. To revert to a former
simile,, we may safely say that the developmental telescope of
the individual is the same as that of the race, but with the tubes
shortened or shifted one upon another so that in many cases
their original order is no longer recognisable. The history of
the development then of any individual animal from the egg it
an abridgment of the history of formation in time of the race ;
or, to state the matter in as few words as possible, ^^developmens
represents descent.'’^

We conclude, therefore, that the ancestors of every animal have
successively exhibited structural conditions which are repre-
sented in a more or less modified form by the successive
stages of development of the individual. This is the only
logical conclusion to which the study of animal development
leads. Modifying slightly the words of Darwin I would say, to
take any other view^ is to admit that the structure of animals,
and the history of their development, form a mere snare laid to
entrap our judgment.^^

HISTOLOGY OF HYDRA FUSCA.

219

On the Histology ^ Hydra fusca.^ By T. Jeffery Parker,
B.Sc., Professor of Natural History in the University of
Otago.

The few observations I have to offer on this much-discussed
subject are partly confirmatory of, partly supplementary to, those
of foeinenberg f they present a certain agreement with those of
F. E. Schulze;^ while they are, in great measure, distinctly con-
tradictory of the later researches of Korotneff.^

1. The Ectoderm and the Muscular Layer, — The layer of
longitudinal fibres between the ectoderm and the endoderm was
discovered by Kolliker, who believed that each fibre was in direct
connection with an endoderm cell. Kleinenberg, in teased spe-
cimens, saw that the ectoderm cells tapered towards their inner
ends, and that each ^vas continued into a simple or branched
process, of precisely the same character as the fibres seen in sec-
tions : from this observation the important conclusion was arrived
at, that the fibres were in direct continuity with the ectoderm
cells, thus forming a sort of nascent mesoderm.

Schulze figures the elements of the middle layer as fusiform
fibres with somewhat jagged edges. Korotneff, following Kleinen-
berg’s directions as to methods of preparation, came to the con-
clusion that the ectoderm cells were expanded (elargie) at their
inner ends, and that each carried a fusiform refringent fibre,
attached by its middle to the enlarged base of the cell, and pro-
jecting beyond it in either direction, so that the cell appeared as
a lateral appendage {annexe) of the fibre, rather than the fibre as
a prolongation of the cell.

How M. Korotneff can have come to this conclusion as to the
shape of the ectoderm cells it is rather difficult to imagine ; by
any ordinary method of preparation it is perfectly easy to satisfy
oneself that the ectoderm cells of the body are, as a rule, markedly
distinguished from those of the endoderm by the tapering of
their inner ends ; and, in good specimens, that these ends are
continued into longer or shorter filaments.

The question of the exact relations of the fibres is by no means
so easy to decide. Any one working at Hydra for a week or two,
and using various methods of preparation, might readily frame a

* From the ‘ Proceedings of the Royal Society,’ No. 200,. 1880. The
Plate there given is not reproduced.

2 ‘ Hydra,’ 1872.

^ ‘ Ueber den Bau u. die Entwicklung von Cordylophora lucustris,’
1871.

■* “ Histologic de I’Hydre et de la Rucernaire,” ' Arch, do Zool. exp.,
t, V (1870), p. 309.

220

T. JEFFERY PARKER.

dozen different theories on this point, all equally supported by
appearances. But the matter seems to me to be entirely set at
rest by thin longitudinal sections of specimens preserved in am-
monic bichromate, which reagent usually has the effect of causing
a certain amount of separation between the layers. In such sec-
tions the ectoderm cells are distinctly seen to taper off towards
their inner ends ; the fibres to pass from them, at a sharp angle,
towards the endoderm, or, more correctly, towards the supporting
lamella ; and, in some cases, the fibres can be distinctly traced
into the attenuated extremities of the cells.

As to the true nature and functions of these structures. Dr.
Kleinenberg calls the ectoderm cell, with its filamentous process,
a neuro-muscle cell ; M. Korotneff prefers to name it an epithelio-
muscle cell; Professor Huxley^ considers that the fibres ^^are
solely internuncial in function, and, therefore, the primary form
of nerves.^^ This last view is rendered, to say the least, decidedly
improbable, by the great number and the regular disposition of
the fibres. It seems a priori unlikely that an animal devoid of
all muscular tissue should have a layer of close-set longitudinal
nerve-fibres throughout its whole body, while such an arrange-
ment is perfectly intelligible in a set of specially contrac-
tile filaments, developed as a means of rapid retraction of the
body.

The term neuro-muscular” implies, as Kleinenberg explains,
that the process only is contractile, the function of the cell itself
being merely to receive and transmit impressions. But, as Pro-
fessor Huxley points out, it is absolutely necessary to assume
contractility in the cell proper to account for the lengthening of
the body. The fibres merely have a special degree of contractility
assigned to them, in correspondence with the obvious advantage
accruing to the animal from the power of instantaneous shorten-
ing, the general contractility of the cells serving for extension,
this movement being, as observation of a living Kydra shows,
a comparatively slow one. The fibres must also be of use in the
characteristic looping^^ movements of the animal.

The simplest and most reasonable way of looking at these
structures is that adopted by Dr. Michael Poster, and illustrated
in the diagram at the beginning of the third chapter of his
' Text-book of Physiology.’ These show clearly enough that the
ectoderm cell of Hydra, with its muscular process, is the equiva-
lent of what, in the higher animals, becomes sensory cell, sensory
nerve, nerve cell, motor nerve, and muscle cell. So that a fairly
logical term might be made by combining Kleinenberg^s and
Korotneff’s, and speaking of epithelioneuro-muscle cell; but,
fortunately, it is unnecessary to employ any such cumbersome
^ * Anat. of Invert. Animals/ p. 64,

HISTOLOGY OF HYDRA FUSCA.

221

ternij and quite sufficient to speak of ectoderm cell with con-
tractile process.

The interstitial tissue, discovered by Kleinenburg, is quite
readily made out in all parts of the body except the proximal
end, where nematocysts are also absent. It is not mentioned by
KorotneflP, and, indeed, its existence would be impossible if the
large ectoderm cells had the shape described by him.

I have found no interstitial cells in the tentacles ; this would
seem to show that the ordinary ectoderm cells may also be the
mother cells of the nematocysts. The ectoderm cells of the
tentacles also differ from those of the body from the fact that
their nuclei are non-nucleolate, resembling, indeed, the nucleoli
of the body-cells, rather than their nuclei,

2. The Supjiorting Lamella. — This structure is clearly distin-

guished by Schulze and by Korotneff, the latter of whom, how-
ever, figures it^ as almost equal in thickness to the diameter of
an ectoderm cell ! Kleinenberg states that the muscular pro-
cesses are imbedded in a structureless cementing substance, and
that this, continued beyond the muscular layer on the endoderm
side, forms a layer — the Stiitzlamelle^^ of Eeichert — which can

sometimes be obtained as a separate structure.

This description by no means expresses the distinctness of the
supporting lamella. In specimens preserved in osmic acid or
ammonium bichromate, without subsequent treatment with alcohol,
it is easy, by teasing with fine needles, to detach shreds of con-
siderable extent, more or less free from attached muscular fibrils
and from cells of the interstitial tissue.

3. The Endoderm. — The ciliation of the endoderm is a question
about which there has been a good deal of discussion. iSchluze
figures a single fiagellum to each cell, as seen in optical section
of the tentacle. Kleinenberg was unable to demonstrate the
existence of fiagella in the uninjured animal, or in preserved
specimens, but in transverse sections of the living animal he ob-
served one or two cilia in connection with more or fewer of the
cells, and noticed that they were not fixed structures, but were
occasionally retracted, and then protruded again, the cells at the
same time sending out pseudopodial processes.

It is quite easy to confirm this observation ; the slow lashing
movement of the fiagelliform cilia their continual disappearance
and reappearance in fresh places can be made out without diffi-
culty. But the best notion of the characters and relation of the
cilia is obtained by teasing out, or, still better, by cutting thin
sections of osmic acid specimens. Such preparations quite lead
one to think that the endoderm is ciliated throughout ; in the
sections particularly cell after cell is seen bearing one, two, pr
^ Loc. cit., pi, 15, fig. 8.

222

T. JEFFERY PARKER.

three cilia. These latter are of great length, in fact, nearly or
quite as long as the cells to which they are attached ; in some
cases, indeed, they are longer. I have never seen anything like
a collar” at the base of any of the cilia.

The amoeboid character of the endoderm cells, as seen in sec-
tions or teased fragments of the living animal, is a well-known
fact; but the extent and activity of the amoeboid movements
during life has not been sufficiently insisted on. In sections of
picric acid or ammonic bichromate specimens, large rounded
pseudopodia are seen to be given off from the cells into the
digestive cavity, sometimes to such an extent as completely to
obliterate the latter. The length of the cells may, therefore,
vary almost indefinitely ; they may be but little longer than the
ectoderm cells, or may be two or three times as long. This va-
riation in the size of the endoderm cells, and the consequent
variation in the diameter of the digestive cavity, is very marked
in my series of sections, nearly all of which are taken from large
specimens,^ killed in a state of half extension. When the endo-
derm cells are fully extended, it is almost impossible to obtain
them complete by teasing. They nearly always break across, and
can only be obtained in a fragmentary condition.

A very noticeable point about the endoderm cells is the
presence in their protoplasm, especially towards the far end, of
dark-coloured irregular granules, of various sizes. It has been
suggested that these are products of excretion : Kleinenberg
makes the important observation that their number varies with
the state of nutrition of the animal.

I am convinced that these bodies are food particles, taken into
the protoplasm of the cells, from the partially disintegrated bodies
of the Entomostraca in the digestive cavity. They are of quite
the same nature as the contents of the alimentary canal in many
of the common Cladocera and Copepoda ; they occur chiefly in
the free end of the cell, and in some cases they have all the ap-
pearance of being half in and half out of the protoplasm. The
particles of the more transparent parts of the body of the Crus-
taceans will naturally not be so evident in the cell protoplasm ;
even these, however, can be made out in a Hydra in full diges-
tion, when the endoderm cells of the distal or gastric region are
completely crammed with transparent sphe^’oids.

The clearest case of ingestion of solid particles was seen when
a diatom was seen to be completely imbedded in the protoplasm
of a cell.

If this explanation of the dark granules is the correct one
Hydra \Cill have been shown to exhibit a process- of alimentation
identical with that described by Metschinkoff in the lower Tur^
^ Supplied by Mr, ]3oUon, of Birmingham.

HISTOLOGY OF HYDRA FUSOA.

223

hellaria and in sponges.^ The Eussian observer describes the
complete obliteration, during digestion, of the digestive cavity
in the Turbellarians, and of the canals in the sponges ; and,
in the former as well as the latter, he has undoubted evidence
of the actual ingestion of solid particles by the endoderm
cells.

It would seem,. therefore, that Kydra adds another instance
to the two already brought forward by Metschinkoff, of a Meta-
zoon exhibiting what is usually considered to be a distinctively
Protozoan mode of digestion. It is quite possible that a pre-
liminary disintegration of the animals taken in is performed by
juices secreted by the endoderm ceils, but the final digestion
seems to take place in the actual protoplasm of the cells, into
which the food articles are taken in the solid form.

The endoderm cells of the tentacles resemble those of the
proximal end of the body in possessing large vacuoles. Their
nuclei are in some instances, although not constantly, simple
and non-nucleolate like those of the ectoderm cells from the
same region.

Finally, I have been able fully to confirm Professor Huxley’s
statement^ as to the presence of nematocysts in the endoderm
(fig. I, n), a statement which, as far as I am aware, has not
been made, with regard to Hydra, by any other writer on the
subject. This fact is, like the absence of interstitial tissue in
the tentacles, an argument against Kleinenberg’s view that the
tissue is the sole source of the nematocysts.

4. Methods. — For sections, the Hydros were either killed with
hot water, and placed in Kleinenberg’s picric acid for two hours,
or were placed alive in ammonic bichromate, I per cent. — which
always kills them in the half extended condition — and kept in it
for two or three days. In either case they were afterwards
transferred to 50 per cent, alcohol, and then placed successively
in 75 per cent., 90 per cent., and absolute alcohol. The speci-
mens were stained either with carmine or picrocarmine, and
imbedded in cacao butter, after soaking for a short time first in
oil of cloves and then in melted cacao butter. By this means
they became so thoroughly permeated with the imbedding
material that they could be cut without the loss of a single
section ; even longitudinal sections of the tentacles could be
made with ease.

For teasing I employed ammonic bichromate, acetic acid (O’ 5

’ ‘ Zool. Anzeiger,’ Bd. I, (1878), p. 387, and ‘ Zeltsch. f. wiss. Zool.,’
Bd. xxxii (1879). It need hardly be said that the above view of the
physiology of digestion in Hydra was suggested by these papers of Met.s*
chinkoff’s.

2 Huxley and Martin, ‘ Elementary Biology,’ p. 100,

224

T. JEFFERY PARKER,

per cent.), or osmic acid (1 per cent.), and for this purpose the
specimens were not transferred to alcohol, but to weak glycerine
(equal parts of glycerine and water), in which they were teased
out.

Tor sections showing the cilia of the endoderm, the Hydrcs
were kept for twenty-four hours in 1 per cent, osmic acid, then
washed, and preserved in weak glycerine until required for cut-
ting, The sections were cut by Dr. Pritchard^s very convenient
freezing microtome, the specimens being placed in gum water
before freezing.

THE ORTHONECTIDA.

225

The Orthonectida, a New Class of the Phylum of the
Worms. ^ By Alfred Giard^ Professor in the Faculty
of Sciences of Lille. With Plate XXII.

In 1868, in an interesting memoir^ on the -marine Plana-
rians of St. Malo, Keferstein figures an animal which he
designates by the name problematic parasite.” The ex-
planation of the plate (plate ii, fig. 8) indicates that this
parasite has been found several times, and often in great
quantity, in the digestive tube of the Tremellaria (Lepto^
plana tremellaris). The text gives us no information as to
this curious organism, and it is chiefly by means of the
rather rough figure of the second plate of Kerferstein's work
that we are led to refer it to the group which we are about
to study.

Macintosh, in his beautiful monograph of the British
Nemerteans, published in 1874,® recalls the observation of
Keferstein in reference to a parasite which he met with in
Linens gesserensis {Nemertes communis of P. J. Van
Beneden), and concerning which he gives the following
details, accompanied by a few drawings :

“ Another curious parasite is found burrowing in the body-
wall of Linens gesserensis, its presence being readily recog-
nised by the perforated and honeycombed appearance of the
dorsum of the affected animal, whose textures seem to be the
seat of the workings of a microscopic Tomicus typographicus.
When highly magnified, the afiected region appears to be
covered with a vast network of pale, minutely granular
channels, which contain numerous opaque, ovoid, granular
masses.

On rupturing the body of the worm a large number of
the peculiar structures (plate xviii, fig. 17) slide out of the
channels, and swim through the surrounding water, generally,
though not always, with the upper end (in the figure) first.
Externally they are coated with long cilia, whose activity in
the free state is of somewhat short duration, for after a time
the animals remain quiet, and they drop off. The body is

^ This memoir has been corrected and new matter, including pre-
viously unpublished figures, has been added by Professor Giard to the
original memoir, which appeared in ‘ Robin and Ponchet’s Journal,’ for
the present publication.

’ keferstein, ‘Beitrage zur Anatomic und Entwickel. einiger See-
planarien,’ Gottingen, 1868.

3 ‘A Monograph of the British Annelids,’ part i, Ray Society, p. 129,
and pi. xviii, ngs. 17, 18, 19.

226

ALFRED GIARD.

distinctly segmented, and tapers slightly towards the pos-
terior end ; while the surface is marked by very fine longitu-
dinal striae, as in Opalina, though in a much more minute
degree. Anteriorly is a conical portion (a), composed of
three rather indistinctly marked segments. Two evident
annuli (6) succeed, the posterior part of the last being nar-
rowed, so as to cause a constriction of the body-'wall. Be-
hind are six nearly equal divisions (c), each often appearing
double, that is, has a broad anterior and a narrow posterior
annulus. The posterior region (d) consists of three indis-
tinct segments. The body is minutely granular throughout,
and an internal cavity is apparent from the fourth segment
to the last, commencing in the former by a rounded end, and
terminating just within the border of the latter. No aper-
ture is observed at either end. The opaque, ovoid, granular
bodies (plate xviii, fig. 18), scattered profusely throughout
the infected portions of the Lineus, are evidently early
stages in the development of this species, and they, too, are
ciliated. On subjecting them to gentle pressure (fig. 19),
transverse segmentation is apparent, the number of segments
varying according to the degree of advancement. The para-
sites are very delicate structures, and in the free state soon
break up into cells and granules, after discarding their cilia,
as above mentioned. Transverse section of the affected
worms shows that they occur both in the skin and in the
walls of the digestive tract, their ravages in the pigmentary
layer of the former tissue causing the curious appearances
which led to their detection. It is a somewhat difficult
point to determine whether the skin, muscles of the body-
wall, or digestive canal, constitute the common area of this
creature’s depredations, whether it is piercing the former on
its way to the surface, or passing to the alimentary cavity to
be voided per anum. The characteristically segmented con-
dition of the mature specimens and their internal structure
exhibit a higher type of organisation than the ordinary
Opalina. Professor Keferstein found a very similar parasite
in the stomach of Leptoplana tremellarisy but he did not
describe it further than simply mention, under the explana-
tion of the plate, that it is an enigmatical structure. The
centre of the body is occupied by a double row of large cells
in his figure.”

Such are the details, certainly rather incomplete, it must
be conceded, which w’e have been able to find in earlier
writers touching the curious animals which we have
designated by the name of ORTHO^’ECTIDA.^

1 We published in the ^ Comptus Kendus de I’Academie des Sciences ’

the orthonectida.

227

II. — Species Observed.

I have had the opportunity of studying the species dis-
covered by Macintosh. Lineus gesserensis is very common
at Wimereux, as also its variety Lineus sanguineus y under
stones, in oozy spots, which are accessible every day even
during the neap tides.

My attention was drawn to the parasite of this species, a
rare one, on the whole, by my pupil and friend, J. Barrois,
at the time when he was preparing in our laboratory his
w’ork on the Nemerteans. But I could then, in the absence
of sufficient material, do no more than verify a part of the
observations of Macintosh, and rectify certain errors committed
by this eminent zoologist, as I shall point out further on.

In vain I have sought, on several occasions, for the para-
site discovered by Refers tein in Leptoplana tremellaris ;
and yet this Planarian is excessively common at AVimereux,
in very nearly the same spots as Lineus gesserensis. Paul
Hallez, Demonstrator in the Faculty of Sciences at Lille,
who has dissected numerous examples of Leptoplana, has
had no better luck than I in this search.

A happy chance made me acquainted, in 1877, with two
new species of Orthonectida, which I have been able since
to obtain in sufficient abundance to make a nearly complete
study of them, although some gaps still remain to be
filled up.

During the autumn of 1877 I had betaken myself to
Wimereux in order to study the embryogeny of a species of
Ophiura with condensed development. I had a choice
between two species, Ophiothrix fragilis and Ophiocoma
neglectay both equally common, hermaphrodite, and vivi-
parous. Reasons of an entirely technical nature caused me
to prefer the second. In opening hundreds of Ophiocoma,
to extract from them the embryos, I discovered two species
of Orthonectida which have enabled me to undertake the
study of the group. Both are excessively abundant in
the animal infested by them, but it is quite a rare thing
to find an Ophiurid thus infested. According to my notes,
each species may be found once among about eighty speci-
mens of Ophiurids, so that one has the chance of finding
one or other species once in forty specimens dissected.^

(Seance du 29, Octobre, 1870), a short preliminary notice of these
animals.

1 Besides these two species of Orthonectida, Ophiocoma neglecia presents
at Wimereux a certain number of other parasites, which are interesting.

1. A pretty Vorticella with a very short peduncle {Vorticella ophiocoma,

•228

ALFRED GIARD.

In the month of June last I twice found the two species
of Orthonectida associated in the same specimen of an
Ophiurid.

It has appeared to me as though the Orthonectida were
more common in the autumn than at any other season. But
this perhaps depends on the fact that I had been able to
give more attention then to my researches than during the
academical session.

III. — Anatomy of Rhopalura ophiocom^.

Nothing is easier than, with aid of fine needles, to separate
the whole dorsal region of Ophiocomma from the ventral disc
formed by the buccal plates and the arms. The animal
deprived of its dorsal cupola, and consequently of its diges-
tive and reproductive organs, lives nevertheless several days,
walks, and conducts itself generally as though no injury had
occurred to it.

In tearing open the dorsal cupola, in order to force out the
embryos, the glass slip on which the operation is being con-
ducted is sometimes seen to be covered by small white
flakes, which, on examination with a low power of the
microscope, are seen to be animated by a rapid movement of
translation in a straight line, which is highly characteristic.

This movement, common to all the species of this group,
has led me to give to these organisms the name of Ortho-
nectida. These agile little animals resemble large ciliated
infusoria of a porcellanous aspect and a generally cylindrical
form. A certain number, however, have a somewhat shorter
form, irregularly ovoid, and move more slowly.

In the one form or the other they are simple Planulce,

nov. sp.), which is found even on the arms of young individuals still
enclosed in the maternal incubatory pouch (very common).

2. An Urceolaria {Ur. ophiocomce., nov. sp.), also very common.

3. A copepod crustacean of the family of the Ergasillidae, or rather
of a distinct family comprising the annelidocolous species. This crus-
tacean, which is broad and provided with hooked limbs, attaches itself
to the arm of the Ophiurid between the bristles. It recalls the form of
the crab-louse. By a very natural association, I name it Phthiriopsis

and I dedicate this parasite of the star-fishes to M. Emile Blan-
chard, who teaches at the Museum of Paris the history of the Arthro-
poda. The Phthiriopsis is very rare.

4. A curious little annelid of the genus Sphserodorum, which presents
the same relations to Ophiocoma which Chlorhema Dujardinii does to
Psammechinus miliaris. 1 shall call it Spht^rodorum Gi'eejii, dedicating it
to the learned Professor of the University of Marburg. About one
Ophiurid in fifteen harbours a Sphcerodorum.

All these parasites of Ophiocoma will be studied in detail in a special
memoir.

THE ORTHONECTIDA.

229

that is to say, organisms composed fundamentally by two
layers of cells, the one within the other — an ectoderm or
external layer of cells, which are mostly ciliated, and an
endoderm constituted by larger cells,’which are more granular,
and form the lining of a central cavity devoid of either mouth
or anus.

The first ring terminates in a blunt cone anteriorly, which
carries a bunch of stiff cilia. It is followed by a cylin-
drical ring of the same length, the whole surface of which is
beset with papillae, which are disposed in eight or ten longi-
tudinal lines and in four or five transverse series ; this ring
is the sole part of the body which does not present vibratile
cilia.

The third ring is larger by itself than the two first
together ; it gradually widens towards its posterior ex-
tremity.

The fourth metamere is of the same dimensions as the
papilliferous ring; it is followed by a terminal ring provided
with longer cilia, forming a brush at its posterior extremity.
This last ring is conical and subdivided into two metameres,
less distinct than those in front.

The last rings form a sort of club, with which the animal
strikes the water, independently of the movement of the
cilia, giving sharp strokes, which one is immediately led to
attribute to the contraction of muscular elements. We
shall point out below where these elements occur.

Such is the elongated variety. The ovoid variety differs
from it only by being a third less in length and a greater
breadth. It seems, at first sight, as though it had been
derived from the former by a contraction along the long
axis ; but it is easy to assure one^s self that this is not the
case, and that it is, on the contrary, the elongated form
which is the ultimate condition of the ovoid form, which
latter is only the final embryonic phase of the animal.

For a more complete study of the anatomy of Rhopalura
it is necessary to employ very high magnifying power, the
objective 6 of Verick,or better, the objectif 9 (immersion) of
Hartnack.

The ectodermal cells then appear very distinctly, except
in the papilliferous zone where it is very difficult to deter-
mine their number and form. All the other metameres are
made up by a single transverse row of cells, and the differ-
ence in the length of the metameres depends entirely on the
difference in the length of the cellular elements which com-
pose it. The terminal rings are formed entirely by four
cells, as in the Dicyemida, the median rings comprise six or

230

ALFRED GIARD.

eight cells; it is very difficult to count with exactitude,
since optical sections rarely present themselves, and it is
impossible to make real sections.

The ectodermal cells present very long and very dense
cilia. By using osmic acid, followed by picrocarmine, it is
easy to preserve the ectoderm with its clothing of cilia.
Preparations made a year ago give at this day an excellent
idea of the living animal.^

The endoderm is primitively formed of larger cells than
those of the ectoderm, but they undergo in the adult a very
singular modification. The whole of this layer forms an
oval sac, the anterior extremity of which is hidden by the
papilliferous ring, and extends from the penultimate meta-
mere until it is inserted, in the form of a sort of pedicle,
between the four terminal ectodermal cells (Plate XXII,
figs. 3 and 4). The swollen part of the endodermal sac
presents, when examined with the immersion lens, fine mus-
cular bands disposed in a finely granular matrix, and recall-
ing the appearance of the endodermic muscular layer of
certain nematoids.

I do not believe that these muscular elements are derived
from the bodily transformation of certain endodermal cells ;
I consider them rather as a part only of such cells, which are
thus called upon to play a double physiological role. They
would thus be analogous to the epithelio-muscular cells
described by Korotneff in Hydra, but with this difference,
that in the present case it is the external part of the cell
which becomes muscular, the internal part remaining
epithelial.

I would draw the attention of histologists, more skilled
than I am in technical methods, to this very delicate point.
The question involved is one worthy of their skill.

Metschnikoff has recently put forward the opinion that
the striated bands are formed by the tails of spermatozoa,
the entire endoderm being nothing but a testicular gland.^

1 am quite unable at present to accept the opinion of my
learned opponent. These bands are chiefly visible in young
individuals; their number is constant; they are always
disposed obliquely, as I have figured them, on an ovoid
endoderm ; and this disposition is not, as I had at first
supposed, the result of an accidental torsion. On changing
the point of view, the continuation of the spiral is seen

* I have shown these preparations to various persons, and specially
to Dr. Macleod, of the University of Ghent, who spent a few days at
Wimereux during last April.

2 ‘ Zoologischen Anzeiger/ II, No. 43, p. 619.

THE ORTHONECTIDA.

251

on the other side of the body, and the clear interspaces
take the form of lozenges, just as one sees sometimes in
looking through certain kinds of open basket-work. Finally,
it will be seen below, that I have also found these elements
in Intoshia, which Metschinkoff considers as the female form
of Rhopalura, and where, consequently, there would be no
spermatozoa.

Whatever may be the origin of these muscular elements,
they unite, as I have pointed out above, into a sort of fascicle
at the terminal part of the adult animals, and by their
contraction give rise to those brusque movements, those
strokes of the club-like tail, to which the name Rhopalura
has reference.

If the interpretation above given is admitted, there
would not be in the Orthonectida any true middle layer, but
only a splanchno-pleural pseudo-mesoderm, comparable to
the somato-pleural pseudo-mesoderm of the Hydra.

I give to the totality of these elements the name o
pseudo-mesoderm, because it appears to me desirable to
reserve the name of mesoderm, properly so called, for other
structures which do not exist in the Orthonectida, and the
homology of which in the various groups of Metazoa is very
difficult to establish.

I distinguish : — (1) a solid mesoderm, formed very early at
the expense of the endodermic cells of the embryo (rudi-
ment of the notochord of the Tunicates and of the Vertebrates,
skeletogenous cells of the embryo of Echinoderms, meso-
dermal cells derived from the four first spheres of the endo-
derm of Planarians, of Bonellia, according to the researches
of P. Hallez and of Spengel, &c.). (2) K cavitary mesoderm
formed by the diverticula of the endoderm (enterocoels), and
appearing generally at a later epoch (aquiferous system of
Echinoderms, enterocoel of the Tunicates, of the Brachio-
pods, of Sagitta, of Amphioxus, &c.). The solid mesoderm
gives rise chiefly to the muscular system ; the cavitary meso-
derm forms principally the vascular organs.

The physiological role of a histological element has, be it
noted, only a secondary importance for the determination of
phylogenetic homologies. A muscular element, for example,
will always be formed at that point where it is needed,
sometimes in a rudiment having an endodermal origin,
sometimes at the expense of ectodermic elements (Nemer-
teans). It may even be formed by a mere portion of a cell
(plastidule), as we find in the Infusoria, in the Coelentera,
and ill the Orthonectida.

The interior of the endodermic sac is filled, in certain

VOL. XX. — NEW SER. Q

232

ALFRED GIARD,

individuals^ by cellular elements, which give rise to the
genital products. I am at a loss to understand why
Metschnikolf refuses the name of endoderm to a cell-layer
which arises absolutely in the same way and plays precisely
the same part as the part called by this name by all embryo-
logists in the other Metazoa.

I have seen in some Rhopalurse a cloud of agile corpuscles
issue from the sides of the body between the third and the
fourth metameres. Are these corpuscles spermatozoa ? It
was not possible for me to get a distinct notion of their
form, and I also am unable to affirm the presence of a natural
opening at this spot on the body. It is possible that the
specimens which presented this appearance of an emission of
spermatozoa were really the victims of lacerations.

IV. — Anatomy of Intoshia gigas.

The second species of Orthonectida parasitic in Ophiocoma
negiecta is much larger than Rhophalura. Its length is, in
fact, two and a half times that of the latter. I name it
Intoshia gigas.

Intoshia presents in swimming alternative movements of
contraction and expansion in a transverse direction. It does
not possess a papilliferous ring, but simply a ring devoid of
cilia in its place. The body is of a more regular breadth
and less tapering at the two extremities, which are blunt
points, a little more conical than in the Intoshia parasitic in
Nemerteans and Planarians.

Further, the anterior part of the body is strongly flattened
in Intoshia gigas, and the non-ciliated segment presents on
its inferior face, throughout its breadth, a transverse groove
of some depth, so that the profile of the animal is that of a
shoe with a heel to it (Plate XXII, fig. 5).

The metamerisation is less distinct than in Rhopalura.
After the cephalic ring and a cervical ring corresponding to
the papilliferous ring there follow three metameres of de-
creasing breadth (the third (7) is about half of the first (a),
the second (j3) being of intermediate dimensions). To these
follows a ring (S) of much larger size, which seems some-
times to be divided into three, then follow two very small
metameres (e and ?), which have about the same length as 7,
and, finally, the terminal piece.

The variable dimensions of these metameres is no longer
related, as in Rhopalura, to the size of the compound
cells.

Each metamere is, in fact, formed by several rows of cells,
the cells of each row being regularly placed over the corre-

THE ORTHONECTIDA.

233

spending cell of adjacent rows, so as to form a longitudinal
series. It is to this disposition that the longitudinal striae
observed by Macintosh are due, which led this naturalist
to approximate the parasites of Lineus to the Opalinae.

The ring a contains three rows of cells, the ring j3 com-
prises two, the ring y only one, and the others only one.

The ectodermal cells oilntoshia gigas are, accordingly, much
smaller and much more numerous than those of Rhopalura,
All the cells, without exception, have long cilia. The head
carries, as in Rhopalura, a bunch of straight cilia directed
forwards. This character is, in fact, common to all known
species of Orthonectida.

The endoderm forms a regularly ovoid sac, constituted in
the adult animal by beautiful polygonal cells, and enclosing,
in its interior, other cells of a rounded shape, more or less
abundant.

The action of acetic acid is such as to detach the ectoderm
(PI. XXII, fig. 7), and then brings well to view the endo-
dermic sac. The jerking movements of the caudal part are
much less energetic in Intoshia than in Rhopalura, We
should, therefore, expect to find the muscular system much
reduced ; and this is actually the case, and for a long time I
in vain endeavoured to see the muscular bands at all. I
succeeded, however, last summer by producing a gradual
and very slight compression of the animal by allowing the
water to evaporate from beneath the cover glass. The
bands were rendered visible at the anterior part of the
body, where the endodermic cells hinder observation the
least. The bands have the same oblique direction as in
Rhopalura.

V. — Gemmiparous Reproduction of the Orthonectida.

The oldest individuals of Rhopalura or of Intoshia pre-
sent a considerable modification of the endodermal cell-
layer. The cells of this layer are no longer visible, and the
endoderm seems formed by a homogeneous membrane of
granular aspect, very similar to certain tissues of the Nema-
toids. On the other hand, we have seen that the ectoderm
can be detached in these animals with the greatest facility
under the influence of reagents. As a result of the pro-
liferation of cellular elements in the interior of the endo-
dermal sac, this organ swells, becomes spherical, and burst-
ing the ectoderm, which disappears, is transformed into a
true sporo-cyst ” (PI. XXII, figs. 14, 15, 16 and 17). In the
interior of the sporocyst are seen buds, the cellular nature
of which is extremely difficult to demonstrate. The same

234

ALFRED GIARD.

difficulty, it may be noted, is found with regard to the buds
of the sporocysts of the Trematoda, and all zoologists who
have occupied themselves with the study of these animals
know how difficult it is to demonstrate the cells which form
the bud-like embryo of a Bucephalus or of any Distoma.

Occasionally the primary buds produce secondary buds.
When these buds have arrived at a certain size, they are
composed clearly enough by a single layer of cells, which
then proceed later to form an internal layer by delamination.

It appears, from what has been just stated, that the
endoderm alone of the parent animal enters into the forma-
tion of the bud-like embryos, but it must not be forgotten
that the papilliferous ring, or that which represents it in
Bitoshia, is characterised by great opacity, and that it is
possible that at certain points ectodermal cells penetrate to
the interior of the endodermal sac.

With their increasing development, the sporocysts lose
their original form and become very voluminous. Often
there are found several sporocysts in the interior of a spe-
cimen of an Ophiurid. This very active gernmiparous
reproduction of the Orthonectida explains how it is that
these animals are found in such great abundance in the
Echinoderms infested by them. The phenomenon is one
very similar to that occurring in the case of the Dicyemida ;
but with regard to these latter, it is almost impossible to
find a Cephalopod which does not swarm with them, whilst
the Orthonectida are comparatively rare.

VI. — Oviparous Eeproduction.

Side by side with the sporocysts just described, we find
in the infested Ophiurids a great number of cellular masses,
which must be considered as eggs in various stages of seg-
mentation.

In the case of Bhopalura, I have only been able to
observe a small number of these embryonic stages (PI.
XXII, figs. 19, 20, 21). The little which 1 have been able
to observe leads me to the conclusion that the segmentation
is irregular, and that the planula is formed by epibole, as in
the Dicyemida. Each division of the cells in the mulberry
mass gives rise to stellate figures, so well known at the
present day by the name of caryolitic figures or amphiasters.
The use of osmic acid enables one to observe them with the
greatest facility.

In the study of the development of Intoshia I have been
more successful. Here the segmented egg forms at first a
perfectly regular blastula, the cells of which are at first

THE ORTHONECTIDA.

235

very much elongated in a radial direction (PI. XXII, figs.
8 to 13). This hlastula very soon exhibits a very definite
process of delamination ; the internal part of each cylin-
drical cell gives rise to a spherical cellular mass, which
detaches itself from the mother-cell, which is shortened by
the process (PI. XXII, fig. 10). This is absolutely the
same mode of formation as that which can be observed in
the production of the mesoderm in the Ophiurids with a
condensed embryogeny {Ophiothrix fragilis, Ophiocoma
neglecta). In the present case it is the endoderm which
forms at the expense of the ectoderm, but we have seen
that this endoderm contains potentially the mesodermic
elements (the muscular bands).

Soon the ectodermal cells become covered with vibratile
cilia, and elongate in a definite direction ; then the embryo,
which was spherical, becomes ovoid, the metameres show
themselves little by little, and by insensible gradations we
arrive at the adult condition.

VII. — System ATI c.

According to what has been stated in the preceding pages,
the Orthonectida may be defined in the following manner : —

Metazoa retaining throughout their existence the form of
the planula, having a ciliated ectoderm (stiff cilia forming a
bunch on the anterior cephalic region, vibratile cilia on the
other regions of the body), presenting metameres, which do
not correspond to any internal division of the body, with
a sacciform endoderm, which gives rise to a muscular
splanchno- pleural pseudo- mesoderm. Reproduction of two
kinds : (1) gemmiparous in the interior of sporocysts, con-
stituted by the development of the endoderm ; (2) oviparous,
dependent on the combination of male and female elements,
formed in all probability in different individuals.”

The class comprises at present two genera :

I. The genus Rhopalura, characterised by the presence of
a papilliferous ring, by its ectoderm composed of large cells
limited in number, and its definitely muscular endoderm. »

Species unique. Rhopalura ophiocoma. Length 0T08 mm.

II. The genus Litoshiay devoid of papilliferous ring ;
ectoderm formed of small, very numerous cells.

Three species :

1. Intoshia gigas, — Parasitic in Ophiocoma neglecta.
Length 0 270 mm.

2. Intoshia linei, — Lineus gesserensis. Length
0‘157 mm.

236

■ ALFRED GTARD.

3. Intoshia leptoplance. — Parasitic in Leptoplana tremeU
laris. Length 0-135, breadth 0 03 mm.

This last species, like the preceding, is regularly cylin-
drical, and rounded at the extremities. According to the
figure of Keferstein, it presents ten metameres, perfectly
regular, in addition to the cephalic and caudal rings. The
endoderm appears to be formed by large spherical cells.

The general form of the body, the regularity of the
metamerisation in the two species parasitic in the Nemertean
and the Planarian, lead to the supposition that there may
be other differential characters present in them, which would
lead zoologists probably to create a particular genus for
these two types. Such a step in our present state of know-
ledge would, it appears to me, be superffuous.

Metschnikoffi has recently described, under the name of
Rhopalura Giardii, a new species of Orthonectida, w'hich he
has since identified with Rhopalura ophiocom<B.

This species, which is parasitic in Amphiura squamata,
was found abundantly at Spezzia. It presents itself in two
different forms, which Metschinkoff" thinks are probably
identical with Rhopalura ophiocomce and Intoshia gigas,
which, according to him, are merely the male and female of
one species.

I must confess that this idea has often occurred to me
during my researches, and I do not entirely give it up even
at this moment. The strongest argument in its favour is
that the two forms, Intoshia and Rhopalura^ exist in about
the same quantity in Ophiocoma neglecta ; and that it would
be curious to find in this little Ophiurid two different repre-
sentatives of a group so rare as are the Orthonectida. At
the same time, the difference between the two forms is
greater than I had at first supposed. Further, we have not
elsewhere any example of an animal in which some females
produce eggs, giving birth to males exclusively, and others
eggs from which only females issue. We might, perhaps,
try to remove the difficulty by supposing that, in one or the
other case, there was parthenogenesis (arrenotoky or thely-
toky) ; but this would be, at present, pure hypothesis.
However this may be, I have no fundamental objection, to
such a mode of explaining the facts, but I shall wait to
make up my mind for the time when I shall have found in
another species of Intoshia {Intoshia lineiy for instance) an
accompanying form analogous to Rhopalura.

Without doubt sexual differences, such as exist \n Bonellia
viridis, and in other worms, such as Bilharzia hcematohia,
* ‘ Zoologischen Anzeiger,’ II, No. 40, and No. 43,

THE ORTHONECTIDA.

237

are of a nature to make one very cautious on this subject ;
but, in addition to the special characters of the histological
constitution of the ectoderm, I have recently observed new
facts which still further separate Intoshia gigas from Mliopa-
lura ophiocomce.

It appears, as the result of an examination of many
hundreds of adult individuals, that Intoshia gigas never has
its non-ciliated segment provided with papillae, nor even
with refringent corpuscles. The refringent corpuscles’^ of
M. Metschnikoff form actual projections on the ring, which
I call the “ papilliferous ring ” in Rhopalura. It must,
then, be admitted that if such corpuscles exist in the sup-
posed females of the parasite of Amphiura squamata, this
parasite belongs to a new species (as is probable enough a
priori)^ and that the sexual dimorphism is less accentuated
ill this species than in the other.

VIII. — Phylogeny.

With reference to the position of the group of the Ortho-
nectidaon the genealogical tree of animals, it is very difficult
to make a definite statement. But there is no doubt that
these parasites ought to be attached to the phylum of the
Vermes, and take their place at the base of this phylum side
by side with the Dicyemida. The phylum of the worms
would then, according to my idea, be represented by the fol-
lowing diagram :

Cestoda,

Turbellarta, Trematoda,

L . J

Gastbrotricha, Prothelmintha, Dicyemida,

Orthonectida,

Gastrseada.

The Orthonectida must occupy in this scheme an inferior
position to that of the Dicyemida. These latter are evidently
much degenerated by parasitism. Their organisation must
have been formerly much higher than it is to-day. The
epiderm contains very clearly [Dicyema of Sepia) the rod-
like bodies characteristic of the skin of the Turbellaria, and
the embryo presents a very complex organ, the ur?ia : nothing
of this sort is seen in the Orthonectida.^ One of the most
^ See on “ Dycyemida,” the beautiful memoir of Edouard Van

238

ALFRED GIARD.

characteristic features of the group of the Vermes thus
limited is the existence, in all the animals of this group, of
a gemmiparous reproduction (sporocysts, echinococci, &c.).
This peculiarity only disappears in the higher worms, the
Turbellarians, which are related to the rest of the group by
so numerous and so important morphological features that
no serious zoologist will entertain the idea of separating
them from the Trematoda and Cestoda, in order to approxi-
mate them to the Annelids, as has sometimes been done.

Amongst the animals formeily classified with the preceding,
some (Bryozoa, Annelida, and satellite-groups) are intimately
related to the true Molluscs, with which I unite them to con-
stitute the group of the Gymnotoca, whilst others form an
assemblage which we may call the Nematelmia, including
therein the Nematoidea, the EcMnorhynchaj the Desmoscole-
ciday the G aster otricha, &c. The Tunicata must be placed
at the base of the phylum of Vertebrata. Budding from the
interior of sporocysts is found, it is true, in other classes of
the animal kingdom. Leaving aside the somewhat aberrant
cases found among Arthropoda, we see in certain Rotifers
{Callidina) an internal gemmiparity very similar to that of
the Vermes. Further, the Gasterotricha, the parent-stock
of the Rotifers and perhaps of all the Gymnotoka, appear
to me to he readily connected with'the Vermes by means of
the Orthonectida.

We must not forget that the external resemblances between
the Orthonectida, Gasterotricha, parasitic Rotifera, &c.,
are further increased by the similar manner of life obtaining
in all these animals. OpMcoma, like Linens gesserensis and
Leptoplana tremellaris^ inhabits muddy bottoms. The same
is true of the limicolous Annelids and of Nebalia, on which
we find as parasites certain degraded Rotifers (Balatro and
Saccobdella). But the Orthonectida are in any case very
inferior to the most degraded Rotifers, and represent, without
any doubt, after the Gastraeada, the first step of the sub-
kingdom Metazoa.

I must add here that, according to a letter from Leuckart,
the embryo of Distoma resembles the Orthonectida in a most
astonishing manner, a fact which tends to confirm the
position which I have assigned to this group among the
Vermes.

Beneden, ‘ Bulletins de I’Acad. de Belgique/ 2e serie, t. xli and xlii,
No. 7, 1876 ; also Abstract, with Plate, in • Quart. Journ. Mic. Science/
vol. xvii, 1877, p. 132.

THE ORTHONECTIDA.

239

IX. — General Reflections.

At first sight it seems as though the discovery of the
Orthonectida would bring a very solid support to the Planula
theory of Ray Lankester, or, still more, to the Parenchy-
mula theory of Metschnikoff, and I do not doubt that more
than one zoologist will interpret, in that sense, the observa-
tions I have recorded.

1 persist, however, for my own part, in considering the
invaginate Gastrula as the prototype of the Metazoa. 1
base this opinion on the following arguments :

1. The Orthonectida are parasitic animals, and we must
take into account the retrogression which this kind of life
may have brought about in their structure. An organisa-
tion which we consider as one of primitive simplicity is very
possibly simple only in consequence of degeneration, if we
have to deal with a parasite, and especially an internal
parasite.

2 We have seen that the Planula is formed by epibole in
Rhopalura ; in this case we have, then, at any rate, mo-
mentarily, a real gastrula, which closes up and does not
reopen, because the mode of life of the animal does not
require the existence of a permanent digestive tube.

3. The forms which present the embryonic phase termed
parenchymula by El. Metschnikoff cannot be considered, as
that naturalist would wish, as the lowest among the Sponges
and Hydroids.

The forms known as Halisarca are not low sponges, but
sponges which have undergone a degeneration of their
skeletal apparatus. From the point of view of general
morphology I have shown that we may compare them to the
Botrylli, and must assign to them a very high grade among
the Fibrosa, analogous to that which the Botrylli occupy
among the Synascidiae or the Leucones among the Calci-
spongiae.

The Siphonophora, we may remember, are very far from
being Coelentera of inferior position, and it is by no means
astonishing that they present a condensed embryogeny.
The typical embryonic form is found among the Coralligena
and certain Actiniae.

The Echini and the Ophiurids with pelagic embryos have
a small nutritive vitellus. They present a gastrula formed
by invagination, and it is only when this gastrula is thus
formed that the mesoderm takes origin, at first by partial
delamination of the ectoderm, and later by lateral thicken-
ings of the endoderm.

210

• ALFRED GIARD.

In the Ophiurids with condensed emhryogeny, the egg
presents an enormous nutritive yelk, and the mesoderm is
formed by abbreviation at the same time as the endoderm
by a general delamination of the ectoderm, which leads us
to Metschnikoff’s form dubbed Parenchymula, This I have
established most clearly in cases of Ophiocoma neglecta and
Ophiothrix fragilis.

It is clearly impossible for me to discuss here the im-
portant questions of general embryogeny which are raised by
the study of the Orthonectida. I will merely say that,
according to my observations, a large number of calcareous
and siliceous sponges present what one may well call a
biconvex archigastrula. Quite recently Keller has shown
that Chalinula possesses an amphigastrula. Kowalevsky
has described, in the most exact manner, the existence of an
archigastrula in a variety of Actinia mesemhryanihemum,
in Cereanthus, and in various Medusae. I am able to confirm
his statement in reference to Actinia equina. Finally, Ed.
van Beneden has described and figured an amphigastrula
in the Dicyemida, so closely related in many respects to the
Orthonectida.

These examples will suffice, I think, to justify my opinion
that the gastrula by invagination is the primitive mode ;
the gastrula by delamination [Planuba or Parenchymula) a
secondary mode of embryonic development.

I wmuld also direct attention to the metamerisation which
is so remarkable in the Orthonectida,

We have seen that this metamerisation only affects the
ectoderm, and I believe that this was primitively the case
also in the Annelids. What proves this to be the case is
the highly differentiated form of the digestive tube of the
Chsetopods in which metamerisation is only well marked in
those organs derived from the ectoderm, such as the bristles,
the parapodia, and the segmental organs (nephridia).

It is obvious enough that I am not alluding to that kind
of metamerisation which is observed in Salmacinay Syllis,
&c., which is merely the result of gemmiparous reproduc-
tion. This last kind of metamerisation is only comparable
to what one finds in the Cestoidea and the Rhabdocoela.

NOTES AND MEMORANDA.

The Origin of the Red Corpuscles of Mammalian blood. —

A step forward in our knowledge of this subject has been made
by Professor Einddeisch, of Wurzburg. As he justly remarks

Archiv. f. mikrosk. Anatomie/ vol. xvii, August, 1879), in
the introduction to his memoir, not a word is needed as to the
usefulness, in fact the necessity, of continually renewed re-
searches as to the site of the formation of the blood and its mode
of formation. Who among us does not feel it as a wound, a
painful raw in his scientific manhood that we still are unable to
say ' Here and thus do the red blood-corpuscles take their
origin Neumann and Bizzozero deserve the amplest recog-
nition for their discovery that in the red marrow of bones, cells
occur with reddish-yellow homogeneous protoplasm and well-
marked nucleus, cells which accordingly are identical with the
red blood-corpuscles of the earliest period of life. These obser-
vations are easy to repeat and are fully accepted by all histolo-
gists. From these observations we know clearly where besides
in the spleen, we have to look for the great factory of the red
blood-corpuscles.

Haematogenesis is either a temporary or permanent function
of certain regions of the connective-substance apparatus of the
body, which for this purpose and during this period enters into
an open communication wfith the lumen of blood-vessels either
by the loss of their proper walls on the part of the capillaries
and veins, as happens in the bone-marrow or by the thinning of
their walls to such a degree, as in the spleen, that the unrestricted
in-and-out wandering of cells becomes possible. The haemato-
genous connective-tissue becomes a sort of accessory cavity for
the lumen of the blood-vascular system. In this cavity haemo-
globin-containing cells are produced by the conversion of
colourless cells. Professor Eindfieisch compares the process of
formation of haemoglobin in these cells to that of fat in fat-
forming connective-tissue.

Professor Eindfleisch^s special contribution to this subject
consists in : 1st, a description of the vascular plexus of the
marrow of mammalian bones which he has succeeded in injecting
(using the rib of a young guinea-pig) and of the w^all-less
character of its smaller vessels. 2nd, and of greatest import-

242

NOTES AND MEMORANDA.

ance, an answer to the question How do the nucleated red
corpuscles of the red bone-marrow give rise to the non-nucleated
red corpuscles of the blood

It is well known that this question has always been answered
by hypothesis based on very slender foundation.

The old view, as to the origin of the red blood-corpuscles,
was that the nucleus of certain colourless corpuscles became
red and escaped as a free nucleus, the homogeneous red blood-
corpuscle.

Later knowledge as to the red coloration of the whole of the
mother-cell of the red corpuscle led to the assumption that the
nucleus became atrophied and the whole cell converted into the
non-nucleated red corpuscle. The attempts which have been
made from time to time during the past few years to detect a
nucleus in some form or other in the red mammalian corpuscle,
point to a foregone conclusion in favour of this total con-
version.

Professor Eindfleisch has, however, seen, both in embryos and
more advanced individuals, the steps in the transformation of the
red-coloured cell of the marrow into the non-nucleated red cor-
puscle which demonstrate that the nucleus of the red coloured
cell escapes and atrophies whilst the body of the cell contracts
and becomes the red corpuscle.

He gives figures of the red cells with their nuclei in the act of
escaping, lying just on the limit of the cell-body or protruding
from or even hanging by a mere thread to the latter. Then
beside these he has seen and figures the freed nucleus and the
irregular collapsed coloured body of the cell, which will soon be
shaped by pressure and rolling into the disc-form of the circu-
lating red corpuscle.

Professor Eindfleisch has endeavoured, but unsuccessfully, to
witness under his own eyes the actual extrusion of a nucleus
from a red cell. At the same time the intermediate series of
forms observed by him are very strong evidence in favour of the
view which he takes.

It seems also that Professor Eindfleisch's view is supported by
certain facts of comparative anatomy which he has not himself
adduced in its favour. In the Chsetopodous and some other worms
the nuclei of the vascular walls are often loosened and float in
the blood as corpuscles. They are not impregnated by haemo-
globin but the plasma, in which they float, is. Whence comes
the haemoglobin of the plasma ? Clearly the cells forming the
walls of the vascular system in certain regions are in the Chae-
topoda as in Vertebrata, haematogenous ; in them as in Verte-
brata, the body of the cell forms the haemoglobin which in this
case becomes liquid instead of retaining the form of a corpuscle,

NOTES AND MEMORANDA.

243

and at the same time the nucleus is separated from the hcemo^
globin-b earing body just as it is in the Mammalia y but here, as it
does not there, enters into the blood stream.

Any discussion of the mode and significance of the formation
of haemoglobin in the mammalian blood, ought to take cognizance
of the fact that haemoglobin is formed in the blood of the worms
above noted, in Insect larvae, Crustacea, and even Molluscs, and
further that whilst it usually occurs diffused in the plasma of
the blood it does occasionally, as in the Chaetopods Glycera and
Capitella, the Molluscs, Solen legumen and Area, sp., &c., take
the form of special nucleated corpuscles differing from and
accompanied by the usual amoeboid colourless corpuscles : also
it is to be noted that just as fat occurs in other cells than
specialised fat-cells so do we find the muscular tissue of many
Vertebrates and of some Molluscs (buccal mass) impregnated with
haemoglobin. And even in one Annelid (the sea-mouse Aphrodite)
we have the cells of the nervous tissue so rich in it, that the
nerve-cord is of a deep crimson colour (see ‘ Proc. Roy. Soc.,’
No. 140, 1873). — E. Ray Lankester.

I. On the Mode in which Hydra swallows its Prey. By M. M.
Hartog, M.A., B.Sc., F.L.S., of the Owens College, Manchester.
The current idea is that Hydra swallows by taking its prey in its
tentacles and turning tentacles and all into its stomach. How-
ever, the part played by the tentacles ceases as soon as the mouth
comes in contact with the food. The hydra then slowly stretches
itself over the food in a way that recalls to some extent the
manner in which a serpent gets outside ” its prey, or in which
an automatic stocking might stretch itself on to the foot and leg.
No care seems to be taken, however, to present the easiest point
for deglutition, and an Entomostracan may be swallowed side-
ways, for instance. So far are the tentacles from co-operating
in the act, that they are usually reflexed away from the food ;
occasionally, however, they are swung forward for a moment
around the mass as if to ascertain how much remains to be
swallowed.

If the prey be at all bulky, immediately after the whole act is
completed the body cavity is every where filled and on the stretch,
but after a short lapse of time the body contracts forcibly along
the long axis, so that the part containing the food is globular,
supported on a slender foot and with a slender apical process
bearing the tentacles around the hypostome.

II. Additional Note on Hydra. By the Same. Since my last
note I think I have found the clue to the false idea referred to. A
Hydra that had swallowed a morsel larger than itself disgorged, as
frequently observed, on my attempting to take it up for examina-
tion. On finding it half an hour after, three of its tentacles were

244

NOTES AND MEMORANDA.

turned into its digestive cavity, whence they were successively
and slowly withdrawn. As the mouth closes but slowly after
disgorging, I imagine the swallowing them to have been acci-
dental; and a similar phenomenon carelessly observed may well
have given rise to a false interpretation.

It seems that here we have the true explanation of the occa-
sional presence of nematocysts in the endoderm, and this expla-
nation my friend, Mr. T. J. Parker, is now inclined to accept.
As regards the absence of the interstitial cells from the tentacles
of Hydra fusca I am not able to confirm him ; on the contrary,
they are present, though in isolated patches, and not forming a
continuous network as over the body. I find the best way to
demonstrate these is, having killed a Hydra extended on a slide
by letting fall a drop of one per cent, osmic acid on it, to at
once wash away the acid by a flood of absolute alcohol, and then
after a few minutes to stain with ammoniacal carmine or picro-
carmine. If the Hydra is now examined in glycerine under a
power sufficiently high to focus successive layers, the presence of
interstitial cells can be made out. Owing to their dispersion,
the want of them in a section becomes very slight evidence for
their absence.

III. On the Anal Respiration of the Copepoda. By the Same.
In a note on Cyclops read at the British Association I pointed
out that its respiration was exclusively, anal. I have now made
out the same in CantJiocamptus (fam. Harpacticidcc) , and Diap-
iomus (fam. Calanida) . In all three the mechanism is the same ;
at regular intervals, after the backward sway of the intestine,
the anal valves open for an instant and then close, giving just
time for a slight indraught of water after the opening, a slight
expulsion at the close. The necessary pressure to confine the
animal seems to interfere somewhat with these movements, some-
times stopping them if excessive ; hence I refrain from noting
with illusory exactness the intervals between each respiratory
movement.

It is to be noticed that the rectum contains as a rule liquid
only, the bolus of fseces remaining in it but a short time. By
endosmose the liquid in the rectum will tend to be at the
same condition of gaseous saturation as the body fluid around
it, kept constantly agitated by the backwards and forwards sway
of the stomach. During the short interval that the anus is
open an approach to gaseous equilibrium with the external
water takes place, even despite the very slight movement of the
water (shown by the little change of place undergone by sus-
pended indigo or carmine particles). In the absence of any
other suitable respiratory apparatus, no one can hesitate as to the
function of the action I have described.

NOTES AND MEMORANDA.

245

■ In the Nauplms larvse of Cyclops and Diaptomus the working
is slightly different. The rectum is a subspherical muscular sac,
which at regular intervals contracts so as to leave a linear cavity
(along the long axis of the animal), and immediately dilates,
sucking up the water from without.

An anal respiration, such as that of Cyclops ^ is found widely
among Crustacea — even those which have well developed gills
like Astacus, which is one of the highest forms. It has been
demonstrated in nyllopocla and Cladocera, and is probably
the exclusive mode in Leptodora, as shown by Weismann. That
it is therefore primitive, and should be expected to occur in the
primitive or at least very generalised group of the Copepoda, is
an obvious deduction. Hence I anticipate that the homoeomor-
phic zooea larvae of the Decapoda will prove to have this same
mode of respiration.

If there be any connection between Eotifers and Nauplms, it
is easy to make out the origin of the arrangement in the latter.
The ciliated funnels and lateral canals of the former can only be
of service when there is a thin unchitinised anterior surface
through which water can transude into the coelom ; by the ex-
tension of chitinisation over the whole surface these organs lose
their function and abort, while the cloacal contractile vesicle
takes on an inspiratory as as well as an expiratory function, and
becomes more or less confounded with the rectum, from which
probably, even in Eotifers, it takes origin.

Here must be noticed the wide diffusion of anal respiration in
aquatic Insect larvee (alternate inspiration and expiration by the
pumping movements of the rectum). This would point to a
common origin with Crustacea.

A list of the groups in which anal respiration is made out may
be added.

Vermes :

Rotifera.

Gephyrea.

Oligochceio Limicola,

EcMnodermata,

Tlolothuroidea.

Arlhropoda.

Crustacea (general).

Insecta (most aquatic larvae).

Mollusca.

Bentalium,

Dr. G. von Koch’s Method of Preparing Sections of
Corals. — When working at the structure of corals during the
Challenger expedition I found very great difficulty in deter-
mining the exact relations of the hard to the soft parts. It is

246

NOTES AND MEMORANDA,

easy enougli to prepare sections of the soft tissues after decalci-
fication, and of the hard dried corallum by the old process of
grinding, but the results thus obtained have then to be com-
bined more or less by guess work. Dr. G. von Koch has lately
sent me a series of microscopic sections of corals prepared by
him by means of a method which he described in the ^ Zoolo-
gischer Anzeiger,^ No. 2, 1878, S. 36. In these sections the
hard and soft tissues are maintained in their exact relations to
one another, and both are reduced to a sufficient thinness to
exhibit their minute structure in all essential details. Amongst
them, for example, is a complete section across a Caryophyllia, in
which the arrangement of the mesenteries with regard to the
septa is most fully and clearly exhibited. Dr. von Koch has
described the results at which he has arrived by means of his
method of section cutting, in a series of papers published in the
‘ Morphologisches Jahrbuch,^ and elsewhere. His latest paper,
which gives an account of some points in the structures of
Caryophyllia, is contained in the ^ Morpbol. Jahrb.,^ Bd. v, S.
316 Bemerkungen fiber das Skelett der Korallen^^).

My object in writing this note is to testify to the great success
of Dr. von KocKs method. It will yield valuable results not
only in the case of corals, but also in all other problems of
histology or minute anatomy in which the relations of hard and
soft parts have to be determined. It^ might, perhaps, be em-
ployed with advantage in the examination of the structure of
Cortfis organ. Sections could thus be prepared of the undecal-
cified cochlea in which the components of the organ of Corti
would be seen in sitiiy and unaltered by the action of acids.
Sections of injected bone, showing the relations of the blood-
vessels to the Haversian system, could also thus be made.
Sections across the arms of undecalcified Crinoids and Starfish,
and many similar preparations, suggest themselves as likely to
yield valuable results. Dr. von Koch^s method is described in
full in the ‘ Zoologischer Anzeiger ^ in the notice quoted above.
The hardened objects of which sections are to be cut are stained
and treated with absolute alcohol, and then placed in a solution
of gupa copal in chloroform. The objects are then slowly dried
by means of artificial heat till they are stony hard. They are
then cut into sections with a fine saw, and the sections are
rubbed smooth on a bone on one side, then fastened to the slide
with the copal solution, and ground down with a grindstone
and hone on the other, just as in the case of ordinary sections of
bones and teeth. — H. N. Moseley.

MEMOIRS.

On the Structure and Homologies of the Germinal
Layers of the Embryo.^ By F. M. Balfour, M.A.,
F.R.S., Fellow of Triiiity College, Cambridge.

The discovery by Pander and Von Baer that the young
embryos of vertebrated animals were formed of distinct
layers of cells known as the germinal layers, and that there
were special and constant relations between these layers and
the adult organs throughout the group, remained for a long
time an isolated fact. In 1859 Huxley made an important
step towards the explanation of the nature of these layers
by comparing them with the ectoderm and endoderm of the
liydrozoa.

In spite of Huxley’s comparison it was for a long time
the generally accepted view that germinal layers similar
to those in the Vertebrata were not found amongst inverte-
brated animals. The brilliant researches of Kowalewsky on
the development of a great variety of invertebrate forms
lii-st proved that the accepted view was erroneous, and they
led Lankester (No. 16) and Haeckel (No. 9) to publish
certain speculations which have had an incalculable
influence in stimulating and directing modern embryological
research.^

At the time when the essays containing these speculations
were published there appeared to be every probability of
a definite answer being given to the questions raised in
them. The results of extended investigations during the
last few years have, however, shown that these expectations
were premature; and there are very few embryologists who

* Tliis paper, with some modifications, forms part of a chapter iu tlie
fortlicomiii{^ volume of my treatise on ‘ Comparative Embryology.’

^ If I do not refer to the above authors to any very great extent in the
sequel, it is not because 1 undervalue the importance of their work, but
because their view's have been so much before the public that it is quite
unnecessary to discuss them in detail.

VOL. XX. — NEW SEU, H

248

F. M. BALFOUR.

would venture to assert that any hypotheses, which can now
be put forward as to the mode of origin or the homologies of
the germinal layers, have more than a tentative value.

In the following pages I aim more at summarising the
facts, and critically examining the different theories which
can be held, than at dogmatically supporting any definite
views of my own.

In all the Metazoa the development of which has been in-
vestigated the first process of differentiation which follows
upon the segmentation consists in the cells of the organism be-
coming divided into two groups or layers, knowm respectively
as epiblast and hypoblast.'' This process may commence
during later phases of the segmentation, but is generally not
completed till after the close of the segmentation. Not
only do the cells of the blastoderm become differentiated
into two layers, but these two layers, in the case of a very
large number of ova wdth but little food-yolk, constitute a
double-w^alled sack — the gastrula (fig 1) — the characters of

Fig. 1. — Diagram of a Gastrula. (From Gegenbaur.) a. Moutli ; h.
archenteron ; c. hypoblast ; d. epiblast.

which are too well known to require further description.
It is generally admitted, and will be assumed in the sequel,
that the segmented ovum represents phylogenetically a
compound Protozoon ; and following the same lines of
phylogenetic speculation, it may be concluded that the two-
layered condition of the organism represents in a general
way the passage from the Protozoon to the Metazoon
condition. It is probable that we may safely go further,
and assert that the gastrula reproduces, wdth more or less
fidelity, a stage in the evolution of the Metazoa, permanent
in the simpler Hydrozoa, during which the organism w^as

^ 111 a few cases amongst the Mollusca and Chaetopoda, &c., the meso-
blast is differentiated simultaneously with the two other layers. These
cases may for the moment be left out of consideration.

GERMINAL LAYERS OF THE EMBRYO.

249

provided with (1) a fully developed digestive cavity (fig. 1 h)
lined by the secretory hypoblast, (2) an oral opening (a) , and
(3) a superficial epiblast {d)» These generalisations, which
are now widely accepted, are no doubt very valuable, but
they leave unanswered the following important questions :

(1) By what steps did the compound Protozoon become
differentiated into a Metazoon ?

(2) Are there any grounds for thinking that there is more
than one line along which the Metazoa have become inde-
pendently differentiated from the Protozoa?

(3.) To what extent is there a complete homology
between the two primary germinal layers throughout the
Metazoa ?

Ontologenetically there are a great variety of processes by
which the passage from the segmented ovum to the two-
layered or diploblastic condition is arrived at.

These processes may be grouped under the following
heads :

1. Invagination. — Under this term a considerable
number of closely connected processes are included. When
the segmentation results in the formation of a blastosphere,
one half of the blastosphere may be pushed in towards the
opposite half, and a gastrula be thus produced (fig. 2, A and
b). This process is known as embolic invagination. Another

Fig. 2. — Two Stages in the Development of Holothuria tubulosa, viewed in
optical section. (After Selenka.) A. Stage at the close of the seg-
mentation. B. Gastrula stage, mr. micropyle ; chorion; 5. 6*. seg-
mentation cavity ; hi. blastoderm ; ep. epiblast ; hy. hypoblast ; ms.
amoeboid cells derived from hypoblast ; a.e. archenteron.

process, known as epibolic invagination, consists in epiblast
cells growing round and enclosing the hypoblast (fig. 3). This

250

V. M. BALFOUR.

])rocess replaces the former process when the hypoblast cells
are so bulky from being distended by food-yolk that their
invagination is mechanically impossible.

e_p

Fig. 3. — Transverse Section through the Ovum of Euaxes during an
early stage of development to show the nature of epiholic invagination.
(After Kowalevsky.) ep. epiblast ; ms. mesoblastic band ; hy. hypo-
blast.

There are various peculiar modifications of invagination
which cannot be dealt with in detail.

Invagination in one form or other occurs in some or all
the members of the following groups :

The Dicyemidae, Calcispongiae (after the amphiblastula
stage) and Silicispongiae, Coelenterata, Turbellaria, Nemertea,
ilotifera, Mollusca, Polyzoa, Brachiopoda, Chaetopoda, Dis-
cophora, Gephyrea, Chaetognatha, Nematelminthes, Crus-
tacea, Echinodermata, and Chordata.

The gastrula of the Crustacea is peculiar, as is also that
of many of the Chordata (Reptilia, Aves, Mammalia), but
there is every reason to suppose that the gastrulae of these
groups are simply modifications of the normal type.

Delamination. — Three types of delamination may
be distinguished :

a. Delamination where the cells of a solid morula become
divided into a superficial epiblast, and a central solid mass
in which the digestive cavity is subsequently hollowed
out (fig. 4).

h. Delamination where the segmented ovum has the
form of a blastosphere, the cells of which give rise by
budding to scattered cells in the interior of the vesicle,
which, though they may at first form a solid mass, finally
arrange themselves in the form of a definite layer around
a central digestive cavity (fig. 5), or, in the case of some
sponges, around several cavities.

c. Delamination where the segmented ovum has the lorrri

GERMINAL LAYERS Of THE EMBRYO.

251

Eig. 4. — Two Stages in the Developmetit of Stephanomia pictum^ to illus-
trate delamination. (After Metschnikoff.) A. Stage after the dela-
mination ; ep. epiblastic invagination to form pneumatocyst. B. Later
stage after the formation of the gastric cavity in the solid hypoblast.
po. polypite ; t. tentacle ; pp. pneumatocyst ; ep. epiblast of pneuma-
tocyst ; hy. hypoblast surrounding pneumatocyst.

of a blastosphere in the cells of which the protoplasm is differ-
entiated into an inner and an outer part. By a subsequent

Fig. 5. — Three Ijarval Stages of Eucope poly sty la. (After Kowalevsky.)
A. Blastosphere stage with hypoblast spheres becoming budded off
into central cavity. B. Planula stage with solid hypoblast. C. Pla-
nula siage wilh a gastric cavity, ep. epiblast; hy. hypoblast ; al. gas-
tric cavity.

process the inner parts of the cells become separated from

252

F. M. BALFOUR.

the outer, and the walls of the blastosphere are so divided
into two distinct layers (fig. -8).

Although the third of these processes is usually regarded as
the type of delamination, it does not, so far as I know,
occur in nature, but is most nearly approached in Geryonia
(fig. 6).

The first type of delamination is found in the Cerato-
spongise, some Silicispongiae (?), and in many Hydrozoa and
Actinozoa, and in Nemertea and Nematelminthes [Gordi^
oidea?). The second type occurs in many Porifera [^Calci-
spongi(B (^Ascetta), Myxospongi<2~\y and in some Coelenterata,
and Brachiopoda [Thecidium) .

Tig. 6. — Diagramatic Figures showing the Felamination of the Fmhryo
of Geryonia. (After Eol.) A. Stage at the commencement of the
delamination ; the dotted lines show the course of the next planes of
division. B. Stage at the close of the delamination, cs. segmentation
cavity ; a. endoplasm ; b. ectoplasm.

Delamination and invagination are undoubtedly the two
most frequent modes in which the layers are differentiated, but
there are in addition several others. In the first place the
whole of the Tracheata (with the apparent exception of the
Scorpion) develop, so far as is known, on a plan of their
own, which approaches delamination. This consists in the
appearance of a superficial layer of cells enclosing a central
yolk mass, which corresponds to the hypoblast (figs. 7 and
17). This mode of development might be classed under
delamination were it not for the fact that the early develop-
ment of the Crustacea is almost the same, but is subsequently
followed by an invagination (fig. 11), which apparently
corresponds to the normal invagination of other types.
There are grounds for thinking that the tracheate type of
formation of the epiblast and hypoblast is a secondary
modification of an invaginate type.

The type of some Turbellaria {Stijlocopsis ponticus) and

GERMINAL LAYERS OF THE EMBRYO. <«od

that of Nephelis amongst the Discophora is not capable of
being reduced to the invaginate type.

A

Pig. 1 .■^Segmentation and Formation of the Blastoderm in Chelifer. (After
Metscbnikoff.) In A the ovum is divided into a number of separate
segments. In B a number of small eells have appeared {bL), whieh
form a blastoderm enveloping the large yolk spheres. In C the blasto-
derm has beeome divided into two layers.

The development of almost all the parasitic groups, i.e, the
Trematoda,the Cestoda, the Acanthocephala, and the Lingua-
tulida, and also of the Tardigrada, Pycnogonida, and other
minor groups, is too imperfectly known to be classed with
either the delaminate or invaginate types.

It will, I think, be conceded on all sides that, if any of the
ontogenetic processes by which a gastrula form is reached are
repetitions of the process by which a single two-layered gas-
trula was actually developed from a compound Protozoon,
these processes are most probably either invagination or de-
lamination.

The much disputed questions which have been raised about
the gastrula and planula theories, originally put forward by
Haeckel and Lankester, resolve themselves then into the
single question, whether any, and if so which, of the onto-
genetic processes by which the gastrula is formed are repe-
titions of the phylogenetic origin of the gastrula.

It is very difficult to bring forward arguments of a con-
clusive kind in favour of either of these processes. The
fact that delaminate and invaginate gastrultc are in several

254

F. M. BALFOUR.

instances found coexisting in the same group renders it
certain that there are not two independent phyla of the
Metazoa, derived respectively from an invaginate and a
delaminate gastrula.^

The four most important cases in which the two processes
coexist are the Porifera, the Coelenterata, the Nemertea, and
the Brachiopoda. In the cases of the Porifera and Coelen-
terata, there do not appear to me to be any means of decid-
ing which of these processes is derived from the other ; but
in the Nemertea and the Brachiopoda the case is different.
In all the types of Nemertea in which the development is
relatively not abbreviated there is an invaginate gastrula,
while in the types with a greatly abbreviated development
there is a delaminate gastrula. It would seem to follow from
this that a delaminate gastrula has here been a secondary
result of an abbreviation in the development. In the Bra-
chiopoda, again, the majority of types develop by a process
of invagination, while Thecidium appears to develop by
delamination ; here, also, the delaminate type would appear
to be secondarily derived from the invaginate.

If these considerations are justified, delamination must be
in some instances secondarily derived from invagination ;
and this fact is so far an argument in favour of the more
primitive nature of invagination ; though it by no means
follows that invagination contains a repetition of the steps
by which the Metazoa were derived from the Protozoa.

It does not, therefore, seem possible to decide conclusively
in favour of either of these processes by a comparison of the
cases where they occur in the same groups.

The relative frequency of the two processes supplies us
with another possible means for deciding between them ; and
there is no doubt that here again the scale inclines towards
invagination. It must, however, be borne in mind that the
frequency of the process of invagination admits of another
possible explanation. There is a continual tendency for

^ It is not difficult to picture a possible derivation of delamination from
invagination, while a eomparison of the formation of the inner layers
(mesoblast and hypoblast) in Ascetta (amongst the sponges) and in the
Echinoderraata shows a very simple way in which it is possible to conceive
of a passage of delamination into invagination. In Ascetta the cells, which
give rise to the mesoblast and hypoblast, are budded off from the inner wall
of the blastosphere, especially at one point ; while in Eehinodermata
(fig. 2) there is a small invaginated sac which gives rise to the hypoblast,
while from the walls of this sac amoeboid cells are budded off whieh give rise
to the mesoblast. If we suppose the hypoblast cells budded off at one point
in Ascetta gradually to form an invaginated sac, while tlie mesoblast cells
continued to be budded off as before, we should pass from the delaminate
type of Ascetta to the invaginate type of an Echinodcrin.

Terminal layers of the embryo. 256

the processes of development to be abbreviated and simpli-
fied, and it is quite possible that the frequent occurrence
of invagination is due to the fact of its being, in most
cases, the simplest means by which the two-layered condition
can be reached. But this argument can have but little weight
until it can be shown that invagination is a simpler process
than delamination.

If it were the case that the blastopore had in all types
the same relation to the adult mouth, there would be strong
grounds for regarding the invaginate gastrula as an ancestral
form ; but the fact that this is by no means so is an argu-
ment of great weight in favour of some other explanation
of the frequency of invagination.

The force of this consideration can best be displayed by a
short summary of the fate of the blastopore in different forms.

The fate of the blastopore is so variable that it is diffi-
cult even to classify the cases which have been described.

The forms which have been classed together under the

(1) It becomes the permanent mouth in the following forms d

Ccelenterata. — Pelagia, Cereanthus.

Turbellaria. — Leptoplana (?), Thysanozoou.

Nemertea. — ^Pilidium, larvae of the type of Desor.

MoUusca. — In numerous examples of most molluscan groups, ex-
cept the Cephalopoda.

Chcetopoda. — Most Oligochseta, and probably many Polychseta.

Gephyrea. — Phascolosoma, Phoronis.

Nematelminthes. — Cucullanus.

(2) It closes in the position where the mouth is subsequently formed.

Ccelenterata.—Citm\}h.ov^ (?).

MoUusca. — In numerous examples of most Molluscan groups, except
the Cephalopoda.

Crustacea. — Cirripedia (?), some Cladocera (Moiiia) (?).

(3) It becomes the permanent anus.

MoUusca. — Paludina.

Chcetopoda. — Serpula and some other types.

Echinodermata. — Almost universally, except amongst the Crinoidea.

(4) It closes in the position where the anus is subsequently formed.

Echinodermata. — Crinoidea.

(5) It closes in a position which does not correspond or is not known
to correspond^ either with the future mouth or anus. — Porifera : Sycandra.
Ccctenierata : Chrysaora,* Aurelia.* Nemertea ;* some larvae which develop
without a metamorphosis, liotifera.* MoUusca : Cephalopoda. Polyzoa.*
Brachiopoda: Argiope, Terebratula, Terebratulina. Chcetopoda: Euaxes.
Discophora : Clepsine. Gephyrea : Bonellia.* Chcetognatha. Crustacea :
Decapoda. Chordata.

‘ The above list is somewhat tentative ; and future investigations will
probably show that many of the statements at present current about the
position ol the blastopore are inaccurate.

^ The forms in which the position of the blastopore in relation to the
mouth or anus is not known are marked with an asterisk.

256

F. M. BALFOUR.

last heading vary considerably in the character of the blas-
topore. In some cases the fact of its not coinciding either
with the mouth or anus appears to be due simply to the
presence of a large amount of food-yolk. The cases of the
Cephalopoda, of Euaxes, and perhaps of Clepsine and
Bonellia, are to be explained in this way : in the case of
all these forms, except Bonellia, the blastopore has the form
of an elongated slit along the ventral surface. This type
of blastopore is characteristic of the Mollusca generally,
of the Polyzoa, of the Nematelminthes, and very possibly
of the Chsetopoda and Discophora. In the Brachio-
poda and the Chsetognatha (fig. 12 b) the blastopore is
situated, so far as can be determined, behind the future
anus. In many Decapoda, amongst the Crustacea, the blas-
topore is placed behind, but not far from, the anus. In
the Chordata it is also placed posteriorly, and, remarkably
enough, remains, in a large number of forms, for some time
in connection with the neural tube by a neurenteric
canal.

The great variations in the character of the gastrula, in-
dicated in the above summary, go far to show that if
the gastrulee, as we find them in most types, have any an-
cestral characters, these characters can only be very general
ones. This may best be shown by the consideration of a few
striking instances. The blastopore in Mollusca has an elon-
gated slit-like form, extending along the ventral surface from
the mouth to the anusj in Echinodermata it is a narrow
pore, remaining as the anus. In most Chsetopoda it is a
pore remaining as the mouth, but in some as the anus. In
Chordata it is a posteriorly-placed pore, opening into the
neural canal.

• It is clearly out of the question to explain these differ-
ences in connection with the characters of ancestral forms.
They can only be accounted for as secondary adaptations for
the convenience of development.

The epibolic gastrula of Mammalia is a still more striking
case of a secondary embryonic process, and is not directly
derived from the gastrula of the lower Chordata. It pro-
bably originated in connection with the loss of food-yolk
which took place on the establishment of a placental nutri-
tion for the foetus. The epibolic gastrula of the Scorpion,
of Isopods, and of other Arthropoda, seems also to be a derived
gastrula. These instances of secondary gastrulae are very
probably by no means isolated, and should serve as a warning
against laying too much stress upon the frequency of the
occurrence of invagination. The great influence of the food-

GERMIN.\L LAYERS OF THE EMBRYO.

257

yolk upon the early development might be illustrated by
numerous examples, especially amongst the Chordata.^

If the descendants of a form with a large amount of food-
yolk in its ova were to have ova with but little food-yolk,
the type of formation of the germinal layers which would
thereby result would be by no means the same as that of the
ancestors of the forms with much food-yolk, but would pro-
bably be something very different, as in the case of Mammalia.
Yet amongst the countless generations of ancestors of most
existing forms, such oscillations in the amount of the food-
yolk must have occurred in a large number of instances.

The whole of the above considerations point towards the
view that the formation of the hypoblast by invagination, as it
occurs in most forms at the present day, can have no special
phylogenetic significance, and that the argument of frequency,
in favour of invagination as opposed to delamination, is not
of prime importance.

A third possible method of deciding between delamination
and invagination is to be found in the consideration as to
which of these processes occurs in the most primitive forms.
If there were any agreement amongst primitive forms as to
the type of their development this argument might have
some weight. On the whole, delamination is, no doubt,
characteristic of the most primitive types, but the not in-
frequent occurrence of invagination in both the Coelen-
terata and the Porifera — the two groups which would on all
hands be admitted to be amongst the most primitive —
deprives this argument of much of the value it might other-
wise have.

To sum up : in the present state of our knowledge there
are no satisfactory data for deciding which of the two
processes is the more primitive ; nor, considering the almost
indisputable fact that both these processes have in many
instances had a purely secondary origin, can any valid
arguments be produced to show that either of them
reproduces the mode of passage between the Protozoa and
the ancestral two-layered Metazoa. These conclusions
do not, however, throw any doubt upon the fact that the
gastrula, however evolved, was a primitive form of the
Metazoa ; since this conclusion is founded upon the actual
existence of adult gastrula forms independently of their
occurrence in development.

Though embryology does not at present furnish us with
an answer to the question how the Metazoa became developed

^ Vide r. M. Balfour, “ A comparison of the early stages iu the deve-
lopment of Vertebrates,” ‘ Quart. Journ. Micr. Sci./ vol. 1875.

258 F. M. BALFOUii.

from the Protozoa, it is nevertheless worth while reviewing
some of the processes by which this can be conceived to
have occurred.

On purely a priori grounds there is in my opinion more
to be said for invagination than for any other view.

On this view we may suppose that the colony of Pro-
tozoa in the course of conversion into Metazoa had the
form of a blastosphere ; and that at one pole of this a
depression appeared. The cells lining this depression we
may suppose to have been amoeboid, and to have carried
on the work of digestion ; while the remaining cells were
probably ciliated. The digestion may be supposed to have
been at first carried on in the interior of the cells, as in the
Protozoa ; but, as the depression became deeper (in order
to increase the area of nutritive cells and to retain the food)
a digestive secretion probably became poured out from the
cells lining it, and the mode of digestion generally character-
istic of the Metazoa was thereby inaugurated. It may be noted
that an intra-cellular protozoon type of digestion persists in
the Porifera, and appears also to occur in many Coelenterata,
Turbellaria, &c., though in most of these cases both kinds
of digestion probably occur simultaneously.^

Another hypothetical mode of passage which fits in with
delamination has been put forward by Lankester, and is
illustrated by fig. 8. He supposes that at the blasto-
sphere stage the fluid in the centre of the colony acquired
special digestive properties ; the inner ends of the cells
had at this stage somewhat different properties to the
outer, and the food was still incepted by the surface of
the cells (fig. 8, 3). In a later stage of the process the
inner portions of the cells became separated off as the
hypoblast, while the food, though still ingested in the form
of solid particles by the superficial cells, was carried through
the protoplasm into the central digestive cavity. Later (fig.
8, j), the point where the food entered became localised, and
eventually a mouth became formed at this point.

The main objection which can be raised against Lankester’s
view is that it presupposes a type of delamination which does
not occur in nature except in Geryonia.

Metschnikoff has propounded a third view with reference
to delamination. He starts as before with a ciliated blasto-
sphere. He next supposes the cells from the walls of this
to become budded off into the central cavity, as in Eucope

* J. Parker, “ On the Histology of Hydra fusca’' ‘ Quart. Journ. Micr.
Science and El. Metschnikoff, “ Ucb. die lutracellulare Verdauuug boi
Coclenterateu,” ‘ Zoologischer Auzeiger/ No. 5G, vol. hi, 1880.

GERMINAL LAYERS OF TIIE EMBRYO.

259

(fig. 5), and to lose their cilia. These cells give rise to an
internal parenchyma, which carries on an intra-cellular di-

Fig. 8. — Diagram showing the Formation of a Gastrula by Delamination.
(From Lankester.) Fig. 1, ovum ; fig. 2, stage in segmentation ; fig. 3,
commencement of delamination after the appearance of a central cavity ;
fig. 4, delamination completed, mouth forming at M. In figs. 1, 2, and
Z^Fc. is ectoplasm, and Fn. is endoplasm. In fig. 4, Fc. is epiblast,
and Fn. hypoblast. F. food particles.

gestion. At a later stage a central digestive cavity is supposed
to be formed. This view of the passage from the protozoon
to the metazoon state, though to my mind improbable in
itself, fits in very well with the ontogeny of the lower
Hydrozoa.

Another view has been put forward by myself,^ to the effect
that the amphiblastula larva of Calcispongiae may be a
transitional form between the Protozoa and the Metazoa,
composed of a hemisphere of nutritive amoeboid cells and a
hemisphere of ciliated cells. The absence of such a larval
form in the Coelenterata and higher Metazoa is opposed,,
however, to this larva being regarded as a transitional form,
except for the Porifera.

It is obvious that so long as there is complete uncer-
tainty as to the value to be attached to the early develop-
mental processes, it is not possible to decide from these
processes whether there is only a single Metazoon phylum
or whether there may not be two or more such phyla. At

* F. M. Balfour, ‘A Treatise on Comparative Embryology,’ vol. i, p. 122.

260

F. M. BALFOUR.'

the same time there appear to be strong arguments for re-
garding the Porifera as a phylum of the Metazoa derived
independently from the Protozoa. This seems to me to be
shown (1) by the striking larval peculiarities of the Porifera ;
(2) by the early development of the mesoblast in the Porifera,
which stands in strong contrast to the absence of this layer in
the embryos of most Coelenterata ; and above all, (3) by the re-
markable characters of the system of digestive channels. A
further argument in the same direction is supplied by the
fact that the germinal layers of the Sponges very probably
do not correspond physiologically to the germinal layers of
other types. Whether or no the amphiblastula larva is, as
suggested above, to be regarded as the larval ancestor of the
Porifera must be left as an open question.

The question as to how far there is a complete homology
between the two primary germinal layers throughout the
Metazoa was the third of the questions propounded.

Since there are some Metazoa with only two germinal
layers, and other Metazoa with three, and since, as is shown in
the following section, the third layer or mesoblast can only be
regarded as a derivative of one or both the primary layers,
it is clear that a complete homology between the two
primary germinal layers does not exist.

That there is a general homology appears on the other
hand hardly open to doubt. ^

The primary layers are usually continuous w’ith each other,
near one or both (when both are present) the openings of
the alimentary tract.

As a rule an oral and anal section of the alimentary tract
— the stomodaeum and proctodaeum — are derived from the
epiblast ; but the limits of both these sections are so
variable, sometimes even in closely allied forms, that it
is difficult to avoid the conclusion that there is a border-
land between the epiblast and hypoblast, which appears
by its development to belong in some forms to the epi-
blast and in other forms to the hypoblast. If this is not
the case it is necessary to admit that there are instances in
which a very large portion of the alimentary canal is formed
of epiblast. In some of the Isopods, for example, the
stomodeeum and proctodeeum give rise to almost the whole
of the alimentary canal with its appendages, except tha
liver.

The origin of the Mesoblast. — The diploblastic condition of
the organism preceded, as we have seen, the triploblastic.
The epiblast during the diploblastic condition was, as appears
from such forms as Hydra, especially the sensory and pro-

GERMINAL LAYERS OF THE EMBRYO.

261

tective layer, while the hypoblast was the secretory layer ;
both layers giving rise to muscular elements. It must not,
however, be supposed that in the early diploblastic ances-
tors there Avas a complete differentiation of function, but
there is reason to think that both the primary layers
retained an indefinite capacity for developing into any
form of tissue.^ The fact of the triploblastic condition
being later than the diploblastic proves in a conclusive way
that the mesoblast is a derivate of one or both the primary
layers. In the Coelenterata Ave can study the actual origin
from the tAvo primary layers of various forms of tissue Avhich
in the higher types become the mesoblast.^ This fact, as
Avell as general a priori considerations, conclusively prove
that the mesohlast did not at first originate as a mass of
independent cells between the txoo primarxj layers j hut that in
the first instance it arose as histological differentiations of the
two layers, and that its condition in the embryo as an inde-
pendent layer of undifferentiated cells is a secondary condi-
tion, brought about by the general tendency towards a simpli-
fication of development, and a retardation of histological
differentiation.^

In addition to the probably degraded Dicyemidse and Or-
thonectidse, the Coelenterata are the only group in Avhich a
completely differentiated mesoblast is not ahvays present.
In other Avords, the Coelenterata are the only group in Avhich
there is not found in the embryo an undifferentiated mass
of cells from which the majority of the organs situated
betAA^en the epidermis and the alimentary epithelium are
dveloped.

^ The Hertwigs (No. 13) have shoAVii that nervous structures are deve-
loped in the hypoblast in the Actinozoa.

“ There is considerable confusion in the use of the names for tlie em-
bryonic layers. In some cases various tissues formed by differentiations
of" the primary layers have been called mesoblast. Schultze, and more
recently the Hertwigs, have pointed out the inconvenience of this nomen-
clature. In some of the Coelenterata it is difficult to decide in certain
instances {e.g. Syinpodium) whether the cells which give rise to a particular
tissue of the adult are to be regarded as forming a mesoblast, i. e. a middle
undifferentiated layer of cells, or arise as already histologically differentiated
elements from one of the primary layers. The attempt to distinguish by^a
special nomenclature the cpiblast and hypoblast after and before the separa-
tion of the mesoblast, which has been made by Allen Thomson (No. 1),
appears to me inconvenient in practice. A proposal of the Hertwigs to
adopt special names for the outer and inner limiting membranes of the adult,
and for the interposed mass of organs, appears to me unnecessary, and
only likely to introduce confusion into an already complicated nomenclature.

^ The causes which give rise to a retardation of histological differentiation
will be dealt with in a sequel to this paper treating of larval characters and
larval forms.

262

F. M. BALFOUR.

Tlie organs invariably derived, in the triploblastic forms,
from the mesoblast, are the vascular and lymphatic systems,
the muscular system, and the greater part of the connective
tissue and the excretory system. On the other hand, the
nervous system and organs of sense (with a few possible
exceptions), the epithelium of most glands, and a few ex-
ceptional connective-tissue organs, i. e. the notochord, are
developed from the two primary layers.

The fact of the first-named set of organs being invariably
derived from the mesoblast.points to the establishment of the
two following propositions : — (1) That with the differentiation
of the mesohlast as a distinct layer hy the process already
explained, the two primary layers lost for the most part the
capacity they primitively possessed of giving rise to muscular
and connective-tissue differentiations. (2) That the mesohlast
throughout the triploblastic Metazoa, in so far as these forms
have sprung from a common triploblastic ancestor, is an
homologous structure.

The second proposition follows from the first. The meso-
blast can only have ceased to be homologous throughout the
triploblastica by additions from the two primary layers, and
the existence of such additions is negatived by the first pro-
position.

These two propositions, which hang together, are possibly
only approximately true. In the first place, it is quite
possible that fresh differentiations from the two primary
layers may have arisen after the triploblastic condition had
been established, and by the process of simplification of
development and precocious segregation, as Lankester calls
it, have become merged in the normal mesoblast ; or such
differentiations may have taken place in forms, the deve-
lopment of which has not yet been investigated. Had this,
however, been a frequent occurrence, it is hardly likely that
no instance of it should have been recorded for the muscular
system and connective tissue so that it is probable that
the muscular system of all existing triploblastic forms has
been differentiated from the muscular system of the an-
cestor of the triploblastica. In the case of other tissues
there are a few instances Avhich might be regarded as
examples of an organ primitively developed in one of the
two primary layers having become secondarily carried
into the mesoblast. The notochord has sometimes been
cited as such an organ, hut it now appears probable that its
hypoblastic origin can always be demonstrated. The de-

^ The connective-tissue test of the Tunicata, though derived from tlie
cjpiblast, is not really an example of such a differentiation.

GERMINAL LAYERS OF THE EMBRYO.

263

velopment of the generative organs in the Invertebrata is still
very imperfectly known, but it is possible, that although
usually developed in the mesoblast, they sometimes (e. g.
Insecta) retain their primitive development from the epiblast.
The nervous system, although imbedded in mesoblastic

Fig. ^.—Epibolic Gastrula of Bonellia. (After Spengel.) A. Stage when
the four Jiypoblast cells are nearly enclosed. B. Stage after the forma-
tion of the mesoblast has commenced by an infolding' of the lips of the
blastopore, ep. Epiblast ; me. mesoblast ; bl. blastopore.

derivates in the adults of all the higher triploblastica, retains
with marvellous constancy its epiblastic origin (though it
is usually separated from the epiblast prior to its histogenic
differentiation) ; ^ yet in the Cephalopoda, and some other
Mollusca, the evidence is in favour of its developing in the
mesoblast. Should future investigations confirm these con-
clusions, a good example will be afforded of an organ chang-
ing the layer from which it develops. The explanation of
such a change would be precisely the same as that already
given for the mesoblast as a whole.

The actual mode of origin of various tissues, which in the
true triploblastic forms constitute the mesoblast, can be traced
in the Coelenterata.^ In this group the epiblast and hypoblast
both give rise to muscular and connective-tissue elements ;
and although the main part of the nervous system is formed
in the epiblast, it seems certain that in some types nerves may
be derived from the hypoblast.^ These facts are extremely in-

‘ The reader is referred for this subject to the extremely valuable
memoirs which have been recently published by the Hertwigs, especially
(Ao. 13). He will find a general account of the subject written before
the appearance of the Hertwigs’ memoir in p. 1 19 and 150 of volume I of
my treatise on ‘Comparative Embryology.’

’ It would be interesting to know about the history of the various ner-
vous structures found in the walls of the alimentary tract in the hio-lier
forms. 1 have shown (‘ Development of Elasmobranch Fishes,’ p. 172)
that the central part of the sympathetic system is derived from the epiblast.
VOL. X.\. NEW SER. S

264

F, M. BALFOUR.

terestincT, but it is by no means certain that any conclusions
can be directly drawn from them as to the actual origin of the
mesoblast in the triploblastic forms, till we know from what
diploblastic forms the triploblastica originated. All that they
show is that any part of the mesoblast may have originated
from either of the primitive layers.

Fig. 10. — The Transverse Sections through Embryos of Hydrophilus piceus.
(After Kowalevsky.) A. Section through an embryo at the point
where the two germinal folds most approximate. B. Section through
an embryo, in the anterior region wher^ the folds of the amnion have
not united, gg. germinal groove; me. mesoblast; am. amnion; yk. yolk.

For further light as to the origin of the mesoblast, it is
necessary to turn to its actual development.

The following summary illustrates the more important
modes in which the mesoblast originates.

1. It grows inwards from the lips of the blastopore as
a pair of bands. In these cases it may originate (1) from
cells which are clearly hypoblastic, (2) from cells which are
clearly epiblastic, (3) from cells which cannot be regarded as
belonging to either layer.

Mollusca. — Gasteropoda, Cephalopoda, and Lamellibran-
chiata. In Gasteropoda- and Lamellibranchiata it some-
times originates from a pair of cells at the lips of the
blastopore, though very probably some of the elements sub-
sequently come from the epiblast ; and in Cephalopoda as a
ring of cells round the edge of the blastoderm.

Polyzoa Entoprocta. — It originates from a pair of cells
at the lips of the blastopore.

ChcBtopoda. — Euaxes. It arises as a ridge of cells at the
lips of the blastopore (fig. 3).

GERMINAL LAYERS OF THE EMBRYO.

265

Gephyrea. — Bonellia. It arises (fig. 9) as an infolding
of the epiblastic lips of the blastopore.

Fig 11. — Figures illustrating the Development of Astacus. (From Parker,
after Reichenbacb.) A. Section through part of the ovum during seg-
mentation. n. nuelei ; w.y. white yolk ; yp. yolk pyramids ; c. central
yolk mass. B and C. Longitudinal sections of the gastrula stage.
a. archenteron; blastopore ; mesoblast; ec. epiblast ; hypo-
blast, distinguished from epiblast by shading. D. Highly magnified
view of anterior lip of blastopore, to show the origin of the primary
mesoblast from the wall of the archenteron. p.ms. primary mesoblast ;
ec. epiblast ; en. hypoblast. E. Two hypoblast cells to show the amoeba-
like absorption of yolk spheres, y. yolk ; n. nucleus ; g. pseudopodial
process. F. Hypoblast cells giving rise endogenously to the secondary
mesoblast {sms?) ; n. nucleus.

Nematelminthes . — -Cucullanus. It grows backwards from
the hypoblast cells at the persistent oral opening of the
blastopore.

Tracheata. — Insecta. It grows inwards from the lips of

the germinal groove (fig. 10), which probably represent the
remains of a blastopore. Part of the mesoblast is probably
also derived from the yolk- cells. A similar though more
modified development of the mesoblast occurs in the
Araneina (fig. IT).

Crustacea. — Decapoda. It partly grows in from the hypo-
blastic lips of the blastopore, and is partly derived from the
yolk-cells (fig. 11).

2. The mesoblast is developed from the walls of hollow

266

F. M. BALFOUR.

outgrowths of the archenteron, the cavities of which become
the body cavity.

Brachiopoda. — The walls of a pair of outgrowths form the
whole of the mesoblast.

Chcetognatha. — The mesoblast arises in the same manner
as in the Brachiopoda (fig. 12).

Echinodermata. — The lining of the peritoneal cavity is
developed from the walls of outgrowths of the archenteron, but
the greater part of the mesoblast is derived from the amoeboid
cells budded off from the walls of the archenteron (fig. 13).

Enteropneusta {Balanoglossus) , — The body cavity is de-
rived from two pairs of alimentary diverticula, the walls of
Avhich give rise to the greater part of the mesoblast.

I'lG. 12. — Three Stages m the Development of Sagitta. (A and C after
Butschli, and B after Kowalevsky.) The three embryos are represented
in the same positions. A. Represents the gastrula stage. B. Repre-
sents a succeeding stage, in which the primitive archenteron is com-
mencing to be divided into three. C. Represents a later stage, in
which tlie mouth involution {vi) has become continuous with alimentary
tract and the blastopore has become closed, m. mouth ; al. alimentary
canal ; ae. archenteron , bl.py blastopore ; pv. perivisceral cavity ;
sp. splanchnic mesoblast ; so. somatic mesoblast ; ge. generative organs.

Chordata. — Paired archenteric outgrowths give rise to the
whole mesoblast in Amphioxus (fig. 14), and the mode of
formation of the mesoblast in other Chordata is probably
secondarily derived from this.

3. The cells which will form the mesoblast become
marked out very early, and cannot be regarded as definitely
springing from either of the primary layers.

Turhellaria. — Leptoplana (fig. 15), Planaria polychroa (?)

Chcetopoda. — Lumbriciis, &c.

iHscophora.

It is very possible that the cases quoted under this head
ought more properly to belong to group 1.

GERMINAL LAYERS OF THE EMBRYO.

267

4. The mesoblast cells are split off from the epiblast.
Nemertea. — Larva of Desor. The mesoblast is stated to

be split off from the four invaginated discs.

5. The mesoblast is split off from the hypoblast.
Nemertea. — Some of the types without a metamorphosis.
Mollusca. — Scaphopoda. It is derived from the lateral and

ventral cells of the hypoblast.

Gephjrea. — Phascolosoma.

Vertebrata. — In most of the higher Vertebrata the meso-
blast is derived from the hypoblast (fig. 16). In some types
(f. e, most of the Icthyopsida) the mesoblast might be
described as originating at the lips of the blastopore.

Fig. 13. — Longitudinal Section through an Embryo ofCucumaria doHolum at

the end of the fourth day. Vpv. vaso-peritoneal vesicle : ME. mesenteron ;

Blp.y Ptd. blastopore, proctodaeum.

6. The mesoblast is derived from both germinal layers.

Tracheata. — Araneina (fig. 17). It is derived partly from
cells split off from the epiblast and partly from the yolk-cells ;
but it is probable that the statement that the mesoblast is
derived from both the germinal layers is only formally accu-
rate ; and that the derivation of part of the mesoblast from
the yolk-cells is not to be interpreted as a derivation from
the hypoblast.

The conclusions to be drawn from the above summary are
by no means such as might have been anticipated. The
analogy of the Coelenterata would lead us to expect that the
mesoblast would be derived partly from the epiblast and
partly from the hypoblast. Such, however, is not for the
most part the case, though more complete investigations may

26h

F. M. BALFOUR.

show that there are a greater number of instances in which
the mesoblast has a mixed origin than might be supposed
from the above summary.

Fig. 14. — Sections of an Amphioxus 'Embryo at three Stages. (After Kow-
alevsky.) A. Section at gastrula stage. B. Section of a somewhat
older embryo. C. Section through the anterior part of still older
embryo, np. neural plate ; n.c. neural canal ; rues, archenteron in A,
and mesenteron in B and C ; ch. notochord ; so. mesoblastic somite.

I have attempted to reduce the types of development of
the mesoblast to six ; but owing to the nature of the case it
is not always easy to distinguish the first of these from the
last four. Of the six types the second will on most hands
be admitted to be the most remarkable. The formation of
hollow outgrowths of the archenteron, the cavities of which
give rise to the body cavity, can only be explained on the
supposition that the body cavity of the types in which
such outgrowths occur are derived from diverticula cut off
from the alimentary tract. The Iming epithelium of the
diverticula — the peritoneal epithelium — is clearly part of the
primitive hypoblast, and this part of the mesoblast is clearly
hypoblastic in origin.

Fig. 15. — Sections through the Ovum of Leptoplana tremellaris in three
Stages of Development, (After Hallez.) ep. epiblast ; m. mesoblast ;
hy. yolk cells (hypoblast) ; bl. blastopore.

In the case of the Chsetognatha (Sagitta), Brachiopoda,
and Amphioxus, the whole of the mesoblast originates from
the walls of the diverticula ; while in the Echinodermata the
walls of the diverticula only give rise to the vaso-peritoneal
e])itheliuin, the remainder of the mesoblast being derived from
amoeboid cells which spring from the walls of the archenteron

GERMIx^AL LAYERS OF THE EMBRYO.

269

before the origin to the vaso-peritoneal outgrowths (figs. 2
and 13).

The first of these processes suggests the view that the
whole of the mesoblast primitively arose by a process of
histogenic differentiation from the walls of the archenteron.
This view, which was originally put forward by myself
(No. 4), appears at first sight very improbable, but it re-
ceives great support from the enormous development of the
hypoblastic muscular system (Hertwigs, No. 13) in many
Actinozoa. Lankester (No. 17), on the other hand, has urged
that the mode of origin of the mesoblast in the Echinoder-
mata is more primitive ; and that the amoeboid cells which
here give rise to the muscular and connective tissue repre-
sent cells which originally arose from the whole inner surface
of the epiblast. It is, however, to be noted that even in the
Echinodermata the amoeboid cells arise from the hypohlast,
and their mode of origin may, therefore, be used to support
the view that the main part of the muscular system of higher
types is derived from the primitive hypoblast.

Reserving for the moment the question as to what con-
clusions can be deduced from the above facts as to the origin
of the mesoblast, it is important to determine how far the
facts of embryology warrant us in supposing that in the
whole of the triploblastic forms the body cavity originated
from the alimentary diverticula. There can be but little
doubt that the mode of origin of the mesoblast in the Verte-
brata, as two solid plates split off from the hypoblast in
which a cavity is secondarily developed, is an abbreviation of

Fig. 1G. — Two Sections of a young TUlasmohranch Embryo, to show the meso-
blast spilt off as two lateral rnasses from the hypoblast, mg. medullary
groove; ^7?. epiblast ; »/2. mesoblast; hypoblast.

270

F. M. BALFOUU.

the process observable in Amphioxus ; but this process
approaches in many forms of Vertebrata to the ingrowth of
the mesoblast from the lips of the blastopore.

It is, therefore, highly probable that the paired ingrowths
of the mesoblast from the lips of the blastopore may have
been in the first instance derived from a pair of archenteric
diverticula. This process of formation of the mesoblast is,
as may be seen by reference to the summary, the most
frequent.^

While there is no difficulty in the view that the body
cavity may have originated from a pair of enteric diverticula
in the case of the forms where a body cavity is present, there
is a considerable difficulty in holding this view, for forms
in which there is no body cavity distinct from the alimen-
tary diverticula.

Tig. 17. — Section through the Embryo of Agelena labyrinthica. The section
is represented with the ventral plate directed upwards. In the ventral
plate is seen a keel-like thickening, which gives rise to the main mass of
hie mesoblast. yk. Yolk divided into large polygonal cells, in several of
which are nuclei.

Of these types the Platyelminthes are the most striking.
It is, no doubt, possible that a body cavity may have existed

^ The wide occurrence of this process was first pointed out by E.abl. He
holds, however, a peculiar modification of the gastraea theory, for which I
must refer the reader to his paper (No. 23) ; according to this theory the
mesoblast has sprung from a zone of cells of the blastosphere, at the
junction between the cells which will be invaginated and the epiblast
cells. In the bilateral blastosphere, from which he holds that all the
higher forms (bilateralia) have originated, these cells had a bilateral
arrangement, and thus the bilateral origin of the mesoblast is explained.
The origin of the mesoblast from the lips of the blastopore is explained by
the position of its mother-cells in the blastosphere. It need scarcely be
said that the views already put forward as to the probable mode of origin
of the mesoblast, founded on the analogy of the Coelenterata, are quite
incompatible with liabl’s theories.

GERMINAL LAYERS OF THE EMBRYO. 271

in the Platyelminthes, and become lost, and even in some
cases replaced physiologically by alimentary diverticula.
The usual view of the primitive character of the Platyel-
minthes, which has much to support it, is, however, opposed
to the idea that the body cavity has disappeared.

If Kowalewsky^ is right in stating that he has found a
form intermediate between the Coelenterata and the Platyel-
minthes, there will be strong grounds for holding that the
Platyelminthes are, like the Coelenterata, forms the ancestors
of which were never provided with a body cavity.

Perhaps the triploblastica are composed of two groups, viz.

(1) a more ancestral group (the Platyelminthes), in which
there is no body cavity as distinct from the alimentary, and

(2) a group descended from these, in which two of the ali-
mentary diverticula have become separated from the alimen-
tary tract to form a body cavity (remaining triploblastica).
However this may be, the above considerations are sufficient
to show how much there is that is still obscure with refer-
ence even to the body cavity.

If embryology gives no certain sound as to the questions
just raised with reference to the body cavity, still less is it
to be hoped that the remaining questions with reference
to the origin of the mesoblast can be satisfactorily answered.
It is clear, in the first place, from an inspection of the sum-
mary given above, that the process of development of the
mesoblast is, in all the higher forms, very much abbreviated
and modified. Not only is its differentiation relatively
deferred, but it does not in most cases originate, as it must
have done to start with, as a more or less continuous sheet,
split off from one or both the primary layers. It originates
in most cases from the hypoblast, and although the con-
siderations already urged preclude us from laying very great
stress on this mode of origin ; yet, as suggested above, it
appears to me not impossible, judging from the analogy of the
Actinozoa, that the muscular system of the triploblastica
may have primitively mainly arisen from differentiations of
the hypoblast of the alimentary diverticula, which seem to
have given rise to the body cavity.

The great changes which have taken place in the de-
velopment of the mesoblast would be more intelligible on this
view than on the view of the mesoblast having primitively
largely originated from the epiblast. The presence of food-

‘ ‘ Zoologischer Auzeiger,’ No. 52, p. 140. This form Las been named by
Kovvalewsky Coleoplana Metschnikowii. Kowalewsky’s deseriptiou appears,
however, to be quite com))atible willi the view that this form is a creeping
Ctenophor, in no way related to the Turbellarians.

272

r. M. BALFOUR.

yolk is much more frequent in the hypoblast than in the epi-
blast ; and it is well known that a large number of the
changes in early development are caused by food-yolk. If,
therefore, the mesoblast has been derived from the hypoblast,
many more changes might be expected to have been intro-
duced into its early development than if it had been derived
from the epiblast. At the same time the hypoblastic origin of
the mesoblast would assist in explaining how it has come
about that the development of the nervous system is almost
always much less modified than that of the mesoblast, and
that the nervous system is not, as might, on the grounds of
analogy, have been anticipated, developed in the mesoblast.

List of Fallen,

(1) Allen Thomson. British Association Address, 1877.

(2) A. Agassiz. “Embryology of the Ctenophorse,” ‘Mem. Amer.
Acad, of Arts and Sciences,’ vol. x, 1874.

(3) K. E. voN Baer. ‘ Ueb. Entwicklungsgeschichte d. Thiere.’ K6-
nigsberg, 1828 — 1837.

(4) E. M. Balfour. “ A Comparison of the Early Stages in the De-
velopment of Vertebrates,” ‘Quart. Journ. of Micr. Sci.,’ vol.'xv, 1875.

(5) C. Claus. ‘Die Typeulehre u. E. Haeckel’s sg. Gastraea-theorie.’
Wien, 1874.

(6) C. Claus. ‘ Grundziige d. Zoologie.’ Marburg und Leipzig, 1879.

(7) C. Gegenbaur. ‘ Grundriss d. vergleichenden Anatomie.’ Leipzig,
1878. Vide also Translation. ‘Elements of Comparative Anatomy.’
Macmillan & Co., 1878.

(8) A. Gotte. ‘ Entwicklungsgeschichte d. Unke.’ Leipzig, 1874.

(9) E. Haeckel. ‘Studien z. Gastrsea-theorie,’ Jena, 1877; and
also ‘ Jenaische Zeitschrift,’ vols. viiiand ix, 1874-5.

(10) E. Haeckel. ‘ Schopfungsgeschichte,’ Leipzig. Vide also Trans-
lation. ‘ The History of Creation.’ King & Co., London, 1876.

(11) E. Haeckel. ‘ Anthropogenie,’ Leipzig. Vide also Translation.
‘ Anthropogeny.’ Kegan Paul & Co., London, 1878.

(12) B. Hatschek. “Studien iib. Entwicklungsgeschichte d. Anneli-
den,” ‘Arbeit a. d. Zool. Instit. d. XJniver. Wien,* von 0. Claus, 1878.

(15) 0. and R. Hertwig. “Die Actinien,” ‘Jenaische Zeitschrift,*
vol. xiii and xiv, 1879.

(14) Th. H. Huxley. ‘The Anatomy of Invertebrated Animals.*
Churchill, 1877.

(16) A. Kowalewsky. “ Embryologische Studien an Warmern u. Ar-
thropoden,*’ ‘Mem. Acad., Petersbourg,’ series vii, vol. xvi, 1871.

(16) E. R. Lankester. “ On the (terminal Layers of the Embryo as
the Basis of the Genealogical Classification of Animals,*’ ‘ Ann. and Mag.
of Nat. Hist.,’ 1873.

(17) E. R. Lankester. “Notes on Embryology and Classification,”
‘ Quart. Journ. of Micr. Sci.,’ vol. xvii, 1877.

(18) E. Metsciiinkoff. “Zur Entwicklungsgeschichte d. Kalk-
schwamme,” ‘ Zeit. f. Wiss. Zool.,’ vol. xxiv, 1874.

(19) E. Metsciiinkoff. “ Spongiologische Studien,” ‘ Zeit. f. Wiss.
Zool.,’ vol. xxxii, 1879.

GERMINAL LAYERS OP THE EMBRYO.

273

(20) A. S. P. Packard. ‘Life Histories of Animals, including Man,
or Outlines of Comparative Embryology.' Holt & Co., New York, 1876.

(21) H. Rathke. ‘ Abbandlungen z. Bildung und Entwicklungsgescli.
d. Menschen u. d. Tliiere.' Leipzig, 1833.

(22) C. Rabl. “Ueb. d. Entwick. d. Malermuschel,” ‘Jenaiscbe
Zeitsch.,* vol. x, 1876.

(23) C. Rabl. “ Ueb. d. Entwicklung. d. Tellerscbneke (Planorbis),”
‘ Morpb. Jabbucb,* vol. v, 1879.

(24) R. Remak. ‘Untersucb. iib. d. Entwick. d. Wirbeltbiere,’ 1855.

(25) Salensky. “ Bemerkungen iib. Haeckels Gastrsea-tb^rie,"
‘ Arcbiv f. Naturgescbichte,' 1874.

(26) E. Schafer. “ Some Teachings of Development,” * Quart. Journ.
of Micr. Science,’ vol. xx, 1880.

(27) A. Kolliker. ‘ Entwicklungsgescbicbte d. Menschen u. d. hob.
Tbiere,’ Leipzig, 1879.

274

DR. H. W. HUBIIECHT.

Hubrecht’s Researches on the Nervous System of
Nemertines.^ (With Plate XXIII).

The above cited paper, which has just been published by
the Amsterdam Academy of Sciences, and from which we
have copied figures 1 to 12 on Plate XXIII, gives a detailed
account of the situation and the structure of the central
nervous system in the Nemerteaus. After all that has been
said on this subject in the works and treatises of McIntosh^
QuatrefageSj Keferstein, and in a former paper of the same
author, it would seem superfiuous once more to go over the
same ground. Only it should be here remembered that the
histological structure was not, or very insufficiently, con-
sidered by those authors, whereas, in the present paper, due
space is allowed to the description of the histological
details. The minute structure of the so-called side-
organs ” (McIntosh’s cephalic sacs) is here also for the first
time minutely entered into.

A series of experiments is next recorded in favour of the
author’s hypothesis, that in the Lineidse these '^cephalic sacs”
with their internal ciliated canal and with the deep longitu-
dinal slits on each side of the head must be regarded as a
special apparatus serving for respiratory purposes, the oxygen
being taken hold of by the haemoglobin contained in the
nerve-cells belonging to the central nervous apparatus as it
stretches throughout the whole length of the animal, from
the head to the tip of the tail. Other arguments in favour
of this hypothesis are given, some of them derived from the
phylogenetic, others from the ontogenetic development of
these organs.

Finally, the paper discusses an explanation of the origin
of the dorsal nerve- cord of Vertebrates and the ventral nerve-
cord of Arthropods and Annelids out of originally paired
lateral cords. This explanation is suggested by the situation
of the latter in different genera of Nemerteans, and finds
itself in harmony with general views which were already
expressed on former occasions by Gegenbaur, Harting,
Balfour, and others. Two hitherto unknown facts resulting
from the author’s investigations are more specially brought
to bear upon this point ; 1st, the presence in all Hoplone-
mertini?i% yet examined on this head, of a commissure uniting
the two lateral nerve-cords and situated above the intestine

* Dr. H. W. Hubrecht, ‘Zur Anatomie uud Physiologie des Nerven-
systems der Nemertinen/ Mit Vier 4o Tafeln. ‘ Verhandelingen van de
Koninklyke Akaderaie van Wetenschappen te Amsterdam/ Dl. 1880.

RESEARCHES ON THE NERVOUS SYSTEM OF NEMERTINES. 275

immediately before the anus {Moseley describes such a com-
missure ill Pelago7iemertes) ; and 2nd, the displacement of
the lateral nerve-cords in the author’s new genus Langia,
which tend to approach each other on the dorsal side (not
ventrally as they do in the genera Drepanophorus and
Oerstedia)i at least in the anterior portion of the body. This
explanation would restore the homology between the dorsum
of Vertebrates on the one hand and of Arthropods and
Annelids on the other, a homology which of late years has
been put into serious doubt by the researches of Dohrn and
Semper, whose ingenious suggestions have as yet, however,
never definitely overcome certain primary objections inherent
to their views, which correspond as a whole with Geoffroy
St. Hilaire’s saying that insects are Vertebrates walking on
their backs.

After this rapid exposition of the contents of the
paper, we shall give a somewhat more detailed account
of certain parts of it. It should here be remarked that the
nomenclature of the genera and suborders employed is
that proposed by the author in a former paper,^ in which
he divides the Nemerteans into three suborders : Palji.one-
MERTiNi, with the genera Carmella, Cephalotlmx, Polia and
Valencinia: Schizonemertini, with the genera Linens,
Borlasia, Cerebratulus and Langia ; Hoplonemertini, with
the genera Amphiporus, Drepanopliorus , Tetrastemma, Oers-
tedia, Prosorhochmus, Rnd Nemertes.

A. The central nervous system, — As such the author
does not regard the cephalic ganglia only — as was done by
all his predecessors — but the so-called longitudinal nerves as
well, on account of the sheath of ganglion cells which unin-
terruptedly accompanies these trunks from their origin in
the cephalic lobes down to the extremity of the tail, in all
the genera without exception.

Ill the genus Carinella — which appears to be one of the
more primitive and less difierentiated — the whole central
nervous system is situated immediately under the epi-
dermal tissues, outside the muscular body wall (PI. XXIII,
fig- and the cephalic ganglion takes the form of a simple
anterior enlargement of the lateral trunks. No distinct
division into lobes can be detected in transverse sections ; the
ventral commissure is very broad, the dorsal commissure ex-
ceedingly thin; through the ring thus formed the proboscis
and its sheath passes (PI. XXIII, fig. 2). The mouth opens
behind and under the ganglion. The fibrous nerve-sub-

' The Genera of European Nemerteans Critically llevised,” ‘ Notes
from the Leyden Museum,’ vol. i, p. 193.

276

DR. H. W. HUBRECHT.

stance prevails over the cellular in this genus, and is some-
what loosely arranged. It is not surrounded by nerve-cells,
as these form only an external coating to it, whereas in the
higher-developed genera the cellular substance (in the brain
at least, not always in the lateral trunks) does surround it
on all sides (PI. XXIII, figs. 4, 5, and 6). This cellular por-
tion in Carmella is also of a less compact nature than in
those of more differentiated genera, and is ’everywhere in
direct contact with the epidermoidal tissue. This may be
regarded as the more primitive stage, in which the entire
central nervous system, from the head of the animal down
to the tail, has not yet become separated from the ectoderm
through intervening muscular tissue.

The ScHizoNEMERTiNi and the genera Polia and ValeU”
ciniuj among the Paljeonemertini, represent a further stage
of development, inasmuch as the central nervous apparatus
is here situated exteriorly to the circular and interiorly to
the outer longitudinal muscular layer, and so is everywhere
surrounded by muscular tissue ; whereas in the large ma-
jority of these genera the longitudinal trunks (‘^Nerven-
markstamme,” as the author proposes to designate them)
occupy a strictly lateral, opposite position (PI. XXIII, fig. 8)
the genus Langia, with the curious longitudinal depression
along the back, shows the well-marked tendency, already
mentioned above, of an approxim action of these two parallel
trunks on the dorsal side (PI. XXIII, fig. 9). In none of all
these genera could a commissure uniting the lateral trunks
in the posterior extremity of the animal be detected. They,
simply terminate. This commissure makes its appearance in
the Hoplonemertini, which are, moreover, characterised
by the lateral trunks being situated wholly interiorly to
the muscular body wall. So in this respect they form the
opposite extreme to Carinella,

Generally the position of the nerve-trunks is strictly
lateral (PI. XXIII, fig. 10), also in this suborder ; in certain
genera, however, as Drepanophorus (PI. XXIII, fig. 11),
the longitudinal trunks have approached each other on the
ventral side, and would, indeed, if they were connected by
commissures or eventually coalesced, form an oesophageal
nerve-ring and a ventral cord corresponding to the Bauch-
mark” of Annelids and Arthropods. But such commissures
fail in this genus, as they do in all other Nemerteans, and
even in Drepanophorus the longitudinal trunks bend up-
wards in the extremity of the tail, and are in direct com-
munication by a commissure above the digestive tract. This
large amount of variability in the position of the central

RESEARCHES ON THE NERVOUS SYSTEM OF NEMERTINES. 277

nervous apparatus, with respect to the muscular layers of the
body wall on the one hand, and to the axis of the body on the
other hand, seems, indeed, to justify the author’s conclusion
that the Nemertines represent an ancient and primitive stock
in which an apparatus, otherwise so stable as the central nerve
chain, is subject to interesting variations, which point to
higher stages of differentiation, apparently wide apart,
which, however, are reached in other subdivisions of the
animal kingdom.

B. The side organs. — The curious apparatus, accessory to
the Nemertean brain, which has been designated by so
many diverse names (side organs, cephalic sacs, ciliated
furrow, cephalic grooves, &c.), is here for the first time
studied in the light of comparison. The author, who has
had at his disposal all the different genera of European
Nemerteans, has been able to study this apparatus in the
most various stages, and develops his views as to its probable
phylogenetic development, basing his views in the first
place on the comparative anatomy, and secondly, on certain
facts which recent embryological observations have brought
to light.

The lowest stage of differentiation of this apparatus is
present in Carinella annulata (we leave Ceplialotlirix^ where
it seems to be altogether absent, out of consideration), in
which genus the epiderm shows a transverse, shallow fur-
row, which is interrupted in the median line of the back,
slightly bent, and situated about in a level with the middle
of the brain. . This groove is ciliated. Series of transverse
and horizontal sections showed no traces of any complica-
tion of this arrangem.ent (PI. XXIII, fig. 2,). Where this
furrow is situated, the distance separating the nerve- cells
from the external ciliated surface of the body — a distance
which all along the nerve-trunks is very insignificant indeed
— is still more reduced, and the facility for osmotic inter-
course augmented.

Carinella inexpectata, a new species which the author has
elsewhere described,^ agrees with Carinella annulata in all
general respects. It differs from it in showing the apparatus
just described more complicated by a few steps. Instead of
the simple exterior transverse groove this species carries
short parallel grooves placed perpendicular with respect to
the one larger transverse one and confluent with it. In the
middle of this transverse groove, strictly laterally, there is
a small opening leading into a ciliated duct. This duct pene-
trates the ectodermic tissues, and immediately enters amidst
‘ ‘Notes from the Leyden Museum, vol. i, p. 93.

.278

DR. H. W. HUBRECHT.

the ganglion cells of the brain. It there appears to terminate
blindly (PI. XXIII, fig. 3.) In all other respects the structure
of the brain and nervous system closely agrees with C. annu-
lata,

A further degree of complication is attained in Polia,
where the brain no longer lies immediately under the epi-
derm, but is separated from it by intervening muscular tissue.
Exteriorly the ciliated transverse grooves, with numerous
small transverse furrows perpendicular to it, have very much
the appearance of the grooves in Carinella inexpectata.

Sections show that here the ciliated duct traverses the
muscular tissue as well, enters between the nerve-cells of the
brain, at the same time making a double bend (PL XXIII,
fig. 4), and finally terminates in what appears to be a
more or less swollen enlargement. Another complication
has taken place, in so far as the brain is no longer a simple
enlargement of the lateral trunks, but is split up into lobes,
two dorsal ones and two ventral ones. Moreover, the dorsal
lobes are separated into a larger anterior and a smaller pos-
terior portion, applied to the former as a sort of cap. That
such a separation exists is rarely visible by transparency ; it
can only be clearly made out in horizontal sections. Only
into this third portion of the brain, in this posterior dorsal
lobe, the ciliated canal communicating with the exterior
penetrates. In its turn this posterior cerebral lobe is capped
over postero-medially by a layer of larger cells with distinct
nuclei, which in their appearance strongly resemble the cells
belonging to the coating of the oesophagus, which cells lie
in their immediate vicinity in the same section. The situa-
tion of this layer of cells is indicated by a paler tint in
PI. XXIII, fig. 4. Further down we shall mention the
authoPs views respecting the morphological significance of
this group of cells.

In all the Schizonemertini the external appearance of
the apparatus is considerably different, in so far as there is
no more any transverse furrrow or small opening of the
ciliated canal, but two deep longitudinal slits on each side
of the head, situated between the mouth and the tip of the
snout. These slits penetrate deeply into the muscular
tissue of the head, and in the bottom of them the lobes of
the ganglion stand out more or less freely, only protected
by a very thin layer of tissue. Their internal surface is
covered by numerous and long cilia, and in their postero-
medial portion they are continued into the ciliated canal
which here penetrates into the third (posterior) ganglionic
lobe. The author regards the cephalic slits in this suborder

RESEARCHES ON THE NERVOUS SYSTEM OF NEMERTINES. 279

(to which the majority of hitherto known Nemerteans
belong) as entirely homologous with the simple opening
described for Polia, ValencAnia, ^c. ; this opening having
here been widened out and deepened for purposes in direct
connection with the strongly augmented amount of haemo-
globin contained in the nerve-cells, and with the scarcity
of oxygen in those places where the species of this suborder
are found (deep in the mud, under stones, or in decaying
matters on the sea bottom). The respiratory function of
these slits, which was tested by the author in different
physiological experiments, described in the original memoir,
can be exercised, 1st, at the surface of the two pairs of
anterior lobes, where these are externally visible at the
bottom of the slits, and 2nd, in a more complete way inside
the third (posterior) cerebral lobe (which has here become
somewhat more separated from the anterior than was the
case in Polia), where the ciliated duct with its continual
current of^fresh seawater is immediately surrounded by the
hsemoglobinous nerve-cells (see Pl.XXIlI,fig. 5). In some of
these species the ciliated canal divides into two, one penetrat-
ing among the nerve-cells, the other apparently remaining
more externally situated in relation to the ganglionic lobe.
Both have blind terminations, and never show any trace *of
a special sensory epithelial layer. The accessory mass of
large cells (pale in fig. 5) is somewhat more conspicuous but
occupies the same place as in Polia.

In the Hoplonemertini differentiation has gone on along
another tract. The transverse furrow has not been modified
into a longitudinal slit, but in certain genera {Drepano-
phorus, Amphiporus) it is even provided with short longi-
tudinal grooves perpendicular to it, exactly as it is found
in Polia, whereas in other genera these secondary short
grooves are absent. Internally, another important change
is the entire separation of the posterior cerebral lobe from
the anterior dorsal one, with which it is still so closely
united in Polia, and with which it is here only connected
by one or more nervous commissures. With this separation
a greater independence in its position has been gained it
may either be situated behind, on a level with, or before
the rest of the brain. Here, as in Polia and the Sciiizone-
MERTiNi, there is a grouj) of cells of quite different appear-
ance coalesced with the ganglion cells, which latter, however,
form tlie bulk of the posterior lobe ; but here a ciliated
space or canal remains free in the midst of this accessory
mass of cells. The primary ciliated canal ])cnetratos from
the exterior into the midst of the true ganglion cells ; into

VOL. XX. NEW SER. T

-230

DR. H. W. HUBRECHT.

it the second ciliated canal above mentioned opens, as
indicated in the figure (PI. XXIII, fig. 6).

Summarising the different phases in which this accessory
apparatus to the brain is found in the various forms of
Nemerteans, we have — a. A simple transverse furrow in the
epiderm, on a level with the brain, which lies immediately
under the epiderm. h. A slight complication of the furrow,
and an external opening from which a short canal leads into
the mass of nerve-cells from the brain (which in position and
structure has undergone no change), c. A separation of
the brain into three lobes, into the posterior one of which the
ciliated canal, perforating the muscles and opening exteriorly,
penetrates, whilst a mass of cells strongly resembling the
cells of the wall of the oesophagus coalesces with this pos-
terior lobe. At the same time we have sub. c, the three
following modifications : — a. The external furrow as in 5,
the internal canal simply bent. |3. The external opening
modified into a deep longitudinal slit on both sides; the in-
ternal canal either simply bent or divided into two. y. The
external furrow as in h and ca\ the ciliated canal divides
into two, one branch situated amidst the true ganglion cells,
the other in the accessory mass of cells.

As to the development of the apparatus, it has as yet only
been studied in a few species. From the results arrived at
by Metschnikoffj Leuckart, Pagenstecher, Biitschliy and
Barrois, the author feels justified in concluding that — 1st.
The exterior opening and ciliated canal leading amongst
the ganglion cells of the brain are an invagination from the
ectoderm. 2nd. The mass of larger cells (indicated by lighter
shading in the figs.), in which sometimes a ciliated space
persists, is an outgrowth from the oesophagus, afterwards
separated from it ; by coalescence with the invagination of
the ectoderm the internal space sometimes communicates
with the exterior. 3rd. The bulk of the apparatus is directly
derived from the brain, with which it remains in more or
less intimate connection.

The facts mentioned under the second heading would be a
strong argument in favour of this apparatus having a respi-
ratory function. The gills of Balonoglossus develop in the
same way ; whereas here it is the blood fluid with* which
similar outgrowths of the oesophagus become more intimately
connected, it is in Nemerteans the heemoglobinous nerve-
tissue.

The author is not unwilling to recognise that in the
higher differentiated Hoplonemertini the apparatus may
gradually lose its significance as a special respiratory centre,

RESEARCHES ON THE NERVOUS SYSTEM OF NEMERTINES. 281

especially because here the circulation is more perfected and
the blood-corpuscles are haemoglobinous themselves,, the
nerve-tissue, at the same time, losing its deep-red tinge.
Only he adds that, not being able to detect any sensory epi-
thelium in these organs, he is provisionally inclined to doubt
whether, after having lost their primary respiratory signifi-
cance, they ever become secondarily modified into a sense-
organ in the higher differentiated forms.

c. Histology of the neroe-centres, peripheral nerves^ ^c.
— Respecting the further histological and anatomical results,
we may rapidly note the following : — The ganglion cells in
the brain, in all genera above Garinella and Cephalotrix,
surround the central mass of nerve-fibre on all sides. This
fibrous skeleton” of the central apparatus is figured for the
ScHizoNEMERTiNi ill PI. XXIII, fig. 1 ; the right dorsal
lobe is supposed to have been removed in order to show the
ventral one. The fibrous nerve-substance is indicated by a
paler, the cellular by a darker tint. The general shape and
distribution of this fibrous nucleus of the brain may be
gathered from the figure ; it is also continued into the posterior
brain lobe.

A very thin homogeneous layer separates the fibrous from
the cellular nerve-substance ; the processes of the nerve-
cells (very few multipolar, the majority apparently unipolar)
are seen to perforate this homogeneous layer, and to pene-
trate into the central fibrous substance. The larger nerves
for the eyes and the tip of the snout, those for the proboscis,
and those for the oesophagus, are seen, as well on compression
as in sections, to take their origin in this central fibrous
nerve-skeleton. The two latter sets of nerves are'lfor the
first time correctly described by the author. For Carinella
he succeeded in demonstrating the direct passage of strong
nerves, leaving the brain close to the ventral commissure
into the proboscis, where they continue their course longi-
tudinally between the muscular and the cellular coating.
Identical nerves were clearly made out in certain Schizo-
NEMERTiNi, and for some of the Hoplonemertini von
KenneV s suggestion that MeIntosW s beaded layer” in the
proboscis of this group is of a nervous nature was fully con-
firmed, and at the same time the distribution of delicate and
numerous nerve-twigs going to the papillae of this proboscis
was made out.

Next to this proboscidian nerve another strong nerve leaves
the brain, springing from the posterior portion of the inferior
This pair of nerves is present in all Nemerteans, and

lobes.

is intended for the anterior portion of the alimentary canal..

282

DR. H. W. HUBRECHT.

on the surface of which it is subject to further dichotomic
division. To this nerve the author assigns the name of
N. vaguSf a term already employed by Leydig for other In-
vertebrateSj and recently misapplied by Semper (who was
then unacquainted with the presence in Nemerteans of a
nerve such as the one here mentioned) to the dorsal com-
missure of the cerebral ganglia in Nemertines.

A thin longitudinal nerve, originating from the dorsal
commissure and situated medially and dorsally, is looked
upon by the author as a special nerve for the proboscidian
chamber (PL XXIII, fig. 1).

The anal commissure, situated above the intestine men-
tioned above, presents no further particulars, its high mor-
phological interest depending on the fact of its causing
the whole of the paired symmetrical central nervous system
present throughout the whole length of the animal to be
situated, together with all the three transverse commissures,
above the alimentary canal.

The system of peripheric nerves is not exhaustively treated
of in the present paper ; the author’s investigations on this
point being not yet terminated. Still, certain points which
may eventually prove to be interesting are already noticed.
Whereas in the Hoplonemertini where the lateral nerve-
trunks are situated inside of the mpscular body wall, small
pairs of nerves springing from those trunks placed meta-
merically, and distributing themselves to the muscular coats
of the body wall can easily be detected, no such transverse
stems to the longitudinal trunks could be detected in the
ScHizoNEMERTiNi. Only here the author found a cellular
layer, which he holds to be a nervous layer, encircling the
body in the way of a tunic, and situated in a free space be-
tween the outer longitudinal and the circular muscular layers
(PI. XXIII, fig. 12,). The cells of this layer are multipolar,
connected together as a sort of network, and uninterruptedly
passing into the ganglion cells ensheathing the longitudinal
trunks. Further details on the structure and significance of
this layer are postponed by the author to a future paper on
this subject.

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 283

On the Structure of the Nephridia of the Medicinal
Leech. By A. G. Bourne, Assistant in the Zoological
Laboratory of University College, London. (With
Plates XXIV and XXV.)

The investigation into the structure of the so-called seg-
mental organs, or (better) nephridia ’’ of the medicinal
Leech, of which the following pages give an account, was
undertaken at the suggestion of Professor Lankester and
carried out under his direction and with his supervision and
advice in the zoological laboratory of University College,
London.^

A series of very successfully prepared sections of the
Leech, made by Mr. J. E. Blomfield, of Oxford, at present
Demonstrator in Prof. LankestePs laboratory, had sufficed
to draw our attention to a number of interesting points in
the minute structure of that animal.

Amongst others. Professor Lankester especially pointed
out the remarkable structure of the constituent cells of the
nephridia, each cell being either perforated by a simple duct
or having within its substance an arborescent extension of
the system of ducts, common to the whole gland. Further,
each cell was seen to be enclosed in a mesh of the capillary
system filled with red fluid (the heemal system), which is
spread throughout the tissues of the medicinal Leech.

I undertook the more thorough investigation of these
structures with two objects in view. In the first place, since
the community of origin of the renal organ of Vertebrates,
Worms, and Molluscs, has become a well-established theory,
it cannot but be a matter of the first importance that the
complete anatomy of the various modifications of the t) pical

NEPHRiDiuM ” (as Professoi* Lankester^ has termed the
single excretory tube from the multiplication or branching
of which more complex forms, such as the primitive kid-
ney or archi-NEPHRON, have been derived) should be care-
fully worked out.

In the second place, it is known that in Leeches this
organ presents very wide differences of structure in different
genera, difierences which appear to go hand in hand with
corresponding modifications of the haemal system, and I was
therefore anxious to obtain a complete knowledge of the

^ See bis “ Observations on tlie Microscopic Anatomy of the Leech,” iu
the ‘ Zoolog. Anzeiger,’ 1880, No. 49.

^ “ Notes on Embryology and Classification,” this Journal, 1877.

284

A. G. BOURNE.

arrangements obtaining in Hirudo medicinalis, in order
subsequently to compare with the nephridia of that form —
those of Nephelis, Clepsine, and Pontobdella.

I. — Current Statements as to the Nephridia of the
Medicinal Leech.

Omitting detailed reference to the writings of De Blain-
ville, to Brandt and Batzeburg’s Medizinische Zoologie,”
and to Moquin Tandon's valuable ^^Monographie de la famille
des Hirudinees.” I shall here give brief extracts from Gra-
tiolet, from Leuckart, from Leydig, and from Gegenbaur, in
order to show what is the present state of our knowledge
as to the nephridia of the Leech.

In 1850 Gratiolet published a memoir in the French
Academy’s series on the Circulation of the Medicinal
Leech/’ and in 1862 another paper on this subject Ann.
des Sciences Nat./’ Ser. iv, ^ Zool./ 1862).

Gratiolet says : — The segmental organs, as Dr. Williams
has called them, have been considered as appareils aquifer es
(? water vessels), as tracheal or respiratory organs, or as
mucous glands ; further, Dr. Williams {loc, cit.') has sug-
gested that these segmental organs are the true ovaries in
all the annelids. They are seventeen in number on each
side of the body. Each is constituted by — 1. K tube with
thick glandular walls. 2. A vesicle behind the tube, open-
ing on the ventral surface of the animal by an excretory
orifice. The tube is bent up into a narrow loop, the much
attenuated branches of w^hich end by anastomosing; it is
spread into a sling/ and then bent back into a right
angle, one extremity of the ^ sling ’ being nearly vertical,
the other directed horizontally forwards, and ending by being
rolled on itself in a sort of bud, the first and much the
thickest end being more or less dilated according to its
position.” Gratiolet means by this that the t-wo halves of
the lobe folded on itself, which forms the vertical portion,
may lie close together, or be separated by an interval.

He continues : — The two branches of the loop are
similar in structure, the anterior differing from the posterior
in having a canal at the lower end of the vertical part,
which opens on the upper wall of the well-known oval or
spherical vesicle.

In front of the series of testes the upper parts of the
loops are much dilated ; the lower part bends for^vards, and
end in the above-mentioned bud; to this in the segments
containing the testes is added a small coecal process, which

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 285

leads inwards toward the middle line ; these processes end
blindly on the upper surface of the testes. The two ne-
phridia behind the last testis also possess these caeca, which
are directed inwards and end m the same way.

This connection with the testes is very interesting, as it
reproduces in a very significant manner the relations which
certain analogous organs have with the testes in the Oligo-
chaeta.”

Gratiolet continues: — Various erroneous views have been
entertained with regard to the vesicles ; they have been
thought to be tracheae or spiracles by Schlacht and Bibiena,
and Thomas^ Duges, and Audouin have considered them as
lungs. Duges has thought that the glandular part might
be a sort of heart.

It is not absolutely contrary to probability to consider
them as water-vessels, but how they could subserve aerial
respiration is difficult to explain.

“ Nevertheless, these hypotheses, however rash they may
have been, are nothing to that which has recently been put
forward and defended by Dr. Williams.

According to this author, the glandular parts are the
normal ovaries. The eggs develope and ripen in their
branches. Each complete segment of the body has thus its
female genital organ — its ovary.”

Gratiolet agrees with De Blainville, Brandt, and Moquin
Tandon that these organs are secretory — are, in fact, renal
organs. D’Udekem has also shown this to be the case.

Gratiolet considers that they also serve to keep the skin
moist while the animal is out of the water, and correlates
the greater power the medicinal Leech has of staying out of
water compared with that of the Horse Leech with the larger
size of these organs in the former animals. Leydig has
shown, however, that unicellular glands open all over the
surface of the skin, and these would serve to keep it moist,
just as in land Planarians, the frog, and other terrestrial
animals which possess a moist skin.

I see no reason to suppose that the nephridia of the Leech
have any such mucous function.

Leuckart Die Menschlichen Parasiteii,’ vol. i, 1863) has
given a brief description of the nephridia. The most
important new fact he mentions is that the vesicle has con-
siderable contractile power, which is produced by a delicate
muscle which lies outside the tunica i^opria. The inner
surface of this membrane bears a finely granular parchment
epithelium com])osed of large cells.

Leuckart also states that there arc muscle-fibres in the

286

A. G. BOURNE.

walls of the central duct. This I have not found to be the
case.

Leydig Histologie des Menschen und der Thiere ’) added
to previous knowledge of the general form of the nephridia
of the Leeches (which was largely due to his own researches,
especially that entitled Zur Anatomie von Piscicola,”
^ Zeitsch. wiss. Zoologie,’ vol. i), an important fact of minute
structure in which it is now known to agree with the ne-
phridia of some other types, such as earthworm and larval
Pulmonate Gastropods. He showed, in fact, by a figure to
which there is no further allusion in the text, that the
gland-cells of this organ are perforated by ducts, so that the
cells have the form of hollow cylinders ; but he did not
observe the more complicated condition of arborescent ducts
discovered by Lankester, which I shall describe hereafter.
The perforated cells figured by Leydig in his ^ Histologie ’
were from the nephridium of Haemopis. Claparede first
showed that such was also the relation of duct and cell in the
nephridium of the common earthworm (“ Histologie des
Regenwurms,” ^ Zeitsch. wiss. Zool.,’ vol. 1869).

Gegenbaur was the first to point out the complicated
labyrinthine character of the ducts of the nephridium of the
Leech when comparing that organ with the nephridium of
the earthworm (^Zeitsch. wiss. Zooh, vol. iv), which he was
the first to describe accurately. The condition of knowledge
at the time when Gegenbaur’s observations w’ere pub-
lished (five-and- twenty years ago) did not require a minute
histological account of these organs, and, accordingly, w^e do
not find histological details in his memoir. Excepting for
the figure of cells from the nephridium of Heemopis, pub-
lished by Leydig, there appears to have been no attempt to
inquire further into the structure of this organ during later
years; and we have the following account in Gegenbaur’s
^ Elements of Comparative Anatomy ’(English edition, 1879),
which may be taken as representing the actual state of
knowledge wdth regard to them. It will be observed as an
important point distinctive of the nephridia of the genus
Hirudo that no internal opening has been observed leading
from the body-cavity into the duct of the nephridium, and to
the possible existence or non-existence of such an opening
my observations have, inter alia, been carefully directed.

Gegenbaur says (p. 176) : — ‘‘ So far as the structure of the
excretory organs is concerned, few fresh characters appear in
the Annulata (Hirudinea, Oligochseta, Chsetopoda). The
organs correspond to the metamerism of the body, for they
are regularly distributed on either side of almost every one

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 287

of its segments. They have, therefore, though with but little
reason, been called " segmental organs,’ a name which is
just as suitable for many other organs. Each of them con-
sists of a closely-coiled or loop-like canal (Schleifen canale) ,
which has an internal opening, often peculiar in form and
always ciliated, and which opens at the other end on to the
surface of the body. This canal is sometimes similar in
character throughout its whole length, or but slightly dif-
ferentiated ; frequently several segments may be made out
in it, which generally correspond with those already de-
scribed in the Platyhelminthes and Rotatoria. The inner-
most portion, on which the opening into the coelom is placed,
is ordinarily the longest ; it is distinguished by its funnel-
like or rosette-shaped orifice. In the next portion the walls
may be seen to be glandular in structure. The last portion,
which is frequently widened, is provided with a layer of
muscle ; it almost always opens on to the ventral surface.
These organs are no more purely excretory in function than
they are in other Vermes, for we not unfrequently find them
entrusted with other functions.

'^Tn the Hirudinea these organs are preceded, in the em-
bryonic stage, by three pairs of looped canals, which are not
connected with those formed at a later stage ; they are found
in the posterior half of the ventral surface. In structure
they are similar to, but simpler than, the permanent canals,
and disappear after these are developed. This most impor-
tant fact shows that the looped canals of the Annulata
cannot be regarded as the direct hornologues of the excretory
organs of the lower Vermes; at the same time arises the
question. Are the looped canals of those Annulata, which
show no signs of primitive organs of this kind, comparable
to the permanent looped canals of the Hirudinea, or only to
the primitive ones ?

In their more special characters there is great variety
among the Hirudinea, the canals in one division having no
internal opening. Instead of this they begin with a closed
portion in the form of a loop, which consists of numerous
canals united with one another into a labyrinth {Hirud'o).
From these looped organs a single canal is given off which
opens by a vesicular enlargement on the surface of the body.
In others {Clepsme^ Nephelis) the labyrinthine portion is
present, but it has an internal opening, which projects into
the lateral blood-sinuses of the body.”

I may be allowed further to cite a statement of Mr.
Balfour’s relative to the nephridium of the Leech, which will
serve to illustrate the view generally entertained as to its

288

A. G. BOURNE.

structure. After describing (this Journal^ 1879, p. 432)
the segmental organs (nephridia) of the interesting Quatho-
pod, Peripatus, as consisting of 1, a dilated vesicle ; 2,

a coiled glandular tube ; 3, a short terminal portion, he pro-
ceeds to say that though formed on a type of their own, the
segmental organs of Peripatus more closely resembles those
of the Leech than of any other form. The facts of the
structure which I have to record as the result of my studies
do not allow me to accept the view that there is any special
similarity between the simple nephridium of Peripatus and
the complicated one of Hirudo. The high differentiation of
the nephridium of Hirudo appears in fact to have been
overlooked by zoologists generally.

II. — Methods of Investigation^

1. For the purpose of studying the nephridium entire and
in the fresh state. Leeches were chloroformed and pinned
in a gutta-percha trough, and opened by a dorsal median
incision, the trough being filled with normal salt-solution.
In this way any one of the series of seventeen pairs of
nephridia could be dissected out whole and removed to a
glass slide for observation under the microscope.

The transparency of the organ renders this method of
study of great value. ^

2. In order to obtain permanent and stained preparations,
the nephridia, after excision, Avere placed in ^ per cent,
osmic acid for tAventy minutes, Avashed in normal salt solu-
tion or dilute alcohol and then stained for half-an-hour
Avith Ranvier’s picrocarmin. Finally, they Avere mounted
in glycerine.

Other Avhole preparations of great value in the difficult
task of folloAving out the course of the ducts and ductules,
Avere obtained by placing Avhole nephridia in absolute
alcohol, staining Avith hsematoxylin, clarifying Avith clove-
oil and mounting in Canada balsam.

3. Teazed preparations for the purpose of studying the
minute characters of the tissues composing the organ were
used. The nephridia Avere teazed either fresh, in salt solu-
tion or in osmic acid, or after short maceration (tAventy
minutes) in nitric acid (20 per cent, in Avater), or in Weiss-
mann^s potash solution (40 parts caustic j)otash to 100 parts
Avater),or after prolonged maceration (six months) in 2 per
cent, aqueous solution of potassium bichromate. The latter
method yielded valuable results, as also did maceration in
-ptj per cent, chromic acid.

4. The method of silver-staining Avas also applied to fresh

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 289

iiephridia in order to ascertain the structure of the walls
of the blood-vessels and of the ducts of the gland. An
interesting structure in the substance of the gland-cells
themselves was revealed by this method.

5. Injection of the vascular system with soluble Berlin-
blue was practised with some success so far as the injection of
the finer vessels of the integument, &c., was concerned, but I
have not succeeded in injecting the capillaries of the nephri-
dium itself. I have been at present equally unsuccessful in
injecting the ducts of the nephridium. The mode of injection
made use of was that used by Moseley in his injections of in-
sects, which consists in using a small glass tube drawn out
to a capillary termination serving as the nozzle, a caoutchouc
tube being attached to the other end of the glass tube. When
the little apparatus is filled with injecting fluid it suffices to
compress the caoutchouc tube in order to obtain sufficient
pressure and sufficient flow of liquid for the injection of very
minute organs. I also made with Professor Lankester some
injections of indigo carmine into the body substance of an
uninjured Leech by means of a subcutaneous syringe. The
Leeches lived well after receiving a cubic centimetre of the
indigo-carmine solution, but we have not as yet obtained
any definite results as to the excretion of indigo by the
nephridia or other organs.

6. The most important method of which I availed myself
in conjunction with the ^^study of whole nephridia, is that
of section-cutting ; sections of nephridia wxre obtained either
by cutting whole Leeches or by cutting only a single nephri-
dium and its surrounding tissues isolated for the purpose.

In either case the Leech was prepared by killing with
chloroform, and w'as moderately stretched by means of a pin
at each end fixing it in a gutta-percha trough. Chromic
acid of -^th per cent, aqueous solution was then allowed to
act on the Leech for eighteen to twenty-four hours. The
specimen was next removed to half a pint of alcohol (60 per
cent.), and removed to a second half pint on the following
day. After three or four days in the second quantum of
alcohol sections should be cut and stained. The staining
is more successful in recently hardened specimens than in
those which have been preserved many months. Chromic
acid was preferred for the hardening process to Kleinen-
berg^s picrin solution or to pure alcohol.

The sections w'ere cut by the aid of a simple screw micro-
tome, and from forty to sixty could be obtained and pre-
served in their order of sequence in the region of a single
pair of nephridia.

290

A. G. BOURNE.

The sections were stained in some cases with haema-
toxylin, but usually with E-anvier’s picrocarmine, which
has a remarkable effect upon the naturally red-coloured
haemal fluid. This very frequently is stained yellow by the
picrin alone, whilst the surrounding tissues have a more or
less complete carmine staining. Some very beautiful pre-
parations were obtained by Mr. J. E. Blomfield by staining
with two pigments in succession, viz. picrocarmine and
anilin-blue. Oil of cloves and Canada balsam were used
for clarifying and mounting.

Camera lucida drawings, under a moderate power of the
microscope (HartnacVs obj. 4), were now made of a complete
series, giving all the sections in order in which any part of
a selected nephridium was involved. Such a series of
drawings compared with the whole nephridium, and with
horizontal and transverse sections, has enabled me to
thoroughly explore this somewhat tortuous and complicated
body.

III. — Form and Eegions of the Nephridium.

The nephridium of tlie medicinal Leech is not the simple
loop-like body with labyrinthine duct opening into a vesicle
which it has been held to be, and^ such as it was drawn
by Gratiolet. In the testicular region of the body the
nephridium has the form which I have diagramrnatically
represented in Plate XXIV, fig. I. It may be divided
first of all into vesicle and gland, which are connected by
the vesicle-duet.

The gland is in the form of a thick made horse-shoe,
the two limbs of which are elongated and produced till
they meet. The front of the horse-shoe is dorsal and
superior in the living Leech (to the right in the figure),
whilst the produced limbs descend and are respectively
anterior and posterior in position ; in the ventral region
they are twisted out of the straight line in a forward direc-
tion (see figure). From the forwardly placed thickened
point of union of the two limbs of the horse-shoe reaches a
delicate process which I call the testis lobe. It is
more or less rudimentary in those nephridia which belong
to the anterior region of the body where there are no testes,
the further forward the nephridium the smaller is the testis
lohe.

The vesicle relatively to the form and position of the
horse-shoe with its prolonged, and bent limbs may be said to
balance the testis lobe. It lies posteriorly and ventrally rela-

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 291

tively to the main curve of the horse-shoe in the living
Leech.

The vesicle duct, instead of passing into that limb of the
gland which is the nearer to it, crosses the posterior limb
externally, and joins the anterior limb. A portion of the
gland, undescribed by previous observers, is found in the
form of a delicate cord of gland-substance which extends as
a free piece (a sort of third limb) from the centre of the
concavity of the horse-shoe to about the middle of the
posterior limb to which it is united. This is marked in the
figure by the words Recurrent Duot.^’

It would not be difficult to give names to the different
parts of the elongated horse-shoe shaped mass just described
which might serve in further description did the minute
structure of the gland in any way correspond with the super-
ficial appearance. As a matter of fact it does not, and I
have found it necessary to separate the following regions or
lobes which have little correspondence with the general
external form.

The whole curve of the horse-shoe, from the point where
the vesicle duct enters the anterior limb to the point where
the recurrent duct joins the posterior limb, may be known as
the MAIN LOBE.

From the point where the recurrent duct joins the posterior
limb to the point where the two prolonged and forwardly-
bent limbs unite, including a portion of the anterior limb,
is the APICAL LOBE.

The outstanding process of the nephridium which comes
into close relation with the testis, and which is given off
from the anterior limb close to the enlarged commencement
of the apical lobe, is the testis lobe.

The little piece of the anterior limb extending from the
base of the testis lobe to the commencement of the main
lobe I shall leave for the present without a name, whilst the
delicate piece depending from the main lobe and in which
the recurrent duct commences may be called the recurrent

piece ; ” it can hardly be called a lobe.

IV. — The Ducts (Vesicle Duct, Central Duct, and
Recurrent Duct).

The divisions of the glandular portion of the nephridium,
which we have just recognised, are necessitated by the dis-
position of the duct and by the characters of the cells sur-
rounding it.

The duct in its various divisions appears as a cylindrical
passage filled with a colourless liquid, and often contains

292

A, G. BOURNE.

needle-like crystals (fig. 11), which are also found in great
abundance in the vesicle.

The VESICLE DUCT is a tube 4-^ inch in diameter, with
thin walls, formed by numerous cells, several cells surround-
ing the lumen of the tube. It plunges into the substance
of the anterior limb of the gland and assumes a different
character. Henceforth I call it the central duct.

The central duct is circular in transverse section, varies from
aio lo ToVo ii^ch in diameter, and is lined by a very strongly-
marked structureless cuticle, which has a radiating fibrous
periphery, the fibrous irregular radii passing into the sub-
stance of the nephridial cells which surround the duct, and
also passing hetween them, where it becomes the proper
cuticle of the cells (see figs. 7 and 18, c. d.). The central
duct takes a course in the centre of the mass of cells w^hich
constitute the lobes of the gland, and so far from being laby-
rinthine, is entirely simple and unbranched, excepting at one
point where it gives off a lateral offshoot similar to itself.

Previous observers appear not to have distinguished be-
tween this central duct, which is the direct continuation of
the vesicle duct, and the immense plexus of ductules, to
be described below, which excavate the cells surrounding the
central duct, but which most assuredly are not ramifications
of the central duct, and which I hav6 not been able definitely
to trace into communication with it at any point, though I
consider it possible that such a communication may exist at
some one or possibly two points which have eluded my con-
stant and very careful search.

The central duct at its origin from the vesicle duct turns
upwards (to the right in the figure) and traverses the arch
of the main lobe, where it is surrounded by the peculiar
nephridial gland-cells, which are in parts two or three rows
deep. It descends in the posterior limb of the main lobe
and passes on into the apical lobe.

At that point of the posterior limb where the apical and
main lobes are joined there is a constriction of the whole
gland, and a sudden and marked change in the character of
the nephridial cells. This sudden change has no effect at
all upon the character of the central duct, which pursues its
course along the axis of the apical lobe, arriving at last at
the recurved apex of that lobe, where it rests upon the root
of the testis lobe (see Plate XXIV, fig. 1). At the same
time, although the central duct passes across the junction of
main and apical lobes without any change, it gives off a
branch — similar in character to itself — as soon as it has
entered the ajdcal lobe. This branch 1 term the recurrent

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 293

DUCT. This recurrent duct takes an axial course in that
delicate cord of gland-cells which depends from between the
two limbs of the arch of the main lobe, and which, as an inspec-
tion of the diagram (fig. 1) will show, is nothing more than
a free or detached continuation of the apical lobe beyond the
point where it meets the main lobe. The recurrent duct thus
reaches the middle point of the concavity of the main lobe,
when it turns downwards again, running among the cells of
the anterior limb of the main lobe parallel with the first part
of the central duct. It crosses without any junction the
origin of the central duct from the vesicle duct, and runs up
the somewhat narrow portion of the gland which connects
the main lobe with the base of the testis lobe and apex of
the apical lobe.

Here (and I must beg the reader to follow my description
with the aid of the figure) the recurrent duct branches, just
in the same way as the central duct branched in giving origin
to it. One branch runs up into the testis lobe for a short
distance, and there appears to terminate — how I have not
been able to determine — but possibly this is a point at which
the duct and the ductules of the gland are in communi-
cation.

The other branch runs into the apical lobe and joins the
central duct, as shown in the diagram (fig. 1), and in fig. 2,
which is a careful drawing from a fresh preparation of this
part. Thus we find the recurrent duct, which originated as
a branch of the central duct at the other end of the apical
lobe, coming again into full continuity with the central duct
after a long course.

The interpretation of the duct of the Leech’s nephridium,
with its central and recurrent portions, is exceedingly diffi-
cult. It is totally unlike the duct or central lumen of the
nephridium of the earthworm. Possibly the recurved apex
of the apical lobe represents the funnel-like extremity of the
nephridium of Nephelis and Oligochseta. But in the latter
the funnel-like aperture leads into a passage which is not
intercellular as is that of Hirudo, but ^V^^m-cellular, as are the
ductules of the latter. It is at this apical region — where the
recurrent duct rejoins the central duct — that I have looked
most carefully for rudiments of the funnel-like aperture, or
for a small opening to the body cavity ; but I am able to
state definitely that none such exists. The relation, or rather
the difficulty of explaining the relation, between the nephri-
dium of the Leech and, that of the Oligocbrnta will become
more apparent when we have examined the ductules.

294

A. G. BOURNE.

V. — The Ductules and Nephridial Cells.

When I commenced the study of the Leech’s nephridium
I regarded the numerous minute passages which give the
labyrinthine character to that organ noted by Gegenbaur as
necessarily only branches and ramifications of the system of
the central duct. Nevertheless, in none of the many hundred
sections of the nephridium which I have examined have I
been able to find a single instance of a ductule opening into
the central duct or into its recurrent branch.

The ductules of the Leech's nephridium are passages lined
by a firm resistant cuticle, which excavate the secreting cells
of that organ in such a way that every individual cell is
completely bored through by such a passage (the grey net-
work in fig. 1). Further, the passages differ very much in
calibre in the different lobes of the nephridium, and not only
that, but differ further from one another in being in some
cells simple traversing passages (a. l. in fig. 13), whilst in
other cells they are hranched within the cell (figs. 5, 6, T).
Further still, the branches may be all passages lead-

ing into corresponding passages in neighbouring cells (fig. 3),
or most of the branches may be ccecal, in which case they are
often exceedingly minute (fig. 5). ^

Nothing equalling in complexity the system of intra-
cellular passages or ductules of the nephridial cells of Hirudo
has hitherto been described by previous writers on that
animal, nor has anything quite parallel been described in
any other cell-structure, so that the facts which I have
ascertained with regard to these ductules in the Leech’s
nephridium have a certain histological interest of a general
character.

It was, I believe, first pointed out by Claparede, who for
the first time made sections of the nephridium of the earth-
worm, that the tube of which that organ consists is built up
by a single series of cells (throughout the greater part of its
length), and that these cells are simply bored thrpugh by
the duct or lumen of the tube. Each cell accordingly has
the form of a drainpipe. A similar structure has now been
recognised in the embryonic nephridia of larval Pulmo-
Gasteropods (Rabl), and is clearly enough characteristic of
nephridia of a certain grade of elaboration.^ At the same

^ It is suggested by Professor Lankester tliat the structure in question
probably points to the phyletic origin of the nephridia from simple uni-
cellular glands, rather than from pockets or invaginated pouches of the
eiiidermis consisting of many cells surrounding a cavity. A nephridial tube,
consisting of a lineal series of perforated cells, is simply a unicellular

STRUCTURE OF NEPHRIDlA OF THE MEDICINAL LEECH. 295

time it is not by any means always exhibited by nephridia,
e.g. those of adult Mollusca (organs of Bojanus) and those
of Vertebrata. I am not aware that a branched duct has
been described as excavating the cells of any nephridium, or,
indeed, of any gland previously to this. The cells from the
nephridium of Hsemopis, figured by Leydig (Joe. cit.^y are
simple cylindrical cells with unbranched passages passing
through them, and probably come from the walls of the re-
current duct. In the salivary glands of some Insects and
Crustaceans the nearest approach to the kind of relation now
established between cell and duct in the nephridia is to be
found,^ but I believe that hitherto only an excavation of the
gland-cell by the ductules of origin has been observed in
these structures, and not an actual thorough perforation of
the gland-cell.

I have been able tx) satisfy myself,^by repeated observa-
tions, that the whole system of ductules in the nephridium
of the Leech is a continuous network of passages, but I
have not been able to determine any aperture in that net-
work by which it communicates either with spaces in the
general substance of the body or with the central duct of
the nephridium, or its branch the recurrent duct. It seems
in the highest degree probable that these ductules do com-
municate with the central duct and so with the vesicle and
exterior, yet I have watched them in various conditions of
distension, and by pressure have caused certain small cor-
puscles (fig. 12) which float in the colourless liquid, which
more or less distends them, to move along from one ductule
to another, without gaining any indication of a communica-
tion with the adjacent central duct.

The ductules do not exist in the cells which line the
vesicle or its duct. These cells are no doubt homologous
with the nephridial cells, yet it is not until the horse-shoe
shaped glandular mass is reached that the ductules make
their appearance. It is difficult to separate any description
of the ductules from that of the nephridial cells themselves,
and I shall therefore describe together the appearances pre-
sented by these structures in the different regions of the gland.

Cells and ductules of the main lobe (figs. 5, 7, 10, lo). —
As in all parts of the gland, the cells are of large size, vary-

gland which has undergone a repeated cell-division transverse to the axis
of its duct.

* And, it may be noted, that the more archaic exam))les of salivary
glands consist of bundles of unicellular glands, each with its own ductule,
which is so far in favour of the hypothesis that the nephridial cells were
originally unicellular glands.

VOL. XX. NEW SER.

U

296

A. G. BOURNE.

ing from -—-o of an inch to of an inch in diameter. They
are sub-spherical, tending to the form of polyhedrons, owing
to mutual adpressure. The protoplasm is fairly transparent
in the living state. By the use of reagents it is seen to be
very finely granular, the granular matter being more abun-
dant peripherally, so as to form a cortical layer, which is
strongly marked after maceration in potassium bichromate.
In cells which have undergone prolonged maceration in this
reagent. Professor Lankester observed a rod-like structure
or striation of the cortical layer, the rods being set at right
angles to the lumen of the ductules.

By the use of silver nitrate I have succeeded in demon-
strating a similar structure in the cortical substance of the
nephridial cells from the apical lobe, that is to say, in cells
wuth a single large traversing ductule (see PL XXV,
fig. 9).

A delicate membranous cuticle exists on the surface of
each cell, and is continuous with the cuticle of the ductules
and with the abundant cuticular deposit which forms the
lining of the central duct (figs. 7 and 13, c. d.). The
cuticle of the ductules is well seen in fig. 10. After pro-
longed maceration it is possible to obtain the ductules in an
isolated condition owdng to the resistance of their cuticle to
the disintegrating process which 'affects the surrounding
protoplasm.

Each cell possesses a large well-marked spherical nucleus
varying in diameter from^Vir inch to -oVoo- inch. The nucleus
has a strongly-marked capsule, and its contents have, after
maceration in w^eak chromic acid, the appearance shown in
fig. 10 a, viz. a clear substance in which a nucleolar net-*
work and nucleolus can be distinguished.

The protoplasm of the nephridial cells of all regions ap-
pears to be distensible and contractile. Whether it is
actively contractile is uncertain, but there is no doubt that
the volume of the ductules varies periodically. There is no
muscular coat to the nephridium itself, as there is to the
vesicle, and the change of volume is probably due to the
physiologically changed condition of the secreting cells.
Gegenbaur found the nephridia of Tubifex to be actively
contractile.

The cells are set around the central duct of the main lobe
radially, as many as eight being thus shown in some sec-
tions taken at right angles to the duct. Superficially to
those which thus abut upon the central duct, other cells
similar to the deeper ones may be seen (fig. 13), forming
portions of a second or even of a third row.

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 297

In the anterior limb of the main lobe the recurrent duct
runs parallel with the central duct, and is in this part of its
course surrounded by cells similar to those which surround
the central duct adjacent to it.

All the cells of the main lohe have that form of ductule
which I distinguished above as arborescent with some of the
branches continuous (^. e. leading into branches of the duc-
tule of a neighbouring cell) and some of them ccecal. The
best general notion of the arrangement of the ductules and
their continuity in the neighbouring cells is given in the
upper part of fig. 13. In fig. 10, from a different portion of
the main lobe, the arborescent ductules are seen to be
limited to one side of the cell, and a tendency to form larger
ductules is exhibited. The finest arborescence is that shown
in the cell drawn in fig. 5. This is from a particular por-
tion of the main lobe where all the cells have the same
character. Five continuous branches of the ductule are
seen, and a dichotomous series of finer branches which
completely honeycomb the cell-protoplasm. The ultimate
branches of this system are 4-oio-o inch in diameter. They are
so disposed as to open into or (to use another metaphor) to
take their origin in the cortical substance of the cell. The
rod-like striated structure of this layer has already been
mentioned as a general character of the nephridial cells. It
appears that t he ultimate ramifications of the ductules have
a definite relation to the bacillary structure of the cortical
substance — although the bacillary structure is observed
equally in cells which are provided with a simple in place
of a finely arborescent ductule.

The cells of the apical lobe differ from those of the main
lobe in possessing very much larger ductules, varying from
-o'oo- inch in diameter to as much as 3-^ inch, -whilst the cells
themselves vary from tto- ii^ch in diameter (figs.

3 and 6, and lower part of fig. 13, A. L.). The medium-
sized and the largest ductules of this larger sort have a defi-
nite position in the gland, the medium-sized gradually pass-
ing into the largest as we proceed along the apical lobe fro?n
the apex towards the posterior limb of the main lobe, where
that middle piece with the recurrent duct is given off (see
fig. 1). Here the ductules become very wide, and are un-
branched, so that the cells are mere hollow cylinders. A
good view of the relation of wide-branched ductules to the
ceils is given in the drawing (fig. 3), which represents cells
of the apical lobe in a state of distension. The ductules of
the apical lobe are also well seen in fig. 2, a. l.

The ductules of this wider and larger kind arc lined with

298

A. G. bourne.

a very distinct cuticle, which in sections of hardened prepara-
tions is often seen in a state of desquamation (fig. 6).

In these large ductules too, in fresh (and therefore still
living) excised nephridia, I have studied the contained liquid.
It is perfectly colourless and transparent, but frequently
contains structureless globules (fig. 12) which, by pressure,
can be made to move along the series of ductules.

The cells of the testis lohe (fig. 2, t. l.) agree altogether
in structure with those of the apical lobe, having medium-
sized ductules which are branched to a small extent, the
branches being continuous, not csecal. In each cell the
ductule gives off some two, three, or four branches of the
same calibre, and each of these is continuous with a similar
branch of the ductule in a continuous cell.

Communications of the ductules of different lobes with
one another. — The diagram (fig. 1) will serve most satis-
factorily to explain what I have ascertained on this point.
The ductules of the testis lobe do not directly communicate
with those of the adjacent apical lobe, but run on through
the short bit of gland joining testis lobe and main lobe, and
then gradually pass into the finer kind of ductule which is
found throughout the main lobe. There is no junction at
the apex of the main lobe (near th^ word inferior vein”
in the diagram) between the ductules of the main lobe and
the ductules of the apical lobe, which is all the more re-
markable since the central duct is here continued from the
one lobe to the other. On the other hand, at the point
where the recurrent piece ” joins the concavity of the
main lobe, as seen in the figure, there is an extensive com-
munication of the large ductules of the recurrent piece ”
with the ductules of the main lobe.

VI. — Blood-vessels and Tunic of the Gland.

The mass of cells which forms the glandular portion of
tlie nephridium, with its various lobes, is held together by an
investment of fibrous tissue, and is further beset by a large
number of blood-vessels which, running longitudinally on
the superfices of the gland, give off branches which pene-
trate the mass of cells at numerous points, and form within
the cell-mass one of the most complete inter- cellular blood-
plexuses known to exist, and resembling that of the Mam-
malian liver.

The investing fibrous tissue of the gland I shall not
describe in detail. It is dealt with by Professor Lankester
in an article On the Connective Tissues of the Leech,”
published in the present number of this journal. It will be

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 299

sufficient to point out that the investing brown-coloured net-
Avork of fibres consists of what Professor Lankester has
termed vaso-fibrous tissue/’ which is found throughout the
body of the Leech, more or less pigmented, and forms the
pigment fibrils which interpenetrate the superficial epi-
thelium of the body-surface.

In the transverse section (fig. 13) this tissue is seen as
dark granular matter {f) surrounding the lobes of the gland,
and passing in and out amongst the muscular fibres {g) and
the unicellular glands (e), all of which are more or less
closely invested by it.

The peculiarly close relationship of this tissue to the
capillary blood-vessels themselves explains how it is that no
distinct connective tissue penetrates the cell-mass of the
nephridium hut only fine blood-vessels , which form, in fact,
an intercellular connective tissue, which is at the same time
a capillary system.

The large blood-vessels found in connection with the
nephridium have been described with great care and ac-
curacy by Gratiolet. They are as shown in fig. 1. They
consist of vessels with delicate membranous walls which are
given off by the great lateral muscular vessel passing near
the nephridium. A connection is further established between
the capillary plexus of the nephridium and the dorsal longi-
tudinal vessel (which runs along the dorsal surface of the
alimentary canal) by means of the superior vein.” The
arrangement and distribution of the vessels will be best
understood from an examination of the diagram fig. 1. The
smaller vessels which are arranged so as to form a plexus or
mesh work with a single nephridial cell in each hole of the
meshwork average in size — o^o-o inch, but are capable

of being greatly distended with blood, and do not always or
in all parts of the nephridium appear to be fully developed.

This, however, is not due, I believe, to any structural
irregularity, but to a functional state. When the nephri-
dial cells or their ducts are swollen, the intercellular vascular
network is compressed, and many of its channels are tem-
porarily obliterated. But, on the other hand, it may happen
that we find all the blood- capillaries well injected with
their natural red-coloured fluid, as shown in figs. 6 and 7.
These are drawn from such natural injections ” which ex-
hibit themselves in the most beautiful way when sections
are cut from specimens of the Leech taken in full nutrition
(that is, within a week or so of a good meal), and hardened
in dilute chromic acid.

The wall of the finer blood-vessels running between the

300

A. G. BOURNE.

nephridial cells is perfectly distinct. These are not mere
intercellular passages but have their distinct membranous
walls like those of the finer blood-vessels m other parts
of the Leech. In these thin- walled vessels, both in the
nephridium and elsewhere, I have not succeeded in tracing
any cell-structure by the use of various histological
methods. It appears propable from what Professor Lan-
kester has ascertained with regard to the relations of the
vaso-fibrous tissue and the blood-vessels, that the nuclei
are discharged at an early period of its development from
the wall of the blood-vessel into the lumen of the vessel, and
accordingly we find only the metamorphosed cell-substance
left to form the wall which is accordingly structureless.

VII. — The Vesicle and the Secretion of the Gland.

The vesicle is remarkable for its great dilatability, and
for possessing a muscular coat which the glandular portion
of the nephridium does not possess. On this latter point I
am at variance with Professor Leuckart, who has ascribed
a muscular coat to the glandular portion. The vesicle may
be regarded as an expansion of the central duct to which it
is joined by the vesicle duct,’’ but it differs greatly from
any part of the glandular portion vof the nephridium, not
only in the fact that it has a muscular tunic, but also in the
fact that the cells lining it have not a cuticle but a ciliated
surface. Cilia are not found in any other part of the nephri-
dium, which is a remarkable fact, when we compare the
structure of other nephridia such as those of the earthworm
and the molluscan organ of Bojanus. The cilia on the epi-
thelium of the vesicle are exceedingly short, the cells them-
selves being small and short, and thrown into ridges when
the vesicle is in a state of contraction. The vesicle is seen
in section in the drawing (fig. 14), which represents accurately
an actual preparation.

The vesicle does not open directly to the exterior, but is
placed in communication with the body-surface by a short
duct, which is lined near its external opening by an involu-
tion of the epidermis of the integument.

The wall of the vesicle is supplied with a very regular
capillary plexus of the haemal system (fig. 1), and it seems
probable that secretion may be carried on from its walls as
well as from the more distinctly glandular portion of the
nephridium.

The vesicle contains a liquid' in which bunches of needle-
like crystals are found, which are drawn in fig. 11, as seen
when magnified 3500 diameters by means of Hartnack’s

STRUCTURE OF NEPHRIDIA OF THE MEDICINAL LEECH. 301

objective 10 immersion. These crystals are not stained
by iodine ; they are soluble in nitric acid. I have not yet
completed their chemical examination. These crystals are
occasionally to be found in the central duct, but I have never
found them in the ductules. On the other hand, in the
ductules are found minute, structureless corpuscles (fig. 12),
which on one occasion I saw exuded from the substance
of a nephridial cell into the ductule. I have never found
these corpuscles in the central duct nor in the vesicle.

VIII. — Opening of the Systems of Ductules and Ducts
INTO ONE another AND INTO THE BoDY-CaVITY.

As I have before mentioned, I have failed to establish the
existence of a communication between the central duct or its
oflf-sets and the system of ductules ; but I am strongly inclined
to regard the recurrent duct as the seat of this commu-
nication. It seems to me not improbable, from the large
size of the ductules of this portion, and from the way in
which it enters the apical lobe on the one hand and the con-
cavity of the main lobe on the other, that we shall even-
tually find a connection through it of the two systems.

I hope during the present summer, by the study of other
genera of Leeches (Aulostomum, Nephelis, Clepsine, Pon-
tobdella), to gain some light upon this question.

The absence of cilia from the ducts and ductules of the
Leeches nephridium, together with the absence of any large
space in the body — comparable to a body-cavity — seems to
be correlated with the absence of an open internal termina-
tion to the duct and ductule systems. I am inclined to
regard what I call the apical lobe as probably corresponding
with the position in which an internal opening is found in
the typical form of nephridium, such as occurs in Nephelis,
Clepsine, and the Oligochaeta.

This point also will be, I hope, decided by the investiga-
tions which I am now commencing on those genera.

The very wide difference between Hirudo and Clepsine
(so far as descriptions of the latter genus enable me to judge)
in the whole arrangement of duct and nephridial cells leads
me to believe that the comparison of the series of Hirudinean
genera will be a very interesting one.

Summary.

1. The Leech’s nephridium consists of gland and vesicle
connected by the vesicle duct.

2. The gland is a curved, horse-shoe shaped loop, present-

302

A. G. BOURNE.

ing the following lohes 1, main lobe \ 2^ apical lobe \ S,
testis lobe ; 4, recurrent lobe.

3. The cells which constitute the gland are all penetrated
by ductules, which pass through them, sometimes forming
several passages ; the ductules differ in the different lobes of
the gland.

4. The axis of all the lobes is occupied by a large duct
separated from the nephridial cells by a thick cuticle ; it
opens into the vesicle duct, but is not shown to have any
connection with the ductules.

5. Vessels of the haemal system (with red fluid) form a
complete plexus in the gland, each cell being surrounded by
a loop of the plexus.

6. There is no internal opening of the system of ducts or
ductules.

7. There are no cilia in any part of the gland excepting
the vesicle, which is lined by a ciliated epithelium.

8. The wall of the vesicle is muscular; that of the gland
is not, but consists of a fibrous, pigmented investment.

CAPILLARIES IN INTEGUMENT OF MEDICINAL LEECH. 308

On Intra-Epithelial Capillaries in the Integument
of the Medicinal Leech. By E. Ray Lankester,
M.A., F.R.S., Professor of Zoology in University Col-
lege, London. With Plate XXVI.

I am not acquainted with any minute investigation of the
structure of the epidermis of the common Leech. Professor
Leydig has published most valuable accounts of the sense-
organs of the Leeches, and in a drawing of a transverse sec-
tion of the medicinal Leech Bau des Thierischen Korpers/
Taf. i, fig. 6), has shown the epidermis in position with the
remarkably enlarged cells which form unicellular glands
plunging down amongst the muscular bundles far below the
horizon of the other epidermal elements. The same draw-
ing also shows a few scattered vessels belonging to the red
vascular system. These are seen in Professor Leydig’s draw-
ing to come near the integument, but they do not penetrate
it. The drawing is on a relatively small scale, so that such
details as the minute structure of the epidermic cells are not
presented.

In sections prepared in my laboratory last year by Mr.
Blomfield and by Mr. Bourne, and in macerated specimens,
I have studied the structure of the Leech’s epidermis more
closely, and have obtained some interesting results. I was
led to make these observations in carrying out a general plan
of study of the epidermic tissue in the various groups of the
animal kingdom. The integument of the Earthworm two
years ago yielded me some very interesting results entirely
contradictory to Claparede’s statements on that subject — more
delicate and varied forms of epithelial cells than are yielded
by the Earthworm’s epiderm when macerated are not pre-
sented by any other animal — whilst the whole structure of
the clitellum and its periodical variation is especially in-
teresting. Amongst the most important facts established
in regard to the Earthworm’s epiderm are, firstly, the existence
beneath the cuticle of a normal cellular matrix, consisting of
varied forms of goblet cells and excessively delicate elongate
interstitial or “ packing ” cells, instead of the altogether im-
probable syncytium described by Claparede ; secondly, the
penetration in the region of the clitellum of vessels of the
haemal system in the form of loops between the groups of
epidermic cells ; thirdly, the similar penetration of processes
of pigment cells belonging to the connective-tissue system
among the epidermic cells of the general body-surface (espc-

304

PROFESSOR E. RAY LANKESTER.

cially in Lumhricus olidus). Tlie publication of these results
was delayed by other work^ which is the less to be re-
gretted since Dr. Horst^ (1876) and Dr. Von Mojsisovics 2
(1877) had but a short time previously applied themselves
to the same investigation and arrived at results of a similar
character, though their memoirs had not come into my hands
until after I had completed my examination of the Earth-
worm’s integument. The drawings of Dr. Von Mojsisovics
do more justice to the very beautiful structure of the Earth-
worm’s epiderm than those of Dr. Horst, at the same time
there are still some points not fully illustrated or appreciated
by the former which I hope to see treated in this journal by
one of my pupils.

The epidermis of the Leech contrasts very strikingly with
that of the Earthworm.

The cells which constitute it are very nearly uniform in
size and columnar in shape (ttVo ii^ch long --o.-^o-o inch broad
on the average).

Gland Cells. — Only a few epidermic cells are enlarged to
form glands, and these do not as a rule remain in the horizon
of the other cells but sink below the surface, having only a
narrow duct to represent them in the epidermic stratum.

These unicellular glands may be rohghly divided into two
series, those which are more superficial and those which
occupy a very deep position. Some of the very deep-lying
epidermal gland cells are seen in Mr. Bourne’s drawing of a
section of a nephridium (PI. XXV, fig. 13 e), whilst a more
superficial one is seen in (PL XXVI, fig. 3y/). I have not
made a special study of these gland cells, which would well
repay the attention of the physiologist, especially that
variety of them which opens into the stomodaeal invagina-
tion of the epiderm, and which is known as salivary
gland. The small and more superficial gland cells appear
to be most abundant about the generative region of the
body, where they may occur so thickly as one to every six
or seven columnar cells when seen in transverse section.
There is little doubt that the superficial series of gland cells
serve, in this region, to secrete the material of which the
egg-cocoon is formed.

Cuticle. — External to the epithelium of the body wall is a
continuous cuticle, which is more delicate than the corres-
ponding cuticle of the Earthworm, and quite devoid of
striations or lamination. It exhibits at intervals (fig. 5)

1 ‘ Aantcekcningen op de Anatomic von Lumbricus terrestris/ Utrecht,
187G.

^ “ Die Lumbriciden hypodermis,” ‘ Sitzber. Akad.,’ Wien, 1877.

CAPILLARIES IN INTEGUMENT OF MEDICINAL LEECH. 305

perforations which are irregularly placed but tend to a dis-
position such as that seen in the figure. These perforations
are the openings of the unicellular glands. Frequently two
such apertures and sometimes three are placed closely side
by side.

Columnar Cells. — When the epithelium is carefully ex-
amined by means of very thin sections and by teazing, it is
found that the columnar cells which form the stratum sub-
jacent to the cuticle have not a simple columnar form but
are T-shaped or rather mallet- shaped (fig. 1). The broad
expanded portion, the head ” of the mallet, forms with its
fellows a complete mosaic (fig. 6) immediately underlying
the cuticle, whilst the handles ” of the mallets are arranged
side by side so as to leave a certain amount of space between
neighbouring handles (fig. 3). Into these spaces pro-
cesses of the connective tissue here and there make their
way, these processes being outgrowths of the pigmented
vaso-fibrous tissue which pervades the whole body of the
Leech (fig. 3 pg). The variously coloured spots of the
Leech’s integument are produced by modifications of this
tissue, and the amount of it which interlocks with the epi-
thelial cells varies from point to point according to the
pattern of the Leech’s colouring.

Columnar cells which have been teased out, after macera-
tion of the Leech in potassium bichromate, show the ex-
panded head ” having a finely granular or sometimes
vacuolated structure, whilst the “handle” appears very nearly
homogeneous (fig. 1). I have not been able by staining
agents to show a nucleus in any part of the cell. The whole
of the “ handle ” takes the staining rather freely, and it is
possible that it consists chiefly or entirely of nucleus.

When a detached flake of epithelium, which has been
stained by picro-carmine, is looked at from above, the appear-
ances represented in figs. 6, 7, 8, 9 are obtained. A complete
mosaic of polygonal cells is seen, in some of which are dark,
that is to say, well-stained bodies, ^vhich look like nuclei. In
others of the polygonal area there are no such darkly-stained
bodies, but only fainter indications of a differentiation. The
result is that the cells with darkly-stained bodies in them
are arranged in irregular groups (fig. 6). The darkly-stained
bodies prove upon further examination to be 7iot ordinary
nuclei, but the “ handles ” or depending portions of the
epithelial cells.

In such fiake-like preparations some of the cells are seen
to be perforated by one or by two circular holes. These arc
the passages for the ducts of the unicellular glands.

306

PROFESSOR E. RAY LANKESTER.

A similar perforation of superficial epidermic cells for the
transit of ducts of subjacent glands is seen in the skin of the
common frog.

Intra-epithelial Blood-capillaries, — In sections from any
region of the Leech’s body the fine vessels of the haemal
system naturally injected with their own haemoglobin-
coloured fluid may be traced up to the sub-epidermal con-
nective tissue, and thence finer branches may be here and
there followed into the spaces between the handles ” of the
mallet-like epithelial cells. Three examples of this dispo-
sition of structures are given in Plate XXVI. Fig. 2 is
drawn from a very thin section, which allows the observer to
trace a horizontal vessel threading its way among the columns
of the epithelium. In fig. S a section is shown which has
fortunately cut one of the^intra-epithelial vessels transversely,
'whilst in fig. 4 a number of vessels are seen supplying the
intra-epithelial plexus, the section being too thick to show
well the individual form of the epithelial cells.

The sections shown in these three figures are merely
samples of a perfectly general arrangement which occurs
over the whole surface of the Leech’s body. The intra-
epithelial vessels form a continuous plexus, giving to the
whole of the epithelial layer of th^ skin a vascular cha-
racter.

General Remarks, — As a histological fact, the vascularity
of this epithelium is a novelty. The only vascular epithe-
lium, so far as I know, which has been hitherto described, is
that of the clitellum of the Earthworm ; but in that case the
vessels do not penetrate between the cells of the most super-
ficial layer.

The penetration into the epithelium of the processes of
the pigmented vaso-fibrous tissue is a phenomenon which
might be expected to accompany the penetration of the ves-
sels, since the vaso-fibrous tissue and the capillary blood-
vessels are virtually one and the same tissue.

Physiologically the significance of the arrangement I have
described is obvious enough. The true respiratory organ of
the Leech is clearly this vascular epiderm, and amongst
respiratory organs it stands alone in the nearness with which
the absorbent blood-vessels succeed in bringing themselves
through all structural obstacles into direct contact with the
oxygenating medium.

CONNECTIVE AND VASIFACTIVE TISSUES OP LEECH. 307

On the Connective and Vasifactive Tissues of the Me-
dicinal Leech.^ By E. Ray Lankester, M.A., F.R.S.,
Professor of Zoology and Comparative Anatomy in
University College, London. With Plates XXVII and
XXVIII.

The minute comparative anatomy of that group of tissues
which is more or less vaguely indicated by the term con-
nective substance (with which I should associate — as is not
usually and expressly done — the vasifactive [angioplastic]
derivatives of connective tissue^) has yet to be worked out,
and when the task has been accomplished the great con-
nective-tissue question/’ in which histologists were but lately
so keenly interested, will be settled.

There can be no doubt that from the point of view of
general morphology, as well as from the more special point
of view of the histologist, the proper understanding of the
nature and relations of the varieties of connective and vasi-
factive tissue is of fundamental importance. You have in
the ideal connective substance an element the vast possi-
bilities of which first impressed the mind of Bichat — an
element which can assume the widest variety of form and
function, embracing the conditions known as tendon, fat,
capillaries, cartilage, areolar tissue, and endothelium.

To study the origins of this tissue and its varieties in
lower forms of organisation cannot but be a fruitful labour,
since the histological differentiation of the higher Vertebrates,
no less than their grosser morphological composition, must
be traced, if we are rightly to comprehend it, to simpler and
more archaic phases which are to some extent at least pre-
served in those branches of the great animal pedigree which
have not advanced so far from a primitive ancestral condition
as have the Vertebrata.

Ectoplastic and entoplastic tissues. — The most fundamental
distinction which can be drawn morphologically between the
various kinds of connective tissues, is one which also serves
to divide other groups of tissues and depends on the mode

‘ See also ‘ Zoologisclier Auzeiger,’ 1880, No. 49.

’ The term “ connective tissue ” is used very unfortuuately in differing
degrees of comprehensiveness — sometimes cartilage, sometimes fat, and
sometimes bone, being excluded or included as the case may be. I would
propose the term “ skeleio-trophic ” fora natural group of tissues which is
divisible into — (1) Skeletal, including fibrous, adenoid, adipose, bony, and
cartilaginous tissues. (2) Vasifactive, including capillaries and embryonic
blood-vessels. (3) Hjemolymph, including the hanna or hccmaglobiuous
clement and lymph, the colourless element of vascular fluids.

308

PROFESSOR E. RAY LANKESTER.

in which the metamorphosis of the embryonic protoplasmic
cells, from which all tissues are developed, commence that
differentiation which leads to the permanent form of the
tissue which they constitute. The mother-cells of all tissues
are either entoplastic ” or ectoplastic,” or both-— that is
to say, the metamorphosis of their protoplasm is either
essentially one occurring at the surface of the protoplasmic
corpuscle, or one occurring deeply within its substance, or
the two processes may go on in connection with the same
cell.^ As examples of the two classes I may cite fat-cells as
essentially entoplastic, and the corpuscles of areolar tissue as
essentially ectoplastic. No doubt both intra-cellular and
inter-cellular deposits do occur in one and the same tissue,
but in most tissues it is possible to point to one or other
mode of deposition, or of metamorphosis, as that which is
characteristic. Tissues which present both phenomena in a
nearly equal proportion may be distinguished as endecto-
plastic.^^ In the group of the connective tissues we find that
the very same chemical and physiological result may be ob-
tained by a metamorphosis which is either entoplastic or, on
the other hand, ectoplastic. Thus hyaline cartilage is essen-
tially ectoplastic, whether we take the well-known type with
rounded corpuscles embedded in a matrix, or the more un-
usual form with arborescent cells found in the cartilages of
fish and of Cephalopoda. On the other hand, notochordal
tissue, which is often loosely spoken of as a kind of car-
tilage,’^ differs profoundly from ectoplastic cartilage in the
fact that the cell metamorphosis is essentially entoplastic,
occuring loiihin the area of the original protoplasmic cor-
puscle, and not in the form of a deposit surrounding and
embedding the embryonic unit. In that variety of con-
nective tissue which chemically differs from cartilage in
yielding a less dense and resistant product of metamorphosis
than that which is associated with the name of cartilage, we
find equally the two great morphological varieties of ento-
plastic and ectoplastic tissue. Fibrous tissue generally is
ectoplastic — that is to say, the protoplasmic corpuscles re-
main more or less intact whilst surrounded by the fibrous
and lamellar masses to which they have peripherally or
laterally given origin. This is true of ordinary subcutaneous
areolar tissue, of tendon, of mucous tissue (umbilical cord,
&c.), and of corneal tissue. At the same time we find
in various Invertebrate groups (not in the Vertebrata) an

^ When this is the case the ectoplastic and the entoplastic products of
metamorphosis arc usually of a different chemical nature and of different
physical properties.

CONNECTIVE AND VASIFACTIVE TISSUES OF LEECH. 309

entoplastic form corresponding chemically and functionally
to the ectoplastic forms just cited. This is the grosshlasige
Bindegewebe the vesicular connective tissue so abundant
in the Mollusca, in the Nemertines, and other Invertebrata.
The only tissue which in form represents this among the
connective tissues of Vertebrata is adipose tissue. Vesicular
connective tissue stands in the same relation to mucous
tissue, tendon, and corneal tissue, as does notochordal tissue
to hyaline or fibrous cartilage.

Yet further, the tissues of the connective group which are
specially related to the nutrient fluids (such as blood and
lymph), and which form the wall of the coelom or of blood-
channels, may be entoplastic when they give rise by internal
metamorphosis (liquid vacuolation) to capillary vessels, or
ectoplastic when they constitute spongy or lacuniferous cell
aggregates, the cells separated by inter -ce\\\x\?iY channels, such
as we find in the pulp ” of lymph-glands and the spleen
and in the lacunar tissue of Molluscs.^ The recognition of
these two modes of cell metamorphosis appears to me to be
important, as enabling us to bring together as morphological
varieties of one and the same tissue, having the same
chemical and physiological significance tissues, which are
frequently separated from one another without scientific
method, or else are abitrarily placed side by side in con-
sequence of the recognition of chemical and physiological
resemblances, without sufficient emphasis being given to
their very different morphological character.

Ectoplastic and entoplastic connectwe substance of the
Leech. — In the medicinal Leech two forms, and only two
chief forms of skeleto-trophic tissue are to be discovered.
The one is an ectoplastic connective substance resembling
somewhat the jelly-like tissue of the umbilical cord (PI.
XXVII, fig. 1) ; this I shall term ectoplastic connective
jelly ” (in contrast to the entoplastic connective jelly ” so
abundant in Molluscs). The second kind of skeleto-trophic
tissue found in the Leech consists of greatly branching and
anastomosing fibres (PI. XXVII, fig. 1, fb, and fig. 3),
which are often darkly pigmented by fine granules, and are
not unfrequently tubular (fig. 4), passing by insensible gra-
dation and actual continuity of substance into the form of
very delicate capillaries, which abound in all parts of the
Leech’s body and contain hsema, that is, fluid impregnated
with haemoglobin. It is a modification of this tissue which
forms the pigment patches of the integument (sec PI.

' Sec 11. It. Peck, “ On the Structure of the Lamcllibraiich Gill,” this
Journal, 1870.

310

PROFESSOR E. RAY LANKESTER,

XXVI, fig. 4), and an extreme modification in the opposite
direction, which appears under the form of botryoidal tissue
(PL XXVIII, fig. 15) called hepatic by Brandt, mistaken
for cutaneous glands by Leuckart, and compared to the fatty
body of insects by Leydig. I propose to term this ento-
plastic form of skeleto-trophic tissue — which has so Protean
a character — vaso-fibrous tissue.”

The connectim jelly of the Leech. — This tissue may be
best studied on perfectly fresh samples, which are to be
obtained by teasing any portion of the muscular paren-
chyma which intervenes between the integument and the
wall of the alimentary canal.

It consists of a jelly-like matrix, in which are embedded nu-
merous branched corpuscles, varying in diameter from t^-o tfi
to the of an inch.i The branches of the corpuscles are

drawn out into very fine processes, and finally are lost in
exceedingly delicate filaments, which permeate the jelly in
every direction. Treatment with nitrate of silver gives the
usual result observed with such tissues, viz. a dark staining
of the matrix, whilst the corpuscles are left as clear spaces.

The corpuscles vary in size and in colour in different
specimens of the tissue. When young they are small and
contain a few highly refringent granules of large and regu-
lar size. Older tracts of the tissue show corpuscles of larger
size, the granules within them being very numerous, of large
size, and similar to one another ; the granules are in these
older cells of a decided brown or straw colour. In some
parts the corpuscles may be found to be very much elongated,
and suggests the possibility of a transition from the ecto-
plastic connective jelly to the entoplastic vaso-fibrous tissue.

Even in perfectly fresh corpuscles of the connective jelly a
clear space amongst the granules may be observed, which
indicates the nucleus. Treatment with osmic acid, followed
by picro-carmine (on the glass-slip, whilst under a cover-
glass, in the manner recommended by Banvier), gives two
interesting results. First, the granules become very much
stained by the osmic acid, and, secondly, the nucleus, but no
other part of the corpuscle, becomes coloured by the carmine

(PI. XXyil, fig. 2).‘

The highly refringent granules of these corpuscles are
very characteristic on account of their largeness, their colour,
and their regularity of form and size. They are not simply
fatty matter, since they are not dissolved by ether. They
arc possibly of the same chemical nature as the very much

* A very similar tissue oeeurs beueatli the epidermis of Sipunculus
nudus, having, however, a more cartilaginoid matrix.

CONNECTIVE AND VASIFACTIVE TISSUES OF LEECH. 311

more minute brown-coloured granules which occur in various
parts of the vaso-fibrous tissue.

The vaso-Jlhrous tissues of the Medicinal Leech* —
Whenever any portion of the muscular tissue or any organ
of the leech is examined under the microscope^ a large
quantity of dark brown fibres are sure to be noticed^ either
forming an investment to the organ or ramifying over the
field of the microscope (PI. XXVII, fig. 3).

This system of fibres penetrates every region of the
Leech’s body. It is continued superficially into the epi-
dermis, beneath and between the cells of which it forms
the characteristic pigment spots. It occurs between the
bundles of muscular fibres, as may be seen in the section
(fig. 13 g) of Mr. Bourne’s PI. XXV. It forms a
loose sort of feltwork on the surface of the nephridium of
the testes and of the alimentary canal. It is particularly
rich and thickly developed on the wall of the blood-sinus in
which the ventral nerve-cord is placed.

When examined with a high power the fibres are seen
to be of varying thickness, and many of them apparently
hollow, so that the larger fibres are tubular rather than
solid, and moreover, the tubular wall, according to the con-
dition of tension or relaxation of the tissue, may be smooth
or thrown into transverse rugae (PI. XXVII, fig. 1). Treat-
ment with osmic acid, followed by picro-carmine, has the
effect of demonstrating the nuclei of the cells by which this
tissue is formed. The nuclei (PI. XXVII, fig. 4) are seen
to be scattered at intervals in the course of the fibre, and
when this is tubular they project internally into the lumen
of the tube. The substance surrounding the nuclei is densely
packed with very fine granules and does not stain red with
picro-carmine as do the nuclei themselves. In many fibres
(apparently old and fully differentiated) nuclei are either
exceedingly rare or are not to be detected. It appears
probable that in these fibres the nuclei of the original cells
have disappeared and left only the finely granular cell-
substance to play the part of a connective substance.

In many of the tubular fibres, however, the nuclei are
abundant, and show a series of transitional conditions, indi-
cating that they gradually become freed from the granular
cell-substance, and at first project into the cavity of the tube,
and then become detached and lloat freely in it (PI.
XXVII, fig. 4 and fig. 5). Some of the fibres of this
system, which are definitely continuous with the darker and
thick- walled fibres, have a pale aspect and relatively few
granulations in the wall, whilst the nuclei are abundant

VOL. XX. NEW SEK. X

312

l’R(3rESS0R E. RAY LANKESTER.

(fig. 5). The appearance of these pale hollow fibres at once
impressed me with a close similarity to the developing
capillaries of Vertebrata. I had previously been able to
demonstrate (see this Journal, vol. xviii, '' The Vascular
Fluid of the Earthworm a Corpusculated Fluid ”) that the
delicate vessels of the red vascular system of the Earthworm
contain corpuscles, which are nothing more nor less than the
liberated nuclei of cells which line the walls of those vessels,
and, accordingly, I was prepared to find in the vessels of the
Leech (in the congeners of which the existence of blood*
corpuscles has been generally asserted) corpuscles deve-
loping by the liberation of nuclei belonging to the cells
which constitute the vascular wall.

At the same time the structures which I had before me
(PI. XXVII, figs. 4, 5, 6, 7) were not vessels forming part
of the red vascular system of the Leech, but tubular fibres
connected with the general system of brown-coloured fibres
pervading the body of the Leech. The vessels, in fact, in
which I observed this displacement of the nuclei were not
filled with haemoglobinous fluid, but had clear contents, and
evidently had not yet — if ever they were destined to have —
any communication with the red-coloured vascular system.
My next care, therefore, was to search for any evidence of
the connection of such tubular portions of the brown-
coloured system of fibres with the blood-vessels.

Thin-walled blood-vessels of the Leech. — The finer vessels
of the Leech’s haemoglobinous vascular system are remark-
able for the exceeding delicacy of their walls. In no in-
stance have I succeeded, by means of reagents, in demon-
strating any nuclei in those walls ; and, further, by the use
of nitrate of silver I have not succeeded in obtaining any
indication of cell-outlines on those walls which I expected
to be able to demonstrate since Mr. D’Arcy Power had, when
working with me at the histology of the Earthworm, ob-
tained such cell-outlines in the finer vessels of that animal
(this Journal, 1878), where, moreover, nuclei are always
readily demonstrated, even in the very finest capillaries.
The structureless character of the thin-walled blood-vessels
of the medicinal Leech is, then, a peculiarity which separates
them from the apparently similar vessels of the Earthworm,
and renders it not improbable that the two sets of vessels
may have a diverse developmental history. Rarely I have
found corpuscles consisting of free nuclei in the vessels con-
taining red fluid in the Leech. Two such instances are
drawn in PI. XXVII, figs. 13 and 14. In the one case the

CONNECTIVE AND VASIFACTIVE TISSUES OF LEECH. 313

red fluid is still present^ in the other the action of reagents
had caused its removal^ leaving only the corpuscles.

Connection hetioeen the hroicn fibres and the thm~walled
vessels. — Here and there a careful search in longitudinal
sections of the Leech (hardened in chromic acid^ per cent.,
followed by alcohol) revealed some thin -walled vessels
(filled with their red fluid), in which the usual structureless
character of the w'all w^as modified. Instead of being abso-
lutely structureless the membranous wall showed thick-
enings, which were densely packed with fine brown granules,
identical with those of the brown fibrous tissue. After
some search I found several examples of the direct continuity
of thin-walled vessels with the tubular fibres of the brown-
pigmented, system of fibres. Three such instances are repre-
sented in PI. XXVII, figs. 8, 10, and 11. Figs. 9 and 12
represent csecal terminations of branches of the red vascular
system, in which it appears that the csecal extremity of the
branch is continued into a fine fibre. Such an appearance
admits of the interpretation that a prse-formed tube has
been placed in communication with the red vascular system,
wLich has then filled it wfith hsemoglobinous fluid up to the
point where its lumen ceased, and where the tube ceases to
be tubular and is merely a solid fibre.

From these observations I cannot doubt that the thin-
walled blood-vessels and the brown-pigmented fibres of the
medicinal Leech are virtually one and the same tissue, pass-
ing into one another by insensible gradation, and, what is
of more importance, actually in continuity with one another
at certain points, as processes of one and the same set of
vaso-fibrous cords.

It is not possible at present to oflfer a definite opinion as
to whether the development of tubular portions of the brown-
pigmented system of fibres into thin-w^alled vessels is a pro-
cess which is always and normally going on in the LeeclFs
body. It is possible to suppose that the solid browm fibres,
the tubular browm fibres, and the thin-walled haematophorous
vessels, are three permanent varieties of the vaso-fibrous
tissue. But I am inclined to think from the fact that not
even young haematophorous vessels with nuclei in their w’alls
are to be met with in the Leech, and from the fact that the
tubular browui fibres discharge into their tubular cavity
nuiueious nuclei identical in appearance with the corpuscles
which are here and there to be detected in the haemato-
phorous vessels, that there is actually a continual develop-
ment of offshoots of the brown-pigmented system of fibres
into thin-walled haematophorous vessels, the nuclei enter-

314

PROFESSOR E. RAY LANKESTER.

iiig llie blood-stream when communication is once esta-
blished and the brown granular wall becoming transparent
and structureless by the absorption of its granules.

Botryoidal tissue {so-called hepatic tissue), — I have now
to notice another development of the vaso-fibrous tissue, the
characters of which are such as to confirm the view already
advanced as to the potential identity of the brown-pigmented
fibres and the vessels of the red vascular system. Nearly
surrounding the wall of the alimentary canal, but separated
from the epithelium of that organ by a vascular plexus, by
a peculiar series of muscular fibres, and by fibres of the vaso-
fibrous system, is a mass of dark-brown botryoidal tissue
(PI. XXVIII, fig. 15), which has been, on account of its
position and colour, regarded as hepatic.'^ I was not a
little surprised on studying the structure of this tissue by
means of transverse sections and of teased preparations, to
find that it has in no way the least structural resemblance to
a hepatic gland, but is simply a plexus of hsematophorous
vessels, the Avails of which are swollen in botryoidal fashion by
the enlargement of the individual cells which constitute them.
The cells which constitute the wall are a single series ; there
is no external tunic and no internal endothelium. In those
parts of the plexus where its special character is fully de-
veloped every constituent cell is seen to have a hemispherical
shape, standing out on the surface of the vessel (PI,
XXVIII, fig. 19), and its substance is found to be densely
charged with very fine brown-coloured granules. A clear
nucleus can be detected in most of the cells. The lumen of
the vessels bounded by these singular cells is larger than
that of the average thin- walled haematophorous vessel, mea-
suring of inch on an average. The vessels form a

true network in the denser part of the plexus, but every-
where csecal branches of the vessels are found, which appear
to be buds, as it were, of the system, which are growing for-
ward to form unions with other branches, and so extend the
plexus ; near the boundary of the masses of botryoidal tissue
very interesting preparations may be obtained, showing the
connection between the botryoidal vessels Avith thickened
cells to their Avails and the thin-Avalled vessels of the general
haematophorous system. Such a preparation is draAvn in
PI. XXVIII, fig. 18. The botryoidal vessels are seen as
buds or branches carried by a thin-Avalled vessel.

The nature of the granules contained in the cells of the
botryoidal vessels I have not in any Avay determined, and I
am altogether unable to suggest at present Avhat may be the
significance of this portion of the vaso-fibrous system. It

CONNECTIVE AND VASIFACTIVE TISSUES OF LEECH. 315

appears to me that the granules are similar to those of the
brown-pigmented fibres, but of this I am by no means
certain.

Further, it appears exceedingly probable that these granules
are of the same physiological nature as the yellow-brown gra-
nules which load the endothelial cells of the coelom of the
Earthworm and other Chaetopoda in the region of the alimen-
tary canal. In the Earthworm these cells not only form a com-
plete tunic to the intestine, but also invest the large blood-
vessels. At the same time it must be distinctly pointed out
that the cells which thus clothe the large vessels of the
Earthworm and the contractile vascular caeca of Lumbri-
culus are not the cells which form the proper wall of the
blood-vessel, but lie externally to these. Hence their rela-
tion to the vascular space and to the haemoglobin ous fluid is
essentially difierent from that of the granular cells of the
botryoidal vessels of the Leech.

Previous ohservations relative to the brown fibres and
botryoidal vessels of the Medicinal Leech, — It is not desirable
to go over the history of a subject which has not of late years
formed the subject of serious investigation. The opinion of
Brandt that the botryoidal form of the vaso-fibrous tissue of the
Leech was hepatic in character has never been maintained by
any histologist. The two observers who have studied the Leech
with the aid of the microscope, namely, Leuckart and Ley dig,
have both rejected BrandFs view as to the botryoidal tissue, but
have advanced divergent opinions of their own.

Leuckart Die Menschlichen Parasiten,*’ vol. i) studied the
anatomy of the Leech by means of transverse sections, and
arrived at the conclusion that BrandPs hepatic tissue^^ was
nothing more nor less than a mass of deep-lying epidermal
glands. In this he was misled by the circular outline of the
botryoidal vessels when in section, and by the fact, which he
was the first to observe, that unicellular epidermal glands do
travel very deeply into the body- wall of the Leech, and form
groups of spherical cells, which are obvious enough in trans-
verse sections (see Mr. Bourne’s fig. 13, Plate XXV).

Leydig, in his ^Histology’ (1857), states: Relatively to

the liver of the Leeches, properly so called, my observations
compel me to differ completely from the prevailing opinion.
Certain utriculi,of a brownish-yellow colour, which envelope the
stomach and intestine, are commonly considered to constitute
the liver ; these utricles open one into the other by means of
their excretory canals, and pour their contents on to the internal
surface of the intestine. Contrary to this view, I venture lo

316

PROFESSOR E. RAY LANKESTER.

maintain that the tissue of the Leeches, which is regarded as a
‘Miver, has a very different signification; it ought to be placed
alongside of the fatty bodies of the Arthropoda. It is formed
of cells of various sizes and of variable form, round, elongated,
sometimes drawn out, and fibrous ; in other cases the cells are
ramified and the prolongations anastomose with one another ;
frequently they form tubes with hemispherical prominences ; in
short, they reproduce all the varieties of form which are pre-
" sented by the cells which compose the fatty body of Arthropods.
In Hirudo, Hsemopis, and Nephelis, the cell- content is formed
by a brown granular mass in greater or less abundance. Just
as the fat body of Arthropods is in connection with the external
“ membrane of the tracheae, intestines, &c., so also the tissue of
" the Leeches, which has wrongly passed, up to the present time,
for the liver, is in connection with the connective-tissue envelope
“ of the intestine ; it embraces not only the alimentary tract, but
it also constitutes the brown envelope of the testicular vesicles,
the tunica adventitia of the vascular trunks, the loose brown en-
velope of the nervous system, &c. ; in a word, this liver is simply
a form of connective tissue, which, in the absence of a proper
perivisceral cavity, fills up all the interstices situated between
the organs which it invests. Its resemblance to the fat body of
Arthropods has other facts to support jt. Thus, although the
brown granules make up the bulk of the cells, yet it may be ob-
served {HamopiSj for instance) that, among the brown-coloured
networks, there are other fibres, the cells of which have, as con-
tents, colourless granules of a fatty nature ; and, what is still
more striking, is that in Clepsine and Piscicola a well-developed
fatty tissue occupies the place of these brown networks. Where
the cells form, by means of their outgrowths, a system of mesh-
works the cavities are filled by gelatin. Besides observation on
" fresh specimens,! recommend the following mode of preparation :
— A leech is thrown into hot water, it is then dried, and fine
transverse sections are made, which are then soaked in slightly
acidulated water. It is then clearly seen that the connective
^Uissue which envelopes all the organs, taking its point of de-
parture from the integument, and, traversing all the muscles, is
“ filled, in certain parts and in its cellular elements, by brown
granules, and, moreover, that the colouring matter is of the
same nature as that of the integument.’^

Thus, it appears that Leydig had arrived at very nearly the
same conclusion with regard to the relationship of the botry-
oidal tissue and the brown fibrous tissue as that to which I
have been led. He had, however, entirely missed the connec-
tion of both these forms of the vaso-fibrous tissue with the

CONNECTIVE AND VASIFACTIVE TISSUES OF LEECH. 317

hsematophorous vessels, the blood-vessels as they are usually
called.

In a much more recent work, however, namely, in his excel-
lent treatise on the anatomy and histology of the Oligochsetous
wovm, Phreor^ctes menkeamts KxcYliy f. mikrosk. Anatomie,'
vol. i, 1865), Leydig has touched upon various points in the
anatomy of other worms illustrative of Phreoryctes. He there
figures a small piece of the botryoidal tissue of the Leech and
remarks on its vascular nature, showing that since the publica-
tion of his ^ Histologic,^ he has ascertained the fact that the cells
of the botryoidal tissue form the walls of blood-vessels.

The observations of Leydig which I have above cited relative
to the continuity and identity of the epidermic pigment, inter-
muscular and tunic-forming fibres, and botryoidal or so-called
hepatic tissue, show that he had arrived at the conclusion that
what I have called the vaso-fibrous tissue is one and the
same structural element with a variety of modifications, but he
has not included in its area the hsematophorous vessels, as I have
found it necessary to do, nor is the very valuable passage which
I have cited from the ^ Histologic ^ explained by drawings either
in that work or by those which the author has given in illustra-
tion of his most admirable papers, published from time to time
in the ^ Zeitsch. f. wiss. Zoologie.'

Injection of the vaso~fibrous tissue. — In the course of some
experiments made by injecting indigo-carmine beneath the in-
tegument of the Leech in order to observe (if possible) the excre-
tion of that substance by the nephridal cells, I was struck by
the fact that in transverse sections of leeches so treated, the in-
termuscular brown fibres of the vaso-fibrous system were very
uniformly coloured blue by the indigo-carmine, to the exclusion
of other parts of the organism. I am not able to say whether
this was a mechanical injection of the tubular fibres, or whether
it was rather due to a selective vital action on the part of the
vaso-fibrous tissue, similar to that noted in Vertebrate areolar
tissue by Arnold. I am inclined to think that it was not me-
chanical since the vessels of the ha3matophorus system were not
injected. This latter fact could, however, be explained on the
hypothesis that these vessels were already filled with their haemo-
globinous fluid, whilst the tubular fibres of the brown fibrous
tissue were in a non-distended but distensible condition.

318

DR. HENEAGE GIBBES.

On the Use of the Wenham Binocular with High
Powers. By Heneage Gibbes, M.B.

A GREAT many plans have been tried for the purpose of
obtaining stereoscopic effect with high powers, and these
have all required a special stand or special apparatus entirely
removed out of the reach of the ordinary working micro-
scopist, and this led I think to the prevalent idea that the
binocular microscope is entirely unsuited for histological or
pathological research.

Mr. Stewart, of St. Thomas’s Hospital, showed me a -u of
Zeiss’s, which, by removing the front part and screwing it
on to an adapter, gave perfect stereoscopic effect, and this
induced me to try the oil-immersion lenses in the same way.
I unscrewed the front of my -jV oil by Zeiss, and screwed
it on to an adapter made in such a manner that the lens was
brought as close as this form of mount would allow to the
prism, and by cutting off the lower part of the slide below
the rackwork of the coarse adjustment I was enabled to
bring the objective low enough to reach the object.

With this arrangement I found tjiat I could not get per-
fect stereoscopic effect, but I got it near enough to show me
that if I could bring the glass a little nearer the prism the
effect would be obtained, the black bands on either side of
the field being very small. With this view Messrs. Powell
.and Lealand made me a oil immersion in w'hich the front
is made to screw off just behind the back combination, and
a screw is cut on the outside, so that the front can be
screwed into an adapter, and this again into the body of the
binocular microscope, thus bringing the objective almost in
contact with the prism.

AVith this glass the stereoscopic effect is perfect, the whole
.field is illuminated, and the result obtained is really wonder-
ful. Taking a preparation of the tadpole’s tail hardened in
gold solution, the different elements are seen in their true
relations to each other; there is no difficulty in deciding
whether a line nerve-termination passes over or under or
INTO a connective-tissue corpuscle. Cells are seen not as
flat plates, but as spheroidal bodies, with their intranuclear
and intracellular network pervading their whole substance.

The only difficulty I have found is to persuade people that
the power is really so high, objects stand out in such bold
relief, they cannot believe it possible. I have had a
made by Messrs. Powell and Lealand, and this glass, although

USE OP WENHAM BINOCULAR WITH HIGH POWERS. 319

there is a slight shade on either side of the field, gives per-
fect stereoscopic effect.

These glasses are of great excellence, and I think excel
Zeiss’s in definition ; they have also this advantage, they are
perfectly homogeneous, and require the ordinary cedar oil
for either oblique or central light, and do not require any
adjustment of the draw tube for thin covers, as is necessary
with Zeiss’s glasses.

Any ordinary stand made on the Jackson-Lister model can
be used, if the body is made to come down in contact with
the stage, as the objective only projects of an inch.

An achromatic condensor is also necessary. I have used
Powell and Lealand’s and Zeiss’s, but any maker’s with a
large angle would probably do.

320

DR. HENEAGE GIBBES.

On the Structure of the Spermatozoon. By Heneage
Gibbes, M.B.

Since my last paper appeared in this journal in October,
1879, I have had little time to make further researches into
this subject.

I have, however, examined the spermatozoa of several
animals, such as rat, mouse, axolotl, pigeon, in all of
which I have found the long filament formerly described ;
also in the pigeon, where it resembles the Spermatozoon of
an Amphibian on a smaller scale.

I have examined a few Invertebrates, but so far have only
found the filament in the snail and leech. In the snail
it appears to be only a little longer than the tail ; this,
however, is enormously long, which may account for it.

In the fowl the Spermatozoa resemble that of the newt in
shape, but the filament is very fine, and consequently indis-
tinct. I have examined a few fish, but the spermatozoa were so
small I could not make out anything satisfactorily. I have
examined a number of specimens of the human spermatozoa,
some taken from the testes twenty-four hours after death by
accident, others from twelve to twenty-four hours after coi-
tion ; these came by post, and I could detect no material
difference in their structure, with the exception of the vary-

STRUCTURE OF THE SPERMATOZOON.

321

cient data to go upon. In all I found the filament very fine,
and requiring great care in the illumination to show it pro-
perly, as when, after a good deal of trouble it was shown
well with the ^ immersion, I found I could readily see it
with the y’g oil, and even the -^V oil immersion.

The drawing was made from a specimen prepared in a
mixture of glycerine and absolute alcohol, and drawn with
Messrs. Powell and Lealand’s — oil immersion and A eye-
piece. The filament is very fine, and is connected to the
tail by a membrane, which is much wider than in the
Amphibia, and allows it to move further from the tail ; it
is also longer than in the Amphibia, and is more folded in
consequence. I found in one specimen a number of heads
with no corresponding tails. I have not, however, been
able to get at the history of the individual who supplied it.

322

P. HERBERT CARPENTER.

Some disputed points in Echinoderm Morphology. By
P. Herbert Carpenter, M.A., Assistant Master at
Eton College.

The last number of the ^ Zeitschrift fiir wissenschaftliche
Zoologie ’ (Band xxxiv. Heft 2) contains two valuable papers^
on Echinoderms, by my friend Dr. H. Ludwig, of Bremen.
Many new observations of the highest interest are here
recorded, and it is not too much to say that their accuracy
may be relied on with the utmost confidence. But, at the
same time, I believe that some of the conclusions which
Ludwig has drawn from them are essentially unsound ; and
in the following pages I propose to state my reasons for this
belief.

In the first place, however, I desire to state that some of
the observations recorded by Ludwig in these papers have
convinced me that certain views which I have advocated
in the pages of this journal are no longer tenable. Four
years ago Ludwig described some tubular appendages of the
water-vascular ring of the Crinoids as opening below into
the body cavity into which they depend ; and he therefore
considered them together with the water-pores of the disc as
collectively representing the madreporic apparatus and sand
canals of the other Echinoderms.

Like Greeff, however, I did not feel quite satisfied respect-
ing the alleged opening of these tubes into the body cavity,
as I knew from my own observations of them (made like
those of Ludwig by the section-method) that the liability to
error was considerable. But now that Ludwig has found
them to be open in the uncut Pentacrinoid larva I see no
reason to doubt that the same is the case in the adult.
Consequently, the resemblance^ that I, like Professor
Huxley, believed to exist between these (apparently caecal)
tubes and the closed oasa ambulacralia cavi of the Ophiurids
will not bear investigation.

But, while abandoning this position, I must still confess to
a lingering doubt as to whether the disconnected water-pore
and water-tube of the Antedon larva can be regarded as
perfectly homologous with the madreporic system of the
other Echinoderms. I fully admit the similarity of their

* “Ueber den primaren Steincanal der Crinoideeu uebst vergleicliend
anatomischen Beraerkungen iiber die Echinodermen iiberbaupt/’ pp. 310 —
332, Taf. xii u. xiii. “ Neue Beitrage zur Aiiatomie der Ophiuren,” pp.
333—30 5, 'I’af. xiv — xvi.

^ This Journal, vol. xix, New Ser., p. 10.

ECHINODERM MORPHOLOGY.

323

histological structure and of their functions as afferent
channels of the water-vascular system. But this similarity
does not constitute an homology. The dorsal pore, which
is developed so early in the other Echinoderm larvae as the
foundation of the madreporic system, is absent from the
corresponding stages of the Antedon larvae ;• and until the
development of the primary water-pore and water-tube of
the Crinoids are more fully known, I do not think that we
can without qualification accept Ludwig^s view regarding
their homology with the madreporic system of other Echino-
derms. The following remarks, however, are based upon
the assumption of the truth of that view.

Ludwig regards the external opening (water-pore) of the
madreporic canal (water-tube) as a fixed point. Whether it
be actinal or whether it be abactinal does not matter. In
the larvae of some Crinoids and in the adult Ophiurids it
opens on one of the interradial actinal plates (orals) around
the mouth. In the adult Urchins and Starfishes, however,
it opens on one of the interradial abactinal plates around
the anus (genitals). This is far from being the case in
young forms of either class. Nevertheless, Ludwig thinks
the relations of the parts in the adult sufficiently fixed to
enable him to assert the undoubted homology of the orals of
Crinoids and Ophiurids with the genitals of Urchins and
Starfishes. He attempts to strengthen this remarkable
position by other considerations to which I shall presently
advert; and he points out how diametrically opposed it is
to the view now generally held that the Basals of Crinoids
represent the genitals of Urchins and Starfishes.^ He seems,
however, not to attach much value to the evidence on which

^ I have often wondered whether the “ pseudoproct ” of Wyv. Thomson,
a depression at the hinder end of the very young Antedon larvse, can be
regarded as representing an undeveloped dorsal pore. Its position at the
end of the transverse Gastrula-axis is the same as that of the dorsal pore
in other Echinoderm larvae, though I am bound to admit that it is not dorsal.
But owing to the extreme elongation of the Gastrula along that axis it is
very far from the water-vascular ring and not close to it as in other larvae.
This might account for its want of further development. On the other hand,
the primary water-pore of the Antedon larvm leads into the oral ccelom
developed from the left peritoneal vesicle of the embryo. It is not likely,
therefore, ever to have been a simple dorsal pore, or it would have opened
into the aboral coelom developed on the dorsal side of the larva from the
right ])eritoneal vesicle, su])])osing of course that it were not directly con-
nected with the water-vascular ring by a water-tube, as in other Echino-
derms.

“ See “The Oral and Apical Systems of Echinoderms,” this Journal,
New Scr., yols. xviii and xix ; and also the recently published text-books
of Claus, Zittcl, and other authors.

324

P. HERBERT CARPENTER.

this view is based, or he would scarcely have written as
follows (p. 317) :

“ Der ganze Beweis fiir diese AuflFassung hegt darin, dass, wenn man von
dem Mittelpunkt der dorsalen Oberflacbe des Thieres ausgeht, bei den
Crinoideen die Basalia, bei den Ecliinoideen die Genitalia die ersten Platten
sind, welcbe stets und immer in der Bicbtung der Interradien angeordnet
sind. Irgend welcbe unmittelbaren Beziebungen zu den inneren Organen des
Tbierkorpers sind bei dieser lediglicb auf die raumlicbe Auordnungsweise
jener Platten gegrundeten Homologisirung nicbt in Betracbt gezogen
worden.”

The view attacked by Ludwig rests, however, on a much
firmer foundation than the mere anatomical fact which he
calls the ganze Beweis ” of its truth. The evidence of
Echinoderm embryology is all in its favour ; but Ludwig
does not make the slightest mention of this evidence, much
less attempt to controvert it.

It is as follows : — The interradial abactinal plates (basals)
of Crinoids are developed with precisely the same relation to
the vaso-peritoneal apparatus of the larva as are the inter-
radial abactinal plates (genitals) of Urchins and Starfishes.
In both cases these plates first appear in the form of a spiral
around the right peritoneal vesicle (Agassiz, Gotte) ; but
the orals of the Crinoids, which Ludwig considers homo-
logous with the genitals of Urchins and Starfishes, are
developed spirally round the left peritoneal vesicle (Gotte).
Are not these peritoneal vesicles as "important inner organs’^
as any in the whole morphology of the Echinoderms ?
They develop before the water-vascular apparatus, on the
connection of which with certain plates of the adult,
Ludwig lays such stress.

He admits that the oral or actinal side is homologous in
all Echinoderms as being chiefly developed from the left
side of the larva ; and yet he considers plates developed on
the left or actinal side of one Pluteus larva (Ophiurids) as
undoubtedly homologous with those developed on the right
or abactinal side of another (Echinids), and leaves the
abactinal plates of the Ophiurids (genitals) out of consi-
deration altogether.

It seems to me that he assumes too much in regarding
the water-pore of the adult as a fixed point. If he could
show that one of the genitals of an Urchin or Starfish larva
were primitively perforated by the water-pore as is one of
the orals of a larval Crinoid, then his position would be
strong indeed. But this is very far from being the case,
lie admits himself that in young Starfishes (Loven) there is
no connection between the water-tube and any genital plate.

ECHINODERM MORPHOLOGY.

325

The observations of MetschiiikofF and Agassiz are quite in
accordance with this fact. The former figures the apex of
a young Starfish in which the madreporic plate is already
formed, but is situated at the edge of the disc quite outside
the circle of genital plates ; while Agassiz represents a
young Starfish which shows the position of the madre-
poric body immediately on the edge of the disc of the lower
(i.e. actinal) side.^^ Somewhat the same is the case in the
adult Ophiurids. Although developed on the dorsal surface
of the larva the water-pore is usually on the actinal surface
of the adult ; while in Trichaster it is neither abactinal nor
actinal, but intermediate in position, somewhat as in the
earlier stages of the young Starfish.

In the same way the anus of the Starfish is at first upon
the actinal side near the edge of the disc.” But as growth
proceeds it moves towards the abactinal surface together with
the water-pore. In the Crinoids, however, the position of
the future anus is gradually shifted in the reverse direction,
i,e. towards the actinal surface ; while the water-pore must be
developed late, unless we are to suppose that it escaped the
notice of both Gotte and of Wyville Thomson in the earliest
larval stages. But whereas in the other Echinoderms it is
developed very early, before the appearance of the actinal
or abactinal plates, the rudiments both of these and of the
chief organs of the Crinoid appear before it, and then it is
only found perforating the “ seitlichen Randtheil ” of an
oral plate.

As far as the Crinoids are concerned, this is the chief
evidence in favour of Ludwig’s views, but it is not complete
even for this group. For while Rhizocrinus has five water-
pores like some exocyclic Urchins, their openings are not in
the persistent oral plates^ as they should be on Ludwig’s
theory, which compares these orals to the genitals of the
Urchins, because one of them in the larval Antedon is per-
forated by the w’ater-pore. Further, the orals of the Crinoids
have precisely the same relation to the water-vascular ring
and to its tentacular apparatus surrounding the mouth as
the five plates figured by Kowalevsky in the Psolinus larva.
Ludwig will hardly deny that these last are orals, but not
one of them is perforated by the water-pore. On the con-
trary, it is separated from them by the whole length of the

^ If the orals of Rhizocrimis are perforated by the water-pores, surely
Ludvyig would have said so ; and I gather from his figure of the water-j)ore
ill this genus that this is not the case (‘Zcitschr. f. wiss. Zool.,’ xxix, Taf.

V, fig. 8).

326

P. HERBERT CARPENTER.

water-tube, which extends backwards from the water-vascular
ring and not forwards as it should on Ludwig’s theory.

Again, what is the position of the five water-pores of Tri-
chaster elegans ? Not close round the mouth in the position
of the absent mouth-shields, but between the two genital
clefts of each interradius. The interradial plates sometimes
developed here in other Ophiurids have been hitherto re-
garded as representing the genitals of Starfishes and Ur-
chins; and the presence of the water-pores in the same posi-
tion in Trichaster elegans goes a long way towards strength-
enins: that view, while at the same time it diminishes the
probative value which Ludwig assigns to the perforation of
the mouth-shields by the water-pores in other Ophiurids.

Besides the perforation of one of the orals of the larval
Crinoid by the water-pore, the only other arguments adduced
by Ludwig as showing the homology of these orals {vice
the basals, discarded) with the genitals of Urchins are the
following: — (1) The anus is adoral from the basals of a
Crinoid, but aboral from the genitals of a regular Urchin ;
though it is aboral from the orals of a Crinoid. I would
urge two considerations in reply. There are Crinoids in
which the mouth is at the margin of the disc, close down
to the abactinal skeleton, while the anus is central ; and I
do not see why these and the irregular Urchins should be left
out of consideration, especially when we remember that the
mouth of the Pentacrinoid larva is excentric, as Ludwig
himself has told us, and that the exocyclic Crinoids present
other embryonic features.

In the young Starfish the anus is primitively adoral from
the genital plates, and only secondarily assumes the intra-
genital position on which Ludwig bases his argument.
Here, as in other cases, Ludwig’s homologies seem to me to
rest too much on the variable secondary relations of organs
as existing in the adult, and too little on the constant and
primary relations of their rudiments in the larva.

(2) Ludwig lays great stress on the homology of the oral
side in all Echinoderms, as shown in the following interest-
ing facts. When viewed from the oral side the gut always
winds from mouth to anus in the direction taken by the
hands of a watch, and the water-pore both of the Urchin and
of the embryo Crinoid is in the same interradius as the fore-
gut. He gives an excellent figure (Tab. XIII, fig. 7) of
tl)e course of the gut in an Urchin as seen from the ventral
side, and therefore (of course) represents the madreporite
as seen from within in the N.w. corner of the figure. He
gives a similar figure of the course of the gut in Antedon

ECHINODERM MORPHOLOGY. 327

(fig. 9), but without representing the homologous ” oral
plate with its water-pore. Had he done so he would have
been obliged to explain that while in the Urchin figure the
madreporite was represented as seen through the gut, in the
Crinoid figure the reverse would be the case, viz. the gut
seen through the perforated oral plate. But nevertheless
the latter is homologous with the genitals of an Urchin !

Again, the positions of the primary actinal and abactinal
plates, with respect to the oral centre, is not the same in
the Urchins and Crinoids if the latter be orientirt
according to Ludwig’s theory. In an Urchin the proximal
plates to the oral centre are the oculars (radial), and the
distal series the genitals (interradial).

In a Crinoid or Ophiurid the interradial orals are proximal
and the radials distal ; but, nevertheless, the homology of
the orals and genitals is undoubted, while the interradial
basals of the Crinoids, which are beyond the radials just as
the genitals of an Urchin are beyond the oculars, are not
represented in an Urchin ! Further, Ludwig takes no account
of the five interradial plates which surround the mouth of
PalcBostoma, and have the same relation to the water-vascular
ring, &c., as the orals of Rhizocrinus, a relation which, it is
almost needless to say, is not characteristic of the genitals of
an Urchin.

(3) Another of Ludwig’s arguments is founded upon the
peculiar relation of the aboral blood-vascular ring of the
Ophiurids to the mouth* shields. Because it dips down in
the five interradial spaces, and partially rests upon these
plates, Ludwig considers them homologous with the genitals
of the Urchins and Starfishes. But the rest of this blood-
vascular ring remains in the abactinal position that the
whole of it occupies in the Urchins and Starfishes, and it is
from this abactinal position that the principal vessels are
given off to the genital glands. According to Ludwig
(p. 352), “Dieses Genitalgefass entspricht sowohl in seinem
Ursprunge als auch in der Richtung seines Verlaufes
und in seiner Beziehung zu den Geschlechtsorganen den
Genitalgefass der Asterien.” On Ludwig’s theory, however,
one would expect the genital vessels to be given off from
those parts of the blood -vascular ring which lie nearest the
ventral surface, since it is in the plates of that ventral surface
that he recognises the homologues of the genital plates of
the Starfishes. One would also expect that these ventral
plates would be occasionally perforated by the openings of
the genital glands. But is this ever the case? Ludwig

VOL. XX. NEW SKll. Y

328

HERBERT P. CARPENTER

admits that it is not ; and the mere fact that in some Urchins
and Starfishes the genital ducts do not open on the genital
plates does not seem to me a sufficient argument to meet
this difficulty, though, according to hini, Indessen stort
das die Homologie, die nach Obigem zweifellos zwischen den
Mundschildern der Ophiuren und den Genitalplatten der
EcUnoideen und Asterien hesteht durchaus niclit ”(!!!)

Ludwig nowhere says a word about the genital plates of
the Ophiurids, and yet these are developed interradially on
the abactinal surface of the larva precisely like the genital
plates of the Starfishes. The following passage from
Agassiz Q North American Starfishes,’ p. 98) is interesting
as regards this question :

“ In Opliiurans the genital plates are formed from the angles of the five
interradial plates ; similar plates can still be traced in the young Starfishes,
while in the full-grown Starfishes their presence is shown by the interbra-
chial partition, on each side of which the ovaries discharge. Thus there
exists a complete homology between the genital plates of Opliiurans and
the interbrachial partitions of Starfishes, a homology fully carried out in its
details when we examine the relations held by the genital plates to the
ovaries in Ophiurans and by the interbrachial partitions to the ovarian
openings in Starfishes.”

Of course there are aberrant forms with more or less
exceptional peculiarities among the members of each class,
but the general relations of both classes are such as entirely
to support the statements quoted above, while the peculiar
features of the exceptions [Asterina gihhosa, Trichaster
elegans^ &c., as shown by Ludwig’s own observations) seem
to me to be of such a nature that they support Agassiz’s
view rather than Ludwig’s very revolutionary one.

The relations of the primary actinal and abactinal plates
to the blood-vascular system may be looked at from another
point of view than the one used hy Ludwig. A person
standing in the central plexus (heart) of a young Starfish,
with his feet in the oral blood-vascular ring, would have the
genital plates above his head. But if he were similarly
placed in a Crinoid (with the ventral side downwards) he
would have the basals over his head, and not the orals,
although these last are the homologues of the genitals of the
Starfish according to Ludwig’s theory.

Ludwig does not attempt to follow out to any great extent
the logical results of his theory with respect to the relations
of the various regions of the body among the different Echi-
iioderms. But he points out (p. 819) that they involve our
believing in a great difference between Urchins and

ECHINODBRM MORPHOLOGY.

329

Crinoids, one, namely, which has not been thought to exist
since the researches of Agassiz and Loven. While the
Crinoid calyx increases by the additions of plates on the
aboral side of the oral plates that of an Urchin grows very
differently :

Hier entfernen sich die Oralplatten der Crinoideen homologen Genital-
platteii immer weiter von dem Munde, indem die Bildung des Perisoras des
erwacbsenen Tbieres, genauer die Bildung der Interambulacralfelder, an der
adoralen seite der Genitalplatten erfolgt.’’

This must not be understood to mean (as it well might
do) that the genital plates have the same primitive relation
to the mouth of an Urchin as the orals have to the mouth of
a Crinoid and only subsequently become separated from it,
for this is far from being the case.

The old view that the genitals of an Urchin represent the
Crinoid basals involves none of the difficulties inseparable
from Ludwig’s new one. In both Crinoids and Urchins^
new plates are added on the adoral side of the interradial
ahactinal plates, viz. the genitals of the latter and the basals
of the former, which Ludwig, however, considers to be unre-
presented in the Urchins.

The only other out of the Reihe wichtiger Folgerungeii
fiir die vergleichend-anatomische Auffassung der Skeletre-
gionen der verschiedenen Echinodermgruppen” that Ludwig
deduces from the homologies which he asserts is the follow-
ing^ : — Dass das perianale P’eld der Echiniden dem gesamm-
ten Perisom der Ophiurensheibe mit Ausnahme der Arme
und der Mundschilder homolog ist.”

There seems to me to be something wrong here. Perhaps
it is only that Ludwig and I understand the expression

perianale Feld” in different ways. I take it to mean the
ring of genital and ocular plates, and, as Ludwig considers
the former to be homologous with the mouth-shields of the
Ophiurids, I am somewhat puzzled as to the meaning of the
above passage.

The above are many inconsistencies which become appa-
rent upon a critical examination of this new theory of
Echinoderm morphology. It is difficult to understand how
Ludwig can insist on the homology of the oral surfaces
of Urchins, Starfishes, and Ophiurids, and yet insist even
more strongly that plates around the mouth of the latter
are homologous with plates around the anus on the aboral

^ This Journal, New Series, vol. xix, pp. 2G — 28.

’ ‘ Zoologiseber Anzeiger,’ 1879, ii, p. 542.

S30

HERBERT P. CARPENTER.

surface of the two former groups. His theory reminds one
of Kowalevsky^s celebrated attempt to prove that the brain
of a worm represents the hinder end of the Vertebrate spinal
cord; and one cannot help feeling that, despite the un-
doubted value of his observations, some (though by no
means all) of the conclusions which he has based upon them
will not bear a close investigation when compared with
the primary facts of Echinoderm structure.

ORIGIN OF RED BLOOD-CORPUSCLES.

331

The Origin of the Red Blood-Corpuscles. By Professor
PoucHETj of the Jardin des Plantes^ Paris.^

I HAVE paid a good deal of attention lately (since November,
1877) to the study of the formed elements of the blood and
their origin. The fact was, as often happens, that several
histologists were pursuing, both in France and elsewhere, the
same investigation.

I propose to give a summary of what seems to me to be
the actual state of our knowledge on a subject of such
importance both to physiology and to medicine.

A sufficiently large part in these studies, which have been
followed out with vigour in various quarters, has been given
to the history of that singular phenomenon the coagu-
lation of the blood. I shall put this subject entirely on one
side, only wishing to occupy myself here with the formed
elements of the blood in the living state, with their origin
and the phases of their existence, in so far as they are con-
stituent elements of the organism. Coagulation is a post-
mortem phenomenon.

I shall speak in the first person, giving credit where
credit is due to others as well as to myself.

It is not a little peculiar that the formed elements of the
blood, which have been so long known, should be so little
known.

Even to-day we cannot say that we know with absolute
certainty how they originate (at least in the adult), how long
they live, and what may be their fate, for it is certain that
their existence is relatively short.

Every day one sees in the hospitals patients losing large
quantities of blood, and it seems as though no one had asked, or
at least had seriously considered and sought to discover, how
it is that, at the end of a few weeks, the blood lost is entirely
replaced.

G. E. Rindfleisch, in calculating the replacement of blood
in women during the intermenstrual period, estimated that
half a centigramme must be produced in a minute, which
means that about 175 millions of haematids are produced in
the body every minute. 2

How does this enormous proliferation come about? It is
not a little hard to be obliged to confess that we are reduced

* Translated from the ‘ llevue Scientifique.’

2 The term ‘ lioematid ’ is a rendering of the Frcncli term ‘ liematie,’ wliicli
M. Pouchet uses to denote red corpuscle.

332

PROFESSOR POUCHET.

to hypotheses. It is upon this obscure problem that I wish
to throw some light.

One need only look back over the last two years of the
morphological study of the blood^ to see how many organs
have been put forward, one after another, as the birthplaces
of the corpuscles, both white and red.

Theories and hypotheses have not been wanting ; the
lymphatic glands, the spleen, the marrow of the bones, and
other organs, have been looked upon as the seat of origin
of the formed elements of the blood, and even called, on that
account, " haematopoietic organs.” It must be allowed that
the question of haematogenesis is far from being a simple
one, and that among various animals it presents itself under
very different aspects.

We may recognise, in the first place, two great classes :

Firstly. — Oviparous animals, either warm- or cold-
blooded, in which the haematids possess a nucleus.

Secondly. — Mammals, in which the haematids, although
at first like those of oviparous animals, are replaced at a
very early stage by corpuscles, absolutely devoid of a
nucleus.

If one considers all animals with regard to their haema-
topoietic organs, it is evident that many animals have neither
marrow nor lymphatic glands, e.g. fishes.

In the same way there are some in which the spleen is
quite rudimentary (Syngnathus), or even entirely wanting
(Lampreys).

Among mammals, even the Rodents, in which the vascular
area persists and forms blood- corpuscles at a very late period ;
and, on the other hand, the Marsupials, in which the umbilical
vesicle has disappeared, even before any part of the primor-
dial skeleton has become vascular, offer various points which
call for special attention.

In order to solve the problem of hsematogenesis, we want
to obtain a large number of data which would be useful
to us.

By what certain characteristics can we recognise those
elements which are newly formed from those which are
about to degenerate? How long do they live ? For it is
impossible to admit the long life of elements, the regeneration
of which is so easy after lesions, and even in the normal state
of things in woman between puberty and the critical age.

The more or less satisfactory answers to all these ques-
tions, which have been made by recent research, I wish to

1 ‘ Gazette Miidicftle/ Nov. 10th, 1877 i Jan. 19tii, Feb. 2nd, March
10th, April 27th, 1878 i and ‘Journal dc i'Anatomie,* Jaw,, 1879.

ORIGIN OF RED BLOOD-CORPUSCLES.

333

present here in an order which, if not the most logical,
is probably the best for a rapid resume.

It is curious that most of those who have worked at
haematogenesis have been led to recognise from the first the
inaccuracy and incompleteness of the descriptions which
have been given, up to the present, of the formed elements
of the blood. One would have thought that objects so easily
examined would have been well known. Nothing of the
sort. Certain elements, of perhaps capital importance, have
been completely neglected, others badly described, while as
to the origin of all, one is reduced to the vaguest hypotheses.

In short, our first want seems to be to complete here, as
well for the oviparous vertebrates as for the mammals, such
descriptions as are found in the recognised text-books, and
even in a whole host of special memoirs on the subject,
which have been recently published.

Blood of Oviparous Animals. — Among all oviparous
animals — those which are warm-blooded — birds, as well as
those which are cold-blooded — reptiles, amphibians, and
fishes — the method of formation in the adult seems to be
the same. Among embryos it evidently differs according
as the embryo has or has not an umbilical vesicle.
In the adult the question of the formation of the elements of
the blood seemed actually settled. The size of the elements
in Amphibia naturally induces histologists to study these
first. I took the newt as the basis of my investigations.

There are in the blood of the newt certain elements not
corpuscles, which have hitherto all been included under the
general term leucocytes. These I have called, for the sake
of clearness, '^nuclei of origin;” however, by no means in-
sisting upon this term. These nuclei of origin” are
spherical, and about t^-o twit diameter.

They are enveloped by a very thin layer of protoplasm, of
equal thickness all round the nucleus. I will return pre-
sently to the origin of these elements ; it is their further
development which concerns me at present. This develop-
ment may take place in two different methods :

a. The nucleus of origin” enlarges and segments, and
the protoplasm becomes proportionally increased in bulk ;
in short, the nucleus becomes a leucocyte with numerous
nuclei. After this stage, when it may be called adult, it
must necessarily disappear; the protoplasm disintegrates
in the serum, setting free the nuclei, which are nothing else
than nuclei of origin about to recommence a new cycle.
This appears to be the normal state of things.

b. The nucleus of origin may be destined to become i\

334

PROFESSOR POUCHET.

hsematid, to this end undergoing a sort of abortion or
degeneration.

The protoplasm now commences to form hsemoglobin ; the
presence of the latter seems to be shown from the first by
the shape which the protoplasm assumes, which is already a
geometrical one.

The nucleus does not multiply; the element which has
thus undergone haemoglobic degeneration has reached by the
process its ultimate form, and is destined to disappear more
or less slowly. The nucleus, as it takes on an oval shape,
in accordance with the elongated shape of the protoplasm,
seems to lose that power of dividing and multiplying which
it originally possessed. The protoplasm increases and com-
mences to take on a yellow colour, visible with a microscope,
on account of a continued accumulation of haemoglobin.

Soon it has entirely lost its essential vital properties,
sensation, and power of movement. It becomes, with its
nucleus, a passive body like the cells of the horny layer of
the epidermis, retaining its chemical affinities, but devoting
itself entirely to this chemical function of haemoglobic de-
generation. Haemoglobin continues to accumulate in the
protoplasm, which thus becomes more and more dense and
coloured.

The nucleus loses its chemical characters and the proto-
plasm soon constitutes a homogeneous mass, which is finally
dissolved in the blood plasma ; it may be while still in the
circulation ; it may be, on the other hand, after having been
stopped on account of its diminished elasticity in the paren-
chymatous tissue of the spleen.

I believe I was the first to completely describe the suc-
cessive stages of this development, as well as a certain
number of points of less importance which go along with it,
but on which it is useless to dwell in this rapid resume,

M. Hayem has called the young corpuscles, the bodies of
which are already flattened, oval, and still appear colourless,
it may be, on account of the extent to which they are
magnified, haematoblasts.

It seems certain that these young corpuscles are at least
capable of changing their shape in the blood in a state of
rest, as do the nuclei of origin and the adult leucocytes.

M. Hayem admits the transformation of his haematoblasts
into corpuscles.

But M. Vulpian (‘ Comptes Rendus,’ 4th June, 1877)
certainly lias the credit of having demonstrated this develop-
ment ; he has shown by careful experiments tliatin the blood
of dead frogs luematoblasts always appear in large numbers.

ORIGIN OF RED BLOOD-CORPUSCLES.

335

and gradually become transformed into red corpuscles
(h^matids) ; he gives a very accurate description of these
hsematoblasts. Repeated experiments on birds have afforded
me precisely the same results. Neither M. Vulpian nor
M. Hayem have put forward any opinion as to the origin of
these haematoblasts. I have stated before that they are
derived from nuclei of origin. But whence do these latter
come ? Do they all come from the broken and scattered
nuclei of leucocytes which have finished their existence?

Fishes have no lymphatic glands. Some animals have
no spleen. Is it possible to admit that the nuclei of origin
originate, in this latter case, from the cells lining the w^alls
of the lymphatic cavities, and that in the other oviparous
animals the spleen, if not the exclusive place of their pro-
duction, at any rate plays an important part in it ? In fact,
I have shown that among Selachians the spleen is partly
formed of elements which are identical with these nuclei of
origin, and which, detaching themselves one by one, fall into
the network forming the tissue of the organ, and are carried
away by the blood. It must be admitted that the spleen
plays, among oviparous animals, a part which it plays, per-
haps, among the mammals, and which corresponds, at all
events, to that of the lymphatic glands of the latter group.

Lastly, it results from certain experiments which I per-
formed some time ago upon birds. Amphibia, and fishes, that
the removal of the spleen only hinders the reparation of the
blood after death, and that either the nuclei of origin are
continually originating from the walls of the lymphatics, or
that the leucocytes of the blood are sufficient for this
multiplication.

Leucocytes of Semmer. — When the blood of an oviparous
animal, and often when that of a mammal, is examined, it
is possible to find certain elements, the nature of which has
been for a long time misunderstood.

I thought I had been the first to describe them in the
blood of Selachians (Soc. de Biologie, Nov. 7). I have since
remembered that attention had already been drawn to them
in a special manner among mammals by a pupil of Alexander
Schmidt, in a paper written at Dorpat. I propose to call
them, after the author of this paper, leucocytes of
Semmer.

Even if their existence was known, their fundamental
structure had escaped all previous observers ; they may be
regarded as leucocytes (they are the same size, possess
amoeboid powers, &c.), in the body of which haemoglobin is
formed. The latter is in large granules, generally measuring

336

PROFESSOR POUCHET.

from -p'oV¥ to -yw-0 They possess all the physico-

chemical characters of the haematids.

We are now able to understand the full importance of
this discovery, which, perhaps, partially escaped M. Semmer.
It shows us that haemoglobin can exist in other anatomical
elements than in red corpuscles. M. Kiihne, before this,
thought that the colouring matter of the red muscles of ver-
tebrates might be haemoglobin, but he was not able to prove
it. Now, it is easy to prove by all the usual tests that the
granules in the leucocytes of Semmer are made of haemo-
globin. If MM. Alexander Schmidt and Semmer have
described them as an intermediate form between leucocytes
and red corpuscles, this must not be understood in the sense
that they represent a developmental stage through which
the leucocytes pass before developing into red corpuscles :
the leucocytes of Semmer are intermediate between the
white and red corpuscles only in the sense that they par-
take in a certain manner of the constitution of both.

Probably the leucocytes of Semmer undergo a final disin-
tegration, and the large granules becoming dispersed in the
blood plasma soon dissolve.

Their nuclei are always or almost always near the surface
of the cell (which is perhaps important with regard to their
special development), and never present any more than do
those of other white corpuscles — any trace of decay, and
probably become nuclei of origin.

Nothing has hitherto gone to prove that these, originat-
ing from a cellular body where haemoglobin is already
formed, can be specially considered to become haematoblasts.

First Red Corpuscles of the Embryo. — The formation of the
first red corpuscles in the vascular area of birds has been
studied by many embryologists, and still raises a certain
number of important questions on which there are two
opinions. Among oviparous holoblastic animals — the Am-
phibia— it appears certain that the embryonic cells are
directly transformed into red corpuscles, as they show those
vitellin grains and small pigment granules which are seen
in the corpuscles in their early life, and which only very
gradually disappear.

Blood of Mammals. — In spite of the numerous works
which have been lately published on the number of the cor-
puscles (Manassein Malassez, Hayem, Perier, &c.), it is evi-
dent that the descriptions given of the elements of the blood
are incomplete, and even inaccurate ; the characters of the
white corpuscles badly defined, the form of the red cor-
puscles insufficiently described, other elements altogether

ORIGIN OF RED BLOOD-CORPUSCLES, 337

neglected. It is^ therefore, necessary before going further
to go over these various points.

White Corpuscles. — I believe I was the first to describe
that the leucocytes of man and other mammals have
always, when completely developed, four nuclei, regularly
arranged, grouped in the centre of the protoplasm, and
devoid of nucleoli. This form represents the adult state.
There are always found besides these other smaller corpus-
cles, few in number, in the blood, in great abundance in the
lymph. These have a single nucleus furnished with nucle-
olus and surrounded by a much reduced layer of proto-
plasm. These corpuscles are true nuclei of origin, and
represent the young state of the quadri-nucleated cor-
puscles.

They originate in the lymphatic glands, and perhaps also
partly in the spleen (see below). What becomes of these
quadri-nucleated corpuscles ? The constant production of
corpuscles in the lymphatics, the absence, or at any rate the
invisibility, of any nucleolus in the four nuclei of the adult
corpuscle, leave little room for the hypothesis that these
nuclei outlive their protoplasm, and becoming free remake
nuclei of origin. It seems equally improbable that the
adult quadri-nucleated corpuscles leave the vessels, as has
been supposed, to become the differentiated elements of
more or fewer of the tissues. In any case, the first thing to
do to support this hypothesis, would be to search in the
developing tissues for these corpuscles, which are easily re-
cognisable on account of their four, regularly disposed, nuclei.

The leucocytes of Semmer are to be found among mam-
mals as among oviparous animals. They are particularly
abundant in the horse where M. Semmer has chiefly studied
them.

Hed Corpuscles. — I was the first to show that all the red
corpuscles of mammals have not necessarily — apart from
accidental injury — the discoidal shape, commonly described
and figured. Besides the classical red corpuscle, there are
always others ovoidal or even fusiform in shape. Their
long diameter is longer than the diameter of the normal
corpuscle, their edges are slightly turned up. Sometimes
these corpuscles are even more elongated, terminating it
may be in a point at each end, as in the rat.

It must be noted that this is in no way a question of
accidental stretching. I have directly proved its existence
in these hsematids in the blood in circulation in mammals.

AVe shall see directly the importance as regards hfemato-
geneeis of this form of hsematid, which has hitherto only been

338

PROFESSOR POUCHET.

noticed in the group of camels i (Camelus and Auchenia),
where it appears to be universal.

Glohulets. — There is in the blood a third kind of element,
which anatomists as well as medical men have almost com-
pletely ignored, and to which M. Hayem has lately suddenly
drawn attention. These elements were discovered by Donne in
1838, and it seems only right to preserve for them the name^
of the French microscopist who discovered them. In 1846,
a German doctor, Zimmerman, blames his fellow country-
men for forgetting in descriptions of the blood, the globulets
of Donne, and proposes to call them elementary corpuscles
(elementare Korperchen). Although described again by M.
Robin, they have remained almost altogether ignored, or else
confounded with the various amorphic granules always
to be found in the blood.

M. Hayem has given to these bodies the name haemato-
blasts, and has since described them in rather a summary
manner, omitting certain points which seem extremely
important. I believe I was the first to give an exact descrip-
tion of them.

These bodies are really chiefly remarkable for the precise
morphological character which distinguishes them from
simple granulations. Zimmerman was not after all so much
mistaken about this as Donne and others ; he called them
corpuscles.

They are elongated, and, except in the very smallest, one
diameter is much longer than the other. They are slightly
refractive, they appear homogeneous, they are devoid of
nucleus, and are unacted upon by staining reagents (?).
On the contrary, they approach in their physico-chemical
characters the substance of the body of leucocytes. This
analogy is even more striking.^

1 Upon what biological peculiarity does the form of the haematids in these
animals depend ? The problem appears to be hard to resolve in the present
state of science. The ancestors of this group of mammals appear to be
indigenous in regions of the globe, such as the Andes and the plain of
Central Asia, where the depression of the barometer is considerable. This
form of hsematid appears to us to indicate a special property of haemoglobin.

Among fishes, the Syngnathians have regularly nucleated discoidal
haematids like those of mammals, whilst other Lophobranchs have them
extremely elongated and fusiform.

G. E. llindfleisch has recently stated that the form of the haematids is
only the consequence of their rubbing one against the other in the serum ;
the context is sufficient to show what this singular hypothesis is worth.

2 The French word, used by Donne, was not ‘ globulet,’ but ‘globulin.’
— Tuansl.

3 This analogy would lead one to suppose that the globulets were derived
from the body of the leucocytes. One might bring forward to defend

ORIGIN OF RED BLOOD-CORPUSCLES. 339

They possess an extremely marked tendency to run together,
either with one another, with the leucocytes, or with the
hgematids, even in the blood in circulation when under ab-
normal conditions.

The proportional number of the globulets varies con-
siderably according to circumstances. They are exceedingly
easy to study in very young cats ; speaking generally, they
are not easily seen in animals whose blood is being repaired.

Le^igth of duration and fate of the heematids . — It is quite
clear that parts of our organism which are as much exposed to
accidental losses as are the elements of the blood, and even
to periodic losses which accompany certain functions, ought
to be always undergoing a regeneration, this regeneration
only becoming more or less active according to circum-
stances.

A consequence of this constant regeneration is that,
although one must consider some element in the body
as perennial, e, g, nerve cells, it is hardly possible to
think so of either hsematids or leucocytes. The elements
of the blood have therefore a short existence, how short we
know not; various things point to its being only a few
weeks, or at most a few months.

If mammal’s blood is transfused into bird’s blood, the
haematids are found unaltered at the end of fifteen or
twenty days.

Brown-Sequard found this to be the case, and I have
since repeated the experiments, using the dog and pigeon.
On the other hand, the haematid of birds placed in a mammal
are not to be found after a few^ hours only ; this may simply
be attributed to the much greater diameter of the haematid of
the bird. These latter seem to be almost immediately killed
and destroyed by the state of the new medium in which
they are placed.

If it is difficult to ascertain the absolute age of a haematid,
it seems easier to determine their relative age.

As a rule with mammal’s haematids as well as with those
of oviparous animals, the nearer they are to their period of
decline and disappearance the more coloured and refractive
are they. There is no doubt that the haematids end by
becoming dissolved in the serum. They diminish in volume,
and finally take a more or less regular spherical form. They
answer in this state to the description of elements found in

this hypothesis a more or less distant analogy, between this phenomenon
and the formation of the directive corpuscles ; a common character being
the regular segmentation of the nucleus of the leucocytes into four and the
regular segmentation of the vitellus, &c.

840

PROFESSOR POUCHET.

abundance in the blood in certain diseases by MM. Vanlair
and Masius, and which they called microcytes.

These old hsematids have the same general characters as
those of oviparous animals. They lose in great part their
elasticity, and seem to have on account of this a great ten-
dency to be caught, and remain in the reticulum of the
spleen where the blood coming from the arterial extremities
falls into this spongy network, out of which open the large
venous trunks. There is nothing to support the theory that an
active destruction of blood-corpuscles goes on in any special
organ.

Origin of Blood in Mammals. — In mammals, as well as in
birds, the first corpuscles are formed in the vascular area.
I have been able to follow this formation in the rabbit,
and to prove that haematids continue to be formed from the
old walls of the umbilical vesicle, even when the embryo is
as much as ram. long.

If the chorion be removed from the surface of the egg
the vascular area is exposed, this consists of a layer of
similar cells. I have assured myself that the primitive or
blastodermic cells differentiate in two ways, forming on the
one hand the endothelial cells of the vascular walls, and
on the other, the embryonic red corpuscles. The former
envelop the latter, which latter, grouped in caeca opening
on to the spaces already traversed by the blood, undergo a
more or less complete development before being in their turn
taken into the current.

Here then the formation of the first red corpuscles is
certainly not endogenous.

The blastodermic cells destined to become red corpuscles
ordinarily multiply to a greater or less extent by fission
before undergoing haemoglobic degeneration. The older
the embryo grows, the more this segmentation appears to
give place to cells of smaller diameter, and the more rapidly
does this degeneration of the cells proceed.

At first the cells which go to form the haematids are very
large, they pass into the circulating stream before the
nucleus has disappeared, or at least before it has ceased to
present the ordinary chemical characters of unclear sub-
stance, these cells become the large embryonic haematids.
Later on segmentation goes on more rapidly, and the result-
ing cells are even smaller and the nucleus rapidly atrophies.
This may go on in two ways, either the nucleus gradually
diminishing in volume loses its chemical characters, takes
on those of the protoplasm surrounding it (the final process
in the haematids of oviparous animals, as I described above).

ORIGIN OF RED BLOOD-CORPUSCLES. 34l

or else the nucleus undergoes a sort of breaking up, and its
substance becomes dispersed in that of the protoplasmic
body, so that it disappears altogether. This method, how-
ever singular it may appear, I have distinctly proved to be
the case. Such is certainly the origin of the first definite
hsematids comparable to those of the adult which succeed
the nucleated hsernatids.

As to the transformation, or rather the haemoglobic de-
generation of the protoplasmic body itself, it always goes on
in the same way, the protoplasm becomes more and more
homogeneous, more hyaline, more refractive. As long as
the newly-formed hsematid remains immotile its true form
remains obscured by the contact and pressure of the neigh-
bouring elements. The hsematid is not really biconcave and
discoidal either at the moment when it leaves or at the
moment when it enters the circulation.

Very soon after the rabbit^s embryo has attained a length
of 22 mm. the vascular area ceases to be the seat of origin of
the ha3matids. Lastly, it is interesting to note that during
the whole of the first period of intra-uterine life globulets
are not found in the blood, they only appear later on, and
in abundance only in certain animals.

Haematogenesis in the Adult. — Where are hsematids formed
in the adult ? This is the great question which is occupying
the attention at this moment of such a number of observers,
and which has already been thought to have been solved
over and over again.

Usually observers have tried to relegate this so-called
hjcmatopoietic function to certain organs or to certain tissues.
I must now call special attention to these. This function in
mammals has been successively attributed to lymphatic
glands, to spleen, to the marrow of the bones, the lymphoid
patches in the mesentery of the rabbit. There are yet the
supra-renal capsules and thymus, do they also deserve to be
studied from the same point of view ?

Lymphatic Glands. — No hsematids are found in the lymph
stream, the few which have been seen in fish are ex-
ceptional, and probably came into it accidentally. Lymph
taken from the thoracic duct of a dog, with proper precau-
tions, never contains haematids, although in operating care-
lessly on the horse the contents of the lymphatic vessel
appear of a reddish colour ; this is due to the large quantity
of haematids whose presence is caused by the careless oj)cra-
tion. In the lymphatic glands the blood system is a closed
system. Indeed, 1 have shown that it is not the same as the
spaces constituting the lacunar tissue which makes a com-

342

PROFESSOR POUCH Et.

muiiication between the afferent and efferent vessels. If at
certain places these channels are definitely closed in, they
sometimes open directly into the follicular substance ; in short,
there is really no valid distinction to be drawn between the
two tissues wliich are described as forming the Malpighian
corpuscles, the lacunar tissue, and the follicular tissue.
Therefore the following may be taken as a plan of a Mal-
pighian corpuscle. On the lymph paths proper (lacunar
tissue) are arranged certain caecal prolongations (follicular
tissue), which are closed only at the periphery, but open on
the other hand into the paths where they are continuous
with them, by lacunae, which are here large, but get narrower
and narrower towards the bottom of the caeca.

Beyond the point of insertion on the lymph path the
caecum is definitely bounded, and bounds in its turn the
lymph path, or, in other words, the regions of lacunar tissue.
In these caecal prolongations, as well as on the walls of the
trabeculae of the so-called lacunar tissue, certain cells pro-
liferate, and masses of cells are seen developing in them
comparable to the nuclei of origin of the leucocytes, and
which are evidently destined to fall into the lymph current,
and form the latter. But sometimes these same cells,
especially those in the lacunar tissue, undergo a different
development. The protoplasmic body of the cell becomes
rounded, and presents, along with other granulations of un-
known nature, four, five, or six large granules, sometimes
almost polyhedral, of a substance having all the characters
of hsemoglobin. These granules have been constantly taken
for hsematids in the course of formation, or for haematids
collected together in cells to which amoeboid movements
have been attributed, although no one has ever proved their
existence, forgetting that the ultimate origin of the haema-
tids thus observed in contact with the cells in question has
yet to be explained.

The interpretation of this phenomenon seems simple
enough. Haemoglobin is not a special product of haema-
tids ; it also occurs in the leucocytes of Semmer. The large
granules of haemoglobin found in the cells of the lymphatic
glands certainly do not originate elsewhere, and, moreover,
are not simply haematids absorbed by the cells of the no-
dules, just as they are not developing haematids. The
presence of these granules of haemoglobin always has the
effect, when they are abundant, of giving the tissue of the
Malpighian corpuscle a reddish tint, and even a decided red
colour, if the cells so altered are very numerous. We
shall see presently to what errors tliis modification of the

ORIGIN OF RED BLOOD-CORPUSCLES.

343

tissue of the nodules^ which has nothing whatever to do
with haematogenesis^ has given rise.

The Spleen. — The older observers, by experiments which
I have repeated, proved beyond a doubt that the spleen was
not essential in the regeneration of the blood after extensive
haemorrhage. I have already briefly described the struc-
ture of the tissue of the spleen. As T said, it is extremely
probable that a large number of the haematids are normally
arrested in it when they have lost their elasticity, which they
do as they grow old, on account of the large amount of
haemoglobin then present. These old haematids, thus re-
tained in the meshes of the splenic tissue, certainly help to
give it its colour, which it possesses even when there is very
little blood in it.^

The fact that the serum found in the splenic vein
appears to be more yellow than that found in other vessels
(G. E. Rindfleisch), might perhaps be explained by this
breaking up of the old haematids, which are arrested in the
parenchymatous tissue of the spleen.

Those who have attributed to the spleen a definite
haematogenetic function, ought, at the same time, to re-
member that haematogenesis is no less active in mammals
from which the spleen has been removed ; they have gone
further, and supposed that the spleen was helped by the
mesenteric glands, and, indeed, even by the subperitoneal
tissue ! At least, it ought to be proved in this case that the
glands and the cellular tissue have taken on the totally
(lifierent histological structure of the spleen-parenchyma ;
and the stages of such a wonderful transformation, which
would be one of the most eurious known to anatomists,
ought to be pointed out ! On the other hand, it would be a
senseless physiology which ascribed to two organs, of such
essentially different structure and texture as these, the
same function. It is really astonishing that biologists
have accepted this singular idea of a vicarious action,
as it has been called, on the part of certain organs which
are supposed to take on, for the time being, the functions

^ Probably the Malpighian corpuscles ought not to be looked upon as
special structures, but simply as regions where the splenic tissue, more or
less accidentally, has become impermeable to the blood which passes through
the organ. In teleostian fishes these impermeable regions are not isolated,
but form a thick net-work in the organ ; moreover, in Amphibia there is a
development of certain cells of the spleen, which recalls (at least, so far as
one can judge without having specially studied it) the development of those
cells, which 1 shall describe further on, in the marrow of the bones of
mammals.

VOL. XX. NEW SER,

Z

34i

PROFESSOR POUCHET,

of Other organs which have not the same anatomical
structure.

The Marrow of the Bones. — Of all the questions which
touch upon hsematogenesis, perhaps the most delicate and,
moreover, the most difficult to settle, is the part which is
played by the marrow of the bones. Just now this haemato-
poietic function, which has been attributed to so many
organs one after anotlier, is ascribed to the marrow of the
bones ; and it must be admitted there are more or less valid
reasons for it. In the first place, in all mammals, with-
out exception, the marrow of the bones preserves the charac-
ters which it has in the foetus, that is to say, the marrow
remains red, particularly in the bodies of the vertebrae.^

Even in mammals which have a large quantity of fat,
e. g. Cetacea, I have ascertained that the marrow of the ver-
tebrae and the spongy substance of the large bones of the
limbs is red ; lastly, the red marrow cannot be removed from
an animal in the way the spleen can, and the part it plays
in haematogenesis experimentally judged.

In 1868 there was a dispute between MM. Neumann and
Bizzozero, as to who had discovered in the red marrow of
animals anatomical elements, the protoplasmic body of which
presented the same characters as the substance of the haema-
tids, but which at the same time possessed a nucleus.
Stated thus, this fact, as pointed out by MM. Neumann and
Bizzozero, is perfectly correct, but MM. Neumann and Biz-
zozero, each on their own side went further, and' concluded
that the red marrow was essentially haematopoietic, and
that the elements that they pointed out were none other than
young haematids in the course of development. This in-
terpretation may he true, but it is not yet proved. To accept
this hypothesis we must take it for granted that haema-
tids are formed by the atrophy or haemoglobic degeneration
of the true cells of the marrow, and that they fall, after their
nucleus has completely disappeared, into the blood-stream,
just as the nuclei of origin fall (as I have shown above) into
the lymph stream.

The first question that then suggests itself is — have the
capillaries of the red marrow any walls ? M. Hoyer,
and more recently M. G. E. Rindfleisch, have declared
that the medullary capillaries possess a wall. Moreover,
M. Rustizky showed, in 1872, and I have since proved,
that the medullary capillaries are certainly coated with endo-
thelial cells.

’ Excepting in the last caudal vertcbrse, where on the contrary it is ex-
tremely full of fat.

ORIGIN OF RED BLOOD-CORPUSCLES.

345

The first point we wished to know is then well estab-
lished.

It is also certain that many of the marrow cells undergo a
haemoglobic degeneration altogether comparable to that
undergone by the haematids in circulation in the blood of
oviparous animals. The protoplasmic body, at first colour-
less and finely granular, soon becomes hyaline, coloured, and
refractive. At the same time the nucleus gradually loses its
chemical characters and finally disappears. The peculiar
red colour of the marrow is due to the abundance of these
elements.

The question so often discussed as to the essential identity
of the marrow elements (medullo-celles of M. Ch. Robin)
and the leucocytes is here beside the mark. Such identity
is improbable, and in any case the medullo-celles ” never
present the characteristic four nuclei found in the leucocytes
of the lymphatics. We might here call to mind the words
of the German author of one of the last works on the sub-
ject, The term white corpuscle has become a sort of omni-
bus into which everything is thrust.”

It is now well known that the medullary elements undergo
haemoglobic degeneration along with a disappearance of the
nucleus.

In the less modified elements independent masses of
haemoglobin appear almost as large as haematids, but pushed
out of shape by the neighbouring elements.

It has been proved that before this degree of degeneration
is attained the masses of haemoglobin contain a nucleus
which disappears by gradual assimilation into the proto-
plasmic body as in the vascular area in Rodents, and not by
going out of the cell, as has recently been stated (G. E.
Rindfleisch). The question arises do these masses of haemo-
globin, which have been very properly called medullary
haematids,” and which are moreover comparable in all respects
to the haematids of birds, end their retrograde development
by becoming dissolved, or do they fall into the blood stream ?

It is impossible to suppose that these elements can exhibit
spontaneous movements, so as to admit of their displacing
themselves, or becoming detached from the walls of the
capillary, and finally passing through the endothelial cells,
or pushing in between them by a sort of reversed diapedesis.
One of the essential characters of haemoglobic degeneration
is the almost immediate loss of all power of amoeboid move-
ment in the protoplasm.

On the other hand, can we look to external forces to
accomplish for the haematid this migration ? No ! for the

346

PROFESSOR POUCHET.

marrow is quite immovable and specially fixed in the solid
substance of the bone.

Another question arises, if this degeneration were to go on '
at the same time in a number of elements bordering upon a
capillary, the wall which, as we have seen, is formed of a
single layer of endothelial cells, would break through sooner .
or later and allow the blood stream to take in these new
h^ematids while they were yet undeveloped, while a new
endothelial wall would have to form to cover up the space
where they had been set free. In other words, would not the
capillaries of bone be continually undergoing development
or rather displacement in the medullary tissue ? Nothing in
such observations relating to this, as I for my part have
been able to make, furnish any indication of anything of
the kind taking place, or that the medullary cells while
undergoing development come into any special relation with
the capillaries.

Certain anatomists have thought that the marrow of the
bones was modified after great hsemorrhage, and when the
blood is being regenerated. All the experiments — which,
however, have not been many — that I have made in this
matter, have not shown this to be the case.

It must then be admitted, and T think we can go no
further, that the marrow cells undergo in situ a haemoglobic
degeneration comparable to that taking place in the hsema-
tids of birds, and which also takes place in the marrow of
the bones of birds before it has given place to the air cavi-
ties. In fact, marrow exists in the bones of reptiles and
Amphibia, as well as in young birds, and yet no one has
dreamt of ascribing any haematopoietic function to it.
Fishes have no marrow in their bones.

To sum up : — the development of haematids among ovi-
parous animals and that of the marrow cells are two processes
which are in all points comparable to one another, just as
the appearance of granules of haemoglobin in the lymphatic
gland is comparable to what happens in the leucocytes of
Semmer.

It has been attempted to found an argument in favour of
the haematopoietic function of the marrow on the existence,
which has been proved from time to time, of cells in the blood,
with a nucleus and a body containing haemoglobin-cells con-
sequently analogous to the haematids of birds, only often
without a regular form.

It will be sufficient to draw attention to the extraordinary
rarity of these elements which one hardly finds in one out
of a liundred preparations. They arc not moreover more

ORIGIN OF RED BLOOD-CORPUSCLES.

347

abundant in blood which is undergoing repair than in
normal blood. Certainly it may be that they are marrow
cells which have accidentally fallen into the blood current^,
but it is, perhaps, more logical to look upon them as leu-
cocytes which have accidentally undergone a hsemoglobic
degeneration after the fashion of marrow cells or the hsema-
tids of oviparous animals. In any case the extreme rarity
of these elements takes away from them any value in the
solution of the problem of haematogenesis.

Lymphoid Latches in the Ltabhit. — MM. Ranvier and
Hayem have recently held that hsematids being devoid of
nucleus were necessarily endogenous cellular productions.

M. Ranvier has brought to bear upon this subject his
observations on the lymphoid patches in the mesentery of
the rabbit where he has thought he has seen and has
figured hsematids originating in the midst of angioplastic
cells destined to become the walls of the vessels. But
admitting the perfect accuracy of the observations of such a
skilful anatomist, it must be allowed it would be very
difficult to ascribe to them such a character. In this
case it would follow that the constant repair of the
blood is necessarily bound up with the productions of new
capillaries, and that the restoration of the blood after great
hemorrhage is accompanied by a considerable extension of
the capillary system ! It has not yet been shown that this
is the case.i

Blood undergoing Repair, — It only remains to examine the
conditions which blood presents among mammals while under-
going repair after great haemorrhage. Observing the blood
under these conditions one is struck with the extraordinary
abundance of globulets, and above all, by the number of
intermediate forms between the globulets properly so called
and the elongated oval hsematids which I described above ;
and no doubt these are all the same, passing from one form
to the other.

They are very easily seen in the dog. I have also observed
them in the rat. It is only necessary to subject these animals
to copious and frequent bleeding The globulets of Donne
(haematoblasts of M. Hayem) appear to me, as Zimmerman
suspected, and as M. Hayem admitted, to be the true origin
of the hsematids of mammals. They represent the young
state just as the microcytes of MM. Vanlair and Masius
represent the old and decayed state.

' A tiling which [circumstances only very rarely allow one to study, and
which would be of considerable interest in this connection, is the mode, of
repair of the uterine mucous coat during an intermcnstrual period.

348

PROFESSOR POUCHET.

The globulet, which after its appearance in the serum is
distinctly elongated, enlarges in all directions ; its substance,
which at first perhaps appeared finely granular, becomes per-
fectly hyaline and refractive. It passes into the state of an oval
elongated hsematid, the long diameter of which is larger
than the diameter of the discoidal hsematids which have a
slightly pronounced and thickened margin. The normal
discoidal hsematid thus represents a more advanced stage in
the development of the cell, and is, in fact, its adult condi-
tion. To this succeeds a period of degeneration, during which
the hsematid becomes irregularly spherical and darker in
colour, before disappearing altogether.

Origin of Glohulets, — How do the globulets originate?
Are they, as M. Hayem wished to show, an endogenous pro-
duct of certain cells which are at present unknown ? Or,
on the contrary, must we recognise in these elements direct
products, organic concretions, of a special kind originating
in the blood plasma itself? As a matter of fact their mor-
phological characters, which are well known, do not allow
of their being looked upon as amorphous deposits of albu-
minous matter. They have evidently a definite structure, and
on this account are entitled to be called anatomical elements
as well as the crystals of otoconia ” or the laminated fibres
of connective tissue, if it is to be admitted that these latter
form independently of the body of the cells which form the
fibres.

The globulet exhibits the peculiarity common to it
and to many cells of fixing or forming haemoglobin.
The gradual accumulation of the latter explains the growth
of the corpuscle. The proportion and character of the
haemoglobin regulates the form of the corpuscle which
at first is oval and afterwards discoidal. When there is a
large amount of haemoglobin present further growth stops.
Thus the amount of haemoglobin determines the limit of
growth. This limit is, on the other hand, in direct relation
with the minimum diameter of the vessels through which
the haematid has to pass. I am well aware that if the evolu-
tion of the haematids is such as I have indicated, it goes
against certain well-known facts of general anatomy. This,
however, is not always a sufficient reason for rejecting a
theory, which if it be true, would be, on the contrary, an
explanation of what has been for a long time misunderstood.
According to my hypothesis, the hsematids of adult mam-
mals would not be cells, would not even be derived from cells.

I may here remark that haemoglobin should be regarded
as a product of a cellular organism and not as an integral

oaiGIN OF RED BLOOD-CORPUSCLES. 349

part of the same. In this respect heemoglobin behaves
like fat, or like vitellin granules, or starch, &c. It might
be even more exactly compared to chlorophyll, which some-
times appears in the midst of the cellular substance, and
sometimes in solution in the body of the cell itself.^

Sometimes, in fact, the haemoglobin appears stored up in
the cell (leucocytes of Semmer, cells of lymphatic glands),
and at other times diffused throughout the whole body of
the cell, only in proportion as the substitution is more and
more complete the cell body loses more and more its vital
properties, properly so called ; it becomes inactive, and the
nucleus disappears. Soon the whole is nothing but the
wreck of a cell.

It appears, at the same time, that as more and more
haemoglobin is formed this living matter cannot continue to
exist and disappears, and, as a consequence, the cell, which
is reduced merely to the haemoglobin which is deposited in
it, is able to dissolve in the plasma, and so disappear.

All this tends to show that haemoglobin is a very
secondary product of the organism, and it would not be
astonishing if this body, which results from extraordinary
complex chemical actions going on in the blood plasma,
were sometimes to form and be deposited in cells, properly
so called.

As to the first appearance of the globulets in the midst of
the blood plasma, it is not really more surprising than that
of the fibres which appear in the same plasma when drawn
from the vessels — fibres which, it must be remembered, have
a definite morphological character.

To the globulet thus formed of albuminous substance
(globulin) there is soon added a crystalline substance, or at
least a substance having certain characters which belong to
crystalline substances (haemoglobin), and it is to this that
the increase in volume is due. Such an increase of volume
is, therefore, not a case of development in its technical sense,
but simply a growth comparable to that of a group of bodies
similarly formed by the union of albuminous compounds and
cystalline compounds, which M. Halting and others have
studied so carefully.

To sum up, it must be admitted that the origin of the
haematids among adult mammals has not as yet been com-
pletely made out. Anatomists are divided between two

* “ It sometimes happens (among algsc) that the whole protoplasmic mass
is coloured green excepting the most internal layer, the membranous layer,
and certain isolated spots (many Zoosporae, Palmellacece, gonidia of lichens.’*
Sachs.

350

PROFESSOR POUCHET.

principal theories. Some^ with MM. Neumann and Biz-
zozero, distinctly attribute to the red marrow this function
in the economy — the production of the hmmatids and the
provision for the normal regeneration of the elements of the
blood after accidental losses.

I think, on the contrary, with M. Hayem, that the
haematids originate from the globulets of Donne ; only we
differ as to the origin of these globulets.

Whilst M. Hayem looks upon them as the endogenous
product of cells which are as yet unknown, it seems to me
that everything goes to prove that they are formed in the
blood plasma whilst in circulation, in a way which is more
or less analogous to the formation of the filaments of fibrin
in blood which has been drawn from the vessels.

LIMNOCODIUM (e RASPEDACUSTES) SOWERBII.

351

On Limnocodium [Grasp edacustes) Soweebii, a new Tracho-
MEDUSA inhahiting Eresh Water. By E. Bay LA^fKESTEE^
M.A., E.E.S. With Plates XXX and XXXI.

At the meeting of the Eoyal Society on June 17th, and in
^ Nature^ of that date, I described in the following terms a re-
markable Medusa which I had received a few days before :

On Thursday last, June 10th, Mr. Sowerby, the Secretary
of the Botanical Society of London, observed in the tank in the
water-lily house in BegenPs Park some peculiar organisms, of
which he was kind enough to place a large number at my dis-
posal on the following Monday.

The organism proves to be an adult Medusa, belonging to
the order Trachomedusae, and the family Petasidae, of HaeckeFs
system. It comes nearest among described genera to Eritz
Miiller^s imperfectly known Aglauropsis, from the coast of
Brazil.

** The most obviously interesting point about the form under
notice is that it occurs in great abundance in perfectly fresh
water, at a temperature of 90° E. Hitherto no Medusa of any
order has been detected in fresh water.^

'' It is exceedingly difficult to trace the introduction of this
animal into the tank in the BegenPs Park, since no plants have
been recently added to the lily-house, and the water is run off
every year. Probably a few specimens were last year, or the
year before, present in the tank, and have only this year multi-
plied in sufficient abundance to attract attention. Clearly the
Medusa is a tropical species, since it flourishes in water of this
high temperature (90° E.)

" Mr. Sowerby has observed the Medusa feeding on Daphnia,
which abounds in the water with it.

The present form will have to be placed in a new genus, for
which I propose the name Craspedacustes.^

It presents the common characters of the Trachomedusm (as
distinguished from the Narcomedusae) in having its genital sacs
or gonads placed in the course of the radial canals. It agrees
with both Narcomedusae and Trachomedusae in having endoder-
mal otocysts — and further presents the solid tentacles with car-
tilaginoid axes, the centripetal travelling of the tentacles, the
tentacle-rivets (Mantelspangen), the thickened marginal ring to
the disc (Nesselring) observed in many Trachomedusae.

* Except, perhaps, some estuarine forms (Crambcssa and others).

^ Kpmo-TTtSaKovoTTjc, One wlio liears by means of the KpacnnZoVf or
velum.

362

PROFESSOR E. RAY LANKESTER.

Amongst Trachomedusae, Craspedacustes finds its place in
the Petasidse, which are characterised as ^ Trachomedusse with
four radial canals, in the course of which the four gonads lie ;
with a long tubular stomach and no stomach stalk/

Amongst Petasidse it is remarkable for the great number of
its tentacles, which are all solid, and for its very numerous oto-
cysts. Further it is remarkable amongst all Hydromedusae
(velate Medusae exclusive of Charybdaea) for the fact that centri-
fugal radiating canals pass from the otolithic concretions which
they enclose into the velum where they end cacally,

” The genus may be characterised as follows :

“ Mouth quadrifid, with four per-radial lobes.

Stomach long and tubular, projecting below the disc.

'^Disc, flattened.

Eadial canals, four terminating in the marginal canal.

“Marginal (Eing) canal, voluminous.

“ Centripetal canals- (such as those of Olindias, Geryonia,
&c.) absent.

“ Tentacles, solid ; in three sets, the horizons of the insertion
of which are superimposed : —

“ (1.) A highest (nearest the umbrella-pole) set of four large
per-radial tentacles — primary tentacles.

“(2.) A second tier of twenty-eighty or more medium*sized
tentacles placed between these in four groups of seven — secon-
dary tentacles.

“ (3.) A third tier of one hundred and ninety-two or more
small tentacles placed in groups of six between the last — tertiary
tentacles.

“Tentacle-rivets (Mantelspangen) connecting the roots of
the tentacles with the marginal ring (Nesselring) are connected
with all the tentacles.

“ Otocysts placed along the line of insertion of the velum :
about eighty in number, from sixteen to twenty between each
pair of perradial tentacles, arranged in groups of two or three
between the successive secondary tentacles.

“Velar centrifugal canals are present, passing from the
otocysts — one from each otocyst — into the velum, and there
ending blindly. They appear to correspond in character to the
centripetal canals found in other Trachomedusse in the disc.
Their presence constitutes the chief peculiarity of the genus
Craspedacustes, and may necessitate the formation of a distinct
family or suborder for its reception.

“ (The minute characters of the otocysts, and their relation to
these canals, is now in course of investigation.)

“ Ocelli are wanting.

LIMNOCODIUM (CRASPEDACUSTES) SOWERBII, o53

“ Gonads. Four oval sacs depending into the cavity of the
subumbrella from the four radial canals.

“The above characters are derived from the examination of
adult male specimens, which were freely discharging ripe,
actively motile spermatozoa.

“ Species. — C, Sowerbii.

“I name the species in honour of Mr. Sowerby, who dis-
covered it, and to whose observation and courtesy zoologists are
much indebted. The sole character, which I can give as specific
over and above the generic characters summarised above, is that
of size. The diameter of the disc does not exceed half an inch.

“ Locality. — The water-lily tank in the gardens of the Bo-
tanical Society, Regent’s Park, London. Very abundant during
June, 1880. Probably introduced from the West Indies.”

On June 17th, the same day as that on which the above
appeared in ‘ Nature,’ the same organism was described by Prof.
Allman to the Linnsean Society ; and in the following week
(June 24th) an article was published by Prof. Allman in
‘ Nature,’ which was an extension of the paper read by him to
the Linnaean Society, including, “ some facts, in addition to
those contained in his original paper.” Professor Allman’s
observations agree very closely with my own reproduced above,
though we differ on one or two important points, and especially
on the relationship of the new Medusa to Trachomedusae, on the
one hand, and liptomedusse on the other, with which latter
group Professor Allman thinks it has most affinity.

Inasmuch as Professor Allman gave his account of the new
Medusa at the Linnsean Society on the same day as that on
which mine was published in ^ Nature,’ and on account of the
regard which all zoologists must feel for him, I propose as a
mark of respect to him to accept the generic name Limnocodium
which he bestowed upon the new Medusa in place of Craspeda-
custes which has the right of priority. At the same time I shall
maintain the specific title ' Sowerbii,’ in recognition of Mr.
Sowerby’s discovery and his valuable observations on the habits
and on the embryonic condition of this Medusa.

Ill the following pages I propose to describe a few important
facts relative to the structure of Limnocodium, which in my
judgment suffice to establish its position as one of the Tracho-
medusae, though it is quite a peculiar form and very possibly is
either the isolated representative of an archaic type of that order,
or has degenerated in connection with its exceptional life-condi-
tioni, namely, those of fresh water.

The facts to which I shall on the present occasion draw atten-
tion, relate to the structure of the marginal ring of the disc, the
insertion of the tentacles and the structure and development of

354

PROFESSOR E. RAY LANKESTER.

the marginal bodies, and their capsules (referred to above as
otocysts^^ and velar canals I am also able to give some
conclusive observations on the embryonic condition of Limnoco-
dium. I have not, as I have the intention of doing later,
studied this Medusa by means of sections, but in the living state
and with the aid of that invaluable reagent, osmic acid.

The account of the structure of this Medusa, published by
Professor Allman, is as follows :

The umbrella varies much in form with its state of contrac-
tion, passing from a somewhat conical shape with depressed
summit through figures more or less hemispherical to that of a
shallow cup or even of a nearly flat disc. Its outer surface is
covered by an epithelium composed of flattened hexagonal cells
with distinct and brilliant nucleus. The manubrium is large ;
it commences with a quadrate base, and when extended projects
beyond the margin of the umbrella. The mouth is destitute of
tentacles, but is divided into four lips, which are everted and
plicated. The endoderm of the manubrium is thrown into four
strongly-marked longitudinal plicated ridges.

The radial canals are four in number; they originate each in
an angle of the quadrate base of the manubrium, and open dis-
tally into a wide circular canal. Each radial canal is accom-
panied by longitudinal muscular fibres, which spread out on
each side at the junction of the radial with the circular canal.

The velum is of moderate width, and the extreme margin of
the umbrella is thickened and festooned, and loaded with
brownish-yellow pigment cells.

^^The attachment of the tentacles is peculiar. Instead of
being free continuations of the umbrella margin, they are given
off from the outer surface of the umbrella at points a little
above the margin. Prom each of these points, however, a ridge
may be traced centrifugally as far as the thickened umbrella
margin ; this is caused by the proximate portion of the tentacle
being here adnate to the outer surface of the umbrella. It
holds exactly the position of the mantelspangen or peronia,
so well developed in the whole of the Narcomedusse of Haeckel,
and occurring also in some genera of his Trachomedusse. Its
structure, however, differs from that of the true peronia, which
are merely lines of thread-cells marking the path travelled over
by the tentacle as the insertion of this moved in the course of
metamorphosis from the margin of the umbrella to a point at
some distance above it, while in Limnocodium the ridges are
direct continuations of the tentacles whose structure they retain.
They become narrower as they approach the margin.

^^The number of the tentacles is very large in adult speci-
mens. The four tentacles which correspond to the directions of

LIMNOCODIUM (cRASPEDACUSTES) SOWERBII. 355

the four radial canals or the perradial tentacles are the longest
and thickest. The quadrant which intervenes between every
two of these carries, at nearly the same height above the margin,
about thirteen shorter and thinner tentacles, while between
every two of these three to five much smaller tentacles are given
off from points nearer to the margin, and at two or three levels,
but without any absolute regularity ; indeed, in the older ex-
amples all regularity, except in the primary or perradial tenta-
tacles, seems lost, and the law of their sequence ceases to be
apparent.

I could find no indication of a cavity in the tentacles : but
they do not present the peculiar cylindrical chorda-like endodermal
axis formed by a series of large, clear, thick-walled cells which
is so characteristic of the solid tentacles in the Trachomedusse
and Narcomedusae. Erom the solid tentacles of these orders they
differ also in their great extensibility, the four perradial tentacles
admitting of extension in the form of long, greatly attenuated
filaments to many times the height of the vertical axis of the
umbrella, even when this height is at its maximum ; and being
again capable of assuming by contraction the form of short thick
clubs. Indeed, instead of presenting the comparatively rigid and
imperfectly contractile character which prevails among the Tra-
chomedusse and the Narcomedusae, they possess as great a power
of extension and contraction as may be found in the tentacles of
many Leptomedusae (Thaumantidae, &c.) . These four perradiate
tentacles contract independently of the others, and seem to form
a different system. All the tentacles are armed along their
length with minute thread- cells, which are set in close, somewhat
spirally-arranged warts.

“ The lithocysts or marginal vesicles are, in adult specimens,
about 128 in number. They are situated near the umbrellar
margin of the velum, between the bases of the tentacles, and are
grouped somewhat irregularly, so that their number has no close
relation with that of the tentacles. They consist of a highly
refringeiit spherical body, on which may be usually seen one or
more small nucleus-like corpuscles, the whole surrounded by a
delicate transparent and structureless capsule. This capsule is
very remarkable, for instead of presenting the usual spherical
form, it is of an elongated pyriform shape. In its larger end is
lodged the spherical refringeiit body, and thence becomes
attenuated, forming a long tubular tail-like extension, which is
continued into the velum, in which it runs transversely towards
its free margin, and there, after usually bocoming more or less
convoluted, terminates in a blind extremity.

The marginal nerve-ring can be traced running round the

356

PROFESSOR E. RAY LANKESTER.

whole margin of the umbrella, and in close relation with the
otolitic cells. Ocelli are not present.

The generative sacs are borne on the radiating canals, into
which they open at a short distance beyond the exit of these from
the base of the manubrium. They are of an oval form, and from
their point of attachment to the radial canal hang down free into
the cavity of the umbrella. Some of the specimens examined
contained nearly mature ova, which, under compression, were
forced from the sac through the radial canal into the cavity of
the stomach.

While some of the characters described above point to an
affinity with both the Trachomedusse and Narcomedusse, this
affinity ceases to show itself in the very important morphological
element afforded by the marginal bodies. In both Trachomedusse
and Narcomedusse the marginal bodies belong to the tentacular
system ; they are metamorphosed tentacles, and their otolite
cells are endodermal, while in the Leptomedusse, the only other
order of craspedotal Medusae in which marginal vesicles occur,
these bodies are genetically derived from the velum. Now, in
Limnocodium the marginal vesicles seem to be as truly velar as
in the Leptomedusae. They occur on the lower or abumbral side
of the velum, close to its insertion into the umbrella, and the
tubular extension of their capsule runs along this side to the free
margin of the velum, while the delicate epithelium of the
abumbral side passes over them as in the Leptomedusae. It is
true that this point cannot be regarded as settled until an oppor-
tunity of tracing the development is afforded ; but in very young
specimens which I examined I found nothing opposed to the view
that the marginal vesicles were derived, like those of the Lepto-
medusae, from the velum.

If this be the case Limnocodium will hold a position inter-
mediate between the Leptomedusae and the Trachomedusae ; but
as the greatest systematic importance must be attached to the
structure and origin of the marginal vesicles, its affinity with the
Leptomedusae must be regarded as the closer of the two.”

It will be observed that, in opposition to the results which I
had published in the previous week. Professor Allman holds
that the structures which I had identified with the peronia
(tentacle-rivets or Mantelspangen) of Trachyline Medusae are
not true peronia ; ” further, that the tentacles, though solid, do
not present the chorda-like endodermal axis ” characteristic of
Trachomedusae and Narcomedusae, and lastly, that the otolite
cells are not endodermal, as in Trachyline Medusae, but are de-
rived from the ectoderm of the velum, as in Leptomedusae, and
on this ground especially he holds that the affinity of the new

357

LIMNOCODIUM (CRASFEDACUSTES) SOWERBII.

form must be regarded as closer with the Leptomedusae than
with Trachomedusae.

Fig. 1. — Limnocodium Sowerbii, as seen floating. Magnified five diameters.
MR., marginal ring ; Ve, velum, witli tubular otocysts ; PTy per-
radial tentacle.

In the present communication I shall apply myself to these

Fig. 2. — Limnocodium Sowerbii, with retroverted disc. MR, marginal
ring ; MC, marginal canal ; Rc, radial canal ; Si, stomach ; Go, gonad ;
Pe, velum. Magnified five diameters.

358

PROFESSOR E. RAY LANKESTER.

points, and first of all discuss the structure of the peronia and
tentacles.

Naked-eye appearances. — When the Medusa is seen swim-
ming in the water, the observer is at once struck by the manner
in which the tentacles are carried. Though they can be de-
pressed and directed towards the subumbrella surface, they
are, as a rule, carried nearly at right angles to the horizontal
plane of the disc and directed upwards. This carriage of the
tentacles is more particularly noticeable when the Medusa is at
rest after a certain number of pulsations, but is also seen be-
tween each one of a series of pulsations. An opaque crenated
line along the margin of the disc (fig. 1, 3/i2.) is very readily
observed, even without any magnifying glass. This line is the
marginal ring, consisting externally of ectoderm cells charged
with thread-cells, and more deeply strengthened and supported
by the thickened wall of the marginal or ring canal, the endo-
derm cells of which are converted into a cartilaginoid tissue,
having a yellowish-green tint when seen by transmitted light
with the microscope. These yellowish cartilaginoid cells have
led Professor Allman to speak of the margin of the disc as being

loaded with brownish-yellow pigment-cells."”

The layer of thread-cells along the margin of the disc, although
not bulky in this small form, yet undoubtedly corresponds to the
^^nettle-ring” (Xesselring) of the typical Trachyline Medusae,
whilst the subjacent cartilaginoid tissue corresponds to the ‘‘'ring-
cartilage ” (Eingknorpel) of the same group.

The crenations or centripetal projections of the opaque mar-
ginal ring are in relation to the insertions of the tentacles.
They have the same superficial structure as the superficies of the
ring itself, and, so far as any definition can be given to the term
“ peronium,” the centripetal prolongations of the nettle-ring
passing from that structure to the point where a tentacle springs
from the disc, are entitled to the name. Below the peronia we
find in each case an extension of the peculiarly modified endo-
dermal cartilaginoid tissue of the ring-canal, which passes with
rapid gradation into the cartilaginoid tissue of the tentacle-
root.

The power of contraction and expansion possessed by the ten-
tacles is no doubt, as Professor Allman has pointed out, greater
than that possessed by the most highly differentiated of the
Trachyline Medusae. At the same time the stiff erect carriage
to which I have alluded, is quite unlike anything seen in Lepto-
medusae, and as I shall point out below the solid axis of the
tentacles, though not so firm a tissue as the corresponding tissue
of Cunina, is very appropriately described by the term “ carlila-
ginoid,” and more distinctly resembles “notochordal tissue”

LlMNOCODIUM (CRASPEDACUSTES) SOWERS II. 359

than does the single row of thickened cells which is found in the
larger and typical Trachyline Medusie.

The woodcut (fig. 2) represents the Medusa in an attitude
which it not unfrequently assumes when at the bottom of the
vessel in which it is confined. When this retroversion of the
disc is effected^ the velum may take the position of a band ex-
ternal to the margin of the retroverted disc, or may also be
retroverted, as in the figure. In this case the tentacles are all
more or less concealed in the concavity of the retroverted disc,
and accordingly I have not represented them in the figure at all.
When in this position the Medusa supports itself on the manu-
brium as a stalk, exerting a feeble adhesive action by means of
the quadrifid lobes of the oral aperture.

This figure also serves to show the position and form of the
gonads {Go) on the radiating canals, which are omitted in
figure 1.

Fig. 3. — Diagram of margin of the disc to show tentacles and marginal
ring. MRy marginal ring ; Tx, tentacle-axis ; Try tentacle-root.

Insertion of the tentacles, — The tentacles do not arise from
the edge of the disc, but in three more or less distinct tiers,
indicating a relative centripetal movement of the tentacles
during the growth of the disc, as in many Trachyline Medusas.

The four perradial tentacles are furthest from the marginal
ring and their free portion rises from the jelly-like convex surface
of the disc at some distance from the margin. At the same time
the cartilaginoid axis of the tentacles is continued beneath the
ectodermal tissues centrifugally towards the cartilaginoid sub-
stance of the marginal ring. It is this structural arrangement
which favours the peculiar vertical carriage of the tentacles,
since if they are to be directed downwards tow’ards the mouth

VOL. XX. — NEW SER. A A

360

PROFESSOR E. RAY LANKESTER.

a considerable flexure of the cartilaginoid axis at the point
where its free and inserted portions unite^ becomes necessary.

This is shown in the diagram woodcut, figure 3, where tenta-
cles of three horizons are shown bent mouthwards by the
pressure of the cover-glass. The recurved axis-roots {tr) of the
two larger tentacles are seen by transparency through the
substance of the disc passing towards the margin, where they
become continuous with the centripetal upgrowths of the tissue
of the marginal ring {MR),

The smaller tentacles spring almost directly from the mar-
ginal ring, and have therefore no length of axial root. This
enables the small (which are the younger) tentacles to assume
more readily the dependent mouthwardly directed carriage ;
they are not mechanically directed into the vertical upright
position by the elasticity of a vertical axial root.

Tissue of the nettle-ring and peroniay and of the cartilage-
ring and tentacle-roots. — In order to obtain a satisfactory view
of these structures witli the microscope, it is necessary to get
the margin of the disc in such a position when prepared for
examination that the optical plane is tangential to the concave
surface of the disc. Under these circumstances such a view as
that drawn in PI. XXX, fig. 4, may be obtained. This re-
presents the part as seen from the supra umbrellar surface. In
PI. XXX, fig. 5, I have represented the same parts seen with
a deeper focussing, the optical plane now passing near to the
subumbrella surface.

The surface of the marginal ring and peronia is seen to be
formed by a continuation of the large-celled ectodermal tissue
of the supraumbrellar surface, but the cells which constitute
it differ from those of the general surface in being loaded with
thread-cells. The continuous marginal ring of nettle-cells thus
constituted is entitled to the name of nettle-ring,^^ though it is
only one cell deep and does not exhibit the marked specialisation
and thickening observed in some Trachomedusae and Narcome-
dusae. It may be regarded as a rudimentary condition of the
structure which is more highly developed in other Trachyline
Medusae.

The same simple but, as it appears to me, distinct character,
is exhibited by the prolongation of this nettle-ring centripetally
towards the point where the tentacle becomes free of the disc.
The masses of thread-cells forming these little peronia"’^ are not
merely due to the continuation of the character of the ectoderm
cells which clothe the free portion of the tentacle. Por the first
])ortion of the free part of the tentacle is almost devoid of thread
cells.

It is not a little difficult, on a first examination of this region.

LTMNOCODIUM (CRASPEDACUSTES) SOWERBII. 361

to separate the superficial ectodermal cells with their refringent
nematocysts (thread-cells) from the immediately subjacent tissue,
the cells of which have a yellowish-green tint, and are also some-
what refringent.

By careful focussing this difficulty is overcome, and it is then
found that beneath the nettle-ring is a peculiar tissue, which
forms the abumbral wall of the marginal canal (PI. XXX,

The marginal canal itself is wide, and capable of distension on
its abumbral side, where the wall is formed by cells of a very
different character from that of those which underlie the nettle -
ring. In fig. 5 the superior wall of the marginal canal {MCa) is
seen in section, showing the peculiar soft clear (ciliated) cells
containing yellowish-green granules, by which it is here lined
internally. The modified cartilaginoid cells of the abumbral
wall are limited inferiorly {i, e. taking the mouth as indicating
the inferior surface of the organism) by a strongly-marked row
of quadrate cells, which form a slight ridge, and are especially
observable by their complete and uniform pigmentation (PI.
XXX, fig. 3 VMCy fig. 4 VMCy fig. 5 MCp) — yellowish green or
greenish brown. These cells form a dark line, which can alw^ays
be recognised along the insertion of the velum, and accordingly
I distinguish them as velo-marginal endoderm.

The relations of these parts are exhibited, in a diagrammatic
form, in PI. XXX, fig. 3. A comparison of this figure, with
the surface view and the optical section, will, I hope, serve to
make the structure comprehensible.

The true character of the tissue formed by the endodermal
cells of the abumbral wall of the marginal canal is best appre-
ciated by tracing the gradual transition of the tentacle root into
this tissue.

The axis of the tentacle is formed by entoplastic cartilage
cells, that is to say, cells the protoplasm of which undergoes
a cartilaginoid metamorphosis within the cell area, as in
the notochord of Vertebrata.

The tissue very closely resembles that of the notochord of a
young Lam])rey. It differs in the smaller density of the cartila-
ginous deposit, as also in the greater abundance of the cells,
from that of typical Trachyline. Medusae, and resembles that of
tlic solid tentacles of many hydroid polyps. At the same time
there can be no question that these tentacles more closely
resemble those of Trachomedusm than they do those of Le.])to-
medusse, since iji all Leptomedusse the tentacles arc hollow,
whilst in Trachomcdusae the tentacles are very usually solid, as
they arc here.

In that part of the tentacle axis which is plunged in the disc,

362

PROFESSOR E. RAY LANKESTER.

and which is called tentacle root/’ we find the same large cells
as those in the axis of the free part.

The root at first expands a little at its insertion into the disc,
but tapers as it approaches the wall of the marginal canal, and
becomes rounded ofi* (see PL XXX, fig. 5 TRj and woodcut,
fig. 3). There is, however, no discontinuity of the tissue where
the tentacle root ceases and the marginal ring commences, but
rather a very rapid transition or change in the form of the en-
dodermal cells, which equally constitute the tentacle axis and
the marginal cartilage ring.

The cells, in fact, diminish very much in size whilst exhibit-
ing a characteristic polygonal form (PI. XXX, fig. 6). In
place of the large vacuoles filled with gelatinous substance which
are presented by the large cells of the tentacle-axis, we find,
within the small polygonal cells, block-like deposits, one or two
to a cell, of a pale greenish tint and homogeneous structure.
The greenish colour of the contents of these cells shows through
the superficial ectoderm cells, and gives the whole ” marginal
ring ” a greenish-yellow appearance. It is most strongly de-
veloped in the velo- marginal cells which form an inferior limit
to these modified endodermal cells of the ring-canal’s wall.

It cannot be disputed that in the character of these cells, as
compared with those of the adumbraLwall of the marginal canal
(PI. XXX, fig. 7), we have a very well-marked distinction.
These adumbral cells are metamorphosed by the intracellular
deposit of a homogeneous greenish substance, so as to form a
sort of cartilaginous tissue, which acts the part of and morpho-
logically represents the cartilage-ring ” of higher Trachyline
Medusse. In order to appreciate their special character ^ it is
only necessary to compare them with the lining cells of the
adumbral wall of the marginal canal (PI. XXX, fig. 7).

I am not able at present to state definitely whether the simple
rudimentary cartilage-ring of our Medusa is the actual limiting
wall of the ring-canal, or whether a second layer of endoderm
cells exists deeper than it, and actually bounding the cavity of
the ring- canal.

This question I hope to decide by the preparation of sections.

Comparison of the arrangement of the tissues of the marginal
ring and tentacle root with those of typical Trachyline Medusae,
and with structures found in the Leptomedusoe. — If we compare
with what has just been described the arrangement of the homo-
logous parts in such a Trachyline Medusae as Cunina, we find
very considerable differences, which, however, appear to me to be
differences of degree, our new Medusa exhibiting in a simple form

* Cells of the same character as those of the endoderm of the marginal
ring occur in parts of the stomach-wall.

LIMNOCODIUM (CRASPEDACUSTEs) SOWERBII. 363

conditions, which are highly specialised and developed in such
forms as Cunina.

In the first place the tentacle-roots of Cunina, instead of pass-
ing centrifugally towards the margin of the disc, have rather a
centripetal direction. This is simply due to the fact that the
marginal ring and its tissues are so largely developed as to push
the tentacle root towards the centre of the disc ; and is not a
difference of a fundamental character.

Further, the cartilage of the ring is much more largely de-
veloped in Cunina than in the new Medusa, and acquires the im-
portance of a well-marked skeletal ring, which it is only (as it
were) beginning to assume in Limnocodium. The same spe--
cialisation of the nettle-ring and of the peronia is observable in
Cunina as compared with Limnocodium, but does not invalidate
the claim of the simpler structures found in the latter to re-
cognition.

It appears exceedingly probable that the Trachyline Medusae
are the modified descendants of such simpler forms as the Lep-
tomedusae, and if we search amongst the latter for conditions
approaching those exhibited by the tentacle-roots and marginal
ring so frequent among the former, the nearest case which we
can find is that of the genus Laodice, one of the Thaumantidae
of HaeckeFs system. Here the tentacles do not spring from the
margin of the disc freely, but from the convex surface of the
disc, the axis of the tentacle being prolonged as a root to join
the wall of the marginal canal, which has a cartilaginoid cha-
racter. Nevertheless, in Laodice the tentacles are hollow, and
there does not appear to be any structure corresponding to peronia.

From the consideration of the characters of the tentacles
and marginal ring alone, I think that we should be led to the
conclusion that Limnocodium is a representative of the Tracho-
medusae in an early or archaic phrase of differentiation, already
distinctively Trachomedusan, but not far advanced on that path.

We have next to see what indications a closer examination of
the marginal bodies (otocysts) may give us, and we shall find
that they confirm very distinctly the conclusion already enunciated.

Bevelopmerit oj the marginal bodies. — The marginal bodies of
Limnocodium vary in number, as do the tentacles, according to
the size of the specimens examined. I have seen as few as fifty
in specimens measuring one third of an inch across the disc, and
as many as 120 in larger specimens. Even in large specimens
the marginal bodies are continually being developed, so that
there is no difficulty in obtaining that opportunity for tracing
the development^ which Professor Allman has stated to be
desirable.

The marginal bodies consist of a spherical refringent body

364

PROFESSOR E. RAY LANKESTER.

averaging of an incli in diameter, enclosed in an

elongated tubular capsule. The refringent body is situated at
the line of junction of the velum with the marginal ring, within
the slightly enlarged termination of the tubular capsule.

The tubular capsule is continued thence radially to the
extreme margin of the velum, whence I have termed these
capsules ‘Welar canals^'’ (PI. XXX, fig. 1).

The refringent body belongs essentially to the abumbral side
of the velum— that which is adjacent to the tentacles — as will be
described below, whilst the capsule is entirely formed by the
ectodermal cells of the abumbral surface of the velum.

The refringent bodies and their capsules form interruptions in
a cellular band of thickened ectodermal tissue, which runs along
the line of junction of the velum and disc peripherally to the
strongly-coloured velo-marginal cells of the endoderm of the
marginal ring.

The appearance of this ring of transparent, colourless cells,
and the relation of the marginal bodies and their capsules to it,
is seen in PL XXX, fig. 5, ca. The transparent ring (^r) is
undoubtedly the representative of the abumbral nerve-ring of
other Hydromedusse ; but I have not yet made that special his-
tological study of its elements which would enable me to say
how much of the ring is to be regarded as nervous tissue, and
how much is unspecialised ectoderm. ^

The tubular capsules of the marginal bodies are often to be
found in an incomplete state of development, and it is then
possible to observe the process by which they increase in length,
so as to reach finally the edge of the velum. The tubular cap-
sules occupy a position between the abumbral and adumbral
ectoderm layers which constitute the velum, and they actually
stand out on the abumbral surface of the velum as ridges raising
up the abumbral ectoderm of the velum, and separating it en-
tirely from the adumbral ectoderm in the form of a delicate
membrane. This is well seen when the velum is folded on
itself, as not unfrequently happens when a specimen of the
Medusa is placed beneath a cover glass, as shown in PI. XXX,
fig. 2.

If the tubular capsules be now carefully studied it is not
difficult to make out that they increase in length, not by the
simple growth of the already existing capsule, but by the fusion
with the capsule of vacuolated cells belonging to the abumbral
layer of the velum (fig. 2 vac').

The vacuolated cells of the abumbral ectoderm layer of the
velum may be seen in various regions of the velum, ready, as it
were, to fuse with the capsules, should their growth favour such
an occurrence. The tendency to fuse on the part of the cap-

LIMNOCODIUM (CRASPEDACUSTES) SOWERBIV. 365

sules is indicated in a rare phenomenon, observed bj my
assistant (Mr. Bourne), namely, the fusion of neighbouring
tubular capsules, an instance of which is shown in PL XXX,
fig. 1.

Deformities and abnormalities of many kinds are not unfre-
quent in the capsules, the marginal bodies themselves, and the
tentacles. The vacuolated cells of the abumbral velar surface
a])pearto stand out from the stratum in which they originally lay,
and sometimes to form separate vesiculi, which may or may not
fuse with the growing capsule of the refringent body.

The adumbral ectoderm is complete and continuous beneath
the tubular capsules, its circular muscular fibrils being traceable
in every part beneath (that is, on the adumbral surface of) the
tubular capsules (see PI. XXX, fig. 2 muse).

The development of the velar canals or tubular capsules of
the marginal bodies is, then, shown to be due to the fusion of
vacuolated cells of the abumbral ectoderm of the velum.

It remains to be shown what is the first origin of the capsule
and of the refringent body itself.

In PI. XXXI, figs. 10 to 20, I have represented a number of
stages in the development of the refringent body, and it is perhaps
as well at once to say that the refringent body is nothing more
nor less than a modified solid tentacle. It consists in the fully
formed state of a number of cortical cells enclosing four, six, or
eight large refringent cells (PL XXXI, figs. 10, 11) . The cortical
cells correspond to the ectoderm of a tentacle, and the medul-
lary refringent cells to the endodermal axis of a tentacle, the
whole group of cells being tightly pressed together so as to form
a spherical refringent body. The cortical as well as the medul-
lary cells of the free hemisphere of the body are highly refringent.

There is not in the refringent body of Limnocodium any
separate concretion^ any otolith in the strict sense of the term.
The whole structure is purely cellular, and consists entirely of
nucleated cells which can be isolated and recognised as distinct
cell elements by means of reagents.

In this respect the refringent body of the new Medusa appears
to differ from any previously described organ of the kind.
Certainly it differs entirely from any marginal body either of
Leptomedusae or of Trachyline Medusae described by the
llertwigs in their recent work ^ Nervensystem der Medusen.’
There seems very small justification for regarding the refringent
body of Limnocodium as an auditory organ at all. It consists
simply and solely of nucleated cells, the substance of which is of
a higlily refringent nature. Refringent bulbs would be a
sufficiently appropriate name for these bodies. The cells which
correspond to the tentacle ectoderm, and which I term cortical

366

PROFESSOR E. RAY LANKESTER.

cells/^ are very delicate, often much stretched and translucent,
but in order, as it were, to fill up the interstices of the spherical
mass, they are sometimes enlarged in places and refringent like
the medullary cells.

The cortical cells are, however, never so distinctively modified
as the medullary cells. There is a peculiar brownish colour and a
refringent appearance about the medullary cells, especially after
treatment with osmic acid, which is not to be noticed in the
cortical cells.

The medullary cells sometimes contain one or two greenish-
yellow granules (figs. 8, and 15 gr\ which are identical in ap-
pearance with the granules found in the endoderm cells of the
marginal ring canal/ro?^ which they are derived.

By searching along the velo-marginal line it is easy in some
specimens to find refringent bodies and their capsules in all
stages of development. The earliest indication which I have
obtained of this development is a protrusion of the endodermic
cells of the marginal canal into the ring or band (^) of colourless
cells, which is ectodermal in origin, and probably to a large
extent nervous in character. This protrusion (PI. XXXI, figs. 12,
13), lifts the ectodermal tissue in front of it, and one of the in-
growing endodermal cells enlarges and becomes highly charged
with fine granules, and sometimes with coarser, green-coloured
granules also. As far as I have been able to ascertain, this enlarged
endodermal cell, which frequently is coloured like the other endo-
dermal cells of the ring- canal, is the mother-cell, from which the
axial or medullary cells of the refringent body are developed (see
PL XXX, figs. 14, 17, 18, axen). The ectodermal tissue into which
this enlarged cell has protruded now grows around it in a very re-
markable way. In some cases it seems to form at first a complete
investment of small cells — the cortical layer — enclosing the en-
dodermal medullary cell. But this appears from such a con-
dition as that represented in PI. XXXI, fig. 14, not to be the
actual mode of growth. The ectodermic layer, pushed forward by
the axial outgrowth of endoderm, arranges itself in such a way as
to form a cortical layer to the medullary cell, and at the same
time a capsule embracing the central body formed by these two
sets of cells (see PI. XXXI, figs. 15, 16). From henceforth the
medullary cell has merely to divide and give rise to the four or
eight highly refringent medullary cells of the fully-formed re-
fringent body, and the capsule has only to expand and increase
its area by the agglutination of vacuolated ectoderm cells in the
manner described above.

The ectodermic cortical investment of the highly refringent
central or medullary cells of the marginal body is complete ;
the endodermic medullary cells are entirely cut off from their

LIMNOCODIUM (CRASPEDACUSTES) SOWERBII. 367

seat of origin ; but the cortical cells rest upon the tissue of the
transparent marginal ring and have some kind of attachment
to that ring, though I am not able to point out any definite
nerve-fibres. That the attachment is very slight is shown by
the fact that sometimes a marginal body (refringent body) be-
comes broken off from its seat of origin, and falls into the
tubular capsule. I have met with such a detached refringent
body enclosed in a constriction of the peripheral extremity of a
capsular tube (PI. XXXI, fig. 20) . This detached refringent
body was remarkable for the fact that it had undergone an ab-
normal development, and was actually a c^st with an internal
cavity, into which projected the large refringent cells, whilst it
was clothed externally by the cortical cells.

That the refringent bodies in themselves, irrespective of the
tubular capsules, correspond morphologically to solid tentacles
is proved by an occasional abnormality. In place of a normal
refringent body I have found rarely a tentacle-like body one of
which is drawn in PI. XXXI, fig. 18. This small tentacle was
in exactly the same position as a refringent body. It is seen to
have in its axis one large and apparently two smaller granular
cells similar to those which are protruded from the endoderm of
the ring- canal to form the medullary cells of the normal re-
fringent body. ‘At the same time the ectoderm which clothes
these axial cells is free and superficial. Instead of forming a
capsule^ the ectoderm cells have simply invested the axial growth
as they do in the case of a tentacle.

Further, I have observed another abnormality of an exactly
complementary character, namely, the commencement of the
tubular ectodermal capsule without any refringent body to oc-
cupy its proximal extremity (PL XXXI, fig. 19). The exist-
ence of these two elements separately one from the other, jus-
tifies us in regarding them as two distinct structures which have
united to form the complete ‘^refringent bulb and sac of our
new Medusa.

Have we anything parallel in Leptomedusae or in Tracho-
medusae for the two distinct elements, the modified tentacle and
the velar ectodermal capsule, above described ? Assuredly there
is no parallel at present described among Leptomedusae.
Among Trachomedusae, however, we find a very exact parallel.
Although in no other Trachomedusan does the axial ])ortion of
the modified tentacle retain so simply the character of the car-
tilaginoid axial tissue of a tentacle, and although in other Tra-
chomedusae an otolithic concretion is formed by the axial cells,
whilst an otolith is most decidedly not formed in Limnocodium,
yet in the essential character of the refringent body, and in the
remarkable ectodermal capsule, Limnocodium agrees most pre-

368

PROFESSOR E. RAY LANKESTER.

cisely with the Trachomedusse, such as Ehopalonema, described
by the Ilertwigs. It appears that in Ehopalonema and other
Trachomedusse, the marginal body originates as a small
tentacle-like outgrowth with endodermal axis. Whilst the
endodermal axis undergoes modifications, an ectodermal up-
growth (of the abumbral velar surface) occurs all round the
little tentacle so as to form a sort of investing vesicle, which at
first is open, but in the course of development closes up. The
^ little tentacle is thus enclosed in a spherical sac or capsule.

Precisely the same history attaches to the tentacle-like com-
mencement of the refringent body of Limnocodium, excepting
that the capsule commences somewhat early, and that the whole
structure, both tentacle-body and capsule, instead of standing
out freely on the abumbral surface of the velar insertion, keeps
a special course of direction of growth, creeping so to speak
between the two layers of the velum instead of pushing out-
wards at right angles to the abumbral surface.

There can, it seems to me, be no doubt that the refringent
body of Limnocodium is identical with the free auditory bulbs”
(freie Horkolbchen) of Trachyline Medusae, in which phase the
so called auditory organ of Trachomedusae makes its first ap-
pearance. The tubular capsules” or “ velar centrifugal canals,”
which enclose the refringent bulbs of Limnocodium, are iden-
tical with the at first open and subsequently closed auditory sacs
of ectodermal origin, which grow up around the Horkolbchen
of Trachynemidae, Olindiadae, and Geryonidae.

I have observed the development of true tentacles in the new
Medusa, and may state that in them the endodermal axis at
once begins to assume its special character, wLich prevents any
possibility of confusion (apart from their different positions)
between the true tentacle and the tentacular auditory ”
bulb.

Until it is shown (as it appears to me very possibly may
some day be shown) that some Leptomedusse have their marginal
bodies formed by modified tentacles, as in Trachomedusse and
Narcomedusse, and not purely and simply from ectodermal cysts,
as the Hertwigs have described, the features wdiich I have
shown to characterise the development and essential structure
of the marginal bodies of Limnocodium will render it necessary,
even apart from those features of the tentacles and marginal
ring above described, to associate that form, not with Lepto-
medusse, but with the Trachomedusse, as I pointed out when I
gave to it the name Craspedacustes.

Embryonic condition of Limnocodium. — Not the least astonish-
ing fact about the new fresh-water Medusa is the immense pre-
ponderance of males. Among more than fifty specimens wdiich

LIMNOCODIUM (CRASPEDACUSTES) SOWERBII. 369

I have examined^ I have not found one female. Yet females, or
egg-bearing forms, are there.

"when I first received specimens of the Medusa from Mr.
Sowerby, he imparted to me his conviction that young Medusae
were being hatched in the glass jar in his sitting-room, which
contained a number of full-grown specimens. He inferred this
from the fact that among these large specimens which he had
introduced into the jar a number of minute specimens made
their appearance after the lapse of three or four days. On a
subsequent occasion Mr. Sowerby again drew my attention to
these minute Medusae, and enabled me to examine a number of
them.

Had I been preoccupied with the notion that Limnocodium
must have a hydriform trophosome, I might perhaps have neg-
lected the opportunity of examining these minute forms. But as
I had come to the conclusion, on anatomical grounds, that Lim-
nocodium is one of the Tracho medusae, I was quite prepared,
when Mr. Sowerby for the second time mentioned these minute
Medusae, to find very minute Limnocodia hatched from eggs in
the vessels in which the adults were kept.

The youngest specimen which I have at present examined
measured only one thirtieth of an inch in diameter, and I have
examined others very little larger.

They agree in a very striking way with the embryo of
Geryonia hastata, figured by Metschnikow in the ^ Zeitsch. fiir
wiss. Zoologie,’ Bd. xxiv, pi. ii, figs. 13, 15. There is no pos-
sibility of doubting that these embyros were developed from eggs,
just as are those of Geryonia, Cunina, and Higinopsis.

The smallest embryo (woodcut, fig. 4) was of a slightly de-
pressed, nearly spherical, form. It exhibited that very striking
separation of the ectoderm from the endoderm layer which is seen
in the similar embryo of Geryonia. Pour perradial tentacles, very
short and stump-like, were sprouting at a little distance from one
depressed pole of the sphere. Between these, rudiments of four
others were seen. The external ectodermal layer was unbroken
and not invaginated at any part of the sphere. Yet within could
be seen the subumbrellar cavity, the manubrium with its mouth,
the quadrangular stomach, and the four radiating canals.

The ectoderm was, it is very strange to say, continued across
the mouth of the umbrella, so as to close in the subumbrellar
space. Were this covering to be ruptured centrally the sub-
umbrellar cavity would be found ready formed, with dependent
manubrium and mouth. The ectodermal lid thus centrally per-
forated would very readily become the velum.

I am at this moment unable to say whether this is the
mode of development, since in specimens a little larger with

370

PROFESSOR E. RAY LANKESTER.

eight tentacles, fig. 5, the praeumbral lid was still imper-
forate.^

Tig. 4. — Enibrjo of Limnocodium Sowerbii, ^th of an inch in diameter.
A. Surface view of oral pole. B. Optical section of same specimen
in a plane at right angles to the oro-apical axis. Pt. Per-radial ten-
tacle. L. Praeumbral lid. RC. Radial canal. St. Stomach. Ec. Ecto-
derm. G. Jelly of the disc.

Pig. 5. — Embryo of Limnocodium Sowerbii, a very little more advanced.
A. Surface lateral view. B. Optical section of the same specimen in
a plane including the oro-apical axis. Pt. Per-radial tentacles. L.
Praeumbral lid. UC. Sub-umbrella cavity. Mn. Manubrium. RC
Radial canal. Ec. Ectoderm. G. Jelly-like substance of the disc.

* Whilst this is passing tl)rough the press I have ascertained that such
is the development of the velum.

371

LiMNOCODIUM (CRASPEDACUSTEs) SOWEllBll.

The movements of the embryo in this phase are active enough.
A frequent pulsation of the subumbrellar musculature is seen,
which alters but little the surface outline of the spherical embryo.
The spherical form is, however, slowly altered, and occasionally the
embryos become much elongated parallel to the oro-apical axis.

I defer any further description of the development of Limno-
codium until I can submit the more complete drawings of the
stages which I have already recorded and of others for which I
am now in search.

Intra-cellular digestion in Lim7iocodium,—T\\Q exceedingly
important fact that some of the Coelentera, and lower kinds of
worms, digest their solid food by the inception of the solid food-
particles into the substance of endodermal cells, each endodermal
cell behaving as an Amoeba, has now been fairly established by
the observations of Allman on Myriothela, of Metschnikoff on
Turbellarians, and of T. J. Parker on the common Hydra.

Limnocodium exhibits this mode of digestion in the most
striking and obvious manner. I have prepared drawings of the
endodermal cells of the stomach, showing their amoeboid character,
and showing further the presence of such food-bodies as Pro-
tococci, Diatoms, and Euglenae, in various stages of digestion,
within the protoplasm of single cells, and of aggregated groups
of such cells. These were observed in and drawn from living
specimens of Limnocodium with a magnifying power of 800
diameters.

NOTES AND MEMORANDA.

On the Development of the Structure known as the ‘ Glo-
merulus of the Head-Kidney ’ in the Chick. ^ By Adam Sedg-
wick, B. A. — In a paper by Mr. Balfour and myself in the ^ Quart.
Journ. of Micr. Science,' vol. xix, describing the development of
what we believed to be a rudimentary head-kidney in the chick,
we drew attention to a structure which so closely resembled the
glomerulus of the head-kidney of the Icthyopsida that we iden-
tified it as an homologous structure.

Gasser ^ has also independently discovered and similarly iden-
tified this structure.

In the paper just referred to no attempt was made to trace
the development of this glomerulus, but it was merely described
as it appeared at its time of greatest dpelopment.

The following description is taken from that paper :

In the chick the glomerulus is paired and consists of a vas-
cular outgrowth or ridge projecting into the body cavity on each
side at the root of the mesentery. It extends from the anterior
end of the Wolffian body to the point where the foremost open-
ing of the head-kidney commences. We have found it at a
period slightly earlier than that of the first development of the
head-kidney. In the interior of this body is seen a stroma with
numerous vascular channels and blood-corpuscles, and a vas-
cular connection is apparently becoming established, if it is not
so already, between the glomerulus and the aorta. The stalk
connecting the glomerulus with the attachment of the mesentery
varies in thickness in different sections, but we believe that the
glomerulus is continued unbroken throughout the very consider-
able region through which it extends. This point is, however,
difficult to make sure of, owing to the facility with which the
glomerulus breaks away. At the stage we are describing no
true Malpighian bodies are present in the part of the Wolffian
body on the same level with the anterior end of the glomerulus,
but the Wolffian body merely consists of the Wolffian duct. At

‘ I?cad bc-.fore the Cambridg-e Philosophical Society,

2 ‘ Sitzungsbcrichte der Gesellschaft zur Beford d. gesain. Naturwiss./
No. 5, 1879.

NOTES AND MEMORANDA.

373

the level of the posterior part of the glomerulus this is no longer
the case, but here a regular series of primary Malpighian bodies
is present, and the glomerulus of the head-kidney may fre-
quently be seen in the same section as a Malpighian body. In
most sections the two bodies appear quite disconnected, but in
those sections in which the glomerulus of the Malpighian body
comes tito view it is seen to be derived from the same formation
as the glomerulus of the head-kidney.^'

The point which is left in doubt in the above description, viz.
as to whether the glomerulus constitutes a continuous structure,
is at once decided by a study of its development.

I may here state that it is not a continuous structure, but
consists of a series of external glomeruli, each of which corre-
sponds and is continuous with the glomeruli of the Malpighian
bodies found in this part of the trunk.

I will commence the description of the development at the
time when the segmental tubes have reached the stage of deve-
lopment figured by Kolliker^ and myself.^ At this stage each of
them in the anterior region of the Wolffian body has the form of
an S-shaped string, with a narrow opening into the body cavity,
the lower limb of the S being formed by the intermediate cell
mass, and the upper limb by a column of cells which connects
the intermediate cell mass with the Wolffian duct.

In the region where each external glomerulus is afterwards
found the openings into the body cavity, which are homologous
with the peritoneal openings of the segmental tubes in Elasmo-
braiichs, widen out very considerably, and a lumen is continued
from them into the intermediate cell mass on the one hand, and
on the other hand into the column of cells which forms the
upper limb of the S and connects the intermediate cell mass with
the Wolffian duct.^

That part of the segmental tube which will afterwards become
a ^Malpighian body is therefore, in the region where an external
glomerulus will subsequently be formed, connected with the body

* This may best be understood by examining fig. 11, PI. XVII, in my paper
already referred to (‘ Quart. Journ. of Micr. Science,’ April, ISSO). If the
primary Wolffian tubule here represented, were connected with the

peritoneal epithelium at the point where the line from toL^ cuts it, and it
were open to the body cavity at that point, an appearance similar to that
which 1 have attempted to describe would be obtained. Or perhaps a better
idea of the structure may be obtained from fig. G, PI. XX, in Balfour’s
‘ Monograph on the Development of Elasinobrancli Fishes.’ If si were very
short and wide, so that mg were widely open to the body cavity, the figure
would resemble a developing Wolffian tubule in this anterior part of the
chick’s Wolffian body.

’ ‘ Entwicklungsgeschichtc dcs Menschen u. dcr hoheren Thierc,’ p. 201,
2nd cd.

> ‘ Quart. Journ. Mic. Sci.,’ April, 1880, PI. XVll, fig. 1.

374

NOTES AND MEMORANDA.

cavity by a short tube. This tube rapidly widens out^ especially
anteriorly, to such an extent that it soon appears as a shallow
bay in the body cavity. Thus each opening at this stage forms
a bay, wide and shallow anteriorly, becoming deeper and nar-
rower as we pass backwards, until finally behind it is separated
from the body cavity altogether, and there is seen in section a
Malpighian capsule precisely resembling a developing Malpig-
hian capsule in the hinder region of the Wolffian body.^ In this
bay and in the small part behind continuous with the bay, but
separated from the body cavity, which are together serially homo-
logous with a Malpighian capsule and the funnel leading from
it into the body cavity, a small glomerulus soon appears
attached to the dorsal wall. The glomerulus increases in size,
and the bay anteriorly widens out very much, while behind it
remains deep, and finally passes into the closed posterior por-
tion. The glomerulus fills up this passage, which clearly runs
obliquely backwards and dorsalwards, and eventually, as far as I
can ascertain, the opening becomes completely closed, the epi-
thelium on the external glomerulus being no longer continued
through the opening on to the internal glomerulus.

The external glomerulus, then, in the chick which has hither-
to been known as the glomerulus of the head-kidney, is nothing
more than a series of glomeruli of primary Malpighian bodies
projecting through the wide openings of the segmental tubes
into the body cavity. Their extreme antero-posterior extension
may be said to be within the ninth and thirteenth segments.

In the chick the primary segmental tubes corresponding to
these external glomeruli are apparently never fully developed.

I may mention that the external glomeruli are present in
greater numbers, and attain a greater development in the duck
than in the chick.

I defer the details and all discussion of this extraordinary and
unexpected development until I am able to publish a fuller paper
with figures.

Bacterium Anthracis. — Professor Greenfield, of the Brown
Institution, has recently made the following report to the
Royal Agricultural Society :

In my former report, published in the last volume of
the Society’s journal, I gave the results of my experiments
so far as they were completed to the middle of February.

It may be remembered that the chief results of the experi-
ments, briefly stated, were —

(1) That splenic fever may be transmitted to a bovine
^ Loc. cit., fig. 11.

NOTES AND MEMORANDA.

375

animal by direct inoculation from a Rodent ; that the disease
thus criven, though severe and possibly fatal, is not usually
so, and that the m.odified attack of the disease confers a
certain degree of protection from subsequent attacks com-
municated in the same way ; so far, at any rate, as the
experiments had been carried.

(2) I showed also that the fungus which constitutes the
essential contagium, when grown in successive generations
in a cultivating fluid, was still capable of giving rise to the
disease, being in one case fatal to a cow in the first genera-
tion, in another fatal to a sheep in the fourth generation.
My experiments also showed that an attack thus com-
municated, causing severe symptoms, appeared to be equally
protective against results from future inoculations with the
disease given directly from the guinea-pig.

I pointed out that, although I had applied as severe tests
as were at my command, to ascertain the degree of protec-
tion conferred, it was yet desirable to perform the more
crucial test of direct contagion from another bovine animal ;
and that, if the experiments were successful, the precise
method of performing the protective inoculation, and the
durability of the protection, would require further investiga-
tion. Having regard, also, to the small number of animals
as yet experimented on, it would be desirable to repeat similar
trials on others.

Keeping these points in view, I have continued the experi-
ments, and have inoculated two other animals with similar
results. Ill one case also, which had just recovered from the
first attack at the time of my report, I have ascertained by
further inoculations that protection had been conferred.
Thus in all, I have added three more to the cases of success
by this method. In one of these, the disease was given by
the fourth generation of the cultivated fungus, and the
symptoms were severe.

8o far as concerns the crucial experiment, that of ex-
])osure to direct contagion from another bovine animal, I
liave not as yet had any opportunity of making it, and am
still awaiting the occurrence of some outbreak, of which I
liopc wo may receive early information. But I have just
received some material from other animals which is known
to be highly infectious, and am about to make experiments
with it.

Since my previous report, another very interesting and
equally important point has become clear, which may I
hope j>rove of great value in future. It is that, when the
virus of the disease (the fungoid organism known as Raci7^W5

VOL. XX. NEW SEll. B B

376

NOTES AND MEMORANDA.

anthracis) is artificially cultivated in an indifierent fluid,
by the method of successive generations, which I have
described in my report, each successive generation becomes
less active than its predecessor, and when inoculated acts
not only with less intensity, but more gradually, and often
in a somewhat difierent manner. This modification takes
place to such a degree that w^hen the cultivation has been
carried to the fourteenth or fifteenth generation, it may
be introduced with impunity into the system of a mouse,
which is one of the animals most susceptible to the poison.

Apart from its scientific interest, this fact will doubtless
prove to be of practical value, for by its means it will be
possible to obtain a virus of sufficient activity to produce an
attack of the disease which shall be protective, but not of
sufficient severity to be dangerous, or in any way injurious
to the animal inoculated.

With regard to any apprehended ill-effect upon the
animals thus inoculated, I may say that the cows which we
have used have thriven remarkably well, and none so well
as that which has been most severely tested.

I hope in a future report to give the details of these in-
vestigations, which have necessarily been extensive and
complicated.

I venture, therefore, to urge upon the Society the im-
portance of continuing these experiments, so as to bring
them to a complete and decisive result. In order to do this,
further outlay in the purchase and keep of animals, and
other expenses, will be necessary, which will involve the
renewal of the grant for the ensuing six months.

Dr. Carl Rabl on the Pedicle of Invagination in Pulmonate
Gastropoda. — Dr. Rabl has renewed his investigations on the
embryology of Planorbis, and has arrived at a result which
brings his observations of fact and my own (published in
this journal in 1874) into close agreement on a matter of
considerable importance concerning which he was at first led
to differ from me.

He writes relatively to the question of the existence of a

pedicle of invagination^’ as follows (dated June 4th) :

The youngest embryos, of which I have cut sections in
order to determine the above question, were little older than
those drawn by me in Plate XXXII, fig. 20, of my Memoir.
Although by observation of the embryos as whole objects
I could see nothing which could lead to the inference of the
existence of a pedicle of invagination, yet I convinced my-
self by the aid of sections that a cylindrical solid cord

NOTES AND MEMORANDA.

377

exists which leads from the cavity of the mid-gut to the
integument. This cord consists of small endoderm-cells,
rich ill granules, and the ectoderm-cells of the spot where it
comes into contact with the integument are somewhat
longer and poorer in granules than the neighbouring cells.
There is, however, no trace to be seen of an orifice of in-
vagination.

In sections of somewhat older embryos one can observe
that the cord becomes hollowed out, the hollowing process
taking origin from the cavity of the mid-gut, and it is easy
enough to convince oneself that it gives rise to the terminal
gut. Accordingly I am in the pleasant position of being
able to confirm your statements, in so far as that a cord
exists which stretches from the integument to the gut, but,
on the other hand, I consider that my view is confirmed,
that the terminal gut is a derivative of the mid-gut.

Your interpretation of this cord as a ^pedicle of invagi-
nation ’ I am not able now, any more than before, to accept.’^
Dr. Rabl will explain his views and observations at
greater length in Professor Gegenbaur’s ^ Morph. Jahrbuch.’

E. Pay Lank ester.

Development of Muscular Tissue from Epiblast in the Mam-
malia.— According to recent observations of Kanvier this ex-
tremely important and unexpected fact has been established
by him in regard to the muscular coat of the sweat-glands.
The reader is referred to Ranvier’s notice of his discovery in
the ' Comptes Rendus of the Acad, of Sciences of Paris,’
Dec. 29th, 1879.

PROCEEDINGS OF SOCIETIES.

Dublin Miceoscopical Club.

Novemher, 1879.

Chroococcaceous alga from Leicester ^exhibited. — Mr. Archer ex-
hibited examples from a small collection of a Chroococcaceous algal
form, kindly forwarded him by Mr. Holmes, of London, which had
occurred in considerable quantities in the water supply of the town
of Leicester. This, regarded from what might be called a mor-
phological point of view, and taken as it stood, would be referable
to the genus Coelosphserium, Nag., but it was clearly not iien-
tical with the common Ccelos'pliceriwn Kutzingianwn, Nag., of
which Mr. Archer exhibited some examples for comparison and
illustration. He might mention that the latter form, when taken
from running water, as these examples were, seems to possess
notably smaller constituent cells than in the form taken fron
standing pools. In C. Kutzingianum, however, the individual
cells are globose. In Mr. Holmes’ plant they were elongate,
considerably larger, of a different tint and internal appearance.
The fission of the cells took place in the direction of the longer
axis, that is, radially, as regards the aggregate group or
rotund colonies, whereby the constituent cells came to
stand themselves radially, that is, pointing to the common
centre. It may be a question, indeed, how far Ccelosphaerium
can be correctly regarded as a generic type ; for instance, the
clearly allied form referred to Bacterium by Lankester, his
B. rubescens, would appear to temporarily put on a Coelo-
sphserium-like condition — nay, also a Merismopsedia-like, a
Clathrocystis-like, and other phases. Since arrival many of the
“ CoelosphaBrium ” aggregations had become broken up the cells,
still living and subdividing, the fission still in the lougitudinal
direction ; in that state the as yet not disconnected pairs of cells
called to mind the appearance of a pair of cotyledons of some
seed — say a bean — opened, and as yet held together by the em-
bryo. Of course, this plant possessed considerable affinity to
such as Microspora {Clathrocystis') ceruginosa, agreeing, too, with
it, as would appear, in its copious occurrence when met with ;
but the elongate form of the cells and different tint were striking.
A very slight pressure was sufficient to drive away from a colony
a number of the constituent cells, and then, or when apparently
normally broken up, of course the separated cells might pass for
a species of Synechococcus (where the line of fission, however, is
transverse), but just as, well the same might be said for the
separated cells of a Clathrocystis, which might pass for, say,
Chroococcus minor. It does uot seem at all unlikely that under
some different impulses the vegptating cells of this group may

DUBLIN MICROSCOPICAL CLUB.

379

take on dilferent modes of combination, and assume arrangements
in themselves as yet supposed to be of generic import. Here,
however, the cells themselves seemed to present an aspect un-
usual ; if one could judge correctly, they seemed to offer a
slightly tapering figure, but by no means comparable to that of
the related GmnphospJiceria aponina, in which the cells are almost
peg-top shaped, or, when partially divided, “ heart-shaped ; ” but
the present plant agrees with that in the longitudinal direction
of the line of self-division. Probably a characteristic connected
with their slightly unequally-ended figure was that the last point
of coniection between the Wo young cells not yet disjunct was
not at the centre but notably nearer one end, thus increasing the
resemblance (alluded to) to the opened but not sundered pair of
cotyledons of a bean. In this plant the internal arrangement of
the fittle granular and darkish masses of endochrome seemed
peciliar.

drains of Ruby in Australian sand. — Dr. Prazer showed some
fine sand from Australia, taken from a river bed, many of the
giains of which were of ruby, very pretty to look at under a low
power, but far too small to be of any value. The sample had been
^ven to him by the Rev. Mr. Whitmee.

Modifications of structure presented in the cross sections of the
spines of Goniocidaris caniculatus, G. tubariaf and G. geranoides.
— Prof. Mackintosh exhibited cross-sections of the spines of
Goniocidaris canicalatus, A. Agassiz, G. tubaria., Lamk., and
G. geraniodes^ Lamk., and called attention to the interesting
modification of the structure to be observed in them. In all
three the central parts of the spine corresponded with the typical
Cidarid arrangement. In the first named the peripheral crust
was also sufficiently normal, there being no extensions of the
intermediate part of the spine ; but in G. geranioides, as is well
known, instead of the usual shell perforated by fine radiating
tubules, there was an investment of reticulated tissue of the
common echinoid type, supported by numerous extensions out-
wards of the intermediate parts of the spine tissue. (Por figures
of this form see Agassiz, ‘ Revision of the Echini,’ pi. xxx, and
Mackintosh, ‘ Trans. Roy. Irish Acad.,’ vol. xvii, pi. ix.) In
G. tuharia was seen the connecting form between the two ex-
tremes, for there the crust is still more or less reticulated, but
the projections into it are few and comparatively short. Prof.
Mackintosh was indebted for the specimens, from which the
sections had been made, to Prof. A. Agassiz, of the Harvard
University.

Minute quasi-par asitic Callithamnion on Lomeniaria articu-
lata. — Dr. E. Perceval Wright exhibited some small morsels of
the young fronds of Lomentaria articulata^ from the surface of
which were to be seen in many stages of growth a very minute
quasi-parasitic species of Callithamnion. It first made its ap-
pearance as a single filament of four or five cells, protruding
from some of the outer cells of Z. articulata / this soon branched,

380

PROCEEDINGS OF SOCIETIES.

and in process of time the little plant reached dimensions just
visible to the unassisted vision ; at this stage its tetraspores were
found, but no trace was detected of either carpogonia or anthe-
ridia. There were great difficulties in the way of referring this
plant to any of Eeinsch’s species, many of which were indeed
still without specific names. It had been found pretty common
on L. articulata, collected about Howth during the spring of
1879, and doubtless would turn up again, when perhaps its true
fruits might be discovered, and with them a more perfect account
laid before the club.

Branching of the staminal hairs of Tradescantia. — • Prof.
M’Nab drew attention to a singular case of branching hairs from
the stamens of Tradescantia virginica. The penultimate cell
gave off at its base abranch consisting of two cells, the aiis of
growth of the branch being downwards and backwards. This
form and three other hairs, each with a branch consisting of a
single lateral cell, were all obtained from an unopened flo\ier-
bud, from a cut flower-stem, which had been some time in water,
and therefore under abnormal conditions.

December, 1879.

Cosmarium isthmochondrum, Nordst., new to Ireland, exhibited.
— Mr. Archer showed, new to Ireland, Cosmarium isthmochon-
drum, Nordst., from Connemara, and, similarly to Nordstedt’s
own example, as he states, in company with Cosmarium gul
narium. Both these are very rare species, but when met with
sometimes occurring in quantity. As Herr Nordstedt remarks,
these two species might possibly be confounded, but Mr. Archer
thought, if their distinction were but once properly appreciated,
such a mistake could not occur an a second occasion, for it
needed only a little careful observation to see their striking and
strong points of difference ; even with a low power, when one
has got a good grasp of the outline of each, Mr. Archer thought
they might be very readily discriminated. There could be no
doubt but that Cosmarium isthmochondrum is a very good and
distinct species.

Barasitic Florideous alga in Plocamium coccineum, exhibited.
— Dr. E. Perceval Wright exhibited some filaments of a Elori-
deous alga, which he had found in the interior of the main
portion of the frond of Plocamium coccineum. These filaments
were of a distinctly red colour, were frequently branched, and
apparently grew np within the cell tissue of the Plocamium. Dr.
Wright had found them in November, 1878, and was again
reminded of them by the appearance, in the ‘ Botanische Zei-
tung’ for 17th January, 1879, of a paper by Eeinsch, on
“ Entozoische Pflanzenparasiten,” in the plate accompanying
which, fig. B, an almost identical form is figured. Unfor-
tunately though watched from time to time, nothing had yet
transpired to throw any additional light on their life history.

MEMOIRS.

Larval Forms : their Nature, Origin, and Affinities.

By F. M. Balfour, M.A., F.R.S., Fellow of Trinity

College, Cambridge.

Preliminary considerations. — Animals either (1) undergo
the whole of their early development in the egg or within
the body of the parent, and are hatched in a condition closely
resembling the adult ; or else (^) they are born in a condi-
tion differing to a greater or less extent from the adult.
When born in the latter condition they are known as
larvae, till they have approximately acquired the adult
characters of the species. There are no questions which are
of greater importance for the embryologist than those which
concern the nature of the secondary changes likely to occur
in the foetal or in the larval state ; since it is on the answer
to such questions that our knowledge of the extent to which
a record of the ancestral history may be expected to be pre-
served in development depends. The principles which govern
the perpetuation of variations which occur in either the larval
or the foetal state are the same as those for the adult state.
Variations favorable to the survival of the species are equally
likely to be perpetuated, at whatever period of life they occur,
prior to the loss of the reproductive powers. The possible
nature and extent of the secondary changes which may have
occurred in the developmental history of forms, which have
either a long larval existence, or which are born in a nearly
complete condition, is primarily determined by the nature
of the favorable variations which can occur in each case.

\Vhere the development is a foetal one, the favorable
variations which can most easily occur are — (1) abbrevia-
tions, (2) an increase in the amount of food-yolk stored up for
the use of the developing embryo. Abbreviations take place
because direct development is always simpler, and therefore

VOL. XX. NEW SER. C C

382

F. M. BALFOUR.

more advantageous ; and, owing to the fact of the foetus not
being required to lead an independent existence till birth,
and of its being in the meantime nourished by food-yolk,
or directly by the parent, there are no physiological
causes to prevent the characters of any stage of the develop-
ment, which are only of functional importance to a free
larva, from disappearing from the developmental history.
All external organs of locomotion and nutrition will, for this
reason, obviously have a tendency to disappear or to be reduced
in foetal developments ; and a little consideration will show
that the ancestral stages in the development of the nervous
and muscular systems, organs of sense, and digestive system
will be liable to drop out or be modified, when a simplifica-
tion can thereby be affected. The circulatory and excretory
systems will not be modified to the same extent, because both
of them are usually functional during foetal life.

The mechanical efiects of food-yolk are very considerable,
and numerous instances of their influence will be found in
my treatise on ^ Comparative Embryology.^ They mainly
affect the early stages of development, i.e. the form of the
gastrula, &c.

The favorable variations which can occur in the free
larva are much less limited than those which can occur in the
foetus. Secondary characters are therefore very numerous
in larvse, and there may even be larvae with secondary
characters only, as, for instance, the larvae of insects.

In spite of the liability of larvae to acquire secondary
characters, there is a powerful counter-balancing influence
tending towards the preservation of ancestral characters, in
that larvae are necessarily compelled at all stages of their
growth to retain in a functional state such systems of organs,
at any rate, as are essential for a free and independent
existence. It thus comes about that in spite of the many
causes tending to produce secondary changes in larvae, there
is always a better chance of their repeating, in an unabbre-
viated form, their ancestral history than is the case with
embryos, which undergo their development within the egg.

It may be further noted as a fact which favours the re-
lative retention by larvae of ancestral characters, that a
secondary larval stage is less likely to be repeated in de-
velopment than an ancestral stage, because there is always
a strong tendency for the former, which is a secondarily inter-
calated link in the chain of development, to drop out by the
occurrence of a reversion to the original type of development.

The relative chances of the ancestral history being pre-
served in the foetus or the larva may be summed up in the

LARVAL FORMS I THEIR NATURE^ ORIGIN AND AFFINITIES. 383

following way : — There is a greater chance of the ancestral
history being lost in forms which develop in the egg ; and
masked in those which are hatched as larvae.

The evidence from existing forms undoubtedly confirms
the a priori considerations just urged.^ This is well shown
by a study of the development of Echinodermata, Nemertea,
Molluscay Crustacea, and Tunicata, The free larvae of the
four first groups are more similar amongst themselves
than the embryos which develop directly, and since this
similarity cannot be supposed to be due to the larvae having
been modified by living under precisely similar conditions,
it must be due to their retaining common ancestral charac-
ters. In the case of the Tunicata the free larvae retain
much more completely than the embryos certain characters
which are known to have been ancestral.

Types of Larvae, — Although there is no reason to suppose
that all larval forms are ancestral, yet it seems reasonable to
anticipate that a certain number of the known types of larvae
should resemble the ancestors of the more important phyla
of the animal kingdom.

Before examining in detail the claims of various larvae to
such a character, it is necessary to consider somewhat more
at length the kind of variations which are most likely to
occur in larval forms.

It is probable a priori that there are two kinds of larval
forms, which may be distinguished as primary and secondary
larval forms. Primary larval forms are more or less modi-
fied ancestral forms, which have continued uninterruptedly
to develop as free larvae from the time when they constituted
the adult form of the species. Secondary larval forms are
those which have become introduced into the ontogeny
of species, the young of which were originally hatched with
all the characters of the adult ; but which, owing to the loss
of food-yolk in the egg, or some other cause, have become
hatched at an earlier period. Such secondary larval forms
may resemble the primary larval forms in cases where the
ancestral characters were retained by the embryo in its develop-
ment within the egg ; but in other instances the characters
they have are probably entirely adaptive.

* It has long been known that land and freshwater forms develop without
a metamorphosis mueh more frequently than marine forms. This is pro-
bably to be explained by the faet that there is not the same possibility of
a land or freshwater species extending itself over a wide area by the agency
of free larvsc, and there is, therefore, much less advantap in the existence of
such larval ; while the fact of such larvm being more liable to be preyed upon
than eggs, which are either concealed or carried about by the parent, might
render it absolutely disadvantageous for a species to have sucli larva.

384

F, M. BALFOUR.

Causes tending to produce secondartj changes in larm. —
The ways in which natural selection can act on larvae may
probably be divided more or less artificially into two classes.

1. The changes in development necessarily produced by
the existence of a larval stage.

The adaptive changes in a larva acquired in the ordi-
nary course of the struggle for existence.

The changes which come under the first head consist
essentially in a displacement in the order of deve-
lopment of certain organs. There is always a tendency in
development to throw back the differentiation of the embry-
onic cells into definite tissues till as late a date as possible.
This takes place in order to enable the changes of form,
which every organ undergoes in repeating even in an abbre-
viated way its phylogenetic history, to be effected with the
least expenditure of energy. Owing to this tendency it
comes about that when an organism is hatched as a larva
many of the organs are still in an undifferentiated state,
although the ancestral form which this larva represents
had all its organs fully differentiated. In order, however,
that the larva may be enabled to exist as an independent
organism certain sets of organs, e.g. the muscular, nervous,
and digestive systems, have to be histologically differentiated.
If the period of hatching becomes earlier, an earlier differen-
tiation of certain organs is a necessary consequence ; and in
almost all cases the existence of a larval stage causes a
displacement in order of development of organs, the com-
plete differentiation of many organs being retarded relatively
to the muscular, nervous, and digestive systems.

The possible changes under the second head appear to
be unlimited. There is, so far as I see, no possibJe reason
why an indefinite number of organs should not be developed
in larvae to protect them from their enemies, and to enable
them to compete with larvae of other species, and so on. The
only limit to such development appears to be the shortness of
larval life, which is not likely to be prolonged, since, ceteris
paribus j the more quickly maturity is reached the better it is
for the species.

A very superficial examination of marine larvae shows that
there are certain peculiarities common to most of them, and
it is important to determine how far such peculiarities are to
be regarded as adaptive. Almost all marine larvae are provided
with well-developed organs of locomotion, and transparent
bodies. These two features are precisely those which it
is most essential for such larvae to have. Organs of locomo-
tion are important, in order that larvae may be scattered as

LARVAL FORMS : THEIR NATURE^ ORIGIN AND AFFINITIES. 385

widely as possible, and so disseminate the species ; and trans-
parency is very important in rendering larvae invisible, and so
less liable to be preyed upon by their numerous enemies.^

These considerations, coupled with the fact that almost
all free-swimming animals, which have not other special
means of protection, are transparent, seem to show that, at
all events, the transparency of larvae is adaptive, and it is
probable that organs of locomotion are in many cases
specially developed, and not ancestral.

Various spinous processes on the larvae of Crustacea and
Teleostei are also examples of secondarily acquired protective
organs.

These general considerations are sufficient to form a basis
for the discussion of the characters of the known types of
larvae.

The following table contains a list of the more important
of such larval forms :

DiCYEMiDiE. — The Infusoriform larva.

PoEiEERA.— («) The Amphiblastula larva (fig. 1), with one half of the
body ciliated, and ‘the other half without cilia; (b) the oval uniformly
ciliated larva, which may be either solid or have the form of a vesicle.

CcELENTEEATA. — The plauula (fig. 2).

Turbellaria. — {a) The eight-lobed larva of Muller (fig. 9); {b) the
larvae of Gotte and Metsclmikoff, with some Pilidium characters.

Nemertea. — The Pilidium (fig. 8).

Trematoda. — The Cercaria.

lloTiEERA.— The Trochosphere-like larvae of Brachionus (fig. 3) and
Lacinularia.

Mollusca. — The Trochosphere larva (fig. 4), and the subsequent Veliger
larva (fig. 5).

Brachiopoda. — The three-lobed larva, with a postoral ring of cilia (fig. G).

PoLYZOi. — A larval form with a single ciliated ring surrounding tlie
mouth, and an aboral ciliated ring or disc (fig. 15).

CH.ETOPODA. — Various larval forms with many characters like those of
Molluscan Trochosphere frequently with distinct transverse bands of cilia.
They are classified as Atrochse, Mesotrochse, Telotrochse (fig. 12 a and fig.
13), Polytrochee, and Monotrochse (fig. 12 b).

Gepiiyrea Nuda. — Larval forms like those of preceding groups. A
specially characteristic larva is that of Echiurus (fig. 14).

Gephyrea Tubicola. — Actiuotrocha (fig. 17) ; with a postoral ciliated
ring of arms.

Myriapoda. — A functionally hexapodous larval form is common to all
the Chilognatha.

Insect A. — Various secondary larval forms.

Crustacea.— The Nauplius and the Zoeea.

Eouindermata. — The Auricularia (fig. 10 a), the Bipinnaria (fig. 10 b),
and the Pluteus (fig. 11), and the transversely-ringed larvae of Criuoidca

^ The phosphorescence of many larvm is very peculiar. I should have
anticipated that phosphorescence would have rendered them much more
liable to be captured by the forms which feed upon them ; and it is diflicult
to see of what advantage it can be to them.

386

F. M. BALFOUR.

(fig. 7). The Auricularia, the Bipinnaria, and the Pluteus can be reduced
to a common type (fig. 18 c).

Enteropnetjsta. — Tornaria (fig. 16).

Urochorda (Tunic at a). —The tadpole-like larva.

Ganoidea. — A larva with adhesive disc with papillae in front of the
mouth.

Anurous Amphibia. — The tadpole.

Of the larval forms included in the above list a certain
number are clearly without affinities outside the group to
which they belong. This is the case with the larvae of the
Myriapoda, the Crustacean larvae, and with the larval forms
of the Chordata. I do not propose to discuss the significance
of these forms in the present essay.

Fig. \.—Two free Stages in the development of Sycandra raphanns. (After
Schultze.) A. Amphiblastate stage. B. Stage after the ciliated cells
have commenced to be invaginated. c.s. segmentation cavity ; ec.
granular epiblast cells ; en. ciliated hypoblast cells.

There are, again, some larval forms which may possibly
turn out hereafter to be of importance, but from which, in
the present state of our knowledge, we cannot draw any
conclusions. The infusoriform larva of the Dicyemidge, and
the Cercaria of the Trematodes are such forms.

The meaning of the Amphiblastula larva was discussed
in a previous essay.

Excluding these and certain other forms, we have finally
left for consideration the larvae of the Coelenterata, the Tur-
bellaria, the Rotifera, theNemertea the Mollusca, the Polyzoa,
the Brachiopoda, the Chaetopoda, the Gephyrea, the Echino-
dermata, and the Enteropneusta.

The larvae of these forms can be divided into two groups.

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 387

The one group contains the larva of the Coelenterata or
Planula, the other group the larvse of all the other forms.

C

Fig. 2. — Three larval Stages of Eucope 'Polystyla. (After Kowalevsky.
A. Blastopliere stage with hypoblast spheres becoming budded into
the central cavity. B. Planula stage with solid hypoblast. C. Planula
stage with a gastric cavity, ep. epiblast ; hy. hypoblast ; al. gastric
cavity.

The Planula (fig. 2) is characterised by its extreme
simplicity. It is a two- layered organism, with a form vary-
ing from cylindrical to oval, and usually a radial symmetry.
So long as it remains free it is not even provided with a
mouth, and it is as yet uncertain whether or no the absence
of a mouth is to be regarded as an ancestral character.
The Planula is very probably the ancestral form of the
Coelenterata.

The larvee of almost all the other groups, although they
may be subdivided into a series of very distinct types, yet
agree in the possession of certain characters.^ There is a
more or less dome-shaped dorsal surface and a flattened or
concave ventral surface, containing the opening of the
mouth, and usually extending posteriorly to the opening of
the anus, when such is present.

The dorsal dome is continued in front of the mouth to
form a large prceoral lobe.

There is usually present at first a uniform covering of
cilia ; but in the later larval stages there are almost always
formed definite bands or rings of long cilia, by which loco-

' The larva of the Brachiopoda does not possess most of the characters
mentioned below. It is probably all the same, a highly dilferentiated larval
form belonging to this group.

388

F. M. BALFOUR.

motion is effected. These bands are often produced into
arm-like processes.

Cr

Fig. 3. — Embryo of Brachionns urceolaris, shortly before it is hatched.
(After Salensky.) m, mouth ; ms. masticatory apparatus ; me. mesen-
teron ; an. anus ; Id. lateral gland ; or. ovary ; t. tail, i.e. foot ; tr.
trochd disc ; sg. supra-cesophageal ganglion.

The alimentary canal has, typically, the form of a bent
tube with a ventral concavity, constituted (when an anus is
present) of three sections, viz. an oesophagus, a stomach, and

Fig. 4. — Diagram of an Bmbryo of Pleurobranchidium. (From Lankester.)
f. foot ; ot. otocyst ; m. mouth ; v. velum ; ng. nerve ganglion ;
ry. residual yolk spheres ; shs. shell-gland ; i. intestine.

a rectum. The oesophagus and rectum are epiblastic in
origin, while the stomach is derived from the hypoblast.

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 889

To the above characters may be added a glass-like trans-
parency ; and the presence of a widish space, often traversed
by contractile cells, between the alimentary tract and the
body wall.

Considering the very profound differences which exist
between many of these larvse it may seem that the characters
just enumerated are hardly sufficient to justify my grouping
of them together. It is, however, to be borne in mind that
my grounds for doing so depend quite as much upon the fact

Fig. 5. — Larva of CephalopTiorous Mollnsca in the veliger stage. (From
Gegenbaur.) A. and B. Earlier and later stage of Gasteropod.
C. Pteropod (Cymbulia). v. velum ; c. shell; p. foot ; op. operculum;
tentacle.

that they constitute a series without any great breaks in it,
as upon the existence of characters common to the whole of
them. It is also worth noting that most of the characters

B

C

c

Fig. G. — Larva of Argiope. (From Gegenbaur, after Kowalevsky.)
m. mantle ; b, seta; ; d. archentcron.

390 F. M. BALFOUR.

which have been enumerated as common to the whole of
these larvae are not such secondary characters as (in accord-
ance with the considerations used above) might be expected
to arise from the fact of their being subjected to nearly
similar conditions of life. Their transparency is, no doubt,
such a secondary character, and it is not impossible that

I'm. 7. — Three Stages in the development oj Antedon {Comatula). (From
Lubbock, after Thompson.) A. Larva just batched. B. Larva with
rudiment of the calareous plates. C. Pentacrinoid larva,

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 391

the existence of ciliated bands may be so also; but it is
more probable that if, as I suppose, these larvae reproduce
the characters of some ancestral form, this form may have
existed at a time when all marine animals were free-swim-
ming, and that it may, therefore, have been provided with
at least one ciliated band.

The detailed consideration of the characters of these
larvae, given below, supports this view.

This great class of larvae may, as already stated, be divided
into a series of minor subdivisions. These subdivisions are
the following :

1. The Pilidium Group. — This group is characterised
by the mouth being situated nearly in the centre of the
ventral surface, and by the absence of a proctodaeum. It
includes the Pilidium of the Nemertines (fig. 8), and the
various larvae of marine Dendrocoela (fig. 9) . At the apex of
the praeoral lobe a thickening of epiblast may be present,
from which (fig. 19) a contractile cord sometimes passes to
the oesophagus.

Fig S. — Ttoo Stages in the development of Pilidium. (After Metsclmikoff.)
ae. arclienteron ; oe. cesophagus ; st. stomach ; am. amnion ; pr.d. pro-
stomial disc ; po.d. metastomial disc ; c.s. cephalic sack.

2. The Echinoderm Group. — This group(figs. 10, 11, and
18 c) is characterised by the presence of a longitudinal postoral
band of cilia, the absence of special sense organs in the praeoral
region, and the development of the body-cavity as an out-
growth of the alimentary tract. The three typical divisions
of the alimentary tract are present, and there is a more

392

F. M. BALFOUR.

or less developed prseoral lobe. This group only includes
the larvae of the Echinodermata.

3. The Trochosphere Group. — This group (figs. 12, 13)
is characterised by the presence of a praeoral ring of long
cilia, the region in front of which forms a great part of the
prseoral lobe. The mouth opens immediately behind the
praeoral ring of cilia, and there is very often a second ring of
short cilia parallel to the main ring, immediately behind the
mouth. The function of the ring of short cilia is nutritive,
in that the cilia are employed in bringing food to the mouth ;
while the function of the main ring is locomotive. A peri-
anal patch or ring of cilia is often present (fig. 12 a), and in
many forms intermediate rings are developed between the
prseoral and perianal rings.

The prseoral lobe is usually the seat of a special thicken-
ing of epiblast, which gives rise to the supra-oesophageal

A.

Fig. 9. — A. Larva of Eur^lepia auriculata immediately after hatching.
Viewed from the side. (After Hallez.) m. mouth. B. Muller's Tur-
bellarian Larva {probably Thysanozoon). Viewed from the ventral
surface. (After Miiller.) The ciliated baud is represented by the
black line. m. mouth ; u.l. upper lip.

ganglion of the adult. On this lobe optic organs are very
often developed in connection with the supra-oesophageal
ganglion, and a contractile band frequently passes from this
region to the oesophagus.

The alimentary tract is formed of the three typical
divisions.

The body-cavity does not develop directly as an outgrowth
of the alimentary tract, though the process by which it
originates is very probably secondarily modified from a
pair of alimentary outgrowths.

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 393

Paired excretory organs opening to the exterior and into
the body cavity are often present.

A B

Fig. 10. — A. The Larva of a Holothuroid. B. The Larva of ati Asterias,
m, mouth ; st. stomach ; a. anus ; l.c. primitive longitudinal ciliated
band ; pr.c. praeoral ciliated band.

This type of larva is found in the Potifera (fig. 3) (where it
is preserved in the adult state), the Chsetopoda, theMollusca
(fig. 4), the Gephyrea nuda (fig. 14), and the Polyzoa
(fig.l5).i

4. Tornaria. — This larva (fig. 16) is intermediate in
most of its characters between the larvae of the Echinoder-
mata (more especially the Bipinnaria) and the Trochosphere.
It agrees with Echinoderm larvae in the possession of a
longitudinal ciliated band (divided into a praeoral and a
postoral ring), and in the derivation of the body-cavity and
water-vascular vesicle from alimentary diverticula ; and it
resembles the Trochosphere in the presence of sense organs

Fig. 11. — A Larva of Strongylocentrus. (From Agassiz.) mouth; a.
anus ; o. oesophagus ; d. stomach ; c. intestine ; v'. and v. ciliated
ridges ; w. water- vascular tube ; r. calcareous rods.

* For a discussion as to the structure of the Polyzoou larva, vide
‘ A Treatise on Comparative Embryology,’ vol. i, p. 253.

394 F. M. BALFbUR.

on the praeoral lobe, in the existence of a perianal ring of
cilia, and in the possession of a contractile band passing
from the praeoral lobe to the oesophagus.

5. Actinotrocha. — The remarkable larva of Phoronis,
(fig. 17) known as Actinotrocha, is characterised by the
presence of (1) a postoral and somewhat longitudinal ciliated
ring produced into tentacles, and (2) a perianal ring. It
is provided with a praeoral lobe, and a terminal or some-
what dorsal anus.

6. The larva of the Brachiopoda articulata (fig. 6).

The relationships of the six types of larval forms just
characterised have been the subject of a considerable
amount of controversy, and the following suggestions on
the subject must be viewed as somewhat speculative. The
Pilidium type of larva is in some important respects less
highly differentiated than the larvse of the five other groups.
It is, in the first place, without an anus ; and there are no
grounds for supposing that the anus has become lost by
retrogressive changes. If for the moment it is granted that
the Pilidium larva represents more nearly than the larvae of
the other groups the ancestral type of larva, what characters
are we led to assign to the ancestral form which this larva
repeats ?

Fig. 12. — Two Chcetopod Larva.^ (From Gegenbaur.) o. mouth ; i. intes-
tine ; a. anus ; v. prseoral ciliated band ; w. perianal ciliated band.

In the first place, this ancestral form, of which fig. 18 A is
an ideal representation, would appear to have had a dome-
shaped body, with a flattened oral surface and a rounded
aboral surface. Its symmetry was radial, and in the centre
of the flattened oral surface was placed the mouth, and round
its edge was a ring of cilia. The passage of a Pilidium-
like larva into the vermiform bilateral Platyelminth form,
and therefore it may be presumed of the ancestral form which
this larva repeats, is effected by the larva becoming more
elongated, and by the region between the mouth and one

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 395

end of the body becoming the prseoral region, and that
between the mouth and the opposite end developing into

Fig. 13. — Polygordius Larva. (After Hatschek.) mouth; 5^. supra-
oesophageal ganglion ; nph. nephridion ; me.p. mesoblastic band ;
an. anus ; ol. stomach.

the trunk, an anus becoming placed at the extremity of
the trunk in the higher forms.

If what has been so far postulated is correct, it is clear
that this primitive larval form bears a very close resemblance
to a simplified free-swimming Coelenterate (Medusa), and that
the conversion of such a radiate form into the bilateral
took place, not by the elongation of the aboral surface, and
the formation of an anus there, but by the unequal elonga-
tion of the oral face, an anterior part forming a preeoral
lobe, and a posterior part the trunk ; while the aboral
surface became the dorsal surface.

This view fits in very well with the anatomical resem-

liG. 14. Larva of Echiurus. (After Salcnsky.) mouth ; anus
sfj. supra-oesophageal ganglion (?).

blanccs between the Coelenterata and the Turbellaria,^ and

* Vide ‘ A Treatise on Comparative Embryology,’ vol. i, p. 148 and 158.
In this connection attention may be called to Coleoplana Metschnikowei,

396

F. M. BALFOUR.

shows, if true, that the ventral and median position of the
mouth in many Turbellaria is the primitive one.

The above suggestion as to the mode of passage from the
radiate into the bilateral form differs entirely from t hat
usually held. Lankester,^ for instance, gives the following
account of this passage :

Pig. 15. — Diagram of a Larva of the^ Dolyzoa. m. mouth ; an. anus
st. stomach ; s. ciliated disc.

It has been recognised by various writers, but notably
by Gegenbaur and Haeckel, that a condition of radiate
symmetry must have preceded the condition of bilateral
symmetry in animal evolution. . The Diblastula may be
conceived to have been at first absolutely spherical with
spherical symmetry. The establishment of a mouth lead
necessarily to the establishment of^a structural axis passing
through the mouth, around, which axis the body was
arranged with radial symmetry. This condition is more or
less perfectly maintained , by many Coelenterates, and is
reassumed by degradation of higher forms (Echinoderms,
some Cirrhipedes, some Tunicates). The next step is the
differentiation of an upper and a lower surface in relation to
the horizontal position, with mouth placed anteriorly,
assumed by the organism in locomotion. With the differ-
entiation of superior and inferior surface, a right and a left
side, complimentary one to the other, are necessarily also
differentiated. Thus the ‘organism becomes bilaterally
symmetrical. The Coelentera are not wanting in indications
of this bilateral symmetry, but for all other higher groups
of animals it is a fundamental character. Probably the
development of a region in front of, and dorsal to the mouth,
forming the Prostomium, was accomplished pari passu with
the development of bilateral symmetry. In the radially

a form described by Kowalevsky, ‘ Zoologischer Anzeiger,’ No. 52, p. 140,
as being intermediate between the Ctenophora and the Turbellaria. There
does not, however, appear to me to be sufficient evidence to prove that this
form is not merely a creeping Ctenophor.

^ ‘Quart. Journ. of Micr, Science,’ vol. xvii, pp. 422-3.

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 397

symmetrical Coelentera we find very commonly a series of
lobes of the body-wall or tentacles produced equally — with

Fig. 16. — Two Stages in the development of Tornaria. (After Metschnikoflf.)
The black lines represent the ciliated bands, m. mouth ; an. anus ;
br. branchial cleft ; ht. heart ; c. body-cavity between splanchnic and
somatic mesoblast layers ; w. so-called water-vascular vesicle ; v. cir-
cular blood-vessel.

radial symmetry, that is to say — all round the mouth, the
mouth terminating the main axis of the body — that is to say,
the organism being ^ telostomiate.' The later fundamental
form, common to all animals above the Coelentera, is attained
by shifting what was the main axis of the body — so that it
may be described now as the ‘ enteric’ axis ; whilst the
new main axis, that parallel with the plane of progression,
passes through the dorsal region of the body running ob-
liquely in relation to the enteric axis. Only one lobe or
outgrowth of those radially disposed in the telostomiate
organisms now persists. This lobe lies dorsally to the
mouth, and through it runs the new main axis. This lobe
is the Prostomiunij and all the organisms which thus
develop a new main axis, oblique to the old main axis, may
be called prostomiate.”

It will be seen from this quotation that the aboral part of
the body is supposed to elongate to form the trunk, while the
prseoral region is derived from one of the tentacles.

VOL. XX. NEW SER. D D

398

F. M. BALFOUR.

Before proceeding to further considerations as to the origin
of the Bilateralia, suggested by the Pilidium type of larva,

it is necessary to enter into a more detailed comparison be-
tween our larval forms.

A very superficial consideration of the characters of these
forms brings to light two important features in which they
differ, viz. :

(1) In the presence or absence of sense organs on the
prgeoral lobe.

(2) In the presence or absence of outgrowths from the
alimentary tract to form the body-cavity.

The larvse of the Echinodermata and (?) Actinotrocha are
without sense organs in the praeoral lobe, while the other
types of larvse are provided with them. Alimentary diverticula
are characteristic of the larvse of the Echinodermata and of
Tornaria.

If the conclusion already arrived at, to the effect that the
prototype of the six larval groups was descended from a
radiate ancestor, is correct, it appears to follow that the
nervous system, in so far as it was differentiated, had primi-
tively a radiate form, and it is also probably true that there
were alimentary diverticula in the form of radial canals, (wo
of which may have given origin to the paired diverticula
which become the body-cavity in such types as the Echino-
dermata. If these two points are granted, the further con-
clusions seem to follow — (1) that the ganglion and sense
organs of the preeoral lobe were secondary structures which
arose (perhaps as differentiations of the original circular
nerve-ring) after the assumption of a bilateral form ; and [2)
that the absence of these organs in the larva? of the Echino-

LARVAL FORMS : THEIR NATURE^ ORIGIN AND AFFINITIES* 399

dermata, and Actinotrocha (?), implies that these larvse retain,
SO far, more primitive characters than the Pilidium. The
same may be said of the alimentary diverticula. There are

Fig. 18. — T/iree Diagrams representing the ideal Evolution of various Larval
forms. A. Ideal ancestral larval form. B. Trochosphere larva.
C. Ecbinoderm larva, m. mouth ; an. anus ; st. stomach ; s.g. supra-
oesophageal ganglion. The black lines represent the ciliated bands.

thus indications that in two important points the Echino-
derm larvse are more primitive than the Pilidium.

The above conclusions, with reference to the Pilidium and
Echinoderms, involve some not inconsiderable difficulties, and
suggest certain points for further discussion.

In the first place it is to be noted that the above specu-
lations render it probable that the type of nervous system
from which that found in the adults of the Echinodermata,
Platyelminthes, Chsetopoda, Mollusca, &c., is derived, was a
circumoral ring, like that of Medusse, with which radially
arranged sense organs may have been connected ; and that in
the Echinodermata this form of nervous system has heen
retainedy while in the other types it has been modified by the
anterior part, having given rise to supra-oesophageal ganglia
and organs of vision; which were formed owing to the assump-
tion of a bilateral symmetry, and the consequent necessity for
the sense organs to be situated at the anterior end of the body.
If this view is correct, the question arises as to how far the
posterior part of the nervous system of the Bilateralia can
be regarded as derived from the primitive radial ring.

A circumoral nerve-ring, if longitudinally extended, might
give rise to a pair of nerve-cords united in front and behind —
exactly such a nervous system, in fact, as is present in

400

F. M. BALFOUR.

many Nemertines^ (the Euopla and Pelagonemertes), in
Peripatus,^ and in primitive molluscan types (Chiton, Fis-
surella, &c.). From the lateral parts of this ring it is easy
to derive the ventral cord of the Chsetopoda and Arthro-
poda. It is especially deserving of notice, in connection
with the nervous system of Nemertines and Peripatus, that
the commissure connecting the two nerve-cords behind is
placed on the dorsal side of the intestine. As is at once
obvious, by referring to the diagram (fig. 18), this is the
position this commissure ought, undoubtedly, to occupy if
derived from part of a nerve-ring which originally followed
more or less closely the ciliated edge of the body of the
supposed radial ancestor.

The fact of this arrangement of the nervous system being
found in so primitive a type as the Nemertines tends to
establish the views for which I am arguing ; the absence or
imperfect development of the two longitudinal cords in Tur-
bellarians may very probably be due to the posterior part of
the nerve-ring having atrophied in this group.

It is by no means certain that this arrangement of the
nervous system in some Mollusca and in Peripatus is primi-
tive, though very probably it may be so.

In the larvse of the Turbellaria^the development of sense
organs in the praeoral region is very clear (fig. 9 b); but this
is by no means so clear in the case of the true Pilidium.
There is in Pilidium (fig. 19 a) a thickening of epiblast
at the summit of the dorsal dome, which might seem,
from the analogy of Mitraria, &c. (fig. 20), to correspond to
the thickening of the prseoral lobe, which gives rise to the
supra-oesophageal ganglion ; but, as a matter of fact, this
part of the larva does not apparently enter into the formation
of the young Nemertine (fig. 19). The peculiar metamor-
phosis which takes place in the development of the Nemer-
tine out of the Pilidium,^ may, perhaps, eventually supply
an explanation of this fact ; but at present it remains as a
not yet explained difficulty.

The position of the fiagellum in Pilidium, and of the supra-
oesophageal ganglion in Mitraria, suggests a different view of
the origin of the supra-oesophageal ganglion to that adopted
above. The position of the ganglion in Mitraria corresponds

^ * Vide Hubrecht, ‘ Zur Anat. and Phys. d. Nerven System d. Nemer-
tinen Kon. Akad.,’ Wiss., Amsterdam; and “Researches on the Nervous
System of Nemertines,” ‘ Journ. of Micr. Science,’ 1880.

’ Vide Seif, “ On some points in the Anat. of Peripatus capensis'^ ‘ Quart.
Journ. of Micr. Science,’ vol. xix, 1879.

’ Vide ‘ A Treatise on Comparative Embryology,’ vol. i, p. 169.

LARVAL FORMS .* THEIR NATURE, ORIGIN AND AFFINITIES. 401

Fig. 19. — A. Pilidium with an advanced Nemertine Worm. B. Ripe Embryo
of Kemcrtes in the position it occupies in Pilidium. (Both after
liutschli.) ce. oesophagus ; st. stomach ; i. intestine ; pr. proboscis ;
Lp. lateral pit ; an. amnion ; n. nervous system.

common origin. If this view is the correct one, we must
suppose that the apex of the aboral lobe has become the
centre of the prieoral field of the Pilidium and Trochosphere
larval forms. ^ The whole of the questions concerning the
nervous system are still very obscure, and until further facts
are brought to light no definite conclusions can be arrived at.

The absence of sense organs on the praeoral lobe of larval
Echinoderrnata, coupled with the structure of the nervous

‘ The independent development of the supra-cesopbageal ganglion and
ventral nerve cord in Chaetopoda Kleinenberg, * Development of
bricus trapezoid^s ’) suits this view very satisfactorily.

closely with that of the auditory organ in Ctenophora ; and
it is not impossible that the two structures may have had a

402

F. M. BALFOUR.

system of the adult, points to the conclusion that the adult
Echinodermata hme retained^ and not, as is now usually
held, secondarily acquired, their radial symmetry ; and if
this is admitted it follows that the obvious bilateral sym-
metry of Echinoderm larvae is a secondary character.

The bilateral symmetry of many Coelenterate larvae (the
larva of ^ginopsis, of many Acraspeda, of Actinia, &c.),
coupled with the fact that a bilateral symmetry is obviously
advantageous to a free-swimming form, are quite sufficient to
show that this supposition is by no means extravagant; while
the presence of only tw^o alimentary diverticula in Echino-
derm larva is quite in accord with the presence of a single
pair of perigastric chambers in the early larva of Actinia,
though it must be admitted that the derivation of the water-
vascular system from the left diverticulum is not easy to
understand on this view.

Fig. 20. — Two Stages in the development oj Mitraria. (After Metsclini-
koff.) m. mouth ; an. anus ; sg. supra-oesophageal ganglion ; br. and
b. provisional bristles ; pr.b. prseoral ciliated band.

A difficulty in the above speculation is presented by the
fact of the anus of the Echinodermata being the permanent
blastopore, and arising prior to the mouth. If this fact has
any special significance, it becomes difficult to regard the larva
of Echinoderms and that of the other types as in any w^ay
related ; but if the view's already urged, in a previous essay on
the germinal layers, as to the unimportance of the blastopore,
are admitted, the fact of the anus coinciding with the blasto-

LARVAL FORMS ; THEIR NATURE^ ORIGIN AND AFFINITIES. 403

pore ceases to be a difficulty. As may be seen, by referring
to fig. 18 c, the anus is placed on the dorsal side of the
ciliated band. This position for the anus adapts itself to the
view that the Echinoderm larva had originally a radial sym-
metry, with the anus placed at the aboral apex, and that the
present terminal position of the anus arose with the elonga-
tion of the larva on the attainment of a bilateral symmetry.

It may be noticed that the obscure points connected with
the absence of a body-cavity in most adult Platyelminthes,
which have already been dealt with in my essay on the
germinal layers, crops up again here ; and that it is necessary
to assume either that alimentary diverticula, like those in
the Eohinodermata, were primitively present in the Platy-
elminthes, but have now disappeared from the ontogeny of
this group, or that the alimentary diverticula have not
become separated from the alimentary tract.

So far the conclusion has been reached that the archi-
type of the six types of larvae was a radiate form, and that
amongst existing larvae it is most nearly approached in
general shape and in the form of the alimentary canal by
the Pilidium group, and in certain other particulars by the
Echinoderm larvae.

The edge of the oral disc of the larval architype was pro-
bably armed with a ciliated ring, from which the ciliated
ring of the Pilidium type and of the Echinodermata were most
likely derived. The ciliated ring of the Pilidium varies
greatly in its characters, and has not always the form of a
complete ring. In Pilidium proper (fig. 19 a) it is a simple
ring surrounding the edge of the oral disc. In Miiller’s
larva of Thysanozoon (fig. 9 b) it is inclined at an axis to
the oral disc, and might be called praeoral, but such a term
cannot be properly used in the absence of an anus.

The Echinoderm ring is oblique to the axis of the body,
and, owing to the fact of its passing ventrally in front of the
anus, must be called postoral.

The next point to be considered is the affinities of the
other larval types to these two types.

The most important of all the larval types is the Tro-
chosphere, and this type is undoubtedly more closely related
to the Pilidium than to the Echinoderm larva. Mitraria
amongst the Chsctopods (fig. 20) retains, indeed, the form
of Pilidium very closely, and mainly differs from a Pilidium
in the possession of an anus and of provisional bristles ;
the same may be said of Cyphonautes (fig. 21) amongst the
Polyzoa.

The existence of these two forms appears to show tliat the

404

r. M. BALFOUR.

prseoral ciliated ring of the Trochosphere may very probably
be derived directly from the circumoral ciliated ring of the
Pilidium ; the other ciliated rings or patches of the Trocho-
sphere having a secondary origin.

Fig. 21. — Cyphonautes {Larva of Memhraiii'pora). (After Hatschek). m.
mouth ; a, anus ; f.g. foot gland ; x. problematical body (probably a
bud).

The larva of the Brachiopoda x(fig. 6), in spite of its
peculiar character, is, in all probability, more closely related
to the Chsetopod Trochosphere than to any other larval type.
The most conspicuous point of agreement between them is,
however, the possession in common of provisional setae.

Echinoderm larvae differ from the Trochosphere, not only
in the points already alluded to, but in the character of the
ciliated band. The Echinoderm band is longitudinal and
postoral. As just stated, there is reason to think that the
praeoral band of the Trochosphere and the postoral band of
the Echinoderm larva are both derived from a ciliated ring
surrounding the oral disc of the prototype of these larvae
{vide fig. 8). In the case of the Echinodermata the anus
must have been formed on the dorsal side of this ring, and
in the case of the Trochosphere on the ventral side ; and so
the difference in position between the two rings was brought
about. Another view with reference to these rings has been
put forward by Gegenbaur and Lankester, to the effect that
the praeoral ring is derived from the breaking up of the single
band of most Echinoderm larvae into the two bands found
in Bipinnaria {vide fig. 10). There is no doubt a good deal
is to be said for this origin of the praeoral ring, and it is
strengthened by the case of Tornaria; but the view adopted
above appears to me more probable.

LARVAL FORMS ; THEIR NATURE, ORIGIN AND AFFINITIES. 405

Actinotrocha (fig. 17) undoubtedly agrees more closely with
Echinoderm larvse than with the Trochosphere. Its ciliated
ring has the same character, and the growth along the line
of the ciliated ring of a series of arms is very similar^ to
what takes place in many Echinoderms. Its affinity with
the Echinoderm larvae is also shown in the absence of
sense organs on the praeoral lobe.

Tornaria (fig. 16) cannot be definitely united either with
the Trochosphere or with the Echinoderm larval type. It
has important characters in common with both of these
groups, and the mixture of these characters renders it a
very striking and well-defined larval form.

Phylogenetic conclusions. — The phylogenetic conclusions
which follow from the above views remain to be dealt with.
The fact that all the larvae of the groups above the Coelen-
terata can be reduced to a common type points to all the
higher groups being descended from a single stem.

Considering that the larvae of comparatively few groups
have persisted, no conclusions as to affinities can be drawn
from the absence of a larva ; and the presence in two groups
of a common larval form may be taken as proving a
common descent, but does not necessarily show any close
affinity.

There is everyreason to believe that the types with aTrocho-
sphere larva, viz. the Rotifera, the Mollusca, the Chaetopoda,
the Gephyrea, and the Polyzoa, are descended from a common
ancestral form, and it is also fairly certain there was a remote
ancestor common to these forms and to the Platyelminthes.
A general affinity of the Brachiopoda with these types
very probable. All these types, together with any other
types which can be proved to be related to them, are des-
cended from a bilateral ancestor. The Echinodermata, on the
other hand, are probably directly descended from a radial
ancestor, and have more or less completely retained their
radial symmetry. How far Actinotrocha' is related to the
Echinoderm larva3 cannot be settled. Its characters may
possibly be secondary, like those of the mesotrochal larvee
of Chaetopods, or they may be due to its having branched off
very early from the stock common to the whole of the forms
above the Coelenterata. The position of Tornaria is still
more obscure. It is difficult, in the face of the peculiar
water-vascular vesicle with a dorsal pore, to avoid the con-
clusion that it has some affinities with the Echinoderm
larvae. Such affinities would seem, on the lines of specula-

‘ It is very probable tliat Phoronis is iu no way related to the other
Gephyrea,

406

F. M. BALFOUR.

tion adopted in this essay, to prove that its affinities to
the Trochosphere, striking as they are, are secondary and
adaptive. From this conclusion, if justified, it would follow
that the Echinodermata and Enteropneusta have a remote
ancestor in common, but not that the two groups are in any
other way related.

General conclusions and summary, — Starting from the
demonstrated fact that the larval forms of a number of widely
separated types above the Coelenterata have certain characters
in common, an attempt has been made(l) to determine the
characters of the common prototype of all these larvae, (^) and,
the mutual relations of the larval forms in question. This
attempt started wfith certain more or less plausible sugges-
tions, the truth of which can only be tested by the coherence
of the results which follow from them, and their capacity to
explain all the facts.

The results arrived at may be summarised as follows :

1. The larval forms above the Coelenterata may be divided
into six groups enumerated on pages 391 — 394.

2. The prototype of all these groups was an organism
something like a Medusa, with a radial symmetry. The
mouth was placed in the centre of a flattened ventral sur-
face. The aboral surface was dome-»shaped. Round the edge
of the oral surface was a ciliated ring, and probably a nervous
ring provided with sense organs. The alimentary canal was
prolonged into two or more diverticula, and there was no
anus.

3. The ^bilaterally symmetrical types were derived from
this larval form by the larva becoming oval, and the region
in front of the mouth forming a praeoral lobe, and that
behind the mouth the trunk. The aboral dome became
the dorsal surface.

On the establishment of a bilateral symmetry the anterior
part of the nervous ring gave rise to the supra*oesophageal
ganglia, and the optic organs connected with them. The
body-cavity was developed from two of the primitive alimen-
tary diverticula.

The usual view that radiate forms have become bilateral
by the elongation of the aboral dome into the trunk is pro-
bably mistaken.

4. Pilidium is the larval form which most nearly repro-
duces the characters of the larval prototype in the course of
its conversion into a bilateral form.

5. The Trochosphere is a completely differentiated bilateral
form, in which an anus has become developed. The praeoral
ciliated ring of the Trochosphere is probably derived from

LARVAL FORMS : THEIR NATURE, ORIGIN AND AFFINITIES. 407

the ciliated ring of Pilidium, which is itself the original ring
of the prototype of all these larval forms.

6. Echinoderm larvae in the absence of a nerve ganglion or
special organs of sense on the praeoral lobe, and in the pre-
sence of alimentary diverticula, which give rise to the body
cavity, retain some characters of the prototype larva which
have been lost in Pilidium. The ciliated ring of Echinoderm
larvae is probably derived directly from that of the proto-
type by the formation of an anus on the dorsal side of the
ring. The anus was very probably originally situated at the
aboral apex.

Adult Echinoderms have probably retained the radial sym-
metry of the forms from which they are descended, and their
nervous ring is probably directly derived from that of their
ancestors. They have not, as is usually supposed, secondarily
acquired their radial symmetry. The bilateral symmetry of
the larva is, on this view, secondary, like that of so many
Coelenterate larvse.

5. The points of similarity between Tornaria and (1) the
Trochosphere and (2) the Echinoderm larvse are probably
adaptive in the one case or the other ; and, while there is no
difficulty in believing that those with the Trochosphere are
adaptive, the presence of a water-vascular vesicle with a
dorsal pore renders probable a real affinity with Echinoderm
larvge.

6. It is not possible in the present state of our knowledge
to decide how far the resemblances between Actinotrocha and
Echinoderm larvae are adaptive or primary.

The majority of these conclusions are undoubtedly of a
highly speculative character, but while they cannot be
regarded as part of our stock of embryological knowledge,
they may, nevertheless, serve to indicate an important line
for continued embryological research. A thorough histo-
logical investigation of the larval forms dealt with in this
essay will be likely to lead to valuable results.

408

ALFRED W. BENNETT.

Oyi the Classification of Cryptogams. ^ By Alfred W.
Bennett, M.A., B.Sc., F.L.S., Lecturer on Botany at
St. Thomases Hospital. 2.

The classification of Cryptogams proposed in the fourth
edition of Sachs's ‘ Lehrbuch der Botanik' — differing in
some important points from that in the third edition, from
which the English translation has been made — is an un-
doubted index of a considerable advance in our knowledge
of the relationships of the lower forms of flowerless plants.
That it will in some points be modified as our knowledge
further advances, its author would be the last to dispute;
there are some other points to which I wish to call attention
in the present paper, in which that distinguished botanist
does not appear to me to have been so successful in pointing
out natural affinities.

The first group, the Thallophytes (including Characese), is
divided by Sachs into four classes of equal rank, the Proto-
phyta, Zygosporese, Oosporeae, and Carposporeae, abandoning,
as a primary classification, the time-honoured distinction of
Algae and Fungi. Few will be disposed to dispute the
desirability of setting off the lowest forms from both the
Algae and Fungi into a distinct class; and the Protophyta
may with advantage be retained with the limits defined by
Sachs. But, beyond this, it is doubtful whether the pro-
posed change represents a nearer approach to a true phylo-
genetic scheme. The supporters of the new classification
point to that of Phanerogams as affording them powerful
support. They say, with truth, that genera destitute of
chlorophyll, like Orobanche, Balanophora, &c., are not
separated into a distinct class of primary rank. But the
cases are not parallel. Without going into the question of
- whether these parasitic Phanerogams do contain chlorophyll
or not, it may be pointed out that they are mostly isolated
genera closely resembling well-known natural orders in all
important points of structure. Were flowering plants divi-
sible into two great series, one containing chlorophyll and
autonomous, the other destitute of chlorophyll and parasitic,
with very few points of resemblance or connecting links
between the two, it is at least doubtful whether this would

^ In the preparation of this paper I have had the most valuable assist-
ance and co-operation of Mr. George Murray, my collaborator in our forth-
coming ‘ Handbook of Cryptogamic Botany.’

2 at the Swansea Meeting of the British Association, Aug. 26tli,
1880.

CLASSIFICATION OF CRYPTOGAMS.

409

not be taken into account in the first line by systematists.
The old division into Fungi and Algae is so engrafted into
popular conceptions, is so convenient to the student, and
rests at the same time, with but few exceptions, on such
well-marked physiological characters, that its abandonment
requires the clearest evidence of the necessity for such a
course, as being obviously in harmony with genetic affinities.
But few will, we think, be bold enough to maintain that
Spirogyra is more nearly allied to Mucor than it is to (Edo-
gonium, or that the Florideae display more affinity to the
Basidiomycetes than they do to the Fucaceae.

The present writer has already stated in detail (^Journal of
Botany,’ 1878, p. 202) his reasons for considering that it is
impossible to place the Characeae among Thallophytes, and
still more among the Carposporeae.

As regards minor points, I am disposed to consider the
anomalous Myxomycetes as a low type of structure, scarcely
raised above the Protophyta, and not exhibiting true sexual
conjugation. The researches of Carpenter, Cohn, and others,
clearly show that Volvox and its allies must be placed among
the Oosporeae rather than the Zygosporeae. The Phaeospore ae
should certainly be separated off as a distinct order from the
Fucaceae.

I propose, therefore, to divide Thallophytes into three
primary classes: — Protophyta, Fungi, and Algje. The
Protophyta are divisible into two sub-classes, Protomycetes
and Protophyce(B. The Protomycetes consist of a single
order, the Schizomycetes, of which Saccharomyces is regarded
as an aberrant form. The Protophyceae are composed of the
Protococcaceae (including Palmellaceae and Scytonemeae),
Nostocaceae, Oscillatorieae, and Rivularieae. The Myxomy-
cetes are treated as a supplement to the Protophyta. The
Fungi are made up of three sub-classes, employing in the
main the same characters as Sachs ; for reasons, however,
which are stated in another paper presented to the Associa-
tion, the syllable sperm ” is used instead of spore ” in
their terminology. The first division, the Zygomycetes (or
Zygospermece achlorojyhyllacece)^ is composed of the Mucorini
only (including the Piptocephalidae). The second, the
Oomycetes (or Oospermece achlorophyllacece)^ comprises the
Peronosporeae and Saprolegnieae (including the Chytri-
diaceae). The third, the Carpomycetes (or Carpospermece
acliloropliyllacece)^ is made up of the Uredineae, Ustilagineae,
Basidiomycetes, and Ascomycetes, the Lichencs being in-
cluded in the last as a sub-order. The Algae are arranged
under three corresponding sub-classes. The Zygophyceco (or

410

ALFRED W. BENNETT.

Zygosper 7716(2 chloy'opliyllaceos) is made up of the following
orders : — Pandorinese, Hydrodictyese, Confervaceae (under
which the Pithophoracese may possibly come); Ulotrichaceae;
Ulvaceae; Botrydieae, and Conjugatae (the last comprising the
Diatomaceae. Desmidieae; Zygnemaceac; and Mesocarpeae). The
Oophyce(B (or Oosperynea chlorophyllacecE) include the Volvo-
cineaS; Siphoneae (with the nearly allied Dasycladeae); Sphae-
ropleaceae; Oedogoniaceae; Fucaceae, and Phaeosporeae. The
CarpophycecB (or Carpospermece cMoropliyllacece) is made up
of the Coleochaeteae and Floride^.

The Characeae will he restored to their old rank as a
group of primary importance. The MusciNEyE are un-
changed, comprising the Hepaticce and Musci (including
Sphagnaceae).

In classifying the Vascular Cryptogams, I am com-
pelled again to dissent from Sachs’s most recent proposal.
Abandoning the primary division into isosporous and hete-
rosporous forms, he proposes in their place three classes, of
which two are newly suggested — Equisetaceae, Filicineae, and
Dichotomi; the two last including both isosporous and hete-
rosporous forms, the Rhizocarpeae being placed under Fili-
cineae, and the Selaginellaceae under Dichotomi. The sexual
differentiation of the spores into those which produce female
prothallia and those which produce antherozoids, seems to
me to lie so much deeper than any characteristics derived
merely from vegetative characters, and to mark so distinctly
a step upwards in the scale towards flowering plants, that
I have no hesitation in restoring the primary division of
vascular cryptogams into Isosporia and Heterosporia. As a
minor point, it hardly seems logical to separate the Ophio-
glossacese as a distinct order from Filices, if the Marattiaceae
are to remain associated wdth them. The Isosporia will
then consist of the Filices (including Ophioglossaceae), Ly-
copodiaceae, and Equisetaceae. The Heterosporia will comprise
the Rhizocarpeae and Selaginellaceae.

In describing the heterosporous vascular cryptogams it is
almost universal to speak of the spores which give rise to the
female prothallium and those which give birth to anthero-
zoids as macrospores” and microspores” respectively.
The first of these terms is doubly objectionable ; firstly, ety-
mologically, the correct meaning of juLaKpoQ being not large,
but long ; and, secondly, from the close similarity in
sound of the two terms, an inconvenience, especially in oral
instruction, which every teacher must have experienced.
Seeing that the correct and far preferable terms inegaspore
and microspore are used by Berkeley, Areschoug, and others.

CLASSIFICATION OP CRYPTOGAMS.

411

it is difficult to understand how macrospore” can ever
have got into general use.^ But those terms are always to
be preferred which convey to the student^s mind important
rather than unimportant facts ; and since, in the present
case, the sexual differentiation is far more important than the
difference in size, I propose that the two kinds of spores
shall in future be known as gpiospores and androspores, the
receptacles in which they are produced as gynosporangia
and androsporangia respectively.

Appended is a table of the proposed classification.

Proposed Classification of Cryptogams.

A. Protophyta.

a. Protomgcetes (Protophyta acliloropliyllacea).

Schizomycetes.

Saccharomyces.

b. Prolophycea (Protophyta chlorophyllacea).

Protococcaceae.

Palmellacese.

Scytouemeae.

NostocaceaB.

Oscillatorieae.

Rivularieae.

Supplement to Protophyta. Myxomycetes.

B. PUXGI.

a. Zygomycetes (Zygospermeae achlorophyllaceae).
Mucorini.

Piptocephalidae.

h. Oomycetes (Oospermeae achlorophyllaceae).
Peronosporeae.

Saprolegnieae.

Chytridiaceae.

c. Carpomycetes (Garpospermeae achlorophyllaceae).

Uredineae.

Ustilagineae.

Basidiomycetes.

Ascomycetes.

Lichenes.

* The equally objectiouable term “ macrolepidoptera ” seems to have
crept into entomological terminology.

4l2 ALFRED W. BENNETT.

C. ALGiE.

a, Zygo-phycecB (Zygospermese chlorophyllaceae).
Pandorinese.

Hydrodictyese.

Confervaceae.

? Pithophoracese.

Ulotrichacese.

UlvaceaB.

Botrydieae.

Conjugatse.

^ Diatomaceae.

Desmidieae.

Zygnemacea3.

Mesocarpese.

h. Oophycea (Oospermeae chloropliyllaceae).
Volvocineae.

Siphoneae.

? Dasycladeae.

Sphaeropleaceae.

Oedogoniaceati.

Fucaceae.

Phaeosporeae.

c. Carpophyeece (Carpospermeae chlorophyllaceae)
Coleochaeteae.

Florideae.

D. Charace^.

E. MuSCINEiE.

Hepaticae.

Musci.

Sphagnaceae.

F. Cryptogamia Vascularia.

a. Isosporia.

Filices.

Opliioglossaceae.
Lycopodiaceae.
Equisetaceae.
h. Ueterosporia.

llhizocarpeae.

Selaginellaceae.

TERMINOLOGY OF REPRODUCTIVE ORGANS OF CRYPTOGAMS. 413

A Reformed System of Terminology of the Reproduc-
tive Organs of the Cryptogamia. By Alfred W.
Bennett, M.A., B.Sc., F.L.S., Lecturer on Botany at
St. Thomas’s Hospital ; and George Murray, F.L.S.,
Assistant, Botanical Department, British Museum.^

We have been led to the following attempt at obtaining a
symmetrical system of terminology among Cryptogams by the
undoubted fact that the anomalies at present existing greatly
retard the prosecution of the study of this most interesting
group of plants. A more striking instance of the want of
accuracy in the use of the commonest terms could not be
afforded than by observing the different meanings, often
quite irreconcilable with one another, applied by the most
approved writers to the term spore.” Thus, Le Maout and
Decaisne (Hooker’s edition of ^Descriptive and Analytical
Botany,’ p. 14) and Professor Asa Gray Botanical Text-
book,’ sixth edition, p. 434) speak of spores as the ana-
logues of seeds Sachs Text-book,’ English edition, p. 203,
but modified in the most recent German edition) defines
them as asexual reproductive cells.” In direct opposition
to each and both of these definitions, Berkeley Micrographic
Dictionary,’ third edition, p. 327) describes the unfertilised
ova or oospheres of Fucus as spores; and Huxley and Mar-
tin Elementary Biology,’ p. 45) call the archegonium of
Characese a spore-fruit or sporangium. A still more sin-
gular example occurs in the most recent text-book of botany
introduced to English readers (Vines’s edition of ^Prantl’s
Text-book” where we read on one page (p. 97) Reproduc-
tion of Cryptogams is effected asexually by cells termed
gonidia, conidia, or spores;” and on another page (p. 115),
“ Fungi are reproduced sexually by means of spores.”

The object kept before us in this attempt at reform has
been to arrive at a system which shall be symmetrical and
in accordance with the present state of knowledge, and which
shall, at the same time, interfere as little as possible with
existing terms. In preparing such a system it has been
impossible to avoid introducing several new^ terms, but these
are associated with one another on an etymological plan
which will not burden the memory of the student, while the
total number of terms in cryptogamy will be greatly reduced

’ Read at Ujc Swansea Meeiinff of the Rritisli Assoeiatioii, Aug. 2Ctli,
1880.

VOL, XX. —NEW SER.

E E

414 ALFRED W. BENNETT AND GEORGE MURRAY.

by the abandonment of a large number that are absolutely
useless.

So far as we know^ the first systematic attempt at placing
cryptogamic terminology on a more satisfactory footing was
made by Sachs in the fourth edition of his ^ Lehrbuch ^ (not
yet published in English), which is deserving of commenda-
tion, though it does not appear to us altogether successful.
Sachs proposes (p. 243) to define a spore ” as a reproduc-
tive cell produced directly or indirectly by an act of fertili-
sation,^’ reserving the term gonidium ” for those repro-
ductive cells which are produced without any previous act
of impregnation. As far as Vascular Cryptogams and Mus-
cineee are concerned, Sachs’s proposal involves no change,
since he regards the spores, which are a part of the non-
sexual generation, as the indirect result of the act of fertili-
sation which closes the sexual generation. In the Basi-
diomycetes he is able to retain the term spore for the
familiar bodies commonly so called, only by the assumption
— for at present it is nothing more — that the structure of
which the hymeniurn forms a part is the result of a
yet undiscovered process of fertilisation on the mycelium.
In the lower fungi the changes involved in the proposal
are considerable. Thus the spores ” of Penicillium,
similar as they are to those of Agaricus, can no longer be
called spores, because they do not result from an act of ferti-
lisation, but become conidia ” or gonidia and for the
same reason the familiar zoospores ” of the lower Thallo-
phytes become zoogonidia.” It is obvious that one practical
defect of this suggestion is that it may necessitate a per-
petual change of terminology as our knowledge advances.
Every fresh extension of the domain of sexual fecundation—
and it is probable that many such will take place — wfill
involve the removal of a fresh series of reproductive cells
from the category of gonidia to that of spores, even though
they may not be the immediate result of an act of fertilisa-
tion. Again, if the spores of Ferns and Mosses are the
indirect result of impreguation, it is difficult to say why the
term should not ultimately include all reproductive bodies
whatever, except the spores of the apogamous ferns ” with
which Earlow and De Bary have recently made us ac-
quainted, and of other similar abnormal productions, which
are certainly not the result of impregnation, direct or in-
direct.

It seems a sounder principle — and is certainly more con-
venient to the student — to base a system of terminology on
facts which can be confirmed by actual observation, rather

TERMINOLOGY OF REPRODUCTIVE ORGANS OF CRYPTOGAMS. 415

tlian on unproved hypotheses. We propose^ therefore, as
the basis of our terminology, to restore the term spore to
what has been in the main hitherto its ordinary significa-
tion, and to restrict its use to any cell produced hy ordinary
processes of vegetatio7ii and not directly hy a union of sexual
ele^nents, which becomes detached for the purpose of direct
vegetative reproduction. The spore may be the result of
ordinary cell division or of free cell formation. In certain
cases (zoospore) its first stage is that of a naked primordial
mass of protoplasm. In rare instances it is multicellular,
breaking up into a number of cells (polyspore, composed of
merispores, or breaking up into sporidia).

The simple term spore will, for the sake of convenience,
be retained in Muscinese and. Vascular Cryptogams; but in
the Thallophytes it will always be used in the form of one
of those compounds to which it so readily lends itself, ex-
pressive of the special character of the organ in the class
in question. Thus, in the Protophyta and Mucorini, we
have chlamy do spores ; in the Myxomycetes, sporangiospores ;
in the Peronosporeae, conidiospores y in the Saprolegnieae,
Oophyceae, and some Zygophyceae, zoospores; in the Ure-
dineae, teleutospores , cecidiospores, uredospores, and sporidia ;
in the Basidiomycetes, hasidiospores ; in the Ascomycetes
(including Lichenes), conidiospores, stylospores, ascospores,
polyspores, and merispores ; in the Hydrodictyeae, mega-
spores; in the Desmidieae, auxospores ; in the Volvocineae
and Mesocarpeae, par'thenospores ; in the Siphoneae and
Botrydieae, hypnospores ; in the Oedogoniaceae, . andro-
spores ; in the Florideae, tetraspores and octospores. The cell
in which the spores are formed will, in all cases, be called a
sporangium. It is obvious that, if greater precision is desired,
this term might be compounded in the same way as spore ;
but the words thus formed would be needlessly cumbrous for
ordinary use.

The male organs of fecundation are so uniform in their
structure throughout cryptogams that very little complica-'
tion has found its way into their terminology. The cell or
more complicated structure in which the male element is
formed is uniformly known among Cormophytes as well as
Thalloi)hytes as an atitheridium ; the fecundating bodies are
almost invariably naked masses of protoplasm, provided
with vibratile cilia, endowed with apparently spontaneous
motion, and bearing the appropriate name of antherozoids
or “ spermatozoids.” The former of these is preferable for
two reasons ; from its etymological connection with anthe-
ridium, and because the use of terms compounded from

416

ALFRED W. BENNETT AND GEORGE MURRAY.

“ sperm ’’ should, for reasons to be detailed presently, be
avoided for male organs. In only two important groups,
Floridem and Lichenes,are the fecundating bodies destitute of
vibratile cilia and of spontaneous motion; in the former
case they are still usually termed antberozoids ; in the latter
" spermatia,” and their receptacles spermogonia.” In
order to mark the difference in structure from true antbe-
rozoids, it is proposed to designate these motionless bodies in
both cases pollinoids ; the term spermogonium is altogether
unnecessary, the organ being a true antheridium.

A satisfactory terminology of the female reproductive
organs presents greater difficulties, from the much greater
variety of structure, and the larger number of terms already
in use. The limits we have placed to the use of the term
spore and its compounds require the abandonment of
‘‘ oospore ” for the fertilised ovum or oosphere in its encysted
state (enclosed in a cell-wall), anterior to its segmentation
into the embryo ; and this is the most important change
involved in the new system.

In devising a term which shall include all those bodies
which are the immediate result of impregnation, it was
necessary to take two points specially into account. Firstly,
the term must be capable of defence on etymological grounds ;
and secondly, it must, like spore, be capable of ready com-
bination. After much consideration we have decided on
proposing the syllable sperm. No doubt the objection will
present itself that the Greek anip/jia, like the Latin semen,”
and the English seed ” (as used by old writers), while
originally meaning the ultimate product of fertilisation,
came afterwards to signify the male factor in impregnation ;
and hence, in zoology, terms derived from these roots are
used for the male fertilising bodies. But the objection applies
to a much smaller extent to phyto-terminology. Some terms
compounded from sperm,” as gymnosperm, angiosperm,
&c., are already familiarly in use in a sense similar to that
we would indicate ; while of those used in the reversed
sense, “sperm-cell,” for antherozoid or pollen-grain, has
never come into general use in this country ; “ spermatozoid ”
is easily replaced by antherozoid ; “ spermogonium ” is
simply a peculiar form of antheridium, and “ spermatium ”
has already been referred to. Accepting this term as the
least open to objection of any that could be proposed, it will
be found to supply the basis of a symmetrical system of ter-
minology, which will go far to redeem the confusion that at
present meets the student at the outset of his researches.
For the unfertilised female protoplasmic mass the term

TERMINOLOGY OF REPRODUCTIVE ORGANS OF CRYPTOGAMS, 417

oosphere is already in general use ; and, though not all that
could be desired, it is proposed to retain it, and to establish
from it a corresponding series of terms ending in sphere.
The entire female organ before fertilisation, whether uni-
cellular or multicellular, is designated by a set of terms
ending in gonium, again following existing analogy.

As will be seen from another paper presented to the Asso-
ciation, we propose, in our forthcoming Handbook of
Cryptogamic Botany,^’ to make the primary division of the
Thallophytes into the three great classes of Protophyta,
Fungi, and Algae, each of the two last being again divided
into three parallel series of Zygospermeae, Oospermeae, and
Carpospermeae, corresponding nearly to Sachs’s Zygosporeae,
Oosporeae, and Carposporeae. In the Zygomycetes and
Zygophyceae, the conjugated zygospheres., or contents of the
zygogonia^ constitute a zygosperm ; in the Oomycetes and
Oophyceae, the fertilised oosphere, or contents of the oogo-
nium, is an oosperm; in the Carpophyceae, the fertilised
carposphere, or contents of the carpogonium, constitutes a
carposperm. In this last class the process is complicated,
being effected by means of a special female organ, which, to
keep up the etymological analogy, may be called a tricho-
gonium rather than a “ trichogyne.” The ultimate result
of impregnation is the production of a mass of tissue, known
as the cystocarp (or sporocarp within which are pro-
duced the germinating bodies, which must be designated
carpospores, since they are not the direct result of fertilisa-
tion, but which must be carefully distinguished from the
so-called carpospore ” — properly an archesperm — in the

Characeae. Any of these bodies which remains in a dor-
mant condition for a time before germinating will be a
hypnosperm.

In the Zygospermeae it is no doubt the case that the two
conjugating bodies are really a zygosphere and an anthero-
zoid (or pollinoid) ; but, as they are at present absolutely in-
distinguishable, it seems best to sacrifice theory, and calf
them both zygospheres.

It may be mentioned, by way of completing the analogy,
that a precisely analogous set of terms will be pro-
posed for the Cormophytes (Characeae, Muscineae, and
Vascular Cryptogams), throughout which the fertilised
archesphere, or contents of the archegonium, is called an
archesperm. The latter term may be objected to as nearly
identical with archisperm,” used by Strasburger and
others as a synonym for gynmosperm ; but the term has not
come into general use, and no confusion between the two

418 ALFRED W. BENNETT AND GEORGE MURRAY.

seems possible. In the proposed system zygosperm will
replace Strasburger’s zygote/’ and the gametes ” of the
same writer wdll be zygospheres ; while his zoogametes ” or
“ planogametes ” must enjoy the somewhat cumbrous name
of zoozygospheres,” the prefix zoo ” or suffix zoid ”
being always used to denote an apparently spontaneous
power of motion.

In the Basidiomycetes, Ascomycetes^ and some other
classes of Cryptogams, the entire non-sexual generation
which bears the spores is frequently called the ‘‘receptacle/’
a term which is for several reasons objectionable. In the
first place it has no meaning etymologically in this connec-
tion; and secondly, it suggests a false analogy with the
receptacle in flowering plants, a structure which supports
the sexual reproductive organs. A convenient expression
for that portion of the non-sexual generation which bears the
spores is the fructification^ corresponding to the German
equivalent “ Fruchtkorper,” and already popularly used in
this sense in the case of mosses, as well as for the sori of
ferns. Except, as far as is yet known, in the case of the
Basidiomycetes, the fructification is always the result of im-
pregnation ,• and, according to Reqss, Eidam, and others, the
same is the case also in that class of fungi.

To illustrate the simplification effected in cryptogamic
terminology by the proposed reform, two comparative tables
are appended ; and the following list is given, which enu-
merates some of the terms in more frequent use by various
writers as respects Thallophytes and Characese, all of which
are disused in the proposed system.

Non-sexual organs, — Gemma, conidium, gonidium, endo-
gonidium, zoogonidium, macrozoospore, acrospore, ectospore,
swarmspore, pycnidium, perithecium, apothecium, theca,
receptacle.

Male sexual organs, — Spermatozoid, spermatium, spermo-
gonium, globule.

Female sexual organs, — Oospore, zygospore, zygogonidium,
zygozoospore, zygote, gamete, planogamete, aplanogamete,
ascogonium, nucule, sporocarp, ceramidium, coccidium,
favella, favellidium, sphaerospore.

TERMINOLOGY OF REPRODUCTIVE ORGANS OF CRYPTOGAMS, 419

ZYGOSPERilEA:.

OoSPERMEiE.

Modes of Fertilization in Cryptogams.

Conjugating Bodies.

Zygogonia, containing Zygospheres.

(fertilised) Zygosperm.

Female Organ.

Carpospermea:.

CORMOPIIYTA.

Male Organ.
Antlieridium, containing
Antlierozoids or ")
Pollinoids. \

Oogonium, containing
Oosphere.

Antlieridium, containing
Antlierozoids or 7
Pollinoids. )

(fertilised) Oosperm.

Carpogonium, containing
Carpospliere.

(fertilised) Carposperm,

Antlieridium, containing Arcliegonium, containing
Antherozoids = Archespliere.

II

(fertilised) Archesperm.

Froductive Organs
Female.

of Thallophytes .
Non-sexual.

Protopiiyta.

Chlamydospore.

Sporangium.

Myxomycetes.

Zoospore.

Sporangiospore.

Mucorixi.

Zygogonium.

Zygosphere.

Zygosperm.

Chlamydospore.

Sporangiospore.

Peronosporea.

-

Saprolegniea.

'"Oogonium.

Oosphere.

Oosperm.

w

Sporangium.

Conidiospore.

Zoospore.

Zoospore.

Uredinea.

UsTILAGINEiE.

r-

Carpogonium.

Carpospliere.

Carposperm.

Teleutospore.

Aecidiospore.

Uredospore.

Sporidium.

Teleutospore.

Sporidium.

Basidiomycetes.

Basidiospore.

Sterigma.

Basidium,

420

ALFRED W. BENNETT AND GEORGE MURRAY.

AscoiiYCETES, Tricliogonium.

including Lichenes. Pollinodium.

ZyGOPHYCEiE.

OoPHYCEiE.

CARPOPHYCEyE.

Zygogonium.

Zygosphere.

Zoozygosphere.

Zygosperm.

Hypnosperm

(Hydrodictyese

zygnemaeae).

Oogonium.

Oosphere.

Oosperm.

Conceptacle.

Hypnosperm.

Carpogonium.

Carposphere.

Carposperm.

Trichogonium.

Tricliophore.

Cystocarp.

Conidiospore.

Stylospore.

Ascospore.

Polyspore.

Merispore.

Zoospore.

Megazoospore (Hydrodictyeae).
Auxospore (Diatom aceae).
Hypnosporaiigium I rj...,
Hypnospore j iiotrydiese.

Parthenospore (Mesocarpeae).
Zoospore.

Parthenospore (Volvocineae).
Androspore (Oedogoniaceae).
Hypnospore (Siphoneae).

Zoosporangium.

Tetraspore.

Octospore.

Carpospore.

ON THE LAMINAR TISSUE OF AMPHIOXUS.

421

On the Laminar Tissue of Amphioxus. By Professor
PoucHET. (With Plate XXIX.)

The chief results in this paper were communicated to the
Societe de Biologic at their meeting on May 1st, 1880.^

I. I was lately fortunate in obtaining large numbers of Am-
phioxus at Concarneau, and I used them in endeavouring to
clear up a special point in their histology, which seems to have
been only very imperfectly touched upon by the numerous
authors who have studied this animal. I intend to speak
of the system of canals and cavities, which from a point of
view, which is perfectly legitimate, are comparable to fin
rays. Both of these systems are differentiations of what
may be termed *naminar tissue.”

In 1849 M. de Quatrefages applied the term “ singular ” to
the laminar tissue of Amphioxus. Since that time the system
of spaces and canals which have been observed have not been
the object of special study. The fact has been recognised
that they do not communicate with the blood capillaries,
but their histological structure has remained obscure,
M. Reichert is the only author who has touched upon it, and
he seems to have guessed at, rather than proved, their analogy
with the constant elements of the laminar tissue. He ex-
presses himself thus, ^^Da das bindegewebige Stroma uns als
pellucide Grundsubstanz angesehen werden Kami und die
dazu gehorigen Bindesubstanzkorper fehlen, so ware es
moglich dass das in Rede stehende netzformige Gebilde
den zellenhalftigen Theil des bindegewebigen Stromas dars-
tellt, unter dessen Vermittelung die in grosser Menge ver-
breitete, ganz hyaline Grundsubstanz gebildet werde.”^

In 1873 Stieda^ contents himself with quoting these words
of Reichert, which in fact contain the true state of the case.

Before Reichert’s papei\ was published, Owsjannikow^ de-
scribed, in 1868, in the region of the tactile corpuscles, that
is to say, round the mouth, connective-tissue cells, gross,
liinglich, zuweilen sternfbrmig,” and compared them to the
cells of the cornea, an error of observation as will be seen
further on.

» ‘See ‘Gazette Medicale’ for May 22, 1880.

2 Reichert, ‘ Arch. f. Anat.,’ 1870.

^ “ Studien iiber den Amy)hioxus lanccolatus,” ‘ Mem. dc 1’ Academic
de SL Petersberg,’ viie Serie, Tome xix, No. 7, 1873.

^ “ Ueber das centrale Nervensystem des Amphioxus lanccolatus,” in
Bullet, de I’Acad. de St. Petersbourg,’ T. xii, 18G8, p. 290.

422

PROFESSOR POUCHET.

P. Langerhans, in 1873^ mentions^ the same cells; ne
describes them as very much branched (reich verastelt), and
compares them, as doesO\vsjannikow,tothe cells of the cornea,
and corroborates the description of Stieda, although this is
limited to the quotation from Reichert, which rather goes
against the existence of such cells.

Quite recently Ant. Schneider^ admits the truth of
Reichert’s ideas, and at the same time describes (pp. 1 and
28) and figures in the laminar tissue of Amphioxus stel-
late fibro-plastic cells, inserted by their prolongations on
the walls of the true capillaries around the mouth.

These quotations suffice to show the confusion and incom-
pleteness of our knowledgeof the laminar tissue of Amphioxus,
and of the farmed elements which it presents. As we shall pre-
sently show, even if it is possible to find isolated cells having
the ordinary appearance of stellate fibro-plastic cells, it is
quite exceptional at least in the adult. In our preparations we
have found one or two in the thinnest region of the caudal
lophioderm, quite at the extremity ; with this exception we
have not come across them. It is in fact a very decided
character of the laminar tissue of Amphioxus that the formed
elements analogous to the fibro-pjastic cells never appear
in an isolated state, and that they are always united again
into scattered groups or in longer or shorter rows which
are more or less regular, and anastomose to a greater or
less extent, forming cords which are sometimes solid iund
sometimes hollow for a certain distance, or even form large
spaces.

It is these cords, which are partly solid, partly hollowed,
and generally slightly elongated, which have been described
as ^'systems of canals.” The supposed branched cells of
Owsjannikow are only the enlargements of this very irre-
gular network in the neighbourhood of the mouth.

II. As these cordons and spaces formed by the coalescence
of the cells of the connective tissue are situated to a large
extent in the layer of laminar tissue which surrounds the
body of the animal, we shall first of all describe this,^ add-
ing, however, but little to the descriptions of Stieda.

Below the epidermis you get, passing from without in-
wards—

^ “Zur Anatomie des Amphioxus lanceolatus,” in *Arch. f. Mik.
Anat.,’ 1873, p. 301.

2 ‘ Beitriige zur vergleichenden Anatomie,’ 4to, Berlin, 1879.

3 Compare with this the strueture of the skin in fishes. See Pouchet,
“Du developpement des poissons osseux,” ‘Journal de 1’ Anatomie,’ May-
June, 1879, pp. 288-289.

ON THE LAMINAR TISSUE OF AMPHIOXUS,

423

1. The dermis. 2. The subdermic layer. 3. The sub-
cutaneous aponeurosis.

1. T/ie dermis^ is formed, as in fishes and Amphibia, of a
thin layer which appears in section to be homogeneous, and
'\Vhich "in Amphioxus presents upon its external surface a
double rectangular striation, wfith the crosses corresponding
to the nerve terminations which Langerhans^ has so well
figured.

2. Subdermic layer. — The subdermic layer^ is charac-

terised in Amphioxus by the presence of a large amount of
structureless substance. This latter is as usual absolutely
hyaline, and keeps this character even after the action of
osmic acid in saturated solution.^ It is traversed by fibres
having the character of laminar fibres, stretching from the
deep face of the dermis to the subcutaneous (cf. fig. 3a).
These fibres are absolutely devoid of any nucleus and of any
division into cells, a fact not without interest as bearing
upon the much debated history of the genesis of the laminar
fibres. We have pointed out in the following terms^ the ex-
istence of similar fibres in the lophioderm of larval Axolotls
measuring 60 to 80 mm. in length : In sections cut la

planchette,^ after treatment with osmic acid, the amorphous
matter is seen to be traversed by vessels and nerves, and to
enclose some fibro-plastic cells with regular spaces between
them, invisible ; but extremely fine laminar fibres are also
seen. These run from the skin of one side to that of the
other, in the same manner as the strings in an instrument
in a very beautiful manner, all parallel to one another. These
fibres seem, therefore, to have no relation to the fibro-plastic
cells.”

It was always difficult in Axolotl to pronounce an opinion
on the independence of the fibres and cells on account of the
presence of the latter. Whilst in Amphioxus there are no
fibro-plastic cells furnished with any prolongations from
which laminar fibres might be derived. These are also
larger than in Axolotl, and are more deeply stained by car-
mine. No migratory corpuscles are found in the hyaline
structureless portion of the subdermic tissue of Amphioxus.

3. Subcutaneous aponeurosis. — This membrane appears in

^ “ Cutis Oder Lederhaut,” Stieda.

- Loc. cit., fig. 13.

3 “ Das Unterhautgewebe ” of Stieda.

^ For the action of this reagent see Pouchct, “De I’emploi dcs solutions
concentrees d’acide osmique,” ‘ Journal dc I’Anat.,’ Sept.-Oct., 187C.

^ G. Pouchet et Tourneux, ‘ Precis d’histologic huinainc ct d’histo-
genie,’ 1878, p. 102.

424

PROFESSOR POUCHET.

section to be finely striated as in fishes. A lamellar struc-
ture must be attributed to it^ and elongated nuclei appear in
certain places. It belongs to a system of fibrous tissue widely
distributed in the body of the animal ; this is quite distinct
from the laminar system properly so called, while in its
physico-chemical characters it approaches the dermis. To
this system belong the aponeuroses which separate the muscles
and envelope the medulla, &c. The skin, the subdermic apo-
neurosis, and the fibrous partitions of which we have been
speaking, when treated with osmic acid and picrocarmine,
take a splendid red staining, while the structureless sub-
stance of the subdermic layer and other varieties of struc-
tureless substance which are sometimes found in the body of
Amphioxus, remain uncoloured or become slightly yellow,
but do not take the carmine.

III. We have said that the special character of the formed
elements of the laminar tissue of Amphioxus was that they
are never present in an isolated state nor as cells simply
united, like the cells in the cornea of Amphibia, by fine
prolongations. We shall first describe these elements and
the network which they form in the lophioderm *of the
caudal fin of Amphioxus, and afterwards their modifications
in other parts of the body of the animal.

The caudal lophioderm ” presents a network with regular
elongated meshes, which had already attracted the attention
of M. de Quatrefages, and which at once calls to mind the
figure which he gave at a time when the technical methods
necessary for such a study were not yet in vogue.

This network is arranged in one plane in the centre of
the lophioderme, which is very thin in this region. The
meshes of the network are elongated in the neighbourhood
of the anus, and measure 60 ju to 80 /x in length, by 10 /x
in breadth (see fig. 1). The substance of the network is
formed of rods, which measure generally 8 to 10 /x .; they
may, however, be narrower or broader than this. They are,
in the region of which we are speaking, flattened out ; they
are solid. At the intersections of the network there are
generally 6 to 10 larger or smaller nuclei. These are oval
or almost spherical, irregularly placed, and separated by a
distance which is generally less than their diameter. Some-
times several are placed in a row in one of the rods. They
are then generally more or less obliquely placed. The rods
after the action of osmic acid stain slightly with picrocar-

‘ Lophioderm is the termed used by M. Pouchet for the dermal fold
which forms the substance of the median fin in Ampliioxus and in Am-
phibia.— Eu,

ON THE LAMINAR TISSUE OF AMPHIOXUS.

425

mine. They represent the bodies of the cells united one
with another, or rather, all fused together.^

The only tissue which approaches in histological struc-
ture that which we are about to describe, is the tissue of
the cornea in the eyes of bony fish. We have elsewhere
described this tissue thus : — All the corneal cells appear
fused together, so that it is impossible to distinguish their
outline, the protoplasmic bodies of the cells forming a
sort of membrane, with nuclei scattered here and there.
The meshes of this network are small and separated by
extremely broad rods, which forms a true lamellar system
mixed up with the fundamental lamellar system of the
organ, &c.”^

In the lophioderm of Amphioxus the meshes of the
network are broader than the rods, and these latter are
disposed in the same plane ; but the two tissues, nevertheless,
present an analogy which enables one to place them side by
side among the large number of varieties which the laminar
tissue of vertebrates presents for our study.

Fig. 6 shows that in posterior region of the notochord the
nuclei are very abundant ; they are very small, and answer
very well to the description and figure given by Stieda.
On the other hand, they are scattered throughout the sub-
stance of this organ, which does not agree with the opinion
of Kossman (see ‘ Verhandlungen der Phys. Med. Ges.,’
Wurzburg, 1874).

In the deep layer under the dermis in the neighbourhood
of the mouth, the network of cells of the laminar tissue has
not the same appearance. It presents large irregular meshes,
with a slightly sinuous contour. The nuclei are no longer
accumulated at the junctions of the network. The rods are
almost cylindrical, and are swollen here and there.® These
swellings present a larger or smaller hollow in the centre,
bounded on all sides by the protoplasm and the nuclei,
which are arranged in a single row. In optical section these
hollows appear to be simply fissures between the layers of
cellular substance.

Sometimes these enlarge and spread to a greater or less
extent, but they always remain very irregular. This is the

^ It sometimes happens in certain preparations that at the inter-
sections of tlic network there is a sort of cleavage, so that the nuclei
appear to be each surrounded by a protoplasmic body with a polygonal
outline.

- G. Pouchet et Tourneux, loc. cit., p. Oil, et fig. 1G8.

•’ These swellings, scattered on the thinnest rods of the network, have
no doubt given rise to the error of Owsjannikow, who thought he had seen
the cellular bodies anastomising by their prolongations.

426

PROFESSOR POUCHET.

‘^system of canals’^; it is very different from a capillary
system properly so called, and its true nature has
been very variously interpreted by different authors. In
the neighbourhood of the notochord these cavities appear
to enlarge and spread out. Fig. 3 b shows one of these
cavities, pyriform in shape, the extremity being narrowed
and ^running into a rod, which binds it to the network.
Thus, this sort of hollow swellings forms a transition to
other cavities which are quite separate from the network of
cells, and which we are about to describe.

ly. The upper surface of the anterior extremity of the
notochord presents isolated groups of connective-tissue
cells, which completely envelope the cavities which they
bound, being arranged in a single row, as in the swellings
of the network in the neighbourhood of the mouth.

These groups of cells, always clearly bounded on the
outside, are generally oval in form, or rather, polyhedral,
where they are crowded together. The nuclei presents
slight prominences in their interior, and, in certain
individuals the cells contain various pigment granules of
a reddish colour.^

The interior of the cavity is filleff with a liquid, which is
unacted on by osmic acid, and does not stain with most
reagents. It appears to be watery. The further we recede
from the extremity of the cord the larger do these laminar
cavities become (see fig. 2). They become also slightly
modified, and finally constitute the organs described under
the very bad name fin rays ” (see figs. 4 and 5). The
structure of these organs is extremely siniple. They
are cellular spaces, larger than the others, almost cubical
in form, one wall of which, that which looks towards
the notochord, is invaginated into the interior of the cavity
itself, so as to fill up two thirds or three fourths of it. We
call the invaginated portion the papilla. It is made of a
structureless substance, presenting special characters which
clearly distinguish it from that of the sub-dermic layer.

Stieda describes these parts very incompletely. He states,
it is true, the analogy between these cavities and the system
of canals, and describes their cellular covering, but the in-
vaginated portion which Muller had already mentioned he
has overlooked. Muller states that in the interior of these
cavities there exists eine durchsichtige Fliissigkeit und eine
consistentere, aber weiche Masse. The cavities are bounded
at the sides and below by the subcutaneous aponeurosis,

^ Spme specimens of Amphioxus present a reddish tint, others a
grcenisli.

ON THE LAMINAR TISSUE OF AMPHIOXUS.

427

and are separated from one another by transverse partitions
(see fig. 5), which are very thin.^ The whole cavity, the
papilla as well as the walls, is lined with a layer of cells,
which closely resemble those lining the other spaces ; there
is no appreciable difference between the cells on the papilla
and those on the walls. The contained liquid seems to be
the same as that in the other spaces.

In the living animal the substance of the papilla can only
be distinguished from the ambient liquid by the optical effects
due to their differing density. In a specimen which has
been preserved in alcohol the papilla had almost disappeared.
Certain reagents have the same action, whilst others appear
to swell them up. If the specimens are put into alcohol,
even after the action of osmic acid, the papilla are altered in
shape and drawn in. Their surface presents depressions,
which gives them a jagged appearance in section.^

The use of reagents brings to view minute cavities inside the
papilla, more or less irregular, and about the size of one of
the nuclei. Suitable methods show that the amorphic sub-
stance of the papillae contains nothing but some very fine and
scattered fibres. Transverse sections (fig. 4) show very well
the relations of the papillae. There is a fibrous lamina occu-
pying the middle line of the body, which, bifurcating at the
level of the upper edge of the two lateral muscular masses,
joins on each side the subcutaneous aponeurosis. The base of
the papilla rests in the groove formed by this bifurcation, and
is clearly marked out by its physico-chemical characters. The
cavity which it contains, is itself bounded laterally and below
by the subcutaneous aponeurosis.

The papilla is traversed towards its base by a layer of very
fine fibres (fig. 4, c.) stretched between the two fibrous
laminae. In front and behind (fig. 5, c.) this sheet of fibres
becomes raised up before being inserted into the partitions
which separate the cavities. On the upper face of the
layer a very loose and irregular network of extremely fine'
laminar fibres rise up and spread out in the centre of the
papilla. None of these are directed downwards. When
treated with osmic acid and then saturated with picro-
carmine, the substance of the papilla assumes a slightly
greenish tint, and moreover becomes slightly granular, and,

* Ant. Schneider, who believes tliat these cavities communicate with the
system of canals, describes the latter as bounded by a membrane, with
nuclei and finely striated fibres, and calls these fibres “ muscular (?),” loc,
cit., pp. 8, 9.

- This it is which Ant. Schneider (loc. cit., p. 8) has described as stalked,
and supposes them to be contractile !

428

PROFESSOR POUCHET.

therefore, to a certain extent opaque. This character
belongs to it alone ; under the influence of the same reagents,
the amorphous matter in other parts of the body of
Amphioxus remains hyaline, and the layers of fibres
become stained bright red.^

In certain cases (fig. 5) a very clear and definite stratifica-
tion of the amorphous matter may be seen in the papilla. The
peripheral portion and the base are clearer and less opaque
than the central portion, and the boundary between these
ditferent portions is clearly marked out. The central portion,
which is also the darkest, rests on the layer of transverse
fibres, and is raised up against the partitions of the cavities
to where the endothelium is reflected.

In the centre of the papilla this dark zone presents a
large projection, wdiich itself may be surmounted by an
intermediate zone.

V. — What histological and morphological signification
can be attributed to the papilla and to the containing
cavity ?

The papilla evidently belongs to the category of connective
or laminar organs.

The amorphous substances are everywhere very abundant
in Amphioxus. In transverse sections (fig. 4) two thick
layers [d, e) of greenish amorphous substance, hyaline and
devoid of laminar fibres, may be distinguished on each side
of the median aponeurosis. Elsewhere the amorphous sub-
stance is very dark g), and contains thick fibres,^ and
although no part is so opaque as the papilla, it is evident
that it is a simple variety of connective-tissue, characterised
by the absence of cellular elements, the great scarcity of fibres,
and lastly, by the probable presence of a certain quantity of
fatty matter in a finely divided state.

The meaning of the cavity which surrounds the papilla is
also evident. It is a space which is more developed than
the rest, and which tends to take on the character of a
serous sac. Silver nitrate gives the same result as on
serous membranes ; in the only preparation we have made,
the network of silver staining is very clear upon the lateral
walls of the cavities, but does not show upon the papilla,
but no doubt could be obtained there also by employing
suitable methods.

^ After the more prolonged action of osmic acid and alcohol the papilla
becomes darkly coloured, and certain coloured oily globules appeared to
be emitted by it. This seems to indicate the presence of fatty bodies
in its composition.

• Comp. ‘ Ant. Schneider, loc. cit., p. 7.

ON THE LAMINAR TISSUE OF AMPHIOXUS.

429

When one observes the gradations of structure between
these cavities, the small separate cellular cavities placed at
the extremity of the cord, the terminal cavities of certain rods,
and, lastly, the rudimentary cavities in the swellings of the
network in the neighbourhood of the mouth, one is led to
regard these various structures as the first indications of a
system of serous cavities and lymphatics. Besides this, the
cellular lining of these cavities has quite the characters of the
lining of the embryonic serous sacs in the higher Vertebrata.
We confine ourselves to merely suggesting this as a probable
analogy, as we have no exact account of the histogenic
development of these parts.

Lastly, what morphological signification shall we give to
the organ represented by the papilla and its cavity

Here all terms of comparison leave us completely at
fault ; no homology whatever among vertebrates is to be
thought of.^ Their number corresponds neither with that of
the spinal nerves, nor with that of the myotomes ; they are
about three or four times as numerous.

The connection which certain authors put forward^ be-
tween the papilli and the fin-rays* ought rather to be con-
sidered as an analogy than a true homology. Amphioxus
is not altogether devoid of either cartilage or bone. It is
cartilage which forms the skeleton of the oral tentacles ;
it is bony substance which forms the framework of the
branchial skeleton. The papillae, looked at from the point
of view of general anatomy, would form a third kind of
skeleton element. They would represent a first stage in
the differentiation of amorphous matter which is about to
form a special organ.

It should not be forgotten that amongst Teleostean fishes
a certain number of bony organs of the dermal skeleton
result simply from a similar differentiation of amorphous
intercellular substance, such as we have shown in the bony
plates of Syngnathians. In the latter case, the amorphous
matter becomes individualised as calcareous salts are de-
posited in it ; in the papilli of Amphioxus fatty matters are
no doubt deposited in it. In any case, both belong directly
to a numerous and varied group of amorphous intercellular
substances.

’ We may here note a peculiarity which presents itself in the sheath of
the notochord of Amphioxus. In the most successful transverse sections
this presents above a slight median prominence (cf. Ant. Schneider, loc. cit.,
pi. XIV, fig. 1), and below two slight prominences on each side of the
median line, which no doubt might be considered as the first rudiments of
the vertebral skeleton.

2 Anton Schneider, loc. cit., p. 8.

VOL. XX. NEW SER.

F r

430

PROFESSOR POUCHET.

We point out this analogy here without giving to it more
weight than seems justifiable, in the absence of special re-
searches. We may add that, from whatever point of view we
look at the subject, either that of comparative anatomy or that
of general anatomy, a considerable difierence still remains
between the papilla of Amphioxus enveloped in its serous
cavity and the dermal bony productions of fishes. Up to
the present day, at any rate, intermediate conditions are
wanting which would show how such a morphological
evolution could have taken place as is taken for granted in
bringing these organs together, from a more or less pro-
nounced phylogenetic point of view, under a common denomi-
nation. All just ground for establishing a real homology
between them is wanting. Although the papillae partake,
along with true fin-rays, of the character of skeletal organs,
they clearly possess a special nature, without homologue, in
other vertebrates.

Museum de Paris, 1 June, 1880.

PERIPHERAL NERVOUS SYSTEM IN NEMERTINES. 431

The Peripheral Nervous System in Pal.eo- and Schizo-
NEMERTiNi^ one of the Layers of the Body- wall. By
Dr. A. A. W. Hubrecht, of Leyden. With Plates
XXXII and XXXIII.

In a paper, which was published in May last, in the
‘ Verhandelingen van de Koninklyke Akademie van
Wetenschappen te Amsterdam, vol. xx,’ and which treats of
the anatomy and physiology of the nervous system of the
Nemerteans (an abstract of the paper with accompanying
plate appeared in the July number of this Journal), the
peripheral nervous system is dealt with very superficially
only. This was done on purpose. At the time when that
paper was prepared for the press the results which I had
arrived at concerning the peripheral nerves in the two sub-
orders of the Pal^o- and Schizonemertini were so much
in contradiction with everything hitherto published on the
subject that I could not venture to publish these results
before I had been able to convince myself of their trust-
worthiness in numerous series of preparations, which should
gradually come under my observation. During the correc-
tion of the proof-sheets of that paper I was happily surprised
in obtaining, much sooner than I expected, different sets
of sections, which, by their superior state of preservation,
allowed me to settle the point. I could only mention this
briefly and insufficiently in the explanation of the plate be-
longing to that article. A more detailed account was at
the same time promised, and is now embodied in the follow-
ing pages.

A. Central Nervous System and Cephalic Nerves,

The central nervous system of the Nemertines is composed
of two longitudinal trunks running parallel to each other all
along the whole length of the body j they are dilated ante-
riorly into a very simple or a more complicated pair of brain-
lobes, situated above the alimentary canal (often in front of
the mouth), and united together by a ventral and a dorsal
commissure, the latter always thinner than the former.
Posteriorly these longitudinal trunks either terminate very
close to the end of the body, or they are again united by a
third commissure, which, like the brain, passes above the
alimentary tract in the anal region. The whole of the cen-
tral nervous system (“ marrow-trunks ” and brain-lobes)

432

DR. A. A. Vr. HUBRECHT.

consists of fibrillar nerve-substance, to which a thick layer
of nerve-cells is uninterruptedly applied. This cellular
coating is nowhere dilated into ganglionic swellings, except
in the brain itself.

An important difference is presented in the situation of the
central nervous system with respect to the muscular body-
wall in three different groups of genera. In the more primi-
tively organised genera of the Paljeonemertini [Carinella
annulatay inexpectata, &c.) the whole of the nervous system
lies directly under the epiderm outside of the muscles ; in two
other genera of Pal;eonemertini and in all the Schizo-
NEMERTiNi it Hes enclosed within the muscular coat of the
body, generally between the outer longitudinal and the
inner circular layer of the muscles, whereas in the Hoplo-
NEMERTiNi the longitudinal marrow-trunks are situated
inside the muscular body-wall in the body-cavity.

Three sets of cephalic nerves originate directly from the
brain : — a. Nerves to the tip of the snout and to the eyes (if
present), h, A pair of nerves for the proboscis, c. A pair
of nerves for the wall of the oesophagus and alimentary
canal (n. vagus). ^

B. Peripheral Nerves,

A rapid recapitulation of what is said about the peripheral
nervous system ofNemertines by the three authors who have
furnished the best original investigations on these worms —
Quatrefages, Keferstein, and Macintosh — may precede the
enumeration of the results arrived at by myself, and, at the
same time, serve to explain apparent discrepancies.

Quatrefages Annales des Sciences Naturelles,’ vi, 1846,
p. 2T8) says, speaking of the longitudinal marrow-trunks :
De ces troncs partent d’espace en espace des filets qui se
portent probablement aux couches musculaires, mais que je
n^ai jamais pu suivre assez loin pour etre certain de ce fait.

Je n^ai distingue le long de ces troncs primitifs

aucune trace de veritables ganglions. Seulement les filets
qui en partent sont assez regulierement epates ^ leur base
et quelquefois on pourrait croire qu’il y a la une sorte de ren-
flement, mais cette particularity ne se reproduit pas d’une
raaniere constante.”

This is all the information he gives about the peripheral
nervous system. I must insist upon the fact that all his
figures (pi. 8, fig. 2; pi. 9, fig. 1) in which he represents
the small branches springing from the longitudinal trunks
belong to the suborder of the Hoplonemertini. In a sepa-
rate figure of the longitudinal nerve-trunks of Carinella

PERIPHERAL NERVOUS SYSTEM IN NBMERTINES. 433

(pi. viii, fig. 3) no lateral branches are indicated in the
drawing,

Keferstein writes (‘ Zeitschr. f. Wissensch. Zoologie,^
Bd. xii, p. 80) ; Aus den Seitennerven treten in regel-
massigen Abstanden (pi. v, fig. 10) feine Nerven, mit
breiter Basis entspringend, aus, die ich nur bis auf unbe-
deutenden Abstand vom Seitennerven verfolgen konnte und
die wahrscheinlich hauptsachlich in die Hautspgehen.”

The figure in which he indicates these peripheral nerves
implanted upon the longitudinal trunks is again, in this
case, one of the Hoplonemerteans, viz. Amphiporus
splendidus.

Macintosh is the first to make separate mention of the
state of things as he finds them in the armed and unarmed
species.

Of the first he says A Monograph of British Annelids,
A. Nemerteans^ p. 83) : The branches given off by these

trunks are generally pale and indistinct, but by the use of
dilute acetic acid in Amphiporus lactifloreus, and in others
without such aid, they can be satisfactorily observed. They
are easily seen, for instance, in Amphiporus pulcher, the
reddish colour which tinges them at their commencement
shining through the translucent integuments. An elaborate
plexus of branches from the lateral trunks has also been
noticed in these species” (pi. xvi, fig. 3),

Of the unarmed species he writes as follows (ibid.,p. 110) :

“ Branches probably exist, but only faint traces of such
are seen in the longitudinal sections, for the opacity of the
textures in the living animal prevents their being satisfac-
torily made out.”

Macintoshes very hesitating affirmation 'concerning the
presence of peripheral branches in the unarmed Nemerteans
is certainly a step in advance of his predecessors, who did not
notice any difference in this respect, who principally ex-
amined and only figured armed species, and regarded the
facts as they found them in this subdivision at the same
time as normal and typical for the whole order. In this way
the error has for a long time persisted, which I now hope to
clear away definitely, that the peripheral nervous system of
the Nemerteans corresponds in general to the type of the
Hoplonemertini, in which it was first described and
figured.

My own investigations of the Hoplonemertini did not
furnish any new results worthy of note. I found the small
branches given off by the longitudinal trunks as the above-
cited naturalists have described them. These branches

434

DR. A. A. W. HUBRECHT.

follow each other at regular intervals, probably corresponding
in number with the successive internal metameric divisions
of the body. They remain undivided for a certain distance,
and, spreading out in the muscular tissue, are quickly sub-
jected to a dichotomic division, which gives rise to numerous
smaller twigs. They can be clearly made out by compres-
sion of the living animal, with or without the use of dilute
acetic acid, as well as in longitudinal and horizontal sec-
tions. I generally noticed two branches springing from the
same region of the longitudinal trunk, one turning to the
dorsal, the other to the ventral half of the muscular body-
wall (see PI. XXXII, fig. 8, and in another paper, ^ Aan-
teekeningen over de Anatomie enz. van eenige Nemertinen,^
Utrecht, 1874, pi. i, fig. 2). The elaborate plexus of
branches observed by Macintosh is worthy of note, as I
will try to show hereafter ; it was sometimes observed by
myself, although in a lesser degree than is represented in
Macintosh^ s figure.

The two other Nemertean suborders, the Paljeo- and the
ScHizoNEMERTiNi, of which I examined several hundreds of
sections, most of them forming uninterrupted series, gradually
puzzled me more and more by the absence in all these sec-
tions, longitudinal as well as transverse, of any small branch
springing from the longitudinal marrow-trunks. Even the
faint traces which Macintosh believes to have observed now
and then did not make their appearance, and the better the
sections were preserved and stained the more conspicuous
was their absence. On the other hand, well-preserved sec-
tions of Schizonemerteans always show the thick coating of
nerve-cells surrounding the central, more fibrous, tissue of
the longitudinal trunks very clearly (PI. XXXII, fig. 6).
The whole of the trunk and cellular coating is again sur-
rounded by another layer of homogeneous connective tissue,
which separates at the same time the external layer of longi-
tudinal muscular fibres from the internal circular muscular
layer, not only in the vicinity of the nerve-trunks, but
throughout the whole extent of the body (cf. PI. XXIII,
fig. 12, in the July number of this Journal). In this homo-
geneous layer is situated another somewhat thinner layer of
tissue of an entirely different histological character.

ft is this last-named layer which we now have to ex-
amine in detail. Staining reagents, as picro-carmine, &c.,
have very little effect upon the bulk of the tissue, although
they render conspicuous in it the presence of numerous
nuclei, which distinctly stand out with a more or less deep red
tinge, hiach nucleus is provided with a nucleolus, and now

PERIPHERAL NERVOUS SYSTEM IN NEMERTINES. 435

and then traces of the circumference of a cell enclosing this
nucleus are visible. No difference can be detected between
these nuclei and those of the nerve-cells of the brain- and
marrow- trunks. The bulk of the tissue which remains
colourless has a finely punctate-fibrillar appearance, resem-
bling in this respect the fibrillar nerve-substance as we find
it in the centre of the longitudinal trunks of these worms.
A characteristic difference is, however, noticed in the fact
that the separate cells appear to be more numerous in this
fibrillar plexus than they are in the fibrous centre of the
longitudinal trunks.

This curious layer forms a cylindrical sheath throughout
the whole body between the longitudinal and circular mus-
cular layers. The median dorsal nerve (which in a former
paper I proposed to call the nerve for the proboscidian
sheath) appears as a simple thickening of this layer, and
laterally the longitudinal marrow-trunks may in the same
way be looked upon as parallel thickenings of the same
tissue, as the ganglion-cells in the longitudinal nerve-trunks
gradually merge, without any considerable difference in
histological character, into the cells of this layer. In the
sheath of ganglion-cells belonging to each longitudinal
trunk the boundaries of the cells are, however, more con-
spicuous and always visible, and this is not so distinctly the
case in the circular layers. This cylindrical tunic of distinct
tissue, which in front is applied to, when not in direct com-
munication with the tissue of the brain (PI. XXXIII, figs. 9
and 10), must no doubt be regarded, after what we have
noticed about its histological character, as a nervous layer.
Its constituents are nerve-cells as well as fibrillar nerve-
tissue.

PI. XXXIII, fig. shows a small part of this nervous
tunic in a horizontal section through the back of the animal ;

is the median dorsal nerve ; nl, the fibrillar plexus
with the separate cells, forming together a layer, differing
in thickness according to the size of the animal, and always
only partly visible in a single section. The longitudinal
{LM.) and circular {TM.) muscles are only sparingly indi-
cated in the figure, as are also the thin fibres w’hich
directly traverse the nervous layer (cf. figs. 6 and 7).

The hypothesis of the nervous nature of this layer is strongly
confirmed by processes which are sent out by it into the
adjacent muscular tissue, and which generally carry one (fig.
8) or more nerve-cells (fig. 11) ; sometimes none. At other
times these ramifications have a still more marked fibrillar
appearance ; for example, those wdiich leave the median dorsal

436

DR. A. A. W. HUBRECHT.

nerve and stretch vertically downwards towards the pro-
boscidian sheath. I believe I am justified in regarding these
processes as the equivalents of the smallest terminal twigs
of the nerve-branches iu the Hoplonemertini and other
animals with a dichotomically divided peripheral nervous
system. At the same time I must regard the nervous tunic
just sketched as representing a more primitive type of peru
pheral nerve-system, in which it has not yet come to a locali-
sation into transverse branches, metamerically placed, hut in
which one of the layers of the hody-wall is yet in function as
the recipient and conductor of nervous stimuli.

The facts observed and above described for the numerous
species and genera of Schizonemertini are so palpable and
unmistakable, and at the same time so constant,^ that the
question immediately presents itself ; — How do the less differ-
entiated and older types of Nemerteans, the Pal^onemer-
TiNi, and more especially the genus Carinella, behave in this
respect ?

PI. XXXII, fig. 4 and 5, may serve to give the answer to this
question. We find the whole of the central nervous system
directly under the epiderm, exteriorly to the muscles of the
body-wall, as was noted above. A thick basement membrane
separates the cellular parts of the skin from the subjacent
muscular tissue, and in this basal membrane we find in
favorable preparations (those in which this basal membrane
is somevvhat distended, being preferable) that the sheath of
nerve-cells {nc.) belonging to the longitudinal trunk is directly
continuous with a layer {nl.), which here again is provided
with nuclei of exactly the same character as those of the
nerve-cells and ensheathes the body exteriorly to the muscles.
The histological character of this layer corresponds to what
we have found in the nervous tunic of Schizonemerteans.
Here, again, a fibrillar structure and the scanty absorption of
staining reagents would suggest the idea of nervous-tissue,
even were the direct passage into the nerve-cells of the
lateral trunks less perfectly demonstrable. In front of
the brain this layer is continued in the heail, being more
conspicuous ventrally and laterally than dorsally. Here,
moreover, the fibrillar structure is far more apparent, and
the enclosed nerve-cells are less numerous.

In Polia and Valencinia the same layer is present, only

’ Tlie opinion on the nervous nature of this layer, first emitted in my
above-cited paper, ‘Zur Anatomie und Physiologic des Nervensystems der
Ncmertincn,’ p. 47, has only lately been confirmed for another genus of
Schizotiemcrteans by Herr Hewoletzky of Yienna (‘ Zoologischer Anzei-
gcr,’ No. 02, p. 399).

PERIPHERAL NERVOUS SYSTEM IN NEMERTINES. 487

enclosed between two muscular layers, as it was described
for the Schizonemerteans, and, moreover, corresponding with
the latter in the fact that this layer does not occur also in
the head in front of the brain, as it partially does in Cari~
nella. In all genera higher differentiated than the latter
this plexus-like arrangement in the head (which I first
mistook for a considerable number of parallel nerves, on
account of the appearance in transverse sections) appears
to have developed into a smaller number of distinct nerve-
stems, dichotomously dividing and innervating the tip of
the snout, the muscles of the head, and the eyes, when
present.

The fact of the presence of the nerve-sheath in Carinella
on the outside of the muscular body-wall and under the
immediate covering of the ectoderm is of the more impor-
tance, as we have indeed to look upon this genus as one of
the less differentiated of all Nemerteans^ which must serve
as a starting-point for comparisons by which to determine
the morphological significance of any structure which is
common to this and other genera.

In the preparation figured (PI. XXXII, fig. 4) I find that
processes {np.) are being sent out from this layer as they were
described for the Schizonemertini. Here I only succeeded,
however, in demonstrating such as were turned towards
the muscular layers ; in other sections I find that processes
towards the epidermic cells are also present (fig. 5). As
the nervous layer is generally closely applied to the muscles,
it is difficult to observe the centripetal processes, and only
in such a case as the one figured, where by some reason or
other this layer lies somewhat removed from the subjacent
muscles, can these important ramifications be noticed.

The diagrammatic figures I, 2 and 3, will now be more
intelligible after this rapid exposition of the facts. Fig. I
applies to Carinella. The lateral trunks and the nervous
layer are wholly outside the muscular body-wall, processes
from this layer being sent outwards and inwards. Fig. %
stands for Valencinia, Folia, and all the Schizonemertini.
The trunks and the layer are enclosed loithin the muscles, the
peripheral processes are sent out in both directions. Fig. 3
finally represents the Hoplonemertini, a stage of higher
development, in which the continuous cellular layer uniting
the longitudinal trunks and ensheathing the whole body has
become localised and differentiated into separate nerve-
branches springing from these trunks at regular intervals.

' Cf. A. A. W. Hubrecht, “The Genera of European Nenierteans
critically revised,” ‘Notes from the Leyden Museum/ vol. i, p, 19.

438

toR. A. A. W. IIUBRECIIT.

That a tendency to a metameric arrangement of the other-
wise indifferent layer is nevertheless already present in the
ScHizoNEMERTiNi is shown in fig. 13, which illustrates a
thin section through the longitudinal muscular layers, im-
mediately above the nervous sheath. Three of the peripheral
processes taking their course through interstices between
the^ muscles have been cut in this section ; they are
regularly arranged along a transverse row, and numerous
other sections show that indeed an approach to regularity
may in this respect be noticed all over the body. If this
regular arrangement gradually combines with a thickening
of the nervous layer along the points from whence these
processes originate, and with a corresponding reduction along
the intermediate spaces, the formation of transverse indi-
vidual nerves (which at first would have the appearance
of commissures,^ and in a later stage would take the aspect
of transverse branches springing from the longitudinal
trunks), carrying perpendicular smaller twigs, would be the
inevitable result. And it is in this way that we must
form our conceptions of the intermediate stages lying
between Carinella and the Hoplois^emertini. As such must
also be regarded the plexus-like arrangement between two
successive peripheral branches, as observed by Macintosh
and noted above.

A certain variability in the aspect of the nervous tunic in
small and large specimens, which, though generally a con-
tinuous layer (even if not of considerable thickness), on
other occasions has the appearance of a plexus with very
narrow meshes, confirms the supposition that local varia-
tions of the thickness of this layer, finally leading to the
formation of separate nerve-stems, may occur. Generally
bundles of radial fibres (^.), which in other places are more
isolated, take their course through these meshes.‘^

^ In a recent paper of von Kennel’s (‘ Die in Deutschland gefundenen
Land-planarien,’ Wurzburg, 1879, p. 39) he speaks (quite aphoristically)
of transverse commissures uniting the longitudinal trunks of Nemerteans
across the back of the animal. It may be that what he took for commis-
sures are portions of the nervous sheath above described ; still it appears
improbable that this excellent observer would have overlooked the nervous
tissue which unites the longitudinal trunks ventrally as well. Surface
sections (PI. XXXIII, fig. 12) show that this questionable arrangement of
the tissue in transverse commissures, now and then met with in longitudinal
sections, is merely apparent.

2 When I showed some of my preparations to Prof. E. Ray Lankester
last April, he drew my attention to the apparent agreement between this
nervous layer and Strieker’s nerve-layer of the epiblast of the embryo-
frog. I am not disposed to believe that any deep significance can be
attached to this mthcr distant resemblance; but the parallelism, such
as it is, is interesting.

PERIPflElUL NERVOUS SYSTEM IN NEMERTINES, 439

Another question which future investigations will have to
solve, and which at present I can no more than briefly call
attention to, is this — How are we to conceive the nervous
sheath to have gradually become enclosed between the
muscular layers ? More especially can the difference be-
tween Carinella and the Schizonemerteans be explained ?
If I venture to emit an hypothesis at present for
which I cannot as yet furnish sufficient or conclusive evi-
dence, and which presented itself to me in studying the
beautiful investigations of 0. and H. Hertivig, on the struc-
ture of iheActinice (^JenaischeZeitschrift,’ 1880), it is because
it appears at the same time to explain the fundamental
difFerence in the muscular system of different groups of
Nemerteans, which was first clearly pointed out by Macin-
tosh. The hypothesis might be briefly formulated thus : the
external longitudinal muscular layer in Polia, Valencinia,
and all the Schizonemertini must be regarded as a special
development of the ectodermal musculature.’^ If this be true
(and the curious arrangement of the external longitudinal
muscles in Polia compared with Carinella^ on the one hand,
with the Schizonemertini, on the other, appears to furnish
strong arguments in its favour), the wandering inward of the
nervous system can be more clearly understood, those muscles,
which from the first were situated exteriorly to it having
gradually attained this strong development and fascicular
arrangement. The inner longitudinal and the circular
muscular layer, as they are present in Carinella and the
Hoplonemertini, would thus represent the typical mus-
cular system of Nemertine organisation, whereas the ex-
terior layer of longitudinal muscles, which is added to it in
the Schizonemertini, Valencinia^ and Polia, would appear
to be of accessory character.

The curious distribution of nervous tissue above described
is, I believe, of importance towards the solution of two
problems touching the physiology of these worms.

These problems are — 1st. The extraordinary restorative
power of a great number of Nemerteans. 2nd. Their intense
sensitiveness to outward stimuli, Avhich may even occasion
spontaneous rupture of the animal into two or more pieces.

As to the restorative power, it has often been noticed (Pa-
lyell, Macintosh, Barrois) : and detailed observations are on
record of severed portions of the trunk having regained a com-
plete central nervous system — brain, respiratory lobes, and
eyes included — when only a cylindrical fragment, open at both
ends, was left. The tail end is always sooner restored, but
under favorable circumstances the restoration of the head

440

DR. A. A. W. HUBRECHT.

and brains does not seem to be rare. Although I never
had occasion to study the phenomenon myself, the descrip-
tions oiBarrois and Macintosh do not leave any doubt as to
the way in which it progresses. Three months after the
fragment was separated from the parent it was again pro-
vided with the normal brain-lobes and all their accessories,
which had developed at the anterior ends of the nerve-
trunks after the cicatrising process at both ends of the frag-
ment had been concluded. Another fragment was severed
from the parent animal in September, developed perfect
spermatozoa and normally deposited them in February next.

Looking upon the lateral trunks as local thickenings of the
more indifferent and primitive nerve-sheath in the body-wall,
these curious facts appear in another light. For if we have
succeeded in demonstrating that nervous tissue, which per-
sists in a more primitive stage, the 'cellular elements being
scattered indiscriminately among the fibrillar, is diffused
throughout the whole circumference of the body-wall, then
the appearance of local thickenings in this tissue finally
leading to the formation of lateral trunks, and eventually of
brain-lohes sensu strictiori, is less startling than would be
the reproduction of the central apparatus out of lateral
trunks, which hitherto have been regarded as essentially
peripheral.

I must here lay stress upon the fact that in the Hoplo-
NEMERTINI, where this primitive^ nerve-sheath has already
been replaced by separate diverging peripheral branches,
and, which, therefore, have attained a far more specialised
stage respecting the structure of their nerve-system, no case
of similar reproduction of the head and brains has ever yet
been noticed, as far as I know.

All such observations were made upon Schizonemertini,
and I do not believe it will prove a hazardous prophecy to
predict that later investigations will show that in the armed
species it really never occurs. At the same time the latter
die much quicker after having been severely injured than do
the Schizonemerteans, the families of Ampliiporid<2 and
TetrastemmidcB enjoying a still more restricted longevity than
do the family of Nemertidce s. str.

All this appears to me to be to a great extent explained
by the above described fundamental differences in the dis-
tribution of the nervous tissue in the body-wall.

As to the extreme sensitiveness to outward stimuli, it has
been noticed by all former observers, and Leuckart applied the
name of somatotomus to one of the species, because it regu-
larly broke itself into pieces upon being handled or touched.

PERIPHERAL NERVOUS SYSTEM IN NEMERTINES. 441

This sensitiveness, with which the Schizonemertini are
none the less endowed than the Hoplonemertini, would
appear the more remarkable in the former in the absence of
peripheral branches springing from the longitudinal marrow-
trunks. The presence, however, of a sheath of nervous
tissue in every region of the body, either immediately under
the epiderm or inside the muscles, and sending out processes
as well to the exterior surface as to the internal muscular
layers, is quite as sufficient an explanation for this
extreme sensitiveness as we before showed it to be for the
persistence of life in the several fragments, which even de-
veloped normal spermatozoa months after they had become
separated.

x\s will be understood from the contents of the foregoing
pages, my investigations into this curious distribution of
nervous tissue in the body- wall of the Nemer tines are far
from being either complete or exhaustive. Not having as
yet had occasion to examine fresh specimens with special
regard to this nervous layer, I must reserve the more delicate
histological questions for a future publication. The direct
connection between the fibrillar plexus and the exterior
cellular layer of the body, the presence of any distinct
sensory cells in this ectodermic layer, as well as the connec-
tion wdth the muscular fibres, on the other hand — all this can
only be studied in careful preparations, teased out from the
tissues when quite fresh, according to the delicate methods
which have, for example, been so successfully applied by
0. and R, Hertwig, in their splendid researches upon the
nervous system of the Coelenterata (‘Das Nervensystem und
die Sinnesorgane der Medusen,' and further ‘Die Actinien,
anatomisch-histologisch mit besonderer Beriicksichtigung
des Neuromuskelsystems,’ 1878-80).

Having as yet only had series of sections mounted in
balsam at my disposal, I could not expect to penetrate into
all these histological details merely by their aid ; and if,
nevertlieless, I venture to publish these preliminary re-
sults it is because they appear to me to possess an increased
interest now that the researches above mentioned ’ of
O. and R. Hertwig have come before the public. The very
lowest stage of differentiation in which nervous tissue is at
present known in the animal kingdom was found by these
searchers! in the Actinia, wffiere it appears as a fibrillar
plexus, with nerve-cells indiscriminately scattered in it

* Might not this word take the place of the barbarous savants^" so
often met with in English scientific papers of the present day 'r* It would
correspond to the German “ Forscher.”

442

DR. A. A. W. HUBRECHT.

throughout a great part of the ecto- and entoderm. In the
oral plate (Mundscheibe), where the nerve-cells are more
numerous and situated closer to each other, there is a ten-
dency towards a certain degree of centralisation.

It cannot be denied that this primitive arrangement offers
strong analogies to the nervous plexus above described for the
Nemertines, which, on the other hand, approach, in the his-
tology of their nervous system, in many respects to what
A. Lang has so carefully described for other Platyelmi^jthes

Untersuchungen zur vergleichenden Anatomie und Histolo-
gie des Nervensystems der Plathelminthen.’ 1. Das Nerven-
systems der marinen Dendrocoelen. 2. Das Nervensystem
der Trematoden ; Mitth. aus der Zool. Station zu Neapel,
1879-80). And so I thought that giving my results, even in
their incomplete form, might be suggestive of further re-
searches into these questions; the more so because, in a former
paper Verhandelingen v. d. Kon. Akad. v. Wetensch. te
Amsterdam,’ vol. xx, 1880 ; cf. July number of this Journal),
I insisted upon the phylogenetic importance which the
nervous system of the Nemerteans appears to possess when
compared to that of the Anxulata, on the one hand, and
of the Yertebrata, on the other.

In the monograph on this group of worms, which I have
now in preparation for Professor Dohrn’s series of publica-
tions, illustrating the Fauna and Flora of the Bay of
Naples, I hope to be able to' furnish evidence about
those points in w^hich the present paper confesses to be
deficient.

THE EYE OF PECTEN,

443

The Eye of Pecten. By Sydney J. Hickson, B.Sc.,
Scholar of Downing College, Cambridge. (With Plates
XXXIV and XXXV.)

The general absence of organs of vision amongst the
members of the class Lamellibranchiata meets with a curious
and interesting exception in the genera Pecten and Spon-
dylus.

These genera have long been knowm to possess a great
number of eyes of considerable complexity, situated on the
border of the mantle. The number of these eyes varies
considerably in different individuals, ranging in the genus
Pecten from eighty to one hundred and twenty. Their
position also varies ; for, although they are always situated
on the border of the mantle, yet sometimes they are placed
at equal distances from one another, and sometimes they are
clustered together in certain localities.

Notwithstanding this indefinite element, both in their
number and position, which might be expected to run
parallel with a primitive and simple organisation, their
anatomy is exceedingly complicated, and exhibits all the
most important structural elements of the eyes of the higher
Vertebrata.

The earliest investigations into the anatomy of Pecten’s
eye are those of Krohn,^ who gives a drawing of the course
of the optic nerve. This drawing is copied in many of the
subsequent papers on the subject by other investigators, and,
as far as it goes, is correct. Duvernoy,^ in his description
of the nervous system of the Pectens, gives a short descrip-
tion of the anatomy of the eye. This paper, however, is
chiefly valuable for the excellent figures and descriptions
of the distribution of the nerves in the mantle, and the
filaments which are given off from the main trunks of these
to supply the tentacles and the eyes.

The researches of Blanchard^ and of Keferstein^ which
followed did not add very much to our knowledge on this
subject, and it was not until 1865 that any careful histo-^
logical inquiries were carried on. It was Hensen*’ who first

' Krohn, ‘ Muller’s Arcliiv/ 1840, p. 301, pi. xi.

- Duvcrnoy, ‘Menioires de I’Acadeinic de Sciences,’ t. xxiv, 1853.
‘ Mernoire sur le systeme uerveux des Acephales,’ p. 73, pi. ii.

^ Blanchard, ‘ Organisation du regne animal : Mollusques Acephales.’
Kcifcrstein, ‘Zeit. fiir wiss. Zoologie,’ 1803, p. 133.

* lienseu, ‘ Zeit. fiir wiss. Zoologie,’ 1865, p. 220.

444

SYDNEY J, HICKSON.

gave figures of the characters of any of the histological
elements. But as the eye of Pecten forms only a very small
part of the paper^ his figures and description are by no
means complete^ and in many respects they are incorrect.
Finally, J. Chatin^ has contributed two short papers, with-
out figures, on this subject.

Of the scanty literature Hensen’s paper is by far the
most important, and he alone gives any good figures of
sections of the eye, or of its elements ; the other observers
give remarkably few figures, and consequently I have had,
owing to an imperfect knowledge of the German language,
some difficulty in making myself acquainted with the sub-
stance of their papers.

I have been encouraged in publishing the following re-
searches chiefly by this scarcity of good figures, but also
because I believe, and will give my reasons for believing,
that these eyes deserve more mention than is usually made
of them in our zoological text-books.

My investigations were chiefly carried on upon Pecten
maximuSj but I have also had the opportunity of making
sections of and studying the eyes of two other species,
Pecten Jacob ecus and Pecten opercularis. The eyes of these
three species differ from one another in one or two not
altogether unimportant particulars, and, as I shall after-
wards point out, they form an interesting gradation, the
points of difference between P. ^maximus and P. oper-
cularis passing through intermediate stages in P.jacohceus.

The eyes of Pecten maximus — are situated amidst a number
of tentacles, which run all round the border of the mantle.
These tentacles are capable of considerable movement, and
frequently overhang the eyes and protect them from the
light. The eyes themselves are situated upon short stalks,
which resemble very closely the basal part of an ordinary
tentacle.

This similarity caused Duvernoy to name a tentacle a
tactile pedicel, and an eye an ocular pedicel, thus to a cer-
tain extent implying that they are morphologically homo-
logous organs respectively modified for a tactile and an
ocular function. This homology is justified by certain
points in their anatomy, such as the course of the nerve and
the arrangement of the muscular fibres, and I believe that
when the development of these eyes is studied the homology
will be still further confirmed.

The border of the mantle which bears the tentacles and

^ J. CliatiD, ‘Bulletin de la Societe Pliilomatique.’ Paris, 1877.

THE EYE or PECTEN,

445

eyes is covered with an epithelium, consisting of columnar,
non-ciliated, and slightly granular cells bearing nuclei,
situated near the base of the cells. As this epithelium
passes over the eye-bulbs, it undergoes two interesting modi-
fications. It becomes considerably thicker and filled with a
dark brown pigment (PI. XXXIV, fig. 1 c) as it passes round
the sides of the eyes, but immediately in front of the
eye (PL XXXIV, fig. 1 a), it again diminishes in thickness,
and becomes perfectly transparent. By thus surrounding
the eye on all sides with a dark-coloured pigment, leaving
only a round spot in front, clear and transparent, the epi-
thelium, by limiting the entrance of the light to a small
diaphragm in front, here performs the function of an iris.
The epithelium which runs over this transparent part, and
which forms the epithelial layer of the cornea, differs from
the ordinary epithelium covering the rest of the mantle in
that their cells are rather larger, are perfectly transparent in
the living condition, and the nuclei are large and spherical,
and situated in the centre of the cells.

The eye consists of the following parts, which I shall now
describe in order. The cornea, covered externally by its
transparent epithelium, protects a large elliptical lens.
Close up to the lens is the retina, but separated from it by
the optic nerve, which spreads out over the anterior surface
of the retina. The retina rests upon a tapetum, and behind
this, occupying all the posterior concavity of the eye-cup,
there is a red pigment.

The cornea — consists of two parts, the outer epithelium,
which has already been described, and a basement mem-
brane, consisting of a thin layer of connective tissue. As
before stated, this epithelium is merely a modification of the
general epithelium of this part of the mantle; and the pig-
mented epithelium surrounding the eye-bulbs (in like manner,
a modification of the same tissue) is continuous with it
all round its edge. The passage of the cells of the pig-
mented epithelium into those of the corneal epithelium is
signalised by two important changes in the characters of the
cells. In the first place the pigment entirely disappears,
and the nuclei, which in the former case were obscured by
the pigment, now becomes apparent, and in the second place
the cells are considerably diminished in their longitudinal
axis. The diminution in size of the cells causes the edge of
the cornea to be sunk below the level of the pigmented e])i-
thelium ; and a shallow trough runs round the line of its
juncture with it (Bl. XXXIV, lig. 3). The convexity of the
cornea is not great, and the dome of it frequently only just

VOL. XIX. NEW SEK. G G

446

SYDNEY J. HICKSON.

reaches the level of a line drawn from the highest points of
the pigmented epithelium on either side of it. This appear-
ance is not often seen in sections, as the pigmented epi-
thelium rapidly shrinks, when the tissue dies, and under
most reagents ; hut I am fully persuaded of the accuracy of
this statement from an examination of the eyes of living
specimens of Pecten maximus and sections of Pecten oper-
cular is.

The delicate epithelial cells of the cornea, in consequence
of being entirely unprotected by any membrane similar to
the conjunctiva of the higher animals, are quite naked, and
very liable to injury from the rough edges of the tentacles
which surround them. The arrangement just described,
however, probably prevents the tentacles from coming into
immediate contact with them. The little trough which runs
round the margin of the cornea always contains a little
liquid, even when the eye itself is removed from the water ;
and the pressure of the tentacles when folding over the eye
causes it to spread out as a thin layer over the cornea, and
thus the cells are prevented from coming into immediate
contact with the tentacle.

Thus, the two remarkable modifications, namely, the pre-
sence of a large quantity of pigment, and a greater longi-
tudinal axis of the cells which the pigmented epithelium
exhibits, are of considerable value to the eye, firstly, to pre-
vent very divergent rays from entering, and secondly, to pre-
vent any damage to the cornea caused by the rubbing of the
adjacent tentacles over the sides of the eye.

The second layer of the cornea is about half as thick as
the epithelial layer, and, like it, is perfectly colourless and
transparent. It consists merely of a thin continuation of
the connective tissue of the stalk. It may be called the
basement membrane of the corneal epithelium, as from the
absence of any definite cellular elements its only function
probably is to support these cells.

Beyond the cornea this membrane becomes much thicker,
and supports the pigmented epithelium, and at the same
time structural elements make their appearance in it. From
thence it passes into the connective tissue of the eye-stalk
without further modifications.

The lens — is one of the most interesting parts of the eye.
It is comparatively large, and is composed of a number of
nucleated cells. In the fact that the lens is formed by
more than one cell tire eye of Pecten bears an interesting
resemblance to thatqf the Vertebrata. The shape of the lens
has been a subject of much dispute amongst the authors

THE EYE OF PECTEN.

447

Avlio have written on this subject. Krohn and Keferstein
believed it to be spherical. Hensen has figured it filling up
the space between the cornea and retina, and consequently
of an irregular bi-convex shape.

It is difiicult to see how a controversy on such a simple
subject could have arisen, unless it is because different
authors have examined different species, and described them
for the genus.

x\s regards Pecten maximuSy^n examination of the fresh eye
has convinced me that in this species the lens is elliptical, the
major axis being parallel to the plane of the mantle. A section
of the eye made in a plane at right angles to the plane of the
mantle and the direction of its margin — that is, the plane
which is most convenient for section-cutting, and the one
which is apparently usually adopted — would consequently
cause the lens to appear circular in section. In the dia-
grammatic representation of the eye (fig. 1) I have for con-
venience sake represented the lens as being at right angles
to the plane of the mantle in order that the true shape of
the lens may not be overlooked.

A fresh examination of the lens, when teased out from the
rest of the eye, exhibits one or two interesting points. The
lens is not, as in most eyes, perfectly colourless, but possesses
a well-marked brown colouration, and a number of fine striae
may be seen running in the direction of the major axis.
The lens does not appear so perfectly elliptical in the fresh
condition as in certain sections I have made; it is drawn
out somewhat longitudinally, so as to be more like a double
cone than an ellipse. This is probably due to the lens being
released from the ligaments and connective-tissue pressures,
which cause it to retain its proper shape.

Hensen says that the lens is very soft, and the cells are
light, polygonal, and nucleated. A careful examination
of the lens of P. maximus has led me to a very different con-
clusion. The lens seemed to be of exactly the same nature
as in the higher forms, and when teasing it out I found some
difficulty in holding it with a needle, as it slipped away from
under it when a slight pressure was exerted. As regards
the shape of the cells composing the lens, they are not all
polygonal, as wmuld be inferred from Hensen^s remarks on
the subject. In the centre they are polygonal, but as they
approach the periphery they become more and more flat-
tened and elongated, until at the periphery they are strap-
shaped. They are nucleated. As Hensen, 1 could find no
membrane covering the lens, and no muscular fibres con-
nected w'ith it; but in a few cases I liave observed a liga-

4i8

SYDNEY J. HICKSON.

ment, such as 1 have represented diagrammatically in
fig. 1/, which, I believe, forms a support for the lens. This
ligament is usually broken by the action of reagents, and
then hangs down by the side of the cavity, and thus becomes
difficult to observe ; at the same time the lens sinks down,
and rests upon the anterior surface of the retina.

The lens is suspended in the space which corresponds
with the vitreous humour in the higher animals. This space
is filled with an aqueous humour in Pecten. The lens is
larger, and, consequently, the space occupied by aqueous
humour relatively smaller in P. maximus than it is in either
P,jacoh(2Us or P. opercularis, and in P.jacohceus it is larger
than in P. opercularis.

The retina — does not line the concavity of the eye-cup, as
it does in most well-developed eyes, but is nearly flat, and
a considerable space is left between it and the floor of the
cup, which is filled up by the red pigment. In conse-
quence of this the retina appears in section to be a thick
band crossing the eye from side to side. Thus, just as the
lens was remarkable for the way in which it approached the
retina by hanging back into the cavity, so the retina is re-
markable for the manner in which it leaves the posterior
concavity of the eye-cup to approach the centre. The eye
of Pecten, in fact, presents the interesting peculiarity of the
approach of the lens and the retina towards the centre, so
that in P. maximus they almost touch.

The anterior surface of the retina is convex at the sides
and concave in the middle, but these convexities and con-
cavities vary in different species. The different layers of the
retina will be described from behind forwards, as it will be
easier to trace the transitions in that way than if described
from before backwards. They are — 1°. Posterior limbs of
the rods. 2°. Anterior limbs of the rods. 8°. Spindle-
shaped nucleated rods. 4°. Molecular and nuclear layer.
5°. Nerves.

The posterior limbs of the rods stand upon a membrane,
which runs along the posterior side of the retina; at their
anterior ends they pierce a very delicate membrane, and pass
into the anterior limbs of the rods. The anterior' limbs are
about twice as long as the posterior limbs, and are usually
smaller in diameter, and situated farther apart than the pos-
terior limbs. That they are circular in section may be seen
from PI. XXXV, fig. 7<z, which is a drawing of a section
made at right angles to the eye-stalk. The anterior limbs
of the rods are sometimes swollen so as to appear oval ;

THE EYE OF PECTEN.

449

this condition occurs especially in the rods at the side con-
vexities. Fig. 6 represents an isolated rod in this con-
dition.

The anterior ends of the rods contract considerably, and
again expand into spindle-shaped bodies, each of which con-
tains a nucleus ; so that in P.jacobceuSy where the retinal
elements of this region are difficult to distinguish, there
may be seen a single row of nuclei running from end to end
of the retina, and following its sinuosities (Plate XXXV,
fig. 115).

In some of the rods at the side of the retina a second
spindle-shaped body follows the first one, as represented in
the isolated rods in figs. 5, 6, but usually the anterior end
of the spindle is drawn out into a delicate thread, which
occasionally possesses nuclear swellings. Finally, this thread
breaks up into a network, which bears a number of nuclear-
like bodies at its nodes, and several round molecular bodies
appear to be caught in its meshes. These bodies are so
much like the ordinary nuclei of the network that I am in-
clined to believe that they are, in reality, merely modifica-
tions of them, and in some way connected with the network
(fig. 6 a). Anteriorly the fibres of the network bend at right
angles and enter the nerve layer, which covers the anterior
surface of the retina. This nervous layer will be described
with the description of the optic nerves.

The above is a description of the retina as I found it in
P. maximus, and I believe it holds good for the other mem-
bers of the genus. The elements of the retina are so much
larger in this species, and the spaces between the rods and
network, &c., so much more considerable, that it is a great
deal easier to investigate ; but I believe careful examination
of the other species would show that they do not differ from
this in any important detail.

The tapetum — is placed immediately behind the retina, and
may help in its support. When fresh,^ the tapetum exhibits
a display of colours, and it is this membrane which gives
the eyes their beautiful metallic lustre. When examined
with a 4th-inch obj. it seems to be composed of a great
number of little black specks separated by a fine yellow
membrane, but careful examination with a higher power
shows that it is composed of a great number of fine fibrils
crossing at right angles.

The space between the tapetum and the posterior part of

‘ 1 have one series of seetions stained in osmic aeid, and mounted in
Canada balsam, whicli lias retained this display of colours.

450

SYDNEY J. HICKSON.

the eye-cavity is filled with a red fluid pigment. In the fresh
condition the pigment readily flow's on to the slide when the
eye is pricked, but in sections of the eye which has been
hardened by alcohol or other reagents the pigment adheres
to the tapetum or posterior w'all of the eye-cup.

Hensen figures a layer of cells in this position, but I have
never been able to observe anything of the kind ; the pigment
contains no cellular elements at all, nor is there a layer of
cells lining the cavity which contains the pigment. The
pigment consists of a number of bright red granules floating
freely in a colourless fluid.

The nereous supply — of the eye of Pecten is perhaps the
most interesting of the many peculiarities of this eye. The
nervous system of Pecten is well described by Duvernoy in
the paper referred to above. The mantle is supplied by a
number of branches given off from the principal ganglia.
These branches all fall into a large nerve, which runs round
the margin of the mantle, and which Duvernoy calls the
circumpaliaP^ nerve. This nerve is figured in section in
fig. 1, PI. XXXIV, one of the nerves joining this nerve being
figured at fig. I, q. This circumpaliar^ nerve gives off
filaments to supply the tentacles and eyes.

Krohn first gave a drawing of the optic nerve, and described
it as a single nerve passing off from this trunk, and dividing
into two branches as it aproaches the eye. Later observers
have, however, drawn and described two nerves passing off
from the circumpalial” nerve. My researches have led
me to believe that Krohn is right, and that such a figure as
Hensen gives in his paper, representing two main trunks
passing up to supply the eye is erroneous. Plate XXXV,
fig. 9, of P. maximuSi shows the division of the single nerve
into its two branches. In fig. I the course of the optic nerve,
.before its division into two branches, was carefully drawm
from one of a complete series of sections, and in none of the
other sections could I find a trace of any other nerve pro-
ceeding from the circumpalial.” The branching of the
nerve takes place in a plane at right angles to the plane of the
mantle. When the optic nerve approaches the eye it divides
into two branches, which may be called the ‘^retinal nerve”
and the complementary nerve.” The former passes up the
side of the eye cavity, and spreads over the anterior surface
of the retina ; the latter soon loses its sheath, and divides
up into a number of branches, w'hich supply the tissues sur-
roundiijg the eye. The course which the retinal branch
takes may be seen in PI. XXXIV, fig. 1, and in PI. XXXV,

THE EYE OF PECTEN.

451

figs, 8 and 9. In figs. 8 and 9, the first section is cut through
the optic nerve, and shows the manner in which the retinal
branch runs up the side of the eye-cavity ; the second section
shows the manner in which the branch bends over on to the
retina and spreads out. The distribution of the comple-
mentary branch is diagrammaticaily represented in fig. 1 n ;
it seems to divide into a number of branches which enve-
lope the eye-cup, and probably send filaments to the cornea,
lens, tapetum, and epithelium.

Comparison of the eyes of the three species, P. maximus,
P. jacohceus, and P. opercularis. — The eye of P. maximus
is undoubtedly the most highly developed, the eye of P.
opercularis is the simplest, whilst P. jacohceuSy although
more like P. opercularis than P. maximus^ shows many
points in which it is intermediate between the two.

The lens in P. opercularis is separated from the retina by
a considerable space (PL XXXV, fig. 10), and consequently
the chamber containing the humour is relatively large. In
P.jacohceus the lens is larger than in P. opercularis ^ and
the chamber consequently smaller ; and in P. maximus the
lens is very large, and nearly touches the retina, the cham-
ber of the eye being sometimes very small. A gradation is
thus observed in the character of this part of the eye in the
three species. In P. maximus but a small space is filled
with humour, in P. jacohceus a much larger space is filled
with it, and in P. opercularis there is a larger space still.

Again, when the retinas of the three species are compared,
a similar gradation is found. The retina of P. opercularis is
comparatively thin, and the concavity and convexities of its
anterior surface slight. In P. jacohceus the retina is de-
cidedly thicker, and the anterior surface is more convex at
its sides than in P .opercularis ; moreover, it may be noticed
that the delicate membrane which separates the anterior
from the posterior limbs of the rods has become bent up in
the regions corresponding with the anterior convexities of
the retina. In P. maximus all these variations become
much exaggerated. The retina is much thicker than in
either of the other species ; and the side convexities of its
anterior surface are much bolder (PI. XXXV, fig. II, a, 5, c).
The anterior concavity does not undergo much variation.

The shape of the membrane separating the anterior and
posterior limbs of the rods is greatly altered. InP. opercularis
this membrane is observed, in section, to stretch from side
to side without any well-marked curves ; in P. jacohceus two
'well-marked curves, corresponding with the anterior con-

452

SYDNEY J. HICKSON.

vexities of the retina, are observed ; but in P. maximus these
curves are converted into two distinct folds, which run up
into the substance of the retina. The membrane between
the folds does not sink again as low as it is at the com-
mencement of the folds, and consequently the central
part of the retina is raised in the form of a table above
the level of its sides. This elevation of the central part
of the retina may be also seen in P. jacobaeus, though it
is not nearly so well marked. The folds which occur in
P. maximus cause th; rods to appear to be given off in a
pinniform manner at the sides of the retina, and before
I found the intermediate condition in P. jacohmis I had
some difficulty in determining the true relationship between
the retinas of P. maximus and P. opercularis. (Compare
a, h, c, fig. 11).

In addition to those just mentioned there are other minor
points in which the eyes of these species differ from one
another, such as in the shape of the cells composing the
lens and in the distribution of the retinal nerve, &c., but
they are comparatively slight.

General considerations. — Plaving thus described, in some
detail the anatomy of the various parts which compose
the eyes of Pecten, I shall, before leaving the subject, point
out some of their interesting morphological peculiarities.
It is, in itself, a remarkable thing to find a large and
variable number of eyes situated oii an area at some con-
siderable distance from any central nerve-ganglion ; and,
when it is remembered that the class and even family (with
one other exception, e. g. Spondylus) to which the genus
belongs, possess no organs of vision at all in the adult con-
dition, it is altogether surprising that they should be of such
extraordinary complexity as they have proved to be. The
high structural development that this eye h.is attained is,
however, not so remarkable as the fact that in many ways
it differs from the ordinary Invertebrate eye, and resembles
that of the Yertebrata.

In the first place, the lens is built up of a large number of
distinct nucleated cells, which undergo a flat'ening at its
circumference very similar to that found in the eye of the
Yertebrata. Whether the lens is developed from the cells
of the epiblast, as in the Yertebrata, or from the mesoblast,
must at present be left unsettled, but it will probably be
found, when the development of the eye is studied, that in
this respect also it resembles the eyes of the Yertebrata.
The tapetum, a structure which is of considerable im-

THE EYE OF PECTEN.

453

portance to animals which are nocturnal or aquatic in habit,
has hitherto been described only in the Vertebrata. That
Pecten possesses a tapetum as highly developed as any
found amongst the Vertebrata is anatomically a point of
considerable interest; but it also indicates to a certain
extent the physiological capability of the eye.

The chief interest, however, lies in the relative positions of
the optic nerve, the retina, and the pigment. In the eyes of
the Cephalopods the pigment layer is situated in front of the
rods, and the nerve-libres enter the rods from behind. In
the eyes of the Gasteropoda, the Crustacea, &c., down to the
simplest form of eye, such as that of the Rotifera, the same
relationship of these parts holds good. In the Vertebrata,
however, their relative positions are reversed ; the optic nerve
pierces the retina, and distributes itself over the front of the
retina, whilst the pigment’is situated behind it. In Pecten
the relationship of these parts is the same as that in the
Vertebrata; the nerve passing up the side of the eye-cup
bends over, and spreads itself over the anterior surface of the
retina. The pigment also is situated behind the retina.
Pecten is not*, however, the only Invertebrate whose eyes
are built up on this type. Semper^ has recently pointed out
that on the backs of certain slugs (Onchidium) a number of
eyes are found, and that in these the nerves pass to the front
of the retina before being distributed. On account of this
distribution of the optic nerve he says that they belong to
the Vertebrate type of eye (typus der Wirbelthieraugen), so
that two animals are now known, each belonging to a large
and important class of In vertebrata (Gasteropoda and Lamel-
libranchiata respectively) that possess eyes which are built
up on this type. The eyes of Pecten are even more deserving
of the name of Wirbelthieraugen than those of Onchidium,
for they are much more highly developed, and possess, in
addition to this relationship of the nerve and retina, other
Vetebrate peculiarities. The lens is multicellular, a character
which, although not unknown amongst the Invertebrates, is
much more characteristic of the Vertebrata. The tapetum,
too, a stru(:ture which doubtfully exists in any other Inverte- .
brata is found in Pecten and some Vertebrates. But, although
the application of this word Wirbelthieraugen to these eyes
is convenient for the adult condition, it must be carefully
remembered that the development of these eyes is essentially

* Semper, ‘ Uber sehorgane vom Typus der Wirbelthieraugen auf dcm
Riicken der Schnecken.’ Wiesbaden, 1877.

• 454

SYDNEY J. HICKSON.

different from that of the Vertebrate eye. The Vertebrate
eye is formed in the embryo from a hollow process given^off
from the brain^ and the future eye-cup is formed by an in-
vagination of this process. It is impossible for the eyes of
Pecten or Onchidium to be formed by any process similar
to this. Thus^ in the young state these eyes are essentially
different from those of the Vertebrata^ and the resemblance
in the adult is merely accidental, and by no means due to
morphological identity.

Little is known and little can be said concerning the
function of the eyes of Pecten. The presence of such a
well-formed tapetum makes it probable that they are
capable of appreciating very diffused light, and the close
approximation of the lens to retina makes it exceedingly
improbable that any image is formed upon the latter.

A few experiments have been made on the extent of their
visual power, which make it very doubtful whether they are
of much value to the animal in avoiding its enemies. The
most reasonable theory of their function seems to be that,
when on the ebbing of the tide, a probability arises that
they will be left higli and dry on the shore, they can appre-
ciate the fact by the growing intensity of the light, and,
by that peculiar flapping motion of their valves the
Pectens are so remarkable for, move away into deeper
water. '

These researches were entirely carried on in the morpho-
logical laboratory of the University of Cambridge, and my
best thanks are due to Mr. Balfour for his valuable advice
and encouragement during the whole course of my researches.
Owing to his kindness, also, I have been enabled to ex-
amine some of Semper’s preparations of the eye of
Onchidium, to which reference has been made in the text.

Methods, — For a general examination of the eye the best
method is to harden in alcohol and stain by immersion in
haematoxylin for twenty-four hours. Of the osmic-acid acid
preparations the best were obtained by immersion in
a 1 per cent, solution for fifteen minutes, followed by abso-
lute alcohol for three or four days. This method is of great
value for studying the retina and lens. I have also used
gold chloride for staining the nerves with some success.
For examining the tapetum the best preparations I have were
made from some eyes given me by Mr. Haddon, which had
been treated with picric acid. This reagent seems to have
dissolved away the red pigment, and consequently left the
tapetum free from the numerous little red granules which

THE EYE OF PECTEN.

455

generally cling to it. For examining the isolated rods of the
retina I have allowed the eyes to remain in a solution of
chloral hydrate for four or five days. I have then dissected
out the retina with needles as carefully as possible, and
poured a drop or two of heematoxylin on to the slide. When
the retina had been standing in hsematoxylin in this manner
for some hours it was washed with water, teased out with
fine needles, and mounted in glycerine.

• 456

L. RANVIER.

On the Terminations of Nerves in the Epidermis. By
L. Ranvier, Professor at the College de France. With
Plate XXXVI.

It is well known that by the employment of chloride of
gold it is possible to show the ultimate nerve-endings in
tissues. But this reaction is not always followed with suc-
cess if the original process of Cohnheim is pursued ; and in
attempting, during the last few years, to render it more
regular by combining the action of the gold with formic acid,
Pritchard and Loewit have effected a very notable progress.
At the same time the process of Loewit, although it gives far
more constant results than any previous process, is not
without its disadvantages. The solution of a third part of
formic acid, in which the tissue is placed before being sub-
mitted to the action of the gold, alters the ultimate ramifica-
tions of the nerves, so that, on being fixed by the action of
the gold, they are more or less altered.

It occurred to me that, in the simultaneous action of
formic acid and chloride of gold, the latter reagent would
retain its selective power, and by the rapidity of its action
would prevent the nervous terminations from being seriously
injured by the acid. It appeared to me also important that
the reduction of the gold should be as rapid as possible.
These considerations led me to the following process.

The tissues, with the nerve terminations, are placed in a
mixture of chloride of gold and formic acid, which has been
boiled and then cooled. After remaining in this mixture
between two and four hours they are removed and washed,
and the reduction of the gold is effected either by the action
of daylight in slightly acidified water, or in the dark in a
solution of formic acid.

By this method of treatment the terminations of the
nerves in muscles appear continuously arborescent, instead of
being frequently interrupted, as they are when the process
of Loewit is employed. At the same time they contain some
irregularities which prove that the process is still insufficient.
For this reason it became necessary to invent a fresh process.
I attempted to replace formic acid by an acid which would
not have an equally deleterious effect on delicate elements,
and believe that I have found it in the juice of the lemon.

Lemon juice, though its protracted action alters nervous
tissues, yet preserves their form sufficiently long for it not
to be altered in the time requisite to procure the whole effect

TERMINATION OF NERVES IN THE EPIDERMIS. 457

of the chloride of gold. Preparations of the white or red
muscles of the rabMt, treated successively with lemon juice
and chloride of gold, preserve the nerve terminations, not
only continuously arborescent, but also remarkably regular.

I have employed this process in the investigation of the
nerve terminations in the epidermis in general, and have em-
ployed for this purpose organs in which numerous authors
have already studied the terminations of nerves ; for example,
the snout of the pig, the nose of the mole, and the skin of
the human finger.

In the pig’s snout (fig. 1) the nerves which penetrate and
ramify in the epidermis have a variable diameter ; the rami-
fications they form have an equally variable distribution,
though on this point it is not necessary to insist. It is,
however, to be noticed, as is shown in fig. 1, that in propor-
tion as the nerves approach their termination they become
moniliform, and that some of them break up into granules,
which become completely free. Some of these granules are
to be seen completely isolated in the corneous layer of the
epidermis.

As regards the termination of the nerves in the nose of
the mole [Talpa europea)^ I will assume a knowledge of the
investigations of Eimer and Moisisovicz. I will add that at
the base of the epidermic thickening, amongst the nervous
tubes destined for them, there are small Paccinian corpuscles,
and that in the deeper layer of the epithelial mass forming
these papillae there are five or six small rounded nervous
corpuscles, as to the signification of which I do not propose
to give any further details. As regard the true inter-epithelial
nervous branches (fig. 2), it is important to notice that those
which, to the number of two or three, are placed in the centre
of the organs of Eimer, form zigzags, which become the more
pronounced as they approach the surface of the epidermic
covering. At the angles of the zigzags there are at first
thickenings of the nervous fibre, and finally veritable
buttons. These buttons become more and more conspicu-
ous, become stalked, and in the neighbourhood of the cor-
neous layer become completely free. The marginal nerve-
fibres {vide fig. 2) of the organ of Eimer remain straight
through their whole course, but as they approach the sur-
face swellings appear on them, which are placed in the same
transverse line for all the fibres. These swellings enlarge,
soon form a prominence turned towards the centre of the
organ, then become stalked, and finally become completely
free near the corneous layer.

The same process may be observed with the peripheral

458

L. RANVIER.

fibres of the organ of Eimer (fig. 2). These fibres divide so
as to form rudimentary arborescences, which end in buds and
button-like thickenings, which can equally be separated,
travel towards the surface, and be thrown off to the exterior
with the other products of the evolution of the epidermis.

In preparations of the human epidermis (fig. 3) made with
chloride of gold, the terminal branches of the nerves have
an arrangement analogous to that of the peripheral nervous
ramifications of the organ of Eimer. Only the nerves which
penetrate into epidermis have a very inconstant diameter,
and the ramifications they form are subject to great variations,
upon which I need not insist further. I must content
myself with saying that in proportion as they approach the
corneous layer the ultimate nervous ramifications become
more and more moniliform, and finally break up into
granules, which are no longer connected with the nervous
system, and which, mingled with the epithelial cells,
appear, like these latter, to be subjected to a process of
elimination.

The theory, or rather the hypothesis, which I propose is
founded on the facts which I have just briefly expounded.

The nerves which enter the epidermis, whatever may be
the form or extent of their ramifications, are subject to a
continuous evolution. They grow while at the same time
their terminations undergo a gradual degeneration ; this
degeneration leads to the formation of granules of nervous
substance, which become completely free, and are soon
transported into the inert layers of the epidermis.

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 459

On the Termination of the Nerves in the Mammalian
Cornea. By E. Klein, M.D., F.R.S., Lecturer on
Histology at St. Bartholomew’s Medical School. (With
Plate XXXVII.)

There is hardly an organ in which the distrihution of
the fine nerves can be so easily observed as in the cornea,
thanks to the invaluable discovery of Cohnheim of staining
the organ with chloride of gold. Since his publication,
November, 1866, On the Termination of the Sensory Nerves
in the Cornea,” in ^ Virchow’s Archiv,’ vol. xxviii, a very great
number of observations on the same subject have been pub-
lished, all of which have been obtained by Cohnheim’s
method of chloride of gold.

On a perusal of all these publications, one arrives at
the remarkable conclusion that, notwithstanding the excel-
lence of the method, notwithstanding the distinctness with
which the final nerves are traceable in all parts of the
cornea, notwithstanding the great transparency and rela-
tively simple structure of this organ, notwithstanding the
absence in it of a variety of structures, such as glands,
blood-vessels, &c., capable of materially interfering with the
observation of the fine nerves — notwithstanding all this, there
exists the greatest variety of opinion as to the termination of
the fine nerves.

The arrangement of the microscopically coarse nerve
branches in the cornea of the frog and mammal were
understood before the use of the gold method, thanks
to the researches of Saemisch, Arnold, and Hoyer; and after
the gold method the observations of Cohnheim, Kolliker,
Hoyer, His, Waldeyer, and others have, one might almost
say, been exhaustive on the same, viz. the arrangement of
the coarser branches ; it is the finer and finest fibres whose
termination, nay, even whose distribution and arrangement
is still a matter of discussion.

The last publication on this subject is by Izquierdo,
(‘ Beitrage zur Kenntniss der Endigung der sensiblen
Nerven,’ Inaugural Dissertation, Strassburg, 1879) and by
Waldeyer (Llrchiv f. mikrosk. Anatom.,’ xvii, p. 367),
chiefiy relying on Izquierdo’s researches carried out under his
(Waldeyer’s) direction. They are to the effect that the fine
nerves entering the anterior epithelium of the cornea termi-
nate in tlie suporfn-ial lay«us, in the manner described by

460

DR. E. KLEIN.

Kolliker and Hoyer^ viz. with free ends of the same nature as
maintained by Hoyer^, while most of those in the substantia
propria terminate in the processes of the corneal corpuscles.
I shall return to these observations below in detail, but
wish to point out here already that they are diametrically
opposed to those that I described in this Journal in 1871
(October, 1871), and in the ^ Monthly Microscopical Journal ’
(April, 1872), to the eifect, first, that the fine nerves having
entered the anterior epithelium and having branched and
run horizontally (Kolliker) for a longer or shorter distance,
terminate in a network, which I called the intraepithelial
network ; and secondly, that those in the substantia propria
do not anastomose with the processes of the corneal cor-
puscles, as was mentioned by Kiihne, Cohnheim, Mose-
ley, Lipmann, and others, as will be referred to below
minutely, but although in their extremely long course many
times they come in close contact with the corneal corpuscles,
do not terminate in them but as a network on them. For
the reason of this discrepancy between my results and those
of Izquierdo and Waldeyer, I have again made the nerves
of the cornea a subject of investigation, and I am able to
prove that neither do they terminate with free ends in the
anterior epithelium, nor are they connected with the corneal
corpuscles. ^

On making this renewed investigation, I have observed
several other points which appear to me to be of importance
in the discussion of the mode of termination of the fine
nerve-fibres.

In order to get at the consideration of the intraepithelial
fine nerves and those of the substantia propria, we shall
start with the axis-cylinders, which in the front layers of the
substantia propria are most numerous, forming what are
called by Arnold the subepithelial, by Hoyer the subbasal,
and by Waldeyer the fine stroma plexuses. Vv^e shall refer
to them simply as the stroma plexus. The nature of the
branches of this plexus, as bundles of primitive fibrils, the
endothelial perineural sheatli of the larger branches, the
great difference in thickness of the different branches and in
the different planes of the same branch, the angular plate-like
enlargements, where two or more of them join, are so con-
spicuous that there are hardly any differences of opinion
about them, and it is therefore unnecessary to refer to
them more than in j)assing. There is no difference amongst
the recent observers as regards the position of these plexuses.
They arc all agreed about their being hehmd Bowman’s
membrane.

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 461

First, as to the relation of this stroma plexus to the sub-
epithelial nerve expansion.

Cohnheim (1. c.) traced from the most anterior branches
of this plexus isolated twigs, which immediately underneath
the epithelium split up into bundles of elementary nerve-
librils. Hoyer Archiv f. Anat. and Physiolog.,’ 1866, Heft
2) already saw these twigs ; and Kolliker Ueber d. Nerven
Endigung in der Hornhaut Physic. Med. Gesellsch., in Wiirz-
burg,’ 30th Juni, 1866), says of them that they pass in
an oblique direction through the membrana Descemeti, in
order to arrive at the posterior surface of the epithelium,
and he therefore called them rami perforantes. Hoyer

Archiv f. mikrosk. Anat.,^ ix, ii Heft, p. 236) was the
first who described in the most anterior layers of the
substantia propria, but underneath the anterior basement
membrane, a special nerve termination, the subbasilar
plexus, which consists of fine and finest fibrils derived from
the branches of the stroma plexus. Most of the finest
fibrils after a long, straight or wavy course and after
branching terminate with free ends in the tissue. All fibres
of the subbasilar plexus remain underneath Bowman’s mem-
brane, and this plexus is densest in the peripheral zone of
the cornea. While confirming Hoyer’s observation as to
the great abundance of the fine and finest fibrils in this sub-
basilar plexus in the cornea of the guinea-pig and rabbit,
I am able to add to his description on some important
points.

I will here state once for all what in the following pages
1 shall speak of thick and thin fibrils ; as I have attempted
in my former paper (1871 and 1872) so also here I distinguish
the nerve branches of the stroma plexus, each of which is in
fact only a bundle of fine fibrils, as nerve bundles of the first
order ; these give off fibres, which are to be considered as a
small bunch of two or three primitive fibrils, are the nerves
of the second order or the thick fibres^ and the individual
fibrils constituting them are the fine fibrils or the fibrils of
the third order. The latter are characterised by the smaller »
and larger varicosities more or less closely placed.

The fibres of this plexus are of various thicknesses be-
tween that of a smaller branch of the stroma plexus and the
finest fibrils marked by regular varicosities ; the thicker
fibrils have only a short course, since they soon branch into
the finer ones. But these latter are of various types, a.
Such as take at once a bold straight course, as if they at-
tempted to run to the furthest possible regions ; in this way
they bend off from their course once or twice at a right angle,

VOL. XX. NEW SER. II II

462

DR, E. KLEIN.

but in most cases remain in the same level, h. Fibrils which
take a long and wavy course^ altering their direction and
level many times^ running repeatedly upwards or dowmwards
in a vertical direction. In sections that are placed horizon-
tally, and comprise the anterior layers of the cornea, these
fibrils can be traced in their entire course, but it requires
great attention to follow them through all their changes of
level.

Now, the nerve expansion that Hoyer called the sub-
basilar plexus can be separated into the following three
systems : — 1. A plexus of thicker and finer fibrils, which lie
in the same level as the most anterior branches of the
stroma plexus, that is, immediately behind, and closely to
Bowman’s membrane. This system we call the suhhasilar
plexus proper. The thicker fibres of this system branch,
and they and their branches run a more or less wavy
course. The finest fibrils of this system pass out of this
level behind or in front. 2. From the branches of the
stroma plexus we trace fine fibrils, which at once pass into
the layers behind, and here run for a long distance in
straight lines ; they branch repeatedly and anastomose by
these branches in rare places wdth similar fibrils ; they ter-
minate apparently in the corneal substance with free ends.
These fibrils constitute the deep subbasilar fibrils. We
shall return to them below. As mentioned just before,
some of the fine fibrils of the first system leave the level of
this latter, and run behind it as straight fibrils identical
with those of the second system. 3. From the level of the
first system fine fibrils pass anteriorly into the basement
membrane ; here they may be followed in their winding
course for a long distance, changing their direction repeat-
edly ; they are not very numerous, and within this mem-
brane may be seen to branch here and there. These fibrils
may be called the intrabasilar fibrils. They and their
branches have a tw^o fold destination : a, either they again
pass backwards and associate themselves to the deep sub-
basilar fibrils, or, as is not unfrequently the case, they pass
anteriorly through the basement membrane and enter the
subepithelial network to be described presently. In the
Plate XXXVII, accompanying this paper, I have given
an accurate representation of these different fibres in figs.
1—3. •

The nerve-fibrils in the immediate proximity of the epithe-
lium form what is known as the subepithelial netw'ork ; it is
made up of fine and finest varicose fibrils, and they are derived
from two sources : — (<7) from the ratni perforaiites ; this source

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 463

was already known to Cohnlieiin, Kollikev, and Hoyer ; as
Cohnheim very beautifully showed, they come olf from the
rami perforantes in bundles,” except in a small central
zone. But 1 would add here that there exist ‘ bundles ’ also
in the central zone, but they are not so conspicuous, because
smaller. All fibrils are possessed of varicosities, and those
of the same bundle diverge the further away from their
ramus perforans ; they form triangular groups, the apex of
which is directed towards the periphery.

(d) A few fibrils are derived from the intrabasilar fibres,
as mentioned above.

As regards the course of the fibrils of the subepithelial
network, they all have a linear, but more or less winding
course. Their number is everywhere very great, but there
exist remarkable differences in this respect in different
specimens. Both rabbits’ and guinea-pigs’ corneae, prepared
with the chloride of gold, with or without subsequent re-
duction by formic acid, tartaric acid, warmth, &c., show
some striking variations. The best results, e. the greatest
number of these subepithelial fibrils, I have noticed in
cornea) prepared, after chloride of gold, with tartaric acid
(see my paper in this Journal, October, 1871, p. 408), or
with glycerine (in the ^ Monthly Micr. Journal,^ April,
1872, p. 157).

In successful specimens so obtained the number of
fibrils is here astounding ; it is so great and dense that we
can truly speak of a special nervous layer formed here by
fine fibrils. Such specimens are not common ; they are, in
fact, rare ; perhaps only one out of ten will show them to
perfection ; but, since we cannot assume that this one would
in reality differ from the others, we must necesssarily con-
clude that this is due to a lucky unknown condition in the
method of preparation, and I would again insist on this, as
1 have done in my former papers, viz. that the description
of the distribution of the fine nerve-fibrils is probably the
correct one, which refers to the most perfect preparations.

In the s])8cimens that we are now having before us, and
of which figs. 7 and 8 give a faithful representation of the
nerve-fibrils of the subepithelial network, it will be seen
that the individual fibrils are of very great length, but there
are differences in this respect. In fig 8 I have drawn, very
accurately, a great number of the fibrils (by no means all) of
the subepithelial network of a portion of the cornea, amount-
ing ill length to about 0*4 mm., and it will be found that
some fibrils exceed this length, while others fall far short
of it.

464

DR. E. KLEIN.

Two facts are very conspicuous in this figure, viz. first, that
most of the fibrils branch and really anastomose with another,
and secondly, that, without exception, they leave the subepi-
thelial network sooner or later, according to their smaller or
greater length. With regard to the first of these facts, it will
he noticed, from figs. 7 and 8, that not only do the fibrils of
the same “ bundle” anastomose with one another, but also
those of neighbouring ones; that consequently the subepi-
thelial fibrils are connected into real networks, as Cohnheirn
distinctly stated in his paper, and as I have also described
and figured it. Hoyer (1. c., p. 2S2) agrees with Engelmann
that the fibrils cross only, but do not anastomose, and
Izquierdo (1. c., p. 26) mentions anastomoses between the
fibrils of the same bundle, but is not certain whether
this is the case also between the fibrils of neighbouring
bundles.

The second fact above mentioned, viz. that all of the
subepithelial network pass out of the level of this latter
is easily ascertained in the preparation represented in fig. 8.
It will be found here that many^of the Jihrils possess short
lateral hranchlets, which, soon after they are given off, curve
hook-like, and enter the anterior epithelium. Some of the
fibrils of the subepithelial network are possessed of a great
number of these minute side brancblets (in the figure these
are marked by a small cross), and when viewed from the
surface (see fig. 8) they resemble a sort of fishing line, to
which are fastened right and left a series of hooks ; in our
case the hooks are the little brancblets ascending into the
epithelium. But also the extremity of the chief fibril itself
ascends into the epithelium. In accordance with this a ver-
tical section through the cornea show these fibrils of the sub-
epithelial network, creeping along the lower surface of the
epithelium, and giving off short brancblets, vertically as-
cending into the epithelium, between the cells of its deepest
layer. I do not find in the description given by Hoyer,
Cohnheirn, Kolliker, Engelmann, Izquierdo, and others,
that these minute side brancblets, as marked by a small
cross in fig. 8, have been known to them in the form as
they are represented in this figure. Kolliker and Hoyer
have pointed out that the intraepithelial network is situated
immediately underneath the epithelium, between it and the
basement membrane ; Krause and Izquierdo, however, think
that its fibrils are sunk in between the extremities of the
deepest epithelial cells. I cannot admit this latter view,
since I possess preparations in which the whole epithelium,
inclusive of the deepest layer, is clearly removed vvithout

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 465

disturbing the subepitheiial network, which is left untouched.
Cohnheim (1. c., p. 41) mentions this already of the cornea
which has been macerated in acetic acid after chloride of
gold, and I can fully confirm him herein. This could not
possibly be the case if the subepitheiial network were not
situated underneath the deepest layer of the epithelial cells.
From vertical sections of the rabbit’s cornea we obtain the
conviction that the greatest number of these fibrils are
situated above (^.e. anterior, to) Bowman’s membrane.
Cohnheim does not admit for the rabbit’s cornea a structure
comparable to the similar membrane in the human cornea
there can be no doubt about its presence, and in gold
specimens it is occasionally well marked as a conspicuous
layer of the thickness of a lamella of the ground substance,
from which it differs by its being differently coloured, but it
is of the same tint as the membrana Descemeti.

Having passed into the epithelium, the nerve-fibrils ascend
in a vertical manner, winding their way between the columnar
cells of the deepest layer,, as was first seen by Hoyer, and
afterwards, by means of his chloride of gold method, minutely
traced by Cohnheim. As regards the fate of these intra-
epithelial fibrils there is even less agreement than of that of
the subepitheiial ones. With the exception of Inzani, Than-
hoffer, and Ditlevsen, all observers agree that the intra-
epithelial fibrils extend to near the surface of the epithelium.
According to Cohnheim (‘Virchow’s Archiv,’ Band 88),
having arrived in the superficial layers, some of them branch,
others not, and then they change their vertical course into a
horizontal one. Some of them terminate with a minute end
knob freely floating in the precorneal fluid. Kdlliker
followed the last termination of the intraepithelial fibrils as
horizontal fibrils situated between the most superficial layer
of the flattened epithelial cell. Here they run for a longer
or shorter distance, branch repeatedly, and anastomose but
rarely, and finally each fibril terminates with an end-knob
underneath the most superficial layer of the epithelial scales,
Engelmann (‘Die Hornhaut,’ Leipzig, 1867), whose obser
vations refer only to the cornea of the frog, saw only free
ends of the intraepithelial fibrils between the epithelial cells
of the most superficial layers. At a similar conclusion
arrived also Tolotschinow (Inaugural dissertation (St. Peters-
burgh, 1867). According to Petermoller the intraepithelial
fibres form networks in all layers of the epithelium. In my
paper on this subject in this Journal I have described them
as giving of lateral branchlets, and as forming a deep and

466

DR. E. KLEIN.

a superficial intraepithelial network^ the former situated in
the layer next to the deepest columnar cells, the latter below
the most superficial layer. The fibrils of the latter branch
repeatedly, and run a horizontal course sometimes for con-
siderable distances ; they join, cross each other, and wind
their way between the epithelial cells, changing their level
several times. 1 have also stated that the end-knobs that one
meets with on some of these fibrils cannot really possess this
character, since they are absent in others ; and, besides, they
are identical with those larger varicosities that occur in the
xjourse of most of the fibrils.

Hoyer Archiv f. mikr. Anat.,^ Band ix, p. 234) describes
the intraepithelial fibrils in the same way as I had done,
but, while admitting anastomoses of these fibrils, main-
tains that they terminate with free ends amongst the
superficial layers of the epithelial cells. The small and large
varicosities which these fibrils show in some specimens are
absent in others, and, therefore, must be regarded as pro-
duced artificially. Rollett (^Strieker’s Handbook,’ article

Cornea,” p. 1136) does not see any anastomosis amongst the
intraepithelial fibrils, but says, that having branched they
terminate in free ends, often with a slight enlargement
amongst the superficial epithelial layers. Waldeyer (Graefe
and Saemisch, ^ Augenheilkunde,’ p. 210), on the other hand,
maintains a terminal network of the intraepithelial nerves.
Krause Allg. und Mikr. Anatomie,’ 1876, p. 539) saw the
intraepithelial nerve fibrils forming plexuses near the surface;
they terminate with small end-knobs. Izquierdo (1. c., p.
27) finds that the intraepithelial nerves, having ascended in
a vertical or oblique direction into the superficial layer, bend
off into a horizontal course and branch. They do not
anastomose. All fibrils terminate with free ends, or with a
slight swelling.

Waldeyer (^Archiv f. mikrosk. Anatom,’ Band xvii, p. 379),
while accepting Izquierdo’s conclusions, admits the incorrect-
ness of his former view of an intraepithelial terminal net-
work. When speaking of a terminal network (Graefo and
Saemisch, p. 270), Waldeyer then observed also free ends, but

the possibility always remains that in such apparently free
ends there is only an imperfect action of the chloride of
gold.” This is just what I have urged in my first paper in
this Journal, 1871, knowing well the great differences in
the number of the intraepithelial nerve-fibrils that one meets
in various specimens, and in various places of the same
specimen, prepared after the same plan, and seeing that in

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 467

many cases the nerve-fibril can be traced beyond an apparent
end-knob.

In his last paper (’ Archiv f. mikr. Anat./ xvii), however,
Waldeyer thinks (p. 329) that the reason for his formerly
maintaining an intraepithelial terminal network lies perhaps
in the fact that he had chiefiy to deal with the human cornea,
in which the cement substance becomes readily stained with
the chloride of gold, and hence the appearance of anastomosis
is easily accounted for. In the preparations that were the
basis of my memoir in 1871, and in those of the present
occasion, the intraepithelial nerves are the only fibres stained
black with the chloride of gold. The epithelial cells and
their interstitial substance is unstained, except a faint
greyish tint in the deepest layer. The nuclei of the
superficial epithelial cells are just visible, being faintly
stained violet. I must repeat that in my descriptions of
the intraepithelial nerves I always had to deal xcitli such
specimens only. Those in which also the interstitial sub-
stance of the cells in the superficial layers are distinctly
stained have always been discarded as unsuitable, since a
confusion of the interstitial cement substance with fine nerve-
fibrils is then quite possible. This has been ably urged
by Cohnhein (1. c., p. 31), and in the drawings of the intra-
epithelial nerve-fibrils accompanying his memoir (figs. 7 and
8, plate xii) ; in the figures given by myself in this Journal
October, 1871, figs. 2, 3, and 4, Plate XIX \ and in ‘ Monthly
Micro. Journal,’ April, 1872, figs. 5 and 6, plate xiv ; in
those represented by Hoyer (^Archiv f. mikr. Anat.,’ ix,
fig. 2, plate xiii,) the unmistakable and conspicuous nature
and course of these fibrils can at once be recognised. To all
those who have seen good specimens of the intraepithelial
nerves it must be evident that, judging from the repre-
sentation given in Rollett’s fig. 390, 1. c., p. 1136, and in
Waldeyer’s fig. 21, 1. c., p. 210, both these authors had not
seen the intraepithelial horizontal nerve-fibrils of the
superficial layers, and consequently were not in a position to
judge of their final distribution.

Following, then, carefully in a horizontal section — the
anterior free surface being directed upwards and a place being
selected that comprises only the epithelium — of a well-stained
and well-reduced gold cornea, the intraepithelial nerve-fibrils,
after they have passed between the columnar cells, can be
traced us more or less horizontal fibrils in all layers anterior to
the deepest columnar cells. The length of these fibrils varies
very greatly ; some of them are very short, not longer than the
breadth of two or three epithelial cells, while many others may

468

DR. E. KLEIN.

be traced for a very considerable distance,even through several
fields of the microscope. On theirway they give off right and left
similar horizontal fibrils; everyone of them, if long enough, can
be seen to change repeatedly its level, now ascending to almost
the free surface, then, again, dipping down and pursuing its
course among the middle layers of the epithelium. In this
manner they form a very intricate texture, such as 1 have
figured ill my -former papers, and as is also figured by
Hoyer in his fig. 2, plate xiii. We find these fibrils in all
layers, between the most superficial layer of scales and the
deepest layer of columnar cells, and I must, therefore, correct
accordingly my former statement as to the presence of these
fibrils only in two layers (a deep and superficial intra-
epithelial network), although I must adhere to my former
statement that they are most numerous in the superficial
layers. The fibrils are of various thicknesses ; some are con-
spicucusly thicker than those of the subepithelial network.
This is especially the case with those in the superficial layers,
as I must persist in maintaining against Hoyer (1. c., p. 234) ;
but in all layers in which they occur we find fibrils of com-
paratively great thickness crossing and anastomosing or
giving off respectively fibrils of such fineness that it is just
possible to trace them. Whether very fine or not, they almost
invariably possess minute varicosities, some fewer, others
more ; especially the very fine ones show them better and
more regularly disposed than the thick. Besides these fine
varicosities there are other larger ones more irregularly
distributed. They are either staff-shaped, or spherical,
or pear-shaped, or angular. Any of these large vari-
cosities may be met with in the course of the fibrils
where they join others or where they give off one or two
lateral branches, or they occur at the apparent extremity of a
fibril. I say purposely apparently,” since we shall presently
see that they are not the real ends. Such apparent ends
occur on the fibrils or their branches in all layers, as Cohn-
heim very correctly described (1. c., p. 28), except, of course,
the deepest layer of columnar cells, and the most superficial
one of fiattened scales.

The nerve-fibrils do not pass into this last-named layer, and certainly do
not extend beyond it. 1 have expressed myself in the same sense already in
my former communications, in accordance with Kolliker. Hoyer and
Izquierdo did not see them pass into the most superficial layer of scales.
When looking at a horizontal section comprising the complete epithelium,
it will at first appear as if some of the most superficial intraepithelial
fibrils did really extend as far as the free surface. To determine this it is
necessary to use high powers, such as Zeiss’s yV oil immersion, or Hartnack’s
10 immersion ; it will then be seen that one layer of nuclei {i.e. those of

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 469

the most superficial scales) cau be traced above the most superficial nerve -
fibrils. Oil account of the great thinness of the superficial scales even a
moderately high power, such as Zeiss’s D and E (-|- and ^ inch), or Hart-
nack’s 7 and 8, does not suffice. Amongst the great many specimens which
I have examined — they amount to hundreds — only in very rare instances
have I seen one or the other fibril at the same level as the most super-
ficial nuclei.

Before coming to their ends I must say a few words as
regards their relation to one another. Cohnheim (1. c.j pp. 27
and 28) does not speak of any anastomosis of the intraepi-
thelial nerve-fibres, while Kolliker (1. c., p. 4) observed such
anastomoses, although, as it appears, not frequently.”
Anastomoses of the intraepithelial nerve-fibrils, although, on
the whole, rarer than I thought them to be in my first
memoir, do no doubt occur. My mistake of assuming them
to be of very frequent occurrence arose from my not having
examined with sufficiently high powers.

It must be borne in mind that these fibrils in many places,
when crossing, are in very close contact, and if they happen
to cross at the point of a varicosity the appearance of an anas-
tomosis is produced. When examining such places with an
oil immersion -pV of Zeiss I can ascertain that in four out of
six there is really no anastomosis between the fibrils ; in the
other two it is impossible to say that there is not an anasto-
mosis. But there exist real anastomoses, about which there
can be no doubt whatever ; these refer to such places where
two of the long fibrils are joined with one another by a
shorter or longer side branch. In fig. 10 are represented
such undoubted anastomoses.

Now, as to the larger or smaller knob-like structures which
are met with on the apparent extremities of the shorter as
xvell as the longer fibrils, as mentioned above, and as observed
by all who have worked at this subject, there is no doubt
about their presence, and the question is whether they are
the real ends or not. Kolliker (1. c., p. 4) would not say
that these knobs are natural or artificial, Hoyer (1. c.,
234) considers them as artificial products, and Izquierdo
(1. c., p. 28) asserts that some fibrils, at any rate, possess
them. Cohnheim, Tolotschinow, Krause, and others, con-
sider them as natural end-knobs.

It cannot be denied that they possess the same appear-
ance, whether at the end of a fibril or in its course ; but the
great differences in their number, size, distribution, and
shape, seems to point to their being artificial products. That
some of them are not “end-knobs” this I am going to prove
presently ; that others are such is doubtful, since it cannot
with any show of reason he asserted, as Izquierdo does, and

470

DR. E. KLEIN.

as Waldeyer confirms, that there exist two modes of termina-
tion— one with end-knobs, and the other without.

What is, then, their real termination ?

In preparations which would pass the muster of good
nerve preparations, there is nothing to be seen, even when
examined with a high power, beyond the fibrils as figured
by Cohnheim, myself, Hoyer, and Izquierdo, and as now re-
presented in the figures of the Plate accompanying this present
memoir. But after they have been kept mounted in glycerine
for some time — I refer here to specimens of the rabbit’s cornea
prepared after the simple (Cohnheim’s) method of chloride
of gold, as well as that where afterwards a' reduction is
effected by means of tartaric acid (see my former paper in
this Journal) and mounted in glycerine for one or two years
— and if they are then carefully examined, it will be found
that in some of them a great many exceedingly fine fibrils
are seen, which are given off by the above ordinary intra-
epithelial fibrils, both as lateral branchlets and from the
knobs that previously appeared as end-knobs. These
fibrils, which at the time the preparation was mounted
could not be seen, are now distinctly traceable, some
for a long, others only for a short distance. In some
places they appear as a row of minute stained dots with
unstained joints between, while in^ others both the dots
and joints are stained and then easily perceived. Fol-
lowing under a high power, e. g. Zeiss’ oil immersion
which by its exquisitely fine definition is admirably suited
for this purpose, one of the fibrils in the superficial layers —
of course, only thin horizontal sections of the epithelium
being referred to here — we are at once struck by the very
great number of short lateral rod-like projections which are
seen on most fibrils ^ both the ordinary ones as well as the very
fine ones just mentioned. Generally, but not by any means
always, they are given off at the varicosities. Especially
interesting are, in this respect, the knobs which were above
referred to as end-knobs ; some of these appear beset with
these projections on their whole circumference, and resemble,
then, somewhat the thick extremity of a stamen with its
radicles. Most of them are very short, rod-like, others can
be traced for long distances, and in this case they are in-
variably dotted, and present themselves as regularly varicose
fibrils, giving off likewise numerous lateral branchlets. Also
in this case, either only the varicosities appear stained and the
joints unstained, or the joints are faintly stained. Now, both
the short rod-like as well as the longer filamentous branchlets,
ramifying dendritically and uniting by their branchlets.

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 471

break up into a network, which is of exceeding minute-
ness. The network appears composed of small rods of equal
length, and at all the nodal points there is a minute dot or
granule. These rods appear generally faintly stained as
compared with the dots, but in some places they are stained
as much as the latter. Only in few places have I seen this
network very perfect, and then only over a small space, but
then it appeared with sufficiently great distinctness as a
reticulated plate-like expansion of some of the fine nerve-
fibrils. This minute network, which I shall call the ter-
minal nettoorh, is shown in figs. 11, 14, 15, 16, and I must
refer the reader to these figures which give a more compre-
hensive idea of its nature than any lengthy description ;
it is situated hetioeen the epithelial cells, since we see
it from its broad surface and in profile (see figs. 1 1 and 16).
As mentioned on a previous page, fibrils with end-knobs are
met w'ith in almost all layers, and so, I presume, is also
the terminal network, but for obvious reasons it is only
distinctly traceable in the superficial layer.

In the above figures I have, for the sake of comparison,
added a few nuclei of the flattened epithelial cells ; these
nuclei are just visible as faintly outlined oval or spherical
discs, of a pale grey or bluish colour ; of the outlines of the
cells themselves nothing can be seen.

Thus, it is easily recognised that the nerve-fibrils we have to
deal with here as connected with the end reticulum are con-
siderably finer than those drawn by Hoyer (fig. 2, pi. xiii), or
Izquierdo (b c., fig. 5, pi. i). The fineness of the terminal net-
-vvork is by no means inferior to that of the terminal network
which I described of the nerves in the skin of the ttadpole,
and situated immediately underneath the epithelium Sit-
zungsber. d. k. Akadem. der Wiss.,^ Vienna, 1870, vol. 61),
and also observed by Lavdowsky Arch. f. norm, und path.
Histolog.,^ 1870, and ^ Centralblatt f. med. Wiss.,’ 1872,
No. 17), nor that discovered by Gerlach in the grey matter
of the spinal cord (^ Centralblatt f. d. medizin. Wiss.,’ 1867,
Nos. 24 and 25), and by Rindfleisch (^Archiv f. mikr. Anat.,’
viii, p. 453) and Gerlach (‘ Centralbl. f. d. medizin.
Wiss.,’ 1872, No. 18), in the grey matter of the cortex
of the cerebrum. If Waldeyer in his last article (Vlrchiv.
f. mikr. Anatomie,’ xvii Band, iii Heft) sums up by say-
ing (p. 379) that there exist no terminal networks, and that
the fine nerve-fibres ending in the epithelium, either run out
freely or end with a minute knob, it is necessary to remind
the reader that Waldeyer bases himself on Izquierdo’s speci-
mens, which, as far as refers to the intraepithelial nerve •

472

DR. E, KLEIN.

fibrils, I consider imperfect, and Waldeyer^s assertion must
therefore be taken to have the value of negative evidence
only. That the reduction of the gold in Izquierdo’s speci-
mens has not been complete is further proved by his saying
(1. c., p. 29) that, in vain have w^e searched in the epithe-
lium for the cells of Langerhans, 'which Ribbert pretends
to have recently found.^^ These cells — elongated cells with
very many finer and thicker branched processes — it is true,
are not met with in ordinary specimens, but in those in
which the reduction of the gold has been complete they are
very conspicuous by their colouration and size. Their number
generally varies in the middle and superficial layers of the
epithelium in different places and their processes, although
crossed by the intraepithelial nerve-fibrils, appear nowhere
connected with them.

The assertions of Inzani (quoted by Izquierdo, 1. c.,
p. 16) about special terminal ends situated amongst epithe-
lial cells of the deepest layer and possessed of minute fibrils,
as well as those of v.Thanhoffer (^Virchow’s Archiv,^Band 63,
1875) about the fine nerve-fibres terminating in so-called
tactile corpuscles, situated between the cells of the deepest
layer of the epithelium, are obviously due to imperfet speci-
mens. This is well illustrated by t^he fact that Thanhoffer
did not see any nerve-fibrils beyond the deepest layer of
cells. To a similar cause, viz. to imperfect specimens, must
be also attributed the assertions of Beale about the absence
of nerves in the epithelium. Beale did not see any nerve-
fibres in the anterior epithelium of the cornea, but then
this observer did not make use of the chloride -of-gold
method.

The last point which I wish to discuss here is the termina-
tion of the nerve-fibrils in the substantia propria. As is
well known, Kiihne was the first to maintain that in the
frog^s cornea the finest fibrils terminate in connection with
the processes of the corneal corpuscles. The number of ob-
servers who made this a special point of investigation,
although great, range themselves in those who confirm
Kuhne, Moseley (this Journal, July, 1871), Konigstein (^Sitz-
ungsber. d. K. Akad. in Wien.,’ 1877, Band 76), Izquierdo
(1. c., p. 25). and Waldeyer (1. c., p. 378) ; both these last-
named observers assert, however, that some fine fibrils termi-
nate also free in the substantia propria. Then those who
simply deny such a connection, and accept a free ending of
the fine nerve-fibrils in the substantia propria, Kolliker (1. c.),
Kngelmann (1. c.), Dwigbt Monthly Micr. Journal/ 1869),

TERMINATION OF NERVES OF MAMMALIAN CORNEA. 473

Tolotschinow (1. c.),E,ollett (1. c., p. 1138) — who is, however,
not able to say in what way they terminate — Hoyer, and
Krause (1. c.). And, lastly, those who maintain a special
termination of the fine nerve-fihrils, Lipmann Virchow’s
Archiv,’ Band 78, p. 218), who says that the fine fibrils ter-
minate in the nucleoli of the nuclei of the corneal corpuscles.
Lavdowsky Archiv f. Mikr. Anat.,’ viii) assumes two kinds
of terminations of the nerve-fibrils, (a) in rhombic plates,
supposed to be present in the wall (?) of the lymph-canali-
cular system (dog), and (b) with peculiar enlargements near
the nucleus or in the nucleolus (frog, dog, cat, calf). In my
former papers I have stated that the finest nerve-fibrils in
the substantia propria, in good gold specimens of the frog’s
cornea, are conspicuous by their great length, their peculiar
rectilinear course, their right-angled bending, their few
branchings ; these fibrils I described as the nerves of the
third order, and as being nowhere in connection with the
corneal corpuscles, although they come in close contact with
them in many jilaces ; they are not the finest nerve-fibrils,
as assumed, since in some places they are seen to give off very
minute short fibrils — nerve-fibrils of the fourth order — which,
on the surface of the corneal corpuscles, anastomose in a sort
of network.

I have now extended my observations on the fine nerve -
fibrils in the substantia propria of the cornea of the kitten and
rabbit, and I must also for these maintain that the fine
nerve-fibrils that one sees in ordinary gold specimens are
not the finest and last fibrils. Especially easily followed are
the fine nerve-fibrils in the anterior strata of the substantia
propria, where they form what we considered above as the
deep subbasilar fibrils, and which have been very beautifully
described already by Hoyer. But also in the deeper sections
of the substantia propria similar fine nerve-fibrils can be met
with. Their general character is that they may be traced
for very long distances ; that they possess now a rectilinear,
then, again, a wavy course ; that they keep close to the
corneal corpuscles, always running in the lymph-canalicular'
system, as was already known to v. Recklinghausen, and as
was very extensively described by Lavdowski and Hoyer ; that
they give off from place to place a lateral branchlel ; that in
some parts they are seen to bend off suddenly at a right
angle, either after a long rectilinear or curved course, or at
short intervals ; further it will be noticed that tliey continu-
ally change their level, now passing into a stratum above or
below, then again returning into their original plane; and.

474

UR. E. KLEIN.

filially, that they anastomose, in a few instances, with their
neighbours.

These fibrils differ in thickness, and they all contain small
or large varicosities, more or less regularly disposed. I must
draw the attention of the reader to figs. 4 and 4a, and 5,
in which the nature and course of these fibres is drawn very
accurately. In what way do they terminate ? As is well
shown in the figures, after having given off one or more
branches, they, as well as their branches, appear to terminate
freely either at a varicosity or beyond it, at the side of a
corneal corpuscle. In this respect I can add little to the
description given by Hoyer of these fibrils with admirable
faithfulness.

In the cornea of the frog these fibrils are much more diffi-
cult to trace, since their great number is compressed into the
posterior strata of the substantia propria, that is, into a very
limited space. For this reason we find that anastomoses
between neighbouring fibrils are more common than in the
cornea of the rabbit and kitten.

I possess specimens both of the cornea of the rabbi t and kitten,
and also of that of the frog, wherein I find places in Avhich
these fibrils, when examined with a high power (Zeiss’ F, or,
still better, oil immersion off ’^'ory minute short

fibrils, which close to the surface of the body of the corneal
corpuscles give off short, exceedingly fine, dotted fibrils, which
themselves are connected into a network. It is, of course,
easily understood, that in many instances it is impossible to
distinguish between dots that are contained in the substance
of the corneal corpuscles and the dots that mark these ter-
minal fibrils ; but in certain other instances this distinction
is possible, viz. in those instances in which the corneal cor-
puscles and their processes are stained only a greyish tint, lohile
the nerve-fihrils and their varicosities, owing to the complete
reduction of the gold salt, possess an almost black colour.

In some specimens, however, we find also nerve-fibrils
in connection with the processes of a corneal corpuscle, such as
is described by Moseley, Kouigstein, and Izquierdo, and there
seems no mistake about the nerve-fibrils being here directly
continuous "with a corneal corpuscle. But let us for a moment
inquire. What are these specimens? As is known to everybody
who has examined a number of cornem, either prepared
after the simple (Uohnheini) method of chloride of gold, or,
in addition to this, after using various reducing agencies,
such as oxalic acid, formic acid, methylated spirit, or simple
heat, hardly a single cornea is obtained in which the sub-
stantia propria appears throughout of the same tint. In

TERMINATION OF NERVES OF MAxMMALIAN CORNEA.

475

some cornese in a greater or smaller portion, the corneal cor-
puscles and their processes are of a deep reddish-purple or
reddish-black colour, while in others the corneal corpuscles
are hardly at all stained, or only of a light grey or bluish-
grey colour. The same difference may occur in one and the
same cornea, and, of course, we then find zones of transition
of the former into the latter. Let us take a cornea in
which the corneal corpuscles and their processes are con-
spicuously brought out, and well and deeply stained, and let
us add that in the majority of these instances also the fine
nerves are well stained, i.e. of the same deep colour as the
corpuscles aud their processes. And let us bear in mind
that the nerve-fibres are situated in the lymph-canalicular
system which also contains the corneal corpuscles and their
processes. Let us further bear in mind that these nerve-
fibrils and their branchlets run out close to the corpuscles
and their processes. Under these circumstances nobody will,
I think, be in a difficulty to explain the above-mentioned
connection of the nerve-fibrils with the corneal corpuscles.
Supposing in the preparations represented in fig. 4, 4a, 5,
and 6, the corneal corpuscles were as deeply stained and of
the same colour as the nerve-fibrils, nobody could fail to find
here connections between the former and the latter ; and,
indeed, in the very same specimens from which these draw-
ings were made there are places in which, owing to the deep
colouration of the corneal corpuscles and their processes, the
distinction between them and the fine nerve-fibres is lost,
and, therefore, an apparent connection between them is
found to exist. It is quite clear that such places are useless
for determining the relation between the nerve-fibrils and
the cornejil corpuscles, and it is to me inexplicable how
Izquierdo omitted to bear this in mind, since Hoyer (1. c.,
p. /^41) had already drawn attention to the great importance
of having specimens in which the corneal corpuscles and
nerve-fibrils are of a different colouration. Izquierdo^s
drawings (plate i, figs. 1, 2, 3, and 4), leave no doubt that
lie had to deal with specimens in which the corneal cor-
])uscles and their processes and the nerve- fibrils were very
deeply and uniformly stained. For this reason, then, his
assertion about the direct connection of the processes of the
corneal corpuscles with the fine nerve-fibrils, and conse-
quently also Waldeyor’s summing up (1. c., p. 378) based
u])on it, loses the value attributed to it.

476

DR. E. KLEIN.

Histological Notes. By E. Klein, M.D., F.R.S.

I. — Ciliated Epithelium of the (Esophagus,

E. Neumann (‘ Archiv f. mikr. Anat./ vol. xii, p. 570)
observed on the surface of the mucosa of the oesophagus of
the human foetus, between the eighteenth and thirty-second
week, ciliated columnar epithelium. This did not, how-
ever, form a continuous lining, since in many places the epi-
thelium was stratified pavement epithelium, such as is found
in the new born and adult. Kolliker Entwickelungsge-
schichte, d. Menschen,' &c., 1879, vol. ii, p. 853), found also
ciliated columnar epithelium in various places in the oeso-
phagus of the human foetus between fourteen weeks and the
sixth month.

Mr. Anderson and Mr. Ricketts, students at St. Bar-
tholomew’s Medical School, when examining sections of
the cervical portions of the (hardened) oesophagus of a new-
born child — full time — found, amongst the stratified pave-
ment epithelium of the ordinary description, places where
the superficial cells were not flattened but columnar;
underneath these were polyhedral cells in two or
three layers, and finally, a deepest layer of columnar,
cells. The superficial cells were either without or with
cilia ; the latter were short and very delicate. The ciliated
cells were seen, more or less continuous, chiefly in the
grooves between the folds of the mucosa ; in the neighbour-
hood of these there were generally only isolated ciliated
cells.

II. — Cilia in the Central Ca^ial of the Emhryo Chick,

Examining with Mr. Anderson the fresh embryo of a
chick of about the stage figured in Kolliker’s ^Entwickel-
ungsgeschichte,’ p. 138, as of the end of the second day, with
seventeen protovertebra3, in warm saline solution and from
the surface, I saw in the whole extent of the protovertebra3
the central canal of the central nervous system lined with
very beautiful delicate cilia ; their length was about half that
of the cells lining the canal, and they were in very active
movement. It was, in fact, this movement by which their
presence made itself conspicuous. It could not be seen in
the caudal part. In a second embryo of about the same
early stage, observed under similar conditions, the move-
ment w'as seen distinctly in the commencement of the cer-
vical region, but was soon lost, owing to the wall of the

HISTOLOGICAL NOTES.

477

central nervous system approaching too close from side to
side, and consequently the central canal becoming alto-
gether lost. A.nd this state, viz. the shutting-up of the
central canal, forms one of the most troublesome impedi-
ments to see the cilia and their movement. In embryos of
a later stage the examination proves unsuccessful, partly
owing to this collapse of the central canal, but chiefly to
the relative opacity of the tissue, an opacity far too great to
allow the very delicate cilia to be perceived. In embryos of
an earlier period, such as possess only about 6 — 10 proto-
vertebrse, I could not detect either the ciliary movement or
the cilia themselves, although in several instances the
central canal had not collapsed.

In the central canal of the earlier embryos, as well as in
those of the later ones, there were always to be met with a
few spherical or oval, coarsely granular, and therefore
opaque, corpuscles, either isolated or two or three together ;
they varied in size, and a nucleus could not be seen in
them.

In transverse sections through the above hardened embryos
nothing of the cilia could be seen, but this will nOt surprise
since the cilia of less delicate structures are lost during
hardening ; one of the best known examples is the mucous
membrane of the human uterus, and of that of mammals,

III. — The Glands of the Nasal Cavity of the Guinea-pig.

As is now well known, the glands occurring in the mucous
membrane of the nasal cavity are mucous and serous glands
(Heidenhain). The latter form, in the thick portions of the
non-olfactory Schneiderian membrane, huge masses, in which
the alveoli — convoluted tubes, with lateral and terminal
shorter and longer branches— are grouped into lobules,
separated, as in other cases, by vascular connective-tissue
septa.

The alveoli of the mucous glands present themselves
in two different forms : either small, with a small lumen and
lined with polyhedral or short columnar cells, each with a
spherical nucleus; or large, with a large lumen, -and lined
with columnar cells of the nature and aspect of mucous
cells. T here are alveoli in which the two kinds of cells occur
side by side. Thus, .they resemble the mucous glands of the
larynx known through Heidenhain, Tarchetti, myself, and
others. The ducts are lined with a single layer of . thin
columnar cells, as in other mucous glands. At the com-
mencement of the nasal mucous membrane, i, e. near ,the

VOL. XX. NEW SER. I i

478

DR. E; KLEIN.

external opening, where the epithelium of the surface is
stratified pavement epithelium, with a superficial stratum
like the stratum corneum of the epidermis, also the mouth
of the duct of the mucous gland receives a continuation of
this latter stratum. Like the epidermis it stains dark
in osmic acid.

The intralobular ducts of the serous glands are iden-
tical with the salivary tubes of Pfliiger, viz. they are
lined with a layer of columnar cells, the outer portion of
which is distinctly longitudinally striated. The intralobular
ducts ultimately branch into several shorter or longer tubes,
which are very narrow and lined with a layer of flattened
cells, and they are therefore identical with the intermediate
portions of the ducts in the salivary glands and the
pancreas,

The epithelial cells lining the alveoli are columnar, and
present themselves in two different aspects according to the
state of secretion. In one aspect, viz. during secretion, they
are more transparent, thicker, and appear therefore less
columnar. Their nucleus is spherical or only slightly flat-
tened, and pressed against the membrana propria. In the
other state they are less transparent, thinner, more columnar,
and their nucleus is spherical, and not quite so close to the
membrana propria; but in both conditions the cell sub-
stance is a reticulum, more distinctly so, on account of the
greater size of its meshes, in the former than in the latter.

There exists, consequently, also in this respect a complete
analogy between these and other serous glands.

In those parts where the mucosa is of great thickness,
covered with ciliated columnar epithelium, and containing
s mailer or larger amount of diffuse adenoid tissue, these glands
form also larger groups, situated in the deep submucous
section.

In preparations that had been prepared, when per-
fectly fresh, with a mixture of chromic acid and osmic acid,
according to the method of Flesch (‘ Arch. f. mikr. Anat.,j
Bd. xvi, p. 300), the alveoli of the serous glands, especially
thcs3 nearest to the surface, appear much larger, their cells
being "distended by the presence of smaller or larger . fat-
globules, isolated or in groups ; consequently there exists, in
this respect, a certain resemblance of these glands and the
sebaceous and Meibomian glands. From this it appears, theii,
probable that tlie secretion of them is not precisely the same
as in the other serous glands, ^.<7. those at the root of the
tongue or the salivary glands.

In the interlobular connective tissue isolated lymph-cor-

HISTOLOGICAL NOTES.

479

puscles may be met with, similar to what has been noticed
by Boll and Lavdowsky in the salivary glands.

Fine threads, some with nuclei, some without, can be
traced in connection apparently with the cells of the intra-
lobular ducts and with those of the alveoli. But whether
they are nerve-fibres, as seems probable, or not, could not be
definitely ascertained.

An interesting fact, observed here in connection with the
large serous gland, is the presence of thin, unstriped muscle
cells running in all directions, and forming a plexus, in
whose meshes the gland-alveoli are situated. In some
places, e. g, the conchse, there exists here also a plexus of
large veins, and the above muscle-bundles form the stroma,
as it were, for them, and thus a certain resemblance with a
cavernous tissue is produced. This muscular tissue is, no
doubt, of some service for the discharge of the secretion,
seeing their peculiar intimate relation to the gland -alveoli.
In the relatively rigid walls of the nasal cavity such inter-
stitial muscle-tissue must be of special value to the glands.

EEVIEWS.

Atlas of Histology. By E. Kleij!^, M.D., F.E.S., and E.

Noble Smith, L.R.C.P., M.R.C.S. Smith, Elder &

Co., 15, Waterloo Place.

We have great pleasure in drawing attention to this very
handsome work, which is certainly the most richly illustrated
treatise on Histology which has ever been published. It
occurred to Dr. Klein that by means of the colour-printing,
which has now been brought to so high a condition of
perfection, it would be possible to produce illustrations which
should be practically fac-similes of the delicately stained
sections which histological students make use of for demon-
stration and investigation. Accordingly Mr. Noble Smith
has executed drawings from the choicest preparations made
by Dr. Klein, which in colour and form present to the eye
precisely the same effect as the preparations themselves when
accurately focussed.

The text accompanying the drawings is by no means a
mere explanation of the latter, but forms an original treatise
on Histology, in which Dr. Klein’s views are clearly stated
in reference to many debatable questions. The medical
student will also find in this work a surer guide to the
knowledge of modern Histology than in the older books,
even those of high authority and reputation, which, on
account of their antiquity, are too often marred by erroneous
theory and insufficiency of observation.

Freshwater Rhizopods of North America. By Joseph
Leidy, M.D., Professor of Anatomy in the University of
Pennsylvania, and of Natural History in Swarthmore
College, Pennsylvania.

It is, perhaps, somewhat astonishing that a treatise on
living microscopic organisms should be published, as this
work is, under the auspices of the Geological and

REVIEWS.

481

Geographical Survey of the United States Territories."”
At the same time^ the authorities can only be congratulated
on the wise liberality which has led them to undertake the
production of so valuable a work. We have here figured
and described by Dr. Leidy, in the most ample manner, a
large series of those freshwater forms which Mr. Archer, of
Dublin, was the first to detect, and to study in a special
manner. Focke, Franz Schulze, Lesser, Hertwig, and Dr.
Leidy, have added to and expanded the rich store, which
was first tapped and offered to the zoological connoisseur in
the pages of this Journal. To the old and well-known
genera of freshwater E-hizopods — Amoeba, Gromia, Arcella,
Difflugia, Actinophrys — a host of remarkable allies have
now been added. In the present fine monograph Dr. Leidy
describes many already known genera and species which he
has detected on the North American continent, but he also
describes several new and very interesting forms, e,g. Din-
amoeba, Ouramoeba, and Biomyxa. The volume is illustrated
by forty-eight quarto plates, and must be carefully studied
by every student of Rhizopod faunae. We should have been
glad had it come into the plan of Dr. Leidy^s studies to
devote more attention to the determination of minute
structure as revealed by powers higher than one of 250
diameters.

A History of the British Marine Polyzoa. By Thomas
Hincks, B.A., F.R.S. Van Voorst, London.

The well-known and highly valued series of works on
Natural History, published by Mr. Van Voorst, is worthily
extended by the two admirable volumes, one of text, one of
plates, treating of the British Polyzoa, by the Rev. T.
Hincks. In an introduction of a hundred and forty pages,
illustrated with numerous cuts, the comparative anatomy
and general morphology of the Polyzoa are very fully and
clearly discussed — the most recent investigations — French,
German, and Swedish, all receiving due notice. In the
systematic part which follows the groups of Polyzoa are
taken up one by one and systematically characterised, until
the whole hierarchy, down to families, genera, species, and
varieties, is exhausted. The characteristics of species, as
well as of genera, and even of suborders, are necessarily
chiefly, if not entirely, drawn from the structure of the hard
parts. These are very extensively and beautifully figured'
in the atlas, so that by the aid of this work the student of

482

RETIEWS.

marine zoology will have no difficulty in referring such
Polyzoa as he may dredge on the British coast to their
proper titles. Naturalists are already deeply indebted to
Mr. Hincks for his beautiful volumes (in the same series)
on the British Hydroid Polyps. In the present work he
has dealt in a masterly way with a still more difficult group.
We hope that other similar works may ere long be added to
Mr. Van Voorst’s celebrated series.

NOTES AND MEMORANDA.

Medusse andHydroid Polyps living in Fresh Water — From
experiments made by Mr. Romanes on Limnocodium, and
published in ^ Nature ’ of June ^4th it appears that the fresh-
water Medusa is very intolerant of sea water. It cannot be
re-introduced with impunity, to the medium in which fts
ancestors originally lived, and this would seem to imply that
a very great length of time has elapsed since the adaptation
to fresh water took place. Curiously enough, Mr. Romanes
has found that marine Medusae are not so injuriously affected
by brine as the Limnocodium is by sea water. The fact,
however, is less astonishing when we remember that the
percentage of saline matter in solution in sea water is many
hundred times what it is in average pond water, whilst
the strongest brine has not a percentage of saline matter
many times in excess of that of sea water.

On the whole the tolerance by marine animals of fresh
-water is a much more frequently observed fact in all classes
than the tolerance of sea water by lacustrine or fluviatile
forms. It is undeniable that existing fresh water forms have
been developed by adaptation from marine forms, whilst it is
difficult to cite any instance in which adaptation in the
opposite direction appears to have taken place ; some few
marine Oligochsetous Chaetopods and Pulmonate Gastero-
pods being perhaps such instances.

The tolerance by Medusae belonging to marine species of
fresh -water under natural conditions was observed by Mr.'
Moseley, at Browera Creek, in New South Wales, and from
Professor Alexander Agassiz I have received some interest-
ing notes recording similar observations.

Mr. Moseley says (‘Naturalist on the Challenger,^ p- ST2),
“ After heavy rain the surface water in all the upper parts
of the creek is so diluted by the torrent of fresh water from
the stream, that it becomes almost fresh ; indeed, at the time
of our visit, it was for three or four miles down so little
brackish as to be drinkable. .At a short depth no doubt the

484

NOTES AND MEMORANDA.

water was salt. Here are the most favorable conditions
possible for turning marine animals into fresh water animals;
in fact, the change of mode of life presents no difficulty.
Below no doubt the water is always salt, but the fish find a
fluid less and less salt as they rise to the surface. We
caught the mullets in the almost fresh water with a net, and
with them the mussels and crabs. I even saw an abundance
of Medusae, and a species of Rhizophora sw'imming in the
creek above the sand-flats, where there was scarcely any
salt at all in the water, yet evidently in most perfect
health.”

Professor Agassiz writes, It strikes me as if the conse-
quences resulting from the finding of this freshwater Medusa
had been somewhat overdrawn. In the first place, w^e have
two genuine freshwater Hydroids, Hydra and Cordylophora,
and in the second place, as far as my experience goes, it is
not conclusive of so fatal an action of fresh water on Medusae
as Romanes would lead us to believe in. We have quite an
estuary leading out back of Boston Harbour, extending on
tlie one side to form what we call the back bay, and beyond
this up the Charles River as far as W atertown, where there is
a dam, about seven miles from the inner extremity of the
harbour proper. Here the Charles River falls into this
estuary as a freshwater stream sufficiently large at times to
affect the saltness of the estuary below it at low tide, so
that at Cambridge, half way from XVatertown to Boston, the
w^ater is salt only at the highest tides, quite brackish during
the first half of the ebb, and comparatively fresh during the
last part of the tide. At W. Boston Bridge, about one mile
from the head of the harbour, the water at the last part of
the tide is fresh enough and tastes but little salt. At this
bridge there is an abundance of Hydroids which thrive
remarkably well on the drainage of the district, and grow to an
unusually large size. The species found there which has no
free Medusa is Laomedea gigantea. While of the Hydroids
which have free Medusae we find Eucope diaphana, Eucope
pyriformis , Ohelia commissuralis . All these species are,

therefore, twice during twenty-four hours exposed to salt
water and to nearly fresh water, and thrive remarkably well
under the treatment, as must of course their free Medusae,
which I have caught both at high tide and low water — in
salt and in nearly fresh water.

“ Other of our Medusae also find their way into this
estuary, and I have found in fresh water at loiv tide active
Sarsiae, Tiaropsis, and also Aureliae, which seemed unaf-
fected by the large quantity of fresh water in which they

NOTES AND MEMORANDA.

485

were found. Cyanea I have never seen in this estuary
except at high tide. I may mention that the scyphostoma
and strobila of Aurelia is found with the above-mentioned
Hydroids attached to the piles of W. Boston Bridge ; but
the scyphostoma of Cyanea I have never foiind.’^ — E. Bay
Lank ESTER.

On the Respiration of the Crustacea. — In the note pub-
lished in the April number of this Journal, a prediction was
hazarded that the Zoea larva of the higher Crustacea would
on examination prove to breathe in the same way as the
Copepoda. Zoeas of Cancer, and probably of some species
of Prawn, received from Mr. T. Bolton, have confirmed this
amply. The respiratory diastole and systole of the rectum
with rhythmical openings of the anus are thoroughly well
marked. It may here be noted that in carmine stainings of
the entire Copepoda the stain does not diffuse through the
integument, but up through the rectum in the first instance.
The power of dialysis through the chitinised integument is
slight if at all existent. Now that another place is found
for the respiratory function, it may be denied to the
expanded pleura of the carapace.

This constancy of function in the anus is remarkable, and
indicates that the gills which characterise so many of the
higher Crustacea are secondary formations, long posterior to
the differentiation of the class. What, then, is their origin ?
Probably they are in all cases modifications of those pro-
cesses of the appendages which primitively bring about
nutritive currents. In Daphnia we have the stage where
these processes are the exclusive food bringers, while they
share respiratory functions with the rhythmically contrac-
tile rectum. And as a parallel to the direction of the bran-
chial current from behind forwards in the Crustacea, we
may cite the Lameilibranchs, where the gills, probably in
origin also parts of the locomotive apparatus (according*
to Lankester’s view) play an equal part in nutrition and
haematosis. To explain body gills we have to invoke either
the principle of the similar modification of neighbouring
parts, or else that of displacement upwards from the appen-
dages. As far as I know, for there are none of the memoirs
on this subject within reach, the development of the body-
gills of the crayfish, which might give a clue, has not been
worked out. — Marcus Hartog.

INDEX TO JOURNAL

VOL. XX, NEW SERIES.

Acanthaster, spine of, 112
Amphioxus, laminar tissue of, by
Pouchet, 421

Amphioxus, spinal nerves of, by
Balfour, 90

Angiosperms, pollen bodies of, by
Elfving, 19
Anthelia Turneri, 111
Araneina, development of, by Balfour,
167

Azolla, root-hairs of, 115

Bacillus of leprosy, Hansen on, 92
Bacteria, biology of, by Waldstein,
190

Bacteria, Fitz on, 106
Bacteria, Koch on, 107

„ of mouth, Butlin on, 109

„ Paul Bert on, 109

Bacterium anthracis, 374
Balfour, development of Araneina, by,
167

„ on larval forms, 381

„ on spinal nerves of Amphioxus,

90

„ structure and homologies of
germinal layers, by, 247
Bennett and Murray on reformed ter-
minology of cryptogams, 413

Bennett on classification of crypto-
gams, 408
Berkleyia, 112

Binocular with high powers, by Gibbs,
318

Blodgettia, Wright on. 111
Blomfield, J. E,, on development of
spermatozoa, 79
Blood-corpuscles, origin of, 241
Blood-corpuscles, origin of the red,
by Pouchet, 331

Bourne, A. G., on nephridia of the
medicinal leech, 283
Bower, on conceptacle in Fucaceae, 36

Carpenter, Herbert, on points in
Echinoderm morphology, 322
Cell-layers of Coelentera, 104
Chroococaceous alga from Leicester,
378

Claus, publications edited by, 110
Coffee-leaf disease of Ceylon, Dyer
on, 119

Conceptacle in the Fucaceae, develop-
ment of, by F. 0. Bower, 36
Copepoda, anal respiration of, 244
Cornea, nerves of, by Klein, 459
Corpuscles, red, blood-, origin of,
241

488

INDEX,

Coscinodiscus Sol, 111
Craspedacustes Sowerbii, Lankester
OD, 351

Cryptogams, classification of, by Ben-
nett, 408

„ reformed terminolody of,
by Bennett and Murray, 413
Cuuningliam on effects of starvation
on animal and vegetable tissues, 50

Development, Schafer on, 202
Disease, coffee-leaf. Dyer on, 119
Dublin Microscopical Club, 111, 378
. Dyer on coffee-leaf disease, 119

Ecbinoderm morphology, disputed
points in, by Herbert Carpenter,
322

Echinothrix, spine of, 115
Elfving on pollen bodies of Angio-
sperms, 19

Embryo sac of Gymnadenia, Ward
on, 1

Endoderm, muscular fibre derived
from, 106

Epiblast, muscular tissue derived
from, 377

Eye of Pecten, by Hickson, 443

Eucacese, development of conceptacle
in, by E. O. Bovver, 36

Germinal layers, structure and homo-
logies of, by Balfour, 247
Giard on Orthonectida, 225
Gibbs on binocular with high powers
318

Gibbs on human spermatozoon, 320
Glomerulus of the head kidney of
the chick, by Sedgwick, 372
Gouiocidaris, spines of, 379
Gymnadenia, embryo sac of, 1

Hansen on bacillus of Leprosy, 92
Hartog on Hydra, 243
„ Zoea, 485
„ Copepoda, 244

Hertwig on cell-layers of Ccelentera,
104

Hickson on eye of Pecten, 443
Histological notes by Klein,

Hubrecht on nervous system of Ne-
mertines, 274, 431
Hydra, histology of, by Parker, 219
Hydra, mode of swallowing prey, 243

Kidney, relation to Wolffian body in
the chick, by Sedgwick, 146
Klein, histological notes,

„ nerves of the cornea, 459
Koch, method of preparing sections
of coral, by, 245

Lankester, E. Ray, on capillaries in
the integument of leech, 303
„ ,) on connective and

vasifactive tissues of leech, 307
„ „ on Limnocodium,

(Craspedacutes) Sowerbii, a new
Traehomedusa inhabiting fresh
water, 351

Larval forms, by Balfour, 381
Leedi, medicinal, nephridia of, by A.
G. Bourne, 283

„ capillaries in the integument
of, by Lankester, 303
,, connective and vasifactive tis-
sues of, by Lankester, 307
Leprosy, bacillus of. Hausen on, 92
Lieberkiihnia, Siddall on, 130
Limnocodium Sowerbii, Lankester on,
351

Lumbricus, spermatozoa of, by J. E.
Bio m field, 79

Micrasterias furcata, 115
Muscular fibre derived from endo-
derm, 106

Muscular tissue developed from epi-
blast, 377

Nemertines, nervous system of, by
I Hubrecht, 272, 431

INDEX.

489

Nephridia of the medicinal leech, by
A. G. Bourne, 283

Nerves in the epidermis, by Ranvier,
455

Nerves of mammalian cornea, by
Klein, 459

Nervous system of Nemertines, Hu-
brecht on, 274, 431

Orthonectida,by Giard, 225

Parker on histology of Hydra, 219

Pecten, eye of, by Hickson, 443

Planorbis, development of, 103

Pollen bodies of Angiosperms, Elf-
ving on, 19

Pouchet on laminar tissue of Amphi-
phioxus, 421

Pouchet on the origin of the red
blood-corpuscles, 331

Pulmonata, development of, by Rabl,
376

Rabl on development of Pulmonata,
376

Ranvier on nerves of epidermis, 455

Reviews,

Rhizopod, new marine, by Siddall,
130

Schafer on development, 202

Sedgwick on relation of kidney to
Wolffian body in the chick, 146
„ on the glomerulus of the
head kidney of the chick, 372
Shepheardella, Siddall on, 130
Siddall, J. D., on Shepheardella, an
undescribed type of marine Rhizo-
poda, 130

Spermatozoa, development of, by
J. E. Blom field, 79
Spermatozoon, structure of the human,
by Gibbs, 320

Starvation, effects of, on animal and
vegetable tissues, by D. D. Cun-
ningham, 50

Trichia clavata, 114
Trichia fallax, 115
Toxopneustes, spine of, 113

Vacuolaria virescens, 117
Van Beneden, new Journal by, 110
Vienna, publications of Zoological
Institute at, 110.

Volvox, spermatic state of, 112

Waldstein, biology of Bacteria, by,
190

Ward, Marshall, on embryo sac of
Gymnadenia, 1

Xanthidium Robinsonianum, 114, 116

rillNTI-D BY J. K. ADLAUD, BARTHOLOMEW CLOSE.

XX.N.sM. I

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES I, II, & III,

Illustrating Mr. H. Marshall M^ard^s Paper on the Em-
bryo Sac ind Development of Gymnadenia conopsea.^^

N. B. — Except vhen otherwise stated, the figures are generally from sections
in absolute alcohol and glycerine, examined under Zeiss’s F, oc. 2 and 3.

PLATE I.

Fig. 1. — Optiial longitudinal section of young ovule, showing embryo-
sac mother-cell, ind axial row of similar cells. This specimen and the next
were examined ii the fresh state with traces of ammonic hydrate.

Fig. 2. — Simiar section of slightly older ovule, of which the integument
is more advance!, and the curvature more decided; the granular contents
of embryo-sac nother-cell resist alkalis more than do the others.

Fig. 3 — Actial section through older ovule seen from before. The
second integument is well established, and following on the first. The
embryo-sac moher-cell has had one “cap-cell” cut off, and a second
division is takiig place to form the other.

Figs. 4, 5, aid 6. — Longitudinal section through the embryo-sac. It is
surmounted b^ two “ cap cells,” wliich, together with the surrounding
layer of nuclei s-cells, are gradually compressed by the enlarging sac.
(Fig. 6 is somevhat oblique, but still shows cells of the axial row.)

Fig. 7. — Sirilar section. Two large nuclei in the embryo.sac have
resulted from dvision. Note traces of cap-cells, and of cells of the nucleus
of the ovule anund.

Figs. 8, 9, .nd 10. — Similar preparations ; showing divisions of the two
nuclei of Fig. 7- Treated (alcohol specimens) with strong acetic acid and
fuchsine.

Fig. ll.-Similar section, showing division of upper nucleus. Two
nucleoli in ech half. The section is slightly oblique.

Fig. 12.— Similar section. The two nuclei at each end, and fixed in the act
of dividing i planes crossing at right angles ; a large vacuole occupies the
centre of th sac. Remains of the cap-cells as a refractive mass under the
micropyle.

Fig. l3,“Longitudinal .section through the ovule. There are four naked
masse.s ofirotoplasm at each end of the sac. Note traces of cap- and
nuclous-ces around. (Alcohol specimen treated with acetic acid and
fuchsine.)

Fig. li— The masses of protoplasm have arranged themselves in the
upper pari of the sac as an “ egg-apparatus,” consisting of two bolster-
shaped “fnergidm” supporting an “oosphere” (germinal or embryonic
vesicle), ad a free nucleus, near which is a second.

PLATE II.

Figs. 15 and 16. — Longitudinal sections through embryo-sac, showing the
“egg-apparatus” above, and a free nucleus below, which has travelled
down towards the “antipodal” mass.

j’jg, 17. — Longitudinal section through a whole ovule during fertilisation.
A pollen tube has penetrated the micropyle, and spread its end on the
“ synergidse,” below^ and to the side of which is the “ egg cell ;” at the
opposite end of embryo-sac are the “ antipodal ” cells imperfectly divided,
aud abutting on these the wandering nucleus from above.

Fig. 17a. — Similar preparation with egg-cell just fertilised.

Fig. 18. — The first division across the fertilised “ egg cell” is completed ;
remains of egg apparatus and cap-cells are seen above, and of “ antipodals ”
below in the sac.

Fig. 19. — A division appears across each of the two cells in Fig. 18 ;
that in the upper cell is not completed. Traces of cap-cdls and egg appa-
ratus above, and of antipodal cells below, as before.

Fig. 20. — The embryo now consists of four cells — a “ suspensor ” of
two cells separated by swollen walls, and an embryo pro>er of two super-
posed cells, of which each nucleus contains two nucleoli.

Figs. 21, 22, and 23. — Embryos a little more advancd. The embryo
proper becoming divided by vertical walls at right auges, comes to con-
sist of three cells ; the lowermost in fig. 23 is dividing.

Fig. 24. — The embryo proper consists of four cells, aid the left hand
lower nucleus has two nucleoli ; the suspensor consisting >f two cells, is in
the act of forming a bipartition across its upper cells.

Fig. 25. — The embryo proper has four cells, the suspusor five. Va-
cuoles are appearing in the cells of the latter.

Fig. 26. — The embryo proper consists of pight cells, anl the suspensor
of four. Traces of “ antipodal ” mass still evident.

Fig. 26a. — Transverse section of embryo at this sta:ge.

Fig. 27. — The suspensor possesses three cells, of which the middle one
has just divided, but the nuclei are not yet completely sepjrated.

Fig. 28. — A slightly older stage than the last. Each->f the four sus-
pensor cells has become doubled; vacuoles are appearing, a.d the cell- walls
are swollen.

PLATE III.

Fig. 29. — The suspensor cells are elongating, sap cavties becoming
established within. The embryo proper still preserves tvelve cells, but
would soon consist of sixteen, as division appeared in its lowr cells.

Fig. 30. — By intercalary growth the suspensor has pmied its apex
through the micropyle ; one more division has occurred abve, but the
limit is approaching, and the exhausted cells are acquiring largisap cavities.
A series of tangential divisions has marked out an epiderma layer to the
embryo proper (optical sec. cleared with KHO).

Fig. 31. — Similar section of older embryo similarly treatd ; a second
series of tangential divisions has appeared more internally, an the “ sus-
pensor ” is shrivelling up, the protoplasm of its cells disappeamg.

Fig. 32. — External view of ripe seed torn from funiculus. Less magni-
fied than rest.

/II

II ivi Ward, del

K Hufh.Lilh' Kdin’’

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE IV,

Illustrating the notice of Elfving^s researches on Pollen-
bodies of the Angiosperms/’

Fig. Anthericum ramosum.

1. — Young Pdlen-body after division, x 300.

2—4. — Devebpment of the vegetative cell. X 300.

5. — Metamorphosis of the nucleus of the large cell. X 300.

6. — Ripe Polbn-body. x 300.

7. — Branchedtube. x 230.

Anthericum liliago.

8. — Ripe Polbn-body, one nucleus star-shaped, x 300. •

Tulipa Gesneriana. X 320.

9 — 11. — Devdopment of the Pollen-body from formation of vegetative
cell to being fuly ripe.

12. — Vegetatve cells pressed out.

13. — Duplicaure of vegetative cell.

14. — Point ol the Pollen-tube.

Ornithogalum pyramidale. X 450.

15. — Vegetatve cell with thick side walls.

16. 17. — Rip' Pollen-bodies.

18. — Pressedout vegetative cells ; at o treated with osmic acid.

Leucojum oestivum, x 400.

19. — Young ?ollen-body after division.

20. — The sane after treatment with osmic acid.

21. — The veptative cell loosened ; at o with osmic acid.

22. 23. — Rip Pollen-bodies ; o, a vegetative cell with osmic acid.

24. — Pressedout nucleus and vegetative cell ; osmic acid.

25. — Pollen-ube. Nuclei after one another.

Narcissus poeticus.

26. — Euterig of the nuclei into the tube.

27. — The poterior nucleus divided.

Iris siberica. X 350.

28. 29.— Cflshed-out Pollen-bodies; in 28 the walls of the vegetative
cell are seen, n 29 the whole cell.

Iris xiphium. X 300.

30. — Tubeend. The nucleus of the larger cell is characteristically
lengthened ot.

Tradescantia virginica. X 300.

31. — Your Pollen -body after division.

32 — 37. -development of the bodies up to ripeness.

Convallaria mulliflora.

38. — For ation of two vegetative cells. X 350.

39. —Rip Pollen-body, x TOO.

Asparagus officinalis.

40. — Aringement of the vegetative cell, x 600.

PLATE -Continued.

Fig. Sparganium ramosum. X 450.

41, 42. — Arrangement and first formation of the vegetative cell.

43, 44. — Cruslied-out Pollen-bodies showing the metamorphoses of the
vegetative nuclei.

45 — 48. — Pollen-tubes. 45. The vegetative nucleus in the act of

dividing. 46. The vegetative nucleus has gone first, and has in 47 divided.
48. The vegetative nucleus has remained undisturbed.

Asphodelus albus. X 350.

49 — 53.— Development of the Pollen-bodies from the first appearance of
the vegetative cell until maturity.

Andropogon campanus. X 400.

54. — Pollen-body after division. 55. Same seen at an angle of 90A

56. — The vegetative cell divided. 57. Same seen fron the side.

58. — Pollen-body with three vegetative cells. 59. Tie same seen at an
angle of 90°.

60. — Both vegetative cells become free.

Bromus erectus. X 350.

61, 62. — Pollen-bodies after division.

63 — 65. — Formation of the vegetative cell.

66. — Division of the vegetative cell.

67 — 69. — Metamorphosis of the vegetative cell and tie larger nuclei up
to the maturity of the Pollen -body.

Arum ternatum. X 600.

70, 71. — First development of the Pollen-bodies after livision.

72. — Pressed-out body ; the division of the vegetativeiuclei is only seen.

73. — Mature body; the spectacle-like united vegetatve cells are still
maintained.

74. — Nuclei pressed out of mature bodies.

Butomus umhellatm. X 300.

75. — First division of the Pollen-bodies.

76. — The vegetative nucleus has in addition divided.

77. 79. — Pollen-tubes; in 78 and 79 the partition wlls are still main-
tained.

Juncus ariiculatus. X 300.

80. — Pollen-body after division. 81. Older body.

Heliocharis palustris. X 350.

82 — 94. — Development of the Pollen-body to maturity. 90 and 91 show
the gradual resorption of the three nuclei situated in theiarrow end.

95. — Normal thickening of the intine of a mature body.

96 — 98. — Peculiar thickenings of the membrane, which ccasion the for-
mation of false partition walls.

99 — 102. — Pollen-tubes.

'fUcTt/owrri^lj^.XX, N. s'^^. V

Fig. n.

Fig.Z

F^.5

Fi^ 4.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE V,

Illustrating Mr. F. O. Bower^s Memoir on the Develop-
ment of the Conceptacle in the Fucacese.

The following system of lettering has been used throughout — Initial
cell = i. Basal oell = b. Lim iting tissue = 1. Cortical tissue = c.
Central column =cc. Mucilage ^ m. Antheridial cell = a. Pedicel
cell —p.

The first six figires are all from material treated first with dilute chromic
acid. The remaiung figures are from material preserved first in saturated
solution of commoi salt (cf. text).

Fucus serratus.

Pig. 1 (^^).— Part of a vertical longitudinal section passing through
the lip which surnunds the depressed apex of a fertile branch. The initial
cell has ceased t( divide, and is surpassed by the surrounding tissue, a
slight inclination i observable in the walls dividing the cells of the limiting
tissue.

Fig. 2 Part of a similar section, with an older conceptacle. The

initial cell has bejun to shrink. The basal cell has not yet divided. The
inclination ot thewalls of the limiting tissue is more pronounced than in
fig.].

Fig. 3 (^^).— A young conceptacle as seen from above, the protoplasm
of the initial celhas shrunk, and its cell-walls swollen.

Fig. 4 Taken from a section similar to 1 and 2. This shows

the initial cell nore shrunk, the basal cell divided loirgitudinally. The
alteration of the>w'ollen cell-wall filling the cavity has begun.

Fig. 5 (-^-^^^J-Part of a transverse section, with a conceptacle more ad-
vanced ; it is (iifcult here to tell the limit between the part of the lining
tissue, derived fpm the limiting, and that derived from the cortical tissue.

Fig. 6 ^pg^t of a transverse section, with an older conceptacle,
the central colum still attached. No formation of hairs as yet.

Fig. 7 (i|-).-Male conceptacle. Mucilage showing stratification and
striation.

Fig. 8 i, ii, i (- *^).‘ — Antheridia in various stages of development
from a single ce of the lining tissue.

Halidrys siliquosa.

Fig. 9. — Ve;ical longitudinal section, with conceptacle.

liymantlialia lorea.

Fig. 10. — ])ung conceptacle seen obliquely from above, showing initial
hair protrudin from the cavity.

Fig. 11. — lightly older conceptacle in longitudinal section. Initial cell
shrivelled, u = Cuticularised outer layer of swollen covering of the
limiting tissu (cf. text, footnote).

Fig. 12 (J^). --Branching antheridial hair, sliowing transition, on for-
mation of aiueridia, from monopodial to sympodial system of branching.
The branchcare numbered according to their ages, ] being the oldest.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES VI & VII,

Illustrating Mr. J. E. Bloomfield’s paper On the De-
velopment of the Spermatozoa.^’ Part I. Lumbricus.”

Fig.

1. — Testis of the earth*worm — a small example as seei under low power.

2. — Cells from the trabecular sustentaculum of the seninal reservoir.

3. — Portion of wall of a seminal reservoir with fusifom cells of the sus-
tentaculum.

4. — Transverse section of the seminal reservoir, showag the penetrating

blood-vessels.

5. — A portion of a seminal reservoir more highly magiified, showing the
spermatospheres packed in the vascular sustentaculum.

6 — 10. — Cells (spermatospores) from a young testis eased : osmic acid
and picrocarmine.

10a. — Drawn on a larger scale.

11 — 15. — Cells from young seminal reservoir elongting to form the
sustentacular fibres.

16 — 20. — Spermatospores, dividing into two (young ‘polyplasts ”).

21, 22. — Into four.

23. — Three segments.

24. — Eight segments or spermatoblasts, drawn in fres state.

25. 26. — Ditto, with acetic acid. 26 showing cental protoplasm or

blastophor {bl).

26a. — Sixteen spermatoblasts.

27, 28. — A polyplast. 27, with acetic acid. 28, fresl

29. — Similar polyplast on treatment with osmic acid ail picrocarmine.

30, 31, 32. — Further stage in segmentation, fresh.

33, 34, 35. — Ditto, treated with osmic acid and picrocrmine.

36, 38. — To show the central blastophor {bl) at this sige, and in the

following.

37. — Showing refractive cap on the spermatoblasts.

39. — Showing protrusion of fine filament.

40. — Same stage fresh.

41 — 46. — Progressive stages, consisting in an elongatioof the nuclei of
the spermatoblasts, h, n, t, head, neck, tail.

J.E BlootnfuW. d«’..

, 'Pfa-r^owrnJlXcURS.^l. W.

PLATES VI & Yl\— Continued.

47, 48. — Spermatozoa mature, resting on central blastophor {bl).

49, — Isolated blastophor with vacuole.

50. — Spermatozoa seen under 10 immersion objective, (?, mature ;
(7, from polyplast, fig. 46.

51 — 59. — Brown corpuscles of the seminal reservoirs exhibiting network,
fresh.

00 — 65. — Ditto with osmic acid and picrocarmine.

06. — Similar brown corpuscle treated with acetic acid.

67 — 71. — Younf brown corpuscle.

72. — Sperm pohplast of Helix aspersa : bl. blastophor, spb. spermato-
blasts.

73. — Sperm pol3>last of Rena temporaria : bl. blastophor, spb. spermato-
blasts.

I

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE Till,

Illustrating Mr. Armaner Hansen^s paper on the Bacillus
of Leprosy.^^

Figs. 1, 2, and 3. — Cells from tubercles with rod-sliped bodies, fresh.
Gundlach, No. 8.

\

Fig. 4. — Such cells treated with osmic acid, Gundlach No. 8.

Fig. 5. — A brown element with adhering articulated breads after four
days’ cultivation.

Fig. 6. — From the border of one of the fungus growhs in the prepa-
ration of April 1st.

Fig. 7. — Brown elements coloured with methyl violei from a tubercle
treated with osmic acid.

Fig. 8. — Bacilli coloured with methyl violet, from a sedon of a tubercle
hardened in absolute alcohol.

i/Ptior. 0owTnylJ^XX, N. S. V//I.

Fi^. L Fi.q. 3.

K HutVUtk’Kir.’

cjfc>u/r. cy^^uyr.

)4‘Fp.rla.n« i- £r sVme Li tti'? Edin'^

Hemileia vastatrix.

j2foc. ^lu.xx.KS.m.x.

Morns De Alwn. dtl

M^Ta-rlane b. ErsVin®, Litl\

1-i IT’ Ky

LMA ..A.;TATP,Y

jsicc .

T^ig 1

-B

-C

--A

Fi^. 2

]^orn t A- De A1 wj a , del

7^'F»rl»T\o &,Erslnne. L.itli’’’ Edin

.HkMH.KTA VASTATRIX

Q^aa/r? cM(yr.

K'Farlan* ai! Er-.l<\ne, LitV Edin’

Fig. 1.

Fii. 2.

Mon-i* A- Jj<> Alwi.^.del

Fi^. 1

P c/(M/p cyjqyr: Ax.. Vk.M^ N.S.^iim

Fig. 1

j ey^,(yr. JZ. N.Z. 0^ Z1V.

Fig. 2.

M'Farl*.Ti€ &L. Er»Vme LitV * £3m

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES IX— XIV,

Illustrating the paper on the Coffee-leaf Disease of
Ceylon/’ by W. T. Thiselton Dyer, M.A., F.L.S.

The whole of the drawings were made by D. Morris, M.A., F.G.S., for
his forthcoming Handbook.

EXPLANATION OF PLATE IX.

Fig. 1. — Under side of coffee leaf, showing disease spots in various
stages of development. The orange -coloured sporangia are arranged in
numberless clusters on each spot. a. An old^disease spot subsequently
attacked by a second fungus {^Aspergillus'), b. A disease spot traversed
by nerves of the leaf. In such cases the filaments cross over the barriers
formed by the nerves from one stoma to another, on the outside of the
leaf.

Fig. 2. — Disease spots coalescing and forming an individual patch.

Fig. 3. — Portion of coffee leaf, twice natural size, with old disease
spots in the centre. The immature sporangia are orange coloured,
whilst the more mature are colourless. The larva of a male dipterous
insect is represented feeding on the sporangia. See fig. 7.

Fig. 4. — Disease spot, magnified about ten times, showing the arrange-
ment of the sporangia in clusters.

Fig. 5. — Portion of under surface of coffee leaf magnified, so as to
show : — A. Stomata, b. Cluster of sporangia coming through and occu-
pying the area of a stoma, x 200.

Fig. 6. — Single cluster of sporangia; The more mature have fallen
off, exposing their points of attachment. X 500.

Fig. 7, — Enlarged drawing of larva of dipterous insect, found feeding
on sporangia of Hemileia. Sec Fig. 3.

EXPLANATION OF PLATE X.

Fig. 1. — Sporangia found on fallen leaves and moist surfaces under
coffee trees, giving rise to wide-spreading mycelial filaments. X 250.

Fig. 2. — Sporangia sown on glass slide, and kept in a moist atmo-
sphere for thirty-six hours. Mycelial filaments abundantly produced.
A. Unbroken cluster of ripe sporangia, b. Detached orange-coloured
sporangia, c. Sporidia escaped from the sporangia, developing fila-
ments. c. Sporidia developing filaments whilst still enclosed in the
sporangia, x 600.

VOL. XX. NEW SEK.

li

y

cy^uyrc/oiLTTi^IXi.XK. KS.^. W

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XV & XVI,

Illustrating Mr. J. D. SiddalPs Memoir on Shepheardella,
an undescribed Type of Marine Rhizopoda ; with a
few Observations on Lieberkuhnia.^^

PLATE XV,

Illustrating Shepheardella tceniformis life-history.

Fig. 1 . — Three specimens drawn natural size.

Fig. 2. — A living Shepheardella^ with pseudopodia extended from the two
end apertures, and also from the investing layer of sarcode. Oval nucleus
near the middle. X 15 diameters.

Fig. 3 a. h. — The middle and two ends of same specimen, showing
nucleus, apertures, integument, and yellowish granular sarcode. x 170
diams.

Fig. 4. — The varying appearances presented by the nucleus as it is
carried along by the rotating protoplasm {sarcode). X 120 diams.

Fig. 5. — A very small and somewhat abnormal specimen. X 60 diams,
a. and b. The end showing aperture and first protrusion of pseudopodia.

Fig. 6. — A many-apertured form, probably ‘‘ Shepheardella ^ in first stage
of “ breaking up.” Pseudopodia drawn to actual length. X 27 diams.

Fig. 7. — Fig. 2 at nine a.m., December 18th. a. Detached portion of
sarcode.

Fig. 8 a. b. — Fig. 2 at seven p.m., December 19th. Viewed from both
sides of the cell.

Fig. 9 a. b. — Fig. 2 at nine a.m., December 20th. Viewed from both
sides of the cell. Sarcode all naked^ having beeu exuded from the
wrinkled empty integument i.

Fig. 10 a. b. c. d. — The naked sarcode of another Shepheardella,^* which,
having passed through preliminary alterations in form, somewhat similar to
those represented by Figs. 2, 7, 8, 'and 9, broke up into four separate portions
on December 17th at nine a.m., twenty-four hours previously having been
just as Fig. 9 in form, but still enclosed in its integument.

Fig. 11. — Fig. 10 at 10.30 p.m.,, December 17th.

Fig. 12. — Fig. 10 at nine a.m., December 18th. a. b. Detached amoeboid
portions. Figs. 7 to 12 X 20 diameters.

Fig. 13 a. b. c. — Fig. 10 on December 27th. Breaking up. a. The
empty integument, b. A protruded mass of still sarcode, containing a
number of nucleated granules, c. One of the latter which has developed
Actinophrys-like rays.

Fig. 14 a. b. — Actinophrys-form some days after liberation. x 250
diams. a. Dividing into three separate individuals, b. Large example,
containing nucleus and contractile vesicle.

Fig. 15. — A minute portion of Fig. 12, detached on December 18th,
containing a large number of granules exceedingly small, yet much larger
and more definite in form than the ordinary granules of the sarcode.

EXPLANATION OF PLATE 'SN .—Continued,

a. As first given off. 1. Its appearance on December 20th, showing
“ lobose ” pseudopodia and granules given off from it. By January 25th
it had discharged all the granules, which could not be distinguished from
the other organic matter on the slide.

Pigs. 16, 17, 18, and 19. — Amoebae, as given off from Shepheardell^e ;
16 containing nucleus, n. Nucleus, c. v. Contractile vesicle. 17.
Another form, occasionally nucleated. 18. Non-nucleated, very active
form. 19. Resting condition, as now assumed, four to seven weeks after
liberation.

PLATE XVI,

Illustrating Shephear della tcsniformis histology, and Lieberkuhnia.

Pig. 1. — Camera tracing of the subcutaneous layer of the sarcode of
“ Shepheardella^'" mounted in glycerine jelly (without any other treatment)
and containing, imbedded in its clear, structureless protoplasm, nucleated
granules of various sizes, and large clear masses of firmer protoplasm.
X 600 diams.

Pig. 2. — Camera tracing of a similarly mounted specimen, after treat-
ment with osmic acid, alcohol, and picro-carmine, showing every granule
contained in the clear protoplasm to be of definite form and nucleated.
X 600 diams.

Pig. 3. — The nucleus of a glycerine-mounted specimen (only), showing,
the nucleus, a, proper, enveloped in a membrane, he. X 300 diams.

Pig. 4. — The nucleus of another specimen, after treatment with acetic
acid and carmine, showing the nucleus, proper, filled with granular
protoplasm (stained deeply), and an irregular edged denser mass near its
centre (nucleolus), the nucleus being embraced by a delicate membrane,
b. X 300 diams.

Pig. 5. — Another nucleus, treated and mounted as Pig. 4, showing a^
nucleus ; the protoplasmic contents contracted upon its walls ; c, the
embracing membrane ; and c?, the very transparent enveloping sac. x 300
diams.

Pig. 6. — Another, mounted in glycerine jelly only, showing Cy the
embracing membrane; and c?, the enveloping sac wrinkled upon it.
X 300 diams.

Pig. 7. — Shephear della tcenia, from a living specimen. X 40 diams.

Pig. 8. — Lieberkuhnia Wageneri. X 55 diameters.

Pig. 9. — Aperture of ditto. Base of the principal stem of pseudopodia,
shown as an oval spot within the square mouth. X 55 diams.

Pig. 10.— The same specimen shortly after being transferred to the cell,
showing the living worm entangled among the pseudopodia. X 38 diams.

Pig. 11. — The same, with the worm swallowed as far as possible.

Pig. 12. — Marginal portion of the same when mounted in glycerine
jelly, viewed in optical section, showing transparent integument, a ; vesi-
cular nuclei, h\ granular protoplasm, c\ and subcutaneous layer of finer
protoplasm, d, x 1000 diams.

■MmIKsM-WJ!.

Muyr.

F. Kuth, lilj.1 Kdiii’

A.r.ed^wi k (id

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XVII AND XVIII,

Illustrating Mr. Sedgwick^s Memoir on Development of
the Kidney in its Relation to the Wolffian Body in the
Chick.^^

Complete List of Reference Letters.

Ao. Aorta. At. Alimentary canal, cl. Cloaca, c. v. Cardinal vein.
ep. Epiblast. hp. Hypoblast, i. c. m. Intermediate cell mass. i. c. mJ
Cell mass, which later becomes the intermediate cell mass. m. e.
Mesentery. M. d. Mullerian duct. k. b. Kidney blastema. k. t.
Kidney tubule, n. c. Notochord, p. Protovertebra. p' . Cell mass,
which later becomes a protovertebra, p. v. Body-cavity, p. e. Peri-
toneal epithelium. t. Testis. u. Ureter. ve. Vertebral body. w.
Wolffian body. w. b. Wolffian blastema, w. d. Wolffian duct. w. t}
Primary Wolffian tubule, w. t? Secondary ditto, w. t? Tertiary ditto.

Fig. 1. — Section between the fifteenth and sixteenth protovertebrse
of a chick with twenty-three protovertebrae, showing the rudimentary
continuation of the body-cavity into the intermediate cell mass and the
connection which the latter has obtained with the Wolffian duct. The
intermediate cell mass in anterior and posterior neighbouring sections
has separated from the peritoneal epithelium.

Figs. 2, 3, 4, and 5. — Sections taken from a duck embryo with about
thirty-two protovertebrae, illustrating the development of the Wolffian
tubules. Hart, cam., ob. 4.

Fig. 2. — Section through the thirtieth segment, intermediate cell mass
continuous with peritoneal epithelium, and containing a rudimentary
prolongation of the body-cavity. Lumen of Wolffian duct doubtful.

Fig. 3. — Section through the twenty-ninth segment, intermediate
cell mass separate from peritoneal epithelium.

Fig. 4. — Section through the twenty-sixth protovertebra, showing
features similar to above.

Fig. 5. — Section through the twenty-second protovertebra ; eom-
mencing diflferentiation of Wolffian tubule.

Figs. 6 — 10. — Sections illustrating the more modified development^of
the Wolffian blastema, as seen in the chick behind the twentieth segment.

Fig. 6. — Section through a chick with twenty-six protovertebrae
behind the four last-formed segment, showing the thick peritoneal
epithelium, the Wolffian blastema in connection with the mass of cells
which will become a protovertebra. Hart, cam., ob. 4.

Fig. 7. — Section through the twenty-ninth protovertebra of a chick
with twenty-nine protovertebrse, showing the thick peritoneal epi-
thelium and the Wolffian blastema in connection with the provertebrai.
Hart, cam., ob. 4.

Fig. 8. — Section through the twenty-fourth segment of a chick with
twenty-six protovertebrm, showing the Wolflian blastema separate from
protovertebrm and thick jieritoneal epithelium. Hart, cam., ob. 3.

EXPLANATION OF PLATES XYII AND Continued.

Fig. 9. — Section through the twenty-fourth segment of a chick with
twenty-nine protovertebrae, showing Wolffian blastema and thin peri-
toneal epithelium. Hart, cam., ob. 3.

Fig. 10. — Section through the twenty. ninth segment of a chick with
thirty-four protovertebrse, showing the commencing development of a
primary Wolffian tubule from Wolffian blastema. Hart, cam., ob. 4.

Fig. 11. — Section through a chick, end of third day or beginning of
fourth, showing earliest appearance of a secondary tubule. Hart,
cam., ob. 3.

Fig. 12. — Section through the thirty-second protovertebra of a
chick with thirty-four protovertebrse, showing the kidney blastema.

Figs. 13 — 17. — A series of sections from the hind end of a chick of
fourth day, illustrating the continuity of the Wolffian body with the
cells forming the kidney blastema. Hart, cam., ob. 3.

Fig. 13. — Last section, in which a tertiary tubule was seen.

Fig. 14. — Last section, in which a secondary tubule was seen. The
tubules in figs. 13 and 14 are contiguous.

Fig. 15. — Next section but one behind fig. 14.

Fig. 16. — Next section but one to fig. 15.

Fig. 17 a. — Section some distance behind that drawn in fig. 16.

Figs. 15, 16, and 17 show kidney blastema.

Fig. 17 shows opening of Wolffian duct into horn of cloaca.

Figs. 18 — 20. — Sections through a slightly older embryo than that
from which above series was taken. Hart, cam., ob, 3. Showing (fig.
20), ureter opening into Wolffian duct, with shifted kidney blastema
lying just internal to it.

Fig. 19. — Showing developing ureter and kidney blastema.

Fig. 20. — Section just anterior to ureter through the anterior end of
the kidney blastema.

Fig. 21. — A longitudinal vertical section through the hind end of a
four-day chick, showing continuity of kidney blastema with hindermost
part of Wolffian blastema, in which the development of Wolffian tubule
is taking place. No line of demarcation can be drawn between the two.

Fig. 22. — Section through a chick of seventh day or late in sixth,
showing the portion of the ureter {u^ and its dorsal dilatation {v. t.)
with regard to the Wolffian body {w.).

Figs. 23 and 24 are from sections of the chick from which fig. 22 was
taken.

Fig. 23. — Section next but one to fig. 22. It shows the kidney tubule
dorsal to the ureter, surrounded by the blastema.

Fig. 24. — Section next to fig. 22, shows the dilated termination of the
kidney tubule and the continuity of its lining cells with those of the
kidney blastema.

Fig. 25. — From a section through the kidney of an eight-day chick,
showing the termination of a kidney tubule. It presents the same
feature as fig. 24.

y

i^Uoi.n.Ns^ic.xx.

Wt '! ^ i

Fi^. 16.

r

hr

f V

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XIX, XX, AND XXI,

Illustrating Mr. F. M. Balfour’s Notes on the Develop-
ment of the Araneina.

PLATE XIX.

Complete List of Reference Letters.

ch. g. Ganglion of chelicerse. c. 1. Caudal lobe. ch. Chelicerse. pd.
Pedipalpi. pr. 1. Praeoral lobe. pp^. pp^. etc. Provisional appendages.
p. c. Primitive cumulus, sp. Spinnerets, st. Stomodaeum.

I — IV. Ambulatory appendages. 1 — 16. Postoral segments.

Fig. 1. — Ovum, with primitive cumulus and streak proceeding from it*

Fig. 2. — Somewhat later stage, in which the primitive cumulus is still
visible. Near the opposite end of the blastoderm is a white area, which is
probably the rudiment of the procephalic lobe.

Fig. 2»a and 2»b. — View of an embryo from the ventral surface and from
the side when six segments have become established.

Fig. 4. — View of an embryo, ideally unrolled, when the first rudi-
ments of the appendages become visible.

Fig. 5. — Embryo ideally unrolled at the stage when all the appendages
have become established.

Fig. 6. — Somewhat older stage, when the limbs begin to be jointed.
Viewed from the side.

Fig. 7. — Later stage, viewed from the side.

Fig. la. — Same embryo as fig. 7, ideally unrolled.

Fig. 8(2 and 85. — View from the ventral surface and from the side of
an embryo, after the ventral flexure has considerably advanced.

Fig. 9. — Somewhat older embryo, viewed from the ventral surface.

PLATES XX AND XXI.

Complete List of Reference Letters.

ao. Aorta. «5. Abdominal nerve cord. cA Chelicerae. ch. g. Gan-
glion of cheliceraj. ep. Epiblast. ht. Heart, hs. Hemispherical lobe
of supra-oESophageal ganglion. 1. 1. Lower lip. m. Muscles, me. Meso-
blast. mes. Mesenteron. mp.g. Malpighian tube. ms. Mesoblastic
somite, oe. (Esophagus, p. c. Pericardium, pr. Proctodaeum (rectum).
pd. Pedipalpi. pd. g. Ganglion of pedipalpi. pr. c. Primitive cumulus.
s. Septum in abdomen, so. Somatopleure. sp. Splanchnopleure. st.
Stomodseum. su. Suctorial apparatus, su. g. Supra-oesophageal ganglion.
th. g. Thoracic ganglion, v. g. Ventral nerve cord. gk. Yolk. g. c.
Cells derived from yolk. g. n. Nuclei of yolk cells.

I g — IV g. Ganglia of ambulatory limbs. 1 — 16. Postoral segments.

PLATE XX & Continued.

Fig. 10. — Section through an ovum, slightly younger than fig. 1.
Showing the primitive cumulus and the columnar character of the cells
of one half of the blastoderm.

Fig. 11. — Section through an embryo of the same age as fig. 2.
Showing the median thickening of the blastoderm.

Fig. 12. — Transverse section through the ventral plate of a somewhat
older embryo. Showing the division of the ventral plate into epiblast and
mesoblast.

Fig. 13. — Section through the ventral plate of an embryo of the same
age as fig. 3, showing the division of the mesoblast of the ventral plate
into two mesoblastic bands.

Fig. 14. — Transverse section through an embryo of the same age as
fig. 5, passing through an abdominal segment above and a thoracic
segment below.

Fig. 15 — Longitudinal section slightly to one side of the middle line
through an embryo of the same age.

Fig. 16. — Tranverse section through the ventral plate in the thoracic
region of an embryo of the same age as fig. 7.

Fig. 17. — Transverse section through the procephalic lobes of an
embryo of the same age. gr. Section of hemicircular groove in pro-
cephalic lobe.

Fig. 18. — Transverse section through the thoracic region of an
embryo of the same age as fig. 8.

Fig. 19. — Section through the procephalic lobes of an embryo of the
same age. ^

Fig. 20 a, 6, c, d, e. — Five sections through an embryo of the same
age as fig. 9. a and b are sections through the procephalic lobes,
c through the front part of the thorax, d cuts transversely the
posterior parts of the thorax, and longitudinally and horizontally the
ventral surface of the abdomen, e cuts the posterior part of the ab-
domen longitudinally and horizontally, and shows the commencement
of the mesenteron.

Fig. 21. — Longitudinal and vertical section of an embryo of the same
age. The section passes somewhat to one side of the middle line, and
shows the structure of the nervous system.

Fig. 22. — Transverse section through the dorsal part of the abdomen
of an embryo of the same stage as fig. 9.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXII,

Illustrating Professor Giard’s Memoir on the Ortho-
nectida/'

Fig. 1. — Rhopalura ophiocorruB (natural state).

Fig. 2. — Short form, or young phase (natural state).

Fig. 3. — Adult animal, treated by reagents and showing the muscular
bands.

Fig. 4. — Immature animal (acetic acid and carmine).

Fig. 5. — Intoahia gigas (profile view).

Fig. 6. — The same seen from the dorsal aspect, and treated in such a
way as to show the ectodermal cells.

Fig. 7. — The same with the ectoderm removed, in order to t>how the
endoderm.

Fig. 8. — Very young blastula of Intoshia.

Fig. 9. — The same more advanced.

Fig, 10. — Commencement of delamination.

Figs. 11, 12, and 13. — Formation of planula.

Figs. 14 and 15. — Very young sporocysts of Intoshia still rotalnin-
remains of the ectoderm.

Fig. 16. — More advanced sporocyst, opened in order to show its buds.
Fig. 17. — Portion of the same very highly magnified.

Fig. 18. — Bud in the condition of a planula.

Figs. 19, 20, and 21. — Stages in the segmentation of the egg of
Rhopalura.

¥"1^ / T^r

. !*"• -s?

^ ••

ifr-‘

*■ » ,axx. aJiiJTlO^ XOfTA'X'jt HXl^f ' '■

i0frW,^P,

4i7T‘^: aiw ■

,if; ■l.-^S

[j ' ' .'-^V 4-

l*-t- ... ,»-!Vt. > ..-j *- 6^,

■/-.

Mo<yr XX,MsM. m//

h. UW sain'

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXIII,

Illustrating Dr. Hubrecht^s Researches on the Nervous
System of the Nemertines.^^

Fig 1. — The brain and the anterior portion of the lateral nerve-chords
in Ceehratulus^ and the LiNEiDiE in general (Diagrammatic).

Ar external sheath of nerve-cells is applied to an internal skeleton of
nerv3-fibre from which latter the principal nerves take their origin. In the
fig\r*e the fibre is indicated by a lighter tint than the cells.

7o the right the superior lobe \0) which is coalesced with the inferior
lolc {U) anteriorly to T is supposed to have been removed. There is a
thck ventral and a thin dorsal commissure. The median longitudinal nerve
f/r the proboscidian sheath (js) takes its origin out of the latter commis-
sure.

H. Posterior or third brain lobe. N. Nerve-chords. F. Nervus vagus.
S. Nervous stems to the tip of the snout, the eyes (when present) and the
muscles of the lateral fissures.

Figs. 2 — 6. — Different stages of differentiation of the respiratory ciliated
duct and of the mass of larger cells which are coalesced with the posterior
brain lobe. This mass of cells is indicated by a lighter tint (in figs. 4, 5,
and 6) than the mass of darker ganglion-cells. In figs. 2 and 3 the nerve-
cells do not surround the nerve-fibre, but are applied externally to it ; a
mass of larger cells is not present here. The ciliated canal and its external
opening are left white.

2. Carinella anmlata; 3. Carinella inexpectata; 4. Folia curia; 5.
Cerebratulus roseus ; 6. Drepanophorus.

Figs. 7 — II. — The different relative situations of the longitudinal nervc-
chords in different genera. The epidermoidal tissues are left white, the
muscles are darker and the nerve-chords of a darker shade still.

7. Carinella; 8. Cerebratulus; 9. Langia ; 10. Amphiporus ; II. Dre-
panophorus.

Fig. 12. — Transverse section (left half) of the oesophageal region of
Cerebratulus roseus. Neither the epiderm, nor the larger portion of the
longitudinal muscles have been indicated.

0. Lumen of the oesophagus; Z. Folds in the mucous membrane; Fr.
Proboscidian sheath with B, dorsal longitudinal blood-vessel ; L. Thin
inner layer of longitudinal muscles ; Q. Circular muscular layer ; Z'. Ex-
terior layer of longitudinal muscles ; M. Median dorsal nerve-cord. iV.
Left nerve-chord with central fibrous portion separated from the ganglion
cells by a hyaline sheath.

Indirect continuous connection with the ganglion cells is a cellular sheath
encircling the body and situated between the outer longitudinal and the cir-
cular muscular layers. It is this sheath which the author regards as a
“ nervous tunic.”

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XXIV & XXV,

Illustrating Mr. A. G. Bourne’s Memoir On the Struc-
ture of the Nephridia of the Medicinal Leech.’’

References.

L. B. V. Lateral blood-vessels. m.l. Cells of main lobe. a. l. Cells
of apical lobe. t. l. Cells of testis lobe. c. n. Central duct. v. d. Vesi-
cle duct. B. D. Recurrent duct.

a. Nuclei. h. Blood-vessels. V. Empty blood-vessels. c. Cufcle
of the blood-vessels. d. Muscular fibres. dJ . Circular muscular fib-es
of lateral blood-vessels. J". Longitudinal, e. Cells of salivary gland, f li-
different connective-tissue fibres. g. Pigmented vaso-fibrous connectivj
tissue. h. Cilia in the vesicle. j. Opening of vesicle duct into vesicle
k. Intracellular ductules. 1. Cuticle of ductules. m. Cuticle of cells.

Fig. 1. — Semi-diagrammatic view of a segmental organ, taken from one
of the right-hand series in the region of the testes. The blood-vessels
are supposed to be injected, and the duct and ductules full. The con-
nective tissue, which covers the organ and penetrates between the cells,
and in which the blood-vessels lie, is not represented.

Fig. 2. — Termination of the apical lobe (‘‘apex”) and a portion of the
testis lobe, showing the origin of the recurrent duct, k. d.

Fig. 3. — Surface view of cell from the apical lobe, with intracellular
ductules, fully distended with fluid.

Fig. 4. — Inner wall of the vesicle -y. opening of the vesicle duct ;
h. blood-vessels ; h. cilia ; v. d. vesicle duct ; n. ridges due to the partially
contracted state of the vesicle.

Fig. 5. — Cells from the main lobe, with intracellular ductules fully
distended.

Fig. 6. — Cells of the apical lobe, as seen in transverse section
h. blood-vessels ; c. cuticle of blood-vessels ; f. indifferent connective-tissue
fibres ; k. intracellular ductules ; 1. cuticle of the ductules, x 800
diams.

Fig. 7. — Cells of the main lobe seen in section, letters as before.

Fig. 8. — Cells surrounding recurrent duct in its free portion; — a. nu-
cleus ; k. intracellular ductules ; /. cuticle of the ductules ; m. cuticle of
the cells.

jIX,

Litero-rtJo^;

Vesicleduct

Fiq. 5.

Ve sicle

Mcor ^owrml^XX.N^s'^. XXY.

Fiff- 10.

EXPLANATION OE PLATES XXIV & XXV— Continued.

Eig. 9. — Surface view of cells from the apical lobe, after treatment with
nitrate of silver.

Eig. 10. — Cells of the main lobe, after maceration in chromic acid per
cent. ; — a. nuclei, showing network ; b. blood-vessels ; c. cuticle of blood-
vessels ; f. indifferent connective-tissue fibres ; k. intracellular ductules ;
1. cuticle of the ductules ; m. cuticle of the cells, x SOO diams.

Eig. 11. — Contents of the vesicle and central and recurrent ducts.
X 3500 diams.

Eig. 12. — Contents of the ductules, x 3500 diams.

Fig. 13. — Portion of a transverse section of a Leech : — l. b. v. lateral
blood-vessel; M. L. cells of the main lobe; a. l. cells of the apical lobe;
T. L. cells of the testis lobe ; c. d. central duct ; v. d. vesicle duct ;
R. D. recurrent duct. a. nuclei ; h. blood-vessels ; V. empty blood-vessels ;
d. muscular fibres in longitudinal and transverse section ; d' . circular fibres
of the lateral blood-vessel; longitudinal fibres of the lateral blood-
vessel ; e. cells of salivary gland ; f. indifferent connective-tissue fibres ;
g, pigmented vaso-fibrous connective tissue; k. intracellular ductules;
1. cuticle of the duct, x 150 diams.

The nephridium here drawn lies on the left-hand side. The cells of the
main lobe are seen to lie above the lateral vessel, and the central duct has
been cut across in three places in this region. The central duct is seen
again cut across twice, in the apical lobe, and the vesicle duct, which in the
next section is free, is just becoming surrounded by the cells of the main
lobe at the commencement of the lobe. The recurrent duct, which further
on lies in the cells of the main lobe, is here surrounded by testis lobe cells
with their peculiar irregularly-shaped ductules.

Fig. 14. — Portion of a transverse section of a Leech, showing the vesicle
and its excretory duct v. d. vesicle duct ; b. blood-vessels ; d. muscular
fibres ; f. indifferent connective-tissue fibres ; g. pigmented vaso-fibrous
connective tissue ; h. cilia, x 150 diams.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXVI,

Illustrating Professor Ray Lankester’s Memoir on Intra-
epithelial Capillaries in the Integument of the Medi-
cinal Leech.^^

Fig. 1. — Mallet -shaped cells of the tegumentary epithelium : — b. shows
a portion of the cuticle adherent to the conjoined “ heads ” of two of the
cells. Macerated preparation.

Fig. 2.— From a transverse section of a Leech, hardened in ^ per cent,
chromic acid, followed by alcohol : — cu. cuticle ; v. intra-epithelial blood-
vessel; ep. epithelial cells (mallet-shaped). '

Fig. 3. — From a similar section: — cuticle; t?. intra-epithelial vessel
in transverse section; pg. intra-epithelial pigment process (modified
vaso-fibrous tissue) ; ep. mallet-shaped epithelial ^cells ; gl. superficial uni-
cellular gland.

Fig. 4. — From a similar section : — c. s. compact connective substance
(sub-epithelial layer of the integument). Other letters as in fig. 3.

Fig. 5. — Cuticle detached by maceration, showing openings of unicellular
glands.

Fig. 6. — Surface view of a flake of tegumentary epithelium, prepared by
maceration in potassium bichromate and staining with picro-carmine. The
dark-coloured “ handles ” of the cells appearing like nuclei (and possibly
representing such) are seen disposed in groups.

Figs. 7, 8, and 9. — Similarly prepared epithelial cells, showing the finely
granular character of the cell-substance ; and in figs. 8 and 9, the perfora-
tion of the cell by the apertures of unicellular glands.

N.sM XXVI.

i

JLor ^M/rMh^.XX.NsM XXV///.

F^. 19.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XXVII AND XXVIII,

Illustrating Professor Ray Lankester's Memoir On the
Connective and Vasifactive Tissues of the Medicinal
Leech.’’

Eig. 1. — Tissue from the neighbourhood of a nephridium, from a freshly-
killed Leech, fb. Brown fibres of the entoplastic vaso-fibrous system, c.
Branched and fusiform corpuscles of the ectoplastic connective jelly.

Fig. 2. — A single corpuscle of the connective jelly ; nucleus stained by
picro-carmine.

Fig. 3. — Network of brown fibres from the neighbourhood of the nerve-
cord. Less magnified than fig. 1.

Fig. 4. — Brown fibre showing nuclei, n, and tubular character. S. Solid
portion of the fibre. E. Hollow portion. Osmic picro-carmine prepara-
tion.

Fig. 5. — More decidedly tubular fibre of the vaso-fibrous system, showing
in one part very thin wall devoid of granulations, and numerous nuclei pro-
jecting into the lumen of the tube. The diameter of nuclei is y^th of an
inch. Osmic picro-carmine preparation.

Figs. 6 and 7. — Similar tubular fibres, showing nuclei and granular
walls.

Fig. 8. — Hsematophorous vessel (blood-vessel) with structureless wall,
distended with hsemoglobinous fluid, and continuous with a brown tubular
fibre, in which are free nuclei. From a chromic-alcohol section, stained
with picro-carmine.

Fig. 9. — Hsematophorous vessel with structureless wall, showing a few
granules at gr, and terminating at fb in a fibre.

Figs. 10 and 11. — Similar vessels to that drawn in fig. 8.

Fig. 12. — A similar vessel to that drawn in fig. 9.

Fig. 13. — Thin-walled haematophorous vessel containing free nuclei (cor-
puscles), stained pink by picro-carmine. The haemoglobinous fluid has
escaped.

Fig. 14. — A similar vessel, with smaller free nuclei — the haemoglobinous
fluid is here in position. Osmic picro-carmine preparation.

Fig. 15. — Botryoidal tissue of the medicinal Leech, drawn from a hori-
zontal section of a chromic-alcohol preparation. I am indebted to Mr. W.
E. Roth, Demonstrator in the Zoological Laboratory of University College,
for this excellent drawing.

Fig. 16. — Surface view of the termination of one of the coeca of the
botryoidal plexus in order to show the cells. Fresh preparation.

Fig. 17. — Transverse section of caeca of the botryoidal plexus in order to
show the relation of the cells to the lumen of the vessel. The haemoglo-
binous fluid was entirely absent in this portion of the preparation. Chromic-
alcohol section.

Fig. 18. — Outlying portion of the botryoidal plexus in order to show the
development of thick- walled botryoidal vessels (cc) as branches of a thin-
walled haematophorous vessel (vv). Osmic teased preparation.

Fig. 19. — Caecum of the botryoidal plexus in optical section, showing the
coagulated haemoglobinous fluid surrounded by the swollen granular cells of
the vessel-wall.

Figs. 16, 17, and 19 arc more highly magnified than figs. 15 and 18,

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXIX,

Illustrating Professor Pouchet’s Memoir On the Laminar
Tissue of Amphioxus.’^

Fig. 1. — Laminar network of the lophioderm.

Fig. 3. — Lophioderm from the region of the extremity of the notochord,
a, cellular rods ; c, <?, walls of the circular cavities with their cellular
covering ; Ci sheath of the notochord ; d, the largest cavity forming the
transition to the cavities of the fin.

Fig. 3. — a. Section of a canal in the subdermic layer, a, dermis ;

fibres. B. Pyriform cavity communicating with a cellular filament.

Fig. 4. — Transverse optical section of a papilla and its lumen. sub-
cutaneous aponeurosis ; space filled with liquid surrounding the papilla,
and completely lined with epithelium ; plan of the delicate fibres which
send prolongations into the papilla ; c?, e, amorphous substance covering
each side of the median aponeurosis (^), and bounding on the other hand
the blood (?) spaces, which are lined with cells ; /, amorphous substance
in connection with the muscles {mm), and traversed by very large laminar
fibres {g).

Fig. 3. — Longitudinal section passing through the median aponeurosis
(A), on which are seen the nuclei of the cells, which cover the spaces shown
in the preceding figure. The substance of the papilla presents concentric
layers, which are clearly visible. The letters are the same as in the pre-
ceding figure.

Fig. 6. — Posterior extremity of the notochord, with the extremity of the
central canal of the medulla, on which arc inserted portions of the laminar
network. In the tissue of the notochord are seen very small nuclei, whicli
are very abundant and scattered throughout its whole thickness.

Fig. 7. — Sensory nerve terminations of the Trigeminus, a, simple termi-
nation with three cells, furnished with an envelope having itself a nucleus ;
h, compound terminations.

^■*- ^ -jir A

. J#3^ai-3® ai-Oiiooeo^oiK ^0 iLkwmly''']

rvii^ - "I ^ '

^(j vroiT^A^'iza

•d^JsL

^ j Jtil •

15^ ■>' fioihn ‘ ': c ’]

■■> ^’)^- ^ iti^ileiiil ^(&kqif{^; 4o<^ - J

; <iM;)^^>)^r iwlii^ l^i^htToV^:) 6ndi ;i’j^ju4|ftb . , :

i " Jl> ■■m i

:' ■ -: J

mb

'.ji

iv: '-'<f ■

00^,

^ . • jj;jy ,

LIMNOCODIUM SOWERBII

LIMNOCODIUM

S 0 W E R B I I

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XXX AND XXXI,

Illustrating Professor Ray Lankester^s Memoir on “ Lirnno-
codium (Craspedacustes) Sowerbii, a new Trachomedusa
inhabiting fresh water.^^

Fig. 1. — Enlarged sketch of a quadrant of the disc of an immature
specimen of an inch in diameter) seen from the subumbral surface.
Per T. two of the perradial tentacles ; Ax. endodermal axis of same ; MR.
marginal ring (nettle-ring, peronia, and endodermal ring tissue) ; MC.
marginal canal ; RC. radial canals ; VMC. velo -marginal cells ; Ve. velum
in winch are seen the ‘‘ velar centrifugal canals ” or tubular capsules of the
marginal bodies. Just above Ve. two neighbouring capsules are seen to have
coalesced ; Go. the immature outgrowths of the radial canals, in the ecto-
derm of which the generative products are formed ; St. stomach.

Fig. 2. — Edge of the velum as seen when reflected, showing the position
of a tubular “otocjstic” capsule between the two layers of the velum; ot.
tubular capsule; Sup. abumbral ectoderm-layer of the velum; Inf. the
adumbral musculo-epithelial ectoderm of the velum ; Vac. vacuolated cells
of the abumbral ectoderm, which by fusion cause the increase of the
tubular capsule ; muse, circular muscle-fibres of the adumbral ectoderm (not
shown in the lithograph).

Fig. 3. — Diagram of a section through the edge of the disc to show the
relation of the tentacle-root to the endoderm of the marginal canal ; Ec T.
ectoderm of the tentacle; EnT. endoderm of the tentacle; EcD. supra-
umbral ectoderm of the disc; EcD'. subumbral ectoderm of the disc;
Ee V. abumbral ectoderm of the velum ; Ec V. adumbral ectoderm of the
velum ; NR. nettle-ring, viz. ectoderm of the marginal ring charged witli
thread-cells ; En R, endoderm of the abumbral wall of the marginal canal,
having a cartilaginoid character and greenish-yellow colour ; VMC. one of
the strongly marked row or ridge of endodermal cells termed “ velo-marginal
cells En L. position of the endoderm-lamella of the disc ; x. ring of
colourless ectodermal cells, being the abumbral portion of the nerve-ring ;
Li. marginal body ; Cap. tubular capsule of the marginal body ; ot. cavity
of the tubular capsule ; MC. marginal canal.

Fig. 4. — View from tlie supra-umbral surface of a portion of the
marginal ring; the ectodermal cells only arc drawn. T. tentacle ; 2'R. ten-
tacle-root ; Ec D. ectoderm of the supra-umbral surface of tlie disc ; Pe.
peronium ; NR. nettle-ring ; Ec V. ectoderm of the abumbral surface of
the velum ; VMC. row of vclo-marginal cells (endodermal) showing through
the ectodermal cells of the nettle-ring.

Fig. 5. — The same preparation more deeply focussed. 2' R. tentacle-
roots formed by notochordal tissue; R End. cartilaginoid endoderm of the
marginal canal (representing the “ ring-cartilage”) ; MCa. adumbral wall of
the marginal canal in section, showing soft ciliated cells with contained
granules (enlarged in fig. 7) ; Mcp. the inferior border of tlie marginal
canal formed by the velo-marginal cells, marked VMC in other figures ; A',
nerve-ring; EcV. adumbral ectoderm of velum; MB. marginal body;
CC. tubular capsule of the same.

Fig. 6. — Transition of the pale notochordal tissue of a tentaclc-root into

EXPLANATION OF PLATES XXX and continued.

the green-coloured cartilaginoid tissue of the abumbral wall of the mar-
ginal canal.

Fig. 7. — Two cells of the adumbral wall of the marginal canal, show-
ing their rounded form and dark greenish-brown-coloured concretionary
granules.

Fig. 8. — One of the marginal bodies as seen in the living condition.
The attached pole is the lower in the figure. A greenish granule similar to
those of the endoderm cells is seen in the body to the left.

Fig. 9. — A similar body after the action of dilute acetic acid. Surface
view.

Figs. 10 and 11. — Two marginal bodies or “ refringent sense-bulbs ”
seen in optical section after the action of osmic acid. tec. cortical cells or
tentacle-ectoderm ; axen. medullary cells or endodermal axis.

Figs. 12 and 13. — Earliest stage in the development of a “ marginal
body” from a medusa nearly half an inch in diameter. ~End. endoderm of
the marginal ring with green-coloured cell-substance ; 'Ect. colourless ecto-
derm ; axen, axial cell of the marginal body derived from endoderm
and coloured by green particles; cc. tubular capsules of neighbouring
“ bodies.”

Fig. 14. — Later stage in the development of the marginal body and its
tubular capsule ; End. endoderm of the marginal ring ; axen. axial endo-
derm cells of the ‘'body”; Ect. ectoderm of the velum; tec. tentacular
ectoderm or cortical cells of the marginal body ; cc. cavity of the tubular
capsule.

Fig. 15 — A later stage. End. endoderm of the marginal ring ; nerve- ‘
ring ; Ve. velum ; gr. greenish granules characteristic of the endoderm ; axen.
axial endoderm or medullary cells of the marginal body ; tec. cortical cells
or tentacular ectoderm of the same ; Cag. tubular capsule.

Fig. 16. — A still later stage ; axen. axial endoderm or medullary cells of
the marginal body ; tec. tentacular ectoderm or cortical cells of the mar-
ginal body; cc. cavity of the ectodermal tubular capsule; cap. wall of the
capsule formed by the velar ectoderm.

Fig. 17. — Early stage in the development of a marginal body, in which
the form of a tentacle is abnormally assumed ; End. endoderm of the mar-
ginal canal ; Ect. ectoderm of the abumbral velar surface ; axen. enlarged
axial endoderm cell, with green granules.

Fig. 18. — Abnormally developed marginal body devoid of capsule and
having the form of a small free tentacle ; End. endoderm ; Cap. capsule of
a neighbouring marginal body; axen. axial endoderm cells forming a
spherical group ; ect. tentacular ectoderm here having the form of a
tentacle instead of closely investing the axial endoderm-sphere.

Fig. 19. — Abnormal capsule developing without any tentacular element,
that is, without a marginal body ; End. endoderm ; cc. cavity of the cap-
sule : Cap. wall of the capsule ; x. nerve-ring.

Fig. 20. — Abnormal marginal body (seen in optical section) which has
broken away from its attachment to the marginal ring and is lying free in
the distal end of its tubular capsule. The body has taken on the form of
a spherical cyst, and shows the cortical cells forming the wall of the cyst,
whilst within project the large endodermal cells.

Fia. 12.

Vf

? Huth,Uh»Bam'

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XXXII & XXXIII,

Illustrating Dr. A. A. W. Hubrecht^s Memoir on “ The Pe-
ripheral Nervous System in Palaeo- and Schizonemer-
tini, one of the Layers of the Body-wall.^’

Fig. 1. — Diagram of a transverse section of Carinella. The longitudinal
nerve-trunks and the nervous sheath (indicated by red) are wholly outside
the muscular body-walh into which processes are being sent out by the
the sheath.

Fig. 2. — The same of one of the Schizonemertini. The nervous tissue
is enclosed within the muscles of the body-wall.

Fig. 3. — The same of one of the Hoplonemertini. The longitudinal
trunks lie inside the muscular body-wall. The nervous sheath has dis-
appeared, and is replaced by metameric branches placed at regular intervals
and double, one being for the dorsal, the other for the ventral half. These
branches divide dichotomously.

Fig. 4. — Part of a transverse section of Carinella annulata in the pos-
terior half of the body. m. The longitudinal muscles of the body-wall, to
which a thin layer [m) of circular muscles is exteriorly applied, b. Basal
membrane of the skin. e. The deep cellular layers of the ectoderm, with
glandular cells {c. c). n. The longitudinal nerve-trunk, with the inferior
coating of ganglion-cells {nc.) applied to it, which gradually pass into the
nervous sheath {nl.), the latter sending out processes {np.) into the muscular
body- wall. In this figure the numerous radial fibres traversing the

basal membrane (b) have been purposely omitted. Drawn with immersion
vii, Seibert Krafft.

PiQ, 5. — Part of a longitudinal section through the region of the nervous
sheath in Carinella polymorpha. Letters as in fig. 4. The processes («/?.)
lead toward the cells of the ectoderm, not indicated in the figure. The
nervous sheath appears more split up by fibrous and muscular tissue than
in the foregoing transverse section.

Fig. 6. — Transverse section through the lateral longitudinal marrow-
trunk in Cerebratulus roseus. l, m. Longitudinal muscular layer, c. m.
Circular muscular layer, separated from the former by a layer of a homo-
geneous structure (like the basement membrane of the skin), through
which, as through the muscular layers, pass numerous radial fibres.
N. The lateral trunk situated in this homogeneous layer. Internally the
fibrillar structure prevails, and this fibrillar nucleus is separated by a layer
of connective tissue from the ganglion-cells {nc). Superiorly is represented
a very large ganglion-cell, which was present in this section, though they

EXPLANATION OF PLATE XXXII & Continued,

very rarely attain this size in the lateral trunks, nl. Nervous layer en-
sheathing the body (see Plate XXIII, fig. 12, of this volume), np. Ner-
vous process penetrating among the muscles.

Fig. 7. — Oblique section (transverse horizontal) through another speci-
men of the same species. Letters as in fig. 6. set. The layer of connec-
tive tissue between the fibrillar and the cellular nerve-tissue. These two
figures and the following drawn with Siebert Krafft’s immersion No. vii.

Fig. 8. — Part of the nervous sheath in a transverse section of a large
specimen of Cerebratulus marginatus. Letters as in figs. 6 and 7. The
radial fibres are omitted in this figure.

Fig. 9. — Longitudinal section through the brain and the nervous sheath
of Cerebratulus urticans, to show the way in which the latter abuts upon
the former, b. Brain-lobes, nl. Nervous layer, p. Outline of section
through the proboscis, c. m. Circular muscular layer, l. m. Outer longi-
tudinal muscular layer, i. m. Inner longitudinal muscular layer.

Fig. 10. — The same for Langia formosa. b. Superior and posterior
brain-lobes, n. Lateral nerve, obliquely cut. 7tp. Nervous process pene-
trating into the longitudinal muscles.

Fig. 11. — Part of the fibrillar nerve-sheath {nl.) with ganglion-cells of
the foregoing preparation more strongly magni^ed. The process {np.) is seen
to have essentially the same character as the fibrillar layer itself, and to be
provided with several nerve-cells as well.

Fig. 12.— The nerve-sheath seen from above in a horizontal section
through the back of Cerebratulus pantkerinus. nps. Median dorsal nerve
(nerve for the proboscidian sheath), nl. Fibrillar plexus with nerve-cells.
nc. Peripheral processes from the nerve-sheath radially penetrating the
muscular layers, rf. Radial fibres (cf. figs. 6 and 7).

Fig. 13. — A surface section through the longitudinal muscular sheath
(L. M.) just above the nervous layer, with elliptic canals {?ic.), more or less
regularly arranged in transverse rows for the passage of the periphera
processes of the nervous sheath. These processes show, also in this sec-
tion a fibrillar structure, with nerve-cells imbedded in it. rf. Radial fibres.

XXXV

i

FHulh.Lilh' B<lm'

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATES XXXIV & XXXV,

Illustrating Mr. Sydney J. Hickson’s Memoir on the
Eye of Pecten.^^

Fig. 1. — A diagrammatic sketch of an eye of Picien maximus. a. The
cornea, b. The transparent basement membrane supporting the epithelial
cells of the cornea, c. The pigmented epithelium, d. The lining epi-
thelium of the mantle, e. The lens. f. The ligament supporting the lens.

g. The retina, h. The tapetum. Jc. The pigment. /. The optic nerve.
m. The retinal nerve, n. Complementary nerve, p. The circumpaleal
nerve (Duvernoy). q. A supplemental nerve from pedal ganglion.

Fig. 2. — Epithelial cells of cornea.

Fig. 3. — Section through the junction of the pigmented epithelium with
corneal epithelium.

Fig. 4. — Vertical section through eye of Pecteti maximus. a, b. Cornea.
c. Pigmented epithelium, d. Mantle epithelium, e. Lens. g. Retina.

h. Tapetum. k. Pigment. 1. Section of optic nerve.

Figs. 5, 6. — Isolated rods. Q>a. — Diagrammatic sketch of a- central

rod. /?./. Posterior limb. Membrane pierced by rods. a. /.Anterior
limb. s.r. Spindle-shaped portion of rod. n. Nerves.

Figs. 7. — Transverse sections through eye of P. maximus. a. Rods in
section, b. Tapetum. c. Pigment, d. Retinal nerve.

Fig. 8. — Vertical section of the eye of Pecten maximus, showing the
nerve dividing into retinal and complementary branches.

Fig. 9. — Vertical section of the eye of Pecten maximus, showing the
termination of the retinal nerve. The retina has dropped out, and the *
frayed-out end of the nerve remains.

Fig. 10. — Section of eye of Pecten opercularis.

Fig. 11. Retina («) of P. opercularis, {b) of P. jacoberus, (e) P. maximus.

Note. — A horizontal section means a section made in the same plane as
the mantle. A transverse section, in a plane at right angles to I lie eye
stalk. A vertical section, in a plane at right angles, both to the plane of the
mantle and the last-named plane.

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXXVI,

Illustrating Professor L. Ranvier’s Memoir On the
Terminations of Nerves in'* the Epidermis/^

Fig. 1. — Section of the epidermis of the snout of the pig, with the
nervous ramifications stained by chloride of gold.

Fig. 2. — Section of an organ of Eimer on the nose of a mole, with the
nerves which supply it.

Fig. 3. — Section of the epidermis of the human finger.

Mu^rJowrTM.TlNs'^LXCm

li Klein, a«l.

F Huth.Uh»

JOURNAL OF MICROSCOPICAL SCIENCE.

EXPLANATION OF PLATE XXXVII,

Illustrating Dr. E. Klein^s Memoir on the Termination
of the Nerves of the Mammalian Cornea.^^

(Figures 4 and 6a are drawn by Mr. Noble Smith, all others by the
author.)

Fig. 1. — From a horizontal section through the rabbit’s cornea. Mag-
nifying power about 260. Showing the subbasilar plexus proper, a.
Branches of the stroma plexus, situated behind and close to Bowman’s
membrana elastica anterior, a. Fine nerve-fibrils, at first in the same
level as a, but, passing into a deeper layer, terminate in the substantia
propria, b. Thicker nerve-fibrils, somewhat anterior to a, being situated
within Bowman’s membrane, x. Fine nerve-fibrils, which enter the sub-
epithelial network, xx. Bundles of pria’.itive nerve-fibrils which come off
from the rami perforantes and enter the subepithelial network.

Fig. 2. — From the same cornea as fig. 1. Magnifying power about 260.
A. Branch of the stroma plexus of the same level as a in fig. A. n. Fine
nerve-fibrils situated within Bowman’s membrane, the intrabasilar fibrils.
At X they pass into the depth again, and finally terminate in the substantia
propria. At xx they enter the subepithelial network.

Fig. 3. — From the same preparation as fig. 2. Magnifying power about
260. a. Branch of the stroma plexus, similar to a in the preceding figures.
n. Fine nerve-fibrils immediately behind the subbasilar plexus of fig. 1.
These are the deep subbasilar fibrils referred to in the text. Most of
them terminate in the substantia propria.

Fig. 4. — From the same cornea as the preceding figures. Magnifying
power about 440. Showing fine nerve-fibrils of the substantia propria of
the eornea, and their relation to the corneal eorpuscles. These latter are
only indieated as granular plates, their proeesses have not been drawn.
At a apparent terminations.

Fig. 4a. — From a similar cornea as in the preceding figure. Magnifying
power about 660. Showing the relation of the fine nerve-fibrils of the
substantia propria to the corneal corpuseles. At a apparent terminations.

Fig. 5. — From a cornea of a kitten. Magnifying power about 660.
Showing the relation of a fine nerve-fibril to the processes of a corneal
corpusele. This latter being stained of a grey colour, and the nerve-fibril
black, the distinetion between the two is easily made. The nerve-fibril
is not aetually conneeted with the processes of the corneal corpuscle, but
lies elose to them, both being contained in the lymph-canalicular system.
This applies equally to figs. 4 and 4a.

Fig. 6. — From the same cornea as in the preceding figure. Magnifying
power about 1400. Showing a very fine nerve-fibril giving off a lateral
branehlet, whieh terminates on not in the corneal corpuscle in a fine
network.

Fig. 6a. — From the cornea of a frog. Magnifying power about 660.
At a are shown the ultimate fine nerve-fibrils forming a network on the
corneal corpuscle. At h an apparent termination, the ultimate fibrils are
not brought out in the preparation.

j'lo, 7. — Subepithelial network of fine nerve-fibrils of the cornea of

EXPLANATION OF PLATE XXXVll— Continued.

rabbit seen, in a horizontal section. Magnifying power about 660. 1 and

2 are two bundles of fine nerve-fibrils coming off separately from two rami
perforantes, not included in the drawing. As is shown in the drawing, not
only are there anastomoses between fi;brils of the same bundle, but also, as
at 3, between the fibrils of the neighbouring bundles.

I'lG. 8. From a similar preparation, showing the fibrils of the subepithe-
lial network in a horizontal section. Magnifying power about 440. All
these fibrils, like those!of the preceding figure, run horizontally close behind
the deep layer of the anterior epithelium of the cornea. The actual length
of the part of the preparation depicted here is 0'4 ram., and all the relations
of the individual fibrils and their branchlets are drawn with great accuracy,
in order to show the relative lengths of the fibrils of the subepithelial net-
work, and* the number of lateral branchlets which pass into the anterior
epithelium. A, B, C are three bundles of fine fibrils coming off from
separate rami perforantes. In the bundle A a thick fibril can be fol-
lowed for a very great length. Of the fibrils with which it anastomoses
/and g are noteworthy on account of their leugth, and on account of the
numerous short fibrils, given off by them, which enter the anterior epithelium
at X ; X indicates all those fibrils which ascend into the anterior epithelium.
k. A nerve-fibril, through which the fibrils of neighbouring bundles anasto-
mose with one another.

Fig. 9. — From a horizontal section through the cornea of guinea-pig.
•Magnifying power about 440, Showing the nature and mode of division
and anastomoses of the intraepithelial^ nerve-fibrils in the superficial layers
of the anterior epithelium. The very minute rod-like lateral branchlets
are very conspicuous. At n a closed network.

Fig. 10. — From a horizontal section through the cornea of rabbit.
Magnifying power about 660. Showing the intraepithelial nerve-fibrils,
and the character of their distribution. At F a very fine fibril. The
numerous minute lateral branchlets are well shown.v At h, g there is a
closed network. In the upper right part of the drawing is an intraepi-
thelial branched corpuscle shown. It has no connection with the nerve-
fibrils.

Fig. 11. — From a horizontal section through the cornea of rabbit. Drawn
with the ^ oil immersion of Zeiss. Showing the intraepithelial nerve-
fibrils, their varicosities, branching, anastomoses, and terminations.

a and b. Two nuclei of superficial epithelial cells, to show the relative
proportions, c. The terminal network seen in profile between two epithelial
cells, d. The terminal netvrork seen en face.

Fig. 12. — Intraepithelial nerve-fibrils in a horizontal section through the
cornea of rabbit. Magnifying power about 440. The numerous minute
rod-like branchlets are shown here.

Fig. 13. — From the same cornea. Magnifying power about 1020.
Showing some of the very fine superficial intraepithelial nerve-fibrils, their
minute branchlets just indicated.

Fig. 14. — From the same specimen. Magnifying power about 660.

Fig. 15. — From the same specimen. Magnifying power about 1020. In
both figures the fine intraepithelial nerve-fibrils of the superficial layers are
shown, and their ultimate networks as seen from the surface. The oval
corpuscle in either figure represents the nucleus of an epithelial cell. '

Fig. 16. — From a similar specimen. Drawn with Zeiss’s oil immersion.
Showing the terminal network of the fine nerve-fibrils. The oval corpuscles
represent two nuclei of superficial epithelial cells, to show the relative pro-
portions. The terminal network is in no connection with the nuclei. •

T

r . «0

, rf

S=^ §

o

,* »t

ii?

—

C-T| O O

^

o

?r<i

-< ’J5 S